From b0e91a4bc7efc02f62f1f50a574f016904c1653e Mon Sep 17 00:00:00 2001
From: gavinsimpson <ucfagls@gmail.com>
Date: Mon, 25 Nov 2024 14:30:48 +0000
Subject: [PATCH] =?UTF-8?q?Deploying=20to=20gh-pages=20from=20@=20gavinsim?=
 =?UTF-8?q?pson/gratia@6f8dabd3204c81a0b9fcc1db332b8556ed352bb1=20?=
 =?UTF-8?q?=F0=9F=9A=80?=
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

---
 articles/posterior-simulation.html            |  15 ++++++++-------
 .../figure-html/unnamed-chunk-7-1.png         | Bin 308785 -> 253157 bytes
 pkgdown.yml                                   |   2 +-
 reference/draw.smooth_samples-2.png           | Bin 170409 -> 177624 bytes
 reference/link.html                           |   4 ++--
 search.json                                   |   2 +-
 6 files changed, 12 insertions(+), 11 deletions(-)

diff --git a/articles/posterior-simulation.html b/articles/posterior-simulation.html
index f5f50271..40dce5be 100644
--- a/articles/posterior-simulation.html
+++ b/articles/posterior-simulation.html
@@ -479,11 +479,12 @@ <h3 id="distributional-gams">Distributional GAMs<a class="anchor" aria-label="an
 <span>  <span class="fu"><a href="https://ggplot2.tidyverse.org/reference/geom_point.html" class="external-link">geom_point</a></span><span class="op">(</span><span class="fu"><a href="https://ggplot2.tidyverse.org/reference/aes.html" class="external-link">aes</a></span><span class="op">(</span>y <span class="op">=</span> <span class="va">accel</span>, colour <span class="op">=</span> <span class="va">type</span>, alpha <span class="op">=</span> <span class="va">.alpha</span><span class="op">)</span><span class="op">)</span> <span class="op">+</span></span>
 <span>  <span class="va">plt_labs</span> <span class="op">+</span></span>
 <span>  <span class="fu">scale_colour_okabe_ito</span><span class="op">(</span>order <span class="op">=</span> <span class="fu"><a href="https://rdrr.io/r/base/c.html" class="external-link">c</a></span><span class="op">(</span><span class="fl">6</span>, <span class="fl">5</span><span class="op">)</span><span class="op">)</span> <span class="op">+</span></span>
-<span>  <span class="fu"><a href="https://ggplot2.tidyverse.org/reference/scale_identity.html" class="external-link">scale_alpha_identity</a></span><span class="op">(</span><span class="op">)</span></span></code></pre></div>
-<p>The resulting plot is shown below in the left-hand panel of the
-figure below. There is clearly a problem here; the simulated data don’t
-look much like the observations in the 15ms immediately after the impact
-and again at ~45ms after impact. This is due to the model we fitted only
+<span>  <span class="fu"><a href="https://ggplot2.tidyverse.org/reference/scale_identity.html" class="external-link">scale_alpha_identity</a></span><span class="op">(</span><span class="op">)</span> <span class="op">+</span></span>
+<span>  <span class="fu"><a href="https://ggplot2.tidyverse.org/reference/theme.html" class="external-link">theme</a></span><span class="op">(</span>legend.position <span class="op">=</span> <span class="st">"bottom"</span><span class="op">)</span></span></code></pre></div>
+<p>The resulting plot is shown in the left-hand panel of the figure
+below. There is clearly a problem here; the simulated data don’t look
+much like the observations in the 15ms immediately after the impact and
+again at ~45ms after impact. This is due to the model we fitted only
 being for the conditional mean of <code>accel</code>.</p>
 <p>Instead, we model both the conditional mean <em>and</em> the
 conditional variance of the data, through linear predictors for both
@@ -517,8 +518,8 @@ <h3 id="distributional-gams">Distributional GAMs<a class="anchor" aria-label="an
 <span><span class="va">plt_gau</span> <span class="op">+</span> <span class="va">plt_gaulss</span> <span class="op">+</span></span>
 <span>  <span class="fu"><a href="https://patchwork.data-imaginist.com/reference/plot_annotation.html" class="external-link">plot_annotation</a></span><span class="op">(</span>tag_levels <span class="op">=</span> <span class="st">"a"</span><span class="op">)</span> <span class="op">+</span></span>
 <span>  <span class="fu"><a href="https://patchwork.data-imaginist.com/reference/plot_layout.html" class="external-link">plot_layout</a></span><span class="op">(</span>guides <span class="op">=</span> <span class="st">"collect"</span>, ncol <span class="op">=</span> <span class="fl">2</span><span class="op">)</span></span></code></pre></div>
-<p><img src="posterior-simulation_files/figure-html/unnamed-chunk-7-1.png" width="700">
-The plot of the simulated response data for the distributional GAM is
+<p><img src="posterior-simulation_files/figure-html/unnamed-chunk-7-1.png" width="700"></p>
+<p>The plot of the simulated response data for the distributional GAM is
 shown in the right hand panel of the plot. Now, there is much less
 disagreement between the observed data and that which we can produce
 from the fitted mdoel.</p>
diff --git a/articles/posterior-simulation_files/figure-html/unnamed-chunk-7-1.png b/articles/posterior-simulation_files/figure-html/unnamed-chunk-7-1.png
index 8623f77926a57982044e1ee99a160ea35a527d62..301eba413c00442fc091722903c81d54b2476a47 100644
GIT binary patch
literal 253157
zcmd>lc{tSX`|qSA6lE=I3t1*q$W9c6><q?ES%wJNw<PtkC4_7v%h<;fL$)Dnc7~zs
zS+b6O?8|wkKELnzoqx`Co$H+c&Rn`?a=qvMzMuQKm)GlcKSDIsA6=kgp@Kjl7nC1A
z)P_LFJt2_O7U#}_-<Xi~n}WBqma2~)f_L!Rdr%z!-Y8!@HgttR*x!-<or;`Ye+GeE
zg(yFi*Y!+Z9`gy$^zfTI=y6473y$F{D!r_iEpA(%7ym}X8Y6n|77Z8g_4>D%ueKKC
z*I2h7OHim@5^v(7p1S_w2X8(bSF~&LL)o^oAIRg_DY;LP`P?>NJ>`&m+4q?=asi3e
z^B6;ptnNq!3_S5^>31#7&aTRu5VD)eE1c6mBCNFP5q*&pn@3*V5<GtlhCu4SufH?-
z_cLU2@7g5!-~YV2q*zb=@9%GfzGD9OqY}d@p?^Py^OBL4PkLR}I#cua6XbUKInPsn
zKbt?iv_|&#WBotZ|9`#lSBO7h$Y0Hw4f@_nW_L2EQILY<lVZg;je-!!D^+<d2L4dX
zA0R})tNgO{|GmqSZcbtb7Xv9cdI=R1g-MQ8%3anvL~oNK&t~WrcBfWjXVf(kC&KQY
z0D&ZzO7{?Bi+mMV{Z@T29;>E$`FiZ)FYa^h_rBZeeL#DmTAE~6J$JopcclfBd3awu
zc8mMyc=ZH}^!lA3>}XUpL=+3y6x;8bIP413WrslKeiMrVYO#A=PAbv<htr(-eYW%a
zOGQHs4Gng!HAe)ep%jntTosf{N&6Yny^Ss3$_DKeR<Gcy$BBidZ>ysR2=2tViW)_a
zQRi00ob!-Zb93W9tN6gv6y29sRB-y(4aTDlp_4Bg(shqx{{)+$=mT2MNS~h_rdd*}
zH98tH#E&?USPbl$A;Qj1+3aM0h}f==b$LEAaZH#PoSK@tm`@Iw+&Z45J-)A~Wjj=8
zQZ<M4DXXgDw?}VC4>i)t6;#v`p4*Sef*k{{kQHe{{CQA@!@K|sc0Iz0_Z->S%YL;D
z>D96kLwn-*4p!e|+$)=FVVd*gt7F~Kc9`Bx2*l>4W!PD}7gHM)a=U?x#Qmy~fQ^{c
z0nX!JoP}d@o@<}CSH`M;x&6r!K>@Iao=4gJeLsGD;<B6TPLXt(`ThY)1~H$kJheI<
zDOY%Ou#@WUGnZ_-*RH?#je_$d*O>nSVSl$(4*Q>*BxNZ!R@#cN*i&?rmX@}Earn-7
z3i2W^$qBm`{;QD73z^;Hzg=p-Q@0i@<nYpvv;(hHHN}_7-gkU?GO<5cu?4ozgjRa(
zmH&Q!Y91c3P)Q#imTpj8TUM57QtfWCtbGb1p5aluWHMJaMjSG$bN8{ZkU#P_!z0`>
z+W#!Y`1M<kHOaZHW+8vAx*eZ7kMaC}N*Bc;=EB}>s-`aLBce%NF)KoXf`}fZ467C$
z$%sih(JiJ66%+o|{yQE4@x@G#7j5F>LdOo$JvwZy_B(bmf1*X)?7;<3PQNWJKj28?
z>VSSVv4z!><5jn{NqIL(_A6FT(Xp<MgbC3hW8tPyU}?2_wyXMx#GJb1H!sT{e@a??
z<z+<J+q(VyI(YR_p+~JMY3HPSq^3VLzN2*_`K0D*r0jl<>1N_kvRgGCkB?dD>FjjS
zRsQR(0Zp4V|AUbU(`vtw=M^cUqt7vLxUuo*{1*tsQ~r2m^>`(@y!zz$V8~HJUHwZv
z%b)etTo9c!F*G!+_7jXp9WSIFFDS*GhSZa<ll-kAceHh-!nu2)e8kbH`s5p}|BH-{
z|MDfv+Hr%>VZ$(?Za1t>e7gK(jrQcT+<u1K@O(4`^7ZQYMqCB2T7SRnF;Bn-kN0FC
z#o$Wa$<b=v<!67h;|}ndQ{`+NW+e4sbnfL6x?*u5J2egii`Hcb=J1-iR12;iVx8D+
zv52LF)O#KLq7BG7O+hmcj7%L^-QiYlxh!+MJMFzwJ%KiIs&#ulvQxY2H(VERQmpc)
zVT9lAk(hf+?25QkwLNT41eGJDf1C4gl7dKZ3NWITJ1pi}$jwIL81;=yqPWQ*FVsyz
zQZ6M;9A<GXM4s$M5<@h66<R)(zV}&BHOu2Xnt!Ac{WL@MZ-sdMRx(KJ9NE#P+(~I+
z<?qB&$Ah0e{;+^;n}96dP(cMyJf#Did`3pLVp0zuPESwUjIN0JPm66eavpzp@uzpK
z)dV*Rn!C7Y%kB3?Zjqi?HT_;{D`nO9(WCcleKRwej$ePBtzB^|>X#Z{ry&eu?0=wu
zo!vBS;-4!kCH!WgW_`6g<M~d$)2v{#2WI2V3e8T4W<#ro-K+DUgw7Us{INFtYg)N2
zX~#(~>3|2%`~H3ev2pz0dgFy!3L!xgFC_FFnX<MP!=FS=Nkw4z?FrxUDO@29FV^bZ
zD@O3VZWffU@8c1AwWb<_!xr(b-M$n0Xjlwa?6X`EwRgsSdlK0CA>KhhUo*L5_-O|Q
ztIpwogQMdQ`4V!mDvI(KbB2~KRDay0Yb}`)j}!6q<L_|!&X|&_lyN88ZD`5O3&~?o
zc)TL4#gIEfpdnqT7{wnOn)z++pXmw^KHZF`gzkmAu4KNd;0?W||I&YvSe5L0L*D}p
zGdie^e<q<VpDS{1G5pRAxBF~we<{~(YK_?B;q$m6jK`FtFD_4|OJIv4rVbR?S~vUx
zW)|P4+Y)vh(d@bVGCevM&)lf2W0P~f88Lp)u<mC)uqeCk+LVK*fk5t2z(xLb-2WMR
z$XQ!XGQS})H-Nxw&1Xa|9v`d@dDI=45WQtgd<dBRAFZ<PgX?IRs>hgHJZOijc3xhF
z`leo}@wK;N^Ozz(f}wtVz%{EssRMzW=RCFgL1wM~^nMeMtaJMi8e_wL@cl;Q{0+{%
z@5&>xhii@AyT9JK^~mhZ_op70@yE84bN-8*%w0b#Hhv(N|IF(6RlZ`}JOB8w=ft>T
z%zZ>6!oFgpicmiC6paT>$=we`J#GlQE$X)(W;THJJJ|YR6}hmPdUE(bZ?<`LlKsn4
zVYNEv@!C6L)ljXvl>4$K8YY$P>@ne25q_s?y@i#C+^Kn_9Q}P$nlM_ale*Vw!*C{W
zrFvDCUB-Ls_4|L(U9s22wf9%&lCk^vB9&-IP#SbNitq^U{-kj(<^66Ctn0Uyf1%qC
z<8z*mH1qM=anGVNWqh^^s=GkZ7gF9o-l+@tU$5$&no8%gv$j@}zX+}>G^yrN){tIr
zB4_=NJ17Cx?4DR$%*(V1`a#sq*HhL;&fRS(2(A=*y6k<tUvvVerlww!bw45wExl!R
z{J<4#k>g$KKRov{+;<{=&g+K$PDnrz9DQ?c^dbmwtep6qE*ka>&F&&Px8ga5vt3zg
z-AT~Ompbo7eVP<LuQH}Dfg;D1tDT$do<#pZ)}AnY#C+FXW@p@|h$G-|-Hnr#lT(gw
zk{y5hx96lxXf)RqcgmiQgzcU#!E>*p#%z=@j(_5?{4QK1{PVQ@!n$eI(w&xqG5?k2
zhB+ywzd>66`);BRf5lhpQa<IxloT1lJ<t(xS9@ig&#q8ksYcQU>^3eEy}{BfS^tsl
ziOpKLl4@U?JGAs=u<%G$NfBfd^@dbR_hXFYx!3mrS{rlR$)vxKG`3!a`gUtiYt)7-
zW&32dD-Vw54^6j-{}Y|{nwQqze_mLyQ^p-09@_cLba!<rM{loAfQ>1ft6D0QB3SWs
z)0GeM_iUERPFQQ_<CVUw&<mCJZ2D%BjZa&c0PqRQ0Y15!ek3w{O4hX|wlxPY(~kaw
z`XFC5zaIOaQG4=)_q~-e?pNcxp|Rg2B*V3UFY+h;w|%`f$u5taTj+A~FXdY3FgxDh
zi`+Tz8}nJMjm{jZ`7O%qaO<y|UUdb(@_1NFL_5XKc>%D2Mb#aR)R|P5c>@eVs``3y
zC{tpdJ_-IXzx>aa8cFDrxg-{4m*;`qOtIfVV<eAxBueB~&=uo(tWGuSnI*I`TOVXW
zaW6cm(KmJumfflBK@$gk3z_e38$UBFAG{%~rsF1=!o20nJkr$kCj1VfVx&=UwnikW
z3u9WYs;YWpD5&D613)|x#<b%hA(YBNNB#rb3G3_5%DBbF#Sc&kIV&rxtzXTG`f(^Y
zda$@dTnKEGGn>7am)8$eay%SeUViRr^_XmK#fkNt7cwh58^7>U$ua6@Lc$g0=sVV2
zI|&F(UUPG^MZAk?9)5Ay5^a}=o$Q%}K+KmcqA+Upmt6;oisBp1T)RX`{psb9iN^x=
zV5J|2i!Q9$P^ksTbD&FXqGp8*!EE^-_XJ+4`Ek2*lAXqS&D8hryXECi-+b0@2YHB-
z-tPo~CRn4n_1{QX*jdPpZ<2N;`nGBD`9=oh$Rca*-g58E5-I^j2*fBTrfgIQucG=4
z6*Kek4fbY!+D$syVUSV|ioR55iWi8~`R{&vd%rBn&KOg$lQ`$u%8zvygI0NG<uPk;
zZ^$^2x@WP4%k<WaOCGocYL`1I?iC%~JM@qa;d2AWNY5B7y9yXG#E9~g&?_|oG4%<#
zvOLb7?Xu|Dymx^&qcsPWJ97IJiZW3CeACG^oda)80AXTHFH?^cBD%fc*0D&=vX9D(
z5|YR@y4<xxMO3-Lv(T&)@(+oBmeT-onl#+Q<g*{YhlPF7)_=m4r`S_YGXbN;42XGK
z<`#vn+W83JbOl%}F<@2UmFk5*)vqmO%1kq&P0FT}Bg8djsJ6CLvVN);wjpvB_&GO8
z_oZSoJY8%TJGfv&?Z<A~<<|UgHuHtPWVwo$z`#?wg<eD<PuwG$U^mb>V_k9DhyfMV
z=S7d;MK_}6lt);j;=EBzReQ}sB*^>VK|Qs3KLVxH;g_K@lh2gme+@&3c#JL>=yP<w
zlz$pBIrSZ_BP15;yBdDv7SkK=tX|vDZql?5zI)pkFz$P3e~XV-@>O2^R)6h0*>;&v
zpi40i(50|eNuH%lS3{|;_@~J+n1>{8ZbVM<GL%MIF);h^R^@ZMWHPh7H`%GP)R6j|
zP+twL+XkLnPu^2}i@mD(LeTCu_=+IqpJ(JEGDB28m)&mfexWHfAs`B~kiP<{NAYdR
zh(goS>-D9ijtqNX6FD3Qh?~i400R9`0Wik_t~I&Po{(5BYAJ&m)rSqoaCoT07gz<h
zsR>uYwx-(JKk$=7%&iCxO~d2=h)h5S1eW<Ci!xELyBW_hkEK}FFY<$X)fe9$yHKCy
zPJE)bIi^6{%_38ooSX0_?`^aj66TuWZ&8+=Fwg9P7uAXU8y=Pgbn~IsiY~Y)_d;$B
zmFs3M+C{}wz8XF}mV~6o-NV%!U}f=4r09Ut?eE9RSmnLNau_Y!WA$@lRHJo7x7^IL
z^5UJG=25~S(vhd(V5#;#GTO^Nh$bvdI|BqUa^Q)AXw3cUR%qTM<IjN27vnFkk^R_e
z;SV&ewRv*DL*M0g7o(GLbxYJFqAcT*-Cat>a%#vAs+)Of;Kw^y3nr<oj_1{+qgw_|
z)<F-rSxJGxb6AGMe#i6H99yZ|DTonVfJ!$*_!6+T>AF2c>EbUC{NPm&`F|q7m}iPB
zSC;UjS>5^m+!peHnWh>75oflu6CDy5SY(}T0T2S8aX$use*_UlHo@v1^7le!c2XcY
zRhlif!OW;-GSuPxG=5i>o^q?{J-{rYwl%mzPYbI2z1E1DbS*lGRi;nH%@#G_**tWd
z`w%Ji32k&4)D2?_Ri2*Bm-r`oozs7Ti4kT07{mebZlv<}aPZBT`#Sg+0;92+ylXFi
z8YotlNDd2>4G6t6_-k{0^CMCMwMXbc1@sK^t?3G?Bj%KaF`DSIfxG{D>4)W?ShTZ?
zjZ??JIC`=36~VwOgvK?PX14#e@gBwhDr4P4b8W>w#O}`~*!<2y>9L(vxo(R4b7}pc
z0t=aWf9PqAE^*CyAGq?+YSB`1Sua#HYBy0_E%mP&_DU51=2vRZJxbHrBQ#Z8nY3Ja
zq+#&vM#7Ucq38LEiFt1y<kSDTq50*<55Q`P--nLd@_c*xeJcj&J94M$<6YHJ&WQzY
zhx`613qTZK`1^F20%^5Ld2jRHimKhLgzckTmS*;lh_s?yL=J)YI8fv3Uz}CW0akP)
zjb+cf@9A}^U%K$N7zXk;Rev98e`AvTwzhANPA44k=QogOgQtI=0D&km{htIEAXKC~
z)VCG5E;7PeRMLd-Q}@S<!$3^eXFUYQW<G7<gQBIC@!$35UB6O0ob8stzy)OX=QFAR
zIVO&+`s)KyGi2(6mfbFiOzC$mgA0tg4Vml%H6?p?_L+&-Y^--9>-c8?j4@QN%^O?D
zl0r(<OH+(N+&P-Ii)Z)l=tcU0`U*Tn=?%*B3%Wl`kBv(tFu>JRpS85@WO8kZn1s3g
zG$;_Ce8GG<i-N=Re(G)$kNRE}W;?k>ZsYrn{qK)tU2xkZmdfG3Ycb~9BeTDiEr$jg
z)wfo%BxSo44l@<E%oKYGQ56M^w7J}3l(iHLbToDAvH<?5)paZ=0<TVyYD%PQd%(UL
zYUhMDwnsYi%lFv#W`(gG0eaLzM?0bE57pIO+4GxCbMc+??YR#z^E2gB-3;x=bw#{K
zeM3)wbH=~c{8bpww}&#)_4>)mP4hmG`X&&XUNl~)I^JJ$BjWX!LU@q9*>Xn<t}G*t
z6>dO(XoKfh)EyD3X5)->b9F_;#IiN~7InOrOIo+~dIC&IY=fMg+Tor?(Tuk@x1M!o
zN7A~oelA1H!egFzw_yIXqFOa|IM_(@SFf1svn)#s<5qVy7Eg9pVz9_tC&ohidmd#w
zzv^%sBgYD}1O5_oWxYbA{yXU-c{_zy-2JvgJWU=k;N(&4qvP4kxjT<Ib+)Ig89Bci
zcAe}hdbRB;9{hZz`V(wyHefG~gyZGqWr@hzgO%!M;e8Qe<2gGub29d36P2#nvYQD#
z^B^uN{$(t>BNa|{=2n4SnMtf&J1%b2`Ow5RzPYOz_5IJRN@&XO;;l5KSd42y$h|ss
zvYZ5|SDY$9?#WHuIp?9_+Q}j)la_n3PO~@W1#~aaR-#tzurc;&`YO>5PKF(3yF1x$
zs(IQu_ju5w>!uPqj!hb5lWY=mTL9kbMZfG2s7T)MM8g0p+PbNXBOy=fdrUtZK#PhU
zu>Hc4vP~A#yCZ@;gRlv5C*{y9Qq<EgoXp!JPUex{*gf_MuUE7p0uF%QmYtUd!)Zf*
zPMTL75PRBg9UT@-2n8$$@cT@znre1IOh3W$QMWp{lzq0B4Ly>Uit0?Gw~kf=PF_cl
z-%A=5pKZx?i8HK>?bUa7FapoIpRRcNm0IS($?>d&Q_|tm(9GCjtBbRV^JH$BHVj^Q
zSKs?&0=c=lwR!;M9;!Py3?qLbe@jqMCED8Bn&Hg1s*&YkjB4~&6TOCUaY+x*&%R;i
zd<v~gbAigJ8Vyu821Og+!mgH`qo{&|raC*o)%EH;9H>_I2G}1dX<h!lc{=#l{y`}h
zxl-tL4QICJ^Q%u8cCJitcA8wlRB4P3$qbT|?w-@6a(Lrp??iI$xl9Z+IpAUQidnR4
z5E19w6Jz3OEVz`sxykU(|CZs6y==KpXN#R-`uf~5tWAa4aI{MFhIITLYk~7#*Q`>P
zgibbu+;+x1iVl9LiH&9U1RQThpOt@7`ct|;vdk_5F^2m**20?y%eOz6{JFeF-(`(e
znI5oB_CJMTj6q~^AX*2$CSoF}@A^$}{`9lfU5Zl@N(Q5xq|*W{5C@v>w%OWEmMO*{
zf)7fIZa+^6>)^feYi05PUU^-n?hZuoQ#vn~a%}s;X*B%<Lwf@G6_8ds7=sw)OnjMc
z-R&jE{x($<3jb6<Ti60g8WEk{eR|H^bxoQ)#^-4q4RZ6M6xL`^qRWHoc~NWP6EgB|
z<D2{C_Y!*3ZsEFn`KiR`$c-H=#CLgJYT{SC!@ce;OZro8Q?nl!gCa?=<E&fE%O7+!
zvX7nHvrDxRrfSrbznHwfQ+6fT;_J4UVCYn4-7JZC01d)xs5XPc7B#ofwx`!6!p{<U
zdJoGpoeV$H?x;2Pxk*RJ+2`MQ(tLIR$rY0M?fTn{s-DaR+mx>|&SqjQ$?F&}@-mkp
z;lGGQsJ}^yQ^0z1xd6|$9m=`#w^2e%jHXn}bv4+H+$x>%XGI^R4r5%sk$bg8{e8j6
zVLL^Ab)L|kFJCXbqF&lX4FDZFdn{{WRn_}mPR&KIk0hQre)wA~6=Q0TmQ2~K(Ki1M
zWF_ey*2*e-%rJZ&(BsF!5v`}kZ*_D-9SW3(6U1XH%gToFX`|omdf(<;+8fr2bTiZo
zeUcZk@5({z`N-J}=og?z%q$kd@_EQ>%Itvq=fe6M!3tWs>Uip0;$*{GH-=jM*8}=p
z5SUtyrO=SjLR(p%5cj!>aeieNNmm2=qsL)&Th;vG{r&j~NaK|b2^cqpVAg*Kwv8l-
zVB>}rYykNxHU!w!dv80D#vy;PzC5uoS>>g<^^1eq5&Oi#yK*G;fd@3(V!JajK4k}T
zCvqEy!fZ8?xgKKyxGmy|yuK6ZYkAumH>nQgKLJ403^mPuFmzlfog!P{hAdjGU93ao
z`Ukt;vt(>Rl-QI-M|5-7*Q3lWEft*ERHJD)WouUM$?bll<x&Q0JYVRCfq4c)3!+?S
zJ}SfH=xJd_s^FHJt*8zh=|ek@m|xnkRwO->*?rKBSf03j51RJ^sJ#%#pggErI#&)*
z6^P#fCqBVG{_7uG4@;i29MGl?VLkj{LK9j5p*)P~NodnG$WH(GyRc3VJ+bS=za*vZ
zg`8c761#?uE03QI6+wm7)p1Wn2UK7PQL9TgCoMOmozS8U@pXr@IvlGz2US3I;Z`p4
z>arIE*4`+{L~+Ot_t&uz7Sev|9WC`CkKO#&R(T#(CoUx$4UOU?Q7Ao&uX!Y0)uvU{
z6ml=ccA~Ct;AEZlu4`nD_|}J~RC~e)41|McUf&W9aC^v^gJ?JHlV?u-J?=+|G%|-K
zc>UDq%?ASS%}^OcD0h!ekxt$vrYgiRAElqbOX_Fb!k&R)1hkW1{z}!7px=_w1P<s{
zTSj$6qtIysphfKuWXqzy+yatBt%{V#O3i1hK<eBWR+;kT#a@*{CARqu>-6;N#M1kO
zf`NW$C6(}ec6BZG*vAkm;8td2D++y9z_5#4j4FCgeB6(J%t23Lr^PfqeJ|)(xjOFZ
z@dsgZ>=NIPzGkMUr@z~hGv-JOG+3O&y&<MK12{YyljNa<a7-Sq)G~#WW3xlN(>tth
z;t?3#xY<qbZk&Y7iySmhoG^hUjQB&_OfF%`zO#q8xaWiM*_O)Nv5DQ&lQ9fPpfMr3
zn;b`*4V1$_XQXHgjhOa_Wk+xBBP}i?Kse#d>o_rf*1o>R(O?E9Iiu3u^W(>lVf*uM
zh17IvH&a%PtDVp$OLl%!Q`RII)3Ii~MW;m$3<&q<AIVj%xSBaPGwzSh$ySu__6M{1
zVB5dw=Hsib_PCV{EV(s17tGGik|YlVCJ}*=H`{ivuFYjObXE6$nwH%fr!5K%yqOo3
z*p=m7iv;DK3x(?Yh1Wi~X<6(0=;))OOsZ+yRwtrx&-9uVI{m7qUBQf3Vs7xTD3krf
zL&f8>*)_t#!qI2L{351gb<i-(*RSHs+TE%cIFQWaMLY#+HH-8~0*m=sp;uwu2H7^l
zmG9lD-R@>INRlH(Oa<_F({7&5-!zev6oWk)4vyOg=q{Ma2jySY3p577)#Lo}{R~?B
zTPZSjo$qJgW!hjfdEWcq->loXrZ=sL|EF8De$%ZgxLf^jf1w-tzm;z4FDd%|*I57{
z<M?Vn_Py3h#;m|2h9E{rD|bCTiX2kDY%5(tQU#8*rC55f{atP@GPTi5wne?~;hFHm
zWgji;g7*c#`OysI*zNbNtVh!zJItE^1^t>6?&j^lhyR5)1|ovNGL-T(5GrhW^e2`O
z4MLcJyRaj+#=)|eK(#6Qsh+0p$UuR;J~8B*t&8(ksE|H1)8(sPqIJg|r}}E4Fe>9)
zT52Vd5^FVF%IVds*aLEz)~&5F*|k90`O}=c^@{A&HJY^zIr=Q3Q^KiFY?}g^yiwl4
zQSVdQZ=w66+|z)fGS+1fJ2|t1RhhJ3eW*yG@6zS&(lwx<_ds0<K(<Lkt}771yheh;
zJAGH8icdQ#%i=OWk&GR_(jJe;k03UBT)6YxaP+FVNYggPeWa;K)c4Iy_>4U_bZeDk
z-zl5cgIhUBaj<ZI=tOnaej<%a+Ra%ENH0xQ-|FVdO*{oBHlLznWGt>9Ags$0ui&RA
zKLQF^zbT?cSr#3KAgb!Merg7+>6MN1K4-mR3~N4MrGP#@o9?UfJ7cq0qA<~~T{qz3
z?ZP?UATB^uWw)D4Tw;KCh5xRuWI71u%4UzSVg4LKrg5-rK`^R1DRA?Oo=I-uI2>U<
zen18(Pga=+n#;4<goAOL7Ji!AnAej)NeR~ym39?3j_`ek(y=y>{nIL}&I;WNR(Ble
zLWjEoJ;o~bK4m<osf%Q#o>&D-{+SK8{nJ0K>1@+t>_r7cO)kuG0rg2_s0-hg@}=SX
zri|VLGs1u^Hx%QvF+;9ARX?<ide@zgotm?u4LSw!EL2*~zeFMVHQN5mA_H1DW=J9z
zFP4~hmZeDg;L@m+zqO0ikUFrVo;&;%^8>sP8enSyBE`{sxLc^NSp-+_6VU<Wx<Q|v
zpnlUlU$~wHp^5^_6B=|s*o{9t#uXQXep&p*v!d6NTbFLnWdk_ZJ1;iZ2u@(hpKo$|
z&L7BQ<1?q1swUMl*GdL)aCV-we{@5er$VwO(cYuf1vlNT>^f$wHEEZ{VOO)UY?Gui
z$wzsbNR_|d?wY!1Slhr)DWliD!2w{&EA%^HVn4LNzr1(om%E@<3J_lSqly~DWlE!S
z@j?Sa1*UgXE^e4!Z>ZOwJh;2v8DGp|!?=rVsPFeJ>J`pU03uxBryO4CG(M+Nx}4Z6
z?1@VYYlu9!vChke>0etbn|VFq1j#zibC6e7f~;XcI>7R<8@I*o=RVQOOJ!PQQUbeE
z%~G<h#?{k394mr+gSeT-bdY=3YKYKi1kZ-m4UC6`-ZVVV05Y@Z8_&b@!EWjr0ydHZ
z-FmcsYIyZ=MxOk@_b;i~S7!2>1Gm+ZEmWwLlw>yaBBj|Kjnwo|Ztyh~mRsUFw0`)8
z`>rVpVSf2z(>(90Th+MJZt@5I_72YS2`(2ynt0wEJ3=IUPOpDDr)ySzG9gox3c`6}
zvn=3Cy7!EIjS>H21&mrDV#dCT36aIL@TFw;W4-~1YCxZ=D*J??!R4X~qSpfi&a5y7
z9TSC#kEK{I2^0XkUr>>mo>zp5!k3vhX4NS@Q1WtBCJ~P{Fn5903ffI2*E-z?APmQ7
zs`_OFo2ES{&VPJfRjI`T#dt`=M7QyFSDLX>g!Y+f<<BWrC7_rNjXUQPfEJ?UCgV)>
z*@{j*+%vR?=JhWgO;saTBhNx!0S%<E#6YT*Ka_PEf!^oeG{$_Fn#Ml{noi}F<nPI=
zOp?QV40}kGIfi{?`H&Sb19R|1h7Fio{Or=2B`?Sp9<jt6B+D8*fWa-$G|2=lDdD^V
z`ZHEr%PEfspeZ!yr;=rf_SY#HD1yo_9Cfx?BL!Ji%RHo`lXc<w-{p?von@%X&>Rcq
z(Ec>Mt`tkuqSn=(nB>e8H+%r>`F87Y?OIU3>^o|Dnf?3|6Q4XQ);0RmkaCuZm_+5w
z&*2~Ue<!yX-(#}BMI)kj0pA@?HOKH6(G;{(NpB=9&coO`%&v$H7=~JsY!?0S$%2>4
zN{__@6dB^f!0fWy{GACt<(4mX$u#uU^cnYXmUX&PV*Kse*i}2}b+<iT6MFeV>t%pI
zfMAbfw4W};@r$Wr=W^-1sr)pg7Ng4XH-I(?&FypGiT7}j*g()N1z@+Sy{C1|2AU#Q
zKG<x;EVc<%t80w!n$4D*pi9KR=G1|ZQ5S(xf}a_#*cQGuiqhAS9T9lrY)z=JO2;>w
z46ZFvP%%rL5eFnn=WyXn&E?Sg9RO@peD+mWQQy+92Q`<^i+S)VD`-%Z;|C-;))K!>
zc-62hDN3~tJTJ<-rHW6#kLm4qXtk3N1$Eag?Avy<4)+$m;7#5`?6w-L%;%Ptz&6O;
z@Yt#Qi%cbP2TGCZ6L~Tx+jTkKYt>XR+Ky&7#`s!oY9cexVZ5~ja3ZBmry$JoJYw*b
z_;yf)953Q-rnT?~(MKR(su+xPMR1t4zbuQ-Pq`A@`g#I)fhq4f3h0wJ^IX3G`Q5zZ
zT35a1zzs8^xF!e@#ie%@K%NqIU5@r48&{3svHQY;eA#BlpvXhy-IB4(y0lO-t1Bmq
zi+#$ti{%d#nl6AZe2y<F!3miR_J-Y@)}e8T_^<A(feLBaIn0~tjmK#72tB)-5Bi|K
z$g~@?GZ6I^?hGzyN(~wF=;xd93#4%Ih4_j>gJ9?5#WYNSd;^OVdbYo(lE-y9d0;Ej
zB-b*z-?C^x2mS4JI)yc&a|Q^dBvJDYLy);pFTHB;aAn;JYMuH>8x+qx7UaD%Ehr&M
zKSLgLCJN#F*&46E?B!yb(_0~l>(-4R17jQ0g#jVe;N#CW7M>1SyRkYUc~M1RKtt%v
ziD%sd2$Kr{_+I3C`)~n}S?|4e#p$4*XcRp~pio(PX<2zeJ4pXKwn7745ddtg>$+&$
zzpYeH%cwP0p0xLwo*`0`_4IFhXM#JrrWn&J=hUh(S%Jnv8VcNDNLyV1EX2@`*tt0Q
z%37L+@r<^qB>JfmGIMPD4h5glsBeYl%z^j9OTA6Iw4jaz91f_^qEq*!SW2|5h+vXr
zYBvzxC|_?xSyszopT(4uVnmM2UVw|+E?iV`p3G;u;LU!whZ0iU_XovQ!h|zAc=)N*
z&%MaFs+&PKpMW<mB^Y;c4LlLhEh^@Zc$pbvk1|7?=(-Hgr4|P7dqH)e!o$<a(lAq!
z&3A##B47>iPxSRD{V>`YFnHd}px~jz*!%q$&4ZhC+}BDyD)P$803cAYXFNM@z8{l~
zL3Y+QAfEB*S)jcu`>_S-(8Q4BWQUH0HkHD9SC+^JjPAV^6)fbSX9%)Vi*y&50M+VC
zFSP6j6q9i&g<*wb&Ijo(VhKQBTlbDg)WgN!sJm9g0hcUO3lYN$qIzM2GIy?cA{Bac
z;BZD8!)cI_e<m8|@$2M@aR;cX%2)}ewR=<d>6LkJraC+)bk1u+`>i?>LUDO4d&6Re
zdh-$j+p03<pB8grc@5%T>ilRuxr$*}<i}X`N0IJ&VGm41Q*1R#uahQg;7TG@22MpP
z)ioCifiNeQsHfC?lqvr_yl@8l5>TA9B>Oz3e2#<GL9Sx>Jab-af_E?;x&G-2NpUuK
zP#=~?Ev+Gp0Y%(G|8-Y`#KH@yp=s4z|GN|Y-(sIk$7;d?Y0#PNDjB9Lzy(}hce~pR
z_(Wl~xJ58bGquyZ44sp>!o!`G{e4FiM)@%g2BrgK6a5CurBmQU*@G){PiN6mt~(D)
z2?`xY3V7aL(Jli_Z@jB^McLwPFw_2Q$<IoxyY$={l(YJ52}HuOPZNL$fnbu@xe4Zb
zJaIYz)$td^Z2CSg3g%A+R_Pi_{3sMr$7ZgJ>mmwnMn5L<_B!yyVAOSpkG<?(eoY~9
z>An*8ZsmCpgm!60^R}V`VWh)7th`jdib!2?vH5#t>c_)EX9;<dm3VC1!;sZBgY0Ji
zG?ZhS`@+*N3Lp&gbEBL^)nE^rzg-nKU5*7HFEQs#=gPkce#(I#?Vr|2Fs80m0CgSY
zUR8pgbqar4w#vo4SnG}CY<I^<6D)$!Q+pC9>^hDq1+j}-pZ!e%=cJ(COX+p>D4?^&
z>g1%1TstlUsHr{%0F}JAK|7i2T)J`<?dTX+b?5oqB%`h!718`O7U!!s78YQI&Hjb%
zj<ZsB5@PU0CP|#2=!5{K`U+TPUYR<kcSuhduVcMU0`EA|^hHNvxJ`7T3xmEVe*gK+
zG*})i7=!SL_=2<L57~MmrJpSls*uYOOjQ7C{j+2p?lruiDx>upCjFy}tILNaWh%1V
z+prc%a^r3)$PZGQgS_QPSZy<Fdp7eiUOk$D90|9}wSl?T`WmU};5B;+o+c7zqBCo|
z45pQHUShK|kRla{h+O^oIcP3RR8dC@)Vuk8o;4zB>Q%EHX<*AjlI7pJs^|YSZ(j3e
z$bF$Lx+QA5&fr->Q)G>p3;J1QXQ3At#L<oU9^)FMs_WOIGpXD{vww-HC6~pEDwTdx
zlIM!pC!#8tuKchie%BY+(1T~UG9BbSJRXkaPI1qzG|m0KGvgEOhI2tYMOg`iX3m<X
z#dh+35HY1>d>0fGFHgr$<=f@1!Vd2e(F@b1^fddoAf9qy8B|7LsAlX$OW|!Q0^2GA
z?h!5YB=E1Jcy(fDhAX&MA#JryGM9-Qj55<!FEGVBYgb(E9Dd#u$Pyc4wjvp<MYqeM
z;d$i@Bw32bE>|CsYnEdsX~?5TW_O(ugXTzaQCR=#=r{l^P?#9%sKD!qSG&8lJ=MlP
zX`Ek(H71PC+g4~++-50y6FrxU?;>s?3UQLb*rKuNK==2}i<~|}`H!$)8s^g~ky-gs
zV7kY=E=E$@EpGgT0BH-$qk$79JPGsDfaYrDr{<@`C$7#WKDn-#Rs>j<Nf_9_eUzS%
zN8TfCHcuCqyWKjMcj1J1Fa#SGHPsUJ-j2qPig$TX6F4}x1kXSwjTu@%Q7Wu-ESxpX
zZJ>eQ+C^rAa|sGXy~k}cA7etJLB&B}jezP>P@?kL-y$~YTc~}>Dj$_}v%H3eAc{Yd
z7_ET_y1!>zr<<?O`(9e_frb^M8UA~W8y29;hje-RaUiI-bfNgLEV?`Iz_AGPrK>+7
zFZ#si282>Tzlq64%s?rR=sDtH`=~gb47`7u(lS(1w&PxD;!>Q7YDRs`^$FI_9}%bZ
z4!?a{K_CVM3aF;bYVUkq#?bAlkZH6S5W}-gEdT`qQz=aTC0gr8-~cX@IxeOo!5<sV
z4^1f-OULXK<5Z}R6abu-@T&4QdeQzPwm`QID4hSydTGnh=mvD}c=kO_-*YSkgF`E4
zuIc3DnBiQ#ijPyh{hoBT1$6B@$cPty8i^{`yUKmw%d!07C}wapzn)3q%c;q9gB515
z%(6t*XQ)^n4dwaVgfqkaPjn-F1&Gub>I{6@00PD<7Kn%{s*mBv>d0)&wDqeX^oq!b
z``J1>1yr-KCHz#`h<UpqgUkhZj>;sIgi-?I_Cc%W+QK+}NjHfIb^n4BhfK1OXfc7Y
zT_sy)x}=7lY_K=peeeb#30>vFYFugV&jZ{YEDETW7`%`*!cWjCPsj)yMZ6n+ZY~)-
z@xlOc1e;!y$Y_Hr(kmCc>#4vb)wr<f^*Wb5H>5Rko%thl*piPAKr5NVkp`fRKtTXp
z?B!EY{YhHu6-fz<By6)*JEdXqi&8<WzKh@L0@6De2;}W(Jw1TaioPyF&llXa64jw4
z4fX>{tJ7xvbAP(0HNp9!7NfFFrbEi{3Ti{wpXa|@oP|}oMaKX<6vCz3NhSh$wOKp>
zEgOYmYDs+=X`96YhTerJ|7)%VenBDz7YE{FoE^uUd04_ucP;-j(5BMmPNjA;Fs=%f
zoS86Ox@uNC+P;H=v++_%#}2kt$``ZV^&v%^D)=*t;xt&PHo#Th)}lk*-UV>q26-Db
zxU#B@$Vh+BYb*_h`pAJ#f9i@NHXEl5lIxNL_b>BH2`x_*Q$`Ux28u!nx$VQ>_@|C=
zD&4uD&>SFvC{MH!PEL%ycC-f0bSo8I?re}L!SuR%jZZFq^tZqZbq)`mp3u2LGwTzK
z!|_rMrJ5!seM^5HAF2QGjh+l%1p{KXKO8Rxt8L)jdf*mmazR*+S~TH2(!SCxk8VG)
zkj8relrE4lB>SyLf`3Z}m*X>H8po`{W2%DtzQr2nUSYdjXIqoy2(#rAMz9+--K6vt
zoz@&-N#ihC&NpId#uhD{6|#c7ikKa}>M1kUIYz%X{yDoQMjbxy0LXuuKB7>^cJ+<W
z$&ZXd2>`j=VT%dwY3rn9FU!Bwb`P3J?JKaQVt9kyv<-7HF0oK4vGCC_ENsRH=Ubs$
zQ9P_gi%BIP1OZdd7JUiAnsr+}-wPZdVsZX#Y7xr{?b&453kujAGxqwWd*L+WpbE>@
z6ybLeIliD2zG5UX$8ZMuM4x}gdCmKYV=|DlT-B8sa=hOrJR|)X!fvOrhPGXb!pYXE
z0yrt5vk73~Z1bf3pfWOMCix3)@!Jg(f+OtX87yG;K)7Qk{N9+&jUjJW&2NoA7&eU*
z)d_n!R#+ANct5@p`?_9d*(*9WZa<jJYr{_i1E7o4Qk6q^2aIGrsd2Gp6{{=q%_@1j
z@age@mL1F0j1{s<84lLEv5ptY{Ji!tzS3GazaG@rK)`kZ+!f!#cSV3fW2Ku)7Cm?0
z;00lyU_8(~;0aoAwu6+jfD{M1i(0DbdU;X2q6R$zUvLIK&p7<^pP{iVz?85>)^z+d
zzO&I_QuJ=PGR`NsDp;?mY;i`3wu=aA8i0y2RG18{s1l_>PFt+eV#%y_mr7P`S18RB
zeInH7=1crblSqMW5K5BBJ4QUBF`D{ax)>lwaLq9Ag8BplvY2_}E)pD!k||AWi)T2C
z<Zos2H|$asa3>*6KYGS1P`ssC;Ho_)6{r?8){nik*>=CQ@c%>R)@gpviF>PJS$<gs
zv9;@B1~6Z+TV60$S_7_(2EjO-(8ow0;w}OXNK!%K!-3`w(9>VZqD!|i27a#~3&rHE
z<dYcX9&y1mApQWwVWRHl$xULS|M-bQ^)NVXS`I>v+2x4VC~#KicXcrEWo-S>LxJ9v
z5`&ZmTX60rs##YE`I00*-`aNj8j~MGqrM`_j|EDR+up-3gFd7aMt@GeT+lJdj%~0l
zj9EhUi{ha@9G>r?nlMRr$GJoT6M%av4c`i~ac5i{UzoFzBdMkV{MIrTN!aBN*Nv#r
zVP!+^)<*Or90!ELllxD4##1Ql-RFpY0Vn=-rLqTO``i4O%!IzXQs>W{eQ^qT`N8D|
zx(DGU!k_M+a!(NFJIC0{n3VY>aQfnHnr6E5fnPdiAuSh~gx`GoaHrGrtaX&N5<;MM
zXKb|;NvN2Kbj2^iDl+4;d<Hsp^icf7+8&v?(d<;&9#`49g9Crs^5>6N*4m~p)Y8@Z
zF4lFHe34h4>C-MIh#_5Ls8GXmwjbK<OwV0&eES&6VcfoU>YlY|&&tr#OSULpG!HIu
zy73c5oQroznseaXHdS||J*@i9(M<5B8(s6LzO_S-^~dq`0r$rA^)0#x?ZA{uRJcrt
zmgngNE#prZ#`+-LjMvxSKwiNT+r61EF|YK`U3+tJii2DKSm7TdYu7<@^qj4b#V@v$
zURa`RQIc&{$nS>CsVIzKqlopQNW%8=ErfT8eB~=Qq+`Lt!^BzXG<s-NRNBkH6-K{{
zDYQ@cTN92#!%9Qj<?Yxp)SnrS3|xIbZLF!N%hevA*&G>bDX_NJ>euKJ{Uvx`Y>;h{
z+o6-?tsU2?qKgWn567C;=Hc%wtbagh9Iul>Q$ERC4pWT;?Zl(Kn!6itszg2TExXeP
zDDF^(xov3lW6tyH_B}txDOxZYcINOQHcoB^%=A=ddeY@VAHoMwTcd61>IqGez+tcV
z7cwcpGtOOWs#PjH48B<LW;4*)uI+kbNYpRfH%Nj7QxuCL<lJ*c$JuLxguDHW=dU@Q
z)L?RcLQhY(-|1wqJKxHr;rheWQq%p#QO`Sk<NXn;snN{exhb!O54feM-w{utzwrk5
zg|>Sk$BHU%=XxIe;^2A9CnU@>#~n~~gu9bJHY+ccp>BSAM$c&p^GWx;A^X>WEgXS)
zKg7bO?{+SG1a;HzaffGMKPnU~8F(ID;dvYQW&KT1QLSh2_cE7GM8}4kBX2jNW5Kz(
zrL2Yi(;!m*w@`o?^m5Z%XS)PZWLIwUd_BspYmSp(f<Nu~frZ~jT(0oW*?qgFNt)Ua
zzBi=xo3c`Rl9qqrdDvU&!2%DfGVNu?3o-3lBd@eW2kQGQLU9yhOuF^9YDn8{Dsp+P
zs<#$dd$=~vHO7E#o%2j{f~y&aZi=2;a?O7DlcxTC-fVZO^0^C}!!fOiDd`q)aoFZR
zs9V;k7#Bz5%>^_W`hslzX;m@P1ran`b}OS#@H3c%!>h-QSM5JVtxu|S6I)kC*<{bC
zEcNo64|KdM5781SNB<6uFrKHr*3@PFF+HhfYY*ehu{4sFzBycS$LWU<yHROkEpB7k
zv^m}RX_U38(Z=Ibp$2@3lj}4ps|gJ6U1OqA(&?#>nuhE|KeBYCnRX72en}RiTcZNb
zRT#D6nq*WMrAsU)YuhV-y@l0k=N0ZmJR(S|VDRcG2Agdcg8Ods-C|tZ8hjm$5gBxK
z`MAlPdaOL|yvSTQrhHCU<521_m-g;nhg_JUmlPbjpKDR~GBe7$mbuhSJPVT@<Eb7_
z=gt&YdO!|Mu=s?ClPZ!9cYsMDhXo|W3Ll4CepGy@)aTh#vn`F0$#8j3@vg$LI3b;Y
znLmWee(ZHe-LF%)cnjp^;6D*9w&EE5Vy^|V^}R^#?$&T3l}1+Nooz}uMODZ((Zn&}
zPM`4C(u0@D`*AvwJ=T4_KkAdF!wO_K&wJ6mQ01|ssf+kLe7hq4^1Zn@WxMrXuBCYo
zNlC9T%JJ6^C0bk2=Qvpq%)qcuiswu7HKbyx*%q2_xmmtV`4stl@X6xa_Oo#t|2z^r
z?j|;lLFc_&tm<|uc4;Tpj0WzVW+)vl84l^AQcAG!>FecV%n(tuQ)Iew@3(0F*gWHD
zQ%*9QZTs#>sFGpRTIhI1$er_dA6Qv-uP|PAsw_H)E}f8ts_<2n+#C!|cb-`%Ld%TL
zeJp4H5~t)6EcE7L(K~+@Er-<UHO}LUX65S#IYDW*5KP-%SFF=7vs->pW)t(un2WyV
z@SpDUIeljs=jReL`wCWNyN6^pt&%^<`r~DdPtH}FZJ!p2^pjEHv&?vLs=#T&BlvGI
zkXIs5U2{x`Rc}(^v9yA@HoFm?$)35FC2C*$siM$h?=a)^xR=9sZKi%zk(l(~s;T?A
zti$nq?h@O(LV*Y$f)YH9w)+LV0^bnpboyDw)q9$TJvEf^JW+K91nCj#j_e95u<N`m
zk3jvibFHM$z*u_p-4<NFz`ki;f(tQRdFUW;>sYdtUBTRS@OHZhG{t})<mZ%|Wx2*W
z_^YV1<1ZJE^n<G*A<*5HF0;wi)Fb)pHdllYzat*xZl9g~oqp}}eAA^>LD}yf0aMjQ
zlpd!#x31gQ5j~twR^C;LIPtQE!LI%+{JUC3e$Qr^Z~OJ#4R_J(5u3Jn^#)Ty)6=NY
zxvzRKsY?fApN75ESK|v_T5UW&{hWMJ-C}_!@x~G8rZ2rD7H&s|lug|UNVIT1e!ElQ
zyMJ_R<%Fh8o7a!nT-HZ5cUtgFGVLy8x_y1iF@);FL&C-&#_<V@o!}G1z26sNa4r_w
zuI}1*9^dG6Y=etov*@mve^Lx*X52ihw1<|c6`Z@YRQsmgyZb$<qoo94BT#P3OiU<-
zcRmYx4?ID@`mUU@`MIBa^15Ku>t}1N-?|poZE0>db~+$QJx{CttHafMT=2dGvl)B2
zV>!#ijtOz;_4U+W4kucI?0YKX#mQEgazTF_A)`oU>O7PLdvCL&>dr(FRN{M|Z@wiF
zrQWn2c*de46c$PmFB}eY=w@v}J@Xldd{88;VII_eI(W1ouX?)K)Pj$t<3R^sD-&xG
zA|U0bVN0xHTH%7s{DTqdA6{-=_lL{M7$P2B(!4K(cjz3>8}gd|VQR3SdsCI61U4}4
z>Zq+#b&t#6+5X+*rJ%Ll!7&j{Ds>KNv!v3uwGA$9-eiGIwjf&&JX7Ch!ygN-k>Wwi
zZ+WOMDHWU6>fg0ev|haw5OD+guFa|um2mT3=*7v7o|C&JP$yfdNdH3~3C!_qCyU1P
zq1(F>N>Z->Qyp`rW~%UB5xWLKP+ZJ$_<l+1hfn3vu5bz9Zx#xwX7AO(3$0c0itzPb
zmlgz%rLk9N7W#Hq`n~WP7hFX;x)|+u_2LiLqA3PceG<bQ<Mfv_de;41o(K`yGyK$J
zX9T<WSvWuSu)?J7IKjKbhc%q9-4<kzbP_Bc2aVyKzEu1JoayjId$p$wBb#S0yzF<s
z-MxvJ3&Bjmn1kg`k_%s6-=<Qv8<~TOCR$h<bl<S+x(c#CV`F`y@>gr3xmV+H>W1!)
z{nvm@8t)&h9Pn<hEsf*(w-u!#es9Nh9RBMpKq5uM%~--02`BM;kIg#LsQ;{9>H}%@
z?vC27&WJ14molJdtrX!UA0NG>>)KO(!17I-qwRjQQ4V|Y*)<#<CtHRVrbBxnveW71
zn&SD~)K(josBRm1g;OZ;r4Da|=bvRJTN?XhH%)o%cK%g~%bMkb5fYM*4J2|JL{&<X
z9~mbPNLT8t4-yYuVm2I~exLs5sKMnc^PTVKr`x|&7qcx?)M=b6))y(reCxko>BS!L
z<)1z`xFRUZo_>2{!bOYgkH*eszqYhE9V9aK3!nTho*vhX(#NETDM}^tui6+(ysynm
zWdP0m%|(y9%4eHQ>`kSB^UfJFu9)%x_vQVBvPp%#C7O0R{`w&#pnRuU;*qTisLy<1
zQR&}}I&7P@Idptl#~ZY{Rg=w5wBMUml~84AOq>vT84U<cyEv`t`qq2Dd54i-1@ABn
z3O~8=eoOoP)>J<K3+SA@JO^w0&D-|UK?A+YA{hg>^KVrLTk$zHjw?<F-W#nI7ihcQ
zg%Hb%-!)p_0w#WakdwKY$C2t$b<5*?EOAuqiC4c$H?Qy;Ud5k_Z!4^#Y2Xj~Jd<g0
zVgF*$Yh9IKMSY&3E&Xby$Ir%9&xf@L7qX@P6x<yce68`EQ5t(^S|^Ec5`f5`9B`DM
z=K5NVPo{l2^$9fn_{;uBGZR@NMZRe@&NHgwK8G@(z)lZq4~BkO-@<*;;&A<%bdfb;
z2c-1Ha%rsOt_PO=`%na9?B?@HpUD?>1ZK8##)*Rj`z(ha2}Z~fr-K_DoYu+4?M^H7
z*Fj5uPoL^Jm+yUej7wU`dMk|+XEqR_`Is(oSk5}~>b1G{@^BIj_To%d+22C*5L8@~
z4vH6N=!hGqiuzdQe*GN#QhT2&@WI$83`%fN(!J$9LGgiw%(Z$}eb<@9h2Kf|_eTlS
zTaJzXh8Yj8iiUcvp7K8~;hgqM-IJFjj#JBheihVxA?aP+n+eYmx$C-w&m|Va+XmfM
zjlWEr)-76Gtj|$(FY7m**`{u-bna?H&A8OTGw$baW9V(osC!hye;;cpAw;Iz?+Oas
zD`W0$J<G>4<ehB1a?wHUsM2S6<y=^c)(`jFCN8~3+Gp!2udP_`G5bR|<#v^?mUT#*
zwy3Y3f^qnXWtbO_JG;X-XS&Ha88+WxL&TB~)}X7m@2~#|P)EBe<SaBPd6)djztIN(
zX?UQ2lnGo`SqfaLuzlrjcwA8a5iZ<*{?R_=;$-?`zIm=hdgq2p<*t?*3%*E$rS|*f
zN!)Mh_w2fw?Hbh=BxKygC*Bq?TZ={nyAYor=3CXZH)<SFtDEkhhP6{0q!ia}T^@<_
zr;jZ1C$}kwzUN;1uF~#$z}Pt8>hK9<U&X1~zAc@`t9LG=a|&!K94<2?;)7&sLi%1_
ze_)~NoqKIie}e5F%v95DZZq5K=*oakYrQVrIY};0G1T{;$<1l}dJbZ;TaAC4X4vq}
z#eXpAsVsv&l<y8gMsHnG7>(^r6D!^yKM5oR9MrwQv$ru*4i}#w1DYD`Wi}oQ+q1##
zMF{nMVYJwIOtyc@ui%rlF9)t;51!O5AN+)iJU7ZUmYw?REY+F*)UWnL`UgUN4qf(f
zo1f+-gKMsxac8=~VVgXsDOAn2Vyev7oaWfsY}e&@NwwaG-Xu{pdA#UHvM3wJ?@=d9
z2R&Q};Ea07!zR7rU5>-EJ)B1z$Mb2cneIqawmY5YGdgdl7=3*0nEv8gv55#T`MWB@
zWDQ=ubGzo+-B5LhAEpC@A*T0svw{>gD)Mqem+>@~9WnN1)UlgSpglpVe36B2!-#>U
z+OI3l{q4?(?8&KuZ{8Y0cKK~ggQX?_0%sH>DU=d!-CzQqq4ArR%jCCy60jOjx3c3m
z`fSxc;B_}Ec6VcFKjfM))jNghlFP{5FuCp<oV)&U-N(O(6){4kC(QmIT+1ndlkeEG
z?<IrWQ#j`=6!l7f6l0koF`)JdhNSKf%{ZFLd-m(v831k}GfGj4XI&L>GpCms;|R)%
z(<5*3Bf-xqY}c;Epd!WOXUXQW8IxQ))-6<cL_exz#NLb)p)Hyb=lm*yo>#im`iN4G
zzC^Sl`AYml$!C3@g6j6tfzAsr(yrm&`k((Pzm|Ae(Sc!*!aF3(>VgldnIfm4_1!C2
zXbX9075fW%`>tl?MqcWtQP)2kh`d*xHC$C<BV=zy#}}Tji#O!-bT67&&b0o(61Op<
zIPuPPnojejW!EoLm!k`m*P2qE*u`Ow18(|v41L)>@kj3U2ZTaz_R2jWPR5t+*6jR}
zg6bcA->&)9P;gT$F{V+lvGX(j+V6?Oo&=fo>ZclqRE-gl7xk@lZ4a83EaPJ*j-9S-
zolrKL?O#)u-Zi~ac=M0@B1!JUzu&S7ZAoKDzA`iub^Q(%^Dwfg_MvMTXQ}OVYyARr
zQL>Gd&A09vOAlDz*Lx7Q__!us_4C^AFGMfF4*myCR~;4A*7l{OL%Nafl5V6!8i@hv
z?ruR^8l;Dk24Uz9X^`&j?sDk-&b{~j)|!7<3uZXyJo|Zmwf97lG-`84D-JO5&=h@T
zlhqx0&vn9P6Q(Q0$H@RGfebu~+Zi=hC5>IK7zDz3?4<~6c@gl7+VreUMx_*sQe4DN
zAqAN<S~c;rMz|Tp_uXxjele=Yf&Q@x=U2BIC+SZ%ZM<Mu+O(sWMV5Vt2>HQoYH7U4
z$YNbk(fWyh%6@&THZk&}4gXBcZde4?Rs|kefKagGy)%9-xVTYquYT7`tli=|=sJ+z
z6NQ+>YZ&A+dp~W3?s#vrH5&IL)Bn}|7Dr_q=XJ<)<6h0{wsC#PQ2swMWbsdn;(_2|
zLU!D6B*m=LcrcD(tJPR`!vK(+>(E==fZ~B~AJDesGZxKEH}>uVO+(Udsql4rsanW0
zlUzGB+%{sq!b-Rhz1+i~B;?CcuiN(b>Cat<@}aeX{l22jpR3en2-mSx8s*%h56jiO
zRPdQ&F>$n7kp;!gjPyYm?)k%%e&!;k#}jqN=LUy_srdQ0y81uB4Fu;2{@s$pL?}Av
zbkT~6yTTyr8ad7(w1|WhG<HjMU5k3t_S-4N5YzlWRDI<hBep7H$u?G$y{q;RoZ5BQ
z%r(!<oVjv$>Z#z{L_nQbEKTDH%zqA{p1=1?$a)Zwp}1evUKzid_m98ZA;2d43RQX~
z`=?qcdr#$I4K4+sKA&5eHV)pKx2JMA3dcYRSJ}`5L`z7_-(^6<tWVRm?5~X*6FELn
z|N4dQZ&_#x!Ev^Wui|M02yrvcbcEx&kpiQ;uQn00^qV@Z)Hru_Zp{j^x*}D0^y%)&
z<$xcRzJCes+ZMyWOQLtueA!mfdgT&zI;oseY-QQyOuo)suKZ3l`nS7wE-Ec4rFJ_~
z&K5VaTX0O7sT@`FxBc+MAt7v)Hi%7`Bo-aJB_h`s&YsZViFAE#{NA0R5Q%c7ZoA?4
zGds_o`Qx#LS_o=ne@DaDb?fn3tgSVRx~;B4GzlYH<Ha-{?uzH=ch;et9G|#>BG9gn
z_AxjQ7V|e1`kQ_K7!bX=e%GAgefKZw+Tx{yscbTj_NF8pu8C$DHn`@M_}1SjvS$Ku
zT;wP!u)I7>($GUl{5ct6QJ9+djUh5kD;7&v8I5bLYROnjmx<)UIw!gCX^wP8sj9B-
z%7>zR7@QFlBS;s)(<5Bukv*;FY9YE+ZK}wCOHf1`XYfS!`)rXy0?~JN9P=0J%hbK$
zMpo-9yV&hqjv1F|#3Yy@f{`^dThI@;R-1{~M482taF4UN!NQGe6o3RAQ3nNKrol;t
z!p5ecNp!~WA8g@|mA@r~lr*NU#>!6&balOV5m*0s*eQI_h?aC~u9q?6Zl^IR#rg8P
zTKynWn715)p)+LuJ|1i9ES0b?L(5eQb>J*7w+=>g8P*Eq%_hU6>vkGWCbCYocDI9m
zUi@<PxQwahLjCEV#kH0F{-3+`^pOK~f*ajjckN<)ZmgEmc7d5VvS?MA+&1NJ+O>h7
zV#*&11>w}_-H0s47$L&)U6I-vx~a%nfSR9Ff0o~YBcP+r&@~dMscgGs4#8Sr25_hO
zB)>kAnGDwm6~=W1Is7eX$6Fb<{<Ur&djB~9voXd=wc1QXzC@ry1BZ3jj-ACfl)4KW
z4o2tqU2)s~Wq^}++Q@Y|o~741{VR!o^{XN&@y(!XPoZvKU7g>ReplX2CSk#hdmp{0
zr6#YFOOi`nafJKEmRX<+x&TpG>E5BD?}R4JF_}p|JH>evdlCdwheD3J=jXa@hXt#&
z+DfG5Y04n<hOTdv{|pN@)uT9bIYXTrCjUJahr*;hFox^kj?{bRns$C0M^;i5-ORpb
z{Qg_d1}K2FLkPgYk18TNA+qgLptoS;OZAO6Oq)BwM^8pIV;8JXKkU=4?@%4R+=b7=
zbeyXM$P!SR4yG>VEDq@xehhE<_e*{qh50)e2xYIaF6wiCQuGwKs1uzSM++C^#ALPP
zt8;sqCG2-HG=%C`r1QkiFbGE$N==Sa%?Z9My_s=WLJDkgU?_|H)pEJ*M0v4d&<AL@
zja4%k?|<T<O@E(t=!H`xY9HV7Y^_mm`Soj=ZsWQyZ7a?{?8x*Lzww&;Bgt02qs4VX
zrR8~`<IbY}a8nR8)YRyw14=p6?^4zu_xZ!-88nt9d(+YXVEI0uFF>r=$FbOsbX7+~
zXI#C=%Z;_zfwW!Dxg=j;il=0w3ZsSdWK_Y!gnvrq&nF|teX0zKC)%rGf7Hfjw1c#R
z1wA>j*9^wB8&*5KK;7*Nm>&#Vci=52G1_^)7f{C+VMAv>uGq(|Itxp!5A3!UN=EIb
z-=k$<K7Vnwc*jr{Cd}(*A}xjjzpUEc@<{WBMxF`~rMDV9t+fEYUTr8C*}v`go)hf<
zOLvf(TxV=B|3>DVgG)bi4(&&o-bzTp?UBbXT&Ba76~q6@YYjU?IxAW>``#9nJ6BCq
zQQY*Y8cSA4!hxshyH`!<x#YUTjy|;KF67V+m;SN5&6j%eHC=tv?}nP6^&Z8|W_Sz(
zXva@KXD(Kz&BfL$e<E-EY6%I?lJ0T!#T!E;7nz~u@f@#F@O{BKdCggmq%ic^jH<{L
zcz#ziO|h9!A23Mqbx~|ra$EEfeTV5y;Or5sEPdSig+GQ;&Q8Y5j?n#G@OnQY`HN8b
ztp=}Uie&qDJ{I~<OmWl`D(DZ99jz?tCj_b#R}%V5_x!Q}+pQN-WNfm-V^hMn#^8O?
zPxk{IBV6bAVpf98QlRZ1)2(8=aa#;eo_t@ow|cv<1Vuc{QDKF`CI<7u)t`SzPeh=3
zn;^l6wrjYQr(Q_kUYn`r0t4K`8FYE#&{ya^%D_W<Mr$?udmR98@}EtEw>Fcb9ab<N
zGc^IC*3*9XSIME#O-z&|cj&9b&h$sWCo%?Vh~vLie};E!1?ctKB(-5N-jev%81CD{
zn{3}?oJ$B2iAMMGj7{fOuAV({l-N0&-dGnL7EZcLt^mquybPK_Ti9Q7;1fr6dkShY
z3F+Lhzk17Q78th^Oe>+0K*OMkLS?X5yLXq;H~tTm`ebLn<=U>{b8a8kW6B7)=O#nM
zw?*17rz@><9DVmv@3$=de~WZ6rhUdBz^zP556Al6Zz=HN_&dw08u}w{m{+u4Z~aw7
zc!H0Vf=0M0vwBL2pO4MU!?SDeX|IF#9KJOwY}9+S&PkK^CoY=0w0N}!Ncl9yj?ISY
z_>e9#zddA!ttA0ysX&tl$Q{WU`7u3<rwMlc_c|)hNCTU+6!e^hw9AN?pe64!k|Lma
zxtWe0K1ZTXgMkZdF5T|#S@)9!Oq7?i+tndMrT;;nf*|b`V1PXjak7%pry92Wa;zt~
zTyg00-{tpD+z+=Z<X=$)YS4B22lZ;%YedMKR%yv?IQkY&=KI-oPf89(vU=Md2+I`U
zIB7+;Pu1vw=6o)9{e9URRw4l0Yu0GR&nXvi9&}-Qvn=*0t;0>{p>6Vxajh2Wc(<K+
z-s7cGF48_|`}(|oou0GzI|~y!xm&qo8ooY!W|OK`6cLkGqkYR)Zug!uUJATM@r2X@
zZri$<j)|MT@9i*oiFJhgnM!y@vMby!@R6q~Q#Qw4Y*%PB?-D?`Nxbo7XxJ_l_U3BM
z-=G|ftkr5(kR}muH~?=T)xg;rUvy-2rHI>AHL6>y0J^oO^wT9fMz^o+=59lP^OLs~
zV}7bt!`}GqQFyj`ui@oar%kx3>z!-Ql5cMYdkmFtDje_ipCbQ<3Mhd&XV1N+0a7WL
z{hZv&UZmD51N^Z4tmHWDR6<f~MfxbZHPR%t2%cdewGY_nwNggIR>r^?!Gl@b1HgTO
z1xS~PLbVR&oFNL(uNa6`%tf*oya&HB|4bxQAYmw{=JmW2n=C4Lf*Q7NhIgE9z4oMg
z`eCNt?4hX#8~SeBuGLdj{eFJ;+vkzm50FB)BnMSE{O+=7K!EhBz8u0IBxT#5Y|a=(
zRElt~(ebtC@sHSjB|Bo<g~aS_T?ODu?LJ|=KKdrpC7Y@XV(06e)hPy3=M#Rp6ck;F
z^7bh<>ZsCWYnv(hM+xC8Tl(g~>YH2nve&jZHB+OZBbY!Hw**rI?T4nHBGhAYi#n}y
z?8=D&Q@R|rS7uWFA>{n|bAQL(Dfjr-yCt%>f7Wj5uU>}(P#*Ti82_1Hz~pQ6W&Vn_
z^_BqvaKJ(b5Bk6{BC^|X5n5=EKsVXDdr#S#N4$Shuy^9$*WXr@Na&J}+}Y#?a0mRD
zd>4C!3+B(Hw(sx6n^0@mTd%pfchSXuH{RkhBadi|j^)vf$2#Y}63P~<_-PGPYZ`0~
zzg`I4)n5#M5*h%+a(=D~-63>QhKK6CgZGrn0tjoX7asYe^j?>Sn&h@O3bKp*O#`rN
zLjuI-hxF{Lw`$Dasbf%8dDMHDLz;ZOmy$)s*U_?PV@I3KEhn8y2n-Dhr|%lfrxW~#
zZrZvSq}k_y)#?*ryygWAcgq31Y|+8SmvyO}d$JLnj)*0H;!5u8qn;$km$!fUUJ$So
z6WT|d|EJ%Mw<n8!Uj~))=oVCqZXmPk_8lX0k0nZjb=JYWQwm*^ANxV{dnQMvS+dO1
ze4O)WV#};7`~7$sQDD>HkgLeZEwwp=v@DYOpG)MG$xv;Py}qG-u92;>6xX%Vps&~M
zA=*Nb(m19_BP8SJ!grot>jNDGG2!D>SIQZTjxT@WE^6bROrvesTN?3GX2iu}CqR>>
zmZZLT%bIHvPWtXlB3|Al8H++Y7zs8?CftQzCtY0-_TgfJKcZFUh1LUh6R+Pld2>eM
zII)98F}YD5kv>z1Ty&8~ATVc=W1pvLRw62Fy3D@5kHhsYKcovgmF{sNubzomCjVUv
zUhS^25~i#xi`cttO&7V5cBB471;-$qT^J*fICL47oaxW>x$WQjxkYJi!*4_{G)f|V
z01PPyfthxHyp5ZYsd2RYnU$HB*7<xqG<CSPE9of!&$w#2Hs$E^HJt+Dzp#nv+|MKv
zPh)|p4*!fW@9mEt#W=M;BdwR_B_{%+tkQ3V??No*uEfU-_8VJALU(<SFTO8jTw;-X
z{bjb5`;bS99kTm0Y5(u2(_!DJc}@}TfOX;L5#re%dD`5CJfWQ<Ly$}zg=Xz`j?k7b
zY?4Ma)L+ZvsB<W#eVx(3R>k9ZKp(lQEQKw9_dcYl_9QjMH4)sNmun{s^F5|0$7Ze-
zdEo8fE_4mR^TlE9-h)2u7HWDnB!61j+dA|pO6-hi=iRm%EITj*t?dQc4V^votJ|$~
z2C3&-IHQ)JrP4h)U%(ndiVnaeelK2tTOFw4kkST~8zmq+uf4&}Dg2N1t^Q+<=;@yK
z0sjEtvh!lXQn@<i?jYNZvlK!zPm0TL4PK*CmJlrJB^m!1(#4aRzte+)7Z}Ph?M7M4
z$Xq`s??V;&;;v0&uMApL&a<T4QzIyR_(rEG+HwaR#V$T`kGVQ)RSQ5CGxf@$PmjBm
zdzyc-)_*@w3j>EEkYrQ&i#Xb6??e<*lfHdhU-Y-?2hr4@s=~Q+$nVg#8+u18z1@tL
zzq0`i+1}gyKdj##wtM=V_5z|vlIt|_`R3Nl4;drX1uQE+5?CR`0a>C|DOdwQVF82$
z0Eulp70|`tw$n&3mXz<G_>Z<>XtDu9Ul=Sv#DT!Usdd*}UKc_I@o(#|fFU=bh(4S}
z@U}_@#8oL=Xy3E?7)CT6(A&iCcn4KP#ZNY!gujuHG+DZTv*w$mzEfnp#*Y-)SiZj^
zreNsORQ7L2YZ6-VMGk3#(8@qJ)RzSuOxK)zI;aGj7^*1&phf@4h{<No_qA!sc9N8Z
zFQCv=4dsVi?OeoRKX_ESr}i;cSRIid{8Y8U^f~QOhB3dj#MAcN_J!P6?A`#0ecf@k
z|I66#=hlVHS%4@nGTR0PPkH~<c}L-k+Trsz=Q26&mE*j{X*+niVo`<16jaa37mJ~>
zS5FO3xM#m`*Sp*i$NuzuPQ#$eK8@X7C~0AM@g?<D7(z-i9SInJrru3QR4)eAnFHn5
z*!}g?laU{{7{4rIP|rs?l&k%+AE?QccH5qRU9SQb`1c_n<v7$a6%vf=A7x{P&%ISi
zvY9zE2aH$PqY3_zlab+Cf-I4d(CNq+>(RPE;hK}rDeZY*q=SN<r#8PDHRG6mHL?)-
z6__mCLVA`Gv!9r{XE$OqRRSSgOZ&`!epnFp6D{E_0d7=3bmlC`*ntD&yw348D_Z~<
z8xlyD%}%!ka$6OXN@$psd(k!DpNG_zcnL?S0AyKyyR}TUg1~_i{q`26-B{9!y{RQ)
zbFv;I-+Ec4(n>hH)1~5oK8*S~Wb?~h8n@_u=vBw1Y}(5arao8viT9Y9--g(|Ve<qF
zC%a+;KF=INlE42UbDJ~m_TBr<k%M#R12}>v$72=4>B|$^@T#%el?b&u$w?0a{9{;R
zLW);x(W|-FSoWu0vF1HXIQytox}%Hpj)MS6_7*q$P|0fH@BE@e{40}a?oe>-sWJgc
z-{nUMa}feMo_t}d2yO-Ipj+Iz>88$V{}E(l62SF|he-DTChn&ZcSmKSLY>22AWMX0
z+vl=Y@!_Z+aNAs4B8uv=0k0%8z3yCK?j!w_6)v#tdsV9XvyMrjTwRgNL~CB$S)uN7
z-26l5^J?1jkb1ZWiFclSw28i8wDsea45+r&>JN3F0{P&t|Bax9)nuB8Rv8m+AlEVg
z$Nd}K%M{}1J-(_9Sa!;0C0~5IuGK$hpdQP|<fYb0>YPy&l9JE>R({$KLU(Pn&%TWJ
zt?S{s0o2g-CrU6E#n(*|h1|6}yv&y%3ZrpyQ(niWDgazyG0G)T#d6=GV`E%nt41ap
zq^`ei1@!NWJPx;A&%B1Rc{twjgPX?FyYvRP2_E=J3NJ)eTgCY%KQRxxV)~I>|IN62
zqW2H-JHBL&)vop0V7{Nwdtp;?O`B|^#Ns@n6Rw)t4BW*?cllI&Yte~XwG;s(ZGsDE
z3F?CSP5??#K_B++9aK}0{RM;sJmal*)8+$ldTb@9bWazu`GE1*S8A>`UJbKmCm0<~
z9mnISKL{kDR=Y2Elm~tIwyeacMHezL1G7SR0SO4R(t=bOC(_%sxn2UPO3%k4xj4g0
zABmgo=jY{{E{c`F6FPg}aPQRye$b%l4a%^AH7yWM0f10blzb<%gRMfl%VY3-qI%U*
z%mOPl+bD7Db)v1rLBqa7j}M<HQU@r(0dCz_`tL+)yx;oymdE&3Lk<kiJBV2H86=M&
z#jImTWo*2;>cYw>sSk~-M;*U<DaJ5Oxx#2c{{BERAoz*-cj|a0h2_oot6(m;eT&oY
zVe%CW>bt5y+WLxQRalZ`S-?#$N8PuVz`i%g7eFg}N;|UVYJ|-zi0RaI4Y-TY=SX8%
z;us9onOJZ8!KvjFw%%hv(&dzEJz*i<tM3VIXlcx;gQ;vrTm}rjGrKv&6?cc7GyD|A
zY4B+%0_+;%Uj6<^Q?K0^p>!;c$n@`ZW<r{<{AZ+4lQo?Wm771~98ZOZ$1Asx0~Z*W
zmNCxkBsw3&u^AO8s>tnxQtQ7cUHQEl<d?J|;wD)8=AWmYMpFLU5S(v+eAE6uKIOx1
ztS$h^g+Q=@#gpnoJ~9e^y=7%x`QAHOM8=+zHDou#NL|rDT~X)Rc`d<r&viXb$S0a8
z8O(t5;B9t5cl68H4I*xW7NZf8cy4jW7ZCBB^H?^KG=HWF1H+c-h5359W0=nZ8)N$&
zgQ=T@013V=T$y;XsWb1=3KtYemfeY=Ql%|87RwBPzMhe)wK&~#**0eZ^UaMb|3LJO
zaLz?m-<vbi2sP5fb+R7VP%Nq`V#rcp<%7}8e%lRTM%s;#q(Hw5&)KosJjjp0DGF3i
zX091daEHlia{<0@IbIyZeUY84cIZPHS>R2kcT+0Rc}I@E38e{DY4gWkp;Ojvy3kK9
zFmW`J6Og`ayahc;uzi~Jn$oFG>SI8l3`sqTr}>XbV*??nOWi3Hz+eAfXJWg?QA+w~
z-YO*)g$peDa9_@Wx7Z%-ng8$>bqq|>#AZ_ni6NBm7|tAL0jyVNQ<koebVB|B%lkUi
zQS|B?kh3|LWBANf)HCh)teow8?CG?35>&el?)R;cgF)Fh2DIY8USpF6Bcawk<(K6i
zCN;qgL~}iJZ5EbmRdcpw5x1z}lPD0po`tF(yO|^eQeA(=<9uEIfPrxX@hQ#Cu%at8
zbyGEOZ7*45V^KuhtQd+#Q~3WZ0H!KmAdBPUR%{D$ZL@{1fh!|l5*rp#-=@Kc0y!sy
z+GbvBjt(@fxXi0;Ne=+)L6e^577N?Ssw8Q%BFGyDJ0l0O?HV$PC4mE|YHMekjpRCD
z2KcZSRb_KenN8u#{;O_Q09r5>-^T(xKvW|=>IYEMkPiR`sx268fsW6tch*~etIl>h
zUh+k%952T45m$wIpxpU{Rk(J^gv!gJW3SsRBlOyiE!KGbx|UQ|c=+K;L#iR~kYcxV
zQm9nkdv7M`d8*nFWV6qy@Q$SKJ<m6)xu41XEe(~%^i!#ict9es&gPMUPNuqK`CAiC
zyJ|w#0N4N<_>2AolS1lZZJ?Rxm3vU72yp_GL$`nf05c+8h=SF0@CydqSGC_KKkl;B
zo!hE7CZ7=8{5tKEsm|=^eUw77g76Crz3(ht4bl*z)syZqu-4^{)jd2YdV36;I8;*=
zDOb<O>RY-OQ%06j^F69ItX_G9BN<QXUmWhqB@ckuh=<Z#b^^7^MG)U6Ix|><#P?~j
zkFm`A)0-Q`;qptss|cct6n6gZW#T$k=cf6aEWTk)vFbQZHQNsrrHR{?VK@vkE2pOM
zf+m5A3q^W4J2!_=#=Cpdp}V$~b`GIN2inOH9Bd9zU}PdGDev&(Y}eKSmTf-4m(1I#
z@2k2>#&|A1ctHF`mfSiP6PuWziK8?k+C)o8m7|O_acMGwr_8(8<~9@>P`*4%H7EHC
zkFAbbWkm5)3_6SrFvxOA8sT){RC&oQ;i@N_Hgt}9z1}M<8P`Q<ag*0+3Acfv{kut{
zp|o-T<i(8X&gK7i?e}%`Ax-vH2ytZO7hT0{mu_$Yq6YC+N}D3v-Pk131;AtYb0n@B
zxrE?%p@TT*4Ft~WuhG;c7x_hfoW~qCo!2<OQ`P*5r-$)-uf!eJ*{mxqHVcGnd&-FN
z%OSm@?S|D!14(mg0#1d6&VfKt+R*1g^M>pSmqf#-j#Ys^D;n;LEK|>i!ahoDQE#aw
z^dVpOw3o#^tmhg2Iq!lu<lAIq;R=TiArMmpq6kCYHp>`Nil0dK&T!G&XR4A&#*}T?
zJ}E@2`>joqgL(`$0e~r7YgpCj!~i)kPMeMW10Rc~4jACILjX7=J?mQ$`(&6y|F7OJ
z;WsAN2@qoYCM<q2MqNPZ+7hY#usP!}xYzs6`hrrG2PT*@4o_+NhPYh+XHEmbpNF3W
zfA{Jvrugw=dRoP?iI@;S3S7WIc%0vGv(Se56DJHr0>~Sa#XK`>#%|08P%uD<L;VTT
zk+Tj`<;HVfH=g)Xz&bYFT4@Znr;=o{)B<Xr8RBjlmcIF?6DE)=5uBKx<nME;B=)V_
z$0Yxt7kur|eb)UqJ?jTFG3*aKG0C?E^LW%imMP)3A8LHraRy*le|5leYT}2wC4S#s
zsmk;Ocp3ml8}?HSEC|iRMhG$7O$8&0aEP>8--Wku%J_8uXWw7bg#ANy8}`rv`H<y9
z+p$xeisQHdUl_LyyYgdcZiD^Aatu#GnDi-;c8WZVIk|I5YE`=xb6IA14Mx(~G`uu1
zE^*8xIPVe=0l%r4OrV%RkQt)2GD!<leVNv^1>Tl!E<VFja6cyJr}6%Tg_mZBvZPHl
zb()q5^qjbQa+1Mv|7InM<(}7!1%O*qGj_hf%5<A|yQ0~W>aEXa1!k2%neG@hFi{p5
zQUp02$~L;|%nX1ysRUzRxr<rC9uPY7#_@7!+nClvkS_pymN7{?mP8g42oAJUy?^p!
zcWGASwLx!d?Db%N173($JaJ4L*?&@^5a6P2eFn<I`_wv<o)>KJxBETY<>E3s%cJyi
z?M1WmpvNYo{d=vdDJ2H`pLO^Yu-iik((!<bVXL4bbU{~_dtC0uPrzW;=`DwL^}_=T
zRPYlS+%DbYV5|n@lW~GW7TTPoS0Z?__ya|g*qj{n|IUmBJ%%&NB65Nl{g6H#cha+s
zHU}gn$biSCN%s1^c`eX=lk@FJ13ge4)ybWka^i`bYY2d_I}n3xa?_zef^XPM=@!Je
z#a5rj&}SwdA6pA@z^-0XV=;eMhz=-%R5)H{^}(pw{q@VyKpJ-;TUWH6#}Wjp(E)P7
zKQVqR<z3p1*=B|x7|AW?8c3D%tRZ5X#>dh8r$*2LY3nAKWYo+x=0#Q(X#XZGAwqjE
zpi!=f4aw1{K9=Lnsqu0casyGrzj6umVN>33)4MdYrSsD?v-3P?1m7;9l&psKJ-{F5
zP{h8~u&oxY!|Sw<jJggKtZR3v*WzZY($o1X$N9$xts-#@L~+6PE$HzrBPvPBUW(lG
znMzl%qKNPYCOO4wH*K}_wl}cfz!bz;vC@S_QS#*{-#j3FBc`Ym{h|)R;#(Z+c5#@>
z_9dv#cL`kbVFPrmbn`UpI`y4|&6ggyw@Xe1`Kl;D&=08XGVgA6@^@B_bGf()HtQ-L
z1n8ExOmpGLzpS;>h1uhH!reyp$|<ApfSkfNUJLyHZpMe^2ZAlUq&Ywx3__hIK8Abl
zjh8_=h6Qx{=P7m2c!rkC)WH%hcniVb?_KMqnJ;JS3x`jY|Lkr3)JZa46tD#VEkXSt
zSW%lo-1@tF>(vY_1_qly9-DY){u_!VvH1!zKYS7nP;CiW3<f{N#W&Y%9p(n6afY~N
z$9T@63UsdlM4TxK&{zZ>AUSpWWI7`51FQQH;~LOTI;_&9?9FE(8xkAo?LVd;@%kA6
z29sJ`9S&HHq~oWlAy{5ls^Omu5`i3w2JpPHSDBmn%MPW&{A~!Cq{TQh$x`jBNJ!2o
zEUwghf{#V84b?v5Q36Fa!aVTcf>w8K7~#E1Dgq&f#gcwY(P<@pxH07GUOl-JVWH)1
zk(_TO41|BR?EhOx+y5*i^AIe4K*ljv$C3Rv=yD!c^Bo|5_wq_M+(5|Eu$a2Nou~Mf
zx_133oO}Ff@Ux-H+C|~j+9T%o*bcI9!s;_hK(gSBqJfo2u$mIGjJnwS4K*^f-pwRb
zEpUe620xJH19x`tT1)IyLyv8q+++lGk$JFWHByRPqX;Gm0Rv><-5)g4qk;KT<JJ-~
z>1{%@cdJeI!4r4;%<hy?)Nivw<gFJo=2^gc1Q!ltOGg%v4JW?eN4^S-DMIfU7Si~l
zi_kGy2Cv*GL+<R7=UgrHU5uxk%;Mt>ouTD8fbYw;UyKgu1XB@!+O2y?H)6pv@!N1l
z6bR_4M_N6~*QmU{y}U`p686|dk5=)+i}8*(_$6Q0Eycmcs>`=Gq`o#|#dJ)Ria`O}
zGo#;#2&fSMP$y4d{PYe3Vm7dHLOUmwT}E${gX)!g7v)DCtkY*>v$;E4749bujUU)w
z@w-}q1B-JWOg#d&%7|XBb2eWs&&Ps!hM_HE9Rb6O+0V_2Yx>N!H4SG)8F{%GSb=Ye
zNeJL|Nj0UAWax1x(~UFJkjP%0iI9=ec_gI!5$NB-xtqN~^!(V0XAPfQ))IAewB*vx
zcm8q$e6L62D8;^ohq<<emc?f+bNpvtX-#RZUDt7F>%Qpi+P>fT?d`n<#og+as}eJg
zK6mW<_p?>%Rv4*aQ-U|Cu{~27TG>lMd4ml4_OzO7337e2)}zMOYVatmR`L<(Fc7>7
z54lA&>oWV!aPeOVUkY^{>7x~7&7Y42t9vHJwyZGAv4S+6sZ+LnMDvnHCu%Qv&yvZK
zQ=AX^SACz4FI$Xry?1}^I9Yp{#jhP)#N;tVu4S$^ZyFgB;U5)kow=(cDZ8n&#V~v9
zr{EYmN$Q|r!3UKcKDTy=Kc)w`LD{2HU&zkycPR?_Zy`@x(*a@1#j+JN{X{qTVaX}K
zo#E8K=L@feZaxoyG7tPki}@mfPsn3<*uq6^Q*%R{9{K<|C9C}uigop2R=6P&VQs!9
zW1RDOP|SIF>kNI2<0V2&<~7_OCQK!7pZcl`f$h!|&G+k4DU~8kE9!99l!j3FZ4eXu
zQ@2xjmMQ~~jmBRsj~MfM@Inm%W*HY-tPOm&G}4ue(k@(~JJLHV>yOOSxS%No+(#%l
zSov=YDpV6$ot7c1xbL*Jz#@6zL3~%Z9`#lvRbWDYMzYD8aD@PyTQ?>dRg-r@VJkEM
zS9(&96Hq*HbZpmOJlj&@ZM)hUSfV=)&y#TL>Ng?`?=<SzpO*C~#5)&JUrBk^cc;!@
zd^5?-kYEuN2lfi+X0ee^Skcq9GEXft$;{B!i^!<eaHWCQWUH*7ETOcgk$$K5na-!+
zz4O$f#UP@uYftM>=0S~)PvUyjuK}-beuJU);q1_Qajc-xwa5E$%Z@n63G<>MVcm~%
zGvre#7ac+PpryBTQ<TyhlJvfIbvR=ft`bO3IJAq|6jcJwG$2*qxaxQtdCH9P-YFk-
zMh@O89<EHo%St^1vFR(9(PU?7G40CZ?~pGWvponotJzbt*mjAb%fsCCt&FRBXb!%$
zXASbqD+g#%m4m>;<f9Jhth)iEm}L3UwKj|P!oz~U4P><uNli}B$3Q?qF$9ei`-8U|
zna{doDCUbj1CR}#ic2aV{Kqg{1k=s+8JU((M6GE9>9t&cKbd7DQ#C(d{#zxHdF;g1
zog~tpKa<7f1KodHk2a$p!zyr-2~NJ#P#1eWKR?QanV8RRKYWeA;^%qc?l@n3O&&Y)
z{F_K05O1tFEziwmDlj*76$Q@<X56u7)pfL8^96EIzb>fveS~X8c7JluR5ZJs?I|@F
z|AL3_0_jaF^0@9deW<voT>PNdOo!Q2L%t8IfW#KZq<jk{VGox+ZRRd*624`CDU%L@
zDmBhF)O@lTX!h^-d4w2-pf+)Gk2WoPUWQ>Y*uS3-`qZc258NT&&&{2oJwJk_YZLdk
zN=#fWCTs@ZPm`V(HIB6NtWDz+vm#J-?m7>Kwj(^E4&!t@+@3$&hmSw;YyW=2oE+vJ
zSVBHN4RN+^<}LHKTi7eT)?zu6`LAmDdSM-LS3e3R)Ts{^xVfEkl-Z~YXhqWWbG3z$
zIAUiumkq3yvXFmsTzh4_N!)(g*XY1JMz@88T&L_wvFE4S&?RqHxkM}YPFHJzkSX}*
zc9VmDWHr;OVf%8ELjGne=jk$X?}rPplWUIH0M?`^e1$Rqh{LQ8OET%P(q>bJ-uBOC
z*5?~rRFZ4ijiX!6zZL$ywB@Cz#zr7M<@Pn^h$I2|cKe!>g}aB!*wfg6@9MlIxUevh
zF--bzMGJ72&v?0q4kidiu(aizlK{9V#{!o?f}0fmEKPoj1M|ojhoNpAkNq&+-O{fc
z2?jFE&wdH>58~Fs5C^WSlcR#2ZsWdW=TRS%gEKMDd0Q(L_RAG!0*1HrI=1wL`)cY=
z4LPpAQfnrg8@C)zVe_Zdm&B0*KPf(>=Y8*4{IAOVHNGRPF|-7KOsM){i;<e5c1XBW
z$v#2Gf+m5a1p+&;T=-|ViDgM|9T=|HLox5-g4bgm(dCkNO0RjF9U=ojboyC@7zBI-
zNf1=R{lUZ#|07-24avvX?Aw@tbmR76^@ipTe&GGpM%8=rWQ_$Zuj{I<p@Kal>jTS3
zg9;=VnxIkLm>)ske)0ys$t1%U{#r8(66%||j@duA0Kc&t2X6~kgW7hX)oTj;Rzz&E
zq`+plYpozrrmY981w%qWOgrHS_>ywbGxX58H&ew`sWSs}oFO42^tl`#Jl%eRZmJGk
z=~=2`{*O_4X)0R!)=|A(s?Q^~12;22>y0cO<O+74I3$rS;2Vl~%KoX1JdZ%;s2``g
zqn8dtcAop$^9gLScQ4LJ*<8=j4!`*4<Ox@d**XbOD4pyr7QmmQQ?sXhVB?t%2#`rm
zL#ErKJtx~c|LOCQc?b9=fqao?;wW9z!O>IR6|uJX{gcDE5rRYh5Ek;TQsr*y=aUMk
zS_ffCm%Nm3D2JHWF(PVwIPLmDK#U6A`k!w83MJ+-bk{cp@D#!?s6K)8$)*;h=|St@
zFPp77&(mmSqZ#DTmrM4`NyXQ10WNz1JIS&E>^ZBtHA@OQpmE^sWiD}lCNjMZ-8-r)
zN|l#jE6Lmk9v$<k9)X5$x<x>>nZ9l`l*W<s!$<ip(=t}^8j;Fo?h3U$#t@SIL*NUu
zf};n*1Qx$CMH27p62?ITH5EY-ZJeXe{-lqE!9PZ=M+;osG(uBgx8YI;@WN4hx_z>B
z7)|&bfnsj;KIPW|*%0ur-9ITUYoY+dRN~uev-u|JG8C-E^ryf`vze|XH_@prqei5W
zG-dQGrfm9$<n)$e0+K-v)L-a+$hqiu^WxJENSUwww((S(jwiJrZ@CuiUqOURObHIm
z(x%6Q4yoylo0j#Fc}`3H*B3Awbv9!1_$K7>R4VYz6kMw6J~4-1E4^5=&zF<p%{V0Y
z-AXGlMONOi;pf~5-JZsiL}?$|F?d*n>C0<HhIlaP5~~#9rT?%SM?>8%A!~D~>h<!<
zX&burKSKPovXwHRBsogaa;JS4s@%OV*pYfRa*?Z%l@ZxGzYD6-va)c;*^RgTm<K-=
zzHHE72{N&(mF8{6R9Z{N7H&@AV!;hJ=YoikLhR$Mc?Ac*GKLWoEM(4|u-f3ut`333
zg_3bWi3}gj$=*ok5$;63h`6P|=%*7_d}4}5%_K47!;76q8NFF?GS^Q~a5fbDnGaOA
znS(HNX?k?#_-!dVDzKU~QE&EwdZwy`5;KnqGtVsT>NJ$yaSE_lVl1_E`xS@#)CRj?
z+0e50CMPS`X9aH_&yD%q)W)Dk#(hA^xFZ;B!%i=$>BeEHT1oRgKA|_sn=gWAPe#x!
zAMTYp2vArL_am9NBO$s_|GkB;e<I`lYbNB;|J?Xi;KrYSVJBo|b=|HOUHOAVhAZgN
zz9Mtn_>47M`%2;9cKw^wQA-F|Z?4<TA0USmDOaw2-S8h+)AYt&g<Mm3fju@E8o?NW
zjt?V-E%wd?Rzdy<eyh-c<W(-i@j#t@BkRMWqm(a;?VqZJxI*%I+I@7oS-1y~F;3CT
zSihQf>hcTzZj19Zs>jVdC<0yp?afVx^J*PDyZu#@or;lRKUBLktw5uILPvKoazhGZ
zra<2Qf{>1n_P40%NlpwH>c)p@^B4s06M`J(a04p|C{pHX8%xsXV_k!v+-(UYswS12
zx!D%=bYeY8S6TXhN~vAlaE_U_7rSS)I=6qCly?nl-X79%@u%ItOb+s|a6sz{q5A-&
zm)T+z4Hop&(>uq@Uts(X*ihSYytf{8Y;8R(QXtl&XWcGE>u7LDFcqUVz^kbKHHgF>
ztg)J8oC|!|&0fm&4aD(*k&=IMv`MgZQ7`SE8~#tT@gaypSUboy`)cK~B{BVcr|dY2
zk;}Nk%GBd$&4?{uLCXvYN%nN3J$I;Smw1hP5sQ8~^W4T{Yunv!&!aFl7GC3nYsMf5
zTlGdq8oS~eTEh9z2U66q8hq(iKlvJ~v2D3_w*jMB;{9+m{x`xkAWi;|e+0b?)n?vJ
z)?8s~6pREDV&u8ji$bay>#z47A*ufxPc-vrse=rC^5{fXc8g;Ivf+VNR==Y(l9REf
zv$uXVcea#5(PxdT1w$Q~WOD-na-y9&oGy@%xv{WNEx6FSfsYv`e~L_6h<o6`OsVLs
z*;tv`I*dS>Sw&36K`#v-_HDC;mL6Sl3^2eV*R&#j@R^u6$fA6#kam@{jbC@!(ESCE
z7zMU$+>2I}<eSxBy0h{3zxm>~eIsV?;_d*&9nZSgG7z|47NCY_4&6%zKuRbA_|qzC
zK6p4Wvb{MsgEw^dn-NIhW-qMpQdn>#Jt<|%TbZtL+*Bo5Trusl61`K=wnM>xz1}k)
zcogJJ*OMt1!J*c4B5>*WSL%Hc4nEp_S>mp_>Pr?qenqrdS7%(3m?<fc(ylnz!fF-*
z#mItm-mfhb`GjY)<+lu}9sBzm9WRnT>`$bSOXfGxk8#x<FL=V2%{g8@rtm+LSPgXU
zAv+4ye3g^G2G{Hv`|8}1<}UxdFHGeTo0C9ta~(a}(YUQ!+dQjG%+$71vbE^vC3s+a
z|4mb|rPQo`0_bVUPzJ_H38$JI%8*XNzO{u~k=l?qtdIjdaI(aD<H}Owa!K6@8=9U*
zFg{1At<Aih-r5&JB&WO#mnx_7o!EMOHw|f_1V0bYLFRBpm#nJExpI6IRvepiB382P
zDRrT7-#NXRG1InU=-VD-4cfF>BS%BWh&+ZCcEQrTM8*a>o*w;m7Xn!%C-*6|8M~V9
z)0i)4wh+4h+(f1ZdfJS{9~$i;$r804+)p#<Vq-P!U3_1jd<z`Br=7Dt-Cy~>AaywH
zJ?c!O(7ylU#LY2NEyUO6qI3ClmZ=61JNuefy;2#)gZ;>6XMG}Vc;T@Ky8LH!|7!;1
zF+RTe+DCUPpkw~CW;``nE@UucnlIbxI?Q~v))%nZ3XB{#o9T@7g~qPA=FY1^FPy9w
zGv}g*fdf&n{?rxZ*eZKyXhrVPp8ItosOAxY+*kMN_NGd<mdY&_G-y7k`i^h*R+Z6k
zZF||Jt-FpL5R^^%7>F=5H%(lLT@9r9^<iE#KO^{}_LTa|oC?<n?^G+kZp99~>l72a
zmUWURI2Vq<{`S)&NJ25mEvX+mW!$$7cd!fA=R=UyRX(#GH4!lOHQ0QI2HCoLUz~~*
zgqbu}%fq8jG(Qr(ZN|5)BV-J09|3EY*Wut!4AR_Kc`aNjcWXm#_+Jpn43be#*gs7t
zkiwS!xcbm{LHYeDO;hPa%31`F;`PA#wE9^yp9+fQ5b~$;LW}}R`Hh9!KW$2OhIjTQ
zY%0dmlKPJBXEI0&shzN*>2XOimmZj*FoMqqgq&dtzIoZwM&>o@F(;JofYlBc()p?p
zDp}(NDZKn02j}XFvKVM!7-<#6#pcUBvBxl7%P|QlRAq(ugh!PmAb4T{Pq$D-=8oWP
zZLzV--7+%;j?-+@`#Z;;#<__(4XZ%hAl*~>x{azUKM%*j+Ib5*W%M2iS91{8FPiPb
z-JMDF1CKmH{J_a6%7Wfz5=N;;a}ldPJ#tv$f4HR@t_Xj%0-gwNIVLd4ccz+(CY<~i
z+$6QwRBhLHGRZZ?x#W(tvf6}PdjOsSNvBLu!q1Z8C<Rxf>h3!11Ss1^L6x)>i!zMA
zl-dY715=CMhw(T#GuWH!NnW9=$%f8C57E#gPnwuLWNo5eck~Rv*_~tlgI?APU%2&7
z4aq*6MeJygKQaEcY#B%l@9<mRep)tN?;GIzow(8IH4YFjdAtZLFIWXRy5#NTXIkPe
z(6Vb@1_0sGJb8hq9zMwQs2F}5cw4@IHI-jki@g&;lbmhXELg~Y#3Z&RKcS!=nbZ?%
ztTsF29)l0d`ZJZ+`}?5n7E#40nl>G6{!}$L`955_4HeoP2rL_ZucRA$UgvDSkVW$I
zaAsD0Bi4ko4(;lLQYnqEYtLF<e-o+dCnl9c*ilAXm%xrHS35|ule4C?4#=lDFWIWU
zuz6c4QTNoUIP!+X?E&CotC(?UE-7<-d|wCfp7j$-EcgbZ5b;`TBenAQr^gwvt%iz&
zyie09w7k|=^w02-d-89Yx2Jf-FcG2^dT895Yi^WwqR85A&6w%=bFO}Rl;eW^ivUsG
z`V=X**|>$T(z+0gBhn=!)a6a6oidtEiNwD=E*BS8{Dn|J(E6tw?ex%O>nKLa5Kl+2
zV5!f*cGZ7yB=3<P-zz`7`)s9(udx7SZkaJ^sg|oUjYP#Ax86u{W1bJ4xC>FG!SUhF
z`?QI>qDhd4)RDe!YMfvoth3uow6<1U{-e-4E5`KM>>l%&%h3jzeX!hjD6@ci=kTK|
zh6N8|E%tGbTGt|Uy`hLk;i3zNuadGQ9W$`Kd2<m$<a)fAP9%XPN3@Wk=YbzK#oyAr
z$DaA1_!Rrnv6a6p))%o`HabA<Rrm<=vfF=qVfWhj9&TnXfk|3u(haAtI<_dw`eqOB
zr&TeKa@vsKf<FfqxXTN5h#}s0C(p-6cF0$}?J4J)3=UmEY)biL8Z3qwOJm^yLB{PN
zDyhT(-=duD#LP=|E@r$abJy*kACQl;LgwGHIw;w)k!tB@%%i!CUMff8sxEgLTc>GZ
z9K)$V>d6=SZpu+uqAlCA*8beKuVpdEz6j03QvR$7JYo>a<p#(OD4h!4iSnma(O_!D
zp{8=dPdS>XCF)S4G-iwx;Aj>{-JPu2?83{;lcFor7YAJF^KY8S*&y2l3iIaiPp(`Q
zgWqm<TAGO(JGf^wRDK@%+tUYqGqP1rq>uFf#lJY7&Vmk++OTSI(LVkn>aja;zp=)y
zdMEH^w=Zaaz9W18_YR`!YAtCBJ|#RQQcEe#bWn_Pe*tLiuDP5BC>=HHMMIsX(k(MS
zkw`d{IvnXd1nqvj2;8R5m_WM-Tr+tbS-0{mmvYy~OaM3hQz-|V1gLG3lE#{#{^adX
zV&?jS70OIEsSOc2{6|TUOiM=x2JF5|hwg~8eRdgOlizByalI-Zo@b>VVK{o=a5&z%
zh}L`QY2>3|@xjhrwZ4MC3<%P{nz)%JtpicB9hThAyufzrsQHl}8+N?*_jsiExX49Y
z4PBfy)qKY<m1XD?qf=t-t*b;;N!`tCAonAO=Nm&f9XZU)t$&W<&xYR8AS}Bo+QU8T
zSF2k^+naEOVAWyEQ@w!~6=$9-Z^UV^<Y`U;p5^hi<njM!0jf1D7mqu_uP337O&0Vx
z*#eiXgmd6}t!^KoK7{dt2D$h1ks6^ERw6m8&MZfJ61qr)xX8g&xfMK7(ync*&DLyn
zS$!&)jM8JXr8`S2PQOyuyzJo)G$$0eJodsM)=KoB0Yr2T)!t<bwJ0I&f|E6*Uc{}1
zKr6j}zW+1s`0P%(Apzu)^Wc-wPuQ5~3X9z^B)CQf5UmP-6#I3h`}cPX0(^KVztS}i
zehv|cHE8)NUvrIJ)n;sI<{f-|ClPOcUQwW4IMeE~3Od$%xzCD+Id|M0u^A|+(*d@y
zR-}Oiq?rikQ>5f0^(C!5_)l0ea*R_6vbjW*I#li4KDe+}Vbn;uSUj;%JYMC{lnwBS
z7@np2IXxO)z67gW9Is#SrDxyUWL{hCzN5Oxj$u<rsu@vFPNBK?gC7IW>8EQZeauU=
z0RG>US%)zYv#|q0M}x?~<F|Nh30FxWcrUy9jT4)U#NaLW{f*Tpuy<nxiTF^HB+I`w
z<@--b!4XI_ajK5`jPKQ%_*wv4ux|)(Z6T<3v5E$QBLG6t%0udGs$+rd_p%DAM&F*W
zqYW=j4$~~-Nd7{A;h$8BhxJL{<C@>IV)XXz=xymtuaBN?PnsO5;fA3Ee^h6d?ylDb
zH1X*`-QHYVKBCmv_R!7Q9@-8U^P%19&B6NoO*e863m)i_vqL-}79(%y|IAW|_0%X|
z*_dtTwa>bk0d-Yzkg}iwEkgkXP=B&?W6+iG9fcA(Q19L>PF_Kkf2Gf*SeswYMo-s9
zIi`I?`q{QwJsBWr=ZAir1rK>c|BgFUA7Zqij;ocIYz7j_#4K6nP8qVz*}JL@Ky=;Q
zK3SwQG5dPQe~9?<-tWK{sZlQZi=(vx84n-7^7Rv31nglWHb{PYcWo4G0>KKRQlt5G
zXlCo<0T-%S&>~O^Lb1V_P@oFg_*)<t>4Ss|n6@A7MV4#nei*yj-2KHtRw>6V02qGN
z2^-7Yy@i|_l>px#uY#Np-<#(rB_$Z88<)RM40~y;L5|-tM9veuUq}A3(L0$l;LvMg
z8}TCuK$_Q(R|(%+fe<duVAqqJm$z&!@Dwy!bmf+j)*Yd#*o@-i9{2VgR}WifwzL*w
z;r^R(E}JC&@YyfQGBQSAj3Dc@Z<v#fOz;o5r8|nS7CcdsWayM$&X2@vI8P=9x$hS=
zN*H@aUwe(N1k9>F`R*?WT)X4;#XA&BHZJ?Df(nw(m*z)W-B&Nh)<?eo+m86c?=Yx*
zHbSuECUJdk2~v6r3sZ;Z@jP<}f%9@r0xSSemYnRDMT;8}jl_zJ`c)eSAWzb30Na$C
zT-<_9Ce#7+UWh(3R$CQoZLe!jOUl9;JmhjY!KJIkKQF~yfK6AwZzs0xWW8WWtW1Ai
z!Isdd_=YXcmN%}Y(s0qRMg+pxZmBMyZ`C4RTraCApN*tmn6CMDJ6DbtBq5|mdQNi!
zY}BXQbQ6=D+NrYmHMqpuUW|haLaxKY0z-bySGv2h0)DRbshr^*f5exS>;22Ct^lL`
ztNB_!zNCf-D}EcVEhFRf&#{H*z&p_;7uIC^C;Tn=qvJb@{`TH5Ad~(6E=ShMm0QaB
z4!e*3g>+U?Y~aQ7>1)cbaO(I!MdGYdTwk9bJT?&3HcrjK+owr@K9RnhpTP)F?R3r2
z_dTEFK@l4|^pqk)s2JRQn%ISDb1BT*U@6|kKE>@n?p$&X2EJ{WpA?2~MgU^-_qV2V
zRo#e9xbvuOK_ZhD3LfIC6egZ6>-Oa-svBm$N2}5gnrNAP#P%fLDVur+T#w>uM)xmb
zRrP>L>U|u)dKrxwCpPvNup2&@BE%=MSNG|iiXVlT=hI{QsNU1uKx`tiq@e&^{dRE2
zW#gZ7)s=STB{1~$EP(Fm$k243<Q3C^+&}HIX;CCMhVqgp)gz_6ahl}Z7vD9wDpJIL
z(U7}U>Zo}?eU$jt_BDF_a*#rX_15tv(`_u7{ShYn6-?2~D#Z78vZA5_cB6g&;^G35
z-Qo#+lMdCmD+UGz`IY*ayc%vDNGV~J5L$689{X>db^R+DOeoNRIzL*M`Wf*k8NB|)
zDk050V9-a3Rt&~e@-Hn|B=%=ccAsGbel4oS!dlpyY5^>00puLNiL1|LKTnwib>4T*
zI!oswq}xR*vB7)n#V#SHjJa01jo1bZ??0KixvM*0R+$A@>pB*rfxyNI@wBl^fsCw_
zol`zF`LQaDT_HtWT2XWslxja>Jo?8*F1GypE|rZy#27=JYk7Ge0Ljs1x0&M?L(<L&
z2faof(H<|K7sUdYsKEkHMc;ojTQqKV*Zc<m?LORd`Gyp`wF$n2k*|aIVU>xkJ{MO?
zgBB?p88hqX9xVA6MlT!M9?@4v0qCWe=fM6T?<>jS>tnDmbdrL4LMrs`Js?V2i>OP`
zCj%Z88fr|_>}ph_JbQ8`$zEGkM?pqUom=#9ey(*V$9z)1slGU}4)K7M^`+>=9moDu
z!CvzzHid90_MN%7JRrIBe<nSq5f=QD(G^;2q40b?vSWIx=XWkUDM`0tUUb{D_29D(
zIsUyG&@o+(DSZ7`Y>(N0Th`(Jc4*%3BD5oF-8&-Zd8*@&uktLeg1T))b(n1BOGfLG
zl1mnU+n;`wuk878FYY6$UztJORTIEuY`C>kVs0s=n*O0{|1Kmy!vc)6T5$j6uc3bu
zz5gNm`omqr&Pe48kH<iz%c;<dV9(RuA}nPHqW{?WtYg^xL#zGie}st?(5-zK%5jlb
z^`U=_mGYX*>CZ~s>aY*?iAliG+Q%8j3|X4gp}BS0`s*#r;}731Ik>V%;F<Oh&5vZl
z3y_FhNft<x<ro>VHmZ|hAvDd{NR;U*7_9I@(4y&!$&V<)Y4-V;iG`Q?8ALZJDY6el
z65o)5l<6*IQ)|b13O;DgCCY_jjt)e;mkTM*viZO|VNivnwApt5=bI`l0eEa|$|q$k
z(YDK&iE<tnbO~&^G7th_2?3@eT{87&qvpmgefJ5La_ZzC^*3tH1caYsp9{?c47e)L
z0Nz-X8u$&8zB$w<nZ2BS0hmi-PZY0DUqt4Y_Sjvv2m;a!z1MSgs_+RhOY!*;kL{J1
z)mW5(X0o~RzLisFzh2)etOEgzv9TkHW^ud^zD}{wDh6(TRSs~j|655^hA#k?RWAU~
zrqF$kbYMEzM?uoh*0fV%3S*U~pJXU3fmm8Ps7h3UxTUwOPt!=oSNMH&pWFcx6xdbO
zY4k)z;r~+EwrSe2bs<*Od|#_C3Y9%z8V3n&R-5?@ScNT%Rqqexyyh^7UGq7|1ti@g
zcEnA$Z(A|)v#{0z2QJlV(mA!|d%LB%)MU7{C0;yn<?PJR2ZX#AUn#O6{B}(c5Y}$_
zJM1BU{joF5_kUKJqf%R5EIA4wgL%<Y&R6l&qPZw}XqMEoaa6GIc3vtEfrN}*)%m1*
z^RYJoC{$^_kWHj~t1)BC)SC{R*W1KfHdxn*IG)#V<vKd$)n5oGs4M)qfkQoB1~?%*
zM!D&XDi<7$*%Ca#4S@W-CVY{;`r~7^(@n|$@$?lAQH9OdN~eS<4bt663rcrOEK7HH
zNeM`IOE-u#ERBRn!;;cnD=xise%JSV-{1WMR=jtfXU?2+W+wDJ=C`S>Fqka7O;+gx
zMz9zGz2>wEGLq};#fnF{4wTM9D@V=0AfpuW@`K>a7^T0LHVukc8(<JuZRG~>{*IT6
z;HVkLEtZK(R{v9W9k}TdrBkOtpZa#pe)J|SM3t44TFg@|f3);Kqbd{;ni^VKct{Vz
zs2OK9S~Ar9Bs8b0uBqAn-jOoc%4T70)@AR66og>&e(*QG{424wGB<%p$h${?4}+e;
zQQ92JO7jma4!Cf!;TX@UjK`fl0zb|*zU8CfJd8tf{%)Em*!Q&L%tL;AP%!M;>2*#~
zk<R>l0G*fN`lN!2EzInJIh}T1<=24wGB$nkOlWFg84u@1C(}%_oLiW50SZ8@<Z7Gc
z3=YU%Ge=XoNn$|#Z>it(tv>YJ-i|&IKi#rDaic3KalSgH?7ZQ)3VC1uB)srMacKPD
z>Lqd6M-S@`Q^sQafFA<;F$nU%85SrPVaZIte%8+vL?qd2K6rA`4G%nXHWfcZSNPb;
z{~xfy{7EqSc&jKP1pnk(5!$cO-|BY2aWM6|i7~$=g=NmYf8Pk}xxY;VsMgwD0Cg*M
z36(RBRVTXaP6Okla39};%I!B1rM()T&-f=hVi)ncZiqaM3nxB)>zgd_+B9$VyC|YR
zl7I-wT|-f07oEjLou0}Gz|K<{gB?RMHCyBdsD>Z}--EYbQOUIq2RQeDwa7D;Ez0#p
zTuEt`UU*t_Sq6iLri>oh<?Hm{lJ#@bW2}R{KLv6tXzGk?=CHzWpL6umEM$_)Dk%xh
zI83m9(w8iJm3G=Wd7~-2&x3XW&6U7>j&eOaaBPu?4Paj^j!g22$LR(>cX<Yu)Pgic
zfY9lPY0oUmV@A`Ll$gnBy*_)F-~GWiPU1FGqBkL6e|=9A7BEz@|HaWd*(=M<Qq$d0
zhmYOEUDa-I6rBN3qx2712&0~jw3%C5k*1|1`tD`(wz89$dtX+f{s1?2L&n&=i3*?R
zYS&LG7<=CzW}0e?gmc6ts-z!J3#|J^g#?yOeQ#Ujt#~Ji5iTsanh0>EpY&d3h`Y`c
zV6VDmm;Zc~_mUh;rYk0XFk@RwxOeA#-|PJb++QE`(0y{C`x9YLO+%2}6knFlLbNrY
zil<P}dJ@$k?j-Q(mGElF!I%Na_&+WSjdK$~F+ZP_Pu&@f_pYM?&dr1J^E_=ukxa9%
z_9a9B0a!(JTeg*q@8XnLSYl#r#0LV&%H_8C()#){SZnyzsN!PJ;0!BTXnj$BMZr$V
zQ4N?!(lFVRwJ`;jSYohB&V|--?6S3Az_PdV_RDQ(0F7cOZDY8J3)&hE_jRpZMQTsB
zSH0xrj03dyVlOmmq*ZM*u!*}4;O)6F9jWk|ac%#GY#?X64v;L^+dlgMZZO?1i8f-?
zw-=;bT-NF<@lviVduBjBs3=;P!7?YB<?GnSTal^PKauxD({<b!bc-vF+3NAQJPZwp
zG_oB|-SOb+6PkF%0B|s@utfR20W*g;Q^|7eEQ&0hc4!^Ku(es?*{;2-t`p*Gd*jEK
zHl+$B7J$Mo?N*gr_qG~<UhBlQ6xv|2kATbI#?(to{J0eYruz8)jTXn!{tp@L&h+Gv
zl-1OitEtmIV2l#QH81SrEoLWEbsL~A&#1&==)JS*S^Md8n^~4-qX<8t0cz=K#ec$Z
zQACb+<l@@90D1eizWQF$ymgkv9CE6sjh4pyiJ;>yc35TpgVI>nE}ez1te_)piP1au
zdU2P)uTnqWXWjgu^ytL2|7Q{Tm$T2E08`PF)Gm;wqJC2_n07({PWMC2P^yJ3(w&l2
znQ3uK-6Lw>*Hj`vL#%4@XmotAS@*7a^+TDv0y*EubL=OYRVkfP*3lHaiIv5KV;+nE
z?o)Y7WFb5dnlyf=M_ox8^Bjmp6GdP!{gFq-b9Ps_nqu`}oOpJvnwMi3QUb|4X9m&z
zCL}1x0C=F+3@A;Z(r74Hwm8oRTN`52)BC1rN=<&G)hA(#`+pyAs@Mbs7fB^mPpvcQ
zN!3@1SPmx?P<EppJ0(6tf&Ojx&>Tcxtir471Eb;TPkH(`WRItd)LR{#4{^&*wE;KR
za+uwS2w~x?)Bzq=To6_&Q7qeFYco*lL}^1(-%xm_vvDN8&tvT*2CGo}&cIyFeUq+K
zg2MWG<-X3phWEuG;K9B?gVRUg()Fd)xBn6NMqgt7vx&DAsm0rb+OCW27f8%Uv}e7(
z38%p_yI)BM=f>)E_+8o3Fk=-5fBI-Loc1CX?0P|LZBtRy!26#}#!B+W+&!<6_+=Gv
zCT-?W0Lj}T4~jg~l2G~jqlwe#bJ2>q%c;c4&r|0wlhUsZNw8lJ%#6OO^i@qH+dMR+
zp4km;>8$xLBh?GIGX4Z-lTWOza0>$N;>BD8;GR-~w~1As{)6J=xnHy%V$Baz-9uvy
z#<G#J4!3g+JUAFfl9)n_Xcb-wG3k*-xCV~|_tAYR%V+I$MyvDG>~$dc=FkIDJ48li
z*e^jMHUX1yCZJ}f6Mf|KIgF|A5gmH_L%aT6<-QdejR{!9j`~k&7LzW6L1T0n4<Ms^
zB~Gg@&vGW(55Li^YI-@hBy6kUB9sT1&)%R`HETlnDd211v2&xu@@Ows($`rNhab4+
z>HpdYWn&P5+$<e5Bu|G^)}b*!%5VPc=5Oo|bIqJ;RZVuP`(r-2o%k(CTrJ;B+rPC%
zlJBjIc-y#?6v=4s4FX*frw8t3&Le#(K%mN~u^{P&Q(L6g@O(S2o`JW^$>ndI-Rhb<
zQCY>Khy`oxlIq6G-L_1)A6O<ol$wHttMFPVpqF0qO*<~8nPh5$zv&Ho*QPi8wcbE^
zdpD<fYXcAuD~{pMf23bL_x9M{%>{V_NNiOe8o+R29rRd31y2?jZMrg33dIjOoj%R<
zC4b@w%6}<=ST9aTclLc(y$bMb{wUk;I5DENU0V6&pRb?T<loNfCK^8O3W}*d^XHMS
z&J7y1{|_q!r1gIiIn`Huq-<bDh3)v`&xUS(2Q`lEks?$Gp90qm3LcgGOJLs<aN)v?
zXhq|&0ql4%j(g`GgKYOw{rmpy_{*?w98D?mG>M=ByF|^pso}-^-xes*+EduUfo0E!
zPvf^NxPaci^tr7c8v*bZQmuz<u%H$|6Iu<pdV<_LQvfrmpF1+Gy04lgGLs)Lr%F;Z
z?Hu^GbxF|Gs(1Z2^s<daVW86M9e-$5bBSW&e%V(lB*CsDC~EqE6b1nBAm`fR+=9UT
z?ZSqFG$731pbYd4(^|6T?+ij8jt`(XzTaGU2j3{9!aNRr0=-oMYJyeGk>f`dz?Vhi
z)d>m$UMcsNCp3wyACdyQdf;NI0tKEwCvqP?biw*1uF9aL>~iDkd@Mr=SVQ6{{-3oK
z?cfAH=agASVfZ^B&E@xpYi&qepAo@7!glqPc&bqadK^q$_`6-sl<Oh!AB;}AF1%8z
zaVR)(J7UVP*KR*2PpZ6yAMS8iV#)z(1y^<L1NBjRPsc^xx|}*@N274e5>Wo%FU{ph
zOF&E45qx5271X;sut(DAY{4ISKQnTU%H>%A+#Y1Qm`9$sZs_9mH7Vfw?)ZBT+vi>M
z?u<D8$EyFz@Ga5DJ4NVwOJjgy(#R)x?k@A49`tiR5R0a+Zc^!00uC>Ih`oCs#7mjg
z4}$v;M+)<G?cVxZVqyBOS{V>zpEnkps`r5bmk^L?la?~cG3vfiX3s<~`J&V%+j*sr
zlTlGK;$W+k(Ze}<mCzX!F=D$r_LsUzb15wVT2!F2O~{Vd%=jI9f*ChTLiJu%Q+%}N
zb3k8rKKlIm0;tjsf1tk1cWLGB>@;w-QlC^L0x1qyru20{@<PM<Ut;w1{&aKH(GpE~
zjoI5*>Iw+G?A3#Q33$=DZKz0he52DSRbzmWl=D3sP@|6w8x2@ry*}XL09<;x+Kq1;
zOR!1^<!QK~Dc`IT$A9{WRQw9>aj;E70!3~0r*x+zbZ~SXr9C0?Cbk*b^zeLX6?QMS
z%1cue&}4o-b~2s}B%#ru;IGV#C$_Nj3!1aSeQzd{+5VJ8Z<qo~KO=Yb!<4#o99}R!
zl5SaeWlHpFTfeVb5&fJWN_c44t~A%~!JCfC-hVZXbH;f^LUPbGp8Dg6+8W82FB_tF
zyYuWEk2im=Q>5lA=_o|UgID1YHI1U-ZZ{}gE$1e%>rf&<u=;HL@S6L_j64g*^H@9-
zBmjl@Y9sdTvGYWpRIZu2GuL2`{gm?^;3vZ5+Pjn?p%dQ!FCrMS0b&z+Q9~YZtS>wF
zJh;rkAlXs|HTuxvEk`>sKWwEMhJ>0O3aQ3-E$8m;^Qdc;T77#}KhXd&%jc%9EZeRM
zyQ-Ogd%97$rb5a@;}gWGojzmE+|a$yqYk2iZow^|m9x>L`g7wGQ7FDV%jt8;>#ZL`
z)0!v0>^_L5U8|q&9|~_z^Rg4FUIOi^q`7d)+M*~=5>p^|1e#mNUj&D}`~z7m>aNSP
zt{G31#P78B+iTshvxCd0(N6wi1fvpS!q8EVS8W=<WjUSnO}{nll2j*+@XV8;{3hvs
zZ}eRr(xZp>K|SJq5vmNWzH%0H12Cq%FStCBH?9+E@8xEFmonYJ!m1O4f^VV!k~oP)
z;iIq7Z4Wif<ifV=u`wuY^Za**ctZEL{Icd?bGe&#VAVhc1q@A?+;Mzbh(34(wP6`Y
z6hWiD^u7>fQS_^`!QE%fU}>MK&kFl>&smANsX8%y@PH)9T?07(>|kyD-xmyL9Uz*Z
z+WlxFvcb#Ff$+(d`~6k%bXq{wab!YSeLAX@oGZQW?;fpYEfa-P?{)kCPN=IQtg2qf
zCnsbeeT|uE=Y_S1zmPpoF+%fn&~h)deqZqv_j?L)RTsZkVGO<_lP$H>F-*ulJDA`A
z$r~*g7KE8>n+K(d?AKqesI8r;Z}U+XJ~nsvzU({77G>g>nDK^%`yTReCNh6uN(1iQ
zGf_uD0Mh@9lB4W@w|`{cG!H9BsB#KxZ<UjqaDumZ{@|h_Jz|Nc@?Ff*l)@{(2wUPc
zTeC<T#FTac0F;@L(?4&dC_KwBHus66jaY4&g4xUy-bFhA@UA+S8w5BtS=yYM+tH5+
zBwgAwP8JkhwV+3>5PCu^4*9MjJ}J3*bJCD<ue@p0Xl=81%B)J?le4>=LX}Ssnn(wF
z6*b(g7Oc!m$4^%pe`|oVX2C*w!2N#c8wc3^*s8oy@QO=|&+_CVP}w+CB+?)$CHvDa
z!oeRM(y16st8*tOkT~VrF>Mxc=9CB(4oz*50Q=4ont%vk(D;2K-`x`qI95Z!XCr~b
zKSct~ougczlba+0(;(e+3|P<`8GgRCSI)l`5x)v{-WU!1>zx$G)wy|d=kXXQp4*k+
zE?(E-5S<^DEXF$|Bsew%u_&Faf)E0-M)_UKk=<(690p2!&*RTY#aA30PX2hEhGl7%
zw(lx)Q)Lbp`SV6TxpDm2#mpq!QFL=Oa<|k({0tnnTl3J(KeDX7%i|>M-_P5b{#GPU
z4?jF_mN^7IQ%fk`_-CbUF_1z@BKYR7#OU>TOVZPTTFRk##M`&oi!9>a2jmM9r*Cl@
zzi4B;P?RfVfT5F?e-ebn5i1P%v~36;uuh0yn0LZ9Bu>w-iG80QnV44&pfOk{W4D9X
zrEZ~%KJl)T3Jjs?09Rkt@|LQVze3O1n<U_Xi`VbAV^|!)w{tf^_rqM+C-v{Ub!GT)
z?e4wehg(KxG^Yui45nA;*rT0ZDrn+~c7V&!KFmswzx(d4t!p>-%YWfx^*?4eSdVOC
z+9%$VP}0UtvT%46<qbb&kYdkI5ms`W^2B{c>l&{-?bRDvQyo8`9mkt!^x2z`!M6AN
z1>Y%cY2};f5p`3e7${!EDon0y8daYBqA93N1Q0C+rWrt&NT_z^yv486EU9XuKQ?LH
zkpfJ4zBeYrd$Y+3dX7w09PCl4sm(W^3}C`jDhlzcv)|5rxR|TAEhNeVJT{p=$El9g
zP3|Z*bbJNCf0QO*9ZoZG5;J-au%^EEQ?GwZW5#=D#|*FfhGmgxf0o9qddbXSV~dK^
zi8pR9h4Ok}E7VHe2Bf<7hySByfsKf%uASPVeRR4ah9rIo8E>`TF+F=M!^dRX&mX=o
z*a%AfVgx!CNbY}mCoc3mPOQ_Dp5akHG(tjXH(aSyFZs|i_bde4T%B|bvdj;^l>Rdg
z0o1Kvy9XP&k898rPAD5u2X=33&Wz%#KCpXP{*>Zc7jXsN^J64U#njKLjKZ$kv+78M
z9EliK1BGE*Ti+~l0B<Kt1)=s_BR3T}V4?>_iIM#*2OMNm(+yJ!YbLM$<)~adU^t$s
zwNvKr*TnC?^c?-<u0>*d7!-c2mk3P?2`#nct;Khb$RX#&Wm>{~lJIkW?QW2j@BOf5
zi|h9|z&Bq}SXtRom(m<}DB$%+2nS252165tSpJ0@3YK&tc~#<9Z2X-NW@dVfFAnGg
zm=w=7X$ecf8`2sfDGNuJljbVc;_I!jw?AGIAB&g6J>ocYy+3-_8do@p=l<yaacQ&;
zx(ir?J&DaGZ&0+9t-Gy8PcB6=oegwauKFD9N%UtJT_Xt^UHb+;Dm^{F8!x`{cXWQ8
zzZKMaYpv%9?01o;&3)-0GGSDh`8q3Dg8~m1*0j6FzkORCl=}BZLjI7J1`UfrnDeb1
za_Ewy>K%TjVEk{HFl+(^dZ6#RaP4i^rh~n&`_zfeYQ3>te7G>sj#C*gibx-{A8l0!
zlajm$&r-;9`<<V?7Z7Eo_Wz#=Ic2VsOMmt+BvFy)`S|ilv@iT4*@3}g$km8hPJFsp
z-u^c!*Z4`1cxf+TlE3=p8B+?Hjx?X^k@tl6G~=>oP!Vet*%~mraV&7nkldGfVmvQa
zo{8)kS0IG{HPopk3URIZ`Z}7V-|53eXvGPLy0yc`r`3xB_igFx9G$Eb$!$rKR2DGC
zm_Y=7NN8|jecmAaMFqX?;G*bjKN?v9!wrj>x++RQZB^=r3=KD#EDE84%%UKiXLCiC
zI?R{oS(e+r_EV=mDZGKDsyukr3Bc@(FpDg<0pvJ5DwgEQAR%sM8Xx<qH`3D4))Ety
zbfJjgL2_&)Clx+L27OUI!FUBzAqtq~NdvIl`pq*1=?~f~6ur&36u7vOpWof%JdLW|
z6B=KSclVTM)pl<`+?{Q>-=876-wSs2y3Ei@lHKL*Mzhg9Xs?8Mto^9VD8f<|q)eEn
z<ShBHIin&6qmxv8F&Z13evJV$7?jyv@QTmCLz@tqlu$&I)gb9#EK}kkX)C)=2j+XB
z(TCV3ZSm|0#Sey~SEl!K>hujp5~-67^Uud(tNcbIy9&*AU7eXhtkOyQJTz>c4!9W!
zE2c1daIy{xa@s+NVq-iUV+byl#;er9`6$qq^Z`^wx=d4>>NU%$a@Zw}BYiDs5a^aI
zYUZa(#SyC2e?54ALH+dV@nrsqH(&H72;Vs-_(T%nUG~KIgjk!<h@|mJND5I_U?r~?
zNUZprj>KSMqp5$`aJoX$H;TUNM}=r`UD1cIgA%}{6dtIP9}n<wCMM3Tk7VZ1sj5V%
zszqxO8p}0stPMuD<1CkbtxC(*90$l~d^`6SJXln>c0-emwh#CJ|8w)_59(T8LYOk9
z8;4C5lUA0Qpm?D5lzPWbna2kJ)kEd}LJ=;jNeb6eEn^X6GdhHj)K>gOk37H8bl=0_
zQj=QlFQz-t&RqxlQ~4_FF|5$^kPT!I6T0bn@&F?ftlBj^X2blwl7cNgPfw@oLB^hx
zHF8ilkFywJfR8_UQ+1LYc59n=R)&B;-F1XjA?5|-@27<)WVVH=Y>vk4&&KLuZyg<l
z_206SPv}mN>Ey0k-Z2hVCx!GCen^siPhG_kGk4li{tB6{t%@>cpwTAThppHo(H86F
zsrU~opl=5)Jkv^t6=k@zx}?A}X_`4)w!uNnP+rC_GkGP)B4<8t+jM|gOWAGsV$8)e
zXr%Yew$B^kTrOYtwKJ37yu)vqmg8%G!LEP9gTko2jdy?ljfwpkowl@n#rPz)&ds%r
z)H;erPUrX;*au$N6bw&^ouEkTY}FO>)VEYTJNwYv+vB3I&k7SZH24nw@MVGmFYh(v
z`9Vg-x5LNpeayp^a-roD7aE#2!<+9LZ7IFeDNCVVGIlIT#a_?h$lKt0V=`V2V8(+W
zjDts-jG81%yY-LxPSIHv*+j`{Ebk3v!q-2*r+n^HXXjh@SFa@xR<^$KdBrO@X^LO`
zq=V_z)jY?bloa5wQwGKm=jFfZY_s|1D2-2Ql{_F5p7|NW+F;PvauFb0dYGd$Sj<e}
zNQS!$f1bl??|mQV4xb_)5bG8P={qy__o}C?S6Mkx%JZWGXuKSxq%y*vZ#bL(03Sc1
z_65mqvPO+h=l(a5r@X-q2a5Gp2cJ3z7T*5qccB{lqDv2sCOpo6z#D1By)$tKmgK-Z
z_9P9p{?!pQ)s_khgpCciVp)}EhFP7H8a7pxD3vq&wzAugiI2zs3)G?OUhU#f;$CH8
zW;ra-=5*J*tWw?F?ds9=C=F?Xdht3tO3j`=#~u3&j=Rtu{1tLir!L3Yr&SWm=t6=O
zV|_C^q2&es_FB{!!<+e2?kyxD*{y(*hG4MbK-iX61YJ!$E*@<7qrdf(tB}JDEMAef
zw4NqS-ce+5+!i_>brbAZu;w-F86b4M$6%RLembcFi{fcG5{rxBpEE@xqY*1lvz$RY
zVi(bg>m%v^cA)q1EmW?#+I-O$BJVWmT602EXVW`G)KZYPv&X)@RaopvMWd;ZNAKoX
z1;hplzCZmak%kB4_3LTA7s(oBLrq2QfiM>I>Yk3Cc`iHu&jX*nJS%>+4ArpfKON$R
zYEf<XFxRL7Mtv@VH8U`@{bI$C{~o#>-8kY3@|0(|9z1QlpBc6$;Emb<y|+fr9-vF>
z3DZ7FCm*)77nW!O8d;%oM(@Z@c8F3WU4$knUhs2D2Z}l@=uI5z4|jMj9i(LmNORgx
z$Q%k$1)gi5_bIyu<hZuiE4wLdzf0gWB*T1hAdS=Cj2KEXb)d&fkN=JhCbg}JYqaJK
z8_cwrLOH4>tBG@OpjHm`S5uqtPJ+s*HdAJ#ZQndlTr}T(Iux^RD&<c`uZ$m6)ND#p
zRsxn#PTAC`@haw0H1$bt=WdzJsFU?sF-5oI52r1uFh~!yZJB)K;Yp6WU`}Ss*nIB%
zHy@BxUxbGb$i9e3@`Xp`BGX{-q&Y=D%Oo0OldBMMY7vGDsQ-z?OS;MmI-zlM{(Wk)
z*kEkO*e_Tv3;7J+jvZY(bunjA>0^u77uQ~Xzm1&IqVk2EY_!2sS?S^YPQT(+oYK{k
z!V6tFQg5K0K5S=cXVrbj@6&&uu7FR$Eim&<g%`_Z?wvf&`uypLxbnXg8BP6uAIQDp
zC)wi!t;lp7cf6@}cNEWx=Og@Y3;ZX-_av1rmHK{Ij^@fJ)g%w?O8vSW>TI4}+YTt8
zuwg|)EV=RXuaN9f#N&c6<K;+O3?_=a9LOff5%H4P5PSe0AR2R)8}CW4YOo2OuM=eK
zFcEae%=_F)`gBiLq32n`7eEQgNw6|fs2nfdxAvhAwj`EGbACFMIK?KrpuOyLs2_cF
zN?iZ#dn@`<)zcRtk+=kAZs-vhh9}pFC|Wdt#+e?X%vN<}!HD@d*cUrfUHTV^25m%j
zy4rv~WZ<Jb%DkL#6R!VF)z90=*GSZKF>@b?fZMbaV(ccyuZF$T<=R}>oR?vRMow#c
z$T0+Y!*{a6-WNW!KR1m=LK=mblvVpJBgV4snq_a*W>z~ZCdkoRC1TWJbHXER?5yI}
z<w<Qi^Ql>YckH|)x<l>jZt_gE`ReU)WKV}KumW%V&C4;~@i|6x&l+0aJX-ath!Gh9
zfoD`v7R$l|zNhhFUHBmiHLWeN*Cc@jd^N!L>JqI24VK|5W8I9s;}{rPY&O>rMT*|8
z4`=?GX4`jbF>~qy1c&3WW%T&s3~9`qsnkx~C>qW5JNs_!oR9GXGXBch0lMhnv@ex&
z_o|v5+CO8{>@Ii*kAB7COM9)hO(jCC*Dha*+^4KTHrylcmi9Ov>bWFFx%>B>(rkLQ
zwUH=R@z>5lqG@0|@6tc6+#}DmX4*s6q=h5vn`23RQs`N>#LMAU<yaOw$<f(z^+kEd
zK0F7=$Nz(jD*qs(!%E&ByWMDZOf=WKl?J<anL1S>qwTKJ(`FyAS_L>sto28f(tvlh
z>#fQPNd!6$I?EG3*=Hf<ItyLj&MBb@aC7oE#pUS6rvQGY_)kd<`f`~l^rbp##2j&P
zv3%n(ehdQA@e$$V+eEwEBd^d!%N1eBOyr{tr~Kk^=XkLe#|DbZUDQ0h{fv6~iqJPv
z%Yp;?WVlTPa8x$9-ZFi%5vlb`7LZlr{YC+GWqyEk`HgRN1Cb&x$99bv<8<5mUAiJK
zWw$6q#Rd+oH!(-G8fo?K=Kct66D@tEvT4TNW_OBo)cJTf@QP+runN4j;}5g9rO{pK
zrVTJ-HF>|MS|K>eHu#f<LCG^}Evxfwl~IHrX|ZyH1h+QE5r>_&2gGx9xOvycggLcD
z95Z%+MbKxeY$0h54e~spCp=Mc9iJpKdNwD4-A)Ns=o!E6V}kiyJjnj1X=e0rUXCVb
z2qO3Zxk!mOjaI4Z32L!O#^B?W?%u$8z0{v0L)j|MtBe(vFihiYv9rY77uHi8N2i+%
zLt_BIR^W|;mGd@jM%wPuf;zJdeCzK1G7osxC2|uy#JRaAtkRs@-ybHEdnK0&<uqOJ
zJ9!_wCVkA_DrpvQ@l55hLxs1OYO@&qv2fh?B86uUY?{l@-)m}FAi<|0qwN<E@eMI~
zRK!3-eX;_Ezm=Ft7F)B_csAm7hst_JnGEX&@Yaj#H*e|7U4sRQDY=e_)wWZ*(yDPC
zM4$4uj$SghnZd;jvIWTb7`R2!%*6?812Eof@wtlRs?lC7Gu1$pdBiwT3W$CFcfE;Y
z{pWU*8c7Ycvmjd>ZKC@{?{-KW{7Ls4|H#`?0-&B4^%88Bw^;U7ZJ|U*9#5O<=pf)@
z@CmHlsJ2wXm|W^%lxBtT5XMJjrG3n)0cGJw@wjQ|V{!aRl|}wI<7S`bZl26Q@Lj#r
zjo)X$vDN9F^B6glb+r9d!RIA@u}SUhSVVlr<tQG}Mwc!rELwvxQ!xEQI`4(q7id4f
z_+K$`Y1QxXP}vc78Cq?DsGr4PE#Zj<SBDUKln?2cv6DD)rI^n`qBI;FqYu?Jx%q(U
zCNWE!$PR`Dvsh-oz6}{=7R%iF{uo8`kdFP#W;af~d~Yb;al>KdszO55&~?Lgf_GEA
zQu#`dNw}6G9fKCmxq)P%*&xU*rq?XC<gsGl<!{&h9-}u+V*r@8Dp#5Q{YbAV53F=l
z1KNx0v-tq~j6eKwB*l{kb6yiekpq*Kmjx&6HCV?fLq9ubYIMNedCE<F4*Y_@at<*a
z7@UrzAUzBjJ_)>jrWon%ARxVAwtQ-`HZ78nx_?1|GuCogv&8kb<qA7q)xFtqM6J1F
z?60JfY^g=6Lb_+kmO^J(Swj-gBvADZ#7Nwk9P$usn#17Q({Ho^p#2*tt<2Ia_PSe}
zep5XDhKtRD03}y-*HAfEU;;aYJwi~A(m&7Ew`X{6GvJd;r0ulDvMiQVdw5tRL5OX|
zA{TD<T<N$VwSIHnNZE=vTvNkgH9Ym<nD#q@y*a(F-c*g8mHbqEt5`my@{2My8mJ8Z
zEy4DTMs|(ow!ZJ?b3(zdJ7eQC{4c9kv?brFZEJk${(gOp#iD6J$kQ)7Wa$5^41f=$
zvd%GXbu=miB#SW}@BaHxzxvU$jHq|j_e*!&xD_)0#!x$M;2()WU|U6Rl4%hgr$U>d
z2B%MzN3{zh1pk|OxURK<N@A4^vlgz7u}}nH{It--)Z|H<y%bd~`xCn~RSgKg?nn%l
zR$SOCJTdi#BbBjjtD(Us+2+Q=qwykXp3DA-Qxr|s&F?rnGhnVf>0VSIExXfA=(@`x
zF7H)z=!by@`_^ylr5ZU-qR^H+u6+I#WA9mk1GAU(baXDy*T5dv_?fu;Nd4P!#h6Ga
z4%Kqly;DQ@r)q|fVsQO8!)RmmkP&o962*>f{o(=!;a_N@u1#!sVVu}f^ZSy&V)?>%
zZv&5@K~kG7>tc}%Cnrz4{r9z}qg?m2i)+LuG(~AW5s}ymU(=D|BS^2(IjqS?%P+Ct
z^ka@;^=(`BH#by~sS#5ckh4I3p$^5F<k-$~+PAn?9D+x+>-fe{`q+j0MeqawM<i>9
zKF}<|T$6xH+yq<u-p||{<A+)d<{P5VP#WEw?6fZ1Y`?(ay-|MoX&dQERUpSvgle3v
zy`^IUJ-kw?Z+yQcBjsD^hE`)psHCgD!JK2kY(=U%mcsVN5`A$MB@LEhwJ?=pS(w#w
zSsWFMi3w4oCY?Mi^b=liQF(bethi{_cH4GQsH}gvHjSW|NCI-}8F()Eq*V6%!I}zI
zx|uI_b$n9#5R@=R5-!Wb9o^W(<Re$+9{WXxIEnXc@!`2t@%57L_)RmNAghWy%87W{
zPnGOn^9qHTJJB&CZr;C(QZ2qJL5PTCO-n46eaK*oY<zANKAp`(ptGOs)`=_vz;5Le
zXzhb+X@e$w^Fg^Y7tf}5h9wuEf^hlfs$soLZjKftB>g?|0s_41E=IyYq!jS0#Yh<B
zcg@7{$`sz&fBL@^E%TogU1c!XuA(xU8<Prh(Gh+fXKt@0>ao!FUM;kr8DE|;vnO)r
z4}xa;`pP?3*=<}%%b`79zh;uP?Q^XvZc~gyRbSY9CwlZ!(?kU_c;>SfE8!ZYkePN*
z(TuUiHvfb<%5sBI?i7j_d9SkK2*K3kR5hZ|=J?Ryhj@$pz!iBKpFpm1?BIDWvcS8N
zqGlH2%g67#MQ5>Y^)LBQF=R~XvpMS=C;^zl9HkRdri#jk)C=%b<kmQl-KjDu%;3o3
z)y3-}o;PN0mw`tk`1E36uN_dm_6uTUGzBM=HqC2%-<uuk6L;tPn=7&OqFh9I*QNE%
z^4$~N#R($zV>IyzkLH2+6*v^zw(704jfh&u;euBl@GaVG>#5urgqUB!8JoiA1$a59
zZR`(`GKs$AR@&PT6@!bVcOHTK7p?nZ)DX?Zt=XPosA*?F&((U-t+|N;7w2CoY{$tq
zy1%z7C}sx-;&$0j7~Qh{+*|x)$+FLvDT@Beu`or+ei`+M%YcmhS#P!NW#I%moX+qB
zOkOt4&%n-8)}5_i_;oirw9qum?}!5!Prtk5)5EOw1;Cv}U;Nui&s<h{c#}=n8|DDU
zhtevC?Gmd`Chcrw(AhmtJm~oNWaepy7{3$Y7lD(W{|R2pmlKrRefrnfVn$fDFT_42
zq`zI}S>w-*Vus|i4d-~i5<>r7$X<h&hj@u?Dl`S|@gsxAMV%Qa!{tqkgpb(mUPOe;
z7jP#_Vy1r}lr*X3U-@dO%-au9Kgb9l%;t<5Jof6$w6Y+|i1XXiTUVtQ{#qN?VEs5=
zZ7T~yP7qA*h<Qyjjdq<onq`_^RtfxA@#+;7F>QhlWdw3xT9Pm0NFi{}{`0`D<o*+$
zdimfaP;~jCbHaSO%;6LkZ#6T79)F>|r?^RlZW}Zm7OCT=%YXKWblf8k{GcV=H^8Xz
zNYiL#f9}jA?Mn;Oc8KQZBtkaiT%~?p&7W`t8jLPNDinp&K7z1f$kE84S>WvF!D^TR
z4T@Y}Yy>=h^zWV4M%cM{H$PjeptB1N?+|_EoSBqKV(G-2+abUd@0CZwFUvT^g}D>0
zP{AY1IM|%<UPO*D42d^Z)d@ty5I)2|Ly;b$LTuAn-lQO8*z_W}U}=0|PqEJN-H6l8
zw)6cwwGTV*r@+gY?xEV%+hvdT0{4oqwJAKik60gw>gwaU@*Ql83z(BPL~B8+x|E!<
zPAa$`Ush0P&+$1^gnfZxV{!0gN<TvkJm1K%wxwGZNKPA8c`Q*@jTvH}A6gCZsg-}x
zXqo+b7eE5bQ1_rk*P!1#Y(#Cz*CY!b!5_EV1fZ?L!$aOG4heJVa`5y93K!3u##P4h
zHX_@5udfP6)gZUDFN;^8`|ezlO^ekt7spKa`E8S8=CEeTLH`dRQL}s(;O4+h9<n?H
zINj6X7LMDDxW3z=M}Bi1|Deai^Ye3mw?s)7_biPzk9Uqa);c7z7j%Kw`43^<ue^iH
zwO$07+$6O5T#t^8#4a1(mhUXTY_|KgJC)#el_t?k9AiIbKhcf=DuT1NO1h8;E^(uB
zTq!uf0)+<4CYrr~B+6qpN;cLfeRp3mfV940{>Q~OYPyNBU+=>PfYL5iwvYkRK~mre
z_dP5sJ59o;4NDrs8V@X7<eGWdaj+M@f1TkRn>0wSR5259f4KNd5ktJmn3p%%mb0S$
z9E4r1=N%936-sgFNd4G(gam}=g~DY#?&&{-{GVr?`afyD&~gOng2&7&ZiYs7M0q<@
zjZ-|6mNZu0Jb*`<l{)hPOB}^E=WG>&$ILrCEjuN44wb4}<cJ+0$tv1=hRe)XKQXoy
z5eCPM)E^>gO6K_r#L>6*bRPM%I$YP2)A2G0)l%4H(D*kuSx9!)zGToEex<*B5~$&`
zj&P{5PPG?z*V7~!F4E`W#$qsCF1Z?$*E?{Fmdm6?V#<y$HXZpxP8QY@BGA)t%4~Y7
z;bNZ5GyWDnHyswAof48C82HPcn4>NHt};H9R!V6Bl4O^$3H)cXq4nS1J&kM~^SF-U
zd9tVceZH98J@5lAF<OQIcsv<LVJa&J!yO05ijJZE0S&8Y_#Gh(352PW3v*kZ0!?i8
zxHy)a@(7#ESC=2rRt_ZD4u8lL;1!qy)+PSxp(h5DmlCYY9O^X(XKVGv4GS7VN;TM*
zPar<*S|J&Dim}Qt*!K&){Ilo?+0%P=*dBk-fP$xnzJTG*rN(2;F;MIGO}i|%UU_=|
z$c=!f`=;<hB4l07lWFPlEO_!9MNtTZM=Nq*ZML$Es7wY*z}yYVVg?_)O6XMk&_&ed
z+^oOxK=!`~J^Q*7opGr-a3f>sTX)oYo8A2dYc=B2J#`m?Z33fLYQyK^t&#6T(Due}
ziShN0+he!pLuN>UdUBs>mX1Du#lgY1u~w039QXd>kAtfp2g!*!R50|Yxk30yBuRlv
zW%S$N(>~pBfMyU&xss^Wx~VMKP9pN)#5e!NJPc8%z~u$_qxNTbx*Xmalr+bal9NIc
zeKLKM08!@dq_XB)i^0<%F(gDnvcd_qv8N+$vRrIky~=xE^Qy%<nyoJup5n}P`M(RZ
z92UC&j~kEEC?g^5|9+~M|C*eOI#tZb;D{ZlTT!)h6O_G+Z|7t<*#^foWMh*6yxr|T
zhdSNP&nZ;PehBT=w;3*(1XEl4?^apOpaG+>oq&`!GU>ADb6-~c_a|opXR^WvxeXwe
zSbSPt!PMv_iI5k2X}b;07Md=O$?rI7|N6VBfT;12u&N_z7-J@`7QdGvl4cwUM~p?C
zTU(A=-@av0eIBfX|L%>ZSYMU;udlpm)d`{+?{;^oQnP%?Nnu%`!S4x#7xli#lQzwO
zf%=1sbQtEW(kjB;+4P5mM5d646XnGy4mG=DU5=B89kJ$@keJS;`?VHNhsK%ptxEJD
zjN)lmu){mNj8~OVLdg<b0vbBbpYv+-6sf}VW#JFNjE3;h5m&LHA;^x8iH%`_cCs%_
z$WXX_gGHOrK<dkPr58)eEN>e2Ddy1KaA&|tKOFh!1|sn%oL&PMY-IB@Tb%i-uL!lb
zesl7J<}nu5%O+fHPHyh4fUp^jGNDz__oD~wl{dRL;!AoUn6GKNdp#TZG;5n1Q=1~v
z`5Z9GA<lK+!gT<s!Ccl{^cg<%98z2qgI&#we4Vm*hVPaE24%uK;fE&Q7(Fb_bXA10
z{OOOB#N`igSL9Gaz)kI>bX?G*nrlGv%tBAKtpbj%`(*f60Wm(A`ceBJN2B^bNJ$=!
zvx;xxEKKMLzh%EZep_oe%Zzy!R>sdV@;6Q<A~s2bvrT|oO|ah}U!Z9n#4r7v>+lr-
z{`5u7zAT~ByO@Jq^vIl48_I7wo>k7dH`G@x@mm(s<20tW_<5GJiSSk}<@R9jDDpyg
z`K}88cOT{@#0U1TAnF0sbE!C<YdP57C`5fd8`xVixdU?Xo^sXjBx*Br;BQmBaYLp%
zq17ZCo_g<o((L;4Biv`ZVbjCD?Hi>ghFR?U5lO?QV2wf&)ZX6;&_2(>@2})p-pIzI
zp&83kRe`|rItF++O$C$Dp_=9;8zzFFR%1Kp5s2zH%iqnVA(w`xv21z}@E4Yq0&ID9
z<H4Qc`j09O?>!D!ljRKR5r+X)O^Q_G7)AMVI#9#q?p%{p7#6rwQER3CoRf8s7jCW`
zK<pJ+CR&puNncsoFS5PmENSS}xBcV4!-tMKA0FN@O3!XgNd{$AG9@DZIQacki27?9
zT6et=ZYa2}<$G84_Gri3QXLXh^w3zpnD~fn5gD}9tUgnabnWu{JT{#<R}y{!dY+VL
zSM!vDUp`<)B_|`kTmCMcw>I2y#n{p08A1_|9>|m)U2V%^yNn3mPAW{a9lKCafX<;|
zajb2c5YuA5^xM5NUJHo*g`wE_WG0F^G;I6yVGN#H7+{&RDLFDJzNl9wTK^#`GTBzU
z6+3u>3b&1CcsOmNQH~*o(Y*1lV@i{Igs4&S-Kx+(!Q#d#nPvFG6>U4V@o*R&MKXZf
z><UHVq6yGOBoKqXn;Q(kUR0m2N3PCbfmVEGNXcY~3QT|Y1vR}anp<fkk72jmxNy+c
zj_%FMZY|Ft6a#}QjqA)$@QK9Oe{<bR{l$J!`D8Ttb}X^g4PT1v&Cvp&nHY6^tLJpF
z&uTO#EAvYsiXcKRR7jF^hHbWQKQ`<_+=QryQhU%{>bG4s^aR3*uduDctaR?h*CGEG
za(Ms0f+6|GdClDPkbAS~RKrY27P#KFaq@ubT9Qy`m{mYv=CmUG7g&9V`areVKhp`F
zLtO*X>xjO}Hj*M{j#`Zr2kMnhyv$3zxACExe8WT^Ybrn!OK1*}2Wez*x*2f2;3P!?
zpK2(>d{zctN~q5aYP|Pg;;=U7B{?3ucLs+6Tu&fRg_y)TTB8Po=FDW3APTH=X*>7l
zN(5j^20C|pM|hzqE!M(>7|(fG+EzZzIa<Jx+6YNFm~#s1L&}tnX7Ew4;v(}DDoA$s
z8*vB#H3=QL|H3sM|Bbi)7p}uy3Oqg_^aHO-Xv68fyu=%9M60lYtW&K&uq(h3@A-K<
zDSY^&Ucj~a%IiU}HqgYM(PZiPPO6w04aw}*7%1lD3n5m?<j^@vPhwISVX-z=X<Ue<
zwf<~;%p9QN(PAaOCqf~pa4@mqp8>o3Irj{R09txlp(%)rs#cT?j>)ryJk}EbqtMi-
zW2coiH=$9}UFdp4C{$S?maq!+)b(@}#W(7Yk<{cByo6SRgPS?su}%lP&yU$)QxG;|
zm}x3zO#;phlz=hJq<eOqIEWlzOV)#i+@hTYsmI+>WA3MG&q~iEtc*ly0rFGiSYIH|
z_4SbFK>yLLxIi>mj`QzlM3SyxcR!cdwG7&0cC8#?n?^S@&iURlaOd=CeEli2dj$Cs
z;{!L#j;L9QSu6{*+`ZFx%C)%(I2I94W;|P$YoC>6_l_3hrVx%x@<E+PtpdX&mMkL7
z2P0||&FN6|RH9DEdgcESIqMv^X&w`UBY_pZHw8Pr?b~eDuiygDo^znoQoTs?XMgcO
zF91tTi1hUJ`!~F`?{venbQ6@ERL`t)0)Tp_*0_Z6uOlpqfR|%Rk-lFYW3JkK#uqRo
zq*Rx3C-bI=bvf{er1?3DOnk_66&iXY4P_e@(oF@KOOe^*&@j$(xma4X@{IdF6z%y}
zeof1QZ_mD<jAZ}Y1Ce)}ON&{_6pU@y<RkT6x0l`k&dlcb-GbRz?`=e>1`aPKVSG3l
z3$3gop!{#X*RoN7(Q5oODkHHQfA~1cJ>Ti}6@gw8w_mhbjd<Ip=!z|i$qI{-u6cng
zSdO|3Jq5zHp?M|i^;mec8OP17%=rGCEB`FJfluAO4fs9nq<mNXQ_B7IL-$Q#0^id?
z_b;Ws<d546ka*j&>%#5lbLdAL95en*TXaM2{fo{0H$Xesj3jO3STw)Y@~7U;Q)XJf
zHdUl6PIMmGUS{IfL}-*g%gxwkWXwoWDhe&q34uTy4AQ(nqTiJ62(t-z4>Xw2pQOhV
zUm?nC*x*`Gi*uj+<-6T2h@?1t4}cA7$1XPSfZa)qnTqi{cp_YyTvU=6Up{Y}M(gMa
z#1HhB58jn*lyg6Pa{}_~720iBe$Ug~(?!y|^}iSE8_q8#hCSPluzfxS<O$|A4m4I!
zX3sV5i%DWh$5MRszIMNQqL8s!L$oE#t#24VTuAiQKG_F_u6IOMEx(8VoCz5xp}U$-
zZ;Z2C)T4b4r>n-RnxpaqwzhpT62C5czkD-P>^bTzz6AJub_=v~97Ld{^aCE$ivhp4
zHor6cI)7J?F#liSc>m^TG6??nW&ub%`5gkk1|fe)EU`FH`oKK?+0V4whKfIqvPCeo
zbX)Fm{Q`b=P<O(f6*gQd8JS8`UTR`OA4cfWff`F27sT~fduuj)Xt812r6hBgj3Xs`
z51_=@;XQ#gdce%l5qX!pBqmerw&ay8lC`C%(NJZ#hI(5KD`Z)$OU3>n63CaB0oUfk
z)>BSPhIKXv&_t0dGfRGS$FoLFo{G0j+3AZtE}maYTfgAz3%sC1wNC5%^mR9D<P${E
z6>U(e4=|+2A>+Itcn6f#lGuAqQcAs*gMH<sqp}JwB1n-_a=x`#hm4>t89ttVzhIG4
zmv&>&_s$E_enc@Q3Ctp2t_|AYX>JJ$C(nPtyfoZ7Z7)V}bEg9$Xn&1kC+>4|1{Dkk
z!;v_n-{~7OjQ!l9qkuufCCITJO83PCRN^jn#2fu1!2-m~A3a*Oz4?8+Q4X0l$r2!9
z(qkM)cyeB6MZUqp*KpzHPZbdc=IaWtdpUsAh)(t&LOfw@9X7yRx@v46YVHn6DjJh4
z7AkfFC)IUdaVR`B|N7qOG~VjMn@%faMV<_nFXDsE!6LiJ_jF`o&H|x_7ZU?|YnL3!
z{ds=L0|u&*uCq_Tnv3PPOrop-8T_QlQk6|WMxIfLKqPs>r!+Tc@9%e?&b$6;9vQ})
zZS#+KN8pDRz;$E^7n<7#`5TEs1=8a@!IUtd0h3wlAT3-{sVaRXQ<oWYUL}j$*L>|{
zu4uQ?`qf-0K~l>ehZFI7qb{>sH3rXg>rjRot8CjALU3?hcS`$z9Cg0-r#tNlrjW|M
zSeP}btfE!ho{vQyfwy5KfU0X^aR#Lj@q^3HjD?raTr6!n=&xC6)HM$jdo(>Ho86U~
z#kpCC6dC7>Y11qTXl{G<2VC{NrD?b|xrgY}REqX24_iAN_U{Fi%1iJMaCx1S`DL-G
zszFrFqRn#z&xld*=F>-)0FBS|fd;Eme0Y{qz2jue8b#qioXYMJofHw*_ibldjD!>k
z;3J!%DN2)sZgg!TBP3+~lg#I;C*+y*I#!zhBq;7^gBh_`JqL=dHF3CYiv32IA8e&R
zQK5oZ%NTRjiotZBft__3PWP~?i=BXkT5==Ho7$%qj3fbcv`qRLb&?Nn`pG?>rvtO<
z^MVN|VsRp*`byOQFlA!`fWG3s`mb0T;1J3kagVidE?DHkDF)NfcOX#?hkB0cLRrC)
zzepCnf)!Z)&Aa{!SRClh>3p5iQI433TiOU$7KE3nYYn<S8J#Z4j2n!iQPasG_XE}p
zqY6Mv^*yvPHxZazG4>P#dxLe-5hHHaQ<{rC)x$^52q~g%fT^I@{U^KxJa8TZ!%ajd
zgtU~S0~6{2650<L@l%af{E11T0oE~ATTuEYaYW8od`y^GlN8oJM@tYJXO?Bl%P<7l
z`21F);r2<~<Mm10{L^b$1v_Hj<biE-3!`#3NMG;=h2+-zW~KWnZbQVx97lpc`adCo
zeM#FwWOn7N6;4BHg82HYT~-=lD!?bfR}=fsYY6S!>uq=#T5J425O!5&Y<T5Q%N)0J
zSw%1H)B2Zr$xsx64{Q~%qXdpVeEYyK_e<q~Ymm9VC%}Q;G@hcPRp9}%Vq4XPj@ud!
z5+&;##G2eKWRq{|VvyuR4joK<OhGQqHT7#R@mF&8QbBI*_0097xvxSj;kN#6S2c6-
zqXQ}A{YHhPmdW}oQ@nA8PNEZi9|x^pj2^F%-xItv^S>1jU_%cjQcJS8;S_62#-|l5
z7lwV`ZXk$_-EpxIB4g!jjvqKxwyM&>qGsomhO)+Jggo#@{mP!>83tU3IK1OOWMc3L
zI85!|WzT8Zsb~kchmL$g(_f-@Ay`Wlyis$0di$7?(*3J4d)<E!?re0MadVh|&(+E!
z7TCU)hUCUEr8x;oz}vTOiXd<MC9i;?zI=WY&l`m|c{@b^(k9cjeP1Xg<U!Khyr7BT
zM83G52I#?IqU8hxNMA?;4qs;A(Z*kOS5F6Z&B^v2fEtx6+j<!<mo*P1=d@2qw?(#0
z3~;~510RxRq!q74P0Atn6_3V4%hN9_>H4cCf%k?JTY_*E6lc%FI$a+TQk!UWqQsj^
zuPlyjVUpm7B5YP&`A1qs1deaS{@n`sDE^VaUb6NJJ!+o^yhr^pnT?3<Q?|%;>)oYn
z-A5TWj<56S!AXFZ<MF*aKm~K!=jG%2?7H;W{%)uzE8gI{Tx+@Ti^woGcc8aUqYFQ8
z{f@0|NH0hT(^{GCbbV%%RK}Ud2S5P#cBcg$4o?`V6(IvrjEKxb71>}J_DY^bmFUEG
z1P=Ld)ucrS@6zZ9!Y8}t=H|+U{r@}#S4jUTx2R2<0*#`A*<_h*TL(e|_i8{vfc<5L
zEeZL}YjgwM`i}3y2#-N%mZ=t?o_1`0&gpA_Q1xXDGvk_%bx(Xw<-h@`-&5}q`+b|x
zI6Z{RVdm@TZLqti1;By;j=XhLdX}e_i{P0!IfdR+<~ZSIPPg_Z4{<l4S&A1#<;(G-
zOAUu&eFvfLDiTo^nu(S4;;Ozdo`Vb)?+FZr$fP-JUflTs29N-Xrto3|Ca^K>l)H2+
zPd7av_w`IJ9^lTXoCb(ESNkm2t|k-|UfI`i)~iASx!ceNnHZuc7bE@4==X14Wjj*(
zXbw315o?M8w@W=X!h@2i3B7%7Z)9qQ-`0vfQAj-U1*OXMJ4Pco%G0+0LOG27(9dzP
z%2}GOw(lIgP4;}b5Q$+JE%&XO&H`XHZF4uK23)!ENva8oHj5OX>?}R#9}X$RKjQ}b
z!U5-4TpDOn7;xdjtWnz(IZw{LX_(Lb@<j=Qv6E!%eVsJL=#In@!npi^3F6mK%pF6-
zfT`3-2nC5u*#_V!woh`a4p4@|70TLpd6n1ugYQ=O`)-MfUA?Cg;X5TQ;!*Ru;llGw
zq`GVUQR_2+TA^S%<GTiefad^r39v`Ah1%KI=l&RBF8GKWcz=G}(0$E2(lncrE!K4-
zN3#Ba-g!OnbQ3)B;B(E=<#vr4E^kMy_cm{-%x&RkN!;P+5-&SSC8OiFa176x06%mB
z%0945LzIK#U&a$2M;`$Jm-c!wBX1mpM*_g8ic9Bw7b3|dGX3a@LgV9_uC***hMFMA
zqWfc(uoMM|1%3VRoS2;~^M-Z0b~-AoqCa^5*53Rtd9Ew`^BKJC56pDan>Pd<NKY>W
z*`n}BIwH3bPNY`K6#NC+90oGYRv9-h4ThU`s(qDqX_4t#-wtzwOenuRyORo?ix%8_
zRc<gL;qTsiI;7r2N%-Lq2~>7E;9AyNW$2bLz~7$F7L^`6HdOvox5jjXA)`r7q0Y!v
z&rXE6wQBXMsz;QCpN<ZY`RAB)VK4LiRnrjc27^<RCdgo&JZ2<Wpb!RVD*B{Tz0(y)
zfotBN!JtCpu0<P7N}Vnlkl^#hnE|ntIVv)#C4Ja-X+it1S-=LwE$4UoQrW#>oBP3>
zIDT=DNCIBSfG;f}@do(~w*SM1Zw(GFTI~Tssu8)yZumP}tEe-eiwlw104CK8`y&NC
zNc;$@TSY66K#l_iah&;-#{x+k?~sovkQGM<%nVDK@}@B$+S-N|SxQ#OGMo&NOvjl_
zZ=$4SGk2Eg+kP{(?b&<-+RGms%-62*K28F2*;h+ppCn}4%<PeyVkS~ITk!!pf72qH
zmi(Oa4CsTSGu4O0lW!kT>QBom7>P3j5DPJ)x%e5+gb1()old%((46`IGl+<DHs`|N
z#IFD%vO**xHN+DxT+6Y{c_MpCF<$<}58%}kN7Pi<d{FrQ7C8m@5$Y#?rFo3#?z%N5
z_#&O$wapLJT6|BXMWG#Qn~a=aCc~IXJ0CK->ApVaI-3Hsr4;KnE-KxqmI;+R@ks%C
z`y&A&6)2x#y=ZHf)5Zm2K6~axg$MsXroK8X%4qvqN;;&w1nI7!LmC7Li2>>EkS=NI
z?g6Bwqy?mFkQR`kb4clK_+Ibt-skzw-}6kobIv|{uf6tK%EHH%>1K)9-wWJ=Cy(Z|
zLez^ff!;(V)SaInfP=v#Hjl>>hfBZ#`OiA<xu`tIQ>RqM_=kOT!(IYx=@mmg5t3=d
zx9C@qWV#9ws_2jeS-KY9Ki)LSpJ$6*PnkqqzA$LiUFsdT`f#rSRM%;GXvx7xM$?&>
ztos1T9k`Zfp%J0<eG#9NLt<Fd)RYiN(#(_DO2cJD#>#LAjUAeYOvz@ZPkpL^qu{2&
zID5pVfaF&4?eN+OrzSn55*{0eEvDBDG-2E1?AA_ynooLDe+N34?|HXQRv?zKyhBAz
zTc4&)i#Kh)mFMm}Ralx3(w<BfeXi4pjq8v6Rp^@fMm%sI(My-3!BE2@Iaq5{4B&*P
zykji|ioG`&?U)uKaRpB8gSVeGFwDMQMUN_6sl5BnQ1j8Vy~)Ps^4n%f2c^<xyBoxb
z1NL?c34qzqn3Evg=L-#gDfgj#9DMPEWL#qyI5C8xZu<5B)joU@D|g4j1jtHBGcQE*
z*JK!EX<y3ApDtCcDSXFaiSxJOgWD9hwmIzjFW9_4el_{xw_ZYO67Is=_w<e~Hzx;(
z4@dP|D-|U_f(=q0$X*56BjKh~NfgJW^z}51vooU14w3=fu{g|@k5Pt_g^T~f8<RXR
z#c(qPkA}W9{0wwY4SqP%;scLkOioq_V3)gj<-!IMyPg#SqbB@MyWo7z568bu@_mk0
z#cvRUE>9#aQkQ*2?&)P2jyiE}fGRle2r!iaNP0@gcbyXr7gtBtZTm}(f^j;D@|c+e
znJ<znUCcyDV++hELos#ri<zYl`a!8tT5`cu<|$!w{U99IuapgQN!SR%bqxK5Nxvib
z@0VKl=O=ucSMoT5#H^1-XO={NzP2U(IF(;fUjEbBt84l3YVT<0-alwHKS6jO*MV^N
zac+q{3BS%ayJHh%Z?zvEzqPUZj-7lWBbOb-Hw<#XYJd_TH%gpjr$OL-v0^??D^(vx
zJA)jomA*3=y_BMLkb5?NPL7m3KlUDc(rT1;<M#cmLeiO4HsQ~QCT%Zn5R-0Ns()x3
z4aC&8>yKqc|F_^yg#_NBoyFXoYf-{3k#X>>P{8-Iz(axYiq2(uz!28Qy7lzhx(qR*
zeB%MMGu$qZH^=dA^b59z*oepID~{C71BRdaZ@8FwoE=PRD$o$#$N_>TQ97}WOzQm5
z0mvVE4a$AS%>@^y<Rp8aO=w!ITX;<&H)&1)y3CjTk6C1D37X^-Aa;uQOQId!Gf_A5
zKo&Q%#Hp6ggn(ax3tg={p(|k4DX*SlD`!g;B4tknSc%_SB(Ozhhc1`~eDw(4KToGm
zTH+hhAHpyaRT7@mRRnZAoKNWl1^3JFX}&fgWb_o)14P~^>5x=lAi*!qa)lLxXqP10
za=%>0LFY|)|NUoe1da_U&r~)|j!am@pr?u&^_JLbS2mJB+~v(0Sy}Xq>S;F^y4`y=
zRL%Hd{}tNNM>bH~!p?&oO?c6Fdk7#8++v-oC9Lk7+O;?r_PC(&vzt}}nFi!Qc`hi3
zASR&b?INHsoRj+(FV{#pnBK^pqQfAgecVnZ&rXx3h}8%KjucLh(W3CqdppkD&(y4Q
z0|amI*bJl>7Eo>#blxjVo}pjYas7!1;KRKQHk1u`%PW1V9mSyK$;ru6yE+>xbsGMI
z5Am|yU`aYH#2wR;FB$^pu;ihe126}F=)vv1K7pphLn)2P`MOt8O;D<9htdKCpZJqi
zpnc7tLuDA0m(}cC-8{bru>xYQqhEYdg^RD_9-eD4N-+qlFR3!%HC{Qbi`ApIY{$*d
zvVgocui-b$`GHJBdVTS3<H((qhUrgQh*;8>Wv2O?xMMjgBY+-G|F!Gd^l1K`n<Ve1
z>vq-nzKFq|3ACu9&ql~}$8w)WH(i*sw7>g*>*n=WU?UoN_<sr61*|Ny=h|Nj&em(k
z!7o4n$6hEMmu|cNV|J4wi?>aONg|6*OxT-KWDL*B2t&4F|G-?RFEqUWCy6KyUHS^w
z;%C?3z7kUya;yiTNOWI{A<r+uR^no?+p!(ATmu?tJE@syJ2Ye(jhm<_ZRyXRB9sRV
ztOrA1exINqryx}~csZ2p&L&(ruzX#{j0~YDyC4{17Vx)gv(PbsVT{SE^Re^amspmC
znTwAN^hNArj{H^S8xn|@+83W5u5WM<7UCo=j!p|yRwvbD{6(Nth&;Q^(V<;GsX}Tq
z)T%MAX4~?U8%s0lH1Q8AW!sa+%LbN3-!v7#Nw^rjx<MPN$q-HdN{@!BvdFLy56^7e
z9nc<!)i>GfmL->6+k2|0*cKuYy#nR}^6k_T7+TZCI(6%g#H@8Nrr51(?7oq!Eb{Z{
z&icTFe5<bUzmFS?t7@L9y4`*?h3S14B9^9<Rd*XJ4;<c)-`WKK7E$MF6*i(n&a^-W
zch|cz)9DRezC%1{_c+jr=PP2^#+yy=adc>Z`o-aC$00Axukf?S)G@0me#WpvaVa3)
zP+@1G%5{{&Z$xZms-;LTYS;goE|F5MoRWqCLR#Fs9H^k6Fj(2s?qQhK+z4~GV{A*>
z=R&@~kHSMBVK^VAx2(b;w>YJ;ocF~cAIWlsA~8Y51bs<pR<c~3DB0rG+S2#kM*<xa
zhv|^?a35t02vteqr+lI1+oQ}#$h^SBb#Az&<r4#vP*)KgtvhNt;q2$71SWjW;0J`<
zk^$`Ikomji#;Lxc5sOcW9z_CKnE>s_mY!sZ9hF&KA4T9++xsnmEwc6{Wm^4IJpTzn
z;w=A(atF#%qzaF(Kbug~eB-q)ss{Xse`|1A{u(-SS^=bVK%g9x8!_5Tq)+#f6m^_4
zAXmfUhpn<hF%T#M#(*}LwXGf-|30ADnP=&~;^tPoXU)S^Q-?rj_GDd%yrb(z+*BqQ
zug!wSQ#SK+4z#~eWw`<_tv3HgD91Wv0q9hO+7fbQym@ahVW5sZx63q24%iCjrN?Rx
zuusQoMqvlrZn=CP@EuZ4cWJK~YO<&<BP~w0w&oyC-1ORdp@`NaFDd;Zr{s+ah8op?
zfF8{p8}@kLWq8StJp1$Xigzw6NFh+)*vOQkSZQLa1*p$=UF;l$Xzki`1)<AEPa8TZ
z49+S`5MtYhk+0%wl@g~3-eXP1(h)l<(kk;ix(Uy6PXlXQA$S+~N)?PfpyCii7EO+<
z<~q{_n?$}!s>PCl$Y_;mVirWE2=uIEq55zy<+n*_0P|R2)&ZbfRTtS(u)-62v;^Zr
zr|Q8r$`o*M9o~0^3^n@4b!`i79x!fuBycZCivbMG6E?j^F3h~2*M?536InyiS1<)L
zI7o2qqh;wnFZhOi_JgwQKqt#DN2UAJZ1uR3bNaPu5NEd|TuVU`ih$OS@Lko9yajI^
zW9~XPHjHYgoEnfbrTWnaLu1ma1pRJ+k-I<tk$gK25cb4OX(#Xz6%pyQ09Bc54F%FW
z+bv|Xxc=eiT`7tKq=7kdV$#{8h-B!I@5(opKyZ&i-y7LlpO&@ZD$|a9xAvOW+H6fn
z103bIXd#>5rmSK82X}#m!UAe~LxiFu>4Vo4fIVZ}wZl!Qs00qeu3dsFP7bE}ZHV4C
zSq$JRWQAwJcshsg@R%qqY2U%!*J=gZ*NdJr4;luR4QeI)QYCLOSG!Ycp7+J>emWzu
z2#b$`gcU_mo&RrpDh$#fc;y%VIUNL^eZ<Ig`V<IwQ%#~A&=~nhyC9$oc-NcAJjC2K
zAKs$FV3;L92}_TJVX9#f#3qc5z@m2*`NyZ{K()fn#Bx;%UTi;zIV2d^!Rq^BP1s^h
z#`UeWb9g{x)2P_9j#l-SZ#9_&3GPHxx2|8<j7FIbgA3+W31Jjq8ohnSR9J&9xlRlr
z9c8=}y{Rdp?hbaSwrf?sS&SaJ&t+{v8BvkpYW?V=lZh!M?4{b376#SFGG!U5+6I0!
z+7CP#A5?}~6nD!)%C5vz;B_l}zq92i6gAcNzc!_Ju-4iD!k!wMrU8V!X`fU2kb`Rb
z+9XblHMP0MUmt`s20h(L!VQn%57n1lBfK{4xM(bR2+9OC*wx)FnQVsogI81iRYoiz
z32&n{LPX#4O^0Mx9~if4k>Yc?<H$gHK}N}iW_W|fW(T*@UJ3d+3>Z=@i$ovvMX&4#
z!(<b7MxUIl2$NFjnb9BFvh9;ToZsEc;dMrkK|9cin2~XxU@rO{ZCVQYhTVi6_5x7S
zOSM(QA&}3q%B-MxMLIzZ!{Q6qNrw4uSDU*v9%Ly6PJ0bJ?2_8|sK|<i;W^!wUk1lS
z3g~Wgglc#EDsXZeo6$$27d+VeS?TrZlYjZzN7|L!UX6wZ(E5jYKv%}c9JHcthLvz$
zto(aA*_GACVa=hSC8+vq1~at;pUF?dv(ri~9OY0vZbd?hj<5y!=Gs(^CC7k1yCgR&
zc96humSB8YnSd^?9kC#}6)tIel{xw?DT@H7US5iqI~Ch|Wq50CGLJoe`u6Rw;^iqR
zv{Y!y%aN>@;|}=oNTvA7$*)!T2ElXQ>eJu+&Ad}|wE&&h<MIDnx)Ew?hu$-p0_lh0
zrYnHc`gCT73gKcbX6$)bAbCy-5tvtXD-LRjN7afT9q`Bh@PXm2iZqw4a1}de|LBxP
zX(nN|`D1Eao|Tb-M6Gc{QzL^yJf%I70VZ(TNlUXezJmnDm5NPtawueUzcY@&ObypA
zxTAw89>YQx>V<_M<fXChWAz?lt!<mNrwWMqukk>_vAqy{7fV>Zrj7sS#|3Wt<Q(4M
z<2O|aSpi&xjA?kJ6nf-+-0w)Yqo=$fn4sfVHqchHIyfMo%Fx{I3Nrbv&JsRwJ16Db
z9P7j?v-ZGHmZ4bcF)I0>)pea`5Z%@tcQzS0qHt3g`3%Vi-`b<GFgNz!e8xqfK-NcW
z@&Nh_O6t?(6NVU^s!rm}M2LdE*@q<s^#}wTUnP<C_%)U|8eh=q9okmFTs>CYJd^8I
zROYI&tf6G&=qrV|T=c3#Q7h;;4b-~ZL@8oM_7fTg6XR3w8Is|l7tHC>E~|F&u4GM@
z$Yy;FBqAlzvS+{L69OydxUf~=w$-m^qWR<?HXw0T<J8EU{hP-$>|!ogrF1Be!nAh!
zJa$ag>k4Z!I*~~xGxzLi5DqJ3nHAL}(f>RmaU>*9CB)Jk;O+=emHl@elH{0vXKA!3
zYXq&<HU=*Uw*$s%Wo~)aIINA@!0o@(1%r=*ELXhcg&|aB5D{NVkl-)x!K_zETdkkw
zT;@OwE_SIYOs83yEWeeLlt4VHS)19i!ROJoTl3{&9{V;SJ42gB>GkTXT*te5*?uc(
z_TuHq#R_raeqi2lAfVo;lJ3;$SK$U*;}`%N3LnLLJky@p-Fo#>W?Wq|I$+cJ{NnF(
z|GWlteo;&vi0l8iI=z>2?KB5zFiYdC7eFwKfVWHGyCj6NUMAle!OH1Stt=_7fVxTz
zN){hYwDI_&FP`rQ6kmK+9kjBW*h+N5(t26G@weSFQ!(|u9%9;Mqu_g$3EH$+cclWx
z@Yp0o;wYVn4h+wXbakf;tm)l;VPQq#DoUE4Vlis(;eeiF!!(9DN{Wp8vE6W%>o`b@
znT7l4R^COoOyIg^(?31w?7K9ToW0($lD60c0v_+z@Lu0YQMcd9Ebr-Y_zm`LKX(7@
zwo5vh{{R6GpnVzi1TI4EVe^Kz$Eo_OgBl)T?~C}WV}tFkQLgGhHcPz%-t>h5dk4l@
ztf{bYbO$ahCI$2khNq)5M(qK54MXG919r2>3-i#_X*O|9jyj;sXWNe3xoelF8F(AZ
zH=wM}<IG{uCc5-I6g0f1^^BpR?(E>ze)>jd^+xDhbWQ84SL>ltQM=>fJc`1SLq`0Q
zr%S?JL|OPtUD;aL(?zlSE#tHSf!)vZfZIwMV1xqW{k6$KQJ7Xec+-7@5@ayJV-I*Q
zcNX&rX=unAHO4;<*BO&tDUTG!rF5fwocJFXfDdDrrZs4ydQ%9pP6ezQh9Xs-aYFT@
z*#mYYxa#Btzj8M>kpl--PR!@a2zV$Ve1qW#6yvI~quY>ENF{j3;w&t%_Mkl8_(QC7
z@#My6h<Sw6G8yWgio)fB16*}QGIEj5=6LwUAw<>aHczWfk|X)6w;5ia*hiY5KH%8(
zaZd|Xd=2|NU34SXBEr{#m*4~IZ;8M|OgqQo9q9i#aWA&gb!qUh4(}RMn6BBnp{3hD
z`p;MW&(~$n?Or^s&}fPWcATG(aYR-;_K2qnCY{yp5jz;4j6%Z_%sWbW^owUfW#j9>
z-5hs?0~G?PRkZ7OXE%Pp)&;}mSb2C)qxTBXDdhmldDwuakMvsYnZK`lPhFsVfp~}P
zdWW?Js*KAhMc*_kb}RnDB^DS15rO2ZgW+gr3VxymI>cb8#hKCiBqTFxQ?b1r<fxI>
zZW+I~+)FzR)HqeAgj_e`nk@P(W(2j%B_%)N<LG<fz}C%)Ar1#@0Shlj7o}F(d=|%1
z!@ihxAdkoLDep7q1+G#Z=i{W;@X)O5PCFMnh$HrJ$7HpqMQt@>GU^Qx$hSgQSzlbE
zq#kO-eyM{{SrKMgO^UeTg1y~1K#lR%xEwo7WK1tvT8SZ)7Qglce7kNAi>G6_7LL4k
z2-+`&p2-^|PanmeB91L>k(7}me0ujxhr?$!7!e7KvgxDVO&)~-P8<@ynB#WNO#4pP
z&nGDm06&(FDL3C#7rV@MMXY9tpC?+kQ>z)}yZC{hn$1@9s$xQ1=?CIk(n1$B7P~|D
z2V%0vJw5@qYaN*-;Ok)%K$}#zXK!zo{u9$R<hWm|UYG3X<Jiqv3X*RWwB_dLbk{TL
z#a-4$=O-y?+7HPDz-~i4hZUAE;8i9-_1kL%1mBCQRG>60xY~A^cDYHJq}n2L3#A7&
z-DTX=-yXXgKlZCo50=aZGR;Of%fHq;9#CrAA>GNwWJ8fb^UR1?o~ktlbE?||u#0wH
z?WDTCLF+T7NCvLaGm(VH6I-i@eXG|cXKv5Yf6M0<m0jeOcwKB)ipXNZ5K~}ie69R3
zd-$a?ycPim5Kv+WRW+l8<XBW)0nYSW!&R1G16G2DcYKUw+tKjH)afEZ7Lx^#74)>*
zH$fHy$8K;Mm593Mvb9GKUFN-hOickl4kN8DVgWKms}-ZJq25rdS8^S9Jku2`If=?J
zjf6mVtKN_2KTb-vlwjrdYX?p&HcnbODZt*0D2upQ^S6k?ISR<q$Wg~_z>aHu?=jkk
zjf`4YB;S3=D(Y+9%N;7kZ>nEp{9n)&8JzEhMzc@yOHXCQ%gR_%-IoB>n$<9v1~>|+
zI+(;WO&LuIgJW-+P6^A4^ZQ5PGMKCczeNOpYOrEV=~^GEIN5jAkJ>9pY7~Qch6o~L
zG{EZf-m#Y#1aaLwiUKM0(rD83l>1!|;&TRH+9qC;2JF)C4o$qz3nHOYw1dwZfr}kL
zqT^4B_1VW{C9e2#f(Ptag0f{lJo1bxKzvHVLPO5@*m;s6-f?YsX6EgSYceKYwu+4=
z$&3*4P8Wk@*YERe*t$C$S;S(EllSc?NB<Hb(SrIs`XmU2C3~D4S`@=dO)B1nrcwp+
zwW=nksvFD?Ju=O_xlxrOl6DP=L0Pj>`1KZOIfX&l#b?3AV<j04B%wOV)snSC{<7T6
zZ(h?t0d97oE}7QyB3VkEO*!YZS^a@@wd)+-Zhl)pa@a+ZeVu<l@-SY0yv@<HHKK+E
zk{mMrbZmHR>#LMdU1$jwMHD!Hv;M#Uf3WLO<|=U_z=@X}9fm1L)VK?0Nendk5GINS
zH~$YLOe^_^%?P-00APkKWzoU5ifc!ij~lKaNqXpl+XS~|b(vC{#R{0ysoCLzA2`jN
z-CFlVvm`u~&H2!?y9E%ZIxt~V<eGP{<No(`^R_&foC@<Nhr&WM3-AVyV!<zbxf29b
z_Npz~?jh57>{^~ty+cMcgg$=G``x1u;9tLslA>w-l;Jo&U>oCFCp-dkl1ozL<xng(
z$+Xk)&0Y|B)7t6S`mp>tJ%q0)3@Y+{91P?$iVS((RhL-dCPg)*rDXQu7Jjz2CA2Dp
z6ahU2vA6Y$BiW(QX{}&a;^ldY$K61xoe2e>SogT&OKvW6n(9#y-w-XJRGCyh2bu`)
zE;mJ#Hmg#=2u=?1*(WW<fQv&1>R}f)+)ukkL(+$UxWr!J0G%G0oxD%4Z-Gtr?#r5L
zCUxX<sEbo-_h2A`-=S8%xq6Y`O>blnq{rZiPb|3v-GO6~g7f%12-zn>v-?(Q7^Hi^
zi5&x5k6NBp$lAD}?WivLk<aQIWit75aa%bO@3k^H&988i4zK;ps3xP{H2sl4qU9?u
zdo_=L*hEkLB_M7dBv9nZ(q<V8(NMs*l~mH(AAlIa#S#oZ;`y6s+l<Ha<*Cry1gzE8
zALK;*NQ2~CIB;+Pbfjmyz?7DIN@b|uijnQ_sIQ9oepk|tNNCcc&hu&FPe*$+>z9{h
zKKtuIn~fb2YP?s_n)P|UpzOX420E!EW$FcHMM=@;-7iM;!d=O$-s*O7JN?|2O(c9_
z0ViUQTgTgnU01q=@Ec2Favw3&lDN3V|7Bw$MjAhp+y*P5470!(+&_YFMb2>hPTcG1
zWbGbMLP=~4biK}oPFNJC$5jh>7Z?oyzLWl$w1|Jk-=Z-nyG<}VJ>Rv9+rb#(BW&ty
z7u4q6_G&FnDMMB3{wBYkEtL8-w={G|nz)l|J=y=|v!A@{iFrV`o_cAafvYVdoc+nq
zk+KBcFrT8dspK-RDbI{iownQ4cKwAKj^=pm#m9X0W=#gMT4IXIkJ*A^9}R%;+_zxb
zvF$;DV$DS6S3-yOOKWJ}@O0N{pf0dCN?InHS-EB|CyklkTAs!JF-+glocatCdEnw8
zhey5rIsV$&{z63FR)#kHVS*x3o_mz2L)NTgXWq6i%Cr}x-Xd0=D4Vp<@|kuU%b63*
zNd^1E>7t%(5^8NdZxUjmZ6jZ9kvwQ$?OIni4lZcz<gb!Eiz)VM>9Px?3OF>+oW|u-
zM3Dn3+J!_8WHz5z8lRF^^nywT9A*yTl5VFf;{nf<`;TEWmc3iHDqI@Pge0EIF@zzq
zy?>MXZasm7Z|Ww4F!o=r+)3I}1kIv&B&<eGVYdoh#4E~ZM4Nk2aw+yhmktd=S<Fn7
z<BgN>#)pq_o|Ht)v%|{Xl(g7YW-(mq3E!3PG?3*N(FFl8T`1dsY{ZX4gE)>pymtFF
z=<_PM1R$M;TSeG@q&MEGG>(p*QO1##DzqFFcdo%KahGRHP?QYJOZwT)>ywvO50r_=
z<MQFNWaayd(gaNVd5;V}KaEIaEDrdvTI)CR{N2zg^!zUZ0@NgtPQ^>76+EMagLg`!
zw-7-HX{m)WKH`CHJaDV7r~_!aQmGO@8bd2_Z2hiHtM}ZC;7fW-`;~f|gkf@Y2Hu}O
z3bvnU)iB0D9|)t&=H;)qw&_I(2v|rN5GSW8b<*|svo5qpXBy&5T-mG5OY&<cRe<|@
zTLX&ffOetv4lWkw!Om!yPbNqb@i<)DYO~=ckx0AC2>6Hoh0!59a2XhJ`FaHql)o4y
z5BkH%_0web9zAdFy%X^8wzB(%>l~<<B3SNA7(axpnVKi^{^GY?f$MJ#6}QltfoY1U
zBe(!Y)XJSBd(b%(9de#~$4Y$12DRWU^_!IU3wY1=0hCgHO6GJ)$&bUXmN2&^X<ISf
z(svI51kzG2KHxzI(-M&6I7nQHWtj98Jy165Lx^trQ*4H;J(a5x;$c4U6H)b!D)$Q9
zJSZfP^aZ1=1FatTitS#3{-hI+rJqXSEufEVmpQf24TK*CT`c@4%@8OujvW#j4CCFQ
z#PwHWm0Hdx?0?x*doqJ^vSO4g5B@jO5~*n1;0130js{ZuKuf(pJkX)b4T}{(^AC5!
z53zr69XLE`p^T_$@n-gTAR6;rCpm~JX4NGxBoc~Jg?bD>Z)(gNZrT0D__d@s9|Nd)
z?AfQGZvKfIZz0|<zg6!nw0BH?hM|J05`2l#dYFke6*S~BaKOo*dg|*SSr%BDV<|Pj
zB)|PNFrgn9yMC}IDv|MQdg7tT5Hk1L^pZG9;$NQrpJhz>Q>Z##H?0AVwS$zS(sQYg
zi3myXsp3TDjIGz>)Z1lV4|nO?BavDA%FiD6Hkuhj<<+^NJ;gT%%3oDipB$n!4dSl}
z-h=~In<a`wA1G`bOcIsx8HEEFvRh4Fbr;q(jw6L=Bmu6bE@qi{@mFv0mt)s|y)w+{
z4xF#;X!JJA@~c4Ex9x4OLqmrSRCP3@#6nT7(j%7Zw*-k&y|8tyu_vhi9DGRUgQo^+
z(#jGlb4fljhpM+{tRE;vc<Z2&v;;ooGVIr^Kh1>Tnbb5QBUAX8FwP!u8vK*d*LL+y
zb(;XS0o+g;zt1bPK{Q*=53-}`E$_@5Kd)sR-%D#3lG#aH_J}9|j%b&L*#Z_N>le)C
z92pMd1~?Q7h)Z3`bQXl)DE3|H%)tXg2(oLno(>lR%F5&gSu^P|_eAB<^HI|a?xu|J
z;F%DySKdZ>U%2|mMy|C6whK$@jqhZ3hNMGYXi_ya)NmD^vwuhinv=&sbMma;Xkh0r
z7r;tp3%;WqU;mdoaiyA8kXj^HHbkV!h#P;TUAv=KTVZ9P?Z01C&U0zc*KF1607|0s
zV1NhPZI}L`Zy+=q=34=q@}6iw6)1xWw+5byAJ#?_sg^15WdEH!j20uI)K0hEGJ|U{
zl9on4>~?zlX4Zq5V@V8`oWD6rvHb#)8BpFr%{-V~8NSPyY`f)cwebE&`aO_|gvKg;
zA4;J7H9MUCSYc_MmE;$L8;rZ<ymjh)BiH#iD}Q~7<^f*RoVR#Bw7z%~y3z+B3bv%U
zfBn9lUQ+*mQ3~~6G00j}yw*7rqEs-?aiKx>!+HfeK1rg(8@+iO@v&WM?Ujs4h;>)v
z1^yT)gI|!lnxo`26VC-FZ>bWX!oCWbrQRP5&O|p9cCZQfbRSVPMoyChQ9TP2Q_&Yp
zUq%_F7&*<eGEbymgLc39P#5bqDFW!9!9b;PB^k;NhbKLJs^-`Grlr7S7#^Ze;V-Qy
z-r(OVwK!L-gS(s@k~OLh^vwqIMOZSgs1#ss522F79W=LOvL`~-=i7rp*^je<m@`6;
z*9T++<5Tsu!6Vbz1ND|&ys>Tr^BO=*CzhL;Fgy9)h4Q1|Awkk(<k5|FM|bt6&ycb!
zce%vU2(UM(llUsjUeF}{RqjFD5sgwX-q+U0MF74sgIJAReSW}!4MSDK)g*?x0=Ib=
z|G`#vn&I?F(KJh1UL`7H(>a8)6o8l%-nk=oY#15w1zp8Y?7Lm|<h=Y{=c2pe{Mssr
zg)VGuliM<tjg$Hh<sY1Vv%aq>p<&bTSUGr09yt((Ziu^%-AA);nt2qDoihA@94bSi
zmAL8>ABSyZd&;Gd=CuWm${;{csI6;+y3e(cuEMeNJ*i)9@NLeAMofGI)VWwNFvrhE
zorcOh!c%KIxD8W6R2tGTQ)2y~!peYxOy7bgctW|rXK{~i)(U$B>h>F?hQ&(^<*On>
z62pi}Oum$9;Q6&{9pVGuxF2g@ALX_-PdvGh9xJi!<NiQtJcKUs)0U4>kHYdG!<F98
zJhAWvi4<m!*1Zb#qW<rcVC{hwgpx}`-*aQ8430s`v#ymZN{MAZj_1XWph%^>B$gpl
z0=-7!BT(_T|0eTA@GUp53MHzbDm+A3f*CFZm-0_~Nu9$u_I(;7pnR*^@vT4$?ftye
zA(9+GJ+-ch)%SVS=F`SUv;d!M2FfKd17)dE^W<Ro=k$}IT*fqPe(2Oqw=#7&A7eLQ
zKajWJs|-&d|40VS>VK)2P_KBe6heqS0!ywy-Zy=%<-i3lPOKsf@o|*ypHh)Eah^{e
zW`3#;`^KqekJ)}U)?kE=V>cpgnYo$m0|!-Mf7tg~Mtyn{YQ5sOc&Q_KUSnI~xEb`s
zht&XSMM&B_c$3!|fjWy}6x1HH_>xCa?s?(?ZE<Hc>NYn>?f6bXag<UBB<*u#e>->U
zyyGTrX5Fyyw=kHNg(8$-avoXZ_e9gi*ZL~r$sT?V11pc%?KwpP&63QSx*@$ou^1?i
zJr-?w_13~@odhwKW?Ta`iCBkhJy;Q_Mi~$p*I!c{ZwM?tuaA%Kok(o6;qv(QbTx5)
zm2G*vy}>u{s{hwv(TSe|$-ID@+q&)}S<6p)c<SkUrgKm{lHfjFv&CT6X0yGKJ7cAk
z12*6DDnRx4_HONO`}&4yAF~z^cl~{vONH&A!lTG9dm7GP3X?DsQM^xwu4lIZ9${_g
z(YX~RiiA=4ie#BUhf5U4ZlbAmeH0{zCS8CWrP;pjoiN>(>Zj`lArY(7{l{Lwgzwra
zIl_^BI003O)UaV>!_NzPjVV<Z{5`0itOWTA9XG?0&Qvk4gPSv0J(Wk&eEKenhD-Nj
z(z9NrRx|!0eN42MMhlT;1*4MCJOu_SbjAw>ZkM%s-Lmvg7Q+L5A=fm|4G%|q&elb<
zMfn)Rg51vFqmbT|e}VkpYRd+?0XNBjcKp>8P}$ZB2#NX4@j7mpSOHCaw}B{A-y&7N
z1wCFy#gebCb?A#z09eLyh61kDPL{6t1_4q+kZ_}6X1`m%?IAPUpyw9a_$25zB9471
z0T1HFS6)4m7T)St-fzk`H>AhrT?x(YcR)hQwh(J~*zXDcu+K~wZs>J;+7-yH>3WDG
zP@W<FG6k~fg$DjB=I!;PVqOhouO7adqL|y!v}@P1tn0vW-tmKtdkvaf{j=srBEZC*
zdza2(psr}!D5DG;z}O4no|=HOPrz&3tFJ!{DY}ySYVPZD%qX|3;Vxxn)<HwPU<fud
z)i3<;bbJz^m&GH97iL<W;ihMh`dKU~6{bPsH0k%Zc@mZ6l@YHNiMQJ&Th?uup^6$#
zQG>VbR~*0`4V}6@GO(1?BKuP2b=cSXh=jPVXd&#?VT$d9r7QpbN~x31$dx{Oq}}Dg
zA<n=TO><^%P$6QEg3xRKcZNUorzEa&GAr|CkRumb5-yrnzfe{M1$XUYp2T16t@ao_
z1o#WQeW#YxzO7!6Abyfud6>qbD|s5YP!^Q%LCD(iQ^?)YY<=d^#HG(j;}3qAyJxZn
z*+i56$jUSg{04TMtnD}F(0g>+T)%YVc0^VR$=T%cq53~CO)+&XxCug^1oXx99b3iX
z$bK-PW@{_qbF>;YvoC(Wj%QCI-FKC!T~Gn=ySWN1^}|_04C;kndLu*`RuSgnfCO{x
z56VBmX)-{@f^@Qw7JHr&n^x4m;M@{4S{GdE7Qf)_PfpPd8*j+7ozh;9KO4FSn@+u;
zt6iC9X3P}I^MjI!mY4!|YivNLF7ySEl3y{W1VU`nt1Z23>zEt=Fq6&gvK^6G0@r0P
zXuHY0qs(^VSt-!z0_SC5j}nf4SV-R|suE<9#$rXnrptqL0nKf~yQ>f6|H~Aqj6wgw
zb*F!JtJ&WCfrVIKQIu8}{SGlldS{NP-0MWv0I$M@Q0)NP%Elt`*oFIE_A66AL5jWw
zEfk;ofbm8CK|hv7mA>(M%Z^#|rr-VpS#-)6)UEYv<)P8l;RX51Rk)NtIOvfv80&aF
zXYye2m6&~_qw`==HJY@w8Y&f_v7ca=TgM+?+0b3@fd=?;3*Mo_4e7$Wedb8&_6#Zt
z+nO`b<!_foZ_w##4ZyLNq%^4MH9>!$k&VCrgJWGGFv8mCIoBbUzX8fyDGq-5gM=Fp
zM0Q1^Q~F2sls5*_IHuX;z6D-KN_#oCb1#$@LHkxgdptDY=b%Y?;K0QZPb9dngH_Rx
zgQq}JUqjbF0=8|<AMd*GR$lx@ReRyA{>J}i<jG#5{wj_E1NeFrGKCEgx&7NeO)N5s
zjUi31T{DV~>(K)LKH#3`SqFXbv!>~~Y$c2ia(!~<JzgFZUA^y!<$Qw1P24Fw&vxDW
zraRwRRa`%7`TMw>ztmQK-d}z9oD+z*N#gk~ONGYr>3lVE+4n#$-_rSx536hI`1O4y
z*Im`m3)jE-enLz?B+vEpA8`L%2fd#5dScEv{5;lbn#fA6iZgf9bo{#YrY9%e2z7L?
znfvKSesR9U4GNXYzjoIMokj>I8yQn1`UK@4;L$K-#Ga8dSsEzd)eV&?{3e|!IJMyr
z7ET^D7`d9FM~-!WSAhan5xmPD_J-Y~Gezdbpb9jhWAQ4n_Pq4A>e?vp$d)bzkRt%f
zSyW3vlP;7sNwmYU?=y8gMZn+lb~&GIt-Tk$sC=i&Yhtn*KlY5ssoKKpIA2rLz1RYJ
z)n-f<^4YYI)GB*e1cwzccb;Zrp;P?+&i=39Zxx16e#^2(hjb2oR#K_~f_d(3UsApy
zqCunIpmAzBH4A0l;R}xQi2tuA_6gB>uyS~O>SWGoi0Owi^<wp$0l9p*RpziBVixni
z+IGm)(l)k?KNyHkarm%YsJ=^qrF~baB7=1Q_!P9SzUiXzjV_&FpN#BmX$;QdFbE$-
z`6W3#i%`R4$pHMRPy^Yw+=4@0v8mrz!JG;z%*)v7%((N^1c?jm{e;JN|05&9^}F6N
z3xo;B<&zloaX}ppqqItoku6Icyw`60)F8ENm17R?Ta~qu{?W&gEKU`uC`}vuGUnFu
zyK&p-RDQ`XxzRq4ygm`2&S#F`L8`0ETb`<SppxhgwHJ<y!~PvTF>L(-5r@2Qw10eA
z5V<&sJ+7ZL@=tOStg#3F@oPRSw7j%F;@JBl(tG3rNvJ-{SaFr|%p>8U`B$Xpy#KTp
zRY1m9MM8<s{cv=^)Q0O<td3@)(hxh@xunDif&bx=pU>b`*Sg^)!;9*p^Zq;Cb4>f7
zs2HnX3C~AfJa*q#@BEGL-ySqcwqFi?;utZ0Mx9$OBMTaoY-MS_+cL?!!$W=E<oayy
zeCu-_^^DqfOQy5(c%0OcwlmvyV`?RNxufXdK5}`+6qIYRqw(PN#H&6NfFs1E`LLA#
zI_r4J?RN98Liuxs<h!S%ph$_+-#Z@c|3XzhS8F<hW}>~EI@?vc0zqR|p<D2b%xgR#
z_Z#S?bu_8@WV5IDCohhkPQxVrr_wffB@^wiBpD{GOp;rdl`Q67YvFnkJJMuQo$unA
zTCXbtakt2_DtW&dMXSy*uwwqe)b$)1n~~wPa+-^aPkoy||4>h}xq-Bb(H#Zw>=58c
zk#Pgw0FlICBmxh4dSOb1VO*M4f&{y6!7`Oa&==h=&Ah7-0FNtM&0tV}pb1@ABBd03
z4k#aIrK7Elau?)(BK!Nlpj*T6pO{-`0G4R|cviQs@`Kp%)6VzpO-+~I{Z?-@k)h>Q
znfAze?+BPR2keC~#$y}i{#1F68Vt%y%Be`th#+c6edNG@SzgSe@p;69(%Udc=pZ#B
zbePxKtX+yLMG;vPlZHwt%S^L)6CcQ9U(J1#sX&3DTn(p>m6*CuCfoLK&p@cH|BSo)
zrWRO6Bu6JWbGAy@zd{5w#LW@r^ub|xRfHJfX3dQXGV9AiX!vL$30eJ-=Zx@)hu%2V
zPEkwG+P9sq5q@_p->YpM*79DPUJK{#vv;=I%!6g=9$(TeF<5cN_4({lrG<--{3@o^
z!YBEfp{U7{sPjIO`K}Cnk~aIy;hYsnOFgSX**hzChkKjGXWT;{K;=Vkquf{M0nr~S
zCl7T6_<g>qmpLrKgu$UAYIqNg{!EeqXStBSnB&*t)qkhGdpzgQ@+yNOn-;&cc`2Wx
zcF8_=&n9=lMUAh17U=Un&L6i{1Guo$y%KTP#VYq*e36xCsnu1GztDyt!2JBSX9?}P
zdVA5ePQ!E8yqeGz2fVm5kLP~=*z{7o>k_3c`Co`J!ij9!ih9gei4;JM8;ArpMZ_t>
zaX|(vFXJf!4uJ|>e8tc31zb#ImK{8D)cc-UI*&O7Ro0)=rU2xCopxEwyWK={Y@eFS
zPL2sQv_^Wp!w9vZ4Z8aAm;4kn9xG=($MQLV%jCzKh)Bu1Yg#>G=ckbu%*auxVK0;w
zl?i}LQW}kX-$i<gTALJ=Eg-HgctJi`7Y#_`gZOU)ek(@k`(NN_${mVHATuBW>b`|m
ziL8Jm{=)iOljQxJy#4=eJt|{Lk>F<dhXRj0Yg~JY*4F;Wf3jhCx!ll~jV@>iIhYDr
zC}P248j6fAoGw-QwZ4R|wUe`Ab*L$-b%MWz40*xw8eU)nXsb!`KhA|cCG_9bpKKqG
zOK8m`Tzz@)hZv4jo=!<qs_G`E4(*u<gfmm9lxoFa3*47jd(EGX2A5t2&sGsZNEv{n
zWk$cR+;q5AX?R9%40xXfJv)kTRd3Ow@VQy=)#5eT^8h;YRb|*YJa%Sfc?PhM&g2s7
zyhC{Ee=RxsZ8>0mky$6f-7L|?A<@-m(7np*UV2Sla(zR(&&Zb4iU+ruH`wUz;S4HW
zU~Y9>-^6Y(L7TD(HhD|AG;W|!$Ux_9w_{tiwO{jiKy_F{>vEOy92*1=3KP><YQP}T
z@x>!VjmduDz(ti6EpraEgfdF8Q7u*WrG81FxpX8S{J~&kmCxFa`%CUD@vyct%GpVf
zeteI9)*LiA;(4k=^W^j3e;=P-EqRAY8Q6lqS^YmQz$I?t_uLYIkPw$+)K+xqtNwuT
z$hWHKdfDgI*uLzay3N<6c6H~qJbCu-e=|l)I~8;D7zenzgI9Lcs~q~|3VF=R4QorF
zw~8m6v?w;{9on?Hu>?I5E56mZ0M_Ej-l3fI1L)Y<AfSG5P|hybYjv9wk<Tj3helNe
zvI5Chw5*$FDvxuapP)tvV9!e-akILtvxJg|>9I3L^Njkig0w0t3PSH@PO$*jHgRi1
z*Aa_(@x%a~F?%c-<S;aTUhYvsxN6E?OTH!=%on{?2dL0qArgYQ#|flrb(?s6dwrmC
zGUnp>OJH7QMy3cd5}mWQQhl5MB1nVFGE66)?W_Qh1)M1f<d;z|kY)NIQvk%>$_-FQ
zEKOGQ?tmdW=55EddD{`-+1q!7NmqabRE5QrZ~G3Td&4sQlt-_Yxv<Nlk&@`s*ki3H
zLB!SY3uZ<w06MPiaM1rL#fg#y7z@W?12pi4^N0mhO{!*Ld-f>2QE+IwaEhV3<ANsh
z&|3<Wv>`meq!zmY+8<$1E~C=Z?n(%d83YD>$x9fl<<a*$kQtFPO5ZuWN&nXN9ddeb
zb$@$?`-t7YGs!+)CEo3Q$RJm!M=e~;8*mTTl}Q9jRHQ-FWTo%(0O=|DB7zyR$9zH@
zMN5=Dj7<FQ>t*nocnLwVlA$2&do*gSSyGtS8wdKer_fdN>fazjW#z_XKsKo=!|gp8
zxw(&9&HC^t&`mG(5EHCsP{!~Do4jJDh)jRK+x>P#A#d@4Bz7@~xcQmg?(ks*JR$aE
zz-50smhk-MT*#m8I$$H<!U*cyr~bS#vQh)ouDiS5CFQX@0qQhoK*OielX-neOZt!j
zDs0Eb{he+)x#DteDaBr{f{4@G0Z-K}oG2LvipT<#H^Do?RpLyjEaf_?B*2q$b%GEe
zE4I)b$;4g(tS5a(eK!8<>g|h^2$+o9338=@m!R~U4%3VB{GGoB&uoBcWs{7ulH*5z
z>TtlI2-AE|X$(pm-JIqM1k6Ckd592%)RG-M5hdOG$R?};PtrWOaT@6LR)x<QlU~`!
z=!D$sVjLNdpi2V`OR*;Fy+a|xT$umB{Hu=Q%3uFw5$xWWO`^9`)nJsI0(hB>NoJWg
z>_y0PLWdCU0(;y3y@Yk^VngTsqTltwl;Z0+VjZPK69weh7mLARQZkwh@0BRE-uG*0
zfII;|Nme3V?v{f>P4w?AkQd(zxj{0Y<MhH_nDQiB$v3Ebsdqgd9XG6QIo!g<rumBW
z|8Q9GlaD=Wzx$wm-7z74Eiuxu<JH-9>^0?mwG%NcQVbN!%p{n;bc*Em%;E5)&}7U?
z6q1)LJ1@zOMGjeV;YoebMN?n#2%}6o6z6yCV3=boONp#emwD0^N|qNrqL~JzO>dFs
zc5UL72Q>B+WiN^rb153}9sY)~2gj%Xfi>8O6|vs!v7$4jnQIDjez*9fC0B})KM!av
zwLE{y0Klkh3Otve`BO^#%DqoZZu~+^bw;!(2k)>$8`r;IK3oRoIFT=_`rID}-NdjS
zAsPnW<C;_{K*5q{K1x3I<OK=T3f_?)JFsUL5eq|Lk>6w&4(zs}XQO{=m)|LV(8kAh
z8<-x<oB6(7n0&xN?rA;Q`&KEzwSS{S8{kbfn0-1A#Yvof4RZ@t!xyU+syv0&O=|J4
zoUc8F42uBG^~O9k);gW{_n+PKvPq%o%u&~QZ;*e^R9EI3qf5^BRU59g#IjoLAVuL{
z@C|^IXpMo7Ve)JK6f5oii!!)l;PGMuHh!ryI5RahbvAG^WX~tzr3c96E;eHrehx=Z
zeb?7&3hi0q5sIMKfKd0>&ay|v#K8n?5GtwUwougH+5hm1rq|Gh!2@XVxCLuPg^I!4
z?1bovo&CL5NoZ=rHYv5sD(DD+2k!dh+vZPdWeLv^R5k@Td(&lV9OG<s!(%MGLR$Hk
zM|31<bh@J5J~O@&ec|9qFEU>3vikf1BP2b<h?T5ZW2lXfkDu-iYqd!O@Kl!n(0<KK
z_gnL9w8czEv^~-Pu+^y1#m4r(>lSmG*xSH}<Y6j@lf4>f8YQyyT-s_k094r#LTK={
z1gk+RFiz5#e2@`e7OygXtrhn9Qmwb#srTuROMu;bO{w>zsf-^G;bLWeWPrJzuN44^
z22BF~(1Vz7xwMLv0nca|=Jw`;al2w<4y1DYR2bhVQLbzqcV3!W$Yt=dQlC7RsVO_=
zzCImsblk{X4w>N3Y=9%|;gHE6A6St%QUNY3jCObW=YldS!1Q74+Yf%uc2h9vB~q8{
zlxpX<fEuG4V(Pgk;=6G^MIx{Y<oXwUQ2%yrB2nU0^vjyF!z$#S_08&u>EH21#WaA6
zq{2pq0DR2%cYi4Y2k}V>v!`CFc?f*DnMXRIi)eXg|27#Q590T)r9)S~g`nb6MEBS6
zFnd4zI!lj(5Dra_;>&#HrwzBr495J87Ab^4_9fq%<%C}9Ju{2mw<i)0D&u%{YL+*i
zA%I^x3Iw-U848q*Znli@`3AvpbFnagZ<WFK%S14d?BX^zmHYV%;)$k~CfBCM#>OR9
zrnIecSHQG;srdtKiPWo@ioudC7bpEO0hp*8BS7%chm+~%X`{bvElg3Cx4c+CEQ@yC
zbNOzS$W%p$F40ZE#5ld+TAJ_b%4n^<knw5yReyKGXDN<q+Hq`N!RUxH;_*suEzas~
zZ76R!5`1`4g{qwR8=!JP(h=rmg?$dy$hJu%60@Odcy*HJC*;2JbFy`vE8$CLPpBAY
zyjN2pGe48%0Tg`9%!$=NQBXwjajZyz2={nm{(zy~gD>c)3H-m+X|sUnJhCrS!I-6~
z`u8PCA$XIa36h#V?5lG|D9zqWhe+SuSnuFL025`IeIsz+@Dg4WeN~<kJ=I-u8FFaU
z+Y8Kk#ju*vwQx)T!)UqseKQ(wgY@+lv-nlxEUSOVvQ1R!S8z}aj=t2!E5=YJNCYut
zFZdLP9mZRAC}xtX4}y?+x)1fqX=0Ma0ee7liDt>2vlg=!F;7VqJvLy7#1>NKiSwQi
ztVyPBDPHbt&TO!`m9TmyVPr$yzJKqt+1b@I6Sbb38#ZQyVLP1lC~0**8-!<TEEe!C
z|5gk3JmBSeMG=_1e&31Pk@~pON_@y0Pn_NAii})_ik|yvyv&P4thV`U3iHo$3vHBG
zB1JOL^n^YKe_DgsavN33h2rQz>G=BOlc=zvUZWc&EH5xwHth#WdWs7N_MoM9umQo_
z+2T~2SLxOD_))UMyE82fPn<2jBN^v%$X``;Z%USlvL9cMd-AQgR`lJPs=>wJ!@L!g
zP+x5$m-<>m(QI?}D#y;8nZdy8(u5hvJ^72#+nQ4Ejb>-xuE`STjf?fMCQ_v0=Ah+w
z&WryEYPCXqlw%zBxm#uqFVxpmxd0l}_X#Yhg52K~b_sO|Rq+KIfNKuWPGe~)CE0Oq
z3NxLDrRa|>iO77$6H~yP&=Fd{nt@as4VN1YSGkftZ*Y>rrB#6Ou1mLdA&oV^#pX{&
zrOjiv(ip{*92>%rQ4WC8>kii&($v9tXDc3=mf)$OxrSEfU5%p8I=5>ThUa}&Yz70x
z@1&;;*G*1()6ja0L~ZVAj5VHO@eLH7Uew=+6&0Hl6}eJ>k5z$T0IuU#tnKa`|6R^b
z|F;Jv(;Pkt%mOx!j{1iTRYd4nI+R@OX&%gUnL@-A$bug(@a4E>s;rXj0awaQfjl;N
zbZ*qTwq!wyqCeUwg#sKA2b67K_?eA~C^9C(sC4dt(L(dR{@`iS)?P{kEv;7ajEt)x
z>zk%HiEr;3O@3+16sE5`0;ejY4yG|mKG>*HwH31I6TJxamM>Y~8+HwF^0v};*A#Ym
z@oMecC(2RAoX6NWnT`UjlD>o8^dHX$YxXynfOq8&FAN^L-<pYbN?ac{w?Cgh-*p8H
zZ5%bAiID^yGwrq?eyn{mqK)VTPSKB_=$y!`1TA6}E!?Bt7F{w3aqgok6H6ixTnL%P
zQY*N2QUj72U$k&KSoTbQ+;ukL^j9qK_>b|R??gL#*c6X4_B~%zFJNZZdaND!2Kl8s
zmF0Ws^;0bTaCe?L4%)=2qVV2Vb2RO_xjl;@KA)5H?|U&duO5c7y3~}UK!x@CP??^+
zKp7?_VIxv%3By{gfr$%m7%)b-KF~Gn2aHE)bl!y+bq^XG4^*wjy%=tORJeFP2~x<s
zC81pPzt8=1e9L7+{meKusb%cqH@`~nHyG*k7eK|p;7&ba=p;y?fC%;65-kgruHbj{
zp^=HEh+B#NocFWHw91}M45Mks@1+h{NeU}WH!^iA<na`uahL^*9|oFjtw4HmNpBjo
zLW};1VBo0ADshiQZzq;bg-h!b{*aRG{DF=CiP-M|p=(;F1@gVWuvs1~Q`<CJ#S$qN
z#S)XE-&hz?_#uZ*y)fD$Pik|ja|}Viv6n7v?$zZljquC($8u9Eq-zI4z)Sh4N?MWG
zu{*4rR@abhb~F~QDp``l{4+`lV;4}uAOdk>??lX3Ziub|=k4D*vnu5{`O(ga#W&_y
ziJ}V?Xya4)UKF_q{!aR#?n8i)s2wUm59MXW-)Vx^d&I`cy##S`3}TMhy<(Dt`q@!P
zq4{PpXaSv{+oDSIV#0RI>b~(RQ%#Y^klb?9Akz~S0X3MCV(M2#22L9XH0^in!x1l_
zC~B9Qf^95|;_qHSu49QKV){5cv`<Hi*Rxy)wlv0=ysQlK$*XXIx(1g$n^f5>&RQFc
zvz^}TK-lic2h~c0)h$eJ*5L78Z9)0<0ba<D=ali;+DaIx8o3-xk84A~EpPj^WT=A?
z9T6pMKYh)8b$*EG(^kyW{C)RyQ8XJo+tsnvo)7+`DJ&dY%$IaEhaV-}!r2MLg6%s`
ze`oFW{KAVSD5QG6h|9|c`J}F?tG7(V7zu%u;t;nYs|V1dL57RGDN&)@wF$2a(H13s
z^J5N!0R}>z<XxXv62o{Jdkl(v-h{6v1cuXJlbDq~6F+rT9~Z<}6y>}J4{oJot182y
zC9={jz;!|&YmLW!CR!1itL0aBMx0`T+>aAZdnTSX*{F+P;(<H4K{@x1Uo%3Et9YKa
zgSv-no=@10c2YRcoga;DLpHOzF8G$NDpzA%wk?}+W$2$4BxQ@6@z-I%pX;{Ip!r@D
z#EP92eeH;TgZExhnZiaabwz`GV!d$vOmA|8q3kD+ptF$Zn62&9{<e^wPC}vUpA`X<
z(z-mCQv<*rLR&+G8ta~Q_M;Er?<5!1{q1a$KqyTcR>a6&wX=H2esf?t{&d%ThON+b
ziZWrF_uch&%GeTAp=8m%O3ej`%D7HgMYg)~FsKa-Of~hqP+0qW3Rz?Ln9E(U(J;!L
z;&9Dhr1`!z!pB?~LeZ#qH-y@=-pdWvo3ajdu*PoF>Z8@datG|GNp$!D+|1Dbhi{Rd
z=D%F!-|}yr`cLv0_E~h+UEVTMa@FAEm#duQ%$Jh-Vh1s0rJwZN2Mk#5;fcYmmyl8R
zG}lpJFX>yw?Hte&>o7INp%@iT{m4<&B<lOaOSiWZ?Zn~px45`5xG&4J^J3r<fir<s
zEJPbq(9c%PY73Mt1WeZvXB)8!&rCD3FvkI44lF_#toZcF8cLwR_|Hc5O3cU5Ens`M
zBw1Dk`Mm%R3v!w}KQ-&VQtH~3YfsB38xFmp3ks`xK+N3Fz<BKMbeH?|{X^RWRccdN
zj^2?TTc$v;57nXICX=yYZGaLof5J0DplirMdu@BWDBr;tmS(pn(~lVuuJWp2JiTMq
z?WUb#2R%Kxf`Sa2M*PIfo4A_iGXLF%pq#i|N$(+~=kdMg%%IL&<JahV`>*JUD^w8C
zln#qD_xIq@fz%|3tounaCLggPF%ghM4B1kXQ=yv0zbB*Rj@6!a{~t|X0n}C(ZQW9|
zxVsj2cPZ{pu~OXKU5Xcn;1&u5YjJmXEl}JE?hZx&`@Q$SJ7Jg!cQPc|=bXLPUVCjZ
zKrXWf6_XL{yehz4ghUG68`hSH;TauVd)r@scuuY(-}^aZZ=*;Nu`c)CFDA>Xn9!?y
zk7S%!VSkC9I!eE?07HactmT&S<^_rT;Y2mLT`J#armMed^<GnHj*#`snT*=xHp|B3
zu1nvF=%v>*TmDY&?bLNQ!Ykh|XEX=q$m4U!bJ|;s-|J~Y!9E+By4wIzLSUb2UF(Zz
zgeWtMz<;@(A@Md>=>9TfkA+yEPhvwYJzinR`>Zs#qy~J=#BjcNRI{dPlT*kW+)-xo
zJDqdb@Zkq(#}+hX^PfqtJ{_fI$aP-^jdQ^Arm!yl!>T&cdqB*aNVkX~M@@h{gxoh5
zmG2&G)ON$y%pH`0JeSPoC#p^TJo7aO6w^WGRf(@ft1s;iI!LyYM5)i6OG_1o+kVSq
zCen_EwfVpbyEo|Kl2MqlW9=wVZVIH1;)Bh#^o02H#K()-M8$(MNYhW9E`V^TZm>ou
zCj2^(_OQ6gabi<)6XLE!9OLqyg=S5wHtV-Hx!wPVOFqf}6Y+eUG!MVr0(>*K-4JLH
z+!T|fR5gx2l!X~4NjbGCZAlpG3%N23yDDQiTNr2(4W+tlIy+h$G^-D1Wp-auQKTAc
z5ly0E`6D7?;&wJPP?-JEqkuP#V@kjPe<A!zL5#gZE+mgqr2jWw56^)bi=~;@Mt6kg
zMXQ#t%ID&zsWxGwFVei9!hY~~k6=Ui2GsFf+0R;+`Uz6(Wz2a?*>%%%t4(XWN@6~_
z8@bt5214MLQ=#UEvLGdsK_llUF_zBgzV*`|gV+5o6ZM{}4eG}2X2OKi<Xz3f_aNAJ
z31XdBAuv_wHM?$FdaM<e<HQLb!5w;iE~ImdnV>|u!A{Dg&p&=Km(zX0%e#A6d<p_T
zWvjY9UO7S0^30o1gz252s(hkyEv`bh<vnb3iw!XYE5b<x`)Y`=>`i2`l2mf_=Iz#>
zyX!r}Fw{7I1&<BZECciWs1bDurm%C{7{qSua=Nb6c`M&2JC5F|qX;bfqbGcOU|RLL
z!&#b!;-J4VzCKFkj5C4dC$CygyZUGQo){D*7y1;)J_N{o(b6OxX9jT7b2k?sF#fZk
zz-zlU+Qk3nL*UVN<f72^`$^G%Rm}m9g!P#&Ap2^sqZfm=hjc;~7*SZFdjrJvaqq0A
zK!hOYy%$K{KU{5eQE9PpYrz$gTpx1TrXViAB4f|$&29Fg<d36u=erjOVltN(RY``g
zk~_~B8+rpdZ~CK6fpn?zgkR!ac6HC|n<~=M(iY&!VFsmfo!C;9vOm+!3)iq#Ta#c_
zi>~*b8!$YyUW;JSAiXCOVI&Oi{)9l|RgQo2<TA(nu5P(js0nC<$j5ECrPgq9Oq`a-
z0osIr*)8P~52k*@Guf@xWyVCLjoUwhh$HK=Q74p=iMbqF{p;z-+YKn(knmp**-*Jx
zdEF2a9alMLX&0Go52$nN+t<PHw$*ZTSiShu6Jm2}W1!5I8v9MUn5G0?0P51$7s*qG
zqhyzGjT;xKPKw_}{9kj4y3I7_dqA&mBK(&>7TKI=J_eR%>j;uxf3YJ!(0-Y7dWgUH
zwFB#V@bac~uepBw=s>R=@-w*~<~-a)ro`dq|8ovL$yc~sny-ycPrT&~RsUQ>uaXkl
z_lCT=k{CyXRBxO8I{H%82i8|F|62FtowJt4e)>|wQd0}YfYFz)m2#f)jRZP}8Bv9>
z``8!Uu}5TNgWr*NKYVV;L5X9W@|@BjlC?1gdgyVG)HHZ&;I1It%^12Y+_1s7#<2+>
zF=;ShYPHmPz%I}O*b#P31A^jmye!pkd6dCI^R7zY3!zzq!5Zp~*a5+LnkTgHeKQN{
z<7Wh^gAL@lr@Jb%J3G%OCT5yHC*S;g@CyP3y<v;RqP$u(iD>_RUOgm_8d;^RW(SYa
zslWKxWYkrIUMtW(cT^LsHxopsA0E(U(1nD%oJ^s<C~;f`FZ;$OocozkDsMI0A6R9T
zxioJIl;PIZ8I3(uqE`#qJsg<P`Ja_i-_x82?HRe++%QB0_RFF8{Cr{+t$Cfm(Tl8&
z+WH<2`YXEmt8`H%l|nNFI-{1|Je)WT_<PL0qzEmv=8zX#BPB0SD{>pkfm<&sPS##2
zYo#`Y0aOC<eCPu;*{_GcbiZ{2X<t|1O@=&FcY>z{XGZ}`ODhgHisDO_x<w{;n(IG3
z`yAPb(rD$Bu-&ie!x_<rE8l0Cp6BPiINC2n+_@&?ZEVV+^~nTIBVTS*;Y)M(jV+Tn
zotb@w-5=Xrz5){$V4`B-E_VLZtj*frVv#s&5VE4Vg=jOMJvrjt2i(W3Z+G4({a3Y`
zce=g4gokwGHPp9@juTBa^RCuS=O;6Q1sPlHN$hgNMg>)D0VN91wi2CUl25P@F9qLP
zwiDlwgMhA}ffRQu9B81SLsM59Z&HDU82pmC1*yN$7$^(u_2jc~AavFyR;BRtMWj@d
z_@hjdK>9`4MMLv|VEk8DuMho~$ULEYlfakNpn(pkSia6FCWuw_<{Wyo>41m@i97a^
zxM)A4=*hK%%asC^gvg>Z-4&zC7D4@GYqhC&G_E%P_G2_kKrg8kX2+-%y6dq{rV2XZ
zJ;-R_q5uw|H+hIA2L7=u1Otd(UPLSgc2VblkYd0fEgqY-!ej9nCdLcs9Prd~@z5wN
zGNwrgHv#NYmalJ0dmcVFtrXE4&L+q!{_i!GPGiIYx1JDqQ%G=qMHbIu1P;UCFC#&4
z>l^2fhE=)eylYKkDvVAwU~5RkK(v0JJ};PaN(XpLCKqOGuPWn&cJ}g5$?xh4v&FSG
zg4b;NELF{bpcvE72fCFV4Q}on0RAo%1hWrQgltZUH-be|%Az%>Ifs@i`#b#9an#{A
zVObcR3Q?;~FNBvw2W;c8l*>DArwT^IQEU57!KfV3QXogIk67H$zrA!o-F~l>&?mj>
z)$ykt({%wBXv7Xh-WaZG3+m?!cnm7mIqkA;o3gyWy5|*B%5~&Q!BYLnWSX(IzUta+
zjp!@q`$(2`oUppulT6a#@`U_48u=NK0*>IEFdc2b5LdzmFlbZwqlIL9JRJqOd`}6R
z8f949rrD;|IcVy9Dy&E4egB;NoUZypEyCs7!?7Xu@~!jy0RLb3${Sjq(uO#rht^42
zge&2%h1n3{X*rW5DhfDzcMX+$;CtC?_bA*vVc%L_q>#EX6$R_U``mz3QcVlf+4<rT
z?)HT{ZZz7>SP&?y$7*aDGKt67V;Jv8eKji7KGtLL56uix<6+><&&mUd#I)9>KcL{F
zm^H}Nx|s=6C+(usSBM41m|rZKM7adFEg)+lo=l=3L**LGgB%yLc~)D}l5Iu`T>kR1
zbD^1uM1fZ8z-~SUT$Ln&7o(Kot^(RY$tDs`?#oYlietx;7KI6Ko_r>4=^zHeHZg_1
z>0Wj_VWyViLfoj20P#i_&xFNr009B59EED`SOaC^qsD)rk2U+QvwxZonmR#4vgdl(
zc>%?7h1|gaM;e}nGj+d#&apPZKegIyQ6hbUSXLn${T(AwaRh1^D%-Z9@XcGJpy@m!
z+1{Oc5`pMxdq9U_bCtrZE!%lTp=Vcp0OINX1g9WJWF}uku>b;V=Md8Jfkz@*0D_Jm
zkwfukZG73Kdy09-ACetF0c08XX*xvddrCEkQ2{Cr_%^(>Y+V*&?YpVm8g}{}j+R<l
zEd#P0yWIUo#gqa$95-N?5C@7Z?pLM`okQ&D|H0D$q8{90D(-&r7k%hKr!pKEU37Zg
zbLzCMn!6$qg=I}3s#LW@!-pp?a5{r{{$1xfp|*GD&P;<fopIA@gP_zN+%sT!lOw{?
zHdtu+%B1FB25Ws34lW#pQMagVs&~=-5<&r^fe53nydCmW4mK0dT~j;@Qoi?bpBXw@
z(6CurtIK^Zn(F#ihF0=Om_VA&Jr6~SC6;w4lx2qvi>)}*E6C7J(}EY2?v<f4sVKY@
z-E0?^A2=++9ccC(kNvMmakYC+r$Y@gVe~oacBKj8%WHl!avtg-cRqS}uf=7lQ+ZAI
zukD}GD_Xr4N)+}?yaq%+(Pp5%5#Spj`n6ZD=T>j&LC?kUEN4$>U@&qllheb+t!5Mg
zYl5vz>aMd1u=@mmZpl&O!}HCQQbeNsNn`J!3HbqM2SCH-y`$1?_Bo)MS=J%8)kgmG
zmF>40S(aJg%{2LL$@z1n0qMFb6&y0dAMrols-(uHaQ6<L{AJAgNGkW5X9Si?4x7vM
zTaZ}M_9wR?U=T0sP`!{-#@~2pr~sc*OHkWb-s$*aCJ~H!h2uA^*GXHz3?9Xm6xXWA
zqSFA@CXG@=k7`;0P|-zUmG5yFRO^AqwE2LNq0K2lYJ6bv(vfu0qkjHkCIJ|%1<?S}
z!7xDgjahV*hyj1_8;uEVK52PC7&(^3nR+{(mi!u6oXLkeySz^C+l~_#DkWoanH_f*
zI6re$M1z4LQ43<~d^Jc|YuRn)7WUAn<KjZW3jSb=t)>8C6o<krltf5l3sLzpWzBY^
ztf<^(PVGY!T#Ks~4<Ky1%{;8m4jLb7X1BH22F`?sx4nYEd(#^Os_2QixPKjt6d&i`
zP&%#*iO7*VT5zn9CWF(}bD0q3U^RTfx2wQFHbd3}+9Qi80d<h3$Vl0LWcSQE#{P$S
z{YsMf#cbk6zbRZVsA!0iqV4`t?p!Qhykqe}Nwfj-M+)V<4pENb#=J5MzBJ>aKlWbO
z+CoTEN{m=EKnuvEVyqiLmnz0+DOI}tpB4bU{b%!&e`^#ZSuP$wqR{6$RiNkp%0E7E
zom6>2hNnm;zXe(xed6sj`{v{)MyPz}Wp0^)ru}+PgY0=~jRFh`xZv-X!J#erp5O)x
zaqC2ec|oKEk`zE-L7`J_{t;?2%|Ch18Es)b+UMs6{0*qqlsP<C)f$k^@wIDm<r@1N
zPwTdO{he2*4rx#Vtl+5NRzxO5h)MD0&Fr<j472s(Em1{(6mbq<3p2OV!74&Ykm(Q!
zKH~Q6z|>?kryX0PiZh!N<5y66f6%=|<5IREZ37DK0-ksvgxe?G@&M9*rU`z(WcBKX
z+8h3n2)+0*V1<J~D#ar>7$kHS-QEYqpa%3;_Mxt~n8h_6?aN2gIJ%90e1xZL?30D5
zknDc;wy{D`Xi-AGK5c}gv7>M*98|dSW=CG>hRiZcr2w#J(L%za1p4L??(lNREm2Z;
zSI8cX45tHOKEn^8fBT2<eAD(u9-_2#_Avy*AaBl^kgm84uiPEGIBvj?fCt<l!QlV{
zBTL0&2Xu&Q+Pgy$D8jl+T@CX15E+^sXm-~pSf8#V9Dg|jCBT}Fg^15UXOW8*t8sv_
zBTVF&@y|d;d5F1{LEXCi0d2HWmoqSWn!r*$oYz4(zT3ZSyJ0>*WyfBuC|Pd3b8UaU
zVsWpHJ$Pg*#4zg*f1g`!H%!N~n<@j)Sm7Ytp_NP*t-i}I^aGi73VdGI)fQ}>SG(rt
zXtF;c<tJum^KT}JB_^$Ixdy~k56U5i0FRpElK4BMT&X|Guzgz7%DCT*6MJ4)jdQFz
z{s0}O{7(D^o(K+|VdokJuvQ3!Izs?joqno&3mxcsY*TM(>t~aBKoILKLlBc_zW4Jv
z`5lz=iTJ=WT5GT}F|(%JO9PgjE6sooN08CnYM!0K>WQ}FMJUR=ki2gSuYpn%i=`*x
zd!S)(lP8DwX|ww{<J35Ow^_i?OXIlx__&;4EbjSPWgd-w70@U@z=Y>&B_a0MCOI4E
zqv59^_u;=yP7x3F_IEGq6yyy&J*@Ia@Od;=v83Kkf0Rf3$fEAlD?JKewEr$PQllie
z^)$91yumu580@@Cyb~qVqFZlr42^1bGN`a>{AiJcR<iBLD&Tu_<2GPonxs*!ZQZI5
zvJwAG84kpaez$Et;ZF1|c-#A|tOI%k*?Jr|?vRb<chS>p#M0(<4$v@@us_UW`U3+4
zN{X#`)gv2D2e)H6;uI-B7qh=cWKvCoCJYCAyVuL}>%rl0=cLh}HNKW71p1e0=aBD|
zvm(#4!a{zCyEwrrKnWHo@tA&pzYmUaj6C0X@>%D4g5ui?i5hI_G`_lt)?S@*v-+M-
z9vz6S%2V5g%M_E&2OXpfJko81ulXHPIeoLgsHu9B{d=3p#Q+p|bjkTENfJ&c^)k9C
z7GTK%3%zvxN+sr}>n!yp3<>7y1yPD0TLQV%#Zh^OvPDM_UVC@b%ljM6u3JzeHrs%2
z$JmOh=;S`{n@FwF_P3uUK-x1qhQsRpk%eAs6v9hz&HzdYZ_q{Wli+e-F7tN;F@+g7
z{2`e{Et~;{0@;Q4Mue?syR}C33J%eMGhG5~y-!L3N7+S(@|Yg}p1Lf$|MolAk?_=5
z@#0Lah{9z|2}y<T?+RxqkaCVJ%XiRZnWc2`FC^IIaPJ*Ghq=plLE^=P=F#e3)R#ZA
zVJl^o#s-yL<OoLrQzzWZ@w1eWz(hRe9~rT$DSvf*tN>`MbOm0)EW^h=ri(^^;5<BF
z3mQY)ix~C2+Gh#WBd_*ukBPCArXoyD8ytAjVT%zrLMQ*5G7qxrY35X78nb;>IMNeO
zlG`Z>3LlVtE72j@IE;K1;63GPz3+k=PyhVlk2bBrS$2BWBCbKCiU>0NML-Lhs2u$F
zwW+yAlzo`{cWw%Frg}-fz(}Z?ucAfx7bayQ4JgKBQfNXFXZME)yl_fKu3m1fx^<%<
zL!U}J{CW@KfsHicbe$K^f&t;&cy_v>l!qw1ujNrHa=U_r^Cfulc=4TwhQ6m2>Z~AZ
zIV0sEB&30A%bl&`P_5;EqdqO|YySV#A#ze8py4Ld-h&p#26`GSiB!d!<f##BWhMrK
zlW1JpwTo8^^3)n~@l1V_ko-J`h}<w*Z+JZ=wZM5l9k%3DY=&lDp4EmVHv_#<%bsCf
z6|dUGVm6J+Sf<fImuBmXjPHOgoMm11PdMt2DtKvSE<1vsRJV2LRH<{dG<d;!P!~}Q
za}=q|_r5JV%`mh53)SMA?Vpj6V2F-&S+E*gIjASuLWfjtllj6#K6C>_H4k0_-GZ~{
zBjOu0SxjbYpKaOmK<2soHg9+P#3b<UX2aB=mB5%zn)!x>Zr1ix*iP(fg<149NaJOR
z&8$jH2@!1R9XvGUpC~5ymWOrqK6hDzct}0d*)vh*?OQk86vZqxFz&(N+Dgc+GM>p6
zztmngqFq{b;e(y2;;*zEV(apl<PolpoHjl|Aey-Hd^Xc6M@VN8yp3;(53N9>p&Fo6
ziDX7y2F{cpcr>E0ywPDcr-Z3^lrRcAVHn%*#NHlvqYJ~$=J|ayslu$kIv3|THzae8
zV_Pa4tg=&rWvOVs284ffdGl)$YTsANi(z6Y1I02a8#0m!WVr1*hxEDCJI6AO<^Cqj
z?#OZcMX%ibk3Z?0Ho8GVHfY-daobn`y!qipYgV2ld6Z+kuvd5FAvf&cX>swhjfhtj
zJ^=apo@j))VlbxTdmavo{C}laH!kh=DtI1~PyQDhUq44T5A!(*3nfRK1}b>@uxTrB
z{$?4nxFNcy^es!UlLgDQroMv(fH@#{8H~H&7Huz8+`eY;9~nG<dtnp){C;2b1^M}G
zG-!DL1YR|ln|nTq=A)gY){+8{uiQ$w2GV>=jaH#TTJ=@uP!E}t;E<z<nmHLKx?t{`
zwnf|HDFH>Fx4o*i1nn<i^sFr|AWaPMb|ohmJcviT<I<S&O_atXZz&8uNwCfF31-D2
zb5COb_(nd(XsB87hcjZVXAFNVF6KT{OAMdrv1o4@vUSvK=`^pl^q@plL;*YHJH+WA
z>{V{brYrn^niwCRhiHb2vy+o{%c=lp=Zmb4yX@3u>4e$6B`Os$AMxYm*K(kksFwLX
z8#X0uu2+*eMv=0=>$o^X9H_20Q%{HVXwaB_V@`=@8iRORzhC1Xdg_+pGtq?5zqbo7
z&SOl$5f-p@9>d@k&q>^kl6uhXHzmB%CLH2Q&cg6T#6(VB)6!T5`u_*{J<}Ld^mY>^
zs6uEkL#Q#Qecw;%CncA1GX`P@UprT`1B1AHlR8cG|HU&6_55WY#>UdQ;>&xCN@qS3
zRebk)b<^Zs;D+AC@vIVi|DP0x1pSPy)LeYQPYN{qZ}1KNvXgu}se@35mPdj<o=LLn
zjv6Z-KmNBxDA9s2*E~H#rgmC|I1Hr<fmof8mDRGE9-=tN*7<!H4n4!L1&SUMEIsC2
z-F4BzDF9}EXR}|-er=_d=j!p5E#QvxGAB|(G*{RP_{`w)36XUgWLV&{4Wa(lxj&DL
zhL8FpvmqyPUJTNbx34B3r;q!jnRw_;2L6s0?Di6yD#1~<?Wr&*jkqQ9ge=Q9XYavt
zsDZRA!OW}rE91y9)M}>xbKW3=Z`s~6yCmWTS-pSdM3;;D<r8xm7J!H7DPSvAY&|VQ
zq@=4(wB)ss!x08kcXNLSTuG_vMQJ&kWKh6)OWh2isbf}}17yGb{W=Nu_Wxjm$C_#R
z(h6C5Kp}LOeqSn!j;Hj?dL)I#0ZuTg|LMWc(_x#RZx~Q=5CB>42My#DjECLxm$B)9
z$BlxqWH#aEcU1hmc@L;}3tjm81*E<HCUPl|bo$9Ea|8VmezD=t{c6e#Er-%jt5B<o
z`(ca&n&?{zE)KGNsOiq@xT}LcSN*5Jog{KgH<w*$%nU}|PMbHKX16Pbz_U}ap?1Ro
zSb{w);%&FhCByMg)$3ekq(HLOkg#SD6+7j%(vhM1X9@>+#!mXG_I-WfXst->8bf|N
zFlqLS04s~mATIh@Isot5UH-BzqkCwMypFT~W(IYO5D&zmRsY{%pc7h^0dqb$I7oce
zuu%@&4)Mg;d}fFK>jnKssJTM)F&ofBMt$P(B4*&=qqfJ{cfoNm5?a%DoD%?d%l<|2
zDUzU^hY{%IYeQPAH?<Vb!~)*E-Xg&`Zzd7mygFan=7*9;4p^FLto?jP;UX7)(;7R+
zt^5wS1beUMYLLsXyxMi;%XyA67Q~VE#_{w~nPb!M@>JsRQRv`N?+|qRdmP}P^f~o8
zX*Q=>$T_pfIqGof%$i5QW8Ey@KdyXZlytj^UJ-lJ4w~8$gElR6l0qzb7OZJ}jI9(t
zal-ck428rm*o8r!1^jsNq^*F>W$s3o`FV9BRkkP-W4hErdlYf_S_QYFoH{W+VUbFX
z$=NS3Wz@5Sv(4X8)Je7UWs`;pD)EHlE@3H>K-)`H)=p1pl~ITG(-YhZ68Siv_DhL}
zURM*DKLFgBTq<=cOc9Ji|MuG!6`p>TGEr-pG_rUBbvS?zuzEF${5AA|n!L8_L@$3m
zsGPZPCSHNqD2PyPkwBI%X|*$JdKx5uO0buu2lak%o|Kfm^sAkoBS4R;m)m?C-|mAE
z-`utheiYvN+pGI%Q?L8|AaYF}2c4XG+iS=dO@SC+F0ElxX`tXrCy&_K25_V8dF{Q7
zKQ%-tu6vc?fc(M2hgqPFFx0QpFC`8?tTO*!LTPMCT!!sSxp|hnS>`+$3tXRs%alHv
zA`?B-G-HW90gqFE@Fy3GsYz2dW3&`!#VKmu%M=}~+7tTP%)=R{>cu+OH`YAvSfYd)
z7mgK7cHYeo2HAgsl=(1Fp>hY6{3g9G(wf1enw=Uu3iUmqaN(${hF+yAih}&)jT<aa
z^62G;(W?zFzRo*CZF2v7-w>ebsN|vR{MFAJRu);D&tdVTpJK%YARY--6hg1x;|l}s
z3i!h*n-I?&atOfcKu2r29QfB%4S{L5l+Ml7h{lE4%pw6!hoIMRr1ze*s6qZ7fCEU2
zXdN?AsMIFhFva2#9SI9ci?+R|*{0+FQLVoRh)bgXUf}{9_WaPFI4fM*hU;<Lp)=+P
zGY1bqy_D3iR%iI@l9K(mFe%FtGk0W>Sz#f@4hb6(nEzJ`XBoj4G$ivVY8@=a&8@!Z
zmYST?5(9&YvT`eff$!DH;bv`lCTZAQ#4yRDwPJkSgZZ#1*92@>?N%DZVE(c)`_0N*
z;m2e7kyZGRgydK)`SM7q!FTKxur?Swh$^>O^;k-%+yD@^o4nC<{5sL%z!N0-uMSrq
zgt*9p>P7Oru6)ktjD5z3&b92zhObkIaA})loCEmOM;N!%gE)SXHr(c%6#+$K6~-Y*
zJw46hfeejmlAKJc<nN*aZw@OBecA?KV1&vqjQ5Z<h~ajKv^?Ql&6xJrW8p!+$IM<`
zRmK6-RJDN15AJU__>PaOy`<#ET=Jh0slmdy-sGpqc%tyL-snw$aGNvy{#}T=gFvOJ
zx>sE`dl{)AENz^LBgs;8v7G7eqmFF2?6ok!LX4Fq_0qsVUZ9fMb1y>+Jm?4c!BX9!
z<ZM%9Y#`fZ_kE0xC$F=!`$w1|t=@4zjLXt=7&LlKudfFNENr98q7gB5om5_4(&*e+
zbzpU=4an>V{RSH>41BPPC<xjUKGIUcM!NM?7ba4v31w(UNVpY#Yk2b!Z6gvFi#b1U
zzO(o2yh3gVBxF@$eFlK$bRGdr#sh2t@+p8ph`Z)MZ?iQG^n<?qN0z1R><8jbt?|27
z20Q6q>`0z9t)$NWj+atyd!Q(6_H{abWkHC(Y%!NN<LBX*h|m@buW1Y^fFdnSXNX14
zIV#i9x@g4!a)#&_C|BR3t%42UkM4ZyJZNH5Mm%+BqVUVicyVpc%4h5#!RU@BZ|aG*
zYe;r4p>CxEd0)f>X>ia#WGLQlblKIVW0NI&0J(j>qyd`g5_ccGqQeP9gl92o`VeXd
zH*2CGdh#;JW@Jempza6882ybD@#e3G)kk(+fE_yy2ZRAbE45N3B@Y&il7KDSTdhfJ
z0$}j}UOK6b-78nEiA>^DX5DEXYRkI)Li*e!EF`S`{gyP;FG{YBM%DWCh-v~;rS7pu
z2e-80o1Ld<2-gu}&B8s82P*;wD!qdGz<?Oi1z_a>1Jl`<zjo{lN^Lh+8xZmlpcjr-
zi}C$M%kiT4aWp#fhYKY9g<1nd-z->WW1!1O4;QLcC<Sm`DU&Yjb^LUfpZUHJAlPMY
zek<{H*BF?rAN==l5wv1ExOQAM)xX2mtIa3=jxN=zsNGZXX}#w?0$@nI#=MwEXiLm5
zGVJ5cx$BH6X|GVHwHR$(rb9H;3qN_k?b4ADebJGb&mCVV%?uJGAFhusTa!qn!GdlB
zX`LyCA>%f6yL%!~jE9DFQz#dBe80X?<4#{sSPB#xS-0xi6qsn-7&y3Dn90+_Yxwlz
zJK}qC*e46gO720uXzs=$-bpT2(p;cG57&UY+K_quw{+VZ6SspbK>`qFL-K4J{ia*D
zRw*85;0qpYgDh>@=`ca3uXR5AFvn0YTQBwCbZ>IN!Y4|7U~ARYq|mq{TP;of>aRR?
ziLc9c3zvhp<&1OSBKvrZcRZlKw&WL);};rstS3{olP#t;A^huUvHI>Z>l3WL;$+nx
zV6S|A@z*8lHkj*0=WgNDcA&H|WpH}E@jtO$iZw&^<n(kj1UWKBSLq~h-Bb%PcvOlw
zXL9DTESV92ftD5%T*(m)60?dl3W&D*0t|(zDvcgbEOBKdhLChQAWwSiZsvt3)Xb%x
zOcK3|$TZjXVbrNvip!N@8%kB5;DTm_qn3N}q~0nqjup6&q{jcIL!K8TOua<mjG6~M
z$pm>Tngy<04*kn%k25SF+yjlmRVovKJyCy$Z-4V~?K^+ad(>->UnoKC&~m$%5vNlp
zVMe*8Wi@4_1onHqV`_k(!>Lja6bJrsxLD-U-QKI~Fa1R>!nIwSNrhN6aFeYTo>B(=
zEENOc-HuRaGBC+)vt6OL^5F9SRuQiDAh`@p?X2z^(tJrdS`ojYS@CJ^oN?$_&MEU8
zze&G#q^K0MQ2Ju3AMS@*82I@{0YjlD;s?WZ>xp)vW0;oWKGU@U9PAE~J|dBw&x3Q1
zwv)G+ur$_-i&b2H4H>bpC?X`*A<pEVAo6()56MDWEfV!dG9$b)UgUuhiJ38+IMmI;
zb1?ExX?vZi6zVS_>K5WsQ|rHzKNe8~Q0!L1%{MUsQ=PtwT)N<ZI?Kj^VMMnhJRrGU
zQY>`}E)Vm2w?5sL`80bO2?fSf8x_^_+exlh#y1_Gs*MIukryTMxL-6a!>;GJX4&A3
zFgpYZI*c=(Ng6&rScwTqEY>hOB);?%MORI^?-`b6-SGDI_GYWjAI55F^t6fPXRTDk
z+A12($o;rM<-E{%|MmSOFqfOBp5KIk-|20|`O}L=9hM1-LZ6mUJ`r^+4;D_eOgSRf
zfWRq&kMgx=d{iT4m`PJxN?UZg>u)z!2}q_E3_D`ILSp>SO?_rID`$wDg@7gMQ*5GE
z<-=Fs&PQ5vHG!R5sX#t4k@M?vVjGPFT(CucYynvjcF?Dvwi31Ra4v!_lY$&E(PVMd
z9xIc}AzS)N<(f!Q(s1dE1VrM!wY94ANe^kM3iN+JrKO7)Fjz}vCCU6am%~f%C^-~a
z{wq09sgPqSvILSzVPE<&P0(7$s}8c&rW0(~c5#qbotLqhNK@#?#&+P0moA^t*d36L
zq@hUwpRAiuvoyrX;V9Kpp%Cd5Ba^t^85N+SZ7zYn#S!SZ_P?u2wE1@|?!PbSe(-0@
zXXssJnBzvi@Z-A=VsEq^v&Yq6vMI#+7q@k49S6QvS1{9Tm{K(Hak-UlVnvWuvSio{
zq3A&WQp~2|h{S-)`;70aW;;)3jd2iuERlsx7gbSd5q<heSTctrPLk|0qhyL2+k^!R
zZ=^?G5rq|Z@19FK12+?8sl=jmaOL2+@F|P7L)NdczfOs0^-GpE#2GRc)j`;rJPmid
zg~g+i?o`M@|6#3jKr3<6_`;#bn+t-~j3B3Ho_o_$M$`{mGj>%wR1l&lqSdc&rYJwD
zT;+#eGo#di6(VI8O9|D-`F2eBMHZN_dKy;aEA=T$q$JV&o4C`w345L|t=2M<>I5wo
z@IGFjUpcV^rt%4Gi<gmvh%x&bta}^||Kv+XqLB)^^ONMhJ0HtXDC)gkFq#UuUmF@9
zNBON&wKrE4$dhj1iZm}y`+kZBOj8Y^R6>E<=q0e0Mmpu6+ZbQM8zoqF(N`d|k)&B<
z3NV(>a2!NI-8&dvq*u<$Jd}fXW1X!CW9OhZ!{}9WX7S}h8@lseZL90C7O|w2qvLu<
zxFSIlK{3s37RN;0n4Bfq;yOQ~XICs4j$>22J+F{$=2xrKfE^Zm7TY&dkWa|quR57!
zJ%C6#?1NW9Klc`wnvXTG*eDmb5EiZx@$ELO92<jCSN$%KhSMiI$#l*7(;RPF^|VNQ
zzW1wMd&O&VrY)T{;6CDj%mCZT0(@VISR3=>^|MpKQkz1c^4bg}9EBpng}KjWdhkQK
zi?ux^w4!}W6EBLgIuT!+O9Sb;_VW7kZURC5XX2A8cO+jmu1K^~L{vz}Ij5ruE9{E(
zqd`>{4l&Kg1Q8p)AHIn_vky!opf#tB^jtSzk592Pz?aH+crKXv3z^V6)fMbatDq2m
z#bT7sATSW(!8{%jNC_A@%QFAcH1&Dj66!uYsV}#lq^q%s6ti9~xl|_?YofPQNW?gp
zjTs<-FohQ;XhN%&ljwzRyQk^Q0z{ktC;9r~+|Q-lV3th#;#&GtOK$wXo(}T!H3t&{
zO7&Q-U<hA!YwkHmCxh?P23=OZ>9PCpyzO_6kk_1;Pc#w`%ji)8EDHjMfrrY0&+C-)
z5t*J!Jwi7kC8T!3UrDB;I{yn9FuH-__Q!$Iyg*Rzqtp34cmWLIzNBa3eMUexeCE4h
z?)2_r)}VxygiB}Lu>3_j8_i=o?_PVL=Wb<>5|b(XBC<pFnEt78w`?7``72n?&MNd<
zT!wxkgw`-V?lU%1sCGnUY$(m=Xg!B}3W0-wMk2;vCK5e$gigiqwlRLS1V8cmxvAn3
z|D5W>Mp+h<awU2Vn_nXzNp4x{&FyN&;HVdj`0>K4eeP)@w4;ulFATNCEitoV-2FoF
z?dO`8i;I!WF{Y+^fBYqFzQPdK6bR?rNI}8a=gCa2k0My`i`qy_Uyw7c2b$CRlR~8E
zVtF;Mr+iCw*OK!^R`^Wp4-Q?e@VUk9LNQ<zKTjstoA!7!nLw~NPyhdd5*{MCeZTJ2
zd7VowU;8s43?oH>M!N!VrK8ImZ`lVcwipD0^&hsYP5R(PAEy%vG%r=(%~u&sJ>DL-
zTn-ZdTiq9p+$R<NeYQjq_KxzBrDwZVRY<$uOuDoiQQ2s|h)CBy(~-*{Zg?4slKRKr
zSle>Uy7m*!jJm@K!8$$Mh~Rd~=_YGBqB!svcuwYL*r4(BuF#JEh&vOD*y^WQjQ%Wk
z-m(z{%XyLki46B%9P3aPxHP5E`oh~E;2hZu`r3Hu!N!BpB^-Zw7B%iu?rZ8slzfFE
zGW{rcrC)lsC>2s%YR<!ccrjOKv!}3>lCG5yvHO{MOFOzOx)K=PDOhSj6PFR_YY$WN
zP1%Z7#U<KQauHXCGd!@t0SjNfhxWWEdHb)q;--AHNOp~-gou_LlEp@^btxTE8!@JM
zfwi>FKZu2?sPMCVywmGd;=TxtcPv|MP>z`4Yg2tn%uIBS!*q6o?%L^Wc7u&H(ol*>
zD<>~BE=-s1dyiM7-1@V}QE~!h(^{Oh>TGS>gY=#|V@HW`7-afmX+1V~7JJgTRMr=Z
z?8<UwA&#3C-by&^30R96y9&6h%2;XO6Q;A`nAcL1vk-yZTq@H84ziPS8N{DVW+i)k
zr~EJF`Zn{6?n3#p{0~q3N7iN;j&v#h$g&@)u@+PmjfRLh2BWQIiLA6odlOkZm838Q
zhN#Y$%g^ovJlG@6iuS}23bAZth%m*?ogL)821PeR)zERc>K^CE#>g}zY5t;21a^Vr
zIS;odK_J;fcJa;h54o4+l9Bk4ASbtzjv(L9&R#~VKYM+eg1%=K@g=%V@o&Dwe$v|}
zd2X?7C9OMdN9W>i$=tpZLidR$eYJg^+D`aX{9#+Fu8j=aPTFfw+QjmFE2Jl<?+PTl
zuB^;gqJJpjRklV=xb}S-Nqd8_{K}Nd0ufuT-m)AMn_)tVs{oft-y}zri6$CkWu};c
zq|v9=bZ@@s+SmT%T<%Xb6S8ZG)7D8UHd?!8Bp=RfF=YuMQ+MEYPDaJxnGTnxUi&Sl
z)3%SFNLm@86vm;MAKahlVZnHsO3?V_lYSI!6mIA#%fTZ*)z?A8dOQ<ygHBTjSxp@t
z=UTL$M)k;~{h)1ZAs0;um?yTLN9Y){lEp@vsG_H4Mq0iCckMLSUA9na#)`V_<LTZW
zt2AN2A2-i~|67Wh@sID$`8qFAx4ljFa1o%tephbC7(G+P;Ia6k>ECW^R4x2l91@L~
z&-z{28qL@BZ^9cqw75u+_i|#j)b7J&^w5j{0Nilb-at7Gzk!94Dh%?BOO_2(n(FT>
z)lRi(%iemuUNVK}cRJ#Q_jDX;`l(NNNptmoS^$Fq=u4I5@4B3R0$~OQt66(##gAd^
ze__Iz;E0r$n)XqYB`h+m`};qJPy?#O_*9EXOPC6Nk0{<=b}i>j+X1b<8}SZIT53U5
z96$N)<IudC#(#>BgSKIPYhj9P-QVtr?1gxg#}+~U^OXBi<r(df!AW@_yM)=^eat8C
zGkOD=Sw_fEJa?#1I*5+GjkglDLm=VX?T=dW*>n%X8NK$t)WjrTsmVrqUo?sgBH?r)
zI->!axv%KNJE)MU4UadnRxFSkb4sv9QAbf`oZOaBS)fTzLuoitn=MBt%dlVBVcZX9
z+6pZ*_8A-AJEozoq02Z)1h4yQ&gEX>lFmt;(euxQ^iODXXgh}^0$L$qAmH3PnN6)J
zo6KrQO0vZCu&|e@ExG&|z#PSCpJ6}q1NEQWY`S7-ekF@bN)J{9g@zJ0&e#}UrqVAV
zQl4Kq?bot56T+k4NkQjddv`!FWjL&q^4u|2WP3TeV)=j8!yAQ(MCdp>j}~n7K9m(#
z<Cv4kvuK5lZrThBfbt6@GrE<8P+<fcGv-!!aD?T9=B1N~NcM8I`kKjTy(@mq_IjY4
zJ#e(lv!?8FeiKcKldRgWQNUCrD(J6nBioV_@2R8rDLXW&r-SNzv^@>eF0Y34H1_9k
z_jTe3#{h{RlREYCpJe&aX(2nzn5)S^z1AeoSYE2p&WX>H0qga+L0=AHtw;si`Eg{(
zq^jO%>sm!4rDz>|9pAW3Pw(;EpV>L;<IHcOgF<eNXx^^hV(e8WJE~TN6PR0~gGRbi
z%lb_aCFhgO?Sl)Mv65NcZOwp0Q~U0*%;~+FGwIDD)cxE2y^?*GnOFs})e<51)Em3d
zfV@^{1`8MEtr~U~UZ=x!ZXE|_$4MqTrFNDKN>oURYt<%y=D=LVrwF5CPoz(z>u&d7
z5AB3`?)_Mp$<704h*iDGtsGxzt=4ZpFn93PeZfsv>$(W~qyH$=$~BooxZhI>kJJX7
zsg&hMY?$2l>rBXGDF&PCY$g>?I>lc#-OiD^KHv4;Ck5P+Q{{AIHP>*)@=(zdJIp)&
zgC9WX46|1`acD~}^j}`_@jpe#aAx858{`{7`o!n9yVzAjzbw!cKxKc%@NJS0VI#rJ
zBgre_kwN!=l2;(dhubR+YHhygMUH}Iu%q}~^dRg9p0~2vEmXp7M$v&O5JOt-Cxx$-
zQZ&kqx&j?JUrwEZvd;tW+k_tv3Ui!$k=Rb8g@^pl{Wc^S1_tnr#@p6!D_b_u(a^5&
zgD&ygf&2H@{mw1?{c~PlU-yepYdCQXvT>neCjU;!`%v$XvRL1%5O%(H=hB{UK3BYr
znklaLw{nR|2-tCHpdXCrn%hS)QO~in+>?*Dod}^1^JDz}>klDh7$uC+#KSF(jMgqE
ztdKn^m7$imJ=tx*!qQOYvDvJ@6MVHkadRK?BR?yHG2@$Ay|^!LyI_mL&+AblcAZ!=
za9@a|R}yYDx}Jg#wfh|z{cDmY-^qVI7AIgdi8Ukib>r>KJsNKTMp9YRJyusPu+diN
za$!?b5B|w(TGCcl#H_0vSkMVf8TT9!m~9$&=yIxVqEht#b*?FF!&db7$B1AS$Z5g>
zql|k~G{x&+67m9SH7&_pnPWF8uGp*1Z>z;ylB%s;*~dNZ5<0i@mb|6yH1HN$6G*V2
zL7&2^r;?seVrKOzKe%#ck>8l*EVBX-iBkO%FZQ1wSiGum&5hyb3`hAnvsn#A6#Ebv
z&IWqQzi5iCtRjUQyEh18k=P6OWbUw|34y7WQ<s~2BXPp@8v;*F&);#;gOcb}{sj=x
zwa1j#y}9d~Rzo0TUKh;b>#6l-%H^|;r(^O8V90@fY!gW9t_M^`sHU;2jPCyfH~Id|
zYuqftY47iCkgDfl&@Onu$pnIV{UAjw8)Btroooq=N;MWA#Ea!KtCz1E#t5+$LcO!?
zFOO;N#==^`Li{ix=XUi|ailKOTIQs3Xwh~4?=VC#x}&c3ZwxZqY3v<-$fWoD*1%7Z
zHpJGF)wL!Xiw!~h>Qn1a&;N90@vo=fZ}_s=x*~gS{PH2+$Pn>8$D62ImpvZe=*e0I
zR)QMk5d0H?%vboJ0%>7V_Bn0uMo2CMSJsm`Q^nLi$#x`?W4s>MN9T`-D|*$xC4`zU
z&fUN4+My<o5gRgZBjnKhD_mh7(wS&%W}B_WkdcV~Ce6@U3We$->R8ETb)~_P>(ipH
zp_Qh$-F!BCUW!+#SCmTs6DUGEOWZajYDHJbMd%Q;`KsHHn-cV3f=BK}9=8w(_RNVE
z+#xys4tH^BIWEOzsjfpdl?MCNPN&tsiP{y@bXtON7ob++v$BC<|A6#O$0%~V7#HnU
zZW@tF<#ps-bM-Fy!wMPSh}eZ{&edO1`#`7;z5BhUJM}r~6GCN%0Q>9;FH%+M>i9?;
z!o%ulm!-Pr=AW1O|Bq1*E>QT7@+O_SUOXM!PI-O({DR#>o*$qI*s$->0hsmQPla|&
z95LV3s7@!=ch{Ejj{9{*f(T&AgX<;*k-ziDx7^GqMim%+TVpmu;ag2i``htv!bx=V
z#h!QU*yzpo-C65#<$eWe8qUpGbE=tBr)AgXSrW#e^VP-e^D>)g&s9O+jr~$SP8mta
zyCZQW_TQI{RSe1RmN_08r6-je3>7;k3PQ$yIVON(QSo+20QNxHX`bd>VN!t%I^`AM
z{kC}~x<k~)zdwJUnEgN#GtO=Gw&{QX>vf+v#}eqYu;0S?K-nRXtxM>s{mHgx-MI3T
zrp^hy&tcj*&nIc7nVoMj?N;g+_RB86r11$N-`Z}>p3lv?y4MDDIBn6@8WilW;`cY&
zHu}yhpE-#?mQZ1oiDzfGM(Icl=zg+%a?gIv=~L^;MW(p1beO-f5a=A*uMZ{}YkU4X
zu$*?YOZs??=7TxYrfT^o=|FupN3Ya%)Hoh&Z~NHPd{Em;*H^P9TIsQQ>QbCr?L%t}
z!7OZ`<bYYh0$vRZb#~&10*2XF7+FYgQhDyt-u#D&8J3OR96zznAZIUfS50Dc#IrN;
zUq2cWo^eZn?qqihl4X}HxN|)lZMXSUZYt@XI{xA@D-S1M(xKZ&AX_FFZmYLa)OK@Z
zG@G$%Xt7-5`R|i5KK`+>`s&plc;7V({aRN`8RB7vdbXkmT_@$C&BEk54P!}7ytGs*
zM{Q)uv>mkJ{lb)*wD@U^T?5_9!)F#Q7#urMvqm;_Zz)i@P$IUkc4Plo%RA(0#b3$S
z_sdait|u^amT*e2J<-~XpfW<3cisF&D&`jI_sZjA-###@&tK)eU&UE0pnN@KMB(pl
zLfCS7!G>4gajpwQfzXoON#;%N=lX5aK}9+ZRq9HX>LSR2P`%m2nQgbbkmaw;^{-!(
z;LQ+mP!?c3=ctN@A_Sn5VsfEv&KK_*I1C;R_MN@N`rVl%U;?DmbL>8`h;zwhl*%eK
z&Wy%rL&_qeEmNG&yNUlz<-f(_A>^sh+dCe<qE42r2I{7Uhr)T?TUgI}JfOcKC&cmS
z@hxvhMd8gER~f|dj%QVJ(6u$Ng$s#}^@OxNHFzu(YE`{#u1=X0Fua|xp%bmH;RpF*
zD+WGC?vRWyF^cU(XF8GJ`Aw%PU2mP{nX2DaNB$qeb=PH=4S=uGh>l@x{eeL4pLjhb
zmM3)He*NwP2PNzfsTWDpnkR9(QJ}zozxTm@tmq;)7<6E7s{2(pbh<o$PGUYsb9eWY
zY{%9pzK%m>wi#RzA;#2?CIm3m<|Cj`N4K+<eOG84pX=Mpz1wc8jwsvdFpY=y>c5&7
z1vq>On-zyrx~@L#vA|H@QXm;k;*fjaAXjH*9$QE3RY$)ZY^JC6@%MxDA<UIZa6w)c
z;Ie_?)EQ&G@?rSrq`W-7*%b;HCk82{r{<?Moxk&fkujmi;+-6>%eLxUhOG%r=H_Pb
zrT}caWgq^7^qGNuQ(pgPCUtHz$sSyKja-u_!e)^!HF`6-*uf-Frd~341)R{Y_j8RB
zCX2u3T?1@TD{|yRed5EdkZaSVm*7n_>e|%^dl#!6Xg3$Z8E!hB>M3|1GGDq%1Z4kP
zSkbI`P+u;T9cFNbHP`$Po1thKgKrleIU97PsU)`IN<FFNr(;D~0SJ`S5nKAhmA1p%
zIRo80g$EV)E}nMqCH%vc`7w)Nk$4%Vx#0!}c?=pU{ZDY-J^XfOZ?is!*yLE0y00-^
zR?j;NHGM*9G}7snvB^Y}<|M^mIiM3^uQ_o{aFuQMN~gN&NLXoJEjN~=`ZcB_AB=wf
zuDShD(&4@$c0Fw4^!WxX(?D8iMLHj#V>-gk?ig$G3m4l?TqL?00}-Wxy~m3>L`zoP
z*jhRoly2g)WF|i?`GZ?-J-Y!Zl~SANYjMx<zpSlxlK=8wL;6_xi6YPW8R@LryrUFJ
z{T>bjB%7fusgmNKvY&d)#0u{6$iG+DT;vwsh48ZZmXS=GbJ1$yt)IXYa+KBR0%V}A
zt%*Kvk_+{yoaqO7$e%$-j4?pOeubFLUwOR6W^*Ta2L^AAetYZB4GoIo#TfXwQ4?jV
zOb=Tl&%}`;zgxZCuO@3O-`sM}7^ue|$v3{*oi)fdOoN-{$d^54<zjc9+J1hCUPR}U
z_WQje*`p9|&LeLzc0!c464!M3$<^Tqq*i&908RVGe9kjDU`K4vBmUu%9Mj3~oz=k7
zk3s&IuIH=vJErs2?9s+H{-9a0cd}&Ar`}wC)o%zxMqE`;um3}-<B^4;-F|Bx3lpie
zuTZg9==<cYfFzBpBDR1iY!&`sS@M9Cs-TzSb|~q2w+qSZ7Wr%N7+V*~ce|>NI~`96
zVV=~Z_}jZXvRbyg15U@K+Fc*J$hxzYjOx4c_$`qOwtmI+VI`qryup*o0rw}6&iUiV
zeW!mqB&26eFH9#S*(xs-*mND(P79$*Z$q-v$$nKXgcm~HFGV_6O&{Jui#4~emQ@d<
z2KVO3x>`}YT2F_GiK?Rn%oe<w&E^n>1w$z2m71t(<7I_-Q7kn@Hqsg6Qx-L99RQA3
zdsJ;0u~$kHooAE)f+3}kZTEiASSh5xpk!xDUVN}2u~*H827qWIb%eu9Ih<P5Hgo7U
zv=)luL6cj!Or!j11Fe^|yNk4lerjpiNB0wMR7+jfguG~Ta@aO~N*Ku==w#(`s1sPN
zcB-{k@nVCdLH_62h26z$oU2~%Hw0gkVGCk8vwF)pmfeN6c9PDG%PxKnf%)o?skS?V
z@}E3{CO)s{jy)S1xZ6k}0gz4TnB!Y#9MkN~vM;0*Av(WcY(@iq0ZXca_S<onz_}*6
z@Ev(93b<b>g9sZ>I3`I(ibw8XY{@&#+(IV}!IL+`)+P;~tJ!neNxWeCI}JswxWd8I
zQVvH-_Uy=ds1;Is-)f1`r2AkU8XL~+X;_kPzY>$B-pI<-yXhtx=_ykymj1Mm0;a=B
zvTs<7TghJPFIR1tQ-2T@JxTrFlgbTAR$$Dp%Q}Dk7!l1{p<B~NHk(&xWq}AQ@i~J&
z;x}c3XNZHMu%a5P?|40SaoA@Y$#T>RIWEfrt*Z3~0h_Bo5e}+W4%P7%D#<RI<^`Ff
zbb1mqvA7x^g+Ii92gk-)$j?kmCyprCo^d?;ouMC0>OW|xA!C!+sF#WkY+Da;9rcxb
zDin;NC8N;wA&ir&sl%dTH$QbJNvab+Ein`O2ZehfJ??5wI>WW<8zlcU{;bSWfzTDQ
z0T#ZBsN3P)fN_Uu2Wk%g6Z-q{K-@nyR0TMk-TK$V>r@$bY;>M`4+*{F-_LQ|3w)~s
zXcz(eEo8-tH25#*R+Y&k5L#??oct6)dyFjyC%+ZS4?X8WZ%;lOPwNQZ&z~MT#qQU~
zFG3O4tSEOZd8cg~t#RQK5g!#okjY+9#`wsEt#_%FReTTFO6SQ6Z~i9ydE7)l+t)rf
zre7QlB7WFlu3{-4ccskv<DZgU6NROWgD3+-e7JIx5}?h(%K7Kz<oLd|tt~G!jZ99R
zkgO`6GqvB{f>hh11!bYc**-gmE~=i2_a}B-5o)gjR}9vtGMTL49N9@X7kl|ZK4X-a
zHZJDlb6PXSa#tV2Q)gEiyjxx$Z=F=n@(hC3jGnEPy#^<@17^+u#fvOVI`cZ+G>C=8
zNUDOHrQz~tmLEu|`IxYanqB6%Uh5%0z9NY1%|!Wm2c51xPuLjASUwI5g4*r2e4>*h
z$g_?1CqvJoA_?_5x_^O5o*41n@2$aE_MFqPrEo~fMKxhRf1B6X{xR|KN_+J`!ejLk
zlTe)M67I4)p_Iu%>BU7!^$)|Kr|ZGmv~)}1LaV{9DX5xoN9pxEU}(Ncr;J4no~bXS
zM%AU@_xr*o`4WJ@QgZ6-EC|($#E*`$>^JAKtlpf2^SZUaP*UJ%`?JTn6ni+bDEsC6
zLTSd)!rLYxf>)f56suTdl;wE6jYdV2TrY39vRJpz&Tr~IbV(A5n9ycBL+P{HjiDov
zh}Rs+3~T1hqs)Y3f+3Jn18$t|bS#fU(P52+?N8cnRpu+ltd8Ya9C&dE>L{Y`YKDP#
zSad-Eki6%R$ohN>)5l;6bvUj+>PHl&!_~k@`?>iw%!RQ!yM`q!QHjq?m=Y*C&Hxnq
z0O+pL>ThIZZ`Kk#gcHSOkbO-Xz=1#k^J8Tkjb@>$(}u_LNH`6X6twKGCU*<&8{*1Y
zOpgAdPkH7iovHr~YoMma9QNOi4kSY39gU}-jLtfH#slsrc3s{6A5ni9*47toeZy^m
z7I(Me?(P)V;_e#U-HR6s?(QzZ-L(+3xEFVKd-Ff{InVon>&hpR?5w@^oMZgPT(u*^
z`>8WeDrT``zyqT8BIiw^X`Sy}y_HFPW+lbE(XXPjX?(jr2>IGX77Ux<D?i?Lu7KR)
zOyo0XM!*NNpw|vsYzN2Mq(wzQOyogs7f6Y1n}fWGYu?@3TOEab;TDU>uB_LUzT|M^
zwo{xMJd7KMZVW9t0F>>m@eT<s%EFRw;-pfB<OmrY&;I_c#IqVcVeK|op5#e_bQkn^
zRw0d$@iS2{R@F>&-d0u>zeXeps&svjpaz6RCdflQcC$WYIHh4RzzGG4iW|lySnTtW
zS}NCgmEorqp%9UYWhQ#!FYnRI$5~w8x*%ph+(z@$RNGNdfT&K8EIZZC(r$u9DyW;;
z-6w2bunkHc;H&HW??<34swxtC0Y!57Xd9OxuZ6Jwb{*G;JUJ9kA7VULtIFgU)e?Gw
zcp(5vSfuLkPixij<SY<t*pjVUrhNy$EJO=Z6cyjPKg=dOS%%0^_jY{xpM@ACPJr)(
zzvfzGRDh+5g0)@6)ajrL(flB;Yji#uZDU}e1XV}V6>^!tP1T&fh28*gEaV5}){jEA
zv5M~>D1-O8br%?+H8JRGAL>EhdStGu1EK<+1B|W43b%AzpSbR%<1K@2;C$sScg=or
z1XWoWkEwNROl0VPjga)rYs^Q(OPbkzTO(cdq|uuyJY42!mJ^h<l!7ltOo4%M(10hE
zp~Q!;8B*g1v|^hy8dYG;L}O8I7GUE?t7rjVqF=7^w0vSU?-`rr!~gcHiB>Wnu*@sq
zmP_n5_=q(`G}Rr@2>Jk*mJr}CCHn!Sf;D?Hw05)q+c0}x@f<^S0vjIsYBVa9>5`E<
z{2JOcD{VLZ9*VUXgt%jT(3UVTmkI}bSdvyMx#{<|&A)v8`qi=prG_(g>fpeDz;xQ%
zfh&6K4qVwKRbZVQFCw2>uJkxY4M6d472Ie3;vso=ZZY3X&WehBlm7D~Ie@u4)aRy_
zL>^PM`qx$10EVeshVjIobB3r$giPpBzb}q2?~J)}fDE+mU&ZFR;#S0jHIN}jqVO^W
z966L@ItdVDVOrXB`j;u5Ep1)3nu{aJg%Moexekx6O2*fs#)WWwlBLwX&R`W{FpF`N
zQ1*+$8|=Zy^-cKt!3?Kn18)GfL0TQnMSbh;wpN0EGS{8^psMl&WaV3uQ<;_m9Q}0t
zrE7$GWIig&UwD|-<bAd%*i`B}#}$Qd<`gGvKFIT({Z1=5wABP2ZyOF}BhTlk7XUp|
zVq5Ou7vC?iY+bm8Gs8xJo^j=Zy5b)R{yiAQPK<qVDAYjT;2my1m>$rJRW)je+?2(0
zfiRiDEq;>s{=js1?%n?wI0FcQY*u+ncuNK`Y4$^Z$zw2&1hq5Mf-ID8u);WeIA)&n
zhLd_C8(Rc!;0u3>)$RZ{Dz#X#rgL4iblkO_=0+}?zYmwmNvQSQf3{ZO-s4PFnp{d^
zzqjovd2T(S)>-Rrk_{r2=1BQOI!lL9+z0Ayu#Gd47oBO)Udbju7DV=Np6gu6kxsQX
z=B46#1opZSQZ){@P}Gn+L6mp6lw|Aj%;f{x;(Ro!5FJue5ki~BB+KgZB%t50pBKx&
zz$Ta4P*klH5|tw|Q_W#O*_KIokloy84G0CnaL5<<9-3vQ><u)#9E54I^SGjD2kZ+E
zQ$r!}`~R;@n?Df$--=AtcgNd&UTFE$x*bN~`OvZHw{+&v>Ez&kV4kYa_29ipvY)|e
zle44I?RW|Y8w~^!J7)?mTCd(4H+X#RejH*ZzV7l19yP|DwLImdlS#>reLM8npWheD
zFaDVig5oAFzFe=2uRfrne+)HqtBV&#d5`{H$?ddRANxx=S|j6@CoHC#3JeG?;%>-e
z7*Q$bz?mr5sC<E0$fc(uKuh*c6p}*y!=YMAL}x2!OiI&M&YQyWwRko`l>kX)OWF|z
zHyQOSscQ9a8Wq=NCT)x_FxE{gr0fF^5(EQ4>%q~g0gq}t_U3<#YNrNF%a<_q;P|&@
z2Oq&T|Fv&kCWnh!*x|S4e%|ZQSR8Oo`5UmyLc|#K16u4RG#CHiB4qM-d`E(wEytD=
z@rVTi<x<RW8Tby9ar!T}^*ay7AA8$uSA`V1#!HzWCh9wN1*!A081B*75*}C<0+?i$
zN_0@h<Zw)l3W!<LAP8F?HbTF<xJu@3eeOFklxi&DEzg%=VP6!n#dk<{CZK<!QrL6)
z407p|cx*ZQXCy5IPA)}_z#4t0RiRpJ)fKoC{HGC{barW0)Td>kgP?ro3qsE2j!!bv
zM)~>2y|hjr6%WoXo5^X`69okTmWzb%-2@jA<eErug>XGDP75BJ@&Oy>WrxOaf#xDr
zh``1=z{p=nzut&}c(mCcALqnauTf#g{0m}O8!qIx-P5h{3-j07c$U353vf$KQiR9`
zD>SsZ_**CNE=SU~bAiIF^DL(Tt~S+Zvs*gYDzK;|F~d#_@-?}Pf(H%G#~U#Bhk=9o
zmA}6!u3vb_mU@5q2nCEo4uTSA!-5MN(SZOlXZdnk$Cyq#<eD?MRy$c4)(Hm+zF6C_
z-j_k-!n)R^%g+7$cx6Hg4(4y$QzGq__}SOGQUdn>-x?O>;hSYd9+=lO6nuDoy9f|I
zS^WOoXERswv-3dkBK_ql!{$EYyX|T-q@<-x)w|AWz6vcWE&cW@A|gTz;Bij7tl#Fm
zce4O|UG9NdXwWr`<J*q>=AS$6f4><YW4O*{3f1X-%&*7yZs+a7lvtex(y&h8T>gz`
zGsx_YfzpfqMOQjHu&+~&H?iISfDmC58&fc|9-;~|I@eGfNE6EC#gNQ|i4y>VLKx>h
z=R;V4k&F$7y@~#yWQbumFb~nj#2ps$`Arku7}gRudWjr5({-7jxEHeDWQcipOxN6w
zK6Yw0+mjEA=*L!q3M@daTa#_c-y-@NDDRY5H>6`VqcC+XzHlitE6s~BM-oO474)xk
zBWi0-M{~q7Jt2vf*5#QJVp_##y>vwXC3DvhBfFdU@~AC5Fp070wu=OVYYciR+RQ)H
z*%gdBN-@!7q|%sHKr95OT>+vogV--Vc1eRfUQ9l{6$+E<Jp^0)eNQEa_}`H2lX0$U
zbl=v?UvU_)xx40G6_I2tMQLV3X*3|7dAIYq!y3&xCmDy?T-?c0WVM!Fe>6AuP0)Yo
z22iq+A(^%{`uB0ddsq0tkI>;+hKgIbKjrAxS@Ne}N_bp9VNNSxAWi6G!`Mnf75SBI
zpp9Hn(yf_oyYtcJSZYwc#m;h2_Q(ep`w+Rs4A&|8BD1$}leF?J1%%!y<M83OSZzd4
z{;C`b**qb<5}n;pv+SrPc>!_JR3Z>sE+{UDg*D{QC77~D%F-PBZ>i!4iRbQwqS4?6
ziV6J!PYTJ!dG#*pmy{%51zj~Gy8vJll=dVK08KU3*V(;xpjP};;LKX`Or&`Nx;34U
zfAZsTdzI6t^=_m){JePNx)9?XmJ<2Hmi|}iGqbVv)lm3)&HqEY4yTRt|ASHFn7fh|
zyUyKyR0#Xb%1~d`cyvYS+P8%w5pdNx?MT{p-DzyRY&<x>UX5TtABo-Vr5a9F8r)U1
z?fhqpW%|5Tp7Izlr8-&2wRl`?_)XjSzu)9SIc|6z2s)h6?B7syjR>FYnCY7C>P%Hf
zj39uC$F^L!{37h*0je4+aH6;?5hcWQw{o_#SuTwAHATe5eM9vY@lLa|2CE5Lvqjnh
zu+GgA3r!XcajF9;T2=+(|M4AquDm`pBcLt><qkP&k3#V8a`*m(f>?(rwQwbPTe*aB
zo-*@;diy7bt3lQtaN-bqx+NT)h$uO$ZEcushAsO{r1A>OdgPT|AZwV`fTy6zs(f#=
z?4YG!34oD-CzJUuFWPuVEi>1?+VP`kH#ao^oiS(K&ST^$ckozOqAyXt8B<XK=Cmx#
zwl*Z1%Vg#1^-H(5RD#~^i~Mbt_ahS0ns0x&^7^<FV?=MAaRK-~{g-5t5(1`fDO66D
zHGA8;1p`e-|K|U(0IPm!D!YD5>|lG}3&kzG47{$j|B=FAqd$+QEU6T*3^qhb595~T
zuolMr-qw+3SE0aPiQg_i;>s*pfqWm~9f{IZ`11@+X3H1dKMqE%`ODfUkd46-Dt@*+
zseiN`hR8<yMW$qhmAokpa;Z{|xcDoqIRD~_d8J_Mwbye6!-N;Z#BNmm)Ep*T_-Z#B
zK)DH!nUR3)YYyA$u<(lqJmmkxgKEi?`nK6?L*3msr1*x>MkGuwLnffBGTFW>C{lY7
z=SGq8OE#|zniV3JB*uhDTEe0SPfz^eF3N?pu}kU*%QQGQn)L5Db}!SWmaBq|po#?K
zO`V(4s_|?E`t;;b7ApNEFwd<SGGf`Y0^R4Ps$EdQDd^v<2b;^-$idf~U_oU&y*Kw9
zEg#Oz(iH)r4U--5Ov^KZrzhHg==AlgqZc@+9xZRIWCUUH1<>+t_4Y^gJ3IQx^c1_|
zy<RVO?8jW&LFs>7<NwnbA`1-)L}2?MI2KV&>p%Gaminje>*U(Y<XYdoxEDR&t0CWy
zm(5MZfcx7y#e>YvSE^4I*=`FrtA__@q|lmrPB4r59z_4uERO@iuLExz4lS!|u5&*>
z>AB8GOxfs8T-64G22$(0^iY3ga*Uxsg5EB8@w4J^ttli68R7$SV^MS`MPK<LB;=UJ
zL}U%(o_w(WJZd$J(dV!wOFbzgIZN@l1_tkIu1X&^2HHmkref_D`VNR5i~|^En*)|l
z0QgqfkeCPt{MlJH(OTp4&}lM2#y*TBHZsTYWYI_~jYkOjglKhThsyVMbb%%nM8$6o
z7mPoFsaa6`{yZ1=F}GQLe*DLc=s&Y?B?(v+S1jsHXAYincLKmh^>xukzse`rhz*|r
z8WI^=1YC;T>9)MP`Up^Fgd)5<EO9PcD!0hGTy`>jvpHn3ug|Gg>u)_;?|<DIh+0#p
zPxdD<gymcJ;k^lE$~aPzlGwSbrMFkGW$7`4BrSQIMq;sje_DI<UrieRV-O^=OGCmh
z)&XlvF^#%jSeCNqxOtNPZt6pWl(}&#EC#V-o(VKE!PXG2Y%_&{`Gv;xaD0`4(Mz6#
zJggF57an-#<&KxhOjS%dQg3tS@UOaO`}e9nm~b3J5Ql`5L-c16M~o8d`m7~XStwqU
z4wW&>MF|#KqE((20WMw9j%mP*`G6JyEs{zRIt*AV(SwpI4%e8cd!%?v+bT4g&CEM2
z&36QeA~qoS^!Wrt%|+rWO5TG=$sjuq)~mZ&d0S!a@}b?-zL6;?#!*5=R#X?I%3w`a
zB2}Vc{GB8M7c9}vk^%c0<7XUhzql!j{5Lo+scW-sc_EkM0O#d%8(yzx1@A&Mbi!uy
z*c}VOS_|(x0&Pw|=7ShL#gE%UEt#kk;ogOAgtm_~570O<3C^%Q$A9FA8b4<~5d)9w
z)IfCe?rNE4K-~H`^I+G-_c5f+n7JG>ce&{M>^_Xj4UwymC?3!uS>TW2sO~Q!JJB94
zc1NxXOOF>-Ic>Qn0zb#;F3sLCiM8&r6U=cyl0^WJsAbM+M@TJG2k@pjAepk`f?^S0
zVp&L4-)6b)pDcvA5Ty_Esu}wKkgADHh_r#Q^&-@==8W*4z4QI;P|E8tKODJz-KsQu
z_(14^*WhJkk=SVMW@WMaKGgCUq5r>`K+bP>i!KmJq|S1Ntfuo?R&UYV**Rg#|Gt4j
zreE74duVHG>t+RbPummY_OMCr)`)u1kc!DS7i=r+vC@;CURq{mC^O?$$DkA;0_#cZ
zy~sfC(}+GaX)Ruaj(wm9&_UZl8cLGBWJK8fx|pE$B@`|0fe{;rP`0M%R4xTe@EHM_
zzC+*?=IeD31sI>sVs;V~(%ZDFhn<-9hHL{>EscgJHVi?4O(BnvnRzYrX2ox==8S*3
z%_KB)R$CU*ttoH!RiIiUG{g&#cKm9M(;osZ`~RS=Vn5VwGdYow-@($(vaObDh?1+V
zndVMRg}<lYtfDx}cE+o&^z05^UU})?=vNB=z@=@TY_y`zC@ooOgont3!DwIBgY1&y
zu*`Ew78owB<8iaZe1BnM=^Y2_pFTG+vvi(Vx3|2d>Dl4#d8&=u;A#7^R(|!yWoU(z
zeh{m(h{KB->6(jNEUl=xEr(GXTbTXe$;=i5W3MTi+e8OMHS<+YbTa;SpAdU*T_MjS
z<v%(hwD*%BZ+I5trJV?)(y4XW1T<Wic6b5v(KS&<gEAehQQ0A@9#Xn3)$IW$<p*UW
zUE8N3RLq)J0Wb)OCy@ZTZ9RH^$dK_J!#ckY-K6=;pb4Keu3bp0JKA~!x{b+B=Tx?9
zHOY<^u1Yx-;_U%ZkoXKNrWb^lhMDh01!ovoS6dAQZA5RUoikK+_&wJ%X>0z|334}a
z6wF;Xx+rmb?&!lc(<;53uN<ASUD1_nL|Lyk-gvqE2D{)JW9Vs7SD3S<9ob#=!|G>M
z;_ccW5=si`mH6R?xgM`7BwUshQdts>NY<X-mQ(|9OL)yNh$E%Wk|G0Ord3XXlFCG;
z^@a}-(EBaX$<u5p)0FaU(AXJ=YIH&z5Fp(D*K1_C$Qwi4^U0-MMZV0j<Wgl6ta<mr
zW?6B;*{r2p(v#D-PJnISzk(4h@{+@4d1ixwfMZ1!B#aQ1#~L|@Nptg5xeX9*jU<HB
z_g$qr!L^<w;r{+W%4Q=3(I^=yWUX`}N>(urOLExDPyP?X_pI^MhV~={`+Cn;8Vm)a
z2uXJtSUjAsN5{rOO)oBvOixdb5C7~`qq|C<@(ok;{fD9FbJhg;dxuOxOJzPF|Nb<3
zvw^IyHMoEvG*r<cbk~`C@+7*t;ZosJ@A_<9Q+1Y?c6Vf#tx)$kzw47?5|L`)INm~8
znW?B`@BpQ=v#N!;!eGKFJf=6=`Uj0Z6g5GMR&S)>Q0Ml*_8EbR`rd*kpL_gYfmFbx
zwxc7?Ec*1aj?(>KeTWT`MP42nS?>@WsFS7?gV}|b>0Mwc4lmF5x%w;XpI%Qfki9@2
z^{N@w@A5_rJ%kn3#vP7AHP}>Lg?9}9w|o>*Wb$;<EW{q3iw4_lI7bEqh+>DQ9E$~%
zXE(qgZ=2(f{T7L%RL2p}z7xyC>v>wY_;PbTzC0a(XTVqh37#~eqb88Ba+n}BEUT)>
zo^L3dx@;Jm50z4%#JNxiYKL!B3F174soA)=DTdVy8Vh)!lXs1$DR8JrUn6HFxZ%;Y
zBJR&2mLItX%?4rV<LSx$VB&1YAY&YN0~-DDq#4GC<1YzbhGXMNp@={-jY&)YF$0SQ
zd0I~$x;S&pLT~|rT%MHP28IG2GZ|7bT2pqNSj)kTPK%jbmoc6rk8W%_J$lA<oFAIQ
zCHPco1G}1By$DK7ZiPaZeifS^9BqaMQJxvm6Gd}VU_S?{KS*&H*fq(=+vgP$(D4P>
zLy+3@6sy+!D(p$!f6zZ29IAR0!O@(=K@Ky#C7pUUpEjc*?=YE9uWHmt%sQh`)`+Q&
zVafu5?SUuv+Wc|A>UF$KJFF-q3_H={gWG+ypngSRDR+SC^9^YjyD>-GNe$rIjavnd
zA;sNG>q`(tGGqdRQS=}z($^sIIoLdY8;o*6?M|gA6)_%C#%ow#Ii*X1_Il)jWUh#X
zzri6{7E_!<zi-(&<2-1*v13^JnrtC>iPrqtoFIP<BjzHy6_@Mh=1Evc8#Y9lc}fbL
zPi<)Pun3lN=f<lZ5$8J_agUq+%~4IWoyPQCHz;lkcJvurTbE5E77xYLBpaan0xaS;
zjc*}BXO<b6{{L5Ys3&u|@3*Cfr;?r1o;gwu4$P3-z~bq8@Am$_yDDqxGepny1lt${
zFJv4YSp|`7Et$sdRXh06FY38~vlK2GL})|pN^M|e5>O>({4cmo!3(EXYF#4F?i+!R
zSQ7xt6wU4C;oOGihY)fGD`5#kJMuO=54Oj_#ZmKyU5%W?_V5X)%?A(w%I?gaA1Op=
zByAsa{=@ToAIb-z#laMWTsFyT4xF=jxU^}?9DP8XITj48GM(95#Uw<SN~t;?yR1lt
zVbU?!U&=v*nOb4qved@zzJL%-I^VnfL3lO*NaqAAz@MG<1VKxb)mHG$5hY<E9W}nP
zXk4#Z90`jwd7MhZq1r+(2u2kglCFeC83xi26_k~WLFoQD(=yhEJ>L%>#y(Q-d+bk0
zL_AGZdVyQ@;2)ex7lms#+?rB^R;(>my_6@@03@P!E78Qa9oc`MrM!ji;Gx>^H!?)%
z^App&+Tr{@Bn^oIsI5dQ*<9y<E6!cXdZbE>t8%_#yDy3|&_X^8eLj{=0mZz<TJ2P^
zk&{PYg2PS}2da>kBQBjM19NdbR$K}^$|6~Yvgni$CCDeJ<J07lX)C!DmzED7zd}z~
z<T}JrkQ-|m<Hd%VD%jP9NHt(_^3c2!E1jsE1Ms@3K@G$+#UZY~%j&+Wg|9g=(O&3W
zVf-as?2Jer^INAZ&6Zkig%xHh|5Oe?9<VAe&92ePhGfszJeVAm?k?&vym29PCGSM|
zZA2y(Lt=&$Y@CZy->vAgXKve)8n=1H&d{}Z$x)e^rSidxGr2aQy38rR;wR2G561c0
zh=wVUDeHw4yz0=0f}(%aU=eX0uloMGI<!eVq6Di_8>;E{$fOa?IGXfAhb9q!y=h0S
zyRU^(aQW{k#|hsfaB#x<iZ3M(&8kEfPGU+FHqcf<rTYlQNJ0C4I*p=;&iQqgKf<(K
zAR4XwDMzO17VXHB4qBWq*=DrE>X+uIYZW`}HBe%?<!JIX;C@43E#ZV|%5PDJ^ZY4V
zu<{!ik`g4o{MGl%&P1$^wRsPhU7#%jkBNH%_l0V(?{ns-Fa$kVxVZ2^QdC0$vJonP
z`z5Y+ebvH}ORf+^+`ZS@!9V}*);@nm{FkxGa%|(`yS3ZJN{7aoYH7tnOZVo(&GHIt
zn!}^3iG+@hWM;frOs}aN<dD#uE*27Tw^#0|MX#qOB7n`i#L$fDfqX^@j$reUao=p2
zncP6n7K8+6dO8t<LUvb;t`<L2<4>4nVd9{cf}t-ADj&7bTpb}hEWxLRZiC7qho|UH
zfF+GcGK<sY1C3PQxHWTq_gT32`?zgrS$MX2-{3Ot4}rs=l5h2;QbaL?oH{&l3HgM{
zE!}t3n&uzXKjL@8m|fq|g8jn1uKG56=yp8WI5c_&J<r|o=^unBbD7<MFM}Pp$x!&4
z_sqEoetJw?XLBzBJy2cOhhBy#cfYxXt{s;BZN}F5H-8vc$mBQyPfASfS7*)c;pJcX
z<E8koweYbxzC>WEFI-Flpm(cEGcZnQ<q8u-05o<A<hKz55x13B(6YwW#oIE`T(AdP
zKV>D0^Ur&~!kg+P)0|dx4g}vY?+AC^imf$Y%j9VKr}IjBX4}cHU)eQwpU14Zn20|F
zicy48s)RLxs1{S)X_iNcCV8t8AK_(pe*4LQ9j>BoJ|cw`?r-qlteW>mM5GsV7jLGU
z9n71$V-<EWe1e!!hB1dhwHTlVyy+!ZV#;E3!I|A*HX?dSn5~L>t6zJ{BE}4|@HN>&
zEuZD7@8i!95P}y1S<efYR&MtfReD~8^%tf<S>3?BeI&g(_(`V<deWs0$W-ttEjiid
zzkbJxr7XvA7~SxoN!CA{QZKwHwIxenlL%w0-)}+Ea0#<)qnw~p+`TZ2fdodk*e+Ex
zmB<k+<lJ+fZ{OPSq<Su$?BAktJGaQr<w5;Z#E&u?pcn2|LwzZF6<pG>nC)E_3%Sy<
zRD-Kkfe(yLMZ3;X(>tnI(qBWFoBfaX7bVI%Kon<R+BPqyPup5}Hj<$ta6GIuhvtwo
z31QwEBSq_xSK{-wUmD5bCY?GB^NO}g+VFLch1x)b7dR6OekHMf`-D>0klf&__F-pw
z#ZXD%K7uHvlbRDYP52B+wis<3I({Gp?fU+k>MShf94Cq$AwfD7BO|Awl2$mgh-{!!
zX4{ZOUd+yx<gl-x*`}JSNa-2Iwi<IULU!da@3}U<OJiz&wrMlAQ5E?jGA#vmUl|Lc
z^LV`faQ6``C1RTs10sAg*O%rg0V-mPVdr|t{~c*A839JEIZ(@|vHZ^(&w<4V^ypU~
z_#7H+VE<T|y*c1U{Xy+AwrB3Sf42T8@>|buw;XSGha`*qkf^@`)oRks)gZqs?se}Y
zxcw-3@TL2lc4efPz5C>VndUraOz!>z{$H3wiS1bdwkOT!o!P5V??2bA;<f;6chMoW
z)9gfyXHO2XA-#3sv$*vY*vsTRZ0UPfGEClean0hyAJp(m8}Orzrh3)@lkc>7(+MK^
zGtGa{Zj&gP8Tkn~t-DFsQ<M#cP#@s|*6=;N;P=Olw3W92$2T`3fvLjvqpd9RnG6Gf
zbkk7Y?=OLWDt))w>nUVOipUp@s)n>QdF`+Sa5pBNC#pmKh(%VcEvE*hV1~dydr@kH
z%hpw5u<G8>Fkr6KanNZI*Z_`+^Qn(4!bI|EV~UAs+?z##_iP%&lnqU?ggt};w3q7h
zS_h&@tM=4h3m(d=AAKVqKa}4N2IW4ms+IIv!o+{ePFz@I_?J+Bxd?$%WJReq_dOr{
zU922|j!s)~I_<A>edrzLHjPNZzxufW6rMEwQ;uskVNZGNX9MdM7UsX2uOnY!a)s+B
z7C+YciCaWRJYa;|ABp`<HoqHH7dD0gP@Ugg#ZToVWAtC&DmSLUF9Df$vHq~;#4gIz
zXI+F{J5pN9FLor*@ukyT)NzrLG*{39jTv=0)?a9Gu~T9uSeO{P6DUfYkvihVbmeyC
z?&U+{v=@dm=39*})PAET6SxV|qE$x8>koUNSyV<l4}knN&Kg`XoBzZy#z-is7fI7{
z;pj*tPel#z(KJ!b2N@enk5{xylKf5@e1Klf6uKKd&}!L|W+qqrGL=Y?4q)VTHHjrS
zFgah+@CgzI-(aZ0$(Ck%`%~KNB$yK5m4iG**(Q-x;C-h^<2+TS;jy}k%#w7GT~)~x
zXKyK2prEUO9HKl~P?ULSG_ogYk)19X84CZql&J)?yRD%&VtseXOXX&IZuom{BK_~f
zg<rL(_vqi24aaZUz%1;41wt<gd@q592Qh}PtR-`ya^3+8?O{!eAEb+ue_21C7*i?u
zgs`B8e(hNVpy(1mkzkIg)`cH;8oZ5f8oy`Ypn7Re5tt_uFXOzgQrgj1Z7-02WcP7@
z5JGlmW@c*0%`fghZ&Zy(<K85;f1=*DpQA9k_v~G36>hnsMVy*hLmh~z?qO{Rm($YK
zS2i7uFusuvO9@?AP^Q<JX=u6?$k4Yv3jgy^W+h+LBt4Qm$ua3GDk@#sH{t<33>uhB
zAh{jUh!5?_PQt)?w7o`sy!HOE5PD!<q$z9?2|SEsaG!CxtS-4C3)b^v|CXH0E<%ZE
zY<Bf0Ls<d1!hW*kA}l&GMOF-YjS<S7**MGbKds_6;ib?qc^R#I)C&9sA!Qj<clENr
zx(Ndp+bfi_Fy{}DTz-;YN;eg+VfsR;!`+YcTpcX)>6QAOIxzeD!-l@!8=9Do$XR;g
z-~@|O;^G+c>yJ=BG*}1~gDiC=@&pp=xfT->d=!8@jq#vLtm%wOY|Vt(IGb^66&1^G
z`Qj0D`;y8-&=kenKe!8tyg2$T&UehM7z}j?$u4@LRwN3rCU-Ob;ZDh&Z#P`I&{mNr
zC1e6@CKR}FsnF*(!j|9cg&OX;kcFMUgMQLv@|S}zsICQBfbs;xp{BazqN{Ykd@S!r
z>p_#AJPaQiT3mZRBZ6)+wUndLZ=X-fjq`I|>c7|f(3#~MPER4N>XI)09C0O!HSHqQ
z?;!NTLGO{UXGEzkmL?rT057RIi46x8qf3)-77|s=Aw+9$!wQy#5-$WbCzoj7Py_~v
z^81n={e=RUMl=zFf3Pw!@-2$z{CiIS?Wao0ytGW#**_&zJz9uG*B$}rb~>LMnQb)_
zoR@}{IU1FX!2r&x^*1F2l`u#6Mhw+)Ff4CeFSn-$`1+DlkC25F9rJv<;mloLDkP>0
z3pK8J->poi`oEoA;cCL~%Ln|%<1n{Min`EB@+6_(`iX(Z^G9x^z+A1m#&CweMqR@m
z#&d%58=ufKV?h5;lsD3rtkx)vX6RpN^HvpEU-NA34|cm3RXTUWmpHk_I*9)Hza~-y
zyzm$0zL48q3~IliIc09G)Q(U>MGoYC_Sj>8#g|Ck7XCn35Ply&<$0I!73{9(N8KWN
z;s|et;|9-9{!w}hFA{j=T{AZ$ksJ{McM<R!>JeUD>d0CVv4f;^=p+M4X&JiY6OU03
z2)P})I!_NOWVFQjuhW2n39|)udxWwGasa|%J~mU{nW^%pgb6MDKvB)BK-_~-4%}~|
zd>Qctjl^&LY?}D)?vpfyu5~;WgR6~yt0*$2h`+NjFe+C|TIKQ6Dc3)DCJtmtYbfNd
z;gV6pDVxobIYO(<!ugeu<=Sh^VkX$deA+Q$x?i3_L{njzK)UHJ!0>?<odwGR@0{UJ
z99ylgZ3$P@722J&ITJgE6CW*p%W)dkb1j%X&nBtZ>F*&DCsVAjjY3X=xH;63EZryl
zRm1^Ha*q}BBS<yc@UFvF_!f=N@Kvnfd|Cf68(xXKD`>LPKWuOK?sqcuOR`&=VPG+O
z+1{dbV$;~$Y5KUNj<bnYJ=?6=@RnSk{AsrPny~)lmMUJyz*>S;6m>o^4YC_M@PQ4$
z&F)Gx4aCfe^OP+pl45kcRS9Nxsj&*@Az7SvTO`Fz&Db**;g7DpHx?pIk)o8_Nyy~t
z_SC*uWp54dQJW%Fi|kFkWgF-o9cUm9tvdhNhKROmZ(+m;R?0V<|A9zyNwJzi%+48G
zRh5xsmtB62A`9o5^Us0M=)@DE#uX!apgySss;OaQl+DQ=ZW5HwjrA$}zf@6DQh$Xp
zwZofO?=#?W&>MW>nQezMl-9UGudXbErG=T~gSY90lSOjML;a#LqQ2jG0q3jukCy8~
zIi;;#Q3d}8`b&^IgVoodJJ4Q+Mt1hupKTpnDfYP2?8r7Ju3?@LNo&-C_QjCYR?7nJ
z3&sPN!^wGH4I3dTA6|lV<74`aFF<%X?gq1B0Qp?fu(7xsL(1`!agt^fL43kr-&n7M
zhU)Ex2CP{)dQdxudYJF7oKB!g^SMiBYU0V*oyY%c0n!}JpDNHisReGN<>YFNb>|9+
zrTIvcVywcL-(-X?{J&ficPalNw?aDDdQydPspd1N{z-kmyk3Sc^7(h2!kcbD*fV_q
ze&)T%L9&ggthgO~$RN+i8};*=$0M%T&de2}zGi9a?po`WmBD>zi}1E7+$>Y<Z(bpn
z)t>yPR}VJRaNoDwu9ViLRyM;FDCX6?%p_M;KM6M5+-a8K)O`S7D3ROKzkk2U-=Zh9
znmH@dpg-TL3`lCIHBeMYTDj~5DZE?{Q-m7H4Oqwxl(GQrbsTLAjf{v0WJDM*%U%-p
zoFbShLqUK2f-!(qu9(G?x9WRaF@l&%H<cwK=^zGV_%D=>$QoShHj~;6EE9+x%)mpO
zlxm`ppEqvzlai1hgXX7fzHw1ygm*q;=)n)ZtkCI7XV2tc_9bZ2`^S-jb7=j2YGCSX
z81z0@{vs?{$GFtw(RLOFNTFvLN-SfHXBFk5Zg4qUHmI49K@B^ZzGSdcRL_zUydGU_
zJSJHZ;TM6fe!ssR1CfFgMib>nh$ppWb$=Gr9)(EbC7k;cv2*tXrc7K#lIlPs&L$zT
z4gt&42moZ(%^6j(2x2E|T3d6`2y$S}LCu*iq3t8Z+QpsV@2@2Plf7PA+JeNfb|5Al
zC9}`+xB_>MFCFjfc$>%E+sZi7lC^QcKE1n#iv8-s*6<Om8x}TQR`8uu-%Lz&vQyu`
zIx};%*(o=5oeM&Cx7=dm$t#huD0hyd67~BN$r4`PX~J{B?g}Y%5@cK82p+8@Ojl9T
zf%VfOwz;pRNQONmxrMII)5-Iun*PZd+>d&)ZQfN2_=-8Ki~^k#hlmg?Xo|LX8>=t6
zezUZUd?+r|R)hT$iqa|s-Vh$LClxJqmUxqQzg#lA4D^p&o|=NKQ*C@6GrNR)^X+OO
zqh*#-8^vPnt<%cY3qa^WLn4zi{*ZP#@0@=3uRV54PxM(7PHp`!>%cdM14A(l{Af^e
z9UZjvr8yS6Q85fuPSv>usPwmApkvA!-|TEWUa&iGjVB3`!Qjs%1ySSeemyZV&zg-k
z$p|di*HmZzwy26+Ma>}=`(tbli}`HYREnO*!ECBLGGl-q?r7fxP(+~$8u48h?m9f8
z@7pLxwIQMohoSz4ms4@xu)y0cBcp8X-ImXyP8sF4g%BsJQP9*tQB27VG|p$O`IBoO
z8mUn>8pF?c8J6=Sg08#3I)fi_U;gjoPcy}0c<$a@1rtU^sNrU_8(V2M-|X%ZL0#$C
z)L!iK28WBAmJYsY7WujSjO|y*<K1TEGx#R?7kprq+vJml9nrT1RSlQEcJ6T{f<_fr
zRdB?_EU?B)WrXdvR2w(tb~d3X&ejv@{4!HhVGGNsz{=Zeub@-sOReRE!NR>@Ef|X4
zG@&14LkAHfN}HBRJMie*O~L(nki9bsYV5w*f7VnF-x^yACVSNMLM<AkK6VB&n4B#$
z5@h@=LPLoyl}8~R-V`>=BYS|m2h6t1wF7_pW(oBsx+4>KnNKACS4@QZwnC<}ti+nW
zFd^)l?#btY2^I#TVEMwBA2Kr9_=rJD@F783v;s#)_TNALj|KQGI5zo&U#%3M6m|x7
zY)EYFg^OaAP{(F8YTjng#cj8esbTcf+=J5H!A%bPA!8jV`YE1y%DIcmJb4CJP=Vr0
z?~Fm;l|Rs_&VsILaJ`=_q&Gr#R~elGH9T3jYTB_%GcGwS)o?W*&+}I}+%asP{<m}j
z1!RQd?%l9ue8|u3t-{8{cz3hwUa|st;ldd^X}<q8fRm*yS0Desu$Du>%PQ!e)#Lg7
z?Pa+qK-gp54d=pZ$_I9!w5ELKU)|P4EK;jOTl)squ##5PdbAbK#?;XP<*Zw))%hPl
zgb;B$*Qw!Qg8&zy_L)mnu8Svm@m~^kH$m%=$YRj8Bll-&-nw{FXa*E`4fgRXe_8Gs
zp1gC-#YuU7@A48vtoVe%?xIX9pzW@wXq=_oKx!t4o4CO(c0IeUeP1Ysb<()nlyFjS
z;pOWFB*~SOsZKs1>Qa?j5cQ={?%&T+i+mz~Ex7z_Y2{Eofv{$EPRUc+!4;7jzmF58
zfnt~Y7E0d!%=?c`-})Ex_tT8;C1~D_WT7zL>E$)4{Z1xYkUn#H1|Bw<f{_uW^U!ea
zX^EhGZ~f-`M9KSO;JWZ|!CKIu(hYqd8#Cf$U;}da5I)34gO)A8<nTkGa2-jmOScMk
zK|X-q>CdBhl${;tRQQF#g5_B>j7HPfU2ZuCr{|oPS<ga~Rt<BS4iA@XB*<Dt2QK3{
z&5WFEXD(WCcI3>@<i}o^om+lKwhT?MKXy5D^=;SIrN3QTfal?*!v&<+r<j%jNpDG3
z_41Mx3d&6)5s_^e`{5`OX0<WZUB&W|te%!!<jbJ9H5r&<<m7hMH<$lMJNvIIOWv(?
zZhlo$_(!G3TVf*w?k&tspQF5KE2l!1)c{WJDa$akbBH;xzs%(^CJloM_n3HlPCbP!
zjI~%85czdJOZ}YuS7u~d<-`Mg^pOrmcAD`K9dR~$_7Z5!nI1<1S)?lqNtN~h)35e8
zV>(EXBiEVpIp)oFthX=xNJYJQN}0!eueCq!!E5>wR<`Q>0bf-)hsqkOs53@ya1Zun
z?ufMUv8t$Gs=>o{!@Sn{WJ*B*nj?$Ir3Q*nQdnqq5DM}$1h(>=;KYO%0Wb!O-n_ae
zN~@#6l{nDKNifFkuCQS}!GiE|r++Ic_g?u=`ZA<fd%3EjB(Ku-|M+nV_hnNo6_{Y{
zti_p+)@Tz6(WvsNwTM%yDAex_tDr+c0!k%+{bFrO#hN)W**PV^7}1R)q!pfU%5AsM
z)7KCb4$V7qR>zl?o5+{)CdNort(B!qzJZTvw8>PHT0no;$2Sck0;$|kbTls<0p<8W
zQ}^ya%AkYgT`p$R?CG?APL%ZeaDkO(rjQ8Asb~@TGlZ<Hcvu-MVItij!KlDyposlp
zao8DSu737mr+KsCnf$p$?35$VF42YOw*{Sz%P?-ogKQ7<Fw*hy@y?g^i~Zg}sMnR+
zH~jwsG&u=xHYcqXPJ;(Pl*3X=-G%f0=9sA&x<$OmuMZMa|F)m2Cxl;NfR$n)c6}4g
zet+$KI5+Msgegt=hK^fz=SIE~9ur6w7xT;VuKk=|a9V%-oWuW2=;DrbGOvBOHe7fd
znl*~q{f!+oG9%J=&0;~Pzm;65=oBLqC~;NRlRDMh9iPuF2O!Farlhp0!Lu6hR#8gA
z#-hXTHesa2!^XEG;+`1Bv8{M?%_CNeGD6`p>}7mCqU^cY`iqbbq#%eeW~Aj^8HX)|
zu0)6~$A*1L-yf{Yi>GyO_8kr_@E`V@;ngC>k{=?T3pBp`-t*A;?>xQpdiT4_;OYQQ
zOJ_!y;WW@s19WO1j(egd$I$>HTf<x+SIj%neR^nlV{+K|U|!D2TJ6b8Lc8eMA9?({
zW^-skIf+j(|C8z(>q}r$3c83D$JF<LzyrVo+kv;4f7II@zgD#2Cs9@h4NtG<hcy;~
zuumz4g1#kpNQmq1jg{At1o=gQgQ>}K+_V*)2g+6uWq^K%(1vj}<8xp_(WGOU?%vJ@
zo8kJ-IJhD)6eJOyvi#gD_1h;l$fZbivA**}7w_lAm#vaz{BCr}gDyYDm%Y1xi#9@P
zZ$gljhSt_Cc0nC-BG#vdbi0eJ0j-YodPBh${1ck$ywsx`CTSf)$kf^tZzWEh%>O4L
zLS*tF`5Y!HJx+j0E>U%yX>ityCI)7fyY8>TO#Fm?5=_0t=%rbR)Mu1MfJ{wUUc+|Q
z%*mXynvpbbW|N<bCM_Ynn5mydI@63kRy6Xl@~-H)7r)tcx<j*rckkeb)dea!m-|Bb
z8)|HG{gQmG`;zo;BKUP(?HoJzO|kONXeAG8ic{vH#HQHuk#Of@*cZe|6mCOEQR#7E
zY!8uaE70G;rl_LRmoQdVGVn3(=Q3D~_bvl0?5Ei^$8X7~jZn38w9P=lY}8_Cgg3$i
zuFL^g5q_1wz7dY-)J9zQR^dV~sUwiXATaAV{(z91x%?o4Wiuxf@ZPz-T(W~OUFW$C
zz~cT_4U0d&y6}uX#1UF_28dPIUmvgbl*OpXwhH^(3`=!OX@c|1>!0aK9i0R`nM1gQ
z{<`TYEn<2T#rjf=8D_-FLUa^O&J~fZEYt37NC40=`DKYuw=GI+2vReZ_GJ{(Z3=<n
zmDA`vJdEQSinnvIJ*r(8?h&5Tw=>FrE5YH%hhuKZiR{XRiB3lggrx(V;u>1?-5fq-
z@dL@`Q`B_#PlOf~G_w~MRlAPx-Dyb6zXj<N<=U*Eu}{TapTm7z`ra0PAHp5r_8s89
z^xJ&!EB)|0S*U)UWqw0<@EX^BoUjlY`11$;=H^Dz_roS)f1Dm}dBxa&tGA<7$m_Cd
z(Qx3K6YOAl<*L)@dH88!kG$uq$8_h-V*J%pgH=`~K40?i%sA2tPey*h8&4F6lKGhb
z@!>DU??g&uR*PSdYp9C|#$W&bVZK@0J#T0Z=!+2E!+PiFVY}HmO=UKRl-fkV$NamG
zUK`2pV*W$(ZMd(*7n$!uLT!8C5Ro4`xGGRF)22T3CQ7(wbY)Ko`q^=ins>IW2%^O`
z=XN9Dtm2c#Q$MP`4RvyCfKzS{ce*VCd<nv>xyc#C{na;m5^!^HytSsBT?G#uxy`-w
z3S#o`XcK1ZJ=%>7sm*-&#R<l}tuKp=<kkz{wqEdR30sR;d){{gcA4=vgND#zW$m+S
z*;Q}q6ymV=s}3kgp{tcZ%Bw@nYp*_?EKdwRnjbjqd+=>lpy~GH55W(+-6p*L*#s=!
z&`@cO8K`*XnFW)ok(z5|f9of~Q+niV&YX)w|6ecx)Gs?0SCQL>SvQ2Y&qJGgwJqyo
zBOV_tD|{tnu_E@o=H~`-a@gT8+D9}o`I)$l)fMU?za&YNRq9z7(#pL~t;4ILIT5;8
ztEb+v7yX|eQ14G8^ZY(ux(Hj3UU=^la!WAOosdwfseIoDVI50zr6_{>`G!L&7m;P|
z_$@;V9ZSLBAITC7ddY%z*$}-RGS82^8k@bu<l@#uQbk8Y5~g~TzcRz>e>-NDZP_bz
zc2j>yTGB(qYCoh@cn)NuJ8<xx%xZNsJsea<oHI8&J<hSxpQA9Sw@!UchgN<rw<76e
zqO?_RUEthJ*sO}^JOC-4=Yw*XtfV5-Gr~!VFzAtM4x~vRXPY+1t9O{|-0sdBa<`bn
z)C8V?_J|%0HHw1kAX8jHpp@BhL4TKLsmF-B&Ca5WolYc0=q*CS?!9o{!_D5L%4@<}
zsu+ilMa*x)!P@?h{RF(8o1DUZ2|X~Bl<^f*0^R71qOnISc&Rzy*AcpbyXdU`aq-nZ
z>!-9{DX<0z6+B^K7s^_43@{CDp_u3=&@jOA>C%0<uWbMuc7X1HMV!qnx-32JK4i|9
z61*PgXATWVXSKQ%OL&Cp3{hASW=6cFgrX2pU5LHl#uy<e<I>RJzQ}vrZKBB!F?pm*
zjJ=W)30j@F_M3^2fVV2IqUhdXjPZ($N(@Mqgc6Fd`__?XNU_Tq9_>~-B}6h-(sGc(
zR)m`#FLl0upvHX=sL^{qx!f;nPlzy1t|7@QM?f3u#1QMOF^qo#JqZU<i`$$iR@AhE
zO4$%8Ee{m9>z#g3(B*sR;j-DWC>U|+4}uKa8laI+){cDt94!^1{naK)MRbXAQcp1z
zRLbmmQM^bvrw8x)yC3l!JYI$iBlc!{Y<NwcH0i`>_B_w8xh?AZ9;@d5yMu(uvAz$?
z))jR)#^o-5h&BHF=v@7Mv!+BCxA|a^)-)jh78OexG*Uo-{f+k<mou{6u0S1MNvgGg
zMSA%+TPulEyKe$Ma@kK|QzYz|nY_82hO@|(Fi9^dQoi?(zvdgw+R^IRCdR@}Iz6_f
zmMc=H_$L9&dk3i-V;vK#Pnj#+*ExG0R|TgC2kQ$SlpnTlC}|CXLai@SGYfW1lQf;+
zwQH#kyq&enD#!p0WyZ+rQb;M}p^ih(ED26+B>%p7^O(JUu}c*DZYWOGwslWn9l5#m
z-axUSg6pV&O5->=J0%Nrh~}LoY7k<INS4Bs&gjl-O!=mP0la6wZxnO)s0giUYoF>q
zO51&p)xS<Fz2nn;d2mr#14hG?Hz&XR)k&%#ELq@#)IcmL%EnirBtYMD1j9enowWn0
z=zE0zd1>V%9AXluy2RQLbCq=+@iK{m9!aGacuI32@N7DG|GEjDbPmkk>-REDoV>gq
z5-v$*y2;n|5A3%G93aDEKl4gG^UA#id6%U*K-=0p!_pp#t7CyB0i*B(b<?NTU;o=3
zfbb<BuV<&c*E2WgsyqXCD6iB&WX1aXl7QIex&Tf}MuvtM7xG_Zl=0{x=$;F9Ih*qy
zwB-vMew1iDjW8WY&KoKy5V=+(F{qG$SJW2pP;qR~-h}4}p7RVe{oh^Ta$>Jia?5Wn
zbHlYMoMoF$S#%2;(hOmudJ(Irp-e&i-4*;KM?{2ys~AF4y2W9=^-?q%fJy9~)0j25
zqpUa6YJ8T)B2C-9Q7nkPoq1|dNTTD9_;}M+u*WNv2l0y=;QRZh*wV}61%>w7cXb8H
z7Go|ISyOKPRQw|fxq+<k9f$pxjof-wW`j$YZo;)uw5G%h>5uEIMOiGDM=#;Aek6ze
zLvKk!nq$r7JJJHG<J^GX@42tIF~PY42X})uS62i98855MrSr_vO(Oik)R;CAB^b$#
zVxcO@?AioU23k5qW3IkV-1ibAL$vzur@5KU(?Ce(z1?bzrKK@M5nP#3;~7*D7)323
zH=uT^Y$?awu*PO+s!6{!Shi)UIRosfc=1)KXJa}l#;4VzR9bkIwQLIGn8RF{F8BH?
zAfcjFgvo&945>D{@uuM^=0J+<*J%%wZJdOM00v~X^AvS=LlVl>(%?C!rR=jhbSDIG
zL_WYls@{Wdc;&b(8-wtaN`OUH8xLnNbdWSHMi5{Vc9apIF(G0_2(ip<LI-Wz%R~)H
zLX5}>8Uz?Brrf7tVcZ+6{+2%=BLs6z-~ExTYeJrasx}s+<*hR?${K2?tL470EU^W!
z8(vQ^O#Xm#Z~U?qkqc$W2<anI6>1mBr(8a{pRajs*cm`Z96Cb~_p7!kzjInNPtC-v
zfY&W%J*BlS)KY(pyMJ%dLhtq&wQn%dO>2J66K~_R4@@zhW>%vv8p4y=!NLc_ugbo%
z3&CdsY#%>;54ATH|1Jk5d~rDPxA%S&Zg4rfN%Bf%J4N-#iL&#<vAMHIv>bWXx(TKZ
zxC#1qZo1>#Y4^664pxM`qoB)c?egO%4{pN=!5No|eI2cZVHinW$zsYyY*J$j_QAxU
zA`Y_^<(u4`FBc@-(DjjQJh<mOBOb3`g_~aK^>T%sCz~I{TW&TOg~f>N$u$c8IJW>L
zf3gzzAYNBwkTMcNyu|cKtOZdT`cfC>d36nTbyKh|SHlvwN{l{V?;eAl{)vKfkS?5`
zOc8W!RmsFW)-a!UxiKdrt{u}?+ue-6TG39YTnL@^SAJ`JkA#d}3RE@>3c{&xFYj=<
zv4%=}HBizt)s!tk(tQ4pOUwq4tEgsBccm%dcx~z%=1{tiAV4c4J1{2FSJx<iLT7DN
zg01vW=_E~>79qGWOc^{;C%`l&EB9O+>wE??sK0efzJ9**{a_mBLm0Sr&6hF)Gk5iL
zvtCM;J6c&`?}$uF`FXw|L5xIv{_kHUPV`%D95Vj+e<@G+^rnysqyo{96f(u)2ofaq
zvQVr1`G1bA9`ApSEZmepd-*V5YY9eBun5GOFHW@@HlI&|U^p`6yj-I)uE|3w*5mg8
zX@0e;7|y~@=l1MBr}BR)cAQF(@=JMU!P*2i|9+sBRfOn&vQ43-$^@q52Yx0->D=9P
zzLE3)Jq+q8>A8empHPlEnHc_}jtElDG<mrhtM!42xEa;Jcup_=oBT;oH>#3KCGcN3
zxBpJTKc?R5p@FqIPp-|bvxmo&Q}$D~DCxZhrJkJIXHAgB2=a3}z(%kaBMuv%DPQ0L
zLi6$wde2`>>-fZNZBgSfeK#L?B&=AUD0E>aSLeK*B{kc#tocSftaDxer(`^)umpN%
zO~o=iE-1VD(rJf_DJ&t`2i*3Dm%vt`Ii!qFQ*mem@HdcrK7|PUYd!n-(l^vh?NCY?
ztD{oNm^|xW7+wkG)28INEDNj{K7`Ao5Om3~&sWe$4Hy5;x~G&XuOXAk{M)JZH=54`
zGV`ErBARFy21Is3V0FwVT4;WJ(n=YT8usQEKlhfx1~Oz(xWJ&+efCj~E_v1o&gy9~
zT~AB9iGdSZgs#qGz3;LDq!5hetrBkq&Y-i>%iES(iZUuVnx>+miztpl9nTJ2OL#A;
z7USqk`TH$y6vq-_T|6;4CE017WJ;A~0h_V{Hu;XD6nnU;1dp>P(rFZ4%A*QzH%ye@
zptVqe{c**4U!nE7s<sx2g`Yo4<)DYjezh6t7bH8D{cuMe#q(rtkH7g6vU%}q_u?b&
zP<M0EkKl4>h}DpfcYpU<#{c+o%E`z7Lwj{<4`G7nqgKG;YT)&FGWTG4D7f=}_rd|W
z<auDqXRYp4v(xXX_VFT{|EbnRSUQd2=K5;WU^dMp^Xjb-Memxm6Y%!7)^pGXiy-+6
zJ)t~9N==K+^r-kNExZ8jVMy}fqpP`>mh*>mkQ>Y9r54WORr8_8(#8jR3&q0>m*2uX
z=EI<eBTGAQ=?0Dq2S$+*5aTj(T_)~HE0jP0ATRk3Sz$?W+D^OL9+n)ut`jHu-Dy2@
z(>I@FuTnonTQH8>a^WLC@;Q5noRM+l83yz{-o~>iGXXPYI(Sr-!*)-p>upfVb8g#E
zA_pxWHIsF*#y`?VQ+Q~y`Xfz>UpP{{j*IT97-EzVj2N1Ec-y;QU~Fe%e}wc{b_b1O
z2-wt&rGzOOGXjdxv{EE9jP2X}srva%|9?dNWmubSv@{I26xZSo#ogTt#ex&uJvc>L
zT#CE9ySqcN;@03^C{QF&ytu#Dy}!Mm?>LhGNv@pdT5D$3OrSsWjV@723|cH+P^Zu9
zR?@9dz!|0US>AF2UYflKEFTrhUwuGtLB6CJonFY`QqJvJ=na*`)6^19<gi85qCcO~
z-kTHq{B`O7A$uA>q_=XN7l{bM%#2f)?-jVm(4V{Q;FIyw#?lww4RZ=y@q9W4<ur?S
zBxz$U%wtPQ*|Gu}k70F?z^rI@T7zIsPO$=0Olu9MBAjXh-g~A%uvVSzeEQGDs^dAy
zL~d5}=wPzd-;&NdSRX#S)?khD`pO7?J=IjSoBY62Q%2{;mw)g5qoL-zGGJ{uF*zFG
z2HG4pkfq!YQ0mEx|NKl5Q>`FEqw1O*o5{lG*2-M3{CR)C$d5TuG`BDG^h){Zy6|m$
z9RnlETokj@pioX_lYv}lDL#nXaYY{E$aefg@)cSz2bqV1z4K9;@OComaDJR$8NjfB
zn|k#2`ofxUJPE_;rut>IwV~z}S&<l%RgKEp`ah^jO*0H`zW5d?huiX0n14)@MTj+R
ze&=b|;VwLs9OYD|N13`Si~9&?O^ZqtE4S|buUORj-1ey-^uYc{_fn56*I(S{@wTbi
zgOV>jVc0;B^AU$)<Jss+^M|a{3MAO=(q4uRZoyZE=SpyG(l#tA>&&Gj-St&rVB0B<
zp=#fCWPp?7qQp{7|DRPX2l-}GeK1~7t4n`H!*QO7k4f%WIk8<@7#;NE+%)fe<ERo8
z*1wlx@k|!@DQY57g!~?e!2q;&NC=&EQ;Cky^BRakXPanKT2m?UR>%*X|FmMs7Mw(H
z;v+j?G;tPhqKbYHxk*yp^_=-}<lhZ3Og*jvw#Vjxv8l(~Eia1s`T25_Y@;q1O6)Tp
z+>c$~``^=Uo;OD;eR&r@{${jOJ_tE+oc3)!{)G(niQY7Op2pP7*3T%DpO^i9#Qtsx
zQZZWUAaVB{|Mfd<{ZpFlAI5@xAHmA8iwJ5^#8#Gyv>97Z%TO4X!9IkY$$L+!+@g54
zz_{LIiOY~3Q_$TZF_rAE`cSX-*aV|>V}7%Gz+1|dj#M)<^*Jq!&XA9cfsJZw%;l}v
zA>S;78qp_1R&G~#mP@d;<JSYfrqt2KQI};=$)?hV7FvtOl~D8yQ2WYozd984@4#O3
z{>+{u+L?Xadb|sqmUVLRk6~boex?$b7O-{FezP~yxVO7`5gY!pA+tCNW8WAyP#*IU
zd1aU=jw;}E-q{+@y|iu;5s|NvJI?y0(uIwEALVQl;oG04{X-wzl$FA(F)6Oj7PjQq
zO3Y<u)282I9>28am)<=sdrOri8Au7=k{x>q8J><vIDE(m^_O*`n-U5#|IvL?oHK&j
zAC*y7o~4WQYwc@Yrfs96f7YRj@9O4YChYr}iT=Co|D>(|Q=)k?k-Ks0NJ=ZSO70c3
z>+*CP6H~8nV__f#3uLjta^ITJkK<J-vjhFtDeh{Wd~&6GfEMvWvVYrY<X=ZY_CZx~
zW97vAbUA>(;%NT4+*5_81`6}vn{x<{e_ln#?|e%sSpxrcx;Wh)P@mjCtoCu|RgqN|
zuX7SUI#|+mRt_DCgavqvZ-x$>*Y02X_Fg<dn>_l{zE9^>W*F5bbn1URM#KE9iBVC<
zpU4m?I{aZ-Iy=+%`R5EWeyu#XK1ll_We9_cC(QUy!na01&o+mGko(bf$2XVo#IB+I
za+l{2n085RL1h?d!8x7Z<NIeP(HRhhGPzO^x$Olz5^=m2J9TdHfz(>^`>THnIZiSZ
zvCso76Ro&)Lj!Hudx&uXu(s;^$;_FxoK8U@#yy66aD62GzW;Gkg~vopte4*zdPima
z{R)4oW0z56Sz;Ou^OaEi{P@;c^_=zCi#{jR<xZio7=Js5Za&P#OsG5Y-|#RMR{b91
z+cm))XWG58bDDx}MSdJfmY={~ZQV<ekdl_gF+DDZnPS?uuR$~_3z0`t)?cf#<V#?^
zcKzFHsrYF&QY;0>1|}a%GoL2=VRaE*kHk(2AulCm39P@ugRim1{JD!4(?q#j88=VR
zN)PUR7-6q}4%2~lW(kbYb99G`l#~?a=g3Qvd+otQiOT>ry=D;zkXjEoc<+H#2YbQw
zdpRsL+4UmC_E2?6>+qpad3N6rdYHi{+77)i8b70a^nV%sN!kI+5=ctX;sGHYAB0E;
zQoAPY^pIH~B-XQw1h?BYTrU`Nd<TusReUt)_no&kv;m)F*uxJ6*3{{NB=v<;-osM}
zvT#n#Rw9^^yL`xq{p71Uqs(gycJ`$y?rn`@s7AEPD_2vXDxtlivvAa1HzdohMgsmd
zNVnZ_hwswU27Nl*-!Tk|DUk-Zy{+CMLqd1^B=T=XV7?%qxTU(twS>*<Wus5G=iY^a
z$++?1?Muyt9l?{mEn<{<oox#2Rm`i%kdUDUKTT7_F0g{yS9E-?p<qU9adfR?_lLI>
z%fCq9alvKAQZ*^HxbNvb)9$O~`>H(5Y_*rtydPUTjueLl()0iLAO-w{?sY1kVcz-N
z|M-~=+VgOI_I;;X*<6RE8UhL#{NIYX9mxrh)rNecOzL1Y5=DX99I;z=2C`Ev>azC)
zj_{`HMa+KDe*LoMx~drA-o#zMu%DqZ&5~uR=_@7@xY(XaB`rWHHO(1J=KP7aroPxQ
zp5Bm{cEzL2rlur=x{b%T$#7nEhiHMJSILArIcSeExNhBRyq95z#u6zjA_q;ra#I#V
z&Wx^t82GG@Pd<p(F~z&&;}EHWqv4lT!hufG;d9uXLIP=Y`H&uQl}@yV(Ynqk!kUn5
zshO_!KwGP+0E~Y`$i0;&smao+2Kyt!zm-|F<IF0TXQiRmc`gwHm`k<z+mdg7K&?jx
zM6#nkLUNA?(Zys;g;ahJi&5Rdxq^w4>TtLnQ~tW97wvGN=CApXf+>=D2dG?E71s>S
zMu?=7u4=VxKp;b>mtw3Ms_3>Zspy3L=`$fbq3usM@xL<Z0-KM;HNPLakZ*IEO}?b=
zbQ~(`7t&sd0$R{Yq>@sIp|FBQVqS^Hr46F-V&b|agq5^X^35$re21T6Y-Uz`a5Gt(
zHMh>C%ha}F|0mFRqh#C{t(UE(vUq;X?J5U3!${nqbzwk!zW?{_CGhzeLDvwJ>NE9U
zKerYUM?$T1iOP1y9c>11M=z@`VVam{_niBU|1P6ZIMnfG9r!u!)-pSI2-cd;|K&6i
znzx*D-lFFXPm}u~wG%$pk6FKbYN{0-moRu>qfiou=Jy@60};<<7?;A*D5DBRnV;_#
z;3UBr(GXbvAEm;{=Xb15>{xExjv2LHF<@oeYWsq(aou0Om}9+@E-g(@`9M1-^k>F`
z{^np0vE~z#q}aK$c0(RP4IM&t@HfYk1ls!33!)9-|I-4bChEw@>LKkIT8M?({a_Tx
zIfn7x_@O=f)tC4ktk*p9G)i4;x<AABGE|0(yL`|)OGh>}v}xmi*Vg*CCf3!v6XIyd
z>G9{Rf-e0f=3mt8>uw*fe~33@q;I}xR#<=XW8HMg#;R2Nml||8(YNxF5{tXC!NF>v
zn${R@DA2UUqWujlIc4!~{z^f1|ApNxzlu+h)1B<h%20qbKNSTlAtlQ!@3OgPbN3gL
zhtauP9ShcFv*OvlzYE79TVIc!<IT|2q6X|yVyIL6fA=%){#tN1=Q%G{AZgJgAT3vo
zg~n<&^H&w&X9bQ{G~|%&hbT1E62F&!kF1|)PRFEsFnH=7(~QYv>ZQraj^33e9?0OL
zNw-s&)IuX;#(Mp6R#8!u$|g+fT~(|SO~?m<phJ01e8o0iIDIIx1-eGw_E5nVKu;SI
zh4l^yozu3w8+SOKovwPu8^Iw8Ehkek%nSbHCEl9^I0m#5&!AUFeACdXd0v<~-#f&a
ztOliW9bIne4{VH%W|MbTqZ{-bnj%WlB-}MI-`|$PzippdzL#58Zf@P?!Z+!b{EbVc
z222$e5Ic%a_ZoV->AdKRRC7&DmEWc6vNSb@rPVZ_c#@Ptqr^1+ljA_b%*-+dB~HHK
z!nN_i0dTUR-O?!C*G&sWu>xYR<1EM7IKQNtAFnHYT@Eg{c)d`s`9hQJWS^IVt3Ki*
zV8EoSTJP7c&?Um74K3X!UDBA8;;`1ufC{2uy~b4gD1)Tv^oVv@<5|&ylVE2R>yz)?
zULs50Yb`&@GmVs5_3kIyOg38vyK4&N`Bc)+V<s&4_LBt-NR^OXe{A|JT*R3=bEYU+
zK`Zr9Vt)6#-3^b~Kyu-R)4PV2mM{H|bG!)5Tz68ZM&H_dP8_>W0N2HJKfBt)Ywz;c
zUhzl6y1s?CCLzl;FW6m=x1uz|8$6v}&I02AI2-Te$w1J*W;#@i=TM1+j6$!hYh=LC
z72=~;t6i*J{WdlF!CdMkb!%lJA~hC!PPm1iH5L_+e6EwsfoTHYg!f}tkA{d&>p7>n
zf0!9bgdk`H7GN{Yft&azGEsjN12U?tt3u}(h5iZ7atVI;uawFCk($ZrVQCMqt3H=D
zlXYFam(Afk8~44H#?WAT`vzLY0is_-KX*+noRRzcLq2k<M9j|wUg_6_og%^!H6@*@
z$TQ>)u^xncduw41N5n%+s=AWK*1)$yV(+hYT_%!mi}gI}&s<rORHg7P>EuyQQL*)R
z4=l&;M(FmqIpFd3(Vynje@{e^JGbHSc$8!9x5CKN=(|_a_!g%RHxxQL8L0n?q!h<4
ziSK{TE~B(nWhgU4M&<>5?p+bzENfw5?_?6->(P2uJaY@)Zss|9ymKz|!dGur<Y%^#
zf!kxQum_)1#xTrsvSTFA@vvcxPf6XjoEx^zm~s+dE~owp`yI=Q#eUY_ywMpJE|*ut
zD9BT{N};;45#=7#4aXrW*N$rJ_N0@~-|^__r0+u+?|Vl`bn-}e_cX!d*~}`!Aqif-
zo)@p)|Cz~SJ(VNm#y%qYLj;CkLA8iSpn}e#-)Py2@8+tTWIJgm*DMyAypv$HI!U$^
z%$GMONMCAY!ca<#CR;ivJgRMGBON7($}yaJjYsfDe&#xB!aL@2NMSAtA(9|v1RFFy
zYo_vv#T&E(B5s0{<d{cSR5T8?I60wZg=A?ePAS<h(ewFab6Maj{r(H5B2f3|y=uPK
zrRP*jU8#GVO4S(kxB^d6cBsP^aI}a=3ylZBJE6;v>Kbc5eKW@~O}n8a|FTtfvZHwt
zG~O0-0w!<eNa}-H)VGQlG6JbB$uNLwU<^{zUxniNtZ{|AUnNf~+n2R|iZbO*&i=4!
z+OsnzHIV2ZEao@~n^G1Mt6N!&nwsPlRS-#I6)<-)#+O!bwxlY>wqNV7jrBZwyJoL?
zA{rocOuDJ~-5=}rk<?Z?Xh5(@zHw!SRls6O*`8trUDobog^S4^GNo>ydO|)pFcU==
zJDQongpjN0Vt%laqirdi(VvqWvZ;$aA(1&JTnA+1>d6c7)Exppg$PQs@DpY=d2@+0
z%1G+YLPZHTMY|gv$!s!DNG=w=l8%0nnDrE{o}nSUsh{67)AiTsJTbH%i8OL~>HHNv
z999|}4K{V%N>3L<z56T^eg1L!*8{GP`}yaYf4c<ddm^P33FagGax-lAB0Q=njFkYm
z|IlT)@g%7>v+xT=;NVJwnxVd&vC#XZIbqQ}$oJxY7E!3~tRh`vT8wp?Jkn|yuQoEJ
zlF9T>W&+ZzP*zPLWU36!G-#$(Co%9V5W(Cnod%9NBlVvSt6wP7DDM26#GZ}Q+D~O4
zvoLyw7_#cOusnJ~tD5u6FsHSeH4Y0)%5&|$M_TIpQ2ZVf5@TAOAP0L^{h5H<L*ZLy
zFLhADAoUt%C^GP36d%x9Y!+aD8vEThTETnO6tQs<VT@yz1N&K0mdcnDtRxZ=;#8DK
z4R5MJZ26r44DEFqc*k*Xoc8)}4vqH>mf>p;YZ>aiGO&2FGZ@qgl3}I;qnI;s?qIx(
zKhzt~c9L~Jre``n(X9DkiYC`({Qu6z{!cIMpYOWej5#4A*yHaxaius}F(?ABgSy@2
ziGu<|zanGr5|q4QqV3Uau223?Mp{(V&_5@*6ps+Yyi=J=lSGx=!kahY^>3~zQBi-%
z;y;J0U)Z9H3V^m;ok+EdT&XY)12xQQ>_C*K>YTbTS=z=nvXwS5NV~>{F3G)Y+&<c+
z!Tk=S1L>7#PQov(2uWdaD=<_iJTjWlQX$bM5D#ECw?&Unq{;r$aiB(qX%t7bx(rJZ
zxJ<G~4mGDRQ~a)5Nl73b{i)mSp(&p64P-KJZ&Su6+M(6J#m6eZhvYMUY2KE)99QJd
zTQmwM{@aH{zo9$f<f{J1yYevnNodAiSy@X`tlmJSSwHaNQ;fX~99#3AQXVXEqnek0
zTP{b}B^<O5xFbqj9KuST(G2qUEKWbb>5lO#>W!WEUS4HfgOo+0JwXx{pGBzK)1(z}
z@ZtLBnNv#anb_}?1z}zb@+!!>j-xROgO%#zQZ(kAwZT_G=yZ)sK9JTuH3=&n9&-zA
zjWhT!q=_XLAq5S1e5CpfsRk`?uJLaFmbkN)J%D~qXS$VMd5irKEURKx$rpzLg2w!-
zTvh*n$#1&<;MaQ^uQ@N82V;Hhi|p&UC!L;r>C^Q(x%%Y$Fm~c;Y9c^AAEdr*!w|JA
zQ_|BTUMnOFT&tD9^P3N<jVvU#58))jS_<8&bS5Yr5K*X+WQrs(Ypfv;58cwP(z9h2
zwv|4`KT&e^K6c?3P5u>GR>cORmvs93QDpz`FnS2=xMjq9`kqStTUwUSZFiuXP5?_e
z65nlIeCEBW`;rQN=9dK(PBwB3HW$zDk^5>KrP)I{x!O5oI+`OVtiB=?+Ff4v_*b)#
zw*26YkWeL>Hql?5dxY-KM6Wly62;q>*6X*U=_cdl+P#_E-cyq|EjFVhOod-0^Nmm=
zzvhAV##2xOk94s?#FqXOSxD^)?yb~M7@v20Ehx7hT5rR(en%5wqHZPjY({im{L*s>
zK}~g=kFt2ze{ljwM>UUAi)YgMjE$pXID0Rzqp|)hzH%8n-J$ybDo7sF<(sn{!Wc8N
ziyzl47-gjvx+E#*f-vI!xZyDKO6;ASPt_kn6J(OioK}~ig6q-aMY%X2${_93uMX38
zt`?b#W;jJ!V-6rKWN8)V@X-W$s2vOaQop5LDRFaGsa`(feAZHfX^u_~iH`1>d<u1J
zu;fGquVi2Q7ZIHIS?VVTLqWq~vdH~u;zQxLegY<Le{?v6SKoa6c7D!+_Ia;nQdb54
z@sSHjnCV&5LhMRIMgr~a31P74Co8dRuNh(S1=ZCIZ})a{a4k_<njU$bP8nG$+OnHS
z{S0_Ufd(%Rn2hq5l&X-AwzGzhL9!<4MZ9s>deDv*BG@c=P61C5L%D3v%cF3VA@a25
zT|tQ_tItAs0j!vmR-vhnR9!Dv#`%j{QOJ*-3*(t-UKVBOw*SN`NLjd2%Ki4RxrO3I
zgW$~Xw)x~wE^ZE3J8c4uS;N9;X6_of=JFWWNR;&q5uj1#NsCb&t4@f`CjI09->5Br
zrn}LRoIT~AHKZh1fn|Ermslz$z-=SzfI&_v(KgJ>!7=m3VK|Fa8;p6Y(Ih_$;<Yb=
zXm_GD1yhzL+q9tTM<yl_@>x?7AfAc!D9y$g-#vOw5ei5l|F^^RFYmTH+@tSt<LMy&
z<p3cc<~2q6B*|IjeO+zx+}%=`FC*%E@arPGFR~-WVlf@tXIWg{&6{K!gVhAO8eDI&
zH{*D0H_Y*6eu@X!RS2VF2<LmeeAM$i!0L24K|@y@)pySjA?(eYZj6^kF{i2!v)P_i
zz+sVeBr`~*)7s9Vrd!#F^Bi}5NB41vsnJ=p)uOU+tmV&f$<SkgRlttx2(&8;D4d-h
zej~Gzk>7IdiqGL4rK4E6l@!9~8Wt2*rZAXM&^!~7T|a?QbG|thbvQ9um8QOzYsgy`
z?M_XWiTd#I&8@q!3Y~?yj*+&+^Ve7V1gRMnJT%)OAn9Hy{`_X~H9P<JlNYA;XkEuy
zb<4DM-;0E4yfWu+H#<WYC^)K-lZkxFCoQY6$-c2LGNpi=sRnIrpq0=eo&}Ysls`|q
zO7LaEL(R`gpk4IN<P-dpizk)W*59-JK#><Aq$cDeF)T-zuUXFzFAsmomei2pUyFyS
z`MBRq0=9@91^%Zs15VhbLne|ZtlKnYM4e>x#c4ueo`*g7x9!Xz2FI^jUGtnFFVzy-
z)a#)}Q?4bl13IAl0^#p8TI7)Oy9QrU?7VsFWz~pT^Pac$2ac2%h}NbYl?fp!L8itE
z`H4+{&ze5$aavjOM@<_*O{VZ2C6ibe<_np>wJztI5?TYz2yRSa6wl$<_eY*2Z65n_
zR9z$2=lSmb$=%u(I}x~ec_F=@a34qvjY(&<Hf+vFIM~A6h@h=yAoA}?P^x-A<VeRX
zPeq%hyPQL$GH$QRqR#j>6=f9^mvEA~oMxu(W^Nexz6EFelL6g_gR-bamW-><Q|Lfi
zN;c16%fA(J@_h)ao=pgN%c&?b)igmOO|sVNuAVziSB!oS+t$mUo4l7nXnj)GcG!ma
zRPLOYMp;6GRe(ALkL;N4if8z4r%(2(TmtQWD^l5SWbC)Gldi4$PDUywT_G+DHB-jw
z`xJ3LW}yXD+tH+++6*&w4dM7|)+H~<fQzR)<)-)abOX1~ZewnKXszmNzBI?Z_p!he
zToC@8d~pxPZD#m6E-JBR^V%#3j|CcWvfjX7xmrw}#9(P`Kv_4~Z!G1+`m3`r*CyqB
z0TZ#)Oxxh*8t29%nGh-_28+)AFA(Kv>g)-V@@WfcwZ}&w0Cx67{znjEVG-;BWIzyE
zbnL{5=o%UM8OLo#&<d=3;s!j;ka})x{ksx2sqq|gBPf{)8yyQ1k!(s1nEGpUHuU?-
z@oMwmXRpR?|2$$H2Ta-ZtHuDy+uqwZGjQ79?C;!qA6DwmN1s3M<W9a4X?oTXV{OPc
zr7-hbn<aWc$h(pN)HkF0feBFJQrVPY!}aSNJGxHR(L-r9b#3Sl=&}aUh);5doAzXf
zTtB(J$_8HM)H~ZSBx%zLe%DH59M97Zkx8gj+tbh;tqIaZCCJ8Rfo8~(X%Me8g}D{~
z>)zIvShP$m(iHpM_psr6bZFxB7vX?B&ckUR4Xxf>)piEVS%;oXEe|47j60X0^krzD
z5+8@oPd6;rx&69ImOh=sc~9B%3}fux)j|?kZx$cay&H-|>CtTDs$I=nTo{>+`y`W$
zzz>lG+dxDkag<ml(^xwnb1WK(XCr8fj{-tgPhJ!3RlDOYA0?TO1c}dJXkEQ;TRjsN
zAwQ+6^zxT{_peginxk0f4BKm5j2q@+fF^8rm&f4S7k}{v=>HB(#ze+spI6*^S+Rr2
zL2&VDLB>@b{z5e1q(=ca+QNkZ%`ny&?V1C6RM_^&9%7dP7Z_C~j+a42Wwnkqn<w~#
zppnlH6rv>p4EaLrspgdh!}bj9%J+;v?LQUaN4odZDuYVSYo<{rCW>8@?4qwE#Rs;5
z0l{V%^^6UwI}?LfD?b|-oT%ei=!)=IhUGeXJXL-Y{&`Gd=<%p;(%d4`K$|qiUe0RU
z`E@vX@5;00QkRq)@)pdN_@l*DiFxqFI7rqR!py0mNC3jYcxHWjb|G%*O+bIp?tf#b
zLI4O^i>le5<w@|u!rR?cV(q^-ng{3<T6M4&4**Ro*kF1QEl8zEP@pnXlYb1z`(~X(
zN5!q9%rkQ-^vb=>GsFG@F=@(4w2gG3=pUv-x(W|SJirQg={z>iiSaD(KzdeECjuGi
zc+v*Xof5%^eS2+aiBVMKSUf%}%gH3JL7T%Xy*6z1Bi!_OjW6)?o84K+f926H?~R@E
zcz&xx{(ExCM#@nPF2FeXcA?_WZbeBpvv(y#Rb)hXx>8*R=(r6?#*`D>I-IzQu|dqR
z)hT-w^iK^^ijpGQz_mbjTCO_BpT`5soI7#9U(J8k^Z&!(ALP+<WYm3RboPDgG8l+A
zPV-vhDNSW`GE}+$%<*ipa+#Uj{pCSt`d4G#=ddQ_>DavX=jyJ2^nQ-q(%bmW;2Dl%
zBI{ghK^CE>m-czF>)B|3InU{%w%>}{4(B{9uSkh0(j?69u&`Lni3tjN(znb^L{Pr(
z{N||@i;rQjY}^eSPFW`VTaYgsXJf)ZFMxp<x)%t|8PI*u!y7SV(6zDo$XScNp@dC!
zAb|IaM^WU$Qh!bn|3uo=x68DWxG{6t0Y5G@j8+3X^cvV-BbxR}?+Z^or0F)l>o0eh
zcRzaH_jPA_RtYb7os$tAM$gA%QK_78xymX=8PyvU&(@nGThig}Dj?zaFlzh#P~-F#
zaF2e!Kep^#QFvIAb(EIn`}Yc-P0UlWiZ&7~xhalG@&eG5q_~4Wb(`e5e(8<rfM;T8
zv%Q8ZQfe4wW`B>$VZHd3Ri~-j661w!<Yw6GbNO_kfKe&d*mbj)q!k+1=7?aK+Wk)&
zMPrZ4KgK7!GbzILx^E@~vi{%Xg{dZStpU6^zR>LZ+M3tO9FZ5lx2UK>Q+(^T0~B~;
z<hk4r>Ma^K88)Ea<S=_fWMp<;>hG0*%8UTzriOQ>EZis}9oA?3nphUHJZ-X|A5uST
zAmjGo2~_iR(eN9*=B@Z#jBMFnP<OL_v=lsa#ppD+9r$tv#G_B{2U2B~x#ii6H($(w
zQ5WP`88hfPt-w>(6bGqUnFk$|N40{cZhv2^9-0hzGJHkQNyT!W&Rf1rN`=haXato)
zo?@=t=Xu3G^QOS=z0%KMQ4_-auAgkz?>}sG>J8HbCrO9Dxh>$#$y`Er!&Ul=Bf~K(
z&3y2VosYLMEhlp!ha^SGuB&*^0h?J?#BG+98@63|5@Zpal&Y;Hpam->fZrCMpzdGm
zkOlSUX)B$G+2aP|P`%gCA8lz;VkW)HMIjdH(&zf^vowNlOS`%Z(sM%G;bj_<!RY^M
zkq^1Dml*RBi-1K;ll!A#o%pjQPw=o}G_N&tRbTw(VI=c%T-}JOH0s#=6)9V7N+QV4
zR;$}!DsZ1T{mKJ=Tu63!qZNCo61VO2;@%LZsR(A9<rYjudz(Br7sIz9J*&(YnP&4g
z1CF0E&^)_k7SfvSQb9;dPQa~9gDSZ_90pmQ8hSA0D7+v;@JOZ=dX34v+=JE`b={SH
zexT2Bb8vK2$3p-Nyf8MejW&mWZN9kwx$8uu7ltuLjkX8{R~AG5;`OEbVn96OswX+7
zxK9>LGVhmo`W^G7>Dh_~O}xg>WYwkdLSfEYTDtZejl+8yB18eZ>3Vm^0X@4OSWBbC
zZ8ymrMvC1QM9k7-o=!G0MPg9-#!S*S)WHYdm|&kr<vJ6E`zk->IYs$7d8&i*EZ`3m
zTRg{8N2ZVbU_0c>La#33-IrLhIU`SHJCbBfEwnOuV2z!6ee4+S@Mk(V(-i`C>qu`}
zh;bj02bGwDeZ*(|F?w5n){VcjbaL-y)3z4%Ju$fJG4RiebVwYIg6bt`lK{;>;^42f
zhs!3(Ow8!o(IM)9EPp;0AbRQ=w7WC%@8tAr_gPu~ZoRh2v@6U0T4{_lXigs3?E^ZH
zJ&3fi2%;6n@8+)LnJB)|a0G+ULUI;SObaUQAhFyGNYMqwk~KlRsRsiMheDYaXD=IT
z3-!1;O;NI4%6s3xyCJ~etEV<%PMF_1t#$w`tg7x4cF67|a{r6b#N-C~v*pj-uM(pZ
zQm@?Yo-vG3QxfGdT$hcG|AB}=r|LHRbCY7f+mzK3`&kL-k>_LZ;<M~-n+sZ(V(4J>
zVe(7IC-jCJs)trlVka3+Bu*Q#da+2~EB8MN6mmC+vt%w|8u*0Z3@PA%B;6Qnw1_5y
zvk{8dsL#!s=m&Srd9WUeB$N7+N0NFlfF&~If@DVFMk(ZeFwLK<3UjdV_sTMCl%;aM
z(RMUyx>cV)(<>TA#t);h_prOklP<mdUJ{G#pCyp#7IRewueeX)(|G{IG8`4Fhy0C&
ziV=fWpZYwewxBIRhl1)uGy*2)x36()p+B5z5M)k>wZhCQiI>wZ3QCf2*P8yw3-o^E
zf4ed(&QB|Bd2AS?$+8w{(ZzXuGDP;*`UCt?au&23*u8BRLX+IuILUz2tY$35uqZO1
z+^+PhT(vx#=D^k|pQ*PEro|&(Dp1c?OMl&+^L@6~)Kxf2_{akmq|8~HTnE-mt;_J0
zYlFy8;5SA~W8RTc!pf$f<!-PfJXS~Jde8kQt{)2<Z$bHA!NaqTlO?<#Gkn=q;;~;Z
zL@G}2X&TJ7Wu=KPpqo-hpP%bzR#dC#9=pbyk5Yu(>b?yR%ly7+Yrkj%@cz$FD<{$y
z-|fFIEM2_y54tgWXDrm?sePcKK9+(c((W?9WihS6!qH4lP2)!}qL)5+OmkLpvGy8C
z2rifLl)J4n={!CQ_;qvUym|V0B*L_%Eoo$kgg(`RY5w?#VcUz`{-%wZz<^7DQWx@h
zzwvs{c7vIz1Hxt1N`7&FO`NTv=!BjKE=ep{R6V4P@+WYy({!HhBpyY?Z^VpF_gq<w
z(Y3Q9XpT^+s&Gu;J@8cJN&H5emLzM*iTPr_r~Kr;bG`nP$;m#RIGVsfQ6Pyw3vUO9
zv5|j_SPv_cKpLF5oHh|E17`aH*U7a&Fh%cke^L6Hn?~+ZEYZ(48FiCqpTT>x)ouM<
z-PuWBLq9&v*;})bzm7OP=-mc<E!`4I_qnJjV*|mIGDhiz{NY#%H8@Tltt5mO$FqLy
zC`uzZNUqHlvkhOb<bY(f5;Q!n9{QoX{2FI?+yyWZ$)QkmB_xoeQzM<0*TW&dsM~A*
zhc*4wrD)tPH+rT@cK^1{_T>AqkKO+sBOX|%CB;_mWf0J^R@Y_>c&ebIrDbD78B_AC
z6HCaMCX@gmYhv*^WjL%1d;F4K?Y+W6JA_XWBk}_Pgf5k5xf~MV0qg7u%#kuPs&uWP
z0F+ZZkiZ10SyTx#SLT=pVk#eMASnhjQC7U~h+B+ABF#(jsdkHrixXs`px>oX`Y|Sz
zr}*vKE;(%vvGR<wwj*0qePgh`m?$u4*R)y8$#b%@`iR6Mw3whvD>N-2b0d|@;rn!P
zDaij@D-2<XMu0Lxz<TC;tDBK7KWZ>>`zCy3L#bl%qBG4k?XmzSFxrM58y6<X#f&}W
za$@e@|0GdQ9@^OQr~E7)LN)<%+Yy?HRima-Pf{mFzak*X0;O*Ce4H#GCuJFbsWb1y
zZf<uwJU+Deg%B(E@`w1k5VUsS>f67W6VY+`BirksjS^yEx|*S7nMkb#w`1h2p%di)
zX{QFZ+hATwyyMQ@Qu==VCiFG^W;Eb}xD_PvMAdP~ruxt&y=9oO@=o*inQ8F#-}BwW
zz?+16w+Yq0?1wywqQ@l@8sUFY)(de$ZoRsCKdFEF&f_3Gp!AjdhfU?a=46+77%jDR
z{`5SneNEa}?u(LpImmbQ+~LTFEs67$P7p=2PhMxjPKca3!Zt*s^o5_g$#=j9tBA8(
z2UY>EKdvsXtPs?PTuZoqtFL^>=WKTV%3)DT{NfLl3bs!tVP$L0RH9#ruHF@;JJEF+
z&Ewyh;H)JZ@uH@i;%JbX+23GCf4@r#pJ0Ps#hS9t<r)q6;6m4D6#3#Yb1_v3KsCYY
zww2@#xAjHh3sln>UTi?*R3Ww^znh{J8wMLNzp!RFLG2)4XU0<D$jmOtejIC#3`v29
zjiz`B4>|Y+SL6dZwM+H?hm`^oO-7nYsvFc9*){Z~H>8qIxl>Q6*vXb!fx~Ldn0z_O
zuG6H^@V0tPwS^EvB^5_I!X%BGvuwd^ku)G$`1!eW9c#OXN{9{%b$zh4l$AnEFfVE(
z(h2Q=*JMG%lkARx+R6M=@P$v_jW^0xa@Q;W?dILH;VU|G{r?|xL^BpuFNP7@v;8+e
z-#>4@`iWm=PB*l*x$Bi4VTqm(>p|{^2Y#!>T)qck8y_(6Nx?gcVhxE{{V@myTn&t?
zbl12#;27-ynkF&HhUu_0%eczrce7A)4oUhDa|#x{J~tD(H3Mfg`S0KhOgjAE;XgB_
z(f;Veg@vWHBu6Z!m0a8GpW+&?D_TT`d`Jq?MY-`XziG9Ixm^FuUths7XPzt9=C-EJ
ziIlCyh2@0bl<GwNe_8;g*qy4}q+3EXfR}1sbM=OoYgdG{mR6DD?&vzo2%hPJ(BEOJ
z`=}PtV$=!>vDgZ>;yUXrh!Fo6S0RW?3EF_BJj&woL-<&r3470sQ^B|CDi5f*DnvSA
zNc?Prb6}G2+n_P%fdC1ao=u}r?P><G0`BoGxb}edZ&Lq;q?s;doDf*Ku3|H8^C@1?
z|1Oqn9_gGE3p@9F^2)Bm=%jDqCAaJ1pDJ*3qK*R;9Svjb=mQ5>Pp%5DM>>fY&w4Zr
z+6~dwbr_xRzU^s|xcLb<eY$&In8vwmw-+L;xM3Od0_>w9m-*e~=BIx&l$IK7c~~k#
z%r;2%y!knqVz7yfcxw*hIpD(hWK?%(GD&%Mk00m%u<W%O>Cr5JMIVkCD2a<p?T@><
zJKwYHynjRaN7>{n&-~m4p~-E2$7|l^pZ(&#{?o%xgvWh2*dD_i!v!5Z568*0_!LlD
zUw%uc3DwPqqxG9Dj!WpggxRc<#H5%pnZPgEiva}BjpK=FMeR>b4uoJypnn?u(7sz*
z?w)5<8xh{FIHFbI?8Oi8I+Z~GgRD*zPmuJkXSe7KN$Q84B0z<CSpn6YmpW2g(v=Yh
zhl(umn=ayNBvLzS)CAR^2$iyapQyDDwCfFX=B>4j+Rm^;>=S7XUCxnh(FKtveayE~
zNp}4TELN4+=#Fxs!`n<T)+_=kOPBZ14>Vuh9RwFtb?t^0_%+pS3o_yL%8U4IP%vuM
z!C<}U+n|YoFLD+p_K}w#0%fW7Y{j)hcZ9KzI=7Ft0jo!rlL&u;;<3zyG-=fHq`9CP
z3^4p^`CNbBQFm%r)33YUi>^+g*CVvuV$mQUQh|#uw1D_=t@^#x+7!ssg7D;jE!_W%
zUegR5c@eXhW6!sn)64!R%S~-<H&OytQ@9CI4dvuK%;W$#n9Oa{=Dj`UJ~1s&#z#6&
zLRXM(tDO}_gGL;QgS@RS--4=_G4*)qT?rK=`k1;z^-a%KFMC0WS<Z<pc_J*`^K$}D
zjmMXHWF%I8z7yG>c`8yxtQ*o|OjUx+tRs>lvtQ;%jSoV<$)h;TbrtmzP4drAaIq?6
z0dsOHeOGprMllYZD6~W7A2mu06H|vb6-_iA%)KuDJT4vcHECJm=@%LCdgwpk+U=-!
zL!w-}-0<ync?X8C-O&vm=3ynbrPZ}#`a?d?PZgvzdQ)SrxrPSU9SJZ~;8Q_+3H(x&
zhi?pF_)}gB=YmM$kNJlb&pcNSxbc?Vo??UceZdzPn~`(a#kmeepuLs(IH1mPj1dF?
zoQzrZ5}7v2Kx%sJjPvmpsi)fU&aQj^$}zRc9i#-kv>sN5$Yl4=E+{!qm+1qmQVer+
zbDpH}JP3zCL)n&f#Ln(WsF82y2CH|cseW^~BJYnsIRRg-^js8v^5*k>ImW4Y)Fe{Q
zV{pLmyZR>(M2nG>01Pn9NAp?1J%r0a!-Tk(7st`Tej6cuUpKh1*k3>73*O!exj1F_
zn<i`24Ym-ljx(`f^lSOtx*X;4Z+3KnI2M1}$vFx8{_M+qZG!!hKaTMhwrmb*sv<fN
z2ir3r9mq!6eThphMm<Bf0Ndewhm4(l?u#8V$z%P2ODq$u*)+~^y*cr^_=LpDvptIV
z8dyJ>uQxoPLG_De`tYR=?`!DjEpk}Xr<)Fhk2~-<kqK}xa&y?mj{5w#wUimzk;Y&L
zs{U;ws&LmN^bXFJ`b>->-Zv}ZrB`Jn&}p2I-}OAm%&F0n46q-$1|W4^C)+d<u0*^J
zWP)Ezxvcv?KgglZNF^Q18^^}dMe}X&YHm8Dhk?v8mp-7-=SC(S+sPPi#Do5DLL^%F
z25Hm=&hE(NRnT4ATqn(G%CZ@SBTL)?N|&G?OwgR-&Ada5SGb{w5h`ubUEUTKrp+qA
zyH`=vB1v7G4zKLDGvKq`;aZ~LT#@hUr0?QxsBsKm;zE(~K>(A=j?8Q}zz`wU>=5DB
zqc<p^8MY{>k)ZWU5I;FGiu*6m1<L%IOLfw*)2L<ojDFt#137x{lJd>4!ax0&cV9L?
z8WZN!_w4@PJIS&R>t4#uPWNBsyv5Rpq)Oop=L7~tzj>MGULgB3Yd5RTQnPripnka0
zPy2LmlRUP4;B6j+m3TkRpJsijlE4xu*5vaO>L)fejkq6KV}{?&4{X2J)K=4%6Il}6
zmUnM5!~-B}TUnf1+9|KHuZ)Tg01Tdc2TFRI`n85*bLwyKS<i8A2haD!+I1lJNiLi&
zUsMP{UWIv+Or2bvz&#bb%BgrjsJoSdoyFZL?rW6DlK)xI87)6-=>*H|=qB9j_TTMA
z-rkuRr`vkpsh)%M6&NZ-(3mnpI5e%nRvA;=L@28w0&+%{;Ho&V?q}SK^u)5?i5dt?
znmSr~%@GT@Dn{1Gi62BI%Z2(J_&!loSp@vJY45{Lzs_8Y<4Hz63i6t?JNh%C6ncTT
zq;V1;G(L-E(o`?1M(20h=A)ahy|ARNFUK(tJjWxR;Kl~Mg`7;SLKCp^2}bbxdyr%a
z0$-lGStcLrGCQyHZ?|jniM&`<?CB+r)wpttjz<>HKO20Nti=;94W!LUf`}e{uT3|5
zfyb8tACIa7k{11g-oqV<1+P0h=$hrou|YVP4-+{M_jgta34zKEOu-$5JYoyOF38B!
z?M_7zmYnI{zOp_vN$U;kA6Y<WG3Jgq>bz@LskBLE22A+1$bcy!jmfUXX!yFHldQf_
z)kI@IPJV>b9^WEEzIxZ0(?%F>+@7{HUqj5nzs;`C-YW09?GBNvZY`l_eN1*oL^C5N
z-?6=J@+Q~z3#V%FhDs=Eg6G0KBIMCX8fjO#a1tWasCG>?L##cGNU|<Ut3UBW!is})
zYV{Hz$QEhk*i69u4191O53n~bTV$%GCb2%h=eq+2+W0zNJ|u5q0C`M9PO&V<2-3#0
zBom#9<WmUqNS7^1kaC6S0@%F^ZDp(livS}XV+>6_Yc*S}bb}>LSSK)dQ_n_8qr|u!
zA4(7DbB)=%9uaF(oRPJ<2_o6bQH7MWrOBg);2~2*gcP@N){>)PKs&f;nY*JGN4Tn4
zqch7Y>54^(V)+5^RPL|yy8OU}S%UGH{iL>;8uE<?O<1X<f+WFVLMK`12j}kMNBbH%
z7RerqqU@0*Cz4qOc$a376?YBcf*>`OZPt4S)l^d*#d**^argh#Ji;)J`_=2d*W1^Z
zC(FCBM4JDs7_7d2y+&|aB>R>;n1n@6?)nM=6(OL-Ms&MUiMZ;U<?Pol(Rcd>ghTLX
z-`}|7EYagzm*WFOsw&^J5^yjNtgx+R@q-6u6nVkO-g-kLKO6Y#B3eq)t4xhUONp7v
zQj4=~vl6A5BqinnJ(OE=)>_y{XGMh=Hzj*Tb%PRg5|2zAUD4{3QA_xo+Z+m=oZJOp
zv+L{*?cU%aBjFvFPi)RQJ66)3q6GH%_Ra6nMP{O1Fh>^MU=%Ng&dX~dK2!TSc`55x
zI<NHb9tf(L&<Bs@SBa^ay`B>R&T7`;xmgfKuC%&AywAG(l2}%x-?q3p9WZQXYUcqW
zl8xGo1DAg6BQfg{{8S9wF`B)e<L5z2M8Xyi4*u3S)lJZAKzDBK*%Kd0qUC3n{oRhT
zLWF3&Fa0l+!)H0WS8ZnuLJ!;JVrO#kKF8nBAZlJy%8LKia6$cJ;pnx}p(s@e*ao2%
zT@{tJBKau0=wC>pG{exV+>7*VWsy;kd{8FK<9Pi44w@vjzChOrO-6s~DOb@>`Ov49
zw&nF~gar5C33=~!7p|4Q|Hh8=<r$uV;EkG8!%9R%tCi!ZZe#La!raKM=*)Z1n_;;d
z^q+{$@&J{E_$UnY;`<JSVQcU}Y%{GbLli)Hg%RM5G`Ewx9obD;FVTxOr~@O7f6pbv
zxcOq7CC$NE>VkpgGcc{lj%P{MjmPd3p9(K6aCe<F8_0WK>8(C}r}XbfUutt51a>nD
z8Y~5#--J)Uj9EF}!XA6W*StqkB^3kHa72eDniFJsQvecsf2Bb>(zLhDo|$P9PGgyu
zy5E%t;=w`<R>g#Kf`!Sop{Zc9Jj4QsUW)}P3AdUmroG*fIybAhtO9o}8<qoJjWu5*
zx?WqZQgcJ|n<~01oZ%B61H7*@YZ#B{TAbyw_)9`lQ1ts598ly%9&78>T!cyyu%iK7
z_X*%FBV9F0ygavN5>P_FLaS*jF3{CxYYcW)T7gqFr)UsZt+QOLGF32gd5hBJg)7$*
z2f;P*EH!A@uv)$0mbP^_r?Bfdy}gLmAA<c5r;LTTuq0mUP}~ObRtFegVu&_%=0cpP
zfDK*7VjMJ+DlN#xmY6V132kDn7>`!LBDcILZvw#!IpGU;tcN6`5{iwCWLY)2Ei2|3
zJ{WmQg}F)I&6CQaHTAv8?%LG<Qf~WC%I9>HThGwVC+Owt%gN@;$-U*BjFc65zOdLf
zS~fyg7B)+5L8^Nbb7bjOnUV5TjC~xY!rDcO$W80i=dN|Y(UEZ_YjM4!dR&2qx}J4#
zu5lrTh>#ChwbTdEr&?7vv?LYSi?E^(t`#YbNkyi7(8|)Lo)uZ8(GLOl_|vX#fEfSf
zKx%aDM&PY<0$cMk8;>1Jj~&a06t=oy+b2oDk%8(#{rNul2Z{d~KL0om)x|@wn57oT
z8_*?CP;?CH&Bs$)I=$~Cn*P=&%(FA4p+P~6XSlcA@qADphQ&^$b-n~O`D4*jUp&Kx
z7k7l{;c}U&+IZG%KwHaxv8@ASI70MAnsk$7RPz`C31!L>&8V}v=4nbSvz_(0y&S_L
z?~WSdci1}#A3*l_WoT4F<stMov2WKPqO}Atg-^9;NoKl|#-WRBwRGO%D-r0+9I+0#
zr1tTPxtL%h>pFbQ%w^9!9g?p_Dy2=t^CYVzV5B>gj+*C@Pucn#nj#-z+D3pPTnQ26
zVYMbf`eK1!!aT&2ucrNdJlEtJv+k7dyv$Vl*X3F~aQpoyg~NCMhaN|j->od7O;ugn
z5Tp2s;Xa8nERP2$kLiTbfIp5&_M9L9I0(gRCA>)x>7R9n+w^tJx98WWGnVg!t;hIG
zK?Q|+V@=vWZ?paC^OuqQeQWeg<Is-fBF};-11hmXW=8FwZVyN>A$$61bpm6GKcwp1
z`z3GkVwApG1WSw6gAV>eC?CadOwnBK4Xo9mglu2-UAFx4^PJB1vbn(LWeNXa8LN*}
z>hk4JLVuU>LQWPQaQYKulah<0{7MVz#W40(GlbgECQH5;_GB0*be@O;5~LiG)S5PP
z5)t9LB`;T8#3f^Ty==`A^#oyIAx7fVumN3)^_s&nY=w{IEr7kWod`svUeCeC$#7gv
z%Gp=5DV6l^8V^^RJ%I*dY2jLBjq~qhTgjq9xOD)-E2+ObUW+*yC<rM@mABi`%U`Tf
zweP;wB@Cf1D=n?0!_I;>?XH)U<cEH{!INX9Yp-bobTYH3=5$;G8pNZAjoXG!@~xI#
z&Va;pSdA^Xa~(`mq!R2aqYg}^+<bAY*t}m91k_<BlH6>ztwjwhj2QFaCN2Gu&VDZd
z43`&dLHB=CdZ8;t?<BBvMbe+(#HKCUprxj#DiT@YkZ=nIzaKZf5~&sqe~kL&(Z2GG
zrzP=!MN}vN*LQrn=B)4a<?_}4P|@T8uFZ04N9enetE=X0#P0o8-Padi_o$SHu4*sm
zW>ZIfF4lovgpZ0d3JC`Bjd_F5c<?#7nmO4dA$amkOfkjDj7|pQM0%=vBBvtui6J0Y
z7Ihnxss4wb3G$LWC?ighnEfr)4;W>Y)^6H>D8@DlF9<>_G^4osD{i%Q{-rrN3s6E<
zaQ>$S={;va4md2RdBbgTp7j^^$?K66!1Dp3<FY_ffPF9o+$VqKF|A^EDlW2p(6IY`
zI3<1E#|wrs<E+s;?i(dX!m(G2^@@fWm#;?Rv-x>j;tF%9Iw0(~a(!2YN;RJIqLczl
z+82%C7Ido=4D7}(CL6k}=BgsYfC$f%e`LpRn|ywMdP(~HR;p~h#ucwM_;`Htq%_N^
zXeEPq?(?0{*aNTlef~D}bHmpIQsSN$lR;WIlbI-l6G3wi5&GF6X%yxoA#v6U1pw5;
z_vOUIOImc7qCpLu<xN^>wN?{!mx~%^%|OlCST?!w63=sad{22_6`&mOim~3*9?Y~G
zr_uRgKFR%`?gtRtvD^@bFJD)UU)K^U`=RkwGo)rbwv_XQ2_s}*^uKbs2P&dudF>;D
z10edHzo9VzFZ~aI=_y6H_6zqY_j7OAFVmXON)Gd1JVc@&MbA?u`ZHgaHS)>0_~UG?
zDr+n{@#pK%U+>oEIV{DZWWeJg$0&;|QZkk%kE*u4r=xj`MGHKoO?g0WghepU+uhbe
z&b;LIS3IO$W&!_~bhXm~=1_g>r})J)n5u^#0|Ww~{+L->c`D3@0=%s#O^EUAhF||9
z62p>4JW;o8?F1JT1w|@_xR_^f>WsV{oPjfV(oMgb=+iLOoJiUR;R{6x)S*L2ZN)?6
z_n;+9YsxsSG8bzHfUPC3sTf6-cxIJYMJIh*fRk>)g+9ckOBTTcJUKgSt#s2`%<2?I
z;uR|dTR~xtm9CzxSTrN|6i_N`tnIfYzov>H228p7ld<59CJ*|u_oP~NO(|M?%<}zG
z`WpVI&#}|fb7uP8l}8V+O6sAqEL-;YOy&y?v<0;ZYiELcf^BeO4#fNT9lJanUrOHm
z&>6lmTBT}Mp((?5VlB}i8fQk8`sS-S0r&7TDUUgmI_l8WT3uSu96TkWDa=UWTMFab
zr2#1(b1(_GwxcS4?sLdr9P0ro*S(E6+xh|BirUf!#7NW60{cgCS6-{9`)H{t;{PWs
zd*@3G?|FOjYZS;JL;ePK+X*H_JK`B^;H+t*bQAJdqP(7cMPdqZv0Y&%+qh~xLHLFh
z`tL%e%t8VK(V=+_F;<o8073%W4#vVF@)qn)0E<lj6pXoC7@-L%xXF_|QEw(W!}X3d
z_h2#eV4!yr;)h`Ga3AZS+6-rvmeK#p{33z{D+`OYg$5eqOb9GX1jzWgcKBxW9iWgH
ztLrOTiF!F!#+K&G@EXI9kW&Z>Blq7gmX%?%tYz!C=Lu-tUgX|BBV{cv$w&~+W7Its
z2Lv5GG0cf=`d^&A0eF*Ch_qACY}D^rc=lKen)>m&2BX>Jx@C=W8V=qd_tTc98Yr1B
zG=9hULVMk)FI2ZH`#UR93ntFS9b(=0y~SglAaBPBYZVqdScj94q&WKzDltoab&|<U
z+sNCQI5R`ykt05+?mP&;k^q3U1d@-B=mBX-|H9dild;fu9kn~@U+dyo{pkXFZk21!
zUJ=vunU=rek|cgeo7MPSVxDBsoBBti&qZ%^tUh2~Bh3^|>c@bXV){LY08k%u%0D3k
zV&wXj<^2aE<x^3iKQz#@J#uT~Y3N{lGeZFKA-dKaAYoAS51C;~-Xj(2JgV;8+2W3s
zLnMN&t|V4g8X->^>4^P^-G5OOfKwBE@j1fiD7EjLAcK80aM{fwwcY#&9h*W+|Jo~5
ztw*|yj`qn>to70|v83fZ@ZIM0<61pzuQ|S%Q%pf3nu)M(e?C{o<&0MHC;GE=k@V0_
z4d!Z|$Q*pEV2sFyCP4X*=+*r#x1p173falF=*G~i0G<GFzM;7u1#(}<F4q95I}ex$
z1~g%-WA<gKH1pgCJ4<n9wqBmqCHc&!EAx=T!nevajV%^HgGof%m`Vz4j5QVlC21!(
zi6=Qqy9hu<V-%(##u|G4BQ=%)+3Ga%#eHpFp7%Er`Ar?cI&=KLDxZJ|W3L<E0NEd{
z%|4y(?q>+k>s9Su_cutMz`KLv`R%FSAr==;+{Vc++;}mBbv45a(pvtdzkq$|LpN=^
zyBw+%7pvoe{h5uuav5*cVWqN?I`Pe|xv@{jBu#MAns<%bZmLz<vNW?i7kYJ;-U+C<
zAYPOidhY?1v<hbDgkuv!bU($CJ0$a2Ln*~Li|MBzC<#&mTH5ob4Q1eKQu}4eWc&|X
z{vQC;htj+Iu6y%>e6ye9X@KJmCq?|>mzV#tpVpdc>Moy-d#1m@d^U*&qsJn~l(>`O
z14ESHvDFmbhU2toy>Hc3G*CpF(E9&)`s%1CqxEY+knS!4>F!2aU}z8l0cmN0p{2Vd
zhi;ITZji2_TWX}cyYqX!_x`@K)~v-p%o><;&ig!j?`Q9?GuDcw!v;%Ht5T(kztHNd
zkF4t_HPF-)G#S5D=P-6D$;-<-MGg=z{}AX9$@(|l)PBbOjwgf&BrEIcC}z{A%dy<z
zmhC;>c5`^~aI|vr@Xc_C<da{rC1VtgaAmMaEPMe141N~!IVB+pWan<nKm%$2){1i<
z(Jw!^WF(2&<<(;|D`dnlgYGC@h*5NpXdy>OB{&hSLqebwlGN%f<FC4bi0nvCm@T1a
zE5q9G4x)x@Iq0v2P_EdK1vVxoNYkZtQ`QV(tK83E+<6_kr~?ldr31tZ1=o|Of>O#2
zUn&opIBivV;U(Xv(WflVSiN4UyGvE1cN>@UzAOTBc(MKD0%C=4t*zm9KtjF#IfRUF
z`VtFB*jHO`SlHM=#q1<iD68{I7zgB~c@WR>XwI79&54+)?7?bT9&M~9a0ArKSjW2m
zPmdq1noPk5vM~5clEaKjMhyi{X2ttdw+Dl?Qg99d#EKX{?dT6#&&bScyUbFjVcT=}
zPuZG(o}XKYUmhrbuMc}UA16K*CAVQk<+0<NezTH^P{D;n*A9y{p54FrueXc$6+R7%
zsR$5~v^_Gg-`yyQ*vH<omMFZupv`z_Rcoi^Zr01SkRQ`8@)$Dg5Vxsha+C2;GOtT$
z=!QB)^YSg?D_+;X`wfz(LM-7zP?!t{cDRGH^U5cRm!x&-=eb=5p$3j^g3z=&yR_VC
zm5Rq+n*k~T@|8ix@;n@c?R{!w9g^E$OUvDG$85`hSQCL$?`CH`7(<X}WFakKWLDbl
zkS>E_%U~OXD#g^~(%`QV?PDJHx=)B?Rt@A)ANwqrmyHM@ewEvEMUr-|c&(TmSX9Ag
zUv;YZ*{d)y>!f<WlYIqKeYb6qAd>?d^vK&V&wbSWQJqu%)wFDKwwn9=MgmC$ry7FJ
zjWJd|A7<4bF&#)PQX(seJ@r6AJ7yz!5~<j2l4*bUM-_LA%u{%L$Na$W@9_&pw`_99
z`+I4knOcrIX`@sd8T5~%4jvr>V{N=2j4{-y0EJB<&Qv*5*;PnpHGE<axgEnIw_cCn
zJb7|4r`h+z0^+Hu?+erj@N!m&E_7jus$*jU()<i*wWd+{oJ?(J`o##iT}o`UmP{~!
zWoHR#rqo7VS({S!B2u<09UY4J?>ZQF@@cta@I7a6^L@S$f4MjijE}eoe=W)RO<6qK
zgDN~><4=y+x3bS<1i)++ta!CoW$Y|7gQKzl7%jd*d7~T*&LG+LI={tyIXR<v9v&3S
zHYB;&b%jFn&0O?r;uowm7PtVv(!%n}+;d6Y_kMk{VTkWaERWPF>+0Hn(}<1}j9f;4
znGq7A#^N6F-m;@Yh*GQO<3f-E=a5V(SNH}rrfAj63?b#cvf3H=6g4@VbaN?~9F_7o
zdZK@XA1B565i1r=z+mQr{)k3rXCUfp0vRmscNv*$KEil<>2x#!AjBxBY!~MLyV?*e
zEZ~esU$8b=9x2zzv#vG~iJ(pew0}C3$r%(`584x|e0hrj7u(jOGSwXvPEBOsX{D%i
z3%q;>a?4;*`-bcMq@((RImdf4-?wh5!}o*7KkkN;+tSI*P3#Pd^AZFir$c_cSy7d_
zO4@zgMT)yxrlZVD|KYWD^@2nXRHi2dU_cnklaTm){7IwSf<Sv1V=3h?ku|w(e|sq%
zO6zkmE;3SjN#WanO5@rqzdrL`eVR*iZ9#|xE~;dHzqZRF;A#&)$SIJq$TXj#fT6G2
z@NtxCI;eX^oT`;qvycr_?mgtv8ZJ+&mOo8_SZ0W<1Vz${nuRSlfToS~;_4M$ZmRq`
zveaHZIFCU;-6ofI+Mtn=3KKurm1yzBodmPbb^q&I)Ync?Xz?wpe7<_sy(nF-S{kC%
z)x-)QT=_J0KX4T{>wJpkHW}O*?}2TGK#+^zY(HXmHx&=yiJpWc1eo&7sH+(0a+@}9
zE&IwP9J4ihitU*D9A2E{eX_m1Izrim+9)=f_Qfp(rCS_~@WaCck=B5(-I7c$ek9tc
z;0>-;xgZq^9hj<0C;yb-5N*`Ywaii$t^kaF4w1BDtP`+0kb!v64n{%OPwIp4d5dIQ
zrckh>1!H+l!m-t`R>uO_j)mvfw?biy&41P_0x=9cs`@C9(wsN^%B?=Qe)#X|yV?|@
z{6Z=09QH5ZsdZ1;e!bu4u;2a2#&G!Y@$o8ac24vnD%JoOa7fyvjX9Wbv5tuM)}ge?
z51P_R5?&SsWQu!|(frX`Wv=y}nNDTVd=_tn3_oW({5FiW9#&GGRFr}*vHTElG~#ID
zDOE_Zl;)=FDi!;AC$OPYBw<dvPwG7ta$g<Y$q*xTEc{rEVgI-%P1N}Feipi2nVJ{(
zccJF2Ar<+~JF|TyWw33b&!?RmU1~(yTv|;?-DM5QaA<q5wBfz@&7~K^X4aJl+<B<j
z*3xVxq)%4jC4>mlh=qrBO|N%4-(_#?`PYVronM+Totg}9RAt+Rw7?JWZ!GZ)b*x;o
z<igzdUkwg5M^5Y|QO@#cQ`{i8c!S8&6QBl40x<u@v(s4wo7_ak1pLj9JZ(xM@FmE>
zSiEi35Skgn(AB#Omo($f#mB_|?+Z}OIU?{j`6FYbDbGyC@2zrLIRk}@jwg|y8@|DR
zmN)HdU1$t-?(L30d+^M^5@G@+tO!nM<pi1PAC}{ah&mGSygYVr-u>{EzQ4oTfTq71
zxE6zmw1>`!_>d>lv0?$w3#mWfQ6d&~v_cS_u428*EgNBNCia>KpZjKh2CYI<Nu!ji
z5=B((Egg?oNH-+7_0Ur75y<zT$kaN;NaTAnOi>-K;m8n)Dx5D%JGN{ortKsemObXW
zj@BYO+o_}=hbKD`9@D}4yjc%XxYF_-LD$L;3INuh4~_dgJj7|Cqf>X5Nr2g}rTa6V
zmM6Ndt{1Tk6D_hH9c@^Orx0wQ9p1@68FfZ5Ju8%dd4s-V`2nKE>JAi!oYTGMcn+8$
z!3%r~>G(-t<8X-LaDAlW58SGbtERhT-|~P&_2_y`njAdEvRB{8dqyTWG-#=mp_9{;
zEzEiC4DXb?F0{nPR+dls$tTk6GLmL<%LTBN;0L?32G;D9w1nFMYbAbWA(%2_`PW#o
zvWquG<uVXP6wJ3vV^vy^pl7!OJZUy91o&&d*ZHUmsl?nHKr~n-VNE9CDt^OdHi|Jh
zk_eZAU|C@Ljy3sl?_qx0gBHV~%?IF4c^JF+5i51bjE!QXlZM7z!@~UERe>~qClZ&K
zNO1i7fc>9Q)0va_jT5<;R|+tv(bjtGQkgfsCt4!B9)7se@j{ZeH%BPvi4qFKxl!U}
z=`j(wO~H?rP}qmFC3=%l={{0dS$0T^o7X1VL&fta4YFn2Wh#P*WvfKI9YUKxek_X*
zn;dWLmCVOiERPLs)OeFB%e+D=$i%90YpsMJ$Wln!5+M(IyOb6i{ZZk;J;1)<ttiv2
zwM>L;@qEhWyoVotK>6}fN>8Q5%OHz3?8$g%)HdV-fsho$$@fP?5CKH>y`dNVN^P<f
zQ68><fP7^PJjph4^y0bLbe_#|*Xq=@7Cwt6B;4dS-p>-6u4<JP2U!%HTOB#@NxJQM
znUI@Hp7f;P$GS-7Ss6T38B@gMPqjg?RZ`%Gdi;9X5x|vV+MR_bB%=oY2f$g}mZ_bH
zbS-Tgw@?2hRaHp>_K2O1e$5KR6s49TZ9=t8Kks1Q{>r7ozBCzZPqP1@tQaUWflOO&
zJ|Oj_ntwL`zMJ;0iBkzxkp-2})3E<K-6u0qawxUdBK+W+EW4QwaT=_eXHsloG2B)8
zP~!-%)$?0U($R-Bdmn*OfhTzWB{+ndSvcyjsJwXvArWSw>dcR5bPC8rCE@E{G@BAE
zCWc%5J~RK0)<<fM06ij7bX3c@`1Jl#rxw5@G5g`L*MBu>fp{-t#eK+|U`Ko3%_LU-
zFbmiFg|FG^2F5=Cc%S(<kEQ+1_2b0Ghjlw+h$k0|G$z|n5Hm1F(y@^TUkf*b*<Q7R
zQn9P}CY`ORb3<eSEkl1<jG|TuN$}_{J;r)xHj~8@D)Ads>Y!vN)Lz8j_i+$b{&z6&
zn0Kj=oF~z;De(bckY3z7^$3vFBi|>o)j^cyMWxZt!4eAKE>A5qx)OfdNZzHimECmw
zG#|T?4-v-Zu^W7LdQTsZNqwUy^bLMnaIjv!M0=0=9cemw<e=mmnQHuuGHt+_Qp5*N
zgoZoCqB(6iJg5R=RFG;KlRAe)QP_VYX-1W^<<V+6wqja0lWy2h+M%Zr7;1y549!vO
z#uYp(e?y=<h=+XJ!PteT1H(JEy81f7F!-6t>n9Y2lu%Eat11a_6}$2$+-Ho{ROeG7
z;SD9!26;9AyY%g|=&bT=`Y{9+B|eK)l&5|rIq}*Nxq0-`PJQ(S51#t=zK4yzV5@61
z!N>fczj1UC%t>k@>b3a+U>i0QR*+MKa3q*Rp~z1UO%Lmh0FDx@X_0SZe(v)3_qcz^
z;(OK!!XRGYoR<|L;-C^pv7p9kU@KD2&%6V%C0LT*l%ww!@tyM))Dec~5Gt4hcfemR
z#Uo1rOVoU$3Y?luD<JdlAq}Ub+H0>sd|V)Utj7Le3N)%ZXE^w>>>)CAx3A+*qh+Ai
zKzJ*2GE{6<_!h^|VsuI#<=<PuxD0jQR>xTf;Xls&@X0c7hpx{Oaomf-EiQTU8QQfS
z{=#i76njY6(M6+`GrU`3wSO=5ICvwrcd`Q+`)(Td(7yFVUf!stF{KF4jS+LSDdWGX
z#;~p5|E-V>2s5-zGg6MCnF%Slog&bj)O@Cyp4IlU8ZM1?)&Z~rlTz;jkM{}0<jt-?
zkAPWcl5kl`gF-i|MvYJwHHtVGCRzBI<nBb7Tt4#ZQx^<Gf5v*nXrzVW!AD8Ufu>{K
zx5yh$6{OnkJ_;>unHAdM!{u4)E3tv9qaoztirQDmvS`>ALXn2XCX<%UEUQVlwXv8R
z&x$1Kvp$J_Lf~A9L>?Np->8fKH^=GH4$Ya!m<I{{DXE=`xjluE-JVRvd^yu$vG$LF
zuJle7Q)zYE3@1mg3j6L@kl`H3$0Zqr6p^q<F@+_X{k4wt_c6c)VUKY!qs#q1<7GOW
zn9Zf5m2#{ja&pa|oK?AUpm2(c^{EizVHJAM*aU`V9~c`KBQ9&Q3M2t$?^9Qwa(A3^
zC~7+roqlE4icLk*DCYE5P}?$~v=VmsAqi;E2g?pIbYzbS@dH_<SI8UWe}ReqXkZX#
zI|OX?nbJnKKoU8WYk3zP!6Fo~Qs)KFXLaNuO!>Yp(=!hZv)*(+*ThM`CXobqaGS@I
z#w8$3yWn=hqg{n9cy!HbQo3naTPsvoUATGO^t{kGEu{7%t%R(d#Y`&t_y=-4WOFK3
zT8OZy!+uh1Lf2R$6%>E5ElF@|qp1gc;Ru@nhod5dLoV0D`pkD*`f&T2s+8(hPTha<
z;@-N<yMaIb(EY!9Z%Zh9wcP!>W<Jb)DT+cE)VU`)cHLVgcIH12XVC3$-ph7Tzcf2&
z`O6-uDv=aguy$B%<n$8|aMwx)LiFda*$^JmarEL7rd|n}FojF?vDLQI%?R~ZoqhrV
z9pI=0czh^Dh2mO(=DP9SN`sDvayt*riQiGih6BU$`#=6#y^M_+Z*O0*v9U48!udY5
z8m@gR5+AV9e(}?4ZhN@Ba)?{8^Wo?)^rh)m_^GB;3ThQwceVbU^1~7-U=%b6^j;I+
z`&$fvW<j(mZXH`bhXj>hdBlF@?4)Vy<k{Z9`x<L+H*=}KGWg*ZEAwPy+@d{@Y>8Ri
zi=cl>&fxCT=qHfm<8R;Y@ayu!hjd@*n(Ko87vkBJ#cT8X%6lu!S&zd;2UpFS_&dUF
zT`G=P$yf65gBl|1mpQA4jB<LR9#SrU`j87z4E!sNma=Z6!UB(CrKAh_ux!mU_8#)j
zsiuKWurBWv#rGmN>Q?+f+-P}}270>T={vhZx}pa}%<Q17azsWT2gh;;8GCw&(XtP2
zROd+RkN`TRCB`-U_UYV}d-OcVNR8O>$(_%LK@0d(>rDDI7`7$IG>-nJ=Se^q!F|lr
zK<9eUFGXYKja!px&TBeoi<=AY?@H34{esyp^29@%oV;-7@F<Lm6Gs?Q?I+K~y%83u
zXdm<*ukKynn#u_5^4VBJcRx-BgbmPFv9&sAVelMrIr%i`2(v<On(1S6*m&`R69T60
znE?Q}y@uZ;mt;~I-dFi8Q5{XNI$clq)Q{J_nl%=NJnKdxK9IruS9Yo8wW*}Dp@<+x
z&M|G0Ins&p;E@?AR2!k@OfIpxU`;0?{c&wsI$XdrJxGC@86C^c93RTX$}3VLaHOQz
z!#SM~BipC*UnQG;4)b-`PyYV9G2qP=-E6;)leH_XuX;YB>1ze_>?jiNuMmLv1cw;5
zc+>-GLhcPg5mw~HRX?#Yk(d`66VI~<lenwzeMQeGM;$jlVwoLjc<HYrtcphNoAP;+
zg3ZIQIhoP@@i*V(J74Zc)Hom7Fp%%{0r<=&^mITAO2xt$l~))2W63|tQJO8`PtQ(k
zIp}`}s5WEAE8u22`}y)JORDdyj-^cNb75?v`ruI2>aY#9l~wb%_HE@q>~Uxi4fprI
zN06-A(z?4pumFw2hyXl&K|wYdJq$-4?Y0v*0wAhgIyj(uj{*(maE{ay@YKiF26Bls
zQ?v0x?}o-bwfh14J%C#jocW3dBw;V4sC5lyW{h*h+^+8jh0h+cK2v22-S;foCZ`h7
zL?7^HlZ>6tSLHDBfjRlU|Dbg#i0z&XqZ4V)?XBUOLjUVI(*DI1AzF)kLIFV?+tD?j
zWG77$)A#QkoKM~zp+SDqXH<;_&6sIxN;15_KOq|?;@MX15!#VrE`1v*O&D&EozvVq
zlWTLQyBL+mW47-eim1zSE$n&`#q`nq9e+j5j|1rwlXmWt)j`-J0#dWGTbkNuLy2((
zXE2{O8Jyx+A9Da?K$c+cn(a-0Nx;<guk2*2|GLU6lBx=OYbj?2n^b{EHT~``V)|VT
z8!5mQf2!iM$`y~w5k64d(IovTq?a44+qcwADH*GK>2O;9{M;ESfX_;=EtU*)-2JYO
zKr^I)YCT=kdQ}bhs&O;3^?RW3d}NBzL6IscEFb;kcvG6IJI-9e2ci<1Qcn3d3yTma
z#Y6mg3{!EL@9~O--i_jvCLZfduH3%ACdRlTRS_YISgRkg|7RO&m*~^T>}yO{6MEqS
zve5QYDLtvZ0w;Ks5-#T+><r0QWTsIrz!>2K6k;XEXza3ySD+=4%LYs(bfis{Fx7#B
z6spD87iRISr%d1x;W1sjZpNiH)(;db3!8E&Ov-bU0l!BHl*(kU<NeQ;!}{;nC06Om
zfS!dZ>I$THBrnya&z4hX;F!l5jU>Rpo9DM6>!$%Z(ws(r%|8UcL0c{X4%FHBR<zQu
z$+c@m*Yp$xZ0NF>Hd{g|z?SW%vWQp8m$F`nNvQp!8Wv)DiMf*r*YDMb^B5=Xm-t2D
zd(ojJZPD3_=X;CWY5PaUKz*E55wn_-<GKVY^+;6vG65#mOwQHd4Nv#wJ@NI`T{u#y
zo)HW{aga6IW53?}r)bUOCfUjgXi@TbP}(ehBx(Kc$jq^h-?hYvoU~qesOGMtGBW8{
z%K=bQ5MaSO$dSJ2GKM6WYbJmio{-V-_lxLlWTJqoVuFB1yvD>hU%C8u?K2%NJv7cS
zaLq|ngG16d2NsD1Np6BO?gwSi;UD}>e?aN{^A_6kO?R5`NgfpGl0ll3L4qHt5aj~k
zlREOz454=>Yo(#rU$HKF^g4!rOnX{2;|zUPAgIMT+Aur8eEZph#=>TNquOWixsZzE
z#!khRjbtn2NHd-hv2b;4RMnwpzX(~D1-%$t^CWd_;7()lTcB2GN}kDm&_BiO?ya5p
z1K3<YTSGRep=pzg98z!p^flVqqJ94ks({VZRmGVy`>wFreX_p*hU75o@iBLpjwT+!
zektYME_i~nA8CG;Kjda$*CkiUwTC}$+jpUSs9=K%k><XecBsOeRzw#`1)_$)#>j~|
z+Gw9GKcH*8ytv{{M0ksw?N^z_+5iA>Xba9pw+CLWd0l$h3_pSeiCM_UzziKw9XPW7
zs|K+ySfYZ>=?IVuIgC|vGh6jV+?mVZ%CkR3yUPZzRr`zqRe%Z4?@oDSqZBd~tVhQ9
zcV?QpvcWHy2;Cip!Lkw}K4Tp3=Yp)?m%7sfxp(9TGq*qF?rjS~KvsepS*sQ6L+8Tq
zVS{%SgZ+0{gMX_qgCJ!X7yuA75l&Eh^@Bm6l5cr{<>y3FZQhidu>8uBnq_}#Nv^IH
zE8mVhTE{E~_iL$GGUd<0W4!oUp6^!-%9^a`sZoGY<WZ;&;19~90KTvt$-N^X9a6&!
zk<CxBm!VCb{Dqvqs>YC870vS$V2A7eksB%4<<e=CQDBibc>)CgYzNq}r!$pqKo;F8
zuBDHV=@4V=3V`7zF?JxekAY1{WGodw32+SG=$`mNa73Gf12Xu29)q%&h*zww_%h#!
zOYXtc%(V$MD(_+H%i#pjtx<VSZLH7xv`U4EGgihx?!g7e+h-zjdCIn58^L{k$>mQZ
z7XrTGtHfJLx>TPjY_#5_=~xr~<*LU2GvI!aywWrG?XfR7Ffu7z#V14r*3kv~@{*?R
zh!&`sXC^_52)n<k$R@+Aqe+5Qp9GW?|93S-YpT2U)xKdu$fV6DyhVkL(K#z41>1)n
z(5}@TEoo!$ZTNA{M50m3XU4cj%dVC(#EjB9zY9;m-N}hZpjF^x<?R<hYhZX2`t_TB
zjSy3K!PjaMlZ<>4WQl5g`Dpu{jK#t4W${jb^)cuWZ_1<Z4pVVrYzumEg~4GxYeAjD
z!_s;rAF|Fi4o-IT8eeWfm28tFY)d=4n{WnB+iuM-r|EYSQ7;KDa9y>u?(>7_00wAs
zB-#Fu<*Q=q@EFjs!u5#0(tdt&7lY&*Ain2A#3R||w)9|N_vPV-Vx&9uZx_FhE01>O
zb}xlReR01DScJn(OxN`^Y}Vq%wP(LuT)L7Qv3hd9#ubeAS({6JI(+C9Gx$>Ox7_yP
z?|6o2>$UUn8RrM^$N|RCT+pqo_GaKc>R*{s0R2!t3|vJJmlzW2qN1zg${hv9v@tee
zwb~b4jIxq}E4mk|y;==OnrQaubk%kA64~m)sI;L7#tmOF+MS^>s8cH<K?LxhF3rGm
zBbIdsH~UX6KzV<V2Wi$l&EGt3cKSlvpA3Ws^1E=$A$Q!FP?ILwD#LTEO#D)sWoFKb
zQc@TO0`|Or0_mC;SFx2V?JvKe?jRq^j&H2txR$2wMm%|hZ+aCRH<Y^9RJ?LCrg(aV
zq~{n<Kb6II8~LlD0(=x6BK+EFZaGS}Ue@XEb_(niVbmnPd*-|RTgea9nKF*~deSN^
z>U!$ElXl(DAzRO*C_H?FXxDt-N5B10f|}OYp2&W!y?P}VCE+3Ti>C+PG&9*DqLtk%
zm%`9}m#aZlTtbX#uT(lrHpXLjAabQl-!f#;jUU)4fW4TjH9)lF!>YvC2Y3;1_p6ih
z1gq}uik`e;cCyA(0sQ10%Yp(Qpq@3qyoIfi+bZz_psAb%u3dH|7b|PO_Sku0#{<J<
z%~p}~V)`4y)r4E0l-c(Tx%oJtkl>Z7U8}w|gYE-LccMo7#~HeQ6r`f>Jw-klNF6sc
z*g!zicT+*8oVOrMH&aX(lkNRaf}hm?v%pU!$MVcspIG^&`KMz)iU?`AkG5Wr5Q<3%
z{fTnZyHBgm&$c#?_mc#0Yj$;w!-j%nsJhCRo_a9(elPhEE%6TVW{475M9Eork2t$V
zdp(DySV@b}tmVciugCQ41#N85P>a=x*&8G;uuC`1Uy^<g7p=UE(M9;Ak=xm8N!@c*
z?Umui=$oX&)}zhPyNWY@-@BR{jgGM9{wpJIeWM1I)0>xr7;J*4A7b=%i1tEPESq1U
znLQnAvu8x?BHK+3ov8)(eM=5Rqsi^53-Vy;gS}=^5Vhb$1PPHrx<fm7ZRZzf^h}%^
zgZO*|F}|3;_ZZiMiRD4l+OeVWHCjgqp1x+FfPP!d!pf5Yz5Nr-bv-5k@eb5|{hBfz
ziPxH$g$2#L5N<*4w`Bve#eQF5o03ZjMPI&qQ}VSP$AmS~X?Pt^*Gq9jhs@^f67%j(
z5_s=Rn@wE)pQ`svf_$-u_K4(`edLF|BQx+xu*oc3Ht79{q8&LBN4)!Z-I}I&Q6@sG
zS4s23dxV##_dJYTnCk$<T$+><5n+J`6OKOF^t#^!>RONpY89rio%^Q^=cQPAkjH4O
z3<%W}*fSRQLJG){E{nW+EkAwk>WzGd$uhpFJcgVWbLPWv89@vh@_X)=xsj=rk8men
z9==^4)V7RghLKWQx5lm5K2zNgggljp8NaRNo&rANc6vMRC*dkW_I6=?@=(=^9oRvS
zn0<q!IFR2(XzJ}^{~rI4{N>#l!!r@kA$Cv~rpKuukB#`g&u%)SU~Neste0LL)umh>
z!FKN%FK|`*#i!6sdjkcF;ElSXstdU*?C{u7(hoJn_F3Mz@!M8;Jp9$^IG|k9%w*+Z
z<?Tj;w7imXFJDc}&9r+PvRr(j*m0i_-}1`%x(Uk<6431jh*bqGPg*$L_AA-b7rnj8
zz8N^5bQIoZ-SCXG9oOIn$kL|T==)y${U0G2;p8<cJ2%$66cIrSV*WN@fpVqhm`}Fh
zR~s0R`Nx4hIl=0soz2BG+8MwVWcX-i8R!@u5Y?#`<}L|p)A6Dp0cy7mg9B)Oz{Rqu
zTpqdRz(7B@O!ZZSt*S!iub(_<4BfF?4Zvf<gL2L3^x|=u0B9>8JIFr=FE|pdva5lg
zU2u9FKws<d2C0&1bE#Mm;ON}Y#d}Km#h?Nz6)JKUi0Qt4ux(%S!e^|kXFQ+8e3*u?
z`%GJh!{Fxe-Q(##687=X6mFC=29Ga~lUsYL>4SO6Uvvw=!*6ZDL!yEv-|s5i+*|*6
z$VvyPmanf}tG7BeNUo7vt3W{5#_(O7kl1&QM&P!8RZV7O7-`u$G}-YT>xcczD|j7S
zS>ReN$qeW#S@_W)_KIRFMABUFPq@%xIC5fsiWX|0|Ef|)r%>fx(VzG$TgpYH=9|M5
zkd=;Y$b8A!f<CT^Lk4U}aC***ZcK2ov|0yCaAMh$;EPoeJx}r(NI|j3bvwSZw%(87
z7*zM}!#p7?78+?>K3VCPjoyeNm4E2f7XX=kt54}<o|0Ej9qvW<oYW-ToS{5(p}}8J
zfHPD=_$&7<2;f)`t8xI|r9&k#21K>65+OJ<zA#v?dY(kb(m9wJN%T0Ez4hrYrSH_m
zKKmAu9OXJ$&G?yV?TK;X_jd(?B4a|-t&hDdn0%?_!TGS*BtR}t?&IVxGHnKPvqg6d
zbPk{e5?brk0PN3hm%atmJ2U=_<K%ma7sg;YiMpes`+zRqde-Q5c;MrYL})m?XRIR5
zHlQGFXIw&{uG7=xK+J;O4Ur#|)#79_`r<6)K(NE<btKaKaH@BAY2I7nGqk=62iqbn
z<nM9V=Eof6{cb!(g$>apgGeaH=cyWW02u3VU`A${Sv9@T<v?hq3B9U0L&nz>vUhZh
zWpvYNqaaWbVJt~yyw}b`%;9%Y?-gH@HP4md_Jw6_zRLgG>o$XPngMRSq-R}NKCxI{
ztFITG#^38axUcUOaCkF`-{*-B{awE1*DarQ$@to*|DIL@V$SHs@&-<~I){P<v%wTL
z&HAwnLBJU|GH}aBr5IsJuveFZt|}?n8RGZ4JT}-ves8tf2Rg*KgT`WNmjy~G3B`wL
zJC4o1e!Daz4=?$Sz?w7&L1QBNbx5#rtmj3dER}ba8mVkB)}Jyz*Ub*CC0z-g1m3y!
zftZY2#OtpN+2=yv&%#%Vob4^X>R>a2ShVH}tZ7-;x@N)G<v!-BF)NCWH{=3hx02DV
zTSdFB)`y3iyG5TJZzV(uNmx8+DlJOc<oyM*N|1n>1)=?SX*QY4O<U689mrsZoUFur
zCD~_Z(Z0f6|N9DM&+Yq(lr1%pZk_%G^{)Nme9q`7@Sv8KV-8961Uj&(ES34J$HzJJ
z5d86xI|c2z-PvWw>b8b$@MdKZFRuIqHEoMR01!Ff<xjldbYK7yKRqkIRGn2-ydH3S
zz&r(gHIVcD6bJNQb22af8Borl9r&SHN#`%%dtd>WU+-!`c~!%Ky1$9AEK2e1@_FHx
za#~tV1Dor(RFO4iROT0YACzD);ML6}*;o)ejZvQJ(FrRyGu78`MBwN&6%k^15c+F1
zX#3u*;%yw6;@#lQ*{`FRjV9-(BX#cx!$lu~6J*~w^|#oFxO(OzSK|F6MuD%FgeSX*
zjP1X8Oc4iH1N75pvgAXvw8XgN5K@8=*LR;SWkv+NLZ`O#%`}Pa(m-L2l>Xw6gA8<l
zbZ7x0;In5~GkNc;gBD#Aulpt#2%9L$HJmQ{Cp)kv7qlM|Vn%;q^FRUeMQe+kc48OZ
zF3T6&K5pb5apjGGb(fIW>EV$^a|<*{+H@oPIL9b9N`6@2%LCH2hPQ`yoU#fz<ur5A
zmJfm-gqEj8+51*c1%3umv9VcMO+qQ%V#yL-F|AGP`ndA7R|N6x@)*AReK%t`)A~QA
zkhy}+kSFd+PV5|K`F^(1Km+OMxaDN~->b-Hzv#An<({qm<pCfGvyA)h>P$a)wyXSN
zvsz;@vX1-dijx|!;Nb^B0o=diNi@FcXf9VVLna~#N_zr2zGOLA$4-v(2I_AFtv*Kq
z>>^+1PYO|c`n7qbZdcu!xX+F>#i}fX8xm6-yf&tlpm}9@Qhy0De-}04u+5DNn;^)J
zg>S!`=xGefQsQALyQ=+$!Li=4X)J=mR3rgVzXq1as=cqB!l+MdsT^#Yguc$t6I)W*
zNsGjqh;Hjj29GhguYsV9qw?T<ve#9@%nL)@5D60s0M|iA<gxKY3xIP-TJPNnv-f?i
zfEs;;`y12=C}dJ>B3+P?6pvDZ&E5U;yM#2pp&;-|8J?vTL%>#+Dh2ej3~8g98xA_u
z^J$j%sRA~C2Ne`(Ec+m-Y44_^0TlopLjzA;y|c*CQZ@UBe1t2t+;9F30NF}JCGd}4
zPE=hVQq<g;mHuGI+U%274(pIDo~@0r@pq~cWhEv|Qz3^O;F_}KCZxz>1xi&njzzF7
z(D4u=+FGOmNXC!y*eSsdAw&aN=)8p&#_M?%KA-~-N!@l3SY8_WvI5&Ah<T)U_zDgd
ziFk0o+pNv7YY!19IQ3bk`afR)$}Nl{pFr8~#umf-hAS#3p1vrUZRn>q14QG|LVS#j
zk<x<ww2k=%#uXLVQk+ca-)Vv1?)gDp;N<v24qs{wWl~sH6cOlm2(WzpV%krA>8GQ?
zxm;RmF|@z|r%D=7ij8)ZZO%<m06l&<`Fq>dbu!JVowOn7a@cY^e3cRfQei>fyP{&@
zEpb&+_#sm@HTSsZ5U52C^s*^D23iAxt#9oJ9jnOH(P(G80gM0rov<}wqDHADGOLSM
zt-GFrf&+}e1mH|rwD(W_SAg|C2;L+e(6os3Z5e)<YX14(wXS=?7q|K}ASc$HzVf(@
z=5zCH<@(zqxSs-<8i*-u@63K2%r+`7U_@O$PilOnDS777DMX2GSC9tE1tXi#0hF+C
za*56nJ=~EF*GOAZR=Ty9!fHIj{bDV*$#*rX0;3ATnht_mF=k?`VHGVP(Yf#~07*w*
z;_^cgZ|uq5{#DuQTV5T}l@sTa!2k@9R*ghVSY~)lKd_+6cjFcyq*m9dCwv$aSh=V*
zba5n-FjDuB(m7Bg?N9Yl5x;oDB`{7uuH&bVMsW7=&)#3O1G0ldarfJ_C|^uEgnjP2
zk>=m3b*?5K#055CNM)dpQ>9El9S3IciCZquatx6>n7H(AU`Y?UTrsv^BT=0uf7x!3
z1LyKic0iv-24%h&sVowh=m3z1*gHZ~3&frgHDWoDIArQp8ZRoQOd1D>vAUY2J$X~S
zRZ=J&L*qv~JxhZ#7l4^F-={0h%Xocs3qrx(0fJaC8>D~NE7}lii@N-}DqDxG#rqe4
z0d)CIY8Nw*uCgfhxR?c0pDVH6NgcLccm?%A3e^4!wB5p$&XxHo&tBe}GW;J$%?m#l
zhqerU@w*RK#aU=0`|0g;&$PKPesLtC%CC&bIOdf)bK{M4MS07oW$*qP%icn?Dap}Q
zBUXNJKi^EB7@+U}lKgqyHGk(Bt@SgtGBkrC(H0X5$k)bioEyj;ruP{@_DqA3RE_W!
z6r=KSL8$)H3Xjd&xMh+G#HIWl;6uZ+u~6SPd+bjJSU{-kS^5iC-qecy4EzUO>wHaT
zT*<~afs5Qs3JsxYt|ogkQrF5D$E?wr;l^L}{rq||!|+6e4)9iHG+=i{8LMnxM@~9c
zmlm06bgL_!$0GA4LL^XpvF6cJLHJK*;!YcQo^SO=^&dRr|F7(*3;~dQH%{XJgmSmM
z_&)qy8S8jzd`U9D=r)6zaj#`R)%rp!pZ6=_PXRxrsJV{kE9aIa3+Hyj(gY~i0B_8(
z8Sqd7S#fFkUHHp0pa4o#xguB#ku<6#1BzhB^F^g9`t6Xcg8201J~e=UD=T7og1WV=
zb+QxwB+8(kXizW-E-F`1Tbh!XkyOVsZuN^#L<<72v}Gko_#>LoMD)D8V*W+tmRX0<
zUT^<yzUmU~rkB22-rv2W>&@{$_$+ylEp<t@1o9&~)nRaj3@-BeSi22(10JN^<JxN_
zEpB~#fi$?feF~L>BPZXewhb%f!1JwQ1BcI8K2R8ifsF+hfWQ)>kw34k{ks|T(Di^>
z284e?7hJ=&`VInh;(mZ`Fj&&2RYe`n;gqMd?1e}u)B)Q`4UMikICF%-zVH`=bG)B1
zFj|=uiR#@q$7P5$Zxximodh)HwNXZmzkD==0cBj&ymo{OMA2e`p&%Crw4O5+(MI}j
z>!!U!@K0c*mX`umzO8yc$xHLW-g*C5h;V-LQy~OfFb9XlvD~T`Ck`JU`tq%kbX8R#
zPbxm~%?xr%!m}`mf~4u~{HLgkpXWZ9MZ?L7PMxICjt5MY_AOJ<vK{xbY?@#rnj5nE
zTn%?c-Y#?N;?RicCrGLwaF0xsB3>q|zH@Fb5<;P>B5`aQYe=V?3weq{Mk<R^$Z$ee
zlr~2P@}snDzou#G9owx8RxX}*oJq&DB{yFVp4^?U&0WlKJPNrwM@i|GIUVNHsrccC
zpins|p~~s%?pI%as*4Xe;?37ai$gZwhx=P%%)&;p6T*`hAR!bzDpR*k_okL3r|tX7
zr#7Ok#d5F9Yu~<}bNcfwI5|BCBetc@03SM!bR}WcL_||A9@zZFCdKo%Hw&uO7Krd*
z-}3A3fAR=T(P_~gyw-naIZTQjKVC9d?lJ)c`FZ;ThUNa1@QLrs!@~0N_KnZU4JY(e
zq~k&a@Sr_$FD!cG`(Q;r+!Fu<0ZXSdS;z^&kgw|j(pGBQ8V=SC$iTcrdWeV1FI?=?
zToK>=!-%2KsvPZ$MCr*p@2zemlK5_jRUoS|%M2?%^^%O^fmh(Gei|E~Vyu?<0)l4}
z6q<SXl16IM=Wo>TwrU>z_S|~tUdHL~`HJYb9V_=;vbRL~ePjuT_8}SbgA4L}4|>->
z%b(@~7Kps?_T=aUAL?ceFm6^r)TWW%uJuQ93Xbg_T~HYw9S?$09z)pl#Rp9W;rnyN
z=P#&&vp>|~XZzE9Di(;7ky}{@{at%a0AZ`h@)I(-^+W%h)S!wk#3s<SeX5CqY5!x<
zdFIHG4}1RQN~1&=ni*4Cu9=1|z|6S@&;88*r<_wzn1wDQfM+Z_0r%UulE)wyAhsJZ
z5xg@{M>mrM9GqDKGiqk`{C`Xm9sKUB)P7tqu^fX&0+f3)DWmkEg7BRLStw4@go0Km
zonm=PPaN%>Puo5%+#F4Z9`W!{&gKXw0}rMJwVE0hZD>I0fv65=dS~JmdC=hJ$44ah
z&Au1h)48WWHhaf~ao&<^*cQMMmjAKowf~w%qk;{RgU(c*NMduU9BDQ#rKf+Lg^<YU
z>PW;gh9{_f1P1c=b2rVN3La-CV;pQBYMr%Ca<!td+g=}%d+@meG`9!|$cSW<lbyVv
zT7GQ~i3x#__-k5RHR~)*<mujtkF|r!1gQ0o8bh3DUEDMW?~)AoKtFInHw;l_@l2)t
z7*BjGQN}SdR@7r!Po2Y^rfX8$!PIU+lt)&fD|xP8o|Y^=z{da*1koaQtV;)O3aq7b
z=|$IdaooL0g7L2#SK_bNQ<HK3Ke$brb>YTB`Fy&e(%NyU*x~%a?d+y?=Y5N~@2)iE
zZ}W>^r-kNNb_JqMLTk@g2{Y=TU%4b`0Ht2!>ZpRH2#*42Z^9vE(P&yBp7EkO6iGn4
zWB!wK%z&F|u@upgB}}>t=$qVs-H|a5=GsMhlrf?MJ8pi2_n-$-u7t-7ko!OR_`F2s
zqpf5kaAT(?8Tqr5E)X2ngT=4&S5VfJcVtt|Q%>4ku1gDqYUN)L?Z3kyEjzl-s;#H}
zMR4oWyr+DCLgKr#$kwx|ecu~xg-{Rc%I8pQgZ6G=O-Q=Aji_%f^nz$#)VF=?f{1jH
zUwZ?Pi`wn>5@irkKkghMqPf@^TAr?Zvqd{wJ6pG>H3AkBzn!q&bi^66j>>bbczM~Y
zqE;onXPA|y16Yymf4`kx34AHt-+-U=iq9b4|55kq)2#ndQMpDaYBe_2d7E@-{{fH=
z6kA-RKe7I2tMVb~Q9E5>f1{A{SaP%>zX>O+u+BgVFcn170Tr?h=RuC@5D<Xwvz?@7
z-fR1-KI(?M8Oz3D)#_of8N*DZVh>cu$x*Ee=CS-KD!Hlc(+vEj$`e6Pw|$`t#7_bx
z0h{+-U(+eQ4YC}AdZbgwf}6@zH!opz{c;waX{d2bRc$9md6-0JeqBBcC_ESDB6nv9
zt|lt+3aWX5Yr?h0=CnCeL~NxsUR=jZgi_}gMXdVN8(mKhAma&GeuF!k7)i|NFeP~~
zfqWWW#uCH;S`D3zD|fiG>Plp+d?Oc4$KQrkqKxzGyklK)DkEL4LD<h3`27aM$@VWO
z?8%GuQs<$KkD|V=7phM7=Psf(AE^*q-F-u~-|n4V0Xk`Z!wslG)IZJV9v%5XFuEf^
zJvqT#@8XF%$k_6R0s9S>6_xWEh4sS!)w>JhPVE6Vwb@UH$K(Cc6fNr;SQNs}lRH|X
zZGt<3)$!opC-L=}GS%@mm;&$CC4QEhn(B8K(`p3-10~`6)ybpCqcXEW*<>Y50%&!-
zmCB)6nJQ`|Ar-|So!F5&m0Gt3mNcn)=Xn(^CN2F0`ASL^AA8&EHxA?Wz0sM=x>4`%
zYE#GrQ3D6CdWfI*Xjp;Pz^)a9?Dn=gn<~W{&PlXcfl&hjvSQ5c`ZcF+JnuC#j?3<w
z1>Jr!ymE){e`<U}9!BHXDO5Ayvbvg4BL+hHj8Z%^VG{SN9u0(t^!pqitXknl&Z~^T
z;PB_hC+?+~_Sb9d<xsNaTLsdU^J&j4Lk5&x{rx}>9Eb!^n&U8ceNkkKbOIisv$Awj
z$Pp!P>-nQK$<lI7d>PE1MZFihg{_Cui;q5;d_<PChY3H*z)&o0Q1S&zoCFgR@TY&Z
z6}S#)h?qA#XBhfqx&A8=bZaM&o4-|zC2RFho9U0`a4u4|`4QB2`qVZQfyGFDdEl%3
zdVoshoO7pF3reA=jHxOhlMTwaAgeEWgD6QQv>1(2G^P0J^PYS09j;wD+myUEhd7~j
z1<d=MwL4)EngW{fstZ_}E+xT-jX6@FBQmj9FcggGZ<?S7;WTm>PT%uc6S#3MYav=5
zD8In?$+=NY!RJj$8z}V}qAzWrn{b)=t^h29YHrd8u&9H;^fsXld9@>eurI`!=N{MZ
zl{*PC5ECRV6nafT%Uq%bu>mX(!N7IbNQ7SFStytb86oxhO=G`DJ({qX2VWmU^wW(O
zyW!?Z_Twk>2(*<Lyh*O-%BK^fB5xvr4Z#@>owwG6>2OFb7zVDz(qBUHp;f|DUhBCO
zDi~-PRhz+I^^+_QbV}P2%9EeptURA`RNFuNAR?FM-?406vncj{!ZMtY{2!5`erCHR
zL}^Uaw=t8o?^vPmG2g4T_)Y_Ji@@;x(?}H;J5$ux5{iontqA^TcbLlCftQ#zftUuC
zmID!S&l8^IJAeRHWB$}UBn|gAFba;tbmY&3MA2(n?AjDl%bs^euUY63@ZjsHF?j79
z(h!~H<)z_+h!5ZJ%q#H+R)5K|j#ZeNj$d}(;W^5Enb)qoRFDRJg4k+jj_c~WF1a0l
zzd7_=S+elGIk|a4O-d8fr_EH@{%Cj?7Xm4B-h@qgb+1nd35s8T5(*q)P_<WOaoh+i
zcsCt#M+<y~y^SSuLya~e{^w#v<zdad5~y80dU#`)PWh15UEpldm^^b!+Iao~6oFw0
zo67}3puN7jcqcA(s-~)!K8LfLtaIY4uc0qa&*?%fx8i5bAETU(QO`c=g+c<z9Lk!$
zCP3qcgZkU|#lf5J#icg7RxSfy@F&2bp<){tt+h9FiyVlG<p<hRq7;Clt0XgHw9G`*
zv$FFqIV&$bFz$tEU><lXbKjhsZIkBz6G!3Pvpp<xQ+v1lLb>?a)DtTS^$KEdbUf@!
z9PrmoIXNt0_10^Mv(b4yfG04cgv$)Z#R}MR7~s});5k;sAHw(nR%<&X5DhMqmZ}4h
zdQGFV+CYj%ZI!#hNxFm$h#s^HS!|14ObGjOA$aAb<<zoA4=1;co8k00dQZIOwYzEE
zI2hyBk(s^ha}n<{dZD*Rbeelgi;CTSV<|Hg%FE1TDYqZ5(e`(ok1ReyL+|*r{P()-
zYOz@az#W<v#Hh2&>we<Y5yRoW;%0cG4}TlK`L&h%<p<gG6T4D)Pb@E|l7b=&BtO?o
z_YQ+I&x@h&&*ybDNU`6LaD(G$yZOMFVtXKz!6(Mj(fh2KvhuVmzH;N4F^3GvIX<>q
zm=N-cfjk&&SyWv10acO;p}<L#(6JOp;U_Dt0?r3WAlu-3G>r8DJ<j}GBL=tLp%u9(
zryhNpd4kqpzad@pXKyUJe1EKwFP1^9{Ey&SA@)c<7-=_yj?$M5VrcMo&yULW+!^ia
zx(n*{H7jm5!Sv-m(?KxvflhrA=qqx*gm8~|M&)jOTHViA7Qn2~Fk4sdT}sS}QmQ6q
zsHvDbGBsFd4Dsf)fv7bqaOSfC&U@(7Ma1k7S3n+zRIxB?UEzP$_H?MlBlAQ1!|}$`
ziLdyp^z?LAHnuM}w>}*m4=>q3;j=V5cWrB~Nh({AoBYPFN2mguBk|J5;q%v&1oOGd
z@%4(|4Wc}(?zU=GT$9cwKuOb^Dmfti7cuV_El<+jp9ypXDO9OFJrudb0s?0E5<z|<
zr*V@&;|gQ+f(CM1*}9*847!VB0`N7t#x}ZB<I;y^8r-ODRNnVl9L5hl;=^7p8^-#{
zw;P)UkDZG*IK3E{<^*IL5oM0Nd2Ob=%;<{cr>NBruJ2CKnsR^KwLSQTT`W~@P`-?I
z7!?S5_Uz1m_U5uUej=UKGJrOKb}bb}Ff9#eBu@7J3I0xQBhFOH1Q#(-@;in4o9fD<
zygVB5172sAqSZ_fK|Z-&r(WU98I1MWKhjJ&`6@d-<f)mX$A4l!_GG#tfA1mJ`^8TS
zAJCK_d2k`MP=|WL5y^!BElhEZvsBk)aDohq2S|cN%Mxuzj_$&$(;B#8A$oLh$D3(a
zV2~B7ieQG*Dv&)`+XGP>v<ab`)lc^jD{aOj-<L%OVo{SLxra5R<36Rmq_w69G#ZX$
zMftuBZnh>kpoG4Jttc4ueA#BD53ADRMkd^)iwG=FQTfFQ44<@Yz}W<LbB<tKMEMC~
zAKLfppy{QWYg$iry2W~xJ>Wu-+JiX^A3o&QYPXHK@rxyAXFOmUG+fuq97ApMhP<;S
zIWWQ!X<0Bd1ch_+Bk2wj`;tPyjFPv<k(>@?QOqk{e<BueoxBczyx1T6^b5pCR*0=c
zrz72Fj(f5E<m$h!lR}ZK#HkXR>^1IucJ+H*2y?8+279P+jbLQza3v$iC`_Val!8G*
z%QC$`7T>RIA(k?Y&Ld5|pB?WKni_C(M0X0rZ6{Me9<2MLbAXIpm*;0TyLH+7kWgM?
zE{4IW5@#bV_|WX+P-qC`4BW8u6BC7~;*W0xhyaD7e*$X(TyBCHELJ-y&xy+diSErX
zv^a;`)y+R10U}!rRBPbRfEQ8XSvr=EBu64TS{I`b^m6|7NM>;lqmA`{me~96P_)*{
z@1r)ZM7{sofL^kdTj;^t7n_c28_z5~Cr_XX{1bimhokr%ofq~%*}E%Tb*hrOzx7lD
z^<Q&w4AzQ(d&wpbQeLvH#sI^J6aEt=G|fRmDqUghq>J5PszIG<JqqpB(3P`c-|B6}
zs@?C6m=e^bCeK#2MD`v%*e#$A)aPMmJCX-HrVm7POH#7m-^%)|PM&;NvOE!q4G+Si
zA7==Mj|{ZhnorY0cU0!_(C1uC(?ZbT1co_a`)M#HMc)hlB=K|4R-;Q|S>{a_Tht0@
zYC-aO+}1clQWfOQCaeaA+H_Yq$IveXC{_i<SwG7YWDft53qD0z2Na1sjPmdtV@29H
zl;szgZ?rWPhyBiXS;De!KD<&2#fw-&Fm0<68hA37wOTwb?9<0RL=pD<Ee{N5#e$7@
z?K3uJuwGE3<d&mIGv$yl_;<4VR|pLUw}%vQKm+;Qwo1&v<lBQTHB!sm4RSk40}o)M
zBeWxnSguw*b!y^KLh&TU!-$>;UaU<qaLKT#P=231vxtIF^#^%jF1na?qzEoJSmN?-
z6l!z(aOz`_{6u76@Q88IpvrE-K7_daEnAzwcdRr*e8s695{so|(A76{l8>{fnCne6
zmf2uAz#?t@7AxzpHzKbsNLG`J8N>PkU+(#N$9)t=iIa-%rywIp9cC*@Z6Zdz(ob2}
z5i5Pq^_-2i;@Y#$E^*yCH=zu#B~1N&tKY~319dHssr<w);x(ukAzXueU*_<jw~SZX
zwI(8Smh1Uk(f3C%)+4vv*rQnaB-5*gbh*3Rc8*ONhG&%5$(%&Dqu?M@!{LWhQq%cC
zDmC1)9zp-k$w+*K?xte*W0>#GCnBG~*VCb|9HcmPyx?ZU?_DpKDhu&geur1Giop~o
zwbb?le2h_8s#Nd*g~NG+T=TUS#u4v4h1&tS-bdv^HaQP2XxDY!Prh2qp}H$ATukPp
zl%%8);yQwsC2hUbj+a4&P?w}Bo$cvZ-UPwb$Kq8rQO}inI<<fTUbAK#u(3XNu*vqz
zNy~SYkv;_fcj^%1HN=y&|1DL4tTXtN1A`l3pG!Gke7}F0L!zghuUoI{E8F&~eXlKi
zp$CR9_wUEco8QaBY21In0HS*H2|#aZ_tO-K4<v4fhSZJKq$a}g<6tc4-;|j7i_*4a
zo8$!b<N*m)ezszyqvghYhNW&bO(|<?120Y2vt93Gb3!0R$2-P)x>BzXLsz=$(o7T3
zBKgLjezJ@U#M(L*LOcU^Y8!Cv_A4D*hIWq#!}riuw2#Rb_s=hV$to*-9b%{UO-4yV
z;_M4UZB*Ky<-|*?kTDx*Fju6&$Ac4qz{2xkuI7(8(ufywwp6>G$7wTpwqn@p6aN-5
zW>CWKp9=R{oF(W_BUA{Uy-UL=iIp<2LFR_qh}XtmiKbpHR$R4YO@sw!xmIIc{f&+d
zh3&#+yqI*YqXYV+fR3_yzRBe%dw4YVyvq&zD9}Oi`TcQy&My?Cu_v5=ubpULlJrU+
zG!WzGZy~IFZDCHOsC&$&-^&ag>@;hp!YpNCVm@M{kgE62)~!Dl=i?}-urKSVeTgc{
z8n-^ipOvJg`w`{$bQBh2uiRE-jfjO5dh0*t78ocbvT1>2zkN5;U-7*nEQ09h5obH|
zWGznniT1diqdidjTu{Tq9%ZNZ*uaDaLXJYS$?zOo;P!}?u>7|l`!_DR?VBHu!!wj$
znMofv@6!c!GMh_Q+PrTsNLH?Nx61+PjDwxplr=&Ev?ksL7%u@>H7`;K*{Hv3x3?lw
z|2E_cbt{en^Tuv^5K@dUp0CHHOWBnZP+z15$-}ABPpZ@|UcP+Gq|$^uB&EqSr6jb|
z{Hj^@$V--Wen+gt{X7@*j6MGMdHQ{&o}irl#>rY^N9y2f!~WF^U*aE??Z@<x9aaII
zZ*mzF36X$aCAQ9drc*<(uz*^FTB|43A1j}wz95qkplPDV`}B#m@u9$cMKuhaDXL&1
zT1ig@3yq(i0H(ui@v5t?E@2jz1+?{;VM9)ZkSW7i!W-BMBXAE5re;CEVJ)yc2z03(
zx@C5x11037GfqiB$a<<K>O1v5{%2iS4e}WLKc>DiEUK_uTaYdR=}ut)rMp337?2o{
zmhSEjMM4^eMnVvf?i#v78l+odK<RG2?Rn38u5W(xW0?Ky^{jQTJJ!Vm?}ePadnmpC
z>aSsSk_JqYp7`Wk`S|~B3}k0f$&|wA>UkkK@-KWO@aZZj=*kChom(3lN)${jG;rJG
zlaXVLk5=L>&A<U7Zt6|sB<IdeQDx%ahdGM9<ag`oa_nutTx|ee^zXXw*SCYA7S!q0
zRQrV^1@vF1BY8g)no-XR`MPv|a6^i+>6~;N@n@9M9|(2nV|>WeU<|KMsi*;%%j{aw
zSUq)oWC(U)&IJg{PtH7`HotIG_VNCam*cm6W_B>X?lBa!-nS`<Xdgn~ou`+VYh}2@
z*7Nio<rdMC_!*ZK#JJ{mjwn_hxzwzsNk)MH>;;mkkKf&fs7r(EKA)YhUA~_`**P>*
zl_M>qP^vo$U@lJ&=<0LX)_`kx=p)b8{4|`Yy@~M7U5Ni%%6^*#Xyj1sRoDT#<lmXU
z9|6$d6?4DP*ty%KASz}0Mt6~vz5*w1ng;S;V*@gPxDB#IvQ`?^XFS#Vv8H-Yx%Q=Y
zu0lxic<{*i*h%lL(TPSKgU9H%?%`^@Tc5uCIvqscv1fh-1*9>N($n=FyQbeG38z&q
z!Rh=rW0pB(h?~NJoWx2oqBcWRj+GwgTjuRybe<~)3IsEgm}4QTwe?!R-Q@Nw)g;1k
zH}hvo=5^1}M)yzk-_JSQvN)eg@;_zxtR9w6;50~#bq77+C4nk#mgU-p;>w@s((PC4
zn%J|iw)b0quT@E$rl7b*$Si%q{tl$!X5<<dr&cZERAZ!k#l^cK#1rp&V=aNZ`zrE9
zSsZzCN~nA%MDW#^|AZ|CMBtSKlfYc;+YdG0gFW-SA6-8eHxMmOadIHggk!)2>|GHZ
z*B})M>E$*mbiSv@xh66Wx7&F^HF5>IT&`HgtS>FTnDCP;lXVAzBqHVd9WV=!79}$-
zQvJKrtGcXChJjhpS3>i!yXp5xm6ZN`npP%F5_kh24<}egH2R{;ErCdg&N;TgvUeyn
z#KdM7K~fw-dScabmVENTI0jrp1|p%{sOT)sr^|Y;j99D^4!EcIPj6*wE3x7YbXU(L
zPhLuP_4oHXw)|;XdvIKRIP@|)yZ*P~+OZ1Uu{!)U?5D+(%H|vKTeBmdSQjScE7_`B
zy1HNz>JeLfNlzzE?Oa9aQkQB#&tdbIONy>_UY{(ycJ+w2lFF@kONY}!t=~N{^R41#
zqwEePedz&y=C>TXH_?zW-GM&EK7gzN1vS*dKDmEEvZhlvy}YBs>YxKiSwxFJ!Z7^t
zGa5)tsMXi?r#~c}9`3OFD4YvpY+TPiavEtbxAj6E$)1#+1iihEz6!zPov&m*tKHZl
z4D;f>xE6fwB?H~#n(j1xcKe+285x~p%Zup+V;ReL8$u6-TAkeJDmX)i7Z3x6j9e8j
z*-0pndZ2D5(3&ZnA+~|rCi9Xm)RY((X$o*5{1qN>Ix2s)88oRpSLB73;VQ{&%f@6|
z3%j0??vXuotoeO#A$^8)`0$j_KirPzaW1FoyD0rnRk)C@rCi^qhrv|$vZKp6KXu>d
zAaRPToCACB0<vVX{$DPD+>%y4&<YfL@*PFLJK+ucpZOyMG8H2`8`{W;k@TF(A@!u~
zd*cc7^wSHMw#yVdI{UsUzrf66-p8f#ph?wkgOW6*=d)qVqu@Ntwc!}aN3S_zFV5;b
z!H<+@zer7FPCHX;?{C&Z#LpQ@MhP|m?6#COI>!nIhG8j|F|og+4V#T%x)UK_*Iu~R
zw++evB%?oE#0N|*ZwaFb>=Es}47e!J&Ye1jB-qO{OmsaYysoC25oRQYPinSk@I0Q*
z*#Y63!65kK#mM8;@!A7bQ2TiwQx-GHbo|+OSa<^u@E*mEF!LcRTHxw#A9VpZfpB$v
zO+1?BEOM|~q_|1pC!)+W)!lw1?3j=0HW;ByoUav`(p8b2g5Hz!{ICh5Pj=f%mIF)E
zp_<U8YXrwRS)NloUMnG260JJTEh=@yl{%_XAmZ8*V#fckBa+J|My`ZL8a>PyZR@vh
z>t8<}Jq51btzH>kG^MufTinmbJ&;xUZ?hjAv-@8XfQCLfDm(gr0duOmo37Q6vJb&c
zx{9i(c6)(urelK!{2JPApbnxGk3a+1Xo&_@u#Lq$kd6&|7V?eqZtF1Fa&V~6-OJ@i
zxtX3V&BEol-CLMnD#iJ`!QunZ(svMW6-b-R1B5>bT42M#+6eF!^K0y_h`(y|%>uHL
zF1OULLP!AJv>4}Y?swuFtxm9R`}w`7lXsb6^#z`U|5M{hW&xUluh&@YhO2=v^Ha;y
z?fKZJb+6d=?5LIT?OqiXR;B|T=X^@HIT_c`gZ^>`XfPP7upfDjgr@eVKBvKW;Kr_j
z%r=!>PcXWm+im)lygV3|wDK&=ZAmzpzfLQ|nV6uR0MZ;`>i0(j=t(#DNt5x?QrczM
z)812xpXjt*(w{GeXZv$m`YK<}JtZ6~qO1lM`dG6dmcP+u*RS|;ei-g^8KRp@B00>C
z^I8)%YjY+T{M(TVkuoW7G%kjN{@j%9s8lav?IGv^Qg{sOHv~_h>0du#UGdwxu~L;b
zQqkrw_MB^Z+P-FhBoV^PP|5-P-Ehp8FzQJ$j{{uTkH}KjjLJuvjH}l86Yr5n&t3zH
zD*o~`t;ZHsOef+5woggHOvhK!bEj6M-P*t3@DVcb*EUC>j7gU<4Uc}2!eYSu`*Kt~
z3Y1gd(f3KEcd^Pi0P0ebGl}v1Re`0SZBf~FDCz3<=pUzN7Fc*WPcIw%jNS=if@P#d
zF@*&O$+g2C4&vW}G7`+|sojcR)(X|GG_O)|h8|y8mB*XprYkYqFm(zqR-cn7K<Hw6
ze`63Rmz`1G^?wR=v1r|vp;LU0=`Ght4g#D(RDdA(xps6i;++x`Xa4{F<h&7MM_E;B
zz2D|Uh#&d9*gQ$AujhYb*!smSc6-|bk4i|YbmrE)qt^<^>&8d4HtvmSx><g!cGMcJ
zj-QGp3U4M*8r^Lsj)i*y#Czb#9gIxGFw1nR6fRI{6D2caF(sVOG~!9C3&SQj+3Gx^
z*^jOoI$j>$=2du_%FFfjP-X<0fRnz}rIe50Lu8Ao2X@SW7UDuZqP`usike+ZpTNiB
zddXcmm${r(N>RFyf7mx%xYn&_lTX{h*{IgIKA?beKdV!Oi&O$_r@zNs>RTv-b4G;y
zY1d`c^~!doNx8?1(yzh;ewbdb6-It+IDXwZ$fE{YjLsbu6`vSe-SDSY4uS5!zDIHp
zXIBYkXz=J>nu&?ET%;kRY4wt6I{`yE{V?zBKuURWA;C}fR!`3I0@~F#K>d8ak_LwO
z#(dni1ee|C1(@l%m_WJT_=^@3P2ezN*FMWv0-fLPOD_LP(^PeBZHOHRW41y^jf!y|
z?}t!;MO->V0WC4t4+3!P=*_q#(a2llzZH}W+7wEx%qAZ|7fJs*ktM+67Ld>hp~u-H
z{FlZ!#T;C&2x%V%B`5sG5P<(gmgszpYZ0Yc7LXBopj=yf$$hzE4&IH)<MhWl^w~HJ
z`tfP`;s*UmT$CgFg=`BFWYZ1h&`5<mmHWpX`R3GQByo%yKImvtbdX;`rY}GV*RU!h
zFQ0aEtje5a&ODhOPX^CgDaclbY;D<Z@sDILBfP5#kaYcX3WC`UH`Tf8x6A(R^@kjZ
z1ple8!bpPgAzRB6Py%FbCPMDo!y2-`4~g|*)fMw;Nf^vFp-98!@#j2I-#qivl!*1y
z*NTn|En;a?!^&#*?G&^heBS-%Gy3ECINPTy!g8&sL>ATA?2sH}pg95RgxhV?0HWX$
z-Y4J0Zu|jzlPc~K`&Du-<y|MCQ7*lj44^VX46J%Ja>c4pRefh;+XN%;GSl}9)zyDv
z!rt2qBCLju{2&-r%PUt+Zx@=xX#H0ID+zGi)k`oBFsR9*EOV)|!lqy}t|upeC~E1V
zg?5Py4zn~`dNXhuoaC)#na0}zW2$;CH(mzm?a10Maw;SLtwtM2543W)%TU?63(xh9
zGCv}WoKyVQjucDt0k$7A%H1Nt=9IIePF2TC0;5-~MyHQa&%0z8zj?*bH7JEZCa{RH
zZKNpJb_nH&P|8LG)uGNN>(9qhDCi%QfuOrHn<~GoS&`ZPKxtn~9jI>jOGB_zEHh<H
zJaGP`3*kcL^O+P3dtZ{AM%#oY_ssz(fF`?DZHXnemSXOqS`h6@TFvsg7V~;yKc@;C
z)`4u0$`MgJA)BWCJ9eIfs@YtoItI?SIftb7mNWLssO8{^NTP=_8T_g7dZTDj&Rqc+
zIl+`@$Z)Qs_e|t{#@Fw_r0K}QULG%mUzR1=r=%db>T_y--*byHN#^{Gnl}fH>W^WY
zEJBMBmFlY!08t){a8WJO*wL*wzpo}`A-3y*nV1aLlX?Rb(1>3Zay>7__p;8uCVV1V
zZb=$xmSTH)Fwn!YQA`3nB=THipC5ct+QG7+gWE~u@}(GvX1Y-w0zPs?nmao%=y$2A
zDZDPgy1*?^6Cd1_@0$N9&mIE}4&+&a!dozh1^0N3G|?QWu>U0(_U>p_CE6{I_|0KR
zQ^+vJ<5ZI|Vjzb;i+v3cc^UJDo53N+-GPO#t<c-2ODPp)w%dZj;U)%pc<u~%9=}=I
zj>?Mv^SRO7)&(8U`r9Um<(6z)&ZlV_P^G$C6^{aJ0GigcjpN}Amkrcvz}O7Woe;3+
z%00cr)z=sfbqTEwHe^WefDESArx+4?6O0JUs8=sz2`wt^_PeJ1#2r-wDhNY}5!c5`
zyx8Xeh2*UXhpT3d0EbUe=66D*Pt$@#&_avet_Sp-sq1Bqv95HJt{e5C9Ze1EC>Y%V
z4<|h0%)W>@6Xy#hZpsq3P*?sh9c<nACy({fO!FIVab62+v1zJKN7|FFymWKi(9XrG
zbBD6pz7GOt%mxnOR9P*jZr($evd?Uaz9bp*m0k+ZM_?vPfoK7YsN;1(wa(_3=hB8g
zoV)T{YX@hWkr(R!B;!9|-`mn;SJJw5R2)+M-Q7as&3f(mgZ$|&>&XumdB_ktw8@#|
zI>?|VKYC^zAzzJ6ZbzD}O?}~q?y5?zx_5)2>O;4eNmC+kV%cX|h_k>|0vy#st^q@V
z((!1?PUE;tiQRyB*oD#f4X!R&iRJ*2`G@r5_ldD5ZV_n;3^s(&as;qf@+&E(o)~BA
zFSyv-fdYZCZA?wy4X1%?&To$yiA7*7&L1ngWzLjQ90DanRtNUs06=}^FLR|RFz-n0
zW&_|eFoi2#e8S8CNN(;NqfQ&5Brb<CMQoN6(2E14Fh)%}I`-E;YGB0;nknouvkZXK
zAKe__$3hOdeO4>>u~-0uo-7%QD$_27iY&aWjwVNvff>Ff1@Kfe-H4P}Hh#4@UN?6q
z`?b#bZzDv)jSVPdnD+`IpC81#SO|+2am_l_yy0`a4~RkDZVM?eZzrglP_tLzD=%VT
z$IXQbv7}j#=GWvWn8I^)Y~`3BRr9fmvfH0uS^|yB${8q^$lfZfd_*lBiy5_iV!S8Z
zkK3*B#xVasE1$w0Pw3jvFMmDWLEN71st#fpWw=Ls8)qpF)Hg@fjv5}?O{SN7rG2)V
z6MLa9Cy47@^>@fm*pY;B8l$aJ-(=YlcCjr-1I1zh<ow)kaH3-3E$I{)<+LcKq^f^m
z%#ZOcF>G*dlAcE1b2{x(jF$)EwLxm!mE`wt%VMxpuCN<wA`9ffyyx%V`uxQgb-Utj
zJX^=jLbQ#J15}SG3?@-vK(A58lFfZEEz6;laCT9qP|R*wUI%kp)_dGdHH6>;<_LU@
zih`JAsr`PpnDcWuy4HvFV$XIF8E1?}ChsZG#}ect4a-_F_U7y90n0N#7PENm=t4uP
zBShF?**i^rY;b>piRe`nbPY;5g~L2O<?5+9!3i+cYXwgCHU<!sz@UY7PzsE~=+6(1
zdc|hP7cy<N^ujJ9#^sqD3?!O!PvWlv=#@r+5hEB7?5^aJI?A;64UTEnii31%MM6G)
z&PppNyuz%oM3PKQ7pk68YuVdH1~YKJ0qR5kC)4&~jF7y#l$Z#43?&Ojz0~>%753r1
zz%i}D^dvNgHr1noKM2OQG_8Xu9gQ<Z+34}L8n+8=BV|7q+UcM1VZokMz$HQm8VX2g
zHX}~z7rEslr@E)w8M{AuE$-!fd{fl7sVyeKfrF4YTo#{*xb$vsWRx2h8jFeiuyNwr
zrLG99c&wG#C93(E=c=+OUCp>Kuz=V}&-b{@Hb>VnM<ymsb`ovf0t(Ln#{yEL`{d*H
zoUabOAD)`aBIm{KZ`O_dO7;*LaLtYQ%mZy@EY9;WbQoV!CarcaWa3?h>TP>Ue>ML_
z`y~YO^QVi6X3b*k{8|iVlCkJCpA7Ak4et<?cdNYsZCuxJA`*G*h0+d1JFkVvvfrcU
zz5_e33n)5oP=H)lB;N2fRg|m!f4u<Nr>ib>+TKw%mVXxTr4o;tP4@=;Ile^wCS^TH
zz{X`+=<tsmn3=WC!a<Xey$@nD$TqNG66=AFE>ZRe&W>9L{9aXwHh6X4#oxE)o-!C(
zQ01N@yAtjbdv!Gu3CK#b<Lz37#kj~(KvJgT6?l3UT)L_zG?V2phnHHOlE&YxhW8q{
zNpC-6Cpv=vrMy{|F?#r>53YEP82GoC>}m5JHyse^>dCq3IO&$W-(e9S*UWD<+qT<%
zEAh+R?9b98GmPmRbJ2y443QF8crz#;mK2!4PrBOgscT%lA)ZqE6S&zkLQ}*~9t>zx
z#ROP&Xy1c3N=K-x#5v~GV0ouBBa6)oO~lHXjgJ^rOi<I9orJ&$RJ_y_73A+zTtE3C
z(iX<Wxz@c%aR=sQar}y`LKBz^Lv>I0e8{hoH|~p2(1Cpb1Pg^LCoCG$LpSH%3>wuA
zda=0{aF!VRDO*(D^GSOuEp*b~+^ic$;(u8!T(S>kjd=@}*42HwS$ju{&irtO2H)u}
zqDvV%>~7<VJ+P4@WXtYYWSU!D(i2r;LiHLQ==~(c0DSZIU=@9d_H-9VO$KK`|D+b&
zV2HD4l>yukZ}P_}5*mRPLSm-M<!Uxl0i_kOdX!HLxeZ<0UtH5NPrji1mfpUX^7uo>
zhAM7<3jk~F#g9L<?l0$j-ygjmF?x<fXA`i{(f%inGZrw^$|$h(o{<^{V`@<l08aw`
z_Jg{5b0zi3{I~VbpCf<jkW9tfU?vNc@jZu>k`af2Z~n{$=7Q>90G{(f7T{9+D%X1-
z9HU#+p{G;r%AgN;G=A&)#Z#z-1r~bb$a|v6KYwoIOCC<W)}#Of98EQyTX=hSRN#1%
zG8@3+@Z{;KtD(DZo9pjYOEaAvqE|--R{GV1z4QPby49FDmmJj~Q<I}!=5n0PyV5eO
zARkCnuW?70F#GTP<&|L--Fesf{Ds<k;?8W;<Lx*};_NZg=AFw;59Y3&I#ZTXYX#T2
z0aXT!0&@xCb3!?QkEb^Ej_7WQz^2H0pxyU0fg1PvPkiVw34J;1H*z*kD20+Jv@VZj
zUTA%nH0<+AnjZ-ia4*TECQ+_qga#iN@K*2vej}`{+g*$&AMU|dE!Vk^`jrcpe?qn-
zsU8Cx5?kiq)Tyqt$0WZKO|PWMQmRxw!AmxF-re?Jhp+WvOPur0i79{>S8HxR+lOMu
z-%N=9<v18@JhMZv^Cq81bpXg<FN{|d^-QtT;Or;r?g1Sx^B#Yoz#w4E$&3hN2AW69
z>!zYzgH9DkA<46`l{C$0tkKY73pgpz4o&Dy$8)Q*rWuY133BM#Z-pQOEjJLi;L2+b
zC$6b{S<72KKi1l`q=bwHuaxlW)=qoPC(}@vabPe#G;_!xjK?IJemT=8bpvxUJ&ox2
zelxk&zT-hUgiITBF<x6Flj)`mQLv$k;iBVwom!!(f7D0`k`}}Sj9AoL8#Br%*;Ou-
z>fBIj$A^LCS&mK0)W;inDH9Ie?QvpOX^LCpdbgh7Ku5A2VuS)QP<r_AIruKMTpAew
zls=O(i3#Fz1U)UQvMU0nH&y=q1XA$>H%*4E1BKPzSdpPKz&7w(OwaG#SFo636V0@x
zH^+JiZ*$Sj^KYFx1Gz6b*O!&r!#@XIyaeCz32KCzZoYS-pMN6Ew!SvYGUNMeZsS$4
zQa$yB$@7nAsWNA!|GVdkB`CS>Au`9-l4ePEi;A7!s+f@WkKQuzVuu!>nJ-pzSLoQH
z%++=7`uD|cgpsn+w1$n>d{&5k8qW-PY<#Asr({FrLRUQtP<hJJty(-Bul$<H717}I
z({ophf#xiT`>MRJ;G9u>bh5wK>!FxEWn0(l&#AZ(tYus7uF)@A%-*tnOAH}H5ly!&
zAP1O0F6$t@z?onF`S>@VMH6`2WoI!#@Y%Ywv=ZKXs)92b@XTIAFi*RylhDA$)?q-h
zebe8ZuP87}B5$Xe?)8tMzc{(3)utkdfztwHpgp;MzTs=#_wrVqF{G+m#@OqPOpL<J
z4K7zc>xwdy50%%KW`t?ZPDYslE_~EGGnpEt{#U!C@W%Vz?7z3x{x6>PI=!lD&(qx@
zJ^5VPn(3>yt_MBZB$V1(ieXS-jHV436RcSg=|6kgDCbaCJrdrzpNjIiJ8e8c3izT4
zMoCeVp?gk3v8r$Z%W)`t7&l!oHV~;RCnHAlLq}oVY>kn(peD5L>!RVDzi{aT;9NLd
zdl}W~f-48jyq2xxg_Cld4%Kbw&L;XTU1rP9%MgI0Kf2h|D{PG}fnrs#O}c?_9n}=Y
zYv2s<Y<rEIvH~qK%+sOOj>Z}`?V+{ElEXN<*rzIs!N~zq=7%E!a)!E%o;X}<s_$Gg
z<L#;ZUI+QCQGQ8$U@Ay&qYcx_uadG^sn|UT2<;Mbu4UUcIHHakRi0J?z?h+f^@m8e
z=fm%e(n;8dW?U09)TIjBUi?<sS-WD1Y)J*opc1F9QO?E6#0l+eB`}%p4iWAMA1d;|
zW03e$?lG$*{Pvan=*#`#hr*Pkz*<!-=_01Z-489POGBQ(ECldg25y5Ct^}WmH+BZH
zXN9w{KyhubQPLomACxR6n)?eJXhEuZCue(!jDSlA01a9aO+?c)Mi9!3xMHL9Eax#Q
z@IqOUKH)P#7($Kh{*<Sr=xx}4UNBhN6#BCFcxFT&v{uRcFx{9tC)s@W-E{lLfBmo$
zai6Q57{iyPqfV}15r96^6rxpBgsi{<^Ap%$T3uQ*+zfaf1MrBq?u!5#C==$JQR?Ho
z=dTaNuOyqEo=d)KT%JE`EuVFfxbGIMaXxoTS9-pv-f7wsekp-{`RsoQBDNJjc`Xa>
ztK$OpOJ)@3d|P+YyZjmRDU-$OKIG<aor?O=x6%o|DI;|;JE||u8eeT<V$t<4_`Ts6
z?G_tj)5j=^D*n>A4%oybYlWcJ@dI}Q%0D4zX)DW7yXRn+uAB-X_z*}w7T~xB^X4X!
zaA0`559Gy=cF9|WcDl+RY`$wdV)0}~UqO`t`7jBMXw~FAtbFmooc?LT`SQ+@)bUdF
z3;U8h(#}_Ut1#vxhxTZ%)^)iuk{?Y`XCczP?k7z8w&|l$X2H^-gm@rg_4VCeERa7&
zRMP6Scf1`%E)y^D68!tiQ1iv**}gW9*o*2H+mxfhtx3n-cm5RnXb-s9G4+Sdp3i^^
zexW#2<es@lN@WBD7j(4S*vylqhCqRls{2u;7Z5`N_(c}$H|kUlK?aK4^IUrp(+h7S
zJ0c8c<d{t;dL;}QL^p?KM-9b#Rge`i9IUuqP7y4XYjk~Ka9>twc9-y?sx2_mvbM+l
z2EN;*q@&j#YN?4*XBsL{TjGq!#^++%m7{;vfWxaK35c#`D_$r6lh&3Q!^wGb3{LB^
z3tIKVy<trZpI+Egl7_hy?Dfv&M2hexco}3^g?8$XTk+@E=Bp0|1&+NqzBsx=<O{Uy
zA##dTPKcMD(v84tgH^|uVoC?`@<sSucW-{jULPHr6F-bN?=K(K9>)sOWd`UWhL0YR
zZ;eimT%?0hf=z~+&MDJVUxGg&pL;xp#kc`@Xq>$j@QrI@S>yQ_hx{o3jXWE$+9yY#
zG5W9TNq+UXVU;$pmI0A%r7#WZTB8sx!n9Rgv!DsBTYtfZrye8B$h@n;JIFyEo)9><
z1tyucr@$InVF9yd)i0cgBqp`sotHbK-9k=4oZu3czz^ajT{%vzmfwfWcx_KsyKP5S
zmA9>-CycGXRog@RPdk1*z=HfA{+9eNj^UGX9IXqY74_UNq^b0=<Vr;;Yha}K3UdRT
zWzJ9Jf|FpaHOi142z!1in*BRIC|8uq8TOEp%mS3rxSIT*kJu2OI)zEJ1^*{U4i}j!
z-tsi-sOkv}l+t%YuI%RZ#0SMy#3LmzCV*>N7>d-Iwtq@cx%B8^DEjz(CuY2JmSMdT
zMtaJA_7q)y{fzxxUxFdPBOEdy#79H+`}&&V&3pZwWM(ZtQkya1emJS|@J|M!XkV6V
zeLvD%LZX*I)M%L}VSn=G@BBQj)k{-NvY(xC5!EDrQO<0HYANuQ>zFHDX9xD)e?D|X
zofW280O?Ss08%P>rilD+d?^RgeV27WybYi@f^32->M8o}b}xCw@2n<x*K}V)U{W7z
z&tdWKw+CpAug7{<KCfVWHWQuW%d@N76_FM@<N1}szUYsHwJDApH&(D3)yTEgKb016
znWeTHFmA>xA#Eh#s1FEG0xPVL+=8VP3oWF}@+(cVJy_T=M|B8)e$=5TOg>503#C^J
z?$a%;Bn)OsYDu9<Abk7jsw4JSJ?g1}=#(0L!IhWR9F5~MWTbxIlg$MwfDh$gLV$wY
z<{eDiarfgUR^gFYY3BA_&L)4N<krij@r}yq%HO~1e$V@NFPAY|>KI^22f#Q#>q}{o
z{~80OJg)oYG&Ety=Xm43@m@A$1rmRRB3Up>px07xsT5@;rA^E^+f2ZQ?0<lzy1Z1h
z-rTqP1Av&Gg?FSN(j-^7?L$1}@J=r*Ipe1=*7E}}K;SXif9)#3Li__AH8)AHUZG46
zyJ!}<pDZN<IFN%o`M<3NO-lHIZ3qzL$;ux!`q9KOO;u*+&0lkuvGC+WJB7J2s%)(k
z8=Hl>RZ&{B7f<wU9k&a^hrcR<Yt1`u&|r)?(oMdn{hSqrI~wt$qob;qO5PBTH*W%K
zjATZCAYC9@3U|g%{-@0lHcz(r2VD1Ugbq07?HUx!>0<~<<N%`dsd4_ViMCpySl!a1
z>a5#K$IQr^bm}m*+tBwmR6x{45oHZv0L||6!o!TB9B-E5bjaMx#J%bzclmw=ry*}M
zkMjZHCe+PRtxAH!MT+x%U7^*@M(x_bqpibu^-miK(K{u8PATyimn?XEwZch1H-Y&a
znLgpm65Ap2Ih1oTi#EA#v&|u;r@KaYMh77@uVgJx7;4Nu9gQ7>gBps2-~hao3kXOH
ziG2#vZH@ZAhaA%kR_74|p%F}YM5CSDtT?(B<RoD?OU?kyW{3_4#NUe=%nkE60P!i4
zJjwcRY5VAVwur3H*wc|QE{Yvo0NXAE9}L3BKxW=8q>qYVawZ-fersw+Ez=Enc8V;y
z3rW7wwGZi?ihD7}JrkRy>}F4oieS%TXSYGY`iqhPi-tHkR4qo%ui<$eFo)pJ5BEl@
z<XaCQk@f>y1H;RklerW$LMml8CC1aEA@^6u0<wyX@yhkzsg+_c3AL8!&upaVP-D~x
zP=f)<V}YfVpH}_MJ7i$-(Vw43JS6@pG_@$&462(#fz-xi9I&rnP#NBUznuH09ayQd
zs|x`T=F@~P-kQhn)SI#xuM_gb4pVif8tcXdr#Gt;KJA`}x$DF}PDU-v;e#()OysEm
z^ewy?<#mwY$Qdk4>rL4$_MpbUWXRlc6?G*cU^f`AJy}9Zxe8hDZrW1AKe|zA;i~X_
z$}+HawttiR_4;TQ(8Tn)^Z{$wy}%)0IrFHZ%pR=1Vf{9E8=R0F8_^)tn)3+oxEhRV
zCU6401~rW7ap&h^xkv;GnOEKdbfRbBs{&#`D4oBANio4;<ZQq2tOBqj=gt9mQQTBF
zC$Y)j60hOCI<$s&z3<w0a@vTSVw?Q-tO`vLSx=(FPy3|VQ}zGtoW&9WYv2r3asc%K
z&Rj8|#lHa@G4S1ge=M*ZOmOs<d!|E&`l76%D0=(8D36jIKNbuk;^JsqTOF&*@To&>
zF=1IsH-VD_p!}V{4LLHcsYUenru{h(=oyAX(|@@Do~87aef^qe>wTC<MwDldZe{{*
z?>J0AB=_r!FAV7q4Q~UU5%&C@3yP$5pfP2<<w(MR`-y382Nbh63aq)5VgTOfzd6l8
zuEne^>%9RPDIH3<2myKvL5$h~TC7Ds*9d>k{73||*ppgRR3Z#O#T$NPxNL*XoWiT7
zurLdxAQ9Xt>Lx<-u>{@W+th7+bMMKA`xJ)&P``=ttasfNiK9_w3>UjUv`bi_VkP?<
z*~juS(*T_*&epbG;0Y04vz=I$HJ#&yLMEC<cDjWWD(G>E6zPNnQui0-fK5R(yOJTf
zM7zaqAre5mn^_>sqD{rY>XBU%h6;JQu9f%!e3Ea})r%Pm!67gm%m%FC%rk+PiBpYR
zTK&fnuSab~yPMMGB0Sn*b9#VVNtuU%A=Od@@d}gikDOZJ>P#Il0GrW6<2YC6<cZyC
z=lu5EfWZZT30;bHDU-_@0LW<p$Y!A>1}smsmI$(#z7eMwWL|eH*}YG@a2xffIK&zP
z-~g`_HO0l;w~B)XojI@DQ`Q^;kv_|u{ry^LB`Nsocr0CF97LN(-4U-vr%Rh^p1dt3
z2o!*3)4&dDrce1kJ8=)GPt(>|&TQsZmZMd7X>Q6`J*E^B7?t<!8A{JD8yV|7G7m`s
z{D2Xfw0I1lBY>H3um4Y1-u#!65^)$n-*nt)C;Zu9I#xCl$<ALvK~)}1Z`~iPekf@E
zLV*C=X3*?2GQhEemSw(A8MWzartzU<q<Y0ZzZTm{8pH&6RR-n)vCcq1-#ZTEPhI*3
z<bT=S{34EuD3VjoHbL-)bt=VOLGLTC`v(KM++Qi~0*0pUpZfd50mS=yx8)ws-l6@!
z6gmzBIp=Zv&8Sm98E4!5VBGzngP+RiGy-VfH9L&W41AqdJiok+w;S%yH#M-Xs|F+`
zhU%lu^J>fDIfUpi8yieyMJ-#N15*9(%XA;7bQ%hiuRg{wp*IuU=@XQe>ud0zPsV_b
zBqev;JJvVKg9CGm>-sRg$NHZ7Ki!{SsIE)>zGguL#x~XkC0KIAgB#p>0H>#(HtV=u
z?^5CAs7aF90cNYPZTsp2?>Zx{L0lG6N{&Y4o#{^PqZDGOP`;Zoc27Rv2tRTPKy?-&
z$YNdZq0|srWXNRHB6LoY|CYabUuBw005FIkOemXEtFCdA>Kmm!dI_yE%Wau(ru-bc
z9SS7sA5B8Vo~W7*ZQAQ9S^zS#kU`-KKq~(T2xN8xxrQU5Cej}_q-6KCEFF~nd9MBc
zw0&VjWg*oNsWufAJa?0&YinAGAX4zfEY~I0q++94%&4iPu9=DzT0JBF0ATq27R22#
zo!as_)G1=#E;1G-P;$2Cn}iK+qu)hQ0kww{O0?^+tA>laL;)qcInmj!ihM9V9?GmV
z+4Y|-!@P-EZZ-~x2*iv$brzer{5Z1rlyQa3ff^7#{P4trN#TQ!1k1{=?0__Hpgv%1
zyjDgFy=pLY4AU-~6nw=D#n{of$TtL%$o|nZ66FXnqU)_UTlkz~)i9ZYA1?cQSCsZ^
zNvojTgAZ4efH_yLW1Zccb4xll-<!2SJp@@j-q|4HK`83ZsCRq%=gzN*NTvg@OQ5E!
z`U~>tS7hH^#t=xoKf@c@K!fbVmzunT{L@oTP1O4#9Q-}(B6L)A`MpB(YidA3oxoXC
zj`e4kz{Ii|5_*&|%&>x^%P%rRiF+-;U1t@Bx-^X?3~N6_L7z$Lv-xLIex^t8+2F`C
zs~4z+tA*B3@@Q0I46O_#G^ZPPUo{w7yWLwr$iM@T8Ebw{Lhi48B?ViMIX|-tc2$P6
zhRvHsD=!LOwEmYmWJNK4J#(tRdP-7Kg6!z;a;KTCadr;S^Sz(o+;O<BDU<}F@IpQS
z-8^}!z1mUt{BY5RusRAT6rD4GCjYE4*FF6>iZ6TLjV*>8qZp3PKI-Q_K<~sYCKD|C
zjrJ9FGwq6^2omfV<I^jA%fa8}h_L<;DHwX%f`BoOKm&I+e+r~zZ|q9h@95FYC|MK8
zYOdiII(;ZBU@o5$zB&}~__N>Aad&kW$ex==1-8@&xgA79id68E^YQIzBX(_juJ8os
zKd2TBz|Xke2PeSBHHD4uoCh_Td$lVFUF+7f@Uc$4g(sAoHjTq`#s}=*+}P6+k6b~4
z1z3Itg1^0bV{VZ^a<EQfQ0`?ai~_OX`{bY6OXe9%qC#MDajh=nJOl>wYvL%7uNX|f
z!;wXegK;j1Y?Zuu!KZ=On6)iUueL?Pvs^#sI?NcwtsPBD&UvHk8Gz@-6Pc}VKUR+u
zsup~lRuW=Qi3hhY8*H~+zS<S9UDL44WF^nc<KZ~YWN~`!K8xm;YKF(BAs-cKXlYY`
z#>|YJwOpS?lGPdUY-K3ji3Mz9^C}~QB|<FK^bNj!NyX}0fN}5vH@&3kPiYn-!5szW
zaJAj|2d!tpBV+=}vMh&q;bU4RxFH`;pi%;e!w$Xtv@#UdjS>&i7EqM<yVtJC@UJB_
zP_P(1*=t`MZN{CYNVspJ<J^LxCURU*eL--%Y2QzJL6vW0Al6q9r0JPslqf<Y%s12#
zMJSQa%aaOJR5C?#{=P1V--{g3nbaq6ZN!u7g_|@a*EA-m;S#sIwkM7X4I~B^N4_M@
z{)4Lb`NqEUhGFP6JX8uKW?{n;wj6xU^nszIPVioDmb1Em<g{*xp$9ci&uCvvJ6bD>
z1HApyNN{j}EJ%Sl$^Ct3n`}pC-0&P1^o!cVVel_9k{029k?}|0%-Y6u89E(Dk2Jf>
z!FuLnl%53B8Y29gNY|pKB9fY4!|{Z#AQFs!F;>zWjHN*vV^@6GE6W~P)=F&%juHk{
z<MKXJGEDI>EE+gB3oduok-3;wobWt^9CSSr-4t`qH@7dkY^KUc-{`yt?tQ!0d*V=x
z<_>K$H7bbN&MipmImj$7QP`q>$b64W9MM2!&1BC&cg`HcWL5l?7kWMs=+Yxe`Yfc=
zH*{Lms`De#b2J<>u65+CU{Y^%F*x^ZO=ZIA<^stT&EW>9D_145eyp<IP}BYmx~6R0
zkn7;8?}&_`SqVipo$IV0dED#15iedIhcu2m6lj9kNrByrDsB>T-{l4`C&`NV+(trZ
zS~{fDdgwbm?X$-QcZ?9?_dd}~%Rn`=4~}G0w4?TCf@P1Qp02=W%d>ie<_nVSGp$gJ
zSHlu7nd5#*an6LaZ#jz=kia8nZag9E{K}WIy*r}h*&k18?B>Q}w!8XhNN144uq28b
zbnLiijm!1j4JFjg3n^+Ggk-RyX7RmGqdCu$-2=OKh_;m$gOkP(p3M|=)g10_c#Up5
zi^6wbx$6hbTFJQd&gTtW=7LK0zvnZNAi0&N!R2K*N$lCsapY7+mIbIbDifGqh1ikC
z{)BA=&%L-F2G{F<QGjq8Xey`N^N&lWSi-)eD8CPg4t*F+D{xd1pY%q~hQq-XxT$fe
z^TH(bp~emAu0>h3KAI=z#b4CdP^%Q_S+9e0+f_|7mjXRf`d+GGwk0+|5hTs&{W?*7
z%@989^s&9|3fTUjtE5`+2QxXWcGsDS;#Z@5P@$l$$-Du%`s0w03*ZT6it?B#^8D?l
ztuOdWx#%KavSAoR`V|&~TBJjtM%x}dEZf&7yCF>M8An$j<KTnldz3{fB5@V9>^t++
zFZca@gQmYGi&b;7YgN|iglSwgROK5{J(vKbXv<N1kfgWEtA%Op*{s7M9A-E!E-S0L
zX_1k~0=kPVj#TBHVab^eJ#$NKg1f#-;xo<B+ZCH!DmGmB<yav6b(mg1TvkSvXKi%u
zg3E~|!|#Qck9+mNPDw2{?7>VNME3eNB-^|p9eVzu@8E=!8MSo3ltwo){m48zRXeIh
z0VCPChOlbht^JyJ25>t|62AyBbqT}lh|Y1apQc-LC;mwZkQ9LfZ2UhbIp@plI=DXZ
zDCetZ*AE`_j0#&PBi_EmV5KmI5EQcyb`J*Z3U*E2U?>h#F3YZEieaJl%~BtHqXLfi
zf}=MHIgXtLQi`(dXsh#QTdami<B}LCuan*BEf1-_!r_a;S~~JdhaTdXS?x`KLn@~A
z2OD_3kBw7@^1|6bVvK2a1!RBOd=`s$U<uvL^iHZo^eDN(RqJIKc#Rsq>6XQ*8gcxI
z#6cEa)*pquqQ+j&d=zS|gJl{PMu@2a8c}sTab(rw=y!5&$U$Y^C9`JlKuMb8!2`CF
zsYC%o8!G*SiZnh4?TwC;sc*B*=X4UP9pxD(i9Vx*nDCvZ7nNNGzlU;ZU{?ozGNl87
z=AkeHJ|ZYK@pm~VS(*E4q@eVMk1NinK6xLKy>^u>gTIImb7wt@=hzIZ4Y8au^(22(
z*|DbwYLmvy$F`WH&!ZtjuFf*B8v+JZd<SQwm0{c$Z_b70A&RcD{O1PDY25oxg>{7Q
z1ZNCHv~*4oxJkwX;NG1YRZ)KP#dVN0oc2**{e(Ad94!-bYIzFmn&F}X`x*#KL&3@6
zd|=~*IC_vE|M1SD)ecHm5$fLX{3#yZ5`B*N%{!x5s}AWnx6Pm=@}|wWX%vGt*14zh
zv2z({?vsIxqic84i$hTYmO_D22j8YKL(PGHOS5}LW&1<Zzfm`fe0ZZ_DI~jCp5xNh
zatSl#BkYGxv|CtS{Qv=YZie1@qqabb2YNiD{8N)zr`mn6&ZK?QMICVE^`Sa!ni${A
zDS%h+JBcLRw#$BinoV@dXivSQF0>^G4dl`nc22=pT1m3}%}Z1+NlRvD^QywlOhm^|
z1qpE%)Z_rz{m$up%Hu3%5zf-&ZrLQZZ+n`iyuJBi<UJD>OZmU26sKu|9HVuhy0x|S
z^hk-b?OZH&tNX`S@|{Hx_Db|ga9lOHLQQf0N4@;p)qLSTI*U@;-26_RD2WP+Gc{2&
zFL7zg9=V<k%sd_Lc(T8f$~-i>vFR0PR+{S${lZzw8kn(K<d*UigE+<9z}*?prB(6e
zXzAE))uHgX86-Nu(4Z+>RzFg2AZpF%SstHInO0oU@lixWImthz1*zpPciRa5wds5Q
zc>WZ_3|>vl;!uhD!BEc_l7K=xnBLYjBvWt$+-0?-R>uGXmcCFyGfP%uSL#nE5o+fa
zYwUtn#i_hC{o~?e{4u>%9)nhOyCGblOrw48p&t}%nMo#dxVI(pG%h>5eCn2d3t4yG
z_!EtZ&U1o-DJr&u(bAaGAT^fxckWO`&@u#P(VFKk$yWW~`$fp#*aT!Id_;xJd}1(!
z&gLTYD~s>N>ncBC%OGCnbCj<?H@GEWk(0p#_UvX2$xs{vyVL3<WRJ1XY;0gPA+WA&
zf;v*qo03RnMu*cmj*|ZIsMP9sdl$8p?{P=b=sP|8{lUDG#!sBl-qUrNIK6x_kPz+a
z-k%7!s|AMJ&>+>TvGD}S^Q^=;I9Ffl52N9Y3ha4<{T=0VQG-`(^r8w;_LQFE)s}fS
zuY|MRXUbxF)l_->ha#K)C^fpwK(YkycU(KoF4FX@S;HgyfbT?YXqNrQ+;P(`k1F2w
zbLFz7hv*E4r<=Tz1hRE+24oinp|wUCbO$z>S^g2Tz0pU5t_OO-3Ea0((%lS;tyyQ*
z6pjH<i@Xv$BI&tYCLFgp5wjIuz15#RMS`hQy?K&sj?npJ=(y~EX%{3VaQE)Me>E{X
zXUt%@*9{%{=znj1<7WoO$VY$z%U?!nI>DLzBRY-uGtivl8Gk>Qq~y5KeV<X8|3M#T
z;K>5{?z3kpx7g{EdE*Si?wP5})m;n)Qd47mqrLcc+It&H($2~)4fPMt%I3?iy(;8{
z<A0e+vrO`BQaS9C+r1!`Io}q#T0Qs$JhE4ct|9n9xJ%4wrM=GjgKkxnxV<S4?|$V#
zQoxYTWRG4HQA)9GXLfTbP6q2uh?TI9fBt$k{7-8Ft%3b#58@BT+*%W(INu9Cj!>kS
zX~X@R66p&ptf~hubK@=Ze4K^pPx*@lIIcKWloM@+Q}jCfWJ5i~Epu3bL>EWf`8hDS
zUg-)i=`E#oK}DX2#)6d+zle$Y;%FWAZc5d!N%jWRfkazb#!35PR@;QTAUlVk$;zQ<
zIV{%O$c^1C67=<w`4<gPJTK^CvF-`5p&AqQZKi0dVl?RcCByQ4br@-0^HOKOIVN1M
ztLNomw!xuI$jAFMYM2xz4{mXTs?Pd{Lkpk>K}k9J9Z~rz^lW<NgQ}Lyz5oNV7%};X
z)$?p+_m@}xiab|<KRO9Jij0)hNL+AflZfBPsPjc6-CEmVNwNn<idCi#9h@e2zAacn
zyo&Ao8d$NuwoG`{=&6jJc-6^2_(ruYd!x+`dm0@(^?2dOgD|c-Gk?a!bCiAE*oC$D
zeJXwf(GP~g?WyFxaKnZCKlwoU0B)fwg&;5Nc5^s_%|LE#@!MB&fn@)hlJQTL@_8{!
z-x6WVO`|QIwoRt3U!<#+hlQ7ifNT7mhxBbrN4Bx^ifsKL;}}l)+u8yWSM}FA{7$35
zFV(<9(o7ELu5<hR<8?NnjsrK(=8sx(7hAX~cpY?A>sf8b5@g9vA0O~^rIBFw8HT)2
z`Sivc7R*A<u2OY=J3f7PUV4HEYS@7fH3aoc1YPR~8&72Yx1f?eBTeh{>m$}KtNUKj
zXOa5BQ?Jsi2Uk#Rk{(wG6R2m1l!RcWOVqw+dc`ko;+UaR5mqiLjvUDo`pnil{e%b#
zW}$+h59^q)Wtc7tfDx)JQrWP(M-ft+Cq2YOsMZH7fa0la2RJ-Yhs8&8wZgi5Fzp<g
z;OC}>&?j`yt(Lg}YHTv5qw)uV_N3l)6E8+2P!sD5%n&9uLS_`%Tm^Pd5`4SR>`WWW
z5+uJr+>4zuf+p37-U8tNV}sNzM|_Dfj2k39=NWbtv}%MDg+_X2)H<*lv$`v;_3MQ(
z<N(uB1sY2^hHoB`*ih^w2D3@=kMtv3o-BHonH66Cob*3`j?=ac`@6K{c_;H1YnF$L
z2gtps2Alj_t880Wl-;a_Eo-Q_-d|w8g_OJoYCLvUwALR(SxlA*JOvtF*^TD+Tm9WV
zQQ~+Fyd*84f=<Tx_^rc5&Z|uAgRyFcTtaYWU#}@AqJ~?0?+wZWlZO=6&)mFJVS47V
z(UL3Jl<vDs9?J)!It?poo;npUe$(2dZILxghE^W=>uhh_v8|oq4o$SPg^w&Rl_Cm~
zt^Jw7ZnBmf{bY|GuINF_-n5K`f!)HwG*>sG*8{t#{>9$ZdZ^0Cs7B|a?o9Py=sEEn
zMa7?>{B@1v@%A-O_DQiR_4{?B2ma`Q&!sz$2?;x&F|yp~=DWQh<D0Vyr0T!u3;0+x
z58?boY{(#QcldysdUa#MdV|R8i|nPrv%B#(4$^i}q+iS~R+8Da8BeX~n1Q74Y;%Nl
zf)q1{$A^07)gY1-|DirvdQx0r)fEL2KVZ<o<0EV&V9LCL3!p_vLB9A8$!&j3FR8T^
zk}1Ru7X_hbV9H?qaZ#Z|NgfP{MZ*6yZqK+=-6w!b>Aa`&7DnmMtO5u9$p1#@%bGqp
zJlagzs2KFUnyRdgTe8N7GON))CiG|x<tmWe=|&wt(*HNl7O!E~R&AiejOEjR%XOzl
z7B*yET*JU~!NPyvjMfY=x<dG|lJd=pB&cF^IZs*%dzxrv;}gYIi{N*^SD>-AG7y0p
z*R<-1!3h0T>8i|s6p?YGMe|>~*7tRq3L>P~a|!HN0(FP2^MFaZWJJY#e3KlD_$xd8
z)%caSJIIs%{U^F(1Gz{Xlu@CrtMI2U9lDR>2AM}jXC8c!k|ht7%iBg{RaIo)l65Tg
z$;0(P`X%%|%rEHa&Y991#$KFon3Y>N^@Oe^KUd7~i%tm$aWUag2=0O{L$Oe84ofN(
zuqC>?pGSQVNS5)$Q!o%#f4kW`02bajF#!8)OS$xUClt)Tced)stdKL}(!$bmpa8D%
zO99gzJ#v79tEzTN9F-5eIUmqPm6M*A<T41Nt)0yasGBsCaKDFg@SD=;LFTpNhGHyO
zPs!e3)VXM3woqb52+SuSo7rJx0?(|rJjt_UK}2(Xo_}9Q-K8k8x+9&Cb!cod(6=dG
zD1GBgW%&}$7q_{ee`5)Z4@v}fJuS&DGq0-NWzX}gxLHvXX3At<6wpmt&K%i^&Ww2R
z=moqmWX?S3)0zBEfmc{EFl;PA>HKP!`Or%rL9+I8YxZ<`^@Y#eDeYA$r)n=F0_IP<
z_99146~$Xf(%<eNk{It}B}xx!pR1r5&`TZUd|1j=rlctqJ$lLw9+#aS-|U#Ju&CQ5
zU>$0g@0P_DXGtpz1MrVsY*XCuCE@Jggh8u@?BfIRVFO`9c||b)*pYzqB?bf|i6~(5
zg4M?HIA4b#nAhw$MCXLM!@sQ4wakWjOm;=X@sS)vMJ&Tf%gHmPQam8swycIHo+3gX
z!5xzAcd4JLeRP79eDG^xh{MHkZ#!cM`i1i@@T^@5L%K`DzL><@y?WIC*|aQ>pU2Bv
zr^jI!SDiO*N{>(zwNT>!?M<zvKip;cQ)dm$MjWCpwYPRpa&u-V2h0QlA6qIEC8MM;
zrMDbLDg*xyl#tm8*5hCEnESv|S7#V-F+;YAQq^+QPFXN}=|v=?#z8_(O<e#j&b+9`
zb4Iq<Rs3zBfk&i~jBjR=n>anjhpNgDvhK_x`z#elL3u9trI=<izNUX%a3$DPNsyOl
zuISlR|5PNunMK@Wu%0f?NWF$9Ckt@4u-fz2F+-fY=X2$iXS`-odFvB7MJh8mP7{F)
z?~K0yG^B=^g)8OAVWA2aPznNg3avTX^%CR5%|;e7En6pVul)tR1!w^$_7^_-Cwq7E
z1u26sN9s+netst;#}go|S;qGcuBnXA^L`ab#bN=PrLr#oii>;{4bU6Sj}^<q#2SBX
zX7&kLHB+e_Gs6Ymwo5505`o`DLC$&JfuNe-FC6&B1Rz-bQ8b%&UrH3Ldxib}a--{!
ze}z1jjD|(P2HNF+K3XErHve)TH)|rRHq%I-G?juejafAiFy<R82@5qJYd%agYMxC1
zOe4<CscJ&G@%Q1$$lC(EA|R!e%qRt!Sk4qe87|$^g*B(jERH*}%+y2cNw1-8<{ap{
zj5lzb=fap4=|bs_m-E&FDl5wD>xk6w#PoiEj`jZbVm>p1%-JqAGM&u%%m$a*pM@9{
zPl?Y%7S<rTyXNOSF)%_0dMLRWGh!TGKv7oX1fTd%+oNMb05vUmiD)3<d5mj&viRU~
zg<-L@p-OSIUB7uGO@9)yv~0uTn>=nWtTxiXx1yZ0H8X~DmK7bU0=QkGWIsd7WdQ=;
z*6+=Ra|irc3hZ8|H}uLgSg7PLA8E79kl~^)WgPTmOa`ku@(r&<%Q~?70>B)ro?^>a
z=$>8LSi=C9!7*$2;h%bqQBD5*Ur)<<`Nn0t<?tIX6eG-rUov6XG?F<-c`?$Ut{qn}
z9qOXU$=T#KP1~~bo^>5mvGjGF**jAlXt#5hgJyraiLk_<N=i1xUD1t)+)x0kl6j>>
zSvx}x^0qIz-kPmxB;YkT1_2%?9vW8Le9+?s157euIpZa$s>YE8)1rgQrK{}fQQgn-
zrJ~oNk=4Ni<~RFRg#U+ki0Q(=p{%l3H2n^#1QlqJ^lBRgT)=6jJ|?O~5!FwaRex2-
zBAS^9eIo~%QsIvf$cIbo%yUz?%$TRY+pc$706O`GjW0RWhgHk9;EnPMy=tzdpH-c5
z_3`hB5odElICmK!^d->OVF}cyB4cQA$S;{b+3*zgsrMBX#Qv@!uHTx>h4Lodi22(z
z&0Hjk+6-D2AvQyD2`{t*6Q6!uQ{tq*`ZT1YCTrg9r6@q+y2-)85l`8bmRiGa&GY*m
zmnJKr4@^8Jq!=QVr+VOa^;URUWuc7<cz2cBqJ3d6t%Pf-cMCIAc91)NGZm6S`jwF6
zR5$$Ssx=w=)TAdBams&jLA?p<-&=AQYAcY0cVVa7SyBBlS59hu|NBi--=3T-eW+Ak
zj77=sR^0EMRz*j`Uu}D21ixpIH{fQGQT!UhaXZM|7D5<{npFQt=eW~E28#9%Dprx=
z0zGG9V>|i~<09H`(-2%{BOWch-5oTC<#G0N1GzN9s+MduJSH=_&`L3!im9@%whcI_
zBfnHQmsw|0WDlaZh4hQ*;>1F`2-y@x>oa=`K6#d;v)&Micli&XGL6;q_D4pJCO`Z1
zGHxk-M~cy7bqP|lN0fX{O85_#PK>C|Q(&TR41@i+){ZhNu+^g!x!PNJFaQ<<&nMs1
zh4!pu!c`h<C6cH9A5UK$RaN&z3kXOGNQZQHcSuWjcQ;6PN+{h(OI@V9Te|zw-Q8X9
z@V)nYj&bx4j^lF9-D~Z+=A3)2U!M4tcE0}zF2u9TL-^U(deP%kt;N{D$2JrIKSS@z
z>b3X)(h!LWqQXI-m>n?T0n_*6wyTn-QeNu*F=Mzy?Nl)5TNYZ|IFvK9R_YO{3Ht|p
zCn5Io`2ca9yHE3}fej(+N=>MJ>eiHfGh;!0h)jm1<p_Rdq_XLA>`t<_lJdDgAA%sj
zT=+&oD`+y}Gm(Z|wTzZgj2Tfoo$+X257ZT8Sz__hq{3B#W6eWD__wNmk?L4)Sb35=
zd_mCg**bC@^XV685gGqAoj{SzKm`w$sokmC`TZa-AuuN8thrC&qB8dacm`xOeCdOJ
zoFwk#D$vdHmnqJl<9z&IEHIKYQU|&-+#WADw}&wsezDFmF5RD$cQ#LlPuK}?`a>N>
z?NyYMW%<cf*)38+S3rxmsF6+my3vC=A8n7o(tM<jRa(2{%<q3*fOgXB3>??St)Og1
zA|`;8D@a0E>plU}OrNG>U88sKS9nSpNFeWQ4Dc0o0*gv>axbMGta)fY;!cLdK=?c6
zm$RHbbQ7kf-9TCAA59EV#uPb-Yf;(AD{(oe?h5m@t;aF!Ce|08f6cc~4^<XbZ^d)f
znoks)%M;S9weVwU7FNZ`-sv=VUrteN_C9LfQj}%=N!f8xmSbYO&-zm)Y;ut=^ZBlT
z{DR&w3)L%L-x>()%p}JwEO7kt%biQdRfDPb+4i^aw1yMyYKzuSY?sSPGgoyu(fqfZ
z%JF_BCsnoR?gF%d?jpTbbhF1LRP-)augtpX<xCXLWr-3Q=(OTYb2jQZL^AtLf3n!Q
z<mqoHk`$G0F@cg_4!OPdr}8(RfhsswlW)#)d_c1%{`x44dBE+I5eg&1Hw&<*&W#C^
z;m%MB#thq>1=~{EzbLWC*e*Q+2@I>!H{iRzZ=-p|hXJs<P6b3O80UVSh6tJT6%9=O
zgvSD7z^r_pHI-xgpGafeiovXW37?GF2UmW&0(cFD=vu;_^7P32s@-`u_QG0hOG3n6
z_H^U5pmv(Q#yAD+fYN)k=p(ktt~YLABKuaPJ*<@Z{;jAUVtYOT9C^@yhKKZLYPm@b
z7q(F+g=rqpp@$03_9guAj|c%m{m6y(q5vwxJEx#tR`p9<#>d1DF~HI`fCrU8>d1?7
z3oH-LtHb~b-hz&TW;7e6G|6yVq<Z}Hpbv0E(pMJ!OW<dk$GbBocgsv+gmFy!AY8Jz
zrQeAnl|`HA5Ao`xvF-V2_<+C@_#J<1DNIy;&m<HhC{=p^NIismzGv>j`|zQNt5TwB
z&L*CwEanPiecHofgM~@kzy_1&U=>(tiO5z5VMqT_hbEO%HJn9(F7ss!1W9!XoSpbS
ztEK_@eO%zK(q^vwia+=N={rTeFihw6tE~?{C`O&Ua3qZ$++sSzx_l-I0wk9i4ccis
zk#)y-43!0Vl?NDh?#giWD3XkwG~u4MJRm}gZGpf-!|LOjg3sT3P0<vZpbvAIta=!s
zO|PW%JN9KFt7?^3D#&%cL&!p|%IO~CBREqg!PjNPY5cVlB&R#ui{-U?I!pdvj37N{
z!ql}BM;|)U{YPcC97@~GH7T)9O*hlsS*woxcdHB@J1i+f)r-fqzb)v}Z54*`9@~NB
z15hv`UTAmNxl~0_46In~_HmKbS@o}D8;eWj#Ixr$ceUa5*76AVWYxwD@C#;f%ieA&
zH;LZ-`(MIH{I5AF3EPl8XC318?g~paw713dWHPLxk8&k-mQeb5E{?*qaj#Y)t@o9#
zVQR`UzQTu+iZdt!%M!~Aw}?}0b+96?3X4Wh+w(6pSzNy|*9S;G3DV|N60Xy%xX*Y3
zqgfIoVl?a37f>eR*;TpzL0^4YmF+xDY0MP_KseO(QDSsmp^MsBykkq6aKpBMm_<vQ
z^Y7?QelCH)&Pb+L4Y@^{S9DOLq-VthJ`rG{*}c*pc@Qvss#)K;b0z?AXvkwUzA2^<
z$j~@sfFDFbzY3RK<_6illT8S1uID)pi=|UGGpzs0!)XZsBbzHbvRIdf2IPj?L045#
zzzS!Ql$Oe{y0h&MQ3absg<VbGC6aTDZ7p9jSf^bm1!};(C&2{k#CTaXGE#PFoI?Ro
zs!0NN_$Y2b7^^J6s&=le+=waxY9tk4yq!x@$Ly=b!nd`toJ?!r#X}9(=ltwUK$2wK
z`m?Y$kC<qv!ju2mjR$a8ek8mW<q1P16pbUjgW{+73yVhPH<oPu{StL40RXlk)7NTQ
zHU%Sf=+MAk@`r)`#tE;=9E7zF-=(|c$*CdQtGLaghdjUZjeACZA#FRxru0J=4tO!?
z`@|MQ=>Jq7i)<>T?CXF+(!h9KUTwoY%Rm_k0Hu|S`>IeF0tBd@35>Zleu?cx(z`z4
z3>tn$NkO{EJt2D@q7ZGh)QK|q%~;cfcfRlR&)tpd{tSibYK^GKc|3>E(Oh96>y%Cy
zq^Dy7(&LtL%is{nF{qlC(!)1qstNe>*RXaSYALZ&vovBHOY+sNO9M=ZJmBU0#0F5x
zea)w#So4>ia=LzRlE7Y&k)4S$1#FrfahLPPF>b|Z|LQ^+-ITSh<XNsQ$$2tIrbg*A
z{f<m?k%1y$dH7&^Oyti#-J;c)6O#~?UE|T@!)0~`^(kRQK$W+k%tIKF;K7mR5!7yU
zhkL7EfI+cF*cykioWLNwz+O~3zG6IPpc+KFbYP1g&&*Z=lOixlVje~E!@M%$g6gMl
zP8x&cA%8wgBeataGZa<vJFA5Fx77_zQ%VDZXCcLD&{c`xvJ?$$_(g7qbyr0&5c_Pp
z*FSxvI25T5=y7hjJ_9TVkD@J6_&ksLpLH3pRFW~h%JD$}OmjE9AXL6fk~Rog+t{2b
z3xc1~sHNC&c*?F)wwasU@Pz|0YHQ66$npqVAk^R^Q=Svt2KkjrUEMf_t3a8%I2j?a
zb;+tasLAT_==eYC#)O2}PMfC3kVmsXncLS>83WEFISAOfZuBB>ebRz}EJ(+d)MUwR
zyrkHZTs;1*i|i5uYVRf8T8?j}@$4MPaP0|f>WaZx_*+%!W0IUkM^)xUO#C0S$Y;%G
zGh6@(UA*omIoGj6fs3+e@ycEE^0bR`H&|PY;S!h5o?i1r)+n%7UqS#ZoA`n`z9nz{
z%jaO10g_ByoXyP*f1Pt7j4VbXL*)z^dKK~mqYGolvr`c9pGodE<bKG)T@#`w9AVV&
z({fyyXF2>kji=2=a6th-KB8ou<=FnLVOiZ1yla9M?#vMl8=JSWPej~U?O3Ptm?Ilz
zy?mY_f<XMhR8bR6r5CZe8?LZiv3!$6?n6*EFVgruhD_MDuQ|N?Y0vvzsf=-Mwvx9d
z9%tZ5u4ekh{~I?P4vf}%>8E`K@B_*wgHMBCGXqv7w=M@51y&5t5qDiW(|8*~z0-JH
z8J_cbZKZ2gQM%F0?=}qOYm;g}2rn#tvxMEFtfXh&wyfSQym?=T5pAE3`G;jvXWrH=
z!iDkSqR_id`WhIPF$8|5rno4(+LB`vC1ZhujAJ?%kYZ|JtOO-6&vbD1i7;GFEHdPd
z?*v`Id2J{)ue}Tmq8vty?<}oK-dVj7s^IZ$#M+y~@k?1D<A+J_D}BnC%RT+r2>}$l
zOk7Bkry1(GG!NDknje=LnDIObrf3|6TJ6h*j`cxSy=f6)|1gEgRf953V0>t_j?Ko}
zh4D2uj--YH#b|`-CdqkoK2wL1)@>Bin=lLu+_f)Bg8dnzF@0~r=s@|LUILjp^5=<E
zjN@9pt+4L29eZ=zJ4*7^TM@OE^BsINt!t&r@OwAQ@=f#WHia74`#5*+y;6I@YJQ9A
zS~sZ!U_wd|)r!Qa;z0d|<qFT4?7EL3Tj%H?U@XWJj{zzH^LR2;By931&FS7hiDEV~
zi+8a45Q!h-Dj)!|&K>FqZ{_1+<2YOtxk**Rl&mwz2C^jx3W~~w%PXxd3d61wpDwG~
zi-#*&;yCMQmikXzqnw8G@{g*|v;e2RSQT0wQ$I%bHeYDlmBMe)4DgWNcr<0??q5j0
zZc9;?powUP0z&|p#=XpLs>qv2#u##_G<L^FG+FK7Viu5)62){*E<ZgI<&YXoa#uUv
z&$wRUG6Ujm)*c<^)`u~Gpsl9Y8T7^ZemYQnY%t^N%h~5G=-<E;?EOYVcCKynu*=mk
zC@J)))P9*%F=1^-PhdJ>>~b0_l2ILh=E5`a?ct(}GlsDj*we~64H??0;zTCG*#Iw+
zPsI<3?-z<5K?CW)XmzT*c&+9a>2xy8HW9dF?=(RA0cy|#>8~wkHN0!1fWlx!o;HC_
zp4tT=LR&h?sCc4T3Q5w)`1&b4N<cMO3U}d(<6ou&H+squw7Cvrz?cTJ`gw+-70eU@
z>j%f(;#6n5fhRipUsv*SfZChsSc#j58j0mw97kSLF8$X{UBbsHzNLM?{0bJ;Q{?(w
z<3RA`QO$r0%(^y+L+2HVA+LMwrOd2<Y1Naqpx&q2D+w=bPv57>PXpM7$hjhupV5O)
zI};)?d_=}qbevC+462nc3Z5FqWM}K{VEa9U0n*-V3vE~++dMwO+JZs(tVu59LB3Cy
zt~Vvf=`Rk$y)Hk2@=E8;08&pA<@vT0-KxNWt-6!${l>ObaWI(ykU^Q|66em31VotF
zptG**37Frp0a1A1rdbMdf@QbfW-tASZ%~!bMdIEHA%8^Jiya)vISlmFXD=v<42tgP
zNvLVex6&sYuFrC8;KUCh7|iNOR((7-V$<@-FbWg<!h9UtU}C4h*m7xg>HUJVP-TxH
z)Ca-#VC!`y{xKO&uGsA{kVR2~V`C<(3bKZ)EcHgI=Vz1q2cH=327qP+aLdu-fkmGh
z9#NBg+@38NRtry^N$E+!cd$Sz2ae4Fd1w%m{JCp(`J*nMidpTvnE_ev<}Ojinii}6
zrv=qpT$O{=2S{5~la*`}rXQVl*I$HuWsF!FEz(hQ-y+d1T=M{n2NfSkhpPQWg<Plu
zMlV}jLhU&sTkfK%LfEeP*$2e|xIs){>|rKmuW@IRw`Md^`me`WGHrz)m|P<g<$3NS
zl7}>0E6_*>yKP_m`8-UBe@ztk`=`O$ojn-=LOnP5Y2|;FXG-w1(z6j2#G~}yNAP4Y
z;?zb*6O&2GOkB@~{?6jDEl#-u-!?2g)S0FM<l=Nhi&01mAVP$1hV_}OOR|6~E{KT*
zyiON5B4^O<y4pflK=mTMDs$OZc;l;%hmxIih-XKCm6=H1Agd;NJEylZ4LCbQnDTJ<
zwz2YE=LQ+1DKdL1E6`Lv@f+J&<pYN0K~T8lJ`Jm#W(ViPas24Zb8WPO`59IFRH<r$
zJ8L%}IO2KACVvQ>D-y{n4t8Pd4wU$p<t4(QO5TLtZIKS26)&p}1sb9X-7@RrzNq_a
z*~0=<cm?|MV3fuZiEtpyC=w$pHYDjn3~TpUw66j43;Pom-aF-?O1;5JmAn-xk*mj{
zSs-#jp5r9xk%1L?q>QX*HlO7+_T6-@HgW<y^^JWRGdpL$=%~Xx_+4!Gahg=@xzUKp
zWANEO_uoD7UJbRFs~|Lp?S9Tq;!3bmocbI}NEQ+xR}gzqZT$4DNA~9W#F!98%Klqz
zjR@^MVRA;dohZ|EeiO-H=1=P?O}fee7Cf3v6n|qRS}B^3kTs*hxl9WjyKL(GyeRX2
zpw91IqVQdSe3rkyF*ut}H!9}*HxnSeJ}%(`i+mI*oena6__cZ5jYhMI8zCwGGp|gL
zL}`(PwV62wSWBUxPg#LoCrm^lqQpQ2t_z$K>-6V$xq0nW2D|&!$c#taFQ)<1!NLXZ
z1~vA;&-X(wt<k7n_G~~Og`vi#p<B9=4-~O4Plf4t2swMZHU2y>if@}dF8io#Y;2ak
zGxpm<bQNMnv)Od3xAhGd7C=W}O`C-}8|)VMv)vFTACk9gg*TU>JCA<8)voNfhir6c
z$cTuD;G7zVr$Gh3e@w5NOi<_ZCNV4m0;A*pug-#(aWUt>e++zh@GNgWq%H4w_0qp<
zo>QW=E8QwvI&r!h<+^(R`I3E0lvmprxJ>6E$>0G~|0-?n{MbUsJ1=D~BVO>{F3tbW
zOV-_U#|C!5;;g{^sgSz;oFUQRGE7n@4;JYjyXPGf#~Ap;?rhn5oboO2s0gKQjn2tm
z-R_O^+mGK<eE`~mbIJQo1VB=41o=1X*faqcM;)tsPS+VZ`(eZ)_Hw_$55PYcz{obc
zogV}5#$NmO2Xo-sy<**kR6w$$4_m|)O0&Cq5n#Bb?{{Gn<!Vq3%9{;)_W?rIO>`#&
z^(PzOa<Kx!WSgq<+Xm;Fher|7Vq8wXu|FF?6q!<Rw5-$BtCC?g^EjW({4&eg9QySc
zK!!mTiqfpwmjmo@bj!xje<yZ!+H@MLUL0R5AB_BJo+v;!P)<K)TO_F;H%sZGVam0s
zCzIp+?&Y|qElsTY$52qVg4%s3&J#^b|4>Uh^|Co|s{)3a$z0=Hj0rhcA(|I{xz@r3
zd$ey|KnRl!v&1k^v73yF@faQ{4gh|;@Cko@<xYxjw3J9*6suW<=i7X7sX^I}%@!Bk
z+2Xon^<1>CbS!X&#dzn}a^hKS1tT?}A;6Y<k@zPGZ@1+LkuH_<xU7Gq-l<O}{32d*
zTgr!m8aDpG`r9NrAQ>Fn^%S;ZwFfKGDEm5<1TS-YM}NzFY)48ot=?u)S3Gv0?0K+e
z8WQ6}+20s2QK2CB3fDOQv|9tI0L!TXRlcHZ?s&QIypeoYJ(sS3)yN0<3;gceMvCCs
zv8<f+mtSBzTFw+64E(OW%AdM)em@R|#g#QVvpXR+s=s+Lb!(rKCqn%y%y5llBG18q
z9`elBp{dk#F_0ZW>RALfPM-N8D+->dEXqm18d+1Ag-Y!nHPoJ)zLWTG7C$gT&wUy2
z6ELwQ`l6W7M{#dW?gWQn1CC>b(lq4!KRDDW9}Sm@ex>IT=~>1--(Tro*S+;5y`9r$
z`5&Ft?3~6_GB;$)nPmgXYrpIz&cAlw@xddogi40rH{t1O<N%KiFnpB!8mG*{m=QrS
zkka<`Ux{Zbf)3Z!*&G?&iV{<~hTeJIH3en<7EnFMp~>nXo-3s7Z>S}-au&E9M}yRa
z>*8Cz_*6~5^&XF@+1WxnHXT4)K$2lr`^Jq>Oh@c!*p@GzZ!vz?dAfcNJI}KQFPPr9
z$2m(Ump_P}1GqXKMrGyX9M^pBGB3R!E>k*2?(exPyk<-O8@;)_OyEdx1@JZJIViqA
z_xSgCT93_`^5%{N#!6h>25w(}KlYa!P@I<k6S+I`7vC#6<7xQ72Lsk_2f7npO{KIa
z#>Jso*Wu0Sxqr!id*0((rhFM7i#dOs>&Pl9qA8xGW%7R*ZrY0G0}JTejLLDX^vLL6
zewL7spiYw31+=v1E~011yEXr}=Dii~v$ocYC<oK0+Y<}~n5^5{!i#?Y=YC)iMw#bx
zm0%a}SQ~N<Ug%2=4peYZ_DwrjyqTYbQ5DP76|C!SZYumUiQzm}j`ZCn`e07)4uQuf
z280NX0_9&e<xQa-uSXq}Qt__iJlV^xm7@;N9e3Hpr{#tlz70g{Ozkai%b$+P`LmvE
z1r(M{&I(pq+^<5pS_2t$>VXgN+@UiVRPcYCPBKYJPW~@Hj%D)EBuTgKi!*KBE;$%n
zy+pD=3x7o|6>{*q?<S&zgL-}MHQK%AoowcFVq>s{3^bn};;y$}8L}~W(mn-7eQ=Dw
zzZke6L$gda0f9EwdSbZyA&qvYe>@5y-fu2VKn2!um3daWJTdNj>ilrq&=I8^KG$~Y
zN|(?_XYe4c|1hNC1lUWyFPNoVY@qo628R5t8O@dJ?K$s%ApGcP56I?5?dcaa-Fmlm
zpPQ>d3AhODrd3q9(HMun)h7;e0#+gftm7o9hUeE*$$3dIApxaTF(}pgYnOYMdK5R+
z7&4nENz&I1I&AY<S)?dUHhm}3MN=Nb{J)R0!t(U9n#FOOW+(oeclbnwwtOXhZ}niw
z#yO$3r*c+Tw^Nmf&Xk}^gD-heBw+c8vQR=1%lzO0m2FB<%{f{g<?n($siD&Q$^(0g
zxD*Og!+IggW*=sAEj~CnQ59b86SnWOG7{0n;zfJ~IGXM_Ne14V`y>qA85BVqN}e`;
z<5E=TzX`0`=L(i7b&W@=w|d0#WL1$^FC1%hW5^$MaC0zb{%z8xhmz6vr+FDudj-XJ
z+<Hhx2zukgM`T92-E0$<E5n5%hD#G`?wdto&y|Dp_IGE160I`7$oiZ4R}OTbvPm4d
z$p4B@g%y*;&zbT8ME3*c2@uiGO-Wd=NQ~_7t?hpbcHI^bIk`To6gQ8y;%#ugr9RJe
zAkmS>I$O18-*LW1MqHU-=_4DS^3gBf`bodt#!w5~xmvW#t>c6TN_M4;xU29M^{Ftc
zAug-{ZPp~E0dJQqB2X$vc4<RSsl6aT!v3fXlBR}B%1!D1G0ZYbZsE*f2$*Q}FBf<*
zs#~&3Jr0N#(2vYRMNbk#%4;6r1*GT(C*OHVnvn1hQ5ZBi1OsYVf+C&YwplPdkBj<o
z5QiP*tqvgdUtStmXI2^>@dfX{94iPxGd++|8k`d;c>0o#@1Q~KwLMy?@chZLyya33
z{=VUePLa*c1=tEq=_mVuBNjM}kG0C9B3u@&06P}-lV56Y#(`~}|1w;oHAti5aTdqq
z{<=_TG>U7r@w5t<IxzRkqyg>4yW`ILi13cf|B9>iY2{4+?k#<ka@=Q^ZK1{jC7sJb
z|LV7b&nEg&xMlw(%G)_gD*!LHz1N;ri{-#qp&Ye>cj^;VHL8vZ`D1f$h?8t@R5Lx%
z&K%9IQqze?<t@<;zDGJG?XR;Hen}O6Hzm5xqZ~fK1XqxzZpY2*+3V?<$8`E??V*jq
z<5bE$kiUAoSFZ@)2X2=?3YBpHnC7bK2CSpu`;z5-y77K7y<IYC>FS!AZd?BTTbGRQ
z&>DBwe%Dqq)^yhX)CgQ2dwgqyWeuxC9CNxghi@?+zx-}%vUfu9?YWHzfZUVP@x0^n
zdX>FbzhuAVIqtohwzdT{AQl8tdV~K{1MpE7(Io&>I;op;0{C*k<v7N7w*7wP5xr0Y
ziFGAyi`zgG@$Dd+&^7Sy(U%`f{EsC|pU%VS`Ws!YpLgYfq=f{&s}*o!T5*|{dp-po
zvv@S-J!klwI%fkZ=G4%~Cxfn{HC}#ipWb7bq2W7loyWZ0$2=bbZNA|vd}%v$$~d%=
zGm~eRbk-i#1<wOfN*@<j+q@sPse}fpgdU_;Z_3>-!u;>UsJ=viz0@oQs?Oe2o9>8E
z6oPOMjxSTkeav^(&Vc8Wx5Va6F4Jkae7)<)=8pF4W6Kh*S>VSOCHNjX5VaGx^IPEN
zxBdIPd!&%?8~{%6pv(Z4F{g|Qudmb@*!#QMJuVcvc(k<bHeLQ;m(<au*o~l$>dwEx
zgaADbnD5EH1j8uAZ>Qb=P8L!4XG%kS4t=#k5YdZ%UcB?x_onYJMB%-8DQfKDRsTA6
zK?_YR@UO)01YWo9MhPgM0q<n|f_i8ov$CRxGI7Bb<Val&{ltsfu`w4w^Ec0I5&RN|
zmoleE8N0iGBy%Oyi`LrT<*%Fvyu9)0Q3@CqLFWfd&TsRbi)z%_lNN1>U-DqeahRNd
zZ6Dx(kbkxGf2QuzoTMlgaxU%x>RmuVUQidIbG2)_;kFoeio^hfefXz{kegMyOY$T|
z<rt3PwmS6rmeibdYuZ8_2OdQi)&_VdTF#^oltYkhBMj9s$aj-q);amXTBF?K;OHpg
zfU>#(;24Hm6eeKr#6=JQj;R24YQPyiFG%1>h%lJtg7hD}Dy?Jp;&HULq)hX&zOTQd
zU)AtJ_E&xp5L9i?YWdOLNJr21h?=eKwPfvhr+FDXHtJC5TMRKmHSpbk%c>|n4nKkl
zGw$=f$RD8FF2f4DdwO-M;}ss34fMaD9#SO?s6k8<allBJ2NOaC-CYpy{vtA{od&-J
zQ^Di_C$hv#V+3_lym^4<=Dt*?wO08Mn0TK>e!O%?aB(h_HCR3Vz01UaY~$9Kf%hZP
zhX3>FkvxX`6I>>7`?Fi~+SC0)W4=X{^j|>DTTs5yI@~=jqv~mTS~e;z$lSF=Dyj#>
z!dl}|)m{r52GiotS&zuD8A_B?R8`>e;^;OYGAqI5kItp{+Vof&&=v0OSD930GtbjP
zFDejz4=Mhsz#rEb@4J>YU{{Utf8Lroa*bpmm-+`lDYExsrt{aw_rXJ8{<qG+4|JQS
z6BDS}uh)gCnVEV;pH6SQmOb|}Ti0QjAl8BXf6qDlX<34CQj2QI9sA+94WD}|h~@h9
zzQV#+7Jmas2J1jvVWo50Z9c8Ny-Hgi199D3Z?|+m8ROS;GiTsGHTRyGdWQ9HS<Njh
zSo$VT+#crp1z|gZ1UdlRTz7b1;d^4U>~&P8>)4C=3{+2Kaw?H5J&q&eez$DzDJ*Lu
z=PnS%C*VClz2CF!_jGdYB@oHNj)(kSsl|e5-V>Zv!_tvo_Iq~Z@i-^$!zrz6z3*j$
zjkeeJIsZYVd$nv{GpNyiiAkF@1P>;Tao4Tze>rN}%krMDGWmG7>doYtl>njvCb@Gh
zG&rom=Po?#o;Gpb5%AvBXV#on`P`m%?-{C6?OiNe;BNq^=L6TO2N_UC;+R-+roRI`
z=;gkJx;cICe`nn-t2c}ge!vKhql}O+yQTW5bo=ICO-)Tn7M6LF>k$Ueq(BC3=y*Bd
zzhD1RGUf?n)~Gr+@x)BvOW+7=y)IzV<GC$zIujyf^aupL%9qojj*8aHA&wXx_2Z`-
zgEv=*P43-<?C0%;FUtXJL%MW3io<H{vT9JYIg)>%wLY~cCPIaXPsk%@Uhj7^jOp!b
zPbMH?3m{|>pRXT|nuYsyWbEL4smAz5s_^=D+~%G^p_cmgCYW@zyTl_q*4>Gu1&s+U
zru9&LiieG!i{p;zCakDr)D>*<Ph3w|&lUSadG0IHdeD9@H3F2VdBET_+sgC>XlU<F
z(sk`1>(lfKI_}|%*UgCUO<A$G{Y??CYt|KiHj1cx#wb4e+8;@ux1N;$c>&f6uRs!D
z<*FO*2?FuT9?$rOYd_~1LX*@<JN<v1@DBZhE+eKDxRa@Dmg8w2!eRyt<PQ7u%_1h!
z{Zkmr#9%rkvR}|cgsRJ^OyPfF(9DGX9Dhow@N#5!n*JI^<<7%Dsj_pCMLTN+jdh2u
zy@YVAvdX`E#uL3E6y)3Yq<D#L;W3gNR8UphWh`Y99+@-yqeUNQGMPhDwbpVRw^2z<
zBKz?+gGIGiSZp?%yAwZH$u_(pg|moMy&SF(&6Z2GYZErY(-Zxdsc3fhE@aB+hbnL-
zs3$EtB78T}RJ5nk{(NW`!)B4S)GB5;IL2CjG6Rfas=VeLZptsiJ=2)8U8En@u00To
z&9V=p%CChI8bnY)hw^Q(U;0f#UY`Hpi`>myOWI!tbG&Od5|h8RAY?oh*jeHp++=%J
z4hk@3Oef;j6B?=C*JM_3O%#Nj+C-8^4yQ?v*Q*ivU>X9CD2YX12R?~I@is%$YB;un
zd_Xw%dgu(Y*$Z_t!8crlgAH##Z4J18KS{W78sLiIx&qLrG!QR-1gH=-Vy$HVGj;U1
zyBrsknD@U=ERzpQt{tnENH37%EL)iOsO89Wp30yv92<7(3xc9cw@YT%q<5&-n#7PH
zrP%7tF>#?ha;er90(M@AiM2H$tA}$SJv0Jg2hq5jW^?N`0-T5lh+u5~=_43}3f0Zb
zy%CJw`1bmw)#k+k(Xi&nr&AwkfYEeVTmnIj6=6GH8Q9Re!%*AMkY`nO7iy+O;QxHF
z?AZV1W_`I|==Gwl{=g8`_Vp;T{*G(KKR}z+*X?v!3#I3Ehsl2fh6xy-3(6^`wE>rI
z0F{cUnwsC^boS--w1R?HhFXcb<55||)aq(9D%4WFCxEZU=Gt#LAQ;=Or&yL^zHz-5
zJSICsN&C&f9n<VuO->Ym4WK}i9^Al`GOX)$P{7W~=>*iIjzctshFVx6a86AG*aki<
z1}}#iMe0DOxEAz}W8;EPw(l(mrodoM#=q{1R)t1dSPjcg!@o<1(fj~N_2J@%^|3k4
zuBExT`To?sBcix?87LxM+idmbc(x+h`uJ@K8J9bJ&&E;n9*&S2giuFnv*l6{nObAt
zAtyUKJ8>{^`sgKatOQM5X};By4H27hdpF&I&{;s&b0=veTPS;VRi6`hnH?R6!SBC!
z>#1u+UR_<CAIg_F1>8x!jswsy>|4pl2h**AtD>{<P%^<zJ}-!IQ`VcgJcsf;s-CE*
zsGB*Vhe3r;-;X}uVk?Vu;Ir<Bx;n?V7ARLDp<ZlHBa*U)ip-)EA%DbtQDohR-X7Id
z7e+%RkxIu)%QgIq5DHokHu+(Rs}jXV9`?gVjj3l!^;oIk+)A{-vNs;pifrt$$#CBv
z^pfer;tIPxF{M%c&?7_Bq>=(YLmZ`v>W!i-CLLI!t=6}}>zJ6;RB2?$oBlmTt&I?6
z^+iwV?*mEOEQAWCi#f9V>Bv;#)Bq0c7p*TsL2^F`$FFD4|NX$;9(`RYWLD1-c!2#}
zx@GVsU6BCdQoYpO!$l@MFKsKbB2Q<EU<J?OjyBRAH5*PGF83R&mM*88DE7Vgucb#N
zTh6}<(_Mlle@ktF6`yEq<s4q0A*?fIX!T?CvhV(`NQ#^U^dqOw{;ZFE>B@%`cl`4w
z3Vmu%YmFp}r!UXMKl@uO^T&IICs#sqpnhuN*zaJRb<xD=P$A&lzcip|K0)s8%JH)D
ze$hhL`!V|R=|gvDj<F_{9>OnFQpR|+Khzf@zbQK>jDHpS_4mV*N$a=9?ut7_(%|&m
zJtjv6mX+8tJ#Y(uBi=G0f!3!!pm}1S$CIlUl)w_!NNOHCg;V}F?aKj1iAkkO<I`eh
z9wnCB-&lTHd)h!E)V)mZ3#Uw<EyInrlB;Uw=pM|}k~fIm3ywzWud~oxNj6uSB()&C
zD0Y@4h`?srZQP3fu>N>dk^Md0!G~;2kO?ayA-N*v9N;C(vWi-$n#<>6V5*=yEq&H1
zGzNWu4X;3vTK17AJog-Ea)%T^mF?s7@6U1<rIQ8LMEq%v!&RpHDtssY>N)K0NGjXu
zOK2$U>f7txhx1{BSEs2hpF@@GG-gznjSy_8^Y`gik@}ZE`TH$Lr-~YZg6U`OPr$bc
zwUpI({00<F9H=s)vOTjJ1_D;~AduA!(g?uZ+ATi;v#c#3-zR4^8!Q4wXm`l-th-i=
z41S2t!`YJlq&gyN+SQ&x`UHC+smtLmQ_zcT^E(^0<yUybTbS*~3qkZSAW3;_L*{Fg
z+gtWybdV|=Np^Xyu>1eVOsdYI<w=2cowmQgW6=F{_N&V}UUH=Y?qJX)CK|pr+A!q^
z0}i_;uP_cWepxT9`q+0jSqF-*a~PNP@?(kvhqTGX%o)0(U_~i>1ioQHJyAkFX~GrE
z>~QQ9SlGB8WU?VYu~pT<sHtRzMZ8g(2;%r{*(6GxE%LoUD}O3IujS=y^Lkv9;mKlT
z8C>o=XP388VBSqvwn&VnffPQivolsiI3*}I_6r!g6xcP*Wuw5g(LB1lX;Z4zAnDbK
z^Iufe&f&0hI<^S)5{rsOk5z<@I+Uvp7QGf055$XlHk2U|uj$7nLJ^Ct9wR!|OVUnn
z-^rQ~+YyEWr{SXuaM*XZWp{Vv7HcE~yoYBtLqdvNK*XHHY@rRAKO=KuSukL&6S9}S
zNlWXC^-Y+w^Khz$G<hcErK^-$$vN8Eq^U7*I$!SAe&~f~qxvs?bYR0=t3&=!*u-;T
zs)~O~XOuo3*zh=W)Zbhlc6!6bpy3%y2W?L<ul9MPd|N8<>*<3GlWZwzH2PUYvn`vs
zn;8A|4{24H874c)&seeUgNo2%!y^RcK9D)`ZKYc|y3$qmMvobf(eD0cANy4VXMK64
zFZG{>&FQ@!P-&Z5Hmpt|N8WTSp7$Fv*)QHhPuM@>@bj$uEbGuCgZT7#Y4iD}tk&cF
z?q`t6l+;K^YccxIGS#lxPrBYGbzk}6*QPP3;DWr??+!~UvXqnI))^c;hZE0;fNkt%
z4nU;yk#liVo<FJH4qWh->W_p#R>o=OE5{Ztm7V|v=KYZTk+6=t$)#7!5}9r3^(*t|
z`?PQ9F2Ob9(dewF9sZZSr6w@~ZvWQOL0aRBt5qfqse5a0FKc;6yg*(2rpe{JRBbk#
zUvNa0`ddTaa|frpa8ihD+$cuic0RaYv)Sb!|K#K06fs%XGFrXEntn=&X0>Ul`K9d=
zhtujf|J8`p5?eFI>vuzYKl3R*XQwdzVvF~%kOdK~E-xeXOu?bE#9y=ITKF-099Q)>
z9#;=*)|exw%7RvzBzM^q_%C`&qn~^ZpK3OK8l37^o5HuXwLvtNN0@1SCCFK;wL(Yb
zLyk5Tjnd!XqZhkE2ZfM-Qjhv%2?u%QM9Cr4lY;g#8_;h3bu{KY2o4jDoV@@kjXnxS
zJOW%hR^$7rjyP)`Mp^lWECZrG#d^Luyr?f8#U4kgQZ+x)^<uH6@ArCg4#H)9Cm7>V
zFM9_WHq|m*c?jNv)t6q#(u*$BDS%Q*c$~xW@)D-hN(P-9vUok9MR$C(e=$f@UaU`*
zrisK4C*^gVN!xLu2!G;CELe<qE3&X}BOr&9mrhFQJP%GN5+Vtd*ca*-)1Ee-f%|Fo
zDLB_S+l%D32>Ubg33>5#j%m6QtK7#5<OJr=zgO7Tr(>u3WX~mIKf_u>BG=1<f8PtL
z)w-rB^NQMwDa;lQg8Oud;u2kw_?<#UWGLM<vy0sSqRQeK%Fn-%Z)yF@`3Fz>Xf0Sr
zZ0gsQdl>dNTnsM6n7#99Xl?g-@e9tezT%-ep653Mmb~$}*}#_-7q?{JdZeb~;(OJ$
zpC{dj_4*h_Hjc5pcGO~DaS)I~jDFOi%As|el#+}5DG7?qF}7L{AqfR-N<!%X^P(^A
zto1PIZE9Ka$IGq1Qj1y#lIPK3*k!_cx5k0+CB&5D!AUIR7y@Q7dYCb=I=v_q)tqd4
zCx5En#_dG@wL6XR?8QnK{8cYm_PH808L9RIs+-vwle-c9jGvxYyK)chxzzMp)xkh^
zO|!`IG#~NK`*K9Z#*>CLo<1~xyl%)S^dZM`z|u?2w@+!0U>)luGj1;x1RBh4ldL`l
zA>>-rEJ&>#?e80|G`pI(aO30Q%~u+IAbOk#yRUeAtnlJ&Fw@d?8l-qfJ+-y9GYv%W
zpCpxP=hk&`<TH4zaa1(S%5|IKZ*SdB+wSz6TUw5?-vG)TA^_M=jF$y#?B)hUhQx3q
z$s(Z<QLyC=G{0&<w<oF$A0s=D#iz;;9XpsWj)R6x$FIB-IDSX1p0$4;<(T_^GpnJ{
zG-^6Gyy_u%^Rn>0XZ&7X;AZCWC@v6s9kaqi{N?5hco&aHlXT5w2V3xALd~SiYJvC?
zCZ|s{qDGECMy0;sCUY#YYr&=!6Qlht{udN9ILUxT<zlJ{ww#U!OGW*bZ-T^sDpp<o
zSxY#KiWA<6Orb}SZ6`{&Bu%I#W2WDg&yg2nqcJ%%xIhadYSD?#J+#eOiLRtnvyBvW
zf-!H&_(y}G2Uqm#V8l0kZGo$+uOtCurNS~2g`490w2l)>zt(K<lK;et82=6U$pNq9
zR0WRZlQwbhQSvd3RSND@G|}sCGb<&ePSmB-ej4f~ZA5hJtfeReFxdhp_v%l1>A|t^
zFc2J`%3;2QEtIBW?e7FOIMP87WJ<er=O6S5m74|bbz>`pFIqpEV2+4=YJz<!-cvqi
zlF`XyWDY5WoYJ|d<*nQ0nlI5hN1Z_w{3%qeSX=AJOLx!53}~j6W+ENrw%LHF51({k
z^^}{XMWx9siu23vy^)$onnA`^c;5ms{#gBqn7@}*>p~%Ig;NzLG3f!7nF)!Y88<Fc
zkvsHywV>zF290G5vXs%tXQuBp-m>GWnArQZT3Jw209&uzfk&WyFmZh13#wcco!PRY
za-9VyJUrXe-7+QlF^Y>`1z(YY`--ax<x%#14;92a`ynQny>?UuLht--o4x>x_X%4J
zXb&cS_8VMpJxw~YrSBV{Q|m4fM4y*!acgQwtp6L2fv1it0XMzt^y@?xrXs)!_yBBd
zj(W+(y9IeGimTeV7eVmlKw)}vva4oXP-o3ejOr#?#~#;$pmA?}!1yWaxXb~8<gnoG
z?MkRBBRjD=Ym!gdUNCK;-o~`!_4QS1$(h@<AO_S0ZoJRGwSsZS#-P{tzTQB~PnQzn
z%Sb(VxhN{k+64BfzG`~=wj1^LK#;Ie9`e=4@czFG_@l=-?@{UF(bH*5#p@$Sg{R!h
z&HJ3icbeF7$*$p|Sfj1@^(lg>{eGRD)hJs@taLURjQMC|qEX=t8rqad-&$hJ;HK(5
zaCXjU$PK(o;c_~yh6N&O{V!H}WKM1g!sr1v-Q`B6nVY(ga_GN!iF6p4h_P1Tyn?s}
z40ZNJ_5yZWxdILW!A8U?S+Ug_G!f-XC_<Tv)LZv>5oXn^19yIQwJt=FAY#U);|iE>
ze=JM3DN|L)py97)eK?UD`B(NTbJol(7P?U=IE0J`Qf8iGv$W|IjEbS;CihDo_`xbB
z(|;5ekc*m0(sCh!n4yG%Cy%Uw%6mqY-tMkU5-yORy=tX9c`QRpQk}}d%9il@aHkbO
zXHh>0!vmUNB0|Kyhr%RuR5VvjYdq##ks<ZTCov1MQrl_=v*Jj!|K*VVx4wljY(}G(
z>7J>3RLNfJ_V9SSGrd>*n;Y&gTd~hX>H0rQ57x&^YAh5A<RwgDHkT0k&c;%}L=;4E
zM$u#nQ+1+<A|!_v1mOqdBKd33D56q-RwXY+MQVnO#<oepPcxJs%V)EZ2abi~iNo<$
z^G4v?YO}%Wg!02?pmIfi@BS&wWqK7L1D>&zIbx6UD*tKvY*IW_8JE5e`J}mxRI1h8
zTR48JN=Z(jDA8AVdV2aeUk@CNZPwrXD0JUt^~K_9w@3~zkJ|X<@dkokyVep&+H42!
znf-*rlKIiXVdw5+#E2(Rucpsp;u3WcX-I=FVvby<k{2`1q)a=&<WsHyPPYf-uX-h(
zq?FU#H~cXli6S}HmlhW(mw^QsEpV&(KHc!$fqxDZ@~O;7?>>mbrHt?Y;5+SWYCiP7
z7Yc?Agg09;Lh>3K*g!*t2&xgMEV~wov8Vmu0D)BLt9!+Zxaqx%%?A~xYYh4%hYs@x
zqrEMbOh323I07+9LBHrA2>tFiaKBsu%p3!R0iYy%%$DejJ3BkCSR&(4^|PsDIr#d6
z2reL3r{$LfCZgQX9IUK1-t`Pr8!TX~?kfYpYuF1|cUaHT>I2dc(Q+T<G5U#C!ftI}
zEItcR?goXXkxoFf&+`^C4PCg?;d@R#)6=Ed;@R3oHc5F)6dE3=w#H^spoGjv$`LbT
z2oI@l$!0w8By-Mt7>RYT`I*T`4+h{XU$)L~Geq7?k?03zj+ttF8>DIx3f2DdRr))r
z6tU<saqf1giz&zM&M2<z86{{}@|^;kJf)4fQT4Q)c{^&Rd#)=PxfM<UMGhWfrH^HV
z6d<~&DZ7z2(-B4!8J0i|Pbk8`k^hkYNfa7+UzoIV?-0+XcDw;mkysU4RRniUKA<Uy
zR@n#;W_i8;T7?%Cf%HFxvAgY(%(+N!Qwh3(FtI|A!M6iyM`u}^v^*VUf|bwdvm?W2
z0@RL`j-pZvh>k&)kw%&#tXUKUye_WuwN?67B%1YCjLA(2eCaXfyzaRD)SDF#VOkB2
z3b2_0foU-XH1o|v0b6YQ!KShytz}bL9KSV~2k~j+R0p(aoiDLGWk~XDOc|mgi()yU
zX+(w5q)XS%#O~wrdCaOy2hELG#7(Td5X(g1cCnGLuo!rqVfAsUY6y-dT7<t@jF~F^
z=>3<=v76CL=>R1Zk74^o=%+{D@Ks8(hWwH94^^^5c!%G;TC(4Me_@+UwY9O?0u~|<
z1i=5m{KtVq45u(<&v4r*yjyQ*)`-5qEwdSNFK!>bt9-SBXO*HH107dbQx<wqkM~3L
zCOZr4x?Ufxmxe?q8ZRR0wf%t%Z?rddff(}R#}A*22m|Q#q>7AZeRtr*iQN2}#A|e6
zrW;<-%r!0byT}LS2)hZlm*Fhmmgz0+eXOv!w8R52XY&AnKANkjF!%s;X;=te8n4$Q
zTf6APcQCcIY@HbgqGPP>ju1!S*|$p(X3hTvfzi`w@3&YE+L*TLdUQkC#^2pkwP8X8
z%0r7V1QO^H^)e_*a4aDK!S>QxRx0G_H}$r1g$#(hrg2~Up%+P6Myla4|Mr`gRHvSg
z9*7}O(Oc`>@_PN|!SB|DcY=Luh?B8-pY-*4Pv}f-L}=su-xM?=BB%2eubPnQR)DC+
zU|z{~RnA&(grQ%TvC<C#GecPo%*AS6WSSqhr7Q%U$kg7DX{7rj`}EX3fg7a-%2y^I
z8YZvRKM)?2e_yL-UHa&opBS98t<Y!fm&WJ6D7>&+nZkwQ8xxPM=)5^@`R#z-x>6JT
zo7C=DhsafQP^FE7^SUK*IkvT6GLFMi+A6rSo3EI^wML{j=65!i=CY-8ZhO+)olkm1
zc2nA&Lq~K&vR@}_T{<XHcx=Mh6H$mWTS>Jm7JWhknZ^@s)G8{uR-a4apd!Q+DM>_{
z>epE`#n>$?H;7^MaAW-#fDzzYb4|rcb6azB8w5F;sQO4@6KTFu^@f7rtnl%?6->fK
zrQMt_@BeH}p5pNhzt+~gMZ3gb&~GTl;c^l%cp8<uqY$j{dBXO{{=&%os9&2bYky`o
z?+t8QOQHPWlSpjFmQY>JTRs;JW8~MF^Twi*APGM38q*?eG?7Ljdx0t|NV-qaQZ5}s
zg|`{ils?4M^GBYxN~@uKC?><jzlG=L9`xUXiou5VV}@!qqe`QkCECiR)>4kD&S`mR
zcDs9{j*E`~*QSLRT0Eh5JE3Yhiu-K7Kw;Z}Eq5u!5|=yez8b~8jmX9q2S6cH7u2$r
z7H^5CXnN&Wzy_LMQWBxyd#KWBk6wRp4)|Gom8Y`+OG`hWn+#4QG@L{weZ;%{1L8kF
z$!W`3meLancCdNuYkJIt&*&D$>V;l)g_g{#9M|N1T!*qp|D7Tp2uvgR8(@L2KwNF`
zKh5$13!VnkhTr8N70*th%DHd$yFN{?Q*S%9#ed%U;a#7}FD(4{lg+*3uI<e6oi=jB
z+OQP7_@(e2aP!r;9)Z33j4{M*S6frxH8qHL!%&HX0X4?E+>hLu#$nUAldMCX<mO$>
z9o_vNNQFW#R;Pe~<jYG%#)=Fif*l~v2<DO}N9y$e$kw{cQPFkh3!r>eaD_okVmx(t
zBOo(*UPRm(P>jCzDPjsb_3;zFb^h|XbEVUK3KxHVT4UU)I1|V^>+qwvLJV8gwF^$Y
ztq~m*$JT@HOb4VXFA6aX@lu_Ja=$NQ!Dw2I!%Q6+ABlV_+aacxmWGow^m)c)zbi;Y
zq65;BgTnp7cz?SPdbCFZ$4@<WDv|%|xq1&r0L@+S$AH0LsRjk^XPQMkP%^25*EOG@
zJ_NCqJYsmgyd?Pd_gM!8RT9f_9QH!h<uu}-Uva??qiI8xNW!gpI8y(Vc#8lCQU9rt
zf7&2#4IWiwN(UMDlC)Lva_3(QPlF47XhzoS$na3dhE0fj&H+MHy}X;Iw6Xth9Xs4l
zLHKEZc<NeoWvIrmR@C4%*Y;xB)I=(zyY=RnEmr0`r;0j{yk8Rx8rriAMm(|Oil_8$
zf1&2P3J*rq_@r=Yx$Md+HwQsKmI(f^9|#H(UxI$@g=YblXsV?BcX=p$Sb}D0K><u6
zxnCcc{HPD&Azz{u2HCVUafq%x1mhzu;KQIqF>CYXU^8RO;}v>iG>>i5Ct0$XldG4#
z@v8dc<%%j3`}I8-ibDF&YCt{<h@SOp1%;bcRlwhPhJS<N@s)2xhuZUL%2?u6j$uCs
zicv!`Mb0y?=OPAzXf>^E%Rw{~@oG{>ynrO4UriiF1XpY%W&!?VIDV?cTTqgJ{p;*R
zSvuqErX(6bh96aNK9z^}XV-(zMF*>4BlOhS8Zs`4TDK>Uc*^(OdI;4}ems~VWj*08
zresI@2HZTNf0j7P_uFO5Ho7^SRY_Zj{Lkphz`)?*?(TfMU;(HA!A3Y4{;v;g*4@D3
zy$3h6_de`pEAVDgB8d`uAG5r#gVT;TKjJZhOd<c*M&Am+<@)YEgV4W>Px(Gbk#Y`b
z59u(7STp@o%t(MKtkj?*>s=qUOzZ1-9_|?V8};#nkvX$EAz@cu(LLae8+*JtY68}{
zR*iXU<Vvwx$zfV;LY3QzHe<^%Yx~>FS&a8tnoapeeu6ySRSQ7*#+Q^gTTg8<XjPl)
zf{y_-DeUFtb-%nP$Ymxz^9C5r<QKqgEb&D?fW;Nkv2(QkM5-a%uI2Sg=q(|P(Ceop
zk*|yHxzE8=!Fzyj1A6rDlR55nAQPuB2wqEAtFWX4=*|Ohr*tpY*=RM`uQYmZ!Bb^)
zQl#iAkAt#5z3-_^o`W*?9dAz^@H=YutrrrjfY2=0{Pu1bc{l!+x7}*LU#e>2N*F*X
z1b*bW^f`H&Ch=S(bHfLKfXQmB=b-659B~01y?v$Tbeg%GcN)hDF1XeVkPtzb%OTvI
z`6W<tXtkzDyiTxoxS6?SfX0YA%ou5l=!mtM44ION$;9b2sa72t4WfplnN$-No##7Q
zXZ%Wnmydcgo%U;3_y@DGs6U`)s@)C#9)aH+G382RU)cIksw^EGJ5M{KoBglsjI{3b
zBRY7ImUR=8uf5+U>lIpn0vpD%Qcazs&}tY5!w!>VGk(ENlni$^9;l>3#GM!YQ-aB2
z?0zsVg&MoxM`w&Jw@T^s8QQO!%Bx%a!KyZ!Fgu#aaWC>?j0IxmV|cV}>o@Dal~$eI
zM&^jl{_HVZqt0&;74Eu2D~PLoXUVVV9Xwm<{8x|JpIrzgO{DW!IG%b1xf1zTbWElL
zM9GbeCQaoT%QY>fZ9LEtmjlX`0}|t_@h@Hcqeedv))9Y|_$-PEozsI16+|r}JY|VX
zZzr3;q4xZ((r%y-OJZaYkDglFggomuX=N6P1AJ0iX>XAf=BBFIkSDu<4D9v!$VuT$
zMUQCWGI@DLzt%Fpf8lCf?_X}eL^w6@LPH_+g%{%80p{?iHB2=hlk--G={vA(3{Yjt
z&+TB?x14a|;p3x;7~;T&8lA_*_?S-lVv28U5xsF&w#eRB4>P7ua=RTAq-=T;V(Wf~
zOqGQtfldQedEFc~*H;u5n~+CKu=&3}duj1^Vt3q&tCwm9nrXq29QM;;MP&F~4!(<T
zwrlOaizhNTF=krJ&f@~l_x(b<8O~!QnerxDSQZ(Muv}|CY$yWPsHw8%Lnf_TU(VL!
zBndZy5E=heRiSk=d-T%jcb)mac?;fF#Y8Q=g3|3b_k`Yf9K4s7G#kH@6c;N~V{ktm
z*F225w|D2o2=bQWNyBvk%ir^`sIc+*a+qiLw~3-!k}c=(FRH8oObuPY6aQ>625`H+
zyhOmKMppcL$E-B(k;a;egd+6J{O*C(YWLwOF0KYjoAQ=ZUP7<^9IrvYf5)kELVT+8
z^`j~Z0=;2E&qlkW>0!mi1n$d@J_j-en}7p_OeUE=b@Cz@_1#&r?6DP9s$GX5uC#xA
zy!{W6$#mra*O~Z#UVvuT)+4>MG^e+`Xx?z|)26ViaiMJEs&6=<_w7Ow%}fVR31r)7
zj82VSH`WCG!qQsUU<GdO#&R8hW@iEm>6GbiWTr%Z9W>kHk|}G61o`#)PoCRyNoou<
zm&rq9D{5IMcl!f}$F$zhjy9>qM2Z(=Zpl8JNHYWKE%!=uA#B#;N4yT8I%mQ?Mmu5~
z>tI$=3u}6NX01hTH#4s6@i`8M$Hzy?S>dBH82;8g5%N?;Aw}uIgf@<Pq~_Aab5VJZ
z^@M5bqm(@Zxmioox;tuQX_%6ngkvQO<2w3YcnUaXHQ}x)%Qdn*?K#)W2mTKqJ|rNt
z5k*_3sm~!g>AB0M4s3iym4eNP4afd0_E}VAU>^ZSM--w-_miQBK<<ZM*wTcnT1HoY
z`deFBau#%`DbitecemxD?WJaK(jluCj<0Bcrbb$XeH2EW$!yD}%b&JF>i#{&<G~9Z
zXL=OdUlwyh>fhhb<8WR$z3TpL@ccH~pK599J?gdcMq68d717pOAY*Z9vY#1grb!nZ
znynQm!_T==3B2|}rWCrVDPan$NcYE(W*s;m&_2|fu@C9@n5<yBs4;iM8Uu!lGEqVS
zOcWuwNi`T172Hb^dN~q@BK8Gd04P@z9i4(Wy5LWEqV4NaHJGVJJ3~A!C+FwCHR<*_
zX0&Ay!P_3YFK}*Gm*CBm;Qsrz2^$}(s@qjlwq0GeZCphAdVXIHCmh_Pj;ZX``lq?%
zb7e6_Z_!ot^!-t!S<eI&Qn&9Z+sAZO+7P>?>)TtxL(m8uprIuA_B&RVUf>mv!@P64
z@&H1FUu)S6r#m#OLuX-U7ZdEz0ht|v%Us;t<R3hNs>R?al<)uW^p4?mb>G);!-kF3
zq_J%@wj0~F)5b<)oY-h=+qTi5v2EksxqtuX-B-`Y<Z8~^YpprQ7;_9jwYdi<3Yd}r
zv^a3Zv{vIo+YHMhD(<wO;c~US*zrr#>-|H$eqTbt^)0h>;uI~#hm7<@a;6gv^$Wgu
zn@BHk9Twi|XI#AWAdv}@DqxVue`iz{)LO{xJgW985)%I*j+4+rO?FZO4U<h(u2jV;
zr7JLu$TZLCdl-mH0Yla7pc+#GH<LOMGaW%ABg6G5@{?7?0L>&#V@_Y;ixBjxZ*D%*
zLDSB-t6_Rd5T+gWFjK{W2eAmPWDb$~+m1_t*Q$%%D}uZF%4;}c*K5`Q&mK=irxtv%
zp(h{6@i{=qXWVQwM$U&W`zQ!fF_*?aqIlt!!s&OBO~8+?7+erbg5+}!^;W-cx2AQB
zu~}y4(%P_VqJ7acdbhv7q}jp_2c1u^TL8Z(GcQ`bw(=hDE5~K?jd-zpwdB%tIrXVU
z0GFGSw!I5^pyG@kJf#OpI56@YT|vqEmzs?~uXRhQ?&oq1^yFPL&y<Q`_K|&0l%C^2
zm(5r)Pf)O2<65^K93Vp9TO={>K{T|;C)X;RRz(i$S@rL}BKL^6zIoGsju~AbO6n+)
z6QAh9yR}9Y0abv+$B|+crxMxCrH)+%@iy34gX#*h5!S^6U?Ay5pNA6eB{_NAy5V)Z
zSSERTJ6~5i9iC4cYrS8eERH$$9JwaGAJd+-UH{Mpn(Q;SrdFVA{*aB@KOGi2eGJ+F
zqN~8OGusbH>TQ$Y&F`aGFs}8`yWl)7tNoCYxvaTS09v^F|D<IA^-y3jv${GqElx0g
zSeP||W0tP47$zg}@o?gCZb7t#W73r*J@t|8OUfTg3?h-?e#UdB=KF!1W97dW|HEiR
zfN)6#krK&_xrBJ((n1oK_3u&g8&jjcVo3_;{8!v(otog_iUG9pEBoNd^tNGt|4?KH
z_FSwHcd-)E1qfO<FfOG(yfzy_sA{P%L_ev`RL@cu?jMczMDZ3=wv!ibs@N_1Dl9{g
zt1O001gBF~eASe;U@w%Jku;P@&h3bl{(V2)j)Fal>hXHmht!Ey@O-b7pP8LCmj27}
zWCbR7-)fEPEO|%`oKh-VR>o7X*v>+$gv3eXhH|l?zq0LnVHhHYu$7QR8!D#MBJBeX
zTOA2PKX|K6A-;S?3*G+rv+vPKx9^^s&k`FEBkz|gcE0idMAj4!#*)jdtM4_G)-Fl#
z_<I0aX%Q$`2w?<K_`FmIRA8X-vK)PolI$L_X{@3nxPGKHCcH1cdQF_XLS<$7&_(Aj
zB7>bSsTbrl?in2)d3WBmF&din8>zAB>#pH7Zsbz|D1Y~!AQpd^@DUJPTX#_Uv?FsH
zvS262>8Vj?j`DHmnYCqYbG`r*3Q);ab>1y_9w*hFz{tcVFs3k47HxnH4d8wBz~`(i
zIyUCxv9IC+mh@u|<p^zySK0?Pil-ro=dGrE40un$C<qZqSXIcZO7fh8A_{VgFO4PD
zF#IKm^TmDW@_0*Yv(4vOVwH*1*V2Q#09ZQzs6UvZ1lidb@vC(P7Z-#RXLf0bYY)bk
z(o7iOHwjTGpeP5#FhXht*kP`ZS<HLfrS>0FV&2oXCnv}jw@g%4P}k*Qup@GwfS0_T
z`0Hv2lh{z69JyFX6c?e%sN*VrZGGw1_~t|0qcf@aH2rzFbC>^ag$a$EJcRV4wL@%S
zEktA<IQs;!sPMSF7||6qlY3wkXUCs$m|ki*@(FSLkS`~%Q$A#D+UAUG-)qSEwe6`G
z@S7l{&v<KjI@ZVrMdg&ne}0mO&wVw-rvB}7*Cq;+&08ABjP6(<rZy}pdevP{VAhY{
z{ck$27d6SQ4TMDhMEzR#hx=Y>g*kA}5uuoY)I5{JV!qp&vKfU;4a7!XX(}I>ZGL#O
z$ipf+r2~r$KJy22KQE$`-BPgc#a~kT`}CTu>)C(vdyEK@`HJEZX>LLvPfzzw2+~Cd
z&>HW@Wku8r`v54LeK!;q-WDbq9%Vy<_SqU|-!|`;U^qI9Z?Y=(FEPL$9JBq$3XbBZ
zv4%3^aZAe&<eEPxC|}aD$A?cBZj>>)SJMTpzAv1I*Lp))2;-O`m>yTp{&XRu7roAf
zkzSOL7OX^@(t`(n6-)x%6CU5(1rgu;z2ulA!lJgpRcKT+O5Wa7)o<9;65nF#CczT~
zK3r#&DbNCrqi_R~oAX{I6EpfJi1`@h>Yu8ESqS$zS*UlqLCa10o%kF?<PoG^^#I#N
zAAPS3OCo_w)FR}?WEi+t*9}l=Iv818k4v|Iyt)@utab9Y#ElDwSW6)t23~T8nEBEr
z25JL)gSgxRf)JsRY}GD;g4rNAev8eMCCuPBLbOQX8<5^>)C?+PARwlXJV41Y(O!dQ
zFrG4eXP<1%L~l3S@n^&E`05TxSc6*`bN_AEy#xp?mUI^G&MzA5eZ4RKN77^vh#RP6
z`2?=sBJdL<mF1G9dqFTd^*oyz95xUI-6qn_ElyOlw1Y7{#-0<H$n1Q(t$YOFB7`mg
z#|SOq>l}Oz{O<wC(r_Gl&AJdf&u!dK+MbvF07&zQ+(@}8Ab%$Wz-C5XS;E3nCK_?F
z-3vq5)v@vAuqC=7{`zWc3nbsVXO#DLXx&HnxJU5MAVE+7f~9X&m-q9l(sI)Vy>B@c
zIxV6o4D>%_S(rdy38@%V6Hc9Jc!*reE!%73wZE<qE-8h`VgHb7h^SpO7<TCBI4(R+
zMS_Fty?QD6et2=?nBiEBoG@%hl?jS35SG&b70hAr6T){eU}?q!=zkzHH9Fzvx*eAB
z8xglr__XRpCm}0c3S}OYUp`cPjdoUZE0{D<qth4e>}B!#(pvJ?WW7{Y44@0twPTR$
zSdM@QX9iHU)=OeiuQKNDh@s0Qw7K;Ghx30Z20TD6)D;i~Woy}FPIvq$!H`yhA(3PY
z*<x}e0NTjz87Ac*K{LaE2{4Mh)*VNw^Gcv5_~pax&iV^QbqMA<_C4oant|kCoqeyJ
z5ya`&he*cqsOry&+}`MfGRzs*NR=PEph8RvVcWMS_`5uSo{e7wnPE+~-aHXlwF_Vt
zf`ACJrQ-TyzH96p4fVgrbOI%-du4te=u|G@_+B=Vu7N3A(Bt*~bvKfC%*PPTfAUSG
zEdRT^yMVcwKre%R`;_Jq0WDK&Nkc%{L1He2E7BDOk`rIR{-aY!C10f?^`j49h?%8p
zleoA7!U%xQv=^56g%Z#{)Fg%#z)3-XDUgO5r>k7ou}AyRd&lFkVT!CwqbHo2Qbbi<
zkU*h`TI1ZfRh9O%`TOPeX+(<vaNN-HcO0)fK?W5RT8k<P9BS@SdJMU(V4_rcbID`i
z1qF0SIkRg455xSEPfmBFUze-bqY{v}AyJM!tgPuC&86^oY-B8eV#bzoN#hbWIN-Ca
z(NhOu1R)HPx37zcTFp_Aa{}*Io1v8I-~z}I+aW+PVWs(p#Oy3d@ob;@qm*i5+zCDi
z`-F43e=|&r|Ch1-b&dyjW|T`zX-X(Gr)vGidWGdXMAYO05dre82W%gFIY8#C@aN{=
z%tb_TYaGY>(ILDn_OrguG86?>!!CzZG$T8xOd67j{#f!TT3o=Ol?8g70#w~%gyZ*n
zWuf3z-%OzAMoFw7CJrT#Y#HHu1p6%Ow=i}Rd5nts1?oeniuRcEzm2Ey0qXBhW&xbR
zEg<dN|B&}Bym`z9%W&RE7z)RNuwv&`?#ga%dI8bPQ0mu!AHKLE%C9hbV#+#_bP*xM
zlF0s8P-vzXKa%QlYMRV*J{|vpcJ=FnH}YyPAwU(|5J$CA|4s}-g`$K{8>}M>M7;>1
zm>?FA5X{=XcJ>L1AhH*xbYlwyaKE7VoEn2JB5<Hw+@l>~av7CBh;K(`i1-<-)FR~m
zx9v1D>RRIG+Q61gXXb4)PNW3J-KYcVuiBqOM&e6@3y`u@85N-46@8216Us)9F*VWT
zkQz#*_ZUEJQtSYY_gOFBPfUx1Qq#O7yYXvP>DC`gWKn%wxD2k0sIpi{>i<s5-xW23
zO0Qo4%9UGFceE7RCzy_c{cqsEr>6G|9|#k@ZX^;~S^Z<`6={4722#*vuH1sM#g9Xl
zL@Wth>hLi3PrSRnYacY01eL=J42~#k0c+jR`e<x{T-_nV4c|+rHQ+dS7A7|z-l;~~
z9o?leihdKqMdioaZ)iO+N4fmHeg3No1F;EZ6F>!`DPC{EWrlE8t)<LTB|xy>c$mXM
zq9TYAV)$W65DT5HtAY~^Yf#@oVu-`S{zePSwV#osZ~(X@hc&nTZ%2Ei|2;frALR<H
zIG-TpC25g=OT}3x3>Zj#e=?d>*mkJ4{R^-kAc;__sW1eb^dc8DOfSFz4Y>-PMLKfk
zsguMpSYpMD%CE!HAQ+(%Z)9ouPcR{drBuW?xKcp|zQm9%)G$4pzRu{MOY;K-DF{Q*
zrMyL4$ogPQZdAN5+11C`F98+`@eXhs#}zp#gXC#B(tAu>#lz&k{D#ie0LhL_otp0L
zEU#rha`H8Wk;s<vu<{Z3<Fy9BxAcFPm&UJhZc7t<Ne9TBZcPDcG4T;j{i*5fST?q{
zqZ|t#D=`YIk{e~O;Pg-Y@dqVB00Z*l*f=6@j+1E9-f$&mZ~`PQwFnb#<C%Ngt6jTo
zITxxL3JC-2vdplxH@7~ch`6M3dpMN(fWlK<D%Kw!VjNgNX^fa)yJ99dO`4FZ(fZXI
zQAt*z&J#M_HCBaEzZfooFt_kz6AM%Sj<4D02PEUeM&)16D_C^8jD|V|j(~_y6d>!9
z%FR~l(X=wj`d|#OjHRYjgB5vNSOs`uU~i@pFd(aWNm_~RFLi{hX$nZG0sN+gmvzCF
zFB<cXI8Q^Zl=!kfpwL?HuoCKd`v5#QfY0ATqYoMU@RrH{>KsOKU|I&r=b<;pR;UL|
zOa(HLSU#{XX*;71QAJP<y$#S`I=qB{QLyvWItSGiXTt^Q%yT}QiZ6ZKq&dh;D^;Kh
zxMk+=qp8;EgvLX#4s9{<XDETxueLm}_=&q@F@!?EA&SN;8Hfzd^kVaZyh<pdf-xgR
zWLHWFqp3&@4+l-gVLY21jt(j?6cAkZS&CBOPfS5h9XU^WMQb#jG+T>m9YV6{OS*_)
z&-xts=kfmA`C$A+-fkHQr_JDIJ-+PLY5RMTE0GF6u}#tK2C#NRY1AhCKzlKSl!kEI
z1oTG?=++F)4h`A^0P{@V@XXB#6xEMzXjgkpR-8oSAG!)GXv<626b_7$%pIhK;kMF4
z_X@nWShBZTqlR*W%gCGKrsfOxJmn&HuvTg!I*U+T5mrD$D~br^M<50C)_{zN2&7HU
zyG>}fEU&Se4d5{h@i1L;kY4pMjPW;gJiqT=u{RK(*=O~SckZX!RD!Ze%&wfvrCj>|
z5amb3#e9{WUQnOYXE}fd_;(|ZC6e;eJk|tgEi*wu2qM?K)-eIPor8tBrhW#|KKObC
z(!+`;YkbGJPj>&UB2ceh2LiHt4`AbdJ0d3?i_z!oZ8!>&=$cWvDhOkeY%U-sv?>N(
z$S^F$D`LliGmwauS`}M&l$1a}(loMbY#xc#oHTiWJTVz9b!wh$!f61Yi^9W1AJgzb
zQV~$nEZi&pt2u<mi;+%98bT^G0Cq_E%Rm?fRTM6U5FH!VP7+K=>V+}v=PhkQGgn+h
zaW0IXK+abY);!ig-;7<~Q(^p|I^N@7=b;5K%1}}j##5K}r<mSL7cMs%d<rmO)J0E3
zR^jhMDqdEF2ophKsdQ%Ea5OQ-raqQvcZ47iC8+2R^Ir_2gCQIMn^K!M6CWO2z@YY)
z@ENCw2SpUmqyPjzXB+q@NqZN=4?5IdP(c|TDjnuUKDXHa=YWFtlqZDrs+q<nZzvgW
zgZb%;C_l8gq(RQWb{g`5rS><)I+d%1Gs64|xKyR@!Nh<0cBVNurqU;waZ_|{nhXaK
ziC-jQ*bH^9WJ}kCdPz8pzI#3IYD1Q=?WGvWrCMhow$nq@>yLb55b*x%I<uex0Q&MF
zWNEa#><);;!)Uy4WRtzP6$dE?tLWB|<z&%sM{jX|{_E`j>+akBJ6k8|+^xU>nikqp
zFx><OTMDnsMi$q3F9OyC8<2Pc*8<WhCCg4A4An5U!%}VujAO$<Fj!KzrG%AO;nKqP
zMn(J_xMU~VjJ_3X;m@zukwCc*bYgd2Q?FyG4u^SIA|Ld2qtf8i>O(tYA1u11Us<c+
zajL)wN^uBh5q0ce&9bZ`Xi!6)pvMBCnMELq@~gTMwo<;n1p)%+&3xM?>isJk`6xCu
zj?$mPtxYJD5^x1lKoi7>I>bmCf8q2@`vR`b<VsQuTZi4d<=N2ovR;INb&j=o6Zr2S
z3=pbQul>C;msU#FW2|-#5~Ki*ix}SmD(5{};7fOMgkB~!>4gOzv}yCwyhwHaMMOa?
ze?kzX;zdY9pm;qQ453Iau|gnW1Tn3sAd0>vTtmSIBw93-o3znT_(3I9e7kQMSMA_@
zhh|$5+fML9ucJgTIb}0ZyF!>=`h_SUXYl$fXrU(u^<yPfayMsdQYD0xb9U;Uz#hD$
z!&tC8&kek|$eTg{<I<(rwc1(O9UK-{RsemVrTyc7_1`mz1o~vxW6(QMGa2mt`x^$5
zRybk@I+6WL2pt4aY|%GrE7uK}l%dllJc`CyB46#$|6VJR^O4mr0q2Oh;o?J7)!xIF
zlpKy^u?UbbH1{uv7ju&O)4hN!6vL!4Vk@u&Z$?KJ!@>5|CQ^_hZzl-QEY}2!X$0@k
zedC662nHwZ4W`WzA%@Kl!Eh6vVjEJgPAq||Gya(xOyN~;ellm%SHE&el1?b1c0ajG
z7i$h<>1XUQOqOo}5Vj@^J-C@wCE&!Qoj<b1vy>>DGMzqCJVh%hQ|EpCKAn^m%5%g6
zZKY2Q7o3{haU#bYEG4FPNcH$|_jjYCm}+b^ZEgky0Fjn!zl+xcIt4~ElaV9<$Y7)}
zW2CSG<Z=0>CyICA{xHV0uF=Y*5`F^6SS1pH#4+G9Dr5*G=unp4oEDgtnrcN)U@8ph
z3cfBYF{6os(HSO4Dk@dVepphOr;bHXL@=Q%uB;q*&spi!Z=>m)t{g{2@&_yfLaF|`
zSJP9v{L<9WUea#_vL64c-X`Pc{6kW)6sySJItohkkKjrmDMF6Ak9ny)P=eX(tF`{`
zd=%=5?Z0O3(kh+*TW_WIFVb^5SM+z2D69&<;b@Gs8rMW>{p~*-DhnL|fwsoILY4p~
z8b<L`=!9S}(P>#l0md0ng50P;q;JYBV2oQ}qRT*nfJp&#+6<UbazH_puU~9bcfv{p
zlW+VLmLrZBl9qD0DF^uKNg_xa;97fP7z5Mc%oe=6M@wC*#(Xekp<%kPzRH=0Kc@V|
z>>O#slVzf6CCVXEFcCR^%9`qUQ&nt_{v(1fXu$V&<y4%0!A^)tvzi54llGZ65Y-H#
z85l6UE`CIbMW<Bs_ay`;g36)BLB{cpfA+8i-X7}&G;T*)o4R2HQCGpPngaBtK31#F
zTx$P@GQnYZ05+dVVHt#U%MYx<1;#1>g;OHCI;mu<IP^!OL1{jUO-q})dU^6m_(uq?
zh=V;xJc-a1AqgoQ5h1cj4x}uhGSCo+El`q>9zpidmV{vOQ-m7)kkFeq8_gC6WIA+K
zF<-yf+u}G+D+{7a<jNBU=KVY~^iFw0&suMY^L%$__55}8%^CMwy5cL^;(5#07$vCz
zYraFp;4;;(DcYI?mjLz^Gjq>(Cd$lirbHe(np%=Qt-6$3gwQ-~Tf3m}kYlVym#2U1
zOsXmI|8Gs)Ob#$zDl-13>FB1Xqf)n<Y4O2!j-I5B8!})z?$d|m<5b=LIMw&O8)Z-T
zQocywy*`6UB{rY@YW>ZTD$y%dN%J3@JEUk=Tbo9b6{QD}F1EoDTsc)nTmGeWXRj~r
zw6PU-NytxYisXdH98wzeIxhr&jZ}S6r?B`6=AB@mZ7hP^I9(oL<K`O?x=(R<#-1^&
ze}^o6rEo3x;9fj<mr7a|381~)TGtgbJJ%buuDs(5a+_4j!*5ABtF`U3bUO^(w`adu
z1P4$RV0a<8ZeXD%3>k$O5=SgKRSG)@m_U&Gff`0=X9d*OlreKcl6LcEy}rFKZ9xBi
zm*y$bW2-p}$fXj^fiV*hvoeAxGI6mhQDOlzhh0^b>xF--Ew;}uY)RmaW<tLq_3+4&
z1i&5P$l`vLZ!N^~%E(}4Fqyma^Lh?p_k^+v{NSM^wD_}0_GyoX-X<AQSa59uBp`4(
zDaZVAbb?2sfre*f+fd5cu+|M{>Ooy!w7=-ICW^1Gp?@`C|7zglu9iO#z5WQnlETD1
z-K5U)?H7%!D>lf(RVOIR*$K{*!O#kNW-h$AJaDWydIt|q7Z_B9!RI?_zo~GgX^FqY
zPWkHNB?-%`l9gMokWNbD!WvA?sO*O;MZT=90-3=TZuX>YDJWsb)+MikAWKMM?;jIJ
zqmXATxR%F~1?h^0nhDR!c3S_1kZ6-j`hQf*WcX6%HYY%$a;<vZ|MU3QpjrN5+yL#n
z`0{@8&n>+`O=IJAoojzBv;8jrp7fujL?2`boLDXsHYSeW5Q3nfgNQv&^1aHc7#<1^
z#%(0YQwbBco(NF{a~C(lA5b5q3~9~plf2eA-I7)-p9K%z{V@W$pC*S#XC-GxXGar{
z-h7Xqna&y4bHycvM?CjG_4ZqI-`y*k_flEfc|IQ)rvknZm3W<mLfMcV=*~{<&X=`U
zqIye7@vjM4g+B2*Gg8m>u4;y5`EgbaL1C6f5*}9X?wRz;{yM!VDQBNs=ye&j-=|ua
z;?KAaS4V3G-Oh7duH%dvFDiQ9Q^qejqPSImnKf^nuZ6XpWL?Z39MZWDYil!NVFgMl
z9(Z4W56z>5BlQO{#5_p<)}nK$KoKI0rG%_za-IdJ#;+FuH!L2gs$C284TmNMG%c8v
z-<!!769pu`{*2BUGLqgti8mRl+ZCWOp{GAEfwn>D04akHn=hF;@jB!56%L)6O^8)>
zrl6RmiY=K{#a#tAued4gZx2S4{!LCQ(A&vB#Coh=8+=^qejoE+mf==tyGSG06^(vl
z>ftM2oqECd_j(uF)q9U`KK994hSwAPQ`e&T;9zT>4<!}0yFouysgh=Kq`)u789yZr
z4~J`5p1%VdEX5a8ZEi#r*ahyQ+L83!yYJ+5ucRI48Kl4~H(AQU{X}e<`EL3}1WQ30
z{#QKla*CJ{4!y@vzV=!f<B8qC17KYSLa0Rk<lP{)GaZ5sn55kbl)FN>H*K*fTxyJw
z*2~9Q$!gy4lLvJvC63p}&<s!`B&=HC3_9QqI;8K1?Th%{10%J%9oQXjPsbkyD7}E+
zK{V4~Ll$6Qg_}sQ4zT-Y{tv7}g&9mYJkgHLZCjjHbTc#|LbAo^85dloL4EK%FI6I3
z+Rs6wi9<E&5_e<$c{jB(Srh8XmShoZBQkR&Qhw2fyFV-E^U=U?_>y?9u&l%6@K<>)
z%{qSTX*UJY$RV%tN}1S5Wxf7<NR$JYr%uR>oo>Wj(1f$7A{KrE4P6(B0Ws5prk7<z
zSLm7kx~3BK{c8A9f@F#kdZ^f=0%TiT)!&6pqE$;m%YR?=V$?=NcbMi<YF^$u{n0$N
zdbMt9QZ#)I0fJANZDMZh+(jscj#NwDt8dj^(ycPpPnhPb`?!=LvIOp|^zozEsC<#Q
ze@hk3MqKe>u_*rY-m`}M{hjLCqI*xXYZU*Luu_+(>Dih)oZ_+eQEzLUgkYop`FyXx
zw((mdb-4^?b4d(b`~TAdY!R#l`&^n`?DUl^#9OTJ*I$<@fj2koEGX)PB~rse3mQBb
zjrtMKCaTla<m#pIQz0yOp>T~5uzTbA9N4fD4+|VP?3xEz4V1b9#z?kB;`n58jI*O|
zhD4%0itW(dch%I9HTq7XnN=t&3K)9Kigy<IsR5@Ep(XDT<jKtqW6pq66WyP@?5aeP
z4DecXQlvcnQTtHYTW{FNdCwpJy$E>wKop7IH^2-`Z%--f3H{TDW7r1+s3;Y$NTZ9a
z2TU1_MxQtD<9`^<WYpK|nrG??bzisb63$4K?1ER8Uj}RhVv*LEQHq1jJBH3sEjY9r
zD*4{Dp6G>08VS?g$SZ;{Q~R5;b?K|~QEI%3{3-fj7uzxOQE$p}Z->?^+|H!qo>h6-
zJKkpx&Ui|gMbI-i^}UIg)#I}^KZEHg$ybWOOd91R-PdNbNSo58=$3Sc_g$0520Csq
zX@pC__2wJ>%<OdgL<Gt5FG3pz4o9B)8^0De!sNP9z7&;I{!Ftj^gLSx?^COL%W(td
z9qS5SD4hawl^3aYg|r5?a1Pz?6{^{46T+Zx!iD#bnK}i#Li=~Mk=vdgr{8qW-OIVD
z2!{=vo>6pp?>R~{IdaTg)sE%dG+x|j2P?JU37w-}cHR-!?|ve`W}k*GFpx<QRrl~1
zrGvns)Tgru9^lBjzY6jy#<%)sq6R7wc?strx*0HmccDJvXZi}<2GkW#<DHoXX69k-
z!+@zpsd3H>116z)Q!qddm^P2XA7aNVTAkZ_^Jf7OgAE|xQ;+4(j`T}@nsG5ZLXsEz
zJb1&LS3xUOZ+`rbWjjp{Y%2D7VKJJw7D$&M<jKdyrzOXZ^2Pu5eB{1A(#0BJJ~A3@
z*Swo(X1rXUr|Csa`!SV}Uj367%e*{^<pe~UGkb{VV+}H_!T($kXH;VuJyLH{k-(Hk
z>`o```Sm=Af=3pAXKWCPg*O0-2P{&aX$xn72TUw(FZEZe*HPAiMu|mIU9-#hHF51z
zxNrpdk`I5vXy@;sprF~2z({T#&C_JgMTt1#NWq-LL;o^gfv=GMLbwRX)Zy`<V211=
z99S~MglWs$*4RNw!4gCkgD9zJ#EYj6xj)m{OV*0${Zy6D*4I4Ietuw|W>|F=7sr<5
zD;L2MeSQ>L64ASQ7KqaT$&8;_v5YVcWm)J#uHVj6;uFP<q*o(h!k<!)toXFbRk<te
zpvw|gCfT7Uud(s*`4&ex$&TyKHbte&K@PV8S(?|YNn}~V)h%9B+z+4H2j7FPDy)j<
zEh+G{7K0CS`_~?mHQVd(Hs%B;axW<(@-@E{twdCV`62KGOY>D<PBWX5pBci1LPf;r
z6JBAwDF)Q}Id(<mFxShY;6@_|=HoH4a(3<fC5$tA{fRz69OSBrlNwY1&oLm@U{smL
zq$Oi!7{2cCp13GDEEGt-<~3}|FbofLq!S5-E+)J|C%=I$B|9oDdh}pc9NB*XqKjet
zx7R<p0B1YiAXg?l0K44(b7A1`xEi5JA|N>b@i-g(GBGxm-W<d=$*6gma`*_OnuuhQ
z-elr5O)UV)(@ctY*yRs23;9~%mf207buu(m2fMF7=KLs%f3G=MJy-nx#lUAY`RdHB
zCoJ#)`87wJH|B5gbF=3p*5Qremm!C{Vmq>VA`7BK2xs7Dpg}nxx#2i<+U<GY;Ugt1
z>gMW)8MfsX`Fa*v?s)`}u%51N)d6jW!UW5}PgKur$#L4(7ZLhpAp66M=E+}tdQGp^
zMvMbf#~G#$ceS(#!tsj74QI^3pt^_&FRyW@1f3PxLDze)co!XM5Gh6br5dq#O2lg|
z*PjIMQ{72*q0$tKe2I)Wh96G1V`?E`m=2sW&A%`t$uFQNCsD2ZAOQ=!No{(QD|hgb
z8D_jL8HSuRoL>lHBJBKUh$oiyQmF+yk8qL8KMbb=27Ggq?gIuo&;e)U?2S!`+7lp7
z_z+zLPO76vebl;O>9Y(&;K^4i7WMD-jgR7XUoxUAZwX-l5w~rlM@3Q%dB-`!dKGP0
z50u-tkM+^01=i=e4^57iU)QQYEww@}W{j&jjA2Rpma1-7pXLuHHQ-N92%K$(ij2r^
zGRj_JG{@I97<e}3HS}^WiWy_}@BWiyC4rS{1PArLqQq(`mypO2I*xEOJPQ}Ai_KTt
z{NkL?Lo9Ev8O&Rqh;4qIMeTlbH__2wY5$EeYeUO+2HfwGkV#38+m_YWDUI)mN!+53
zT5;lPsba$!!R7(y%&r6l#cc3&SR>wCBW6J>*KbW&4@^(;60lpPqQY109pHrQ#uf&V
z?z`hy<;Dw1HQOGqDOX1%5RFZ~SU|<y@QY{eQ+R!g4rIiIyJJ&wmmHB_B2fnUeM&u~
zP2<1!rcZm)3n^52fBmt^!gr(j=KeA$30vI)n1%_dZ{=ohxv_#v?)VAJN)k=L!Pyi&
z$|urv;s+utm#SRx@Ry&rHr^0G&*b@BncrKF#o7)OSXMa;x4pk*?@4M(c4sbq`99g?
z*ybozNE!w0T3)~Z&q_xs^!$1MUrP}3sL$eLAwFc(HfPkfH6E6yuZzCEZ|`p;h)G%;
z`?f#j8i{oLYaL$CDSFwp)=Ip6N+1c?)Z&5^^C}>ijV?0gQmSij`SN%KV_-&#tzE{*
zZz_(JIQoJ$%D3{juI<B!a++5-&M-vQCUb2=_!qDfM24*`=lha@&uNvU*z0a^G+qTK
zo0BuGui`M{)i_EylCIY;zne%s((j|*sJUXWpbu|5&)1n%P51RvhTZuCM!<h_a{<XQ
zir??~(e}sW6$wS(?_G$dNi|t?FSB9|s3GG$hxM0?8nhVaCyIX4hr6(?0)-1YncQXW
z3^8DG7vV$i<GNBZ2rX6RL@6ZMZOyT7&rQOSSmcp>eEwpPJ2A)P6hH1Zzq)R3@`C97
zUc+y!D+#w5yA!O9E7j4Z1|ppwyMrSy1&I}zlP89zP6{~xvXlRIXe{IMg!nhpYZ*2h
z-)(1r85axh(Ae50P#qp#vfEQT1|8an@Jh6kR%MC*G&t$M@0)X?*`@ols~(`xz6h1`
z1V129$VgA;*?cH8lh6SS2Fx5SCg29B`2u3aS?2qKx;0^xJC(bpWu_9wZsCCR(Nsx6
zZS06@pf3nV2TqppdS54p*O_6~_ol{v5*D4Xe&Ls)#r4Rl!DHNx8fB3e{y+SKb_7pk
zmQRpuib0>OaGDRj3Y#q%PjONS-eA>%V$OMe>2du<emBn3eP>+(7F|)D+=?$QAx4Vu
zU^lpok#*L;2H2Q@)fxY9Hl^}@QaWRK@ZBi>ipClSDi8kvpfW455V7pP0&}p%rhOF2
z)PBpLD@QJ}&m0E&ieLjoh^!|nIA8$wq?UFB8ls@NrCU+K>uq%L=EX$M@zK6dd+ARK
zq;Zhx#`#)Mr;t5XrsYjT`#lW*_glTk>u;SAzsbsI>im9z&|F`wRrRF6<%p}TaU=d}
z!(UEprG+cs%k@BNgl`V={Z943^EGtE8Ybh={{#IAf-IT>OK~{`XoT9~VQ1UZgPI5z
zKVg#TXYa6RE@-o2XI}6oe(V6wsM<@Ni}O9(mcxSGL16NCQF3q{a5Dpn1Yi)1rG|XJ
z&yx(B%EP++v$YkZjB)^)uB(iT6Mk0oQf<uCD4U5MJqOm(y37*q;``s_o|3{z&4->h
zOeFk6MeK(T5`x}EtI@Ubz{zKF@1Vh6>BkJZb25cpH1D1x8lRLAm35y~n%@sjKEbkT
zYMp266JCMRpS^V3ng>isFo5EQyoArfz`PJ@S(Drf3=1joFBjfcR=|l4Z*AIw_&Y>x
zXG`3Y%enCr7CXy*>mrKJ+2jT~IQqm0&|U?YWcBCig>xtlnl^kZAJjrZhsYtFPvo54
z*RS6vXw2fm6@A)RSHe>_di|Hr|4LBw?s4aqrpA_kr^6r2QDqTXdYna$UHWkHHR^Bu
z1=LG`p6+~GX(7LGb9?lodK`LUk9r_FEqSo1^`SA8hCz%_m5oIucUq0M((8;(+Xl(H
z>^2L(_-zM&j{ABR<3&oA$;(2L{Nv2$Zq0aPLP$y5zpcp^#w;dG^uWEV4G4;c<@=0z
zvy*_XMCNa8GU&MWm45i6a04z3IpCDK|5@qrO<|6q<_Qo+$C=h<vDM7PE&#qtNm{ne
zP`=&Wf*;PqwT<%o@qNt(dMqX-87vB?i`QjPVPOn7)E?FZFcyWWK;#5TMlJ3T_v)})
zP0I9auyinS@RyhC)IEN5?EyKvC*iZmj*O{*0AaQ3SS+8A{?ql*+Sak+OKFRjXCx)>
zrOC{@mhzy9fcL7S%lzE_g$d!YKCX;1peEt|E!<(!)`RiKq{F1Fr!hjM$(WLu^Yi%p
zOr=j)^Ll&Kjx<4UpPY%naWB&B+fHTGh4?%4N(T4waSQUJR8>Acqw1Z-093ezU<KWP
z=1dRc;^?7C@HA-fLE{Th+5dN@9SZw&-brG<HM~&sX}IHba^Bc#*0}d<zBOPmKLm9C
z45RUsdfGTzq%PZC-r+-LoE=XNfBxAQSr6HvPk0j}P)Ru@z{b=1c{5C|Gc9)V6|;u3
zA*yr8WA6RZ3;1wUHDo!gjw6aM;}kalHRL2G6$<rmfu=9WzA3};fwoiJ@xKLeSXhK2
z*eF+~{&4R41jH6rIqm=8yU9V~{qyIt{eEo}Fo%vZ!!u|HW`h@2`v$mpm({<g=<?Or
zs40x-v(qP^?;<3$VMM#q5G9vEjomq#x<gr&McHc>e5Z11H!Oyx68T4BmFgqDjmU$e
zGp65f)b>?AoF{Bkn)Bg^fBRz6GgMsfl_=Bm)+UetRSp4=`oUIu_CJ-iW*A_q@{PFj
z<ewQVhniL_zCY5D2glY-E{F}e&KG({MS1VZdB!g2wZ8^xv^;iHRy|<x!ws0kR)`wV
z!aD#?<r9EexuIfqq)FaMtrI1nh;BlkqvKqqH!nz?QnIf2Z2c_+FqKslD1szgbhcCZ
zZ&ggshS?$~VIIbp<u2EjC-oyt8ap!xaGyWv(dsyz=pQP$AFg1(9Lf;|prvbSknwif
z$!=4GTUh3hPO!C)#aIf1aO06Bs17ipgNW%7<Y))bpH5i&qzGQPuY4Y_3oWI1R%V{l
zMw0A)+pS6NNHW`KNU$1V*!4SdeKJzXGBr$&X4|vwko<pZGY_Q~fVBZk^dmN|xzpJB
zJg)%ekh>o~H7tB@C(D72|9;^D=7-xF+M7;V27tf327fC<qq)^(em16igc`5d$l&CS
zRUuTy7^5cF8@3*-5ZM9k{4K9m_W)-jX^=tP%sX4Hp-on!Vri`xp<&(*6V`rTbGI63
ztm~|-r|S;r6Fk`?<D-TTcyHNozx-q#c!8tV(Y|f#sRU1q1%bosrCQdy5Si9sBD;rK
z65^DWts?Kq9C==;Rh7c5?s=S(JT6m&=srBvR+U_@J9C(oB+N`8NKU`MiDl-ZP_Zux
z(SMCNFb<|jg$>B*A->4VKDkdrlcX{?`)WlOX}+7rGBNV8E|3L{Pp@DbO1tOVQCFo;
z)sjTra0P9!-rg@9OL>J&-@RfKFlwy^`#mS`J3z}*Lr-P1LbHKo`H@C)@l}}2G%-N*
zdE)BbI=8yjVF9$?Zgq-(T@~#>i(YCGraV``IfJ{)tKfS#wP95~!wXXYhYv##Q=K)x
z*1anbq8Grqm=~14(;95`zl(qW&;FRo?aJC6``zivVA|SE!FE*hVL{s8b`OpZci_pb
zk<1ll8#inA&Fbc4UU;hVvh_{3hNSE9X8h>!0iNb$e2z2@t&#Lt3Rru0Ma5)q-uBx5
zIB+rnT*^Ap(fV(aSxJaeQZZTwCc$BaA$i$p7Qr-pK(_D6SNdD4)AG2-ru9Iwm75-w
z$!ehs-^MtS_VhYm5!?>%ra?8N`-g#8{OGZpcMto5QBTjU0-IyW$rJMR_j#3QxsDqS
zGkZ3yD2Vd-e@?I-SW50eM9Ls}$(9t;NzW0EeM3;sP1>mnL?E(^su%kr1hn)kZH;M3
z%HfX890dEkc5Q3D<8zABS-As4J#<js3A|KhmwI@0Hfdi%`OD}9O|#u~=k`V-?m^L0
zOZnAi3cd&Rq}FVDWk?2K#zNXm&XVXzh!ernfFy#Dz!OXjR<6u~i;JIjd^u!2W>m|B
zG<lPJxJU60S=dExb4AfvcXU~QSjmiH8V?bQPVue*OY~&&?T1<H{}o_dx9LM0yq3VY
zfG1#xzBl>)GMSQ^y0y_AU|x<xKw!4gU=uH8HGFCRBdNa1g~-6Cl%T8KoY7M(n|(gG
zSTz_*%9|hJW+`gHnQbAHaYze0fMQ<b^cS!49e_|hYf-7_5sxh&tt4z)6wK%p$&ln3
z6JBALJotiHg*piv81N!B4mJ}}AzaSe<-YxW?{a*Z%@+6SaQx<R4cD7V$QE2|I-oyH
zF^m{FWG3BPh+}of=2Qn}mnXx^b1i-q-iwqqeK{!_F4+yVkDImwd+rhsb-`|tMOk3K
zisEwO^%vy>t6#0x15WY$M5!bT5Xt0fei9lF3b?B3!ebEw*Zw{ThT7`uItGd=Oa$@+
zCXe10Rk!?-QP~MeV83~CIDCzv*T^nu&~8pj^5AFN4Ywnt*i!SNrJLY~qyte>E^MPq
z-f;Bb?y12z|5bH_#BUHpwe8-$L){Vzu%ypeEM@hOQd1!d_owxTV-+_3_12fmeU5lW
z%s0~mR)eva!P&$x6M|^U|3lx*ym$)1+r*&+yQ02v1LFI^X5A5|WuWU@86B=Skmltn
z7!fEUO*<y~tVuO5z@&X8KM6Ob8t@jR!p*`Og+@soK{lT#MM#y=uUWNX2{tcqSugwz
z=t`ICbvQ9Z1u3TEA^HVKmza^|5lKl*vP>1H#?cgQnyooC8)rUa`Y|P1A2-grzac5S
z4~YNZ^j2Vd7`@CgE<;784yu=j<%KIvIkx=LfZD(Y9z25ILBTmUQT#wPiA$5kYVWyx
z^R+fe2XQSf^=t<y2nR~vSo~YX6FmFPaZI_nddZ>z@E;;}(6c!^?OvMsL2)YN#Yr>m
zQE2OqWnVj{zM3skF9re|Lc>%*@+c>FN0K_TPhDD7pnz?_evHqj<Vzju7Z*iiBhaAk
zL#_5(|2p5H?5KNn7b5(7Ty$V%&nHc-5wd7iTgv`aN5T~C??2_Q5XP0`FoVVBe0%^r
zLjiHF<+H13arGWo36BJC(j0NWXC<dCRjH{WGNmNR)x?pk{r_8p=2ENubO1b6&BdX9
z9;73}r4(FJt7DO+;Aa2C0pKF^&ttcapsl}R6v5tJ>zklS^CkXk!Tv%8R52R9+bNO>
z$1mbqi>&Vo+{4z|*^vf=0+0h!qllTvKr{59$UHB#v?#M^GoC{M%OlCb;khmP9y<8z
z?E-dCO(uGg{JpE8WKcj-5tj5{Sr@i6_hOxrtjFQLDm;4T$>%%n!xW@~lIjdI=NBhg
z1KRSkclWY~!O50#=rDCHO-Vmw|I3oeYcXkUHr>6FQ0n#X3Y)pzQ&2A7_IWV1uRv^(
z=OL&81wrz!nyVOkTQt*?)EK^ud7|%ivL+d+^z7UIy0%5o_<Bos&~I-0lj%3SzFk*w
z;(T58msl5hd8n>`^BYI5Ehc8qtq$|`>oiL$^$^h&VNyd2k}B=C*)(u^a66snx7&yJ
ziKGz+s9*2-+E*Dnh5jr7fyU8n!0jA*v^*lg^}n+RH@vmy*F=i)eXCMd(c@>JhR`q=
zx~ZPu9O*EW=GCvk1RB`nvpwf)Z$f82dRIfD{4E-w?%y_h+%jRBR`wcJK{KHzD79bj
zx-HFoZfXpmc%?Hp8fKOGsg#;xQZiqugutk%YP`(c>5S5ff0UDl;(nqflQOyyuJ#J9
zxBly8z^G&V@h!loo2KK^>So4Y^RHdqycD7$%=bySu_P~1MfEKV|I6d?v^G>W_u-Pk
z-|J5hz|p_o*`p5ks;m*ejRNio+gyz<_nZM_?a6QD-PU{?Tzc4n*Nuj?EX98>J>}oa
zT;jMIbUHT`zv9fO)2F9TJ~YNQiiJUVy@zthiPm!Qkqs31^iTVG$8f^vElNf1H5$HA
zQ57=~iZ(NlwecqGt)-?XCfQrKM)PKFy$-}6S+8*!e0@mWeq&Qg@T|iRcJr>Q$Aev6
zjsmjmUF#3*&4>X0QDCBCDvc+Kbw&Bb9DRpc_z!4uNO#)yr<%T)5=JVYgzB#SnBP2p
zOaGCTy^dG@?!@gqHMr~z5wqPN1OPjk@gh<)GHm5}S^22{;T+X9lZo#33lP!{1|d$p
zCcVH{U+w9-x9;fr6eqKzG??LZ!+lwOgjA(5ITcDiy)gBI4XB$PDw<r2^t6;<1nHEW
zll+U0V^Nl1Fi0?>#mx|b40>TFz3mb&ACDO~hSX<TB>a|OVQ#%}?eAN~#{87Y{B%PU
zLD^M%$+WpFLNF!od%Uc6J}j^2N0<D?91Y=D(rccWm_Oe@;PI$s<hBaQ-tB{Nef}BS
z_RlG`({zfv8@HXq3!OfLCHpY63_m)1*k<pqt%N(rbf)X|p&Am5`~YKngH0<X9nfl(
zQI*Ie>%w0cUQepWflc;VopBd}r)QPP49YJKqE9gQPEZ}_2!Z%cE!lDCb*RzGxKf$=
zWhV-EYVI`cZ=Id_=dNgn9DKre2mdJWuz1c*;yYe{C%$)LQUj!mLNxldLnls2KeO_k
zMPS%4;P*E#^nc*&xeg=PT4`lRDv#d#S-;s9@ya2M8wkX3uon#W^CdjowgQLhIsS*d
z*{{5lX_#lQHBJuor`>N~b6ix9&-gN`^*>PoR2W(S;+y%R-6eeMx+rgkhC|e+vY3~V
zt<Rs#OQ2o8wE6o`zfv}T(=X7fLf#jE^kskoGDMPZ-7?^a{xTt74mpp#innM#Uz`H_
z$V-V}Pp;whlj@CCwWXE3G>)=3Sy0Z-XzjufS=$wcvfiteGvDj+;5U-bjWZ9Io!}Vh
z32Yga;AF*5ukI-nBW>rq&~hWY+anXp&97ri2U|Uku7jf}l`kY>=PxEbZS6<RnRE~J
zkL!D(q3=l$zi>}wlrG7(F|vkBCS~N%X!DGh9e*Cq1+=B%5<w;Mjz@5ESWDiOS-kZ>
z>f%RL2uNP=NG>qZlBb^b%!HsXz>r&ZvwRgDXf~<KrOe2V7^jL0^CNghK8=3oxx_kG
zcaXTT;ZS<-UWa-dn<Q|1m02sljmcwFr;RPtR$}|_$WMnKf<oK)Il<qHk>wXP=x5qV
zH7s?xVyftij39E9hf~r4iiqd<F_O(dB>#Fg7fX>+1kIz^8K~EK!mitK{;R^uJFzA*
zNerI0K`(5IFM)VzO4i%{%&4vI+Kun#1*vK!YQ|IZby-`>lj?GjNrhS7KQ>HUvgcY;
z59k7hPd=R<dd*QZFk!pbIYkx{-t=8*wrX@PtS^mWQtQ1FTr_Bl-k*D?y@dMcyt$yv
zhg7&hB;fksYv6phI_AoXjQ853hWfpNqxE%@NoNTjV)kD|wPTmWk_|(YQ_jqQA?;35
zdsruJ1>y6N&wkC>Ya8J+bg%nn-Zezp_4YW@^;@Wm60B#uzhX8_ykX5zjz}lf{*+;l
ze!IJr6|M7ou$_WH0pd(WxwVkwG{ZF^DMXHV;fcZ*-Xp)?7>ue&*yWI;k;BU=1K^QE
z+?$Apvs>bGwr|2x>$j?)N-xmX#h*{EaY<-8uix9oC;2a($u8vVh5!8z$>8$QVN~0Q
z=6fm{{qcP=zRjECD6a}gSB;40De)5td1JZ=;`|eb)9@wG+9ELInijVT#cfVFy7x4~
zy{O5UaHG3sr}Xsux!nuGIn66%S{OtdBalS(_bCpI-3Lb+>1yxudVi(fn_%E+be#YV
zO>!Bua=>kS@*L(@YBEMEs||L;ee*r7lrpKkIDfyQbz}N@AfF{-I$w<5tX0XS>Np>=
zRX6N(;t$qcfSypKcA$Jk@t9^rycd8wR?t0TDnU!@@+*yNx;GmTQDIX(%Pc?ykDPgU
zlzg;X&VHznaCeQ&nl&R>19n?JWR+WX_p5!O;q<(_ty|Ac?X=jH;xJ(4q=hLDA|1v}
zJJ-f61{FzY`%m%AhDi1BTQV%_s>*eP9s0OhGZd=4of@4mNK?yI$$Rl$lDx=$K?YVN
z#!lmAMLPz5rnxESyVe$v)+HH_onC8+rA4;QM-e{w|Ep;BUmUt2T^U*qa-G^C_tC8@
zg49<guhNe(WL%yHLHWpYLc>wr=FN}4t1-JTtuAd@f!gweJ%&;`jo1vEv}fx{n}CVW
zra)g;)_P=Uzb8>usI((q8<jSF`y3u;XIP;Xqai&Uhme)$*<T#O6N*5MofD&)<=F$0
zGD;SOzb80kR)B}$?Y7pIH}6#^(^vXu*5-I7MHDiEoH@F{Ya-Mo8i3LPl5t_d-1h&q
zFc}7q0e~$s$`I^}I?ds;n|CZ`e@cJD>3uK<fe8`Hp2mE_#&IhA;`%JVhTOi+gMoUg
zTypfn8}Z#>!0NO0j5;zf3(+?+j`He2{^8IhaOIC%UALZQ&sl10F4MSMu~*kPb#OMc
zWRu&u-M#?gHIkXf_!{v~*LJur=R~>zG5Ew>WUd!}ue|<$kf;;B!?{p~lN8L~Wf+_g
zp1rMHJvmh^&-qSYya*ml)N$ojaS$$~{1<d#vgpm4#!V|vTdP20>7`O6yu)cJwp$65
zeNfPy=%D!j(*mG3?0;TT;i*2__qg2Rzlwc8^t{wh3)Sp`Cm*%8Df3}0KaoD#^?i*j
z{oC=7pfYaZCVWG!CFzma4ceN?3E7WK4B-)Fi-}N!5YtxOcaxLszVnNTtr;{_jG<?n
z4(C~#Jm6Ilj{}Mk#;>uzDWFo=UGWi;cv>>*%`ysAW6=|TSpqP@7QBBx8h&Ox`a-5-
z<fFD09Hhy-9qC=A(c<;g*t)&w)s(|N-Z<ae#)a0Qh3Eln13S1>=FU+lHN!fa05{9n
zFP4_(0pg1w07Pce)p<`V|6f@bRjWzgb!-h&FaJ#`sXopEUr!?C(6kOUGG4z-fc~cs
ze(3`MIHQ&sy)NDE(R5dAqRtBd-%`5_nv^u}6_A56diuu3yI!y(*)T8c2^6V<WET@d
ziv;c<O~PiM!Ex*N<s<O6ds}y21!Tq?1hq&}&vz3Nkq$!nu%VT|ey(Px)PA|BO`1_r
z*V=4~cg0U20ZG#p9FA{oHuFw&Q;I+Dj3KNKO_TFygfqsj=6iKUq)@R^HHT*;;jC>=
zk4vG$Bw*cB$gjGH-Oe9;_FjMcgq{vk%@g`|;=dU1@!vQOcn;f3R_9|XB9aD>g)%_>
zF^Qh8`OS`LL)QT+rIy_M4@TeSG@Ef^?!p~ps;#XZ)foYNBe$H|i1(L(xOY!_Qy&_V
zE7BF{p52!6Vs?Ny`T6+i7os@s_qfG4|MJL~V%0DZ@)N!mu^<Oa;P4TIaiaI#4m^ii
zF(m$MMnq6P!ifDhZUbB^3o@wkxhjvH)+HWy5{C50-)Iv{YONp=wO39x!~~59_=@Br
zCe@!yEm2Mdz-`Q>=1KNf8U8ZA>v>P7t<G(+<oFzJgMgYr7(NuN#m$caL+LbLRda(z
zFo!NlRw7e7gsRMhR^2O()rPMA(@3VV3k4AsdULh2zrh;Q<LA}Y7wk$)uVAqDCZ7H7
z3pY>ahjGvRm(w}k?t;`oQDKu`h+^mi@^Cy6L4R5sK;XMVt^bEQKwqZAek1||b|DiV
z*-rSL7u`mDzvc*lq&fS8DWXpwt~Hx(lPPx7!5?h&zf%r2x4Y^$dC3M=p)YmJ1>88j
zCRwpp1Mgh7aNz=qcQT#qP{GCe*CqS#`Z?8coQ?53(j>}tXtFF+*rjE9jJE<be^M~b
zt#IrhJ>Nvn^Y3rJFXF#(8U_u>>EU`(>Agj*>o2MYgx1KB$s>r-*Umajpns48f<Pyl
zBZ52Xg8#%>)v`g~mCn(l$FUZI?k_K?Xi*#l%8@{E4{OfTXS~6V0N1n-WM2G(f&FW)
z9gY34Zh`q99nd7Jhp>_9A)BQNm;0}pe{UvRBx+>yj$gs$m*VA;$n#+R<Iqwv-5FcL
zmew1CB|GQ3rk(<jOtr}y@P3U^97_MI-SNzLc_}LHK%|?1E!WORj*aBtojdRUOyoa;
z`~Zu(JsE~cH-L!1-17fn+T_Zp?GW*4fEr)?JJ)slOB>9ErPT##px>l22ZW95^Xlit
z8$6`QPXDE#K-Dd!FQgcHv6!Wkmfurs3%F*60T2}nQOwhc+@Qs*$>!yoC#x$IgU!Ew
z(L>$9dAy58d%bOsWPEQhG+qK9B7=f7hb8+I2}#Z30X2V#;p+mX%x!CDTYe|@Fu$Qc
zUn$@2JIL+Q-<MPO-WH#<xD06!&{HAGBQ%SO9>Sy!;?PsOhlJGyhq0``Q36S<_CUW5
zx(k%WyKiK^<4`K#=Gv9NY^#Qgp6tgvBE5kVE}<L1M)22MGgk7BDUnTPU~N^2RL0G_
zVpK(Y@E<~}oAC!xScqhfF8=%<KF?kjNr0jC8StXHzUsdKgl&Mn?Hd2X4k?0*o$(it
z&PU7N0SvycLS1sjU<PxZo2`YQi9idJ4Z};@6)X3OsokOjJ&S4CFkCMU=DSbD;p=h9
z{&}VLNI;riKc{_OlEN?sM2=wCRZF=85moXj1}!<49(Q`R9g{&z+Uz)M`OFvFYww6M
zMhVCOvgpe&IZ;7|S<jy(!gR_zL1fAo#!A4Rn^%ia@e&*DfWP@ZVF5sdwF)Acc}Z4#
zEnI$w<rZhO81eC};6a6uk_H+QPM&fzFwvQ-C^wG63=GLdP(04ry=8Vw%1EjRbrXY7
zpVM&uKdRm`D(mfw7MAXm1_|kI1nEuzDJcn&?vU<Mx>H)ZJEXf?8U$&OeCWQL|2gO0
z_jL@1Uvv!J?!A6%%{kYcoGWR7o)$rJGE6B=GBXKYC^wNR2w&p!c0USeDZLt($ry6g
zSBANFdx-G-`A7s#+2zOv(j(R-vU&DOrzE_&TKvD)T5sR1S{C;l68Rid0KN9}Z~4Y^
z_8It9yeJ_m`v2;}GHR+McC>DjbfD=KA0ap@9SnWeuotoYGJ~e_ls)(-i$U#L7t0cf
z*R6+7K`^hn6<BFmnLO}S+=DuGzS3AOY65KmfOnT$3wl3#;NL8Eg6tiCT^#}XiN08T
z_%fh%4_%J7h^X=}WivxT2{E}6>v|bRXm6AJ7gwMy`efHCsx=JBBnTTLa-%eIX*?}r
zbI?RCjrfzb)=BUjSHr9D_vixkzr?|#u-9W-<peM?U~;M|QXCr7t6j?b+j<F%D9&;1
zZDb({N@XMfvAuEth=U(?Zcug#9_W*{x{YZppvE~7Zk{4r0*F1H&ci`X$kN;Vh~m@b
zlj<gHU+m^7w5^Zf#P@3~#lpC8ANAielydGg{eN;Ryj%f$wo>gf7Gdu-KZLRQpJ^K?
z2~;Pr_6;v0Guh)T!#W(=;yuc)L(5<&%G4a+8T(D%4;1g*^7N}o6$K$18Pq3DGLny>
z&1UF@1S{q*G)J?|F*e%ze6M(AGyM)Nz*v$FE-0vj!6Ym#rxKBxt~UE?R)X2>RS0ah
z{{TMDtO`L*+F`u;aJF-B9ibnNm3gVIbGLokOULJ12cFNKV!fJP=f)qu<F&<<r>5d2
z&ff%Ib6yQ}0cGsk9d=kAp(&RtAoZ0m;JUY>rJw>(*i{fAT?RO%wHtYhQ;H~54<ny^
zgBlHZ#){yumE`2YUj62N$S~#7&>4(q_1U~ZAf>?J6N4KeX-T6@|9&m?px$Co9r{)j
z&ETJ&wV&?;%k_)n6_?f4`)iPgtn2OB^5aR#Rt#h$J8t7`lO$EZ0<BpY&S-8HR@psA
zdYB_oOuvZv2HE%8S!{a0x`|M3qnx_@n<2b!Ez+!Lh=MSPp5_A)6f*^+YlmB@wqt!H
zW~MUqW;?-RI`DF#QoHtzag(E{eVdzopQ?V<z%iw2u=~~v9gg_=DsJ1~nB`@DsY|x+
zO{P$e6+z;e69>I$v0=NecqO`4KWH>Io|{zAyNDmQ_-yud;n%mO;EzAPn%a+>Eqa(g
z!H6@LH$mkIO7(l>c}<5L))P(NtH5g5LW(WZ+1VRN^^TAz>g2<`kc0I=1oa8*p#ta9
z=)&p{@5BhAU`XA`s~VsF{DY3-wzk3biK&;}1{}S4(=MDf$M7br_XXh?Xr+4|TmQ8k
z-GGbqS`VR3B>Vbjkpp3x>*LN0Z%JJdDS&hC@Kr^WF}55qyMa=vLjSFSJmvfe+U$DQ
z$);GBSgAP;U;45gRY}<%AM&45JJmm5-X4&!oZg=zYocgOC~|QrT3Z9u=IdL&84<rA
zsy8x7ZL!xGJp|oiIfRwWzw58x>n${Y&(1q2C}%p0%Co#N*oq)Ip?P6OP1vdn3VXGS
zka53(%&x2eQl0XY9I|$iN$vr3OJt44|2`GVnt0Rx%i2WQ>)#U&@3-nRcbJvh$auZ!
zGsY7uc5LNYX-7pZ!`LxpWNlF5<KsQS$2qoZYLKaBB`sTS;8WVP9P(6Gxod#3tRIg5
zg5H>Jg+cj;i9uv!qpBtvW~RB(`)PWgRy=|g5dPNhjqMHj(dZBP?|nLUN^8NxXAIR&
z+!mSW_VX{M8ZSUL4wo7-fk=*4?UWiHL(1wE;;u<*YZyJ4j6lX=D+;~Af)J*|q=;{W
zUzGf%L-!9(b`Lz2{ANPxyuu)CzQ*(*JBOL^vjosVZw6~vBwK79SmuB3`&4Nc>ecE2
z#zvNjSq?RbOl@bSQ_GAtdc(`bCS%L$ukFdV^xlmRt4RUPFL0+fW>&@_F)5jUmSQWw
zKL_unc$NTkUh<x?LS4<)&*m7~!CDGt#fsd(4S^5Om<t`OfkL}lEDpAk=23CGzdnH%
zE5gh*`^HNwNdE=d)&~Ti3_WzM_aIg&+F#SMx=Y0@K+4xw8Hp;<raa|u{$>ybgt`rT
z7uTYRjCmM*jKS03R93Mn|B9L*_%<aw#X_oDyx`&j`Z{V=(&4E<>BjA!&n(G~!oqw}
z)ZaB~RyU+)zj)tAteL^Vmk2W%85CbYbY{k29v>~c8U#^2rIXH-oqhm0EE<!35b!OZ
z0>~~JI^x2R0JX8G?J%qYA4>s&!#Hr~Wc6BeQM@IAzP62++h7do$|*Cx((Q`T=xnep
zQAScc^Z=?8fM#9XNrsofQ8W9geU?%s@XKtt4rtOhRvRb1avS2=zQXqocnZ71#-Z4{
zCI8N&4{-9{MuWo5X@DSIw&U$_WRmQ}H$ESoTpM8IOgs`r2(>}jBk>WV53wzx&_Kc<
zjBe^7s9ATvXxcE_d?O2e;jOvwPN>WP8JY+kAeqgHrlEp?ORYaS$fd;6(_%LKLJKhi
z6Of(<!j5EEB8E&s^{{%vB20ngj7;GAG_v7Xc<s0MUp-L&KGT24Ie89woWrIQrG(JA
z5nl<0{5yFXHm||7ya9=!5V+IlFv4f1^krl_NIy*?1KrH1sgw2CpgMFfwRmWkCWkdv
z&T61F4DBI{PPpUtIso&*^;#K3iqBe?hkrfz%lOUN{I8uxr$>E_G#{?UyVFI>YX>K^
zpeh;qLb%v2ZM-S}?y`gUfaSi+;g7EJ3qBcvQfp9zvXkTcUWc7goX-AhNAF|a-Av!;
zA3Tjy$Yc_1c5K;hK7O>#{tn0WTJp`b8nmQ&9&~8rMq2#y=l4?~&=+0I18I;PbGq8}
z5OO@uwRgvdrc|GciY`rGZ@z5r^)e_b6%el@y%2Zqu<~STStJ>nJvB9|QxYS_rj6EQ
zR@kGU>evFoNI9BYwGJfBh$T6~QG*#2)14Sa>(acGYM%e?aZuZS>ACvBZA~W0q&SHv
z-M=CW%Pm|38jP&*%`tobl*Tk;e9v64egJecI8Z15$YzjAJ##O9dmIBZTpK6@JG*FN
zAe%X0{>4Y3gbbEPHq*&EKl&<~m{I@K4{%|p<_xycwyTUg1JzRP!&yj)_t){LR`eRN
zlpx(}PrEhddz%Mqr!)U90u>{mt_eN%tD{<E((h2L;{-CE1%5Ke#`M_q+?!NZIKS9o
z>X*fSIizkfhDd7dLhDzx$3lM`NtB{l>;~tal8>9tq$^wD+*1HK3R*^Y*Ur@s4r!cS
z&+92zbe#Zy?_<OQ{A$^VK{y;r2G~GlN+_x$xP9}oYz7e)F|0D$GNKBvs7Gb~1KazH
z1$G<`?Mp;~LCu*R$RtrjtWp}Wm=b28(bcfR6>*>q8EQ)DO}y8jIH#a9>-j*PB+77l
z#E0_XIh&ef+EjXT!=j*2zu2#|)F86iW7~){Gb{;u>7U9O>3FGo5|b#G*Y(O}x$`+S
z8xId}zR8gmWC*td+2^p)5pby!0vm4-X`gfH;V0+*S1zSZArTt_9}SrRbDOOp#`%V-
zDK8F(fSj4BRbhj%la|E|-$&i+)#}WHDND`QT^#?4CF|vnKNHQ_!!hxwsk32MU%JZI
zSy_Tu>>AHQ<s`jqAQejSv%t5=i|-sLpmC6rv*^8;5d$E}h!baUgfz9Xns`)aY?K1?
zCjG|EAJyB%C_Qfs7NIUlmnV{_mV2IhR0*qcxSe;tzwLR5o+rP-OTTZU7s(>X;q2FP
zu{dz8T{+BjOd8*|v`@Q=^Ed>3gKY9AXR+n{B$PD+xXH`Nx#6jq<7T3(ipM=!e;mle
z%Rh#GxysuWMmV?Ckc)}ig13gvu9l&9YjL2yLTyhyPT_AQ`YHS6<;$5h!Yu-D*|6x*
z>mAWukSz@RwtwEGJN{Et$DZZeC?~0NEx|grtqIRDn<HamkVKFhNJ!@6Y~2O2zal{3
z@hMaJ#KHiwQOe;KU<?gKkmY+ngloWb;c|pTs)(`83;`8c3~u5?908P^$fmrutRg3S
z{OhP20~8<Pdra}{oOE;aFMzKO#O7c=c|m<gf0eO|U>dhPp}=w3fNk;qsjHzJcB@(`
z$M8y5;a&dbh3}xY-N>x~JwXM$6Gu1T-?!a6?5I)-Z^5pLr#u<wjv6Z2y+E_J{-^bv
zBz8@A#rr|~G|s-~3#@`c-EADz!CK=(v1PC{HHsR-qse%y)hgDZlH8yl_K`fd<MjkQ
z^5ppCD01>U2ZpK~Hh=A1`#~8A8|K4YgBBt-t+Nlh(wnk?gi*=?@}EOBP-as0H*sb$
z8A3u~8<ZZU6*3^LW#cU4A1zC!jwD%iK;SHn|DZ)jo5_$ugQorJzoTIq=axU~q}{K7
zr6?|4sX?IJzHD|m-Yx?|6fND9ni%FjJ`BYxSJ$8Yl>gILxVqA8|5-@3CE2sfjVt4|
ztxu|PVc!}w={mZJch&}My{v=0o^M^EN)}yEXs@244XSwzXZq&IIkJlNKq>V!bZIAY
zSnNYSd+g)se3z!mO2xaDR;)=x+!$K=yU;3&jwi$8v2O0>edtx6Ic*R39Zn2xC8U(%
zSTp|^Sy2fJa%`wW6xSI6UwkCHip4@NYwhldiAzd4>mOW3M9RVQSe=ZEp&Y^VBHP8N
z-07;OLdPxrb+Zo}WofywM0pH+YO_1QJ@w;GY3r5yYFgYy`KsREt0e&v``kG8>A~Hd
zex*4!^@{~%6ap4E603YC2YCRBd?o!6!7ysJU)mD&9o^fnsv70!?byOvAE#;S?oRN8
zs%}kD-_|rPO^MK=b~xc?z8Bp)7!7FX;Wb<8o<$pU3W_&)j7)uU8b<H53G$=LV`mnt
zjdDlIC$VE(?v+b%K&c<7`{Oqg2spWgq|Lsma1;RRlI)b4_%}p7;62$-H($FSrj)uH
z^;D1ezZsj%$WYr>A1HUbKx#jh{U#vF%BaYD^!DFwuIVH<v=R7fr93CqVtB5>fG5Mj
z$r&)4Iyfj>Z?zy+q~@N}yjw~MIN4Q)k;j=phpgmMbmVrk=ejLkmL`UwaTxxAR8QTh
zGJ#`y$knb+Nq8D3$grF)8nG6R9xjHc)M_6$m3Nv74(#4-%S+IG3-eb9FploaCXy*`
z`>N*O5um72nMPpI_W91P=XAeGBI9%IX3g@r*8Yx_xqx$XA&&nmN-Sf26%?GBm2hY6
z=uLkzLgvZEFU;aDo@)p`oJnNYofUt0$B5o!@cs3R$H>#g43`oT(=3GwlwfydOWE%f
zf7e64#z<{hL<Fsr#bgGBSm^3M+&3IP>&ZgMTw~obc5at|214JDV)AK!YjKdc#d~5H
zVVf+`a~`H`@tWqn(VVe7G(dPKsO&j+hjnD58&Q{-r{{;8x@yy!iC_NPv+J)V;mHiy
z6RSWTkK5E$PZG8@LbK0*&E6>_7x_ep3K1Gap!R{E8N21Yf8*=yDpTwl6WDYiW)c`r
z#jYxB4o+VttoSL_{H1ac<3DF7ITj{V&kqyQ-+&J^sa1aMhA1I7JzPf6yv&r?^dPgA
zVWpU|2IjTDT-NR;9}g#@A(mLQx%`}koF92y>{c!LD3j59I|aL&$^Gn4RY{YhiURa_
z9Zt*!IqNlERG)a&NGu9|iDiPUli?n(24%TK@5(o7q{0}L8@$2<;4GDLvnPJw(V;hB
zzfB?B>`Vy2NJ8WA&h|hNo_WGvI;TL5_4@5BL?-yXz5Jeu4}ihJ!=X0$r}3NBgUIck
zTd>4LP=4`T7Pt&s9DQA<SHl7d#NC=&6hWVTICn%CvtQwF!b3*Z<b&|=l(cKBWZ=@(
z;&cL{)_zR4cUpbHXBC7%Nri41m9LPr$1YVo3jU<T|4}CS>bPkCeafub{l*Tm^FIfi
z=C~zGxb`JYb<ZR<AwyErL&5GU=Cu-#x{pfK5Lxyy#x~jXebTSVCauKAr&=^XrK3Q4
z=p4$ZW!vmqXABy049BGt(UCV~UMUqJ2x09^7DW!Oosb~?tT7?S728=S*Y?)@M1wTl
z0IwiM@q?@=rJ64L$2@HJkuRJ_N0JB4`zv6x)ix`>{GRytAe8NGtj}2!cwS?-wb`T`
zS{F^0Rcp$OG-DK){vWqr_|%@`<%iuEE$U~ij&9wo+>1e6mf6nn7ENNi<cO%rsMlgc
zkntnYG%`&W!R^4T*v?&?I=(`h092pUFe&7{s0Ib>;jamKH1qdtTJJSqPANKeP>!L*
zE>H2K>WXU;#N+&eAzKaTi1Rds1bG@?o6aw^e~-9f60buRU#iG@>jm`%g8{5yz<cp#
zbZJg?si5N+Kgq_yjsB~dM##Vc+<D|&q*9&rg~H<7oeJ;ry{%yF%>cpEpV`H%T9Moj
zy3?an>Ax@;<O`{q>d#dpFxfN`H#MRy!-DG6oAIE!Uf}7|1>y!eRLW_huxXG7N`CJw
zCim-zi&Y^<iqH9u;Xgc!u5l}rT8{^HZCdGDzbVf`#P}Wpg73q-e<YpEjo(n=uRN|c
zIUOA%g&k#OV+Q1n(Vi-wrtV7XyWhUDBiRXDcHJ(e1!Q<wpWJfmFWNZ(bfGTL9BW+*
z_Kk#)A?qaz4`i0L1<ffuAkAJAScizomUEkIksm&+nzl_B%*@MrIK5FpzP0xsbCenq
zdWjqLs+;;=8s{w<?no^{khDu=JY6X@3-)XMb=hG#{fR`lm0wv=k%}9}nIWJ2MjyPm
zIJzy&Vealf3TUp3hA)cm(r0KD+sm$)p4o*UXxx%mm`I<kkP&yx%sPONg?p_}6uJoi
ze?`~nH-f9FiOF+n%Zq2Q@TR2wwj{3JXXUS6|FfC)i_7EHaS&}W{slk%%YY^uyns@N
ztVTNnJ$pWZ71>O1>4d~@@>GVP4i3a!WQx;D9B4?_#2}ZPh>ZvZl7Q%%EDeKDpO|iL
zPn8t1D6=w^_8PrBb2O~Vr?Th9+#>mJZjWS2_0#;|XU0i2aanHO8XP1)>;s|@%}5lD
zJ3epow2Fh9k3tQ98?qde^K80;R`{<YQd}}>p_^X3ZT5)?d6$$q`S7aWjp@1MzRQB;
zmzLAz&@g<mkXu7~9KY;WBwF?w7+Zf)5&DGhv+9al#BH_rdw9C-<U13y7lQAfA3h_y
z&Y_vDF=O>R6CrRYD`~j|0w*m7rL{bdJ#aluQuKbbqMClJ2sd7ot4787-#pngRCD3*
zQ(;>6Ix-#2ulW1VE$NdtZFyRQS?4n4TVa1t_tyXUCY3u8foNn{pRS4wVEWf^xlP##
z=0tv<iuXCS1nu{!{=>}h$L=1Cp0k%1PeCKTs9N6+mbvF5!Rkg&FHhQ;P8OBln}IkF
zjv}6bwNh*u3QRGB$a0GWXt_3Ms1n&oT>pVM0$98WcwTMmmdZq`H?`AgBLo1GLj5ci
zhZ8J@A5g!Dy_34E8D2#|+33Tahl?Mzr^7m3dg&-N3<dhw@YR435yshbM{&j~HE|4P
zSeLvX%Lz)5hOHe)XNbN}p~9_wNt7_#gX*U6*(_&U!v=5E7U&=ZiI_cf1xW2XWqLF}
zyXjr|WL`Fn|GkGHYOKECeq8gL<E%eV@2`y+#=!PGNQiH{Cl#Y!_=PBoGVIA}l=pNp
zKPQXg|D#uOV;K`a*dYs1P*yKlbds~=o4|F<rbo-w2{Gx*WjKQ2?VCCTn@#+vdlvy%
z#;R?<nfPZb1F+LUUu8s<dmww;l%0+`3u9XfcR<2i<w)gF;_|O@s(;0eZ0!L=IW#zX
z_Tqj>tA5B3^q7qle!3aYj1%JftIr${gbB{=BAMi90DKHbu>hl;d%UMCU-UT#4C;DD
z8_u<g%B5he?({o&C>?+@Q;Ep{KpS!RlfGkadMr+`F2+$hahP8x8T<7Ta>xs=M7{Eb
z=CsOR<9nY|HB{K9<-ql}vXB5Qf8hQI&3F6qo7-%>tFrUG?SPh!pn6ov(8+3P!SjwT
zQ)&G^TN`iIhS^NXe0ZIXGBpx{pZ-4AF)ntiN>50lI1j{6jJ&{L&_=*<w&WEhHA`l(
zLr8<f!nx923)+HVer&+~9i?q|O_Y~%lN9366GX8Kq5a#7z9CZZ;lbXqe3wg*YfivI
z-&afuZAJ?R5ZENkpL^t(I>X=ZcM=+~7teFSy4+3sCZzZ29qsgiEzNSxFH`O@LV7pd
zw|d)CAsk8J35`xUBY7`zos~4^rx4$lH>Umrc$~rV9;<RKJ1<#V_D4u<#2;GLEiRtA
zwBM~bPSKh^pu~T!#+wL-G11_+Dn#g4>#4`v_8`o_30L<7!WG0$<;|*jMl%{BBY1P7
zV9Fr))Q2=1ZA|=tc=SFn&Qr}AL`bqHXXOV`5Ala@h2?U5;!Ep^i>;=mcMBazI_G?c
z9sr989V|~Ohu+2lBhHLG&vBqygA}2h6OosI?f^&3Kd+&K(Yk@>r*(k8%)YvFcg9}w
zK%#_**9><T8#obnKvJ|;6qw@d&yvMg|6R%Er++&~V(_=q9}IC9-%dlvgCfi@QJqP&
zW#p~@(*o=uY(MwVM3Dk48Q(PB*EZc2qIQzBeI(6d<KDcz&8oZ1dy@P4KudoUS~~FU
z?5oQHYJw+&-)nQ@U*aLLm~NrG4QAFcQ=gdcJ^z|m<a=m)-4($cZ|Hn@5->8(>@Gu9
zl40<Z&Vr?+J`U`i{lC4N^KT!HkKhM?BIFlJBAHg<e$38WC86(M<<6(ls3Q61hv@FG
z^~DUcl8B4_48$ovFDchvI@SObd6esGDt)U}beky2Xmbe`4x-q2!2}UCtQfL{;pFVp
z2-DnoTX#k2@1BhF$R8mj2I^>MH@7I)=M%$&PK1`s)IMx(hdV^ps7t}(ETpV)c&SpY
za;+pBYRG2gcuZFu=yg*@&@F9&Qk$F_m-EYrbN+Mk5B>-mpN>B6PmgmgNyGQf9>HqF
z`H$c*_4-%Sng<FoJ5UA3LZaKq>(qH>VCs3q7_vXtu5*&4Ydz}s+eeN^n2xzq!~TMa
zn!4s;`ec3oHlgi<k<!lh0xvS3M`tw`^pEL8_nu*2eJ~i>KC3csHaiMlGSmJN9~_JM
zrV$ICXH}Y6?WFF?&5-%3tw2h)um{G<K54PvC;Bn1T@x#dzydF{e@n2NK-v{O5e*H^
zY~=L$=F;g4$o-!~Z*g()om{0c3<UZ$|0<4VDa!gc&<uVO$j^+YIjk<P_I6hcb|>W&
z8GvBQW3ZQkDKQgmau7O~8p&C?Vi3OA+C_9}zBL!p91UAbC{sG=cYMkio?ePdA;I{o
zu2*%-{&;;tT$mqbCL-&HO&-)6=P!a<UuskO)yhgDJRT<3H&1WOi;*3pl;+&=9hl{U
zfa+tU_T`a&NRus}6tR5yv+F_4VR*JO*0c5yXrCL|;ss<S-MQ}IQrN?j1cP&Yml3gn
z)?G9gM@)_==948Qu$){f-{&<ABOXt?C{8kubmF)z2HB+_GB4d+q67jgemY5Asf>z$
zUd(}wk`?B9e^K<6G(DGm8eSh&<5H^awHh=9Rou-o;YtOki!cASk$XyZI*wCy+|Kqt
z(n$L>+AdIZTenj!XCtqURy>9-uU37Y7Eh-$E^9$G*!lVu#~-cEcSRNVQbA*$lY5}1
z!L+(-HXHMax5@f%_Im6o$I@5IJ_ab&8DGYN5`{Tv&%GF}F!q4u=4=&$wWiZ{2=QT5
zS}p4m1m?c;o=kZYj?rF`lbJ23*KEEy&2H(B`cax?jcw7mgD}`YKmc=7dRj>FxEgmB
z9^H^1kZe_17fR$aOaJ7nNe`dNHQ;6`3<{k{B0vERpZI86ke*`0L{Hy@BAR?FYNkJ$
zac92!fZZl#DXd}PZRlY)mJ%W3dp<>E#D4on>*vPr1Gwlghp1(6`t!FIOa<PRNg>e{
zHtnR8V^(RE-aRV9K<vedBiMrVj#yrg0IZg<tSU<)c6KWEc`+glSJ9kM>Fh&kXJ}1M
zN(7Qv<Ox)!PRMho2%T}^{0G&Qr-R%19C#vhGpGdpoV^MuQu5l*1!0eu4Hd82uX|aA
z3cMV>UFqlxzZPU#bbRoS88FS=#JYJ_rNER3>1Xcw*O^%kILe#ANm>4$za=a9vJy7s
z;byWEcJFn=hsY#e-#h;5{)MpjPEujNbfrw}Budkh^y8~j^|_jP#zKVH(5Lw}6e!g?
zUzpMGJa0uo2#^!YEKt&;o~?A|@(!Y<ROK#CRDSd}e|#xWccT<uP@_p<re4q9uo1Le
z^A~q{CgvCPAZUtWp8*K1=k}p>-y}82X2?@`M|n%-cl+tn_6(<wXO|ippO8<>sql#b
z`&j2xkk#K+NA0Ugy#a+rTuZ|i0U6Ho=tj14l;L3aj^2&XfHL3S-{gYgxyT^s?$7q0
zTWa0P+lc9(nOL#2C#sm(CU5Cejm}R0Yj)l3e@h$ao#)_VNn*Xod^|63JN8%&tLx9S
zYXQknJ0Qmfy5q`A@3J~c(B{EuOxTYWF}Ekq-`i$>9+8ms4tC${Jy)Jgo#BTk(){yv
zbNwq?_b0K1B;CZ{od_8E<7fCnvVua;z8#p%tKt`Azr{eG(Y^c!q44V?Qbk^1)&Yup
zQ+5*Clq^o4!<@><Lkazz2KHOfzF1huyFi}2pVW2vpM(jOOcAbX60?s2bk84zx7?mi
z99XcVMG)~kZjg@mc*oi!o*yp*Giq7SgaX$g7XWL8`<q!@-(V!Nx+^T%!g;tJZZ_Za
zuMur~@I7(d`cJowKYZ5{p`ffsmp&i}iIjy>OBNOKP?8yHThFP&R$5ir7(&CP84|+P
zkHtx3MH76gJ$A7?OmR|=sEQqTXwS;T?`Kxr|Cg#%sy*|Dvu)Dde$FI%wwihIX4wvg
z>}EjvIcK;B<irJH`aiF<UbXB9E`u#(%N?&>ak3n}3gP!F&SNlQ9j-G!;!zeuXy{?z
zXPqP%6eYdHPgKxx-W4}>S}D7spsDU|aeTZgntw!EQrEb2xDalS?A>ect58>(AB%&w
zhTo$L^tC0maa>hVpymAKg@l)XfZXnL?Nao27O6t@h-9;%{nowhu}f)Ek-?Wy42Zho
zo=<Ej&R;QOIvnLiExW)(W2|uDJ>Sf4)b7@Y&l(Wlh9Exn(2A^e#y&>U)uPsDQ|_mj
z!}R0Kb5eV*u>*D6_0UP|rKq^qgGtazh;M(cWk)qM`DtiRuyO@eg@5ooO7Uh{&3Gaa
zUVNIBJ)TS87UH?k?#m-ThqcdsFUEn>^rGUQR;lJO?FvfIH>qWly5(-=c>A5L`^y)r
zx3dkdAk)FE>+;xruK77$>D2YhIG<9GoJtf9wZ6(3Zdq*>o{Rt-o@hSzqUO2zCW9vx
z<J#>OMg|~+qGK+Fv{!F3PhPfrcku*O1(I@!^~zNk9IME2f{hDVL-H~1cBfq!TC^jf
z+$q7H%`2gSt^r%4KdAaC8@vVU&%J!{S7b6tPDQp^e_XDOv`|w*xgCPRZBZGH3bA}C
z3T3O(JB2wIciB#l`quQPoe7=!sr$=reB2Y|U<A(Egw}BnG94b996Sb-w}e52ID}xc
z<&VsJbetI(_POB!VE5$Tq$nlKeJvJE@Q#17hY08j$ZigA8tm1tW;fb0jBVo^6dowb
z%u*v%s(%gByT!)HOnp;qVeI<{2(ZtfQwn{-+CjW}E%ls(_8TO?IfBSUvWC$wb07P`
zrY<^5%Sl&%lgrpmAG8MI`Z7?6R9BI*sX!KZK5Ol55qzJLb*Qa7IBQ6stTTOJ7HEA+
zn6}Yx-!U<%Xv>(4jdE4%QP+_dZctF8GS;79MrD*06>g}VCefEG9dnbq%T(kyXf}Q_
zZK^wi$NNbbH^a$j0#KErrqdBETd+V>H}K;-z6{lti2cQsFa)x2#jeXlHtjsS+96$F
zoAu_wDxZ-;UZt9~{B-u!Q$o@%>)DSmAEh5|tQpJ(R<tHp@@5}Thn*6!WA#KedYm|0
z@z%;<V`F0b6Qu?u777UQN1Qo;Crwyu_>G>Qnm<?7$d^xxDcQ#fZo@rfkb^%JXXjh(
zMjw7gMf=~_!8FWfbeb}-=9<6eK@z<%dwiHgJ|HS1ClieR4{3dSl8bZ#eXv*;HtxC2
zNhRAgGJ+}{Nz$o3{OcDE2p!TJw%+9x{+7!o(3o2Ji&ZlMCDp!Dm)58lFAFnXJcyPV
zVs_Nw(ejZ*nRubtRPpaE0&z3&SNovUEKGUm<!N8l-zOFFF5<-(CkG2=pvVDp4v<d?
zH#n(8R85SdnL~LaE~U_NoSLszkdkFDSvwYf%*S$*$XFT!y~+3I#|j+aKtb}0b#feK
z+!h+Ct05DzSwg#EYIi(wxtZ%b#<F#T6E_ZmBGpxNHJCnIxDVH-VsxXU_?ClM0)3=^
z5Njc>Z3azE7%AXMJrI$XZceHxD^s`SKlu%<?uF%_qdd|p7Fw#5{7puNHVJ8D3b+R*
z+V}oGDoPufZ-gQmE8_)3nWJ;Hl{JTx{XhON4yKgkpE)2xUtmA*uJ{+}yV#S(UGtK`
zkvv##ic9Br5t{sHKgO9-fACR}cUi8gBQoZj)MnMhX!5c7dhMDFziD=XEHaFkLD^dS
zFa2(h533GCe^=i(#ll>yuzB2iVQb<oA*C;ap>p&hT8$W})tS{4K6vGqK)JB?a&4d*
zp++cUyG0X+Me%K1iKeD4XpE_Go2ssCt$#^#?0xQ***{*T6uRnGoxCege>kCc@e<gG
zd>VO|Ag@f!vO!NteVfvm6o3#8jRtL%_$gP^fPJyL32Tb?XAU0d7^+##Jz^J}r3O_6
zjG11m`QK?6{gf5mM2fG}q%T5%Hgf12rU<I5orX4~R<S^~+`~_lLCTjCqEyd}eM2Xv
z9VNg*T2fi`r;qcWp4!yQT?O9t)s8OTvSP|#6IeM8J;Hl;br?xX2tmmvlau3-?taX6
zR<49r#L1ex?r6lt+?mIruy{Iomiq@gcAx4k;rHsMcy{kBzBhgr54ih)4L)WWI9#uL
z``kM;f$txs1bCru&QYS}*hpkbFU)A*ths}*kj*cOY-z|Zh|gQ$0nggpw`@TRtn$v%
ztCS&B{8Dl)yO&Wf3KaJ!%l_mvKOvBDzX3uHv+K}T-B?DrnY?cdva#IabD|S8G3CT5
z1s>J^X6Jl2*H}+;UI&Yt1dMJ9@t(PXHh{MzLOHLco(<L}fB#zK)2UaK2+*q_wTLRN
z>JoEUo5hQjCo|M|<cJmOy2Jm!1A%qjxc+mvm8Y^-TK6SQdZ^a8DRX)pxgJgUzJ}5l
z6if^%S(|SHQQERE-9k-1v1_a&q_*R+t~ozJ>n%OVXJmMNA24|#uxq_-%fPyJ{M9ET
z6mb`6a3x_yce(kY_BKIy>*jzYk`5u%M7OdCt2z%0m1azOn)cjr6tJ5!tY)+U%fa2*
z($pt(3MosLN?VZ2IYlo{l$!4u^1HG?$G_!m-*^0DgO~Psy7j6CgUwi0|Mu9c0ylWw
z$(y%6l|)7#V80U!DXr+{kuvbi0u1U?p$5%Mp>#?(?_DTb?+Hy5+*XBYR@5M4WnX5s
z<@VHQivXN}3OE&VKi)Suv$4dDjqlfszB9P0Ob(+K@h~}a@}LC?SUaX$i&M;qq@Vo$
zx9L5T;6IKt0&Z8zEJ^f9!}miK_e1nc?MGiKbS!<=I{X}0MR!^zN4e|`i-LdKe_-K`
z^S3E4f`=ySbmhz1_K-0oi1Z{fj(V7cyT3d+K06OlI$J)Cf`k_-ofV@U@1Vr7f=Idd
zV$npy?=~v>(xe7CKN8WjZ9j?0b%KJ6#~}5(SB+Z}<%s3d;Yc@}uQGK~4#izdw4{P>
zd2k?HRCaLH%9A^Y_u?#Tq{MD1sDcJU?Pe$g;Z!<a;~hJF{w!4FVggT}LOr1`8$Vz%
zQ(=W|rsTL#1V>kZTqc$Ntd+!)xeaHe?qy^xcH$j<S+0nV$MRn{6$4Z=m_^M%%yWlP
zi`Y2aME6Y~y(xr89*+aq`zY#Yqd{Aes+p{yvaW!h$&OtQ^-JUO1>KG6kY{S|pTqn6
z`*{Kb;)vNufG5snR6hXmY`e2(&i-bkqd=ektaI77g&$uy-?4K4dYw6duQ1~U{Bp8m
zOZ)iDO+dr@fjqE3l&IUC8w};;$;A_hUJioECu8V`+v3DNXN{u2zvj2`6i?p<VdThD
z20XqR{YFPR*>0j@H|{X<2K~!!F5ip{mRmq+vzuYExiQnF!(|ZvbxlNj03L0|CLY7O
zZG|iCp;;_%!?hf?yA6NIMjG=os#W8@8$;d7Da3mAS}vpK6TtbqI<igr<}h5Y_x<p$
zn0o<5bnsEG(T94=FCkFlc0Qp+%kbnEIjilOY^<kUt1|t5Oim_YYE}Msvvq~IT~V9W
zfjAk%=K{26BG9-<JI35BC2wWh$n~E@7USy~F!6fM5kI^RWi014z7zDgEPdtkc$O(l
z>*X+}lmV-R#Nl;6BN;>p>qn`C3Fhor-MXqVu>KG+A&u-83$2g*TK_#7E+`)9KdAS!
zzhdH1Ae7|FWvt6@u&$T!KgHi!M%^|glJ*|`9ngEChC~Seyxj6O`PR~T!-fbe+?eXH
zq>v<`1D0nU1r2hk0NVCTEGX7rQ6u_u@Ufy$)YXxju*+;km(Xh&tGxAzKGU<N%v)HD
z$VY+@o}Qq5QU)oxM=hVsS3844;1lpOk@n!RPiK5L{x0AZSFyTOajb54rwO!geIoz1
zho^11{NM-Bt%m!J5P(MQt2X^`Yo~>tCG)Hpz}{h^w6hA~LYFmx`<4^)=h6e!OKPoU
z%H2`6t%C3*&!M)QeUIEAYG%~4^t2Rbsom~GDG?K<yYKWKNuQ4NWt0D!d|o_M__tx3
z%{&uAIJRzBsdrvB<5FB`(<mu><#W}%D)_&Hoe_DDMg`iLu4~sQchA?r`O%LAClE{J
zMIz&y#3Cuu;Z_n#rjiGsE0pr=&7Z`^RjOVdmqixmi;frY6Io-_Bv|P-WxYpX(2C1)
z^Ua7+IQ*Grs{b=!AZfw|MwwvPLZz~ZlueiDe$hpB6o4o|o2hj{7%K2uX=<3saV1`R
zmt7B!sUNOSL{RxWn02Fj<_a^PnI%y=T5O7VF*0HwXovCfC4L(?2x^>~N!aN8-%ENh
z8>Ot%x)VOul;;4H%@!F-tg?XX6F&^BE*PlIBXti~6DUR7yD);}WR=2t;2W7N(ofg<
z${>HTF8Ufe`2&L%s@P9Nb|BQh<gNZjj)zc>+L?3DBo0Q+49ZQ?Ueej&X(q+AzX%cw
z35q#2P=1U!JG?E*kF(&gR#s6yrS6XD-_HJ41#x@ATKwm%Zzvpo`@H@Cz=BO)4*(V%
zD*`O;`%}P-FE|ecyZh3TJ-USo$l7p`&$DC4;Nnpjl*h#p$?wWbLEQAJAV}p_J_(ba
zq=yNbL<^(B=<U$~kp2*8*)`cdIU!lY<kDL!vQUwDJ~g7Pj3A-b$j>gqMyixui=B8g
zl!xIRNoJglnHBI*zqh>oV1v6lpOAy)0a`-w0{kb#M$l%Jac5&WP5YN!V2)?L#!{uD
zr$d%c)B#e`Qk=>!bmZsM-}fng*gzlfn*MLP9nwT>yGk;38nb@P^DI7}R=eYi!6xzA
zmgQs_@EmqxCZ7wgWwRm(HFNel%23igz&+;>|IN;H$>CrWqrvv^jF81mK?7z#`}44J
zkEZNt_|ocN)R+K|<Y4fT-s4JJgM<2AiRbpZ@aH?39#yH~kuQ$uZ&WpIQ{CynpM;gm
ziO7yK2TKebMt-p<kpIeRpWD-*)2C1V0SFk-(fRc6ogkhTZW$G?FwoG_r`z~;GWWDa
zgx`pab!te>Bn7rfSjsappX7bcr2+zo(w^Xl5_Ty>$cWwGemmqZf(;eh;Nr#S-$w3_
ze_Jdb|LSV^yd`Vj6=qjsP(=gIzwCMt_Sva$?i4?0-l@^+xx1u=`EMuI@bJh+0}+x?
z^7f+SK+_5A6@ta1zN*}SonTFjLfF`DX31w&qJNimJ|k4Kl=$ns6KW?%aT6vL&DvKj
zmm#6jKQ`Ibk-%@Jzq5g>__y>jr?0af3jR=iZ1ff+%A{(bYNCMPZYh5{B#$chIfp8$
zTa%aNJ~>GclNnVWG-`{hUY@&Fi6$JDORdDjUv4IF-0EYPp7Gro^$rg+wGje>>1&qc
zC)#r6h-_AGVmV}uSpQzVPB`i)v^s;bbXw}{XUPm`Hp$^GDiyNE?yOAzWg{{C&K_0w
zxaV|hoLgLui)hLI|KQ<0b&$c1OwZV#=rVDr7}h~#G|hSpX~2O0$vMT&P_EtZPf!DD
zJ+}H=LGO*uCs#nePl9_Lq`!RS{L@~vn7MnBe6-t>)igWPuq<yK2~1@`2|t3$5QyLR
z8Y?_O5U>a@jiX{K%=%I_blFx%*`kW&gmBz+d9=3o->I_@m4Tts;l?RQFixH4jo3bQ
z)H34sPr{l*{{-(o6;`7<ec+p9Bxvw+A94(IHk;2d7q+Bx-_B<ZC_E!(SwnE^dp`sS
zoZ_XLi`u~7eBrb<%O|2YxN;4t3YE@bPOlYVNG_$LVH9Gt@3fY7_jL35?twwF*~S0l
zInl0qZhYmZ#6IZ*R-sG#;P414NPJKk?_#y-fZJ?BBk#uC;lCKo&i^o2;7dOn2{dMX
zs`bD>X|>JyejP+Yr=1V-waZe>|3X|4eE(-E1iP8G<RGKC+M~LwrhayT2d63(XSq4H
z(ubNRlDpk`Q;?2S2lsxmT>HVap1U1*ikOMgGMg2o{u${1M96S+u%NkNn!(YRu}8F0
zYtWQg^M%)-g@cacrF*P<q3e*Q#3p0n9o*T<{XDVOZho+MS^Uz@64M2@3+v_0Z{5rr
zn;{GRmkfolq%)X+`=%Z@;%14=m^R|(yZCEQvUC2wPD(;E2r^JaStP9m%jpm+7SQ=R
zuj3P@%x#LH59gl_i#|DkAu?`9QoaJCd4fnz{H6@;k8~obAt-);B$q(FRL_j4i=GS}
z7ZWz!;?ES{JLls6Ty%UV(7HSoiX!z|azXf1deeP#l5u46)xS_-<XI2@F-!}q`()T`
zeU<m^bmn&Ju@~#*`x0C-2#B1^1R5oz7pqrp3&VPsjTh5=PuFq)ZAdA-1-EFZO#+dv
zk7H?Y*Oc!&5!QIoXbt6p%i`R_Y2Ns>BJz>mzWtS0<}v}E2cNwS_#;-tEEK+CI<-(J
zRJ}$%+a-SM{xO(Fe+)SDk}#mFXSMNe`L0+lZnhPo0A!dhQG>H{pcC0-T|WNCbji~v
zJF1HV2)}r6svW65_sdAIPA~J!ZG_z|)xK=_IcPtI_sokYVg@F)8nCUU1>4>wO~u!V
z?!&?uuYL;3P5JDEJ^Dvcs*nn}>@1Eq`819H(eUS~oFEp6FNO6VEOt>S-<yD;XNtom
zv)w!ISAWsmnAGGf#}^G&s^fX>YqgR5X5p<B&*1(8S^1oSqF{K7gY4<Lf9}x>>}j3+
zDiBJD@;Wf`1lMqzfAt3aS!dn|I0DyEOTC)eBXP1@%nSF2klkuWQk^4wFhK+3b;>;7
zVVWV`Pe;NRAyI4#EAsq|u6Ep8cKjPMjHX5UYmA-1^Ryhzh1IFisJvT@)wC#ybJgSE
zQcDR;RFYPtkU=yiCr%T7!n;K5qmK%WrhM7N_-d3<W#U5o`_zCA;hP+(b-?52yVD_z
zfe`R%H??57>GEoT;nz!Eu&rUDf7IvI&Vk)VN==5k&+WUlbOF{{<4X4{uR8rcRp=rn
zDa6|Mn?WYonUf=7qTX&y-PD<T?h?%Bn(vSk*VCF5<v3@;U#%oK9H7CHSHhPH*38tU
zQx%Ya=M7bNvU#&n(}i6qI;r|E38jTGcK6!y1*_YjyesIFMuBLY=0)2@ne8bEHiQHo
zt^)J!)6fqs>!<q}XH?bLJ0v!=h$t5_6tIZtnYTUC$~5B0vNKm~Q_2wB8D(%fo(<&q
zNN|-C!(cY<(nc?LN0p{0w&-LpN$(pL;{PfS+&FUyaBB8&Q#G^=YAKMZmV#mOODOzk
zPf$d)WFra0pUiCjAZQrv=OO=@6a!u!z?J$)=N){)KtJ*a^JV)n?B|P>K(rutsO}^B
zmvWAB^t7KFYFWjNC299=o@m_8#g#rR&$=h2??rv3mHA3%@Nc0~W=-AszV&j`_az#1
z9<kkJ@p=}Bu;9(yOXKnmm-WU$=Ngx{9ERVvD_KJsB^M*q-ohE-9tW1Veu(o2n}=}a
z(S`pasX_wgIV{h4oK2UDmTqU?S4T#{t{}Mk0Fyp;?dNYi=|Umf`OT+<_plthnJ7Ib
zyOf7!>J*VMMbpP9<i~(1=Fn*-A*e=TMonTttu~CMBhBb4Hu3jruOQgwJ=-d?vK+_7
zw=Q-SWN=;=6Q;SZQ!IDR9o|t+<p%Xj^qLxt_Nzi$8d9@phB!_wX9XH?)AP>DnZbsJ
zxU<pS)4<P+i_?QT=?wOsBOQ7ESI9<!#y9>7-`VH4L7bgyZ9q@{ZXwI!XG`N1M!)BS
z$Dv@$k4JZ!5$N`s6Cs|@U!!ikpT^+zy@t?Wzw3#m=?sC!F>3dH=SOQMdmhaN9VDDp
zB6oOJyyL!_8+uh39oWo^48^@i8l7OVdUQz=i!Cb~lJCU;)l(7fhyFngk`&-v|CL5u
z{E@}|vQTG<pLsK4Dia%+l0q0s!slO?SX4yMWBa=$acRPS49@b9Tt5>>qD_ogKGBj#
z&~BLU<4&{qc<y(i_2sNGI79OSB1~84T0$#%<{M-`5&BLi?yPj&K*3Js!f%l_jLfLH
z@bwsQ0>&*p>D~(YP-rV$CaT@igPohs*_oV6@3bNvL%Uvk0gocK#3pZaxvVnd7J+%~
zzl<UEiMq`sXpxTf&&L)8pdXW4j>j8X;ek~a(oeM>WYVeD8NyfgETfmCZzPYmPOolc
zfD82ZS8p(N#&iU(A{#hnJbZmx4;p~)s9!Z#oPUVL3z2e!r}7|D@$A>iq11d`rczHB
z-sXFh1qW9<mgA=7)i4$5K)(uxA4wx)hacv!q<*xMeG{_W*|cCmb6yCWLMgW3e=h6*
z|3iov%Kc9ZK&qrv{Nh+VhwjtPSUSK0&q9tu&!}DpW)aB<AZ4r3t+IJv6*Qd>tfpUb
zqa%F3S%N6N?BQc<4ey8pYJ>vPYdYg1xKIV1>MmvUUs-RPUg)E%FggL}DsCXGx^@SO
zE$vIND;n=9NvQ_bHMrI2=rP5}zYNLM0H><_1asJEqA+GA@^9e=bdX(RA|;W9I878k
zpbC<C^8}#5zGrCAuvAc6Q`O<3kH`gXX}+0jFr(Y2CvO2HtCZ)WSvwEiS|tfy+iTFW
zm^%MG15U7D0eYa0`W#mi>|+DNI*aUaKa(M|!JWjx5A1pP*Ml=JqiXakbi2Ri+ZyQ1
z6dcHcagy^eTx#?E1!X)APG45;=b!n&T*JyHFmMRX=HKq__B0puBm><Db@jUM&2#yG
z7N73KWXJY@rq0UjMYMG5^TWuvYzjfV!FXwF$bD7#&A){J)^;hyB>MxK_9dH?GApsR
ztZ1rvJ$Fv0oT0T0vOFix=AuI(^Bz?yuu^d4{=9Cs7(5Ucn9+eBY>6k_&%Jtwx>4rw
zVnDLvV?@c#K@Ve-`bXdT;M&T~#pq37j-3-{J`@5^@h$gr4w14wAM&dmB*0GYoGKH5
zr9Ndgz0<~_ikx+Iy~f+Qf-jGIf$~?&4?C{SJ{cm2Q(YWOvgU!L-@oRF>UqF|G1wDs
zN!oK!w!`Q9E%Q<76T!ItH81~qtL2>g#A=r*(cEcoHoh`Si>L*K((9nC1frvl-$LR)
z)gHVL^W>R@k$N39^#i5(U!$$Dw`<TXQ5^jI!_vaH?|<KnxlCE{RLxqmx{PH2Q3{+9
z7Rp1~c4sr8;<n=5az?RT=DftK2p&?=EIPMX8G6WT)4a@SMedAa&ILO58AN)@Z}A4@
z-|bgbMs!*8;&4$DV8o0EBBayIFIlJ$2#~;V94y&a(*|*;iS*n`zPsbufNX!5pT1=v
zBl6(RENsUae{(%VI~lAdf7@hSA6Mf@C4$@;tla^8)H)d*NOp7`Hd4-9W1$0mzg9y)
zI7`N(wQD`E<&{}2SqB#eJ9ll2=2r<hfc{IrC2almgOnbK&UJ!cI~egzjhzY(rqAr0
z^bXa?c)XJeRZlF82^2TzY=EYw4gx+YgUAsgQ><L_t?@y6A~xj8{!+W7n*U8@aVRIy
z=Bbb%R!_f%)<f%O#!<=1CG9L3+h4wK*y-iy)$w=XACA(Y>ZRPS*T7IC98RfAZKeQ=
ze%qZdhzI;lB{z(NhoRh1wTYprMUnTj`z4L2>oJlFgRza^;fyZdaBFk~QDS{>yKZ)k
zLOvqx*7mkOWz~${@C4~i$kx2`LoyEHYe>NRPP;QN&KRJg4G}Z;q|yM>%Am`@QS%I{
zBP`+xl*7JIdZO6gBT~epn{%oBQrHutIP_5-N%}3&7tSkVEC_0S`iXr~w!rdIoh*#}
ziYt8t`ztq|Kf}t13N|z&6I^008`C99yl;nUU`7^^D{Mx;Yp*mC?=QI+M3xk~KTXU3
z!CG(6a#u`E%C>-HU>ziLW~mgNI0|1m_A3Mz<mX>c3qMTLR(>73Dc+O?A~NN2?y=pL
zHYQ+*;7$j&l=VfkDW;v3M@lOV0LrOMpSy{nLG*=V)IqoanCmNPz;~|3YA;3ptRl^!
zBH7DjTD~!EX%_OB@)qDK^E!mw5nNtOu5%`8Xe$#Wm)`)etS|CU^kA`0>tEiT)MsYr
zQ|SSx_NCS?9)Ts&&`gbY)Sno5&H08YAj6h?(M=r%G&A;Na-}y(0nA4_A%iGkm5IwD
z%s8-ev0Lx~y+mPE2B_y8qbjUF(s4>JQF47nv^#RB!s7+SsI`uEm@XkTkE#FZr4TN1
zbvl<U>aIcm*{id9tAS3!dy;(ov|sVaVC&jV+;|Te#wveX?HCCx!XHn1H#vIV?3t%K
zn%W_WQ%7~Ybb|ORmcAFOTpk|?my{KK!<_t=X1V#7;g#Ae-(K&rfYJE9;(i&~_d<$~
z9U-CP0gOmD)|u;0;?>=~97+u?c3`U}v`a;kfYk#i6YUoyLMA+>goGW^`h%Vi!lZPa
zGXZCF5lHVj1iiMjTH7ueM_`v+?dpfE!2Dui(Uwc<rUo6P=HHAD^xckc%{vpbfv357
zU$6vmY}W*FZU`-VfWT!RI&zPj9Oe!@tlB5ZUR5<u$zHPiqwqvQ^g5OQL3z9-6~q6+
za^A{-N3EDmt?)Xuyq_d{z<So5A}xJ+`>J@RrhT!;o^(*N(Ux!Pa>oC368|KvcbZY;
zY;UOU^WuaoGf#}BWl?Y_$SR_9Ghjl{nTSMbb7_2+rd9lU!+?`o1sLmp+q*rb%~L^q
zt>3xcYI?3+PkpBa$?!`6+g24g@S^Q(z+{GZ$M^Z4ukZrsd;)=<B))Y|!#{h)%Ivqd
zM!C;%BEI9!r@l|IFL&6qAtD_XX_Q!{=P009WW<_T@+}V~7vJE_zQ9wG^DG*Spl--|
zf%dsi3|$U6x0dOWQ?AW`P9^ZS!JBYefbA83WU>60eR<%kpk(`s{%Rc<aeQ`lO*R!e
zO92l;|8^<~vc>8{m`BQz{D-D{tSQmF`y+~18|Pd$p`}pssrz;ST7keZ%2wvu@FX`?
z6ZIq@G-%=YrJ+HGrJ&(SV?hB-PUj?oszRgc1IYhwOLYXh59GW13W%md6o|Lq5}JFP
z^xB0_jyOA_H`H|YXdxZx?LxL5tdDdeS0v{D>7s?1-WJi&!|!2WTVFr1JbvULeCU|_
z7|sJ}&-YI7_;3p_^xVpT4yMGI>aU&E6T^UyP`Q9h|HU>fn{v2WN&{D>$%#Wx!vNJ5
zznPN%fna@^gc-+Tu&(#@u4$rk@t!ff5&TASodMxN;{Ptf&XXzFRbnpLhdF-URT4OB
zy;{yTDW(gBhJt$k4Hhm4Gbf*WN<cQe_JJmK^NZZU8?2lPw<)5kxPk<;SVb}HIK`5c
zD06Kj^!GV!olrQmG@^xmIl=syj>jpNwX^O|yy|c=xWCZU)o@G4&wh!|2zr#&jBhX3
zj6WW(-adeMlSzRc4%axUKPdXw&_35+eFmP@qO=0_Zf7RH?|xrFe$KT<YFjsYg2;#2
zTi1i)GXW#NS7q_(@<?H=Y3{oe8#GVGXAv!5?mYOoZ%!%tS5%r}+x0GVex00dv2D=J
zU`jyKQ{{=|FoR1`?6(Ow%E7AprmK9`m)kg+XU*>hYSG{2+ux~WO=spjfuCtpESX&c
z_bbx(sr7DuHAy!W)I2rRVE_)(oTTDh{}Nhhf|W_9?^BMrVATe%(qQ^{THL4=tEnl!
z;jx3%u58~h6vkfk6Za{N-kU2VhM#A-?C_nm7IT$~$6Hgs55K0ld@x5L<KKcog(Rz4
z0^FkZbD<=uA5Kq(gbV==U!UfBTNWw0Jkw^7B%&zwxfv~A2FYw$qQ#R>-V1hJ+#ggj
zC~{h{%hg`y9VrjVMRFqT6B2E&ol8{zrJ`R>I5KqH%{H_pv|op{cMv$9n2KVaTmlDg
zuyDw1wNSmqdz60W*)Y$Ve7Lo_s1hZB7YiE1gOGgJ@qFQ((x3|a_g8;Q2c6?kG7%nB
zZ#?%Loa4uH8K|f!1|mAizj^MO1Rs{gFB1t+AmJ*g!^}oWfSuu8*=m6TEg~KhOEfSd
zDV109uME!D#jn%Ly7!g*iP4GNy+Cata!DFDttPRXyhYx~Grai<_$)n0j98@yeQ#3b
zmlmhwWy5}Z^xo$aeJA|pKwEC1CXG`a-ArLs_MDY$tzi4uqstklVyL5JA{Nbi*LrIk
zqApN68$;aPoA=?=F~G-AwLV?M3~&8nne$Ojs6bDOpV{nNcy+}@-x^YsGBO9+u(?z*
z1wE{`MIf4aBCf3A0Mp194rC))WO>WzIATPvsQEZp0A|W-n}yfK!^kdN#Y|O~q*v7>
zaOhp8{M}ZFhmkV+l1#bwFSgb{4$j(R`W>*QB4|6P9)6C0Yy`iaJWyH_Z}IjkGSUNm
z^#*m6Gz^9zHRlRRxN#PoehBNIus=$1<{)l~Vj<V%EK}QO2nE4KwL!ktIDjmE?Eet;
z)lpGz-`jMTNSCC5fON^w(g+g5fJmnZA|M^o-AJdDbfeVJCEeZNpmaC<4)@;g`>w_O
zr)y1o&OUoT`-ydIla^NfkD;6Ti5eCy1F9U}5q*w=zmB;zI1TOYI6JN~+n`F-JC#BQ
zv$Kk?)wu{=skVN^69p#Q4oPvs!$WxV*NJn_4U~UwQ>tFO!G5j|PAC!U|NL9I@*Kw+
z$6b+2?QFI=t&)qy^3*|AY43dGQ)IkIrZ96Va9jX*XvxNC)4j&TNB8}TvNqM=FgzH|
zLCo*B7rGvgGNM_m`>@-b+a2oJUPLVrp-8EPiF&AxD%fM$kbO7e(Ce@8u`~h_{+h?q
z;+(yM6+>kLDFZC;vxAs9Tvylod~NeF{iUMUS=*qw1nPiLK^p{`Z;BfXq&zaBK+pi0
zFhv2Hndzhjqf?9Df0IH8T$p%8lokW2tm#N`=*2d>Gxts!SL{MGaiAgr4~6@N*0>7x
zQTuCX<9)kkRxRylf&tCKfJnj}MVAS=$7kIS^3DfEXB-~pYY|N~r)F9fjZfTTab{qo
zb>8IFSA?P7_shY}Jd|=$B>|p`;pw~nvE>9foV^zA!q(Pw#cX1mtvR<TQQTKi(`&;E
z?laRzei~lMK?3?63F@tLb0#e&w+jTWzE$D_b`&9^39-Jmw&bF7=o6nZ&=)QTgmu^p
zI>s=!6Zf;7=I`BmSN0GyI9&E`aCh{MwRg_6<1A1qd}`5DS!-f|0_;jSV(d;NR*I_^
zH-jvJPDdFB7$ZalnmvYWRtmAm^9<fd0Djykx6JIh<x4cxXD^a_jwJjDFc=Qb67khB
zDzL0^bR~gN0Y|HZ$mj%87i&cD>8Z5fv;RqD*LQ{A$bSG<;RvQW-4AT36n^WUTi?Pi
zb}Ym}>=FoaOoKP~l&Pr^K7+L<#!A{kRrqv_`mp`lj;`9okLQ)V13dKX&JQ72Wf_+j
zkNq0Ov+K{D-rrwOb2R|3{j{a<Sss$=IvT?3*k01`&d|c5;qI6%!@yG3h^mAB``)px
zefp?W)g*_8y07#4M~~{>xljY;QejclOeUV5rBboM>nI4t=)sa2Eo!}h>D5R%@U)S3
zfb+slzeO;OQ!r&;MMTl=jC=nM5?I$x5EE|dvkE*k9OU}>zC@Ce%(CA;B?s3K=*Smz
z`8rbb1AJt+c)`Ifxo|;xR%u+8YE&vi8|*0qTt^nN7+;EEu+Qn%kPG;+@!?=eZz!cy
z<`S_ygMFiMFDXL(K>cKi<umSoeEYP=lMc`P&i5Mk&4Hi#g5&tBo4(8&OodoGc)Ih^
z+%$}Yl?h0tT7ErrmP5kuG(Sn!9@jb$$aTQzr?-wN$@DtBJ9M^8#^f+wLQ%3By7sQ$
zJ0PHAx>VG|c;a!q({fw;53#Nb)s~o<#IzR_F;>#*(2{^n<;TN-)>roK0^)L&`X}w^
z=}UKfKm9P*+0J7A7>qG}N%3jNuz0ucuL&h^5`QN^6x$A#V(Ngv{?{!`cqsJeak2wX
zj1*3_;kcRayJg7QC-KE=w58sUMx_!MVVuMXMjCvT;;53e2f3|=phzm=v(lXfQ98yV
z`Q;{(qs<ZOLZ>mnMpNaX$grS1U}u3L@8w0fO?9i&D5O-sMbN$W)y+M~YlX6)hE^5K
zKnS_jTva^Z{ilG;@XOF3Z$c=@aC`RTY9t7%um4-g;EFKuofA1iHyMbic{(I|uBTP5
zKlm`$np;?8j<j}vip={b0+Y8nZ2<~yQxm2qs;q&ixa?YF&$pLV`tTxkHL6aAyj|K`
zkePt3fm^O+0tB$Rh1u}R^4}@)HzsPFIq_gIV5CrvtaBQgeh7W72)3r_8xNMVrHCRV
zkIi2QOW|zqsX)qN-ZGB^o_wD(B1?}BtFxGc3W<~i|Iv56{3sUp^Dk9@)W$^pD_B@D
z@@S8Uq`p*iRLQ7i1-9uU&#b;uw?>U&uFxHy4xEx0@8|Q05BCpMWYKwpEgxaVIvUT=
zrOA{<qu?KKdK#_xy;WBo>}(0$LnSZ(If%$#ky{@RYJiXVRV3)Y8K1chuJ`2}tP$$h
zLfI|=&fC7T!wzCC7#=@<jG(Nldi>yb0mRQWQ|LWE!%W~%iU{I$pvz<e^bwQ^4`7tw
zpCN%2!fz*gwxZ^#?|nsc^i62Pq>`DO?Xok&G(R!My5Eu6tSs}fd#v)?I?ks%`oCtT
z2g!*7V>zMPQ9_=4rrXdE{?q7~nU1*Vy^jRw9d1pF17MLs_W6|+6?}Q%;a+ZomVE!Z
zP8>xE{d^!>5DgseBn;0iP~{ZYFnzkeb=u`AC6Hp!vF_M?9tf0ey^`%jwSM*?nVp-3
z4~Z<ob4pOEPQXOJuClv3hVT^RU;nO#kr1MX%4U5JqG1@V5V)<V@H1poZ!)A`n;4xj
z>-%qyz}y7?@a6olyQjw$F0$Nra>56KjYy~U-FNAJysb9JeR4h11BF6e63$0l#Xxe8
z6`%QdFPDR*s)(5?v$S;>0f_wSrq65`i2QqurNOl<;xRb%`84{?q|tiPHT-I4cY~cN
zP_&;Ff8Z3{jAMiRVYdKY@R8tJ9zM(i?1xpQ?-#L6WCAv+e>Mls9e_Ph@iVdQzjUyE
zFZ;BdcJReKo|NAi*J~uF){Jr)$hp&fshPdwqlsjV62RYaRZkL+gkovUHD_D_f+30A
zG(8lWH@GQW0nOs{^h-oU^dJLwaS_09e3NY>yJD@CVy%XvNH1nB(K$=m?e8>`tGoL(
zO=LFktq<~bA3tM3?>K%tt3BXYeaogBSs@VnF9^O)g+3gHu#YH(@xW?7JommocWiw3
zpcAnFooqql;o;_%;@q@O&6oP>;U<M0eFf~r=Fb^nNK8VW?7!0ru{7RfE<in0N5Y9~
zU#o>Nm0C`!c@8n!R;3jeEESlCyn0Z(SC+?BC7H70U;!$CMR+1q13Ox(DviX$LN)Q)
z;@ZRcV~|`T512?5`K_`+djkF*wX&W@YsfKl8*4X$Rzu+7l)c=R`*qMh{p9e>py&<<
zh7i)J>qFAfZ87J|$+TM0dR?KD0#^$o@`=Atq-B4p{tab9s>^bRTkEJ$E>uHK+E`T`
z8}(G8>MmRLSVPdDPbL%Q8+Lz*doxg4zJ~kbT^HM#m+z>=8bk0`(_Es8GLp(X9pleT
z5=is9toTFt7l$61HFn8||7$Q5pw^k^teEpZK0eU?AI3rtb!U5Lm%H6DJ|-M#1dyRQ
zze@ub)lHVo@tI0dDR7v8tJH2Di|cP-<JWSzC8gIm3nOEuHk==!&o6mRWNgO_P7}cP
zqJ%AqPHN_t4lRQ?#TMG>lb*T9djufvrT+mPW%T<k3)I3&z^nURnKarfjY-COX*yR?
ztMl7;k}1V3Q=)W-CU*aNV*B&lu0WU3FU7QA$Z*c6B|^vfZKcaFl~S!P?w8HY&GjZ`
zHjPp(e6MaH={*mp9gq{zxG1*NSV_NWpG=EkS*zP~Q~%+kPr$t!xw+M{hV2i*rOx?j
zulV`_1~&Ep;J8Gnltlz1)a+RO3mk+fVsam`crYdJ<zdhjy2f&>b5qLF%hKZumsGY&
zsRIbynk}{;=0JTUmnpvEH4tIYTFMEFsl!#1s&ZvFw&s0$76>nIZ~ee2jRAlx;bwA%
zs4P=I0^iC??uNy^I4@FF$3gSQ2&JdhM;XE?ig|>es4^XkN!Nt(^`0Yl><_3BQNu<5
zN&bPG9yk435lwrQD)0#`4oChtUK}O146qqo{#tKV8gfW$9Sep`O-cQ&Y=4N=lzO{&
zI<(T5haLIrn25);XSU_nc$W@(fRTmB>Z(?IdseHZv^p6HnlL!zE{0{pj+_N?Bq*?j
zWg-=>UXFe%$t_ka-hrPKA;gcr8g{k4>-m%+3?9`;K{cA3!|nTEibcB_8<M{T4#$^I
z@(8sLa#7e0m{8Q!k31BC%@7*|QxO=A0{!OqKpKY1o-5tXzly!4{HEV_G;oQ!p;z@r
z`H3bGc{*4^@Qf%iGdog4kG<#m$Lr$8d}plJu@YD(>%X{N2({jLgjWunW0PDl4-d-y
z3s}G}J4(~NsD_UpFsgd+O3uuL$7SAZ94;}zg^Rd4y16V=<CAwScHS4TGZLy|&K3TS
zdRhjLI&X?Qz1-^<8db-YJz1^p4L|s~nTAggKTU8QgkLvwj_p*3&VGWOFM`^vS<S~X
zGe0PgZHbHZ8y&TI;59a55dpB$bLh|$%pp}w&=~9<C9}a<)Ld84c9{MY$^7u$Fn_ei
z&S^O=$AKOGU=+V$|8o-6S8AGH6LPI}%Wm)ymePzYZ9pk&9Hnd7%YOlKR3KQ^kD$4l
z;R#Y82`(eRoIUt_x{Cwdx2l}wnr1h*HOYvpj>c_*Az(Sx{3giT`Q`Pz#|25g01Q#&
zGYP1Vv79W|qiw(6OJSsd1*@&B^$M@PQ-t(?i>$|C*atMS)w3H9JHNvvT)6PBIOF>#
zGjrjrHB6&{0@0f(s5+bX8CQ?Js&57LfHzU@Bzg}?`+8<k`EOOW>7&rK#v8Zmaw~IQ
z)g*82Q&bP5{{4!Armsr5h9}H_SIp;ldij+KS4Z>wXKjZp`%L;RTQitsK(eYuf>-vR
zO;kc4x%AHxWkCo#LHskSz^9ZiK}p8pO4h1t_o{~)-nlYlF?2}lOPJTcB)cv1@wd`^
z!cPkWXXvjUve+&ZBlZaSN8gR=Y}Ls?e_AS|8x|eg_r6z?|0y1$iBA>{63%csmXO7p
z0LCX}LfkmWolFMwh_1^vhLMo@vv^ShbGJod_J^1&z#E_OF7fi*0vvONRmJrGrd4Jd
zQs73h4lr|jS}SIdj`GJ^x{M=RozFN^`K{x~ifV`R(T_ocb<{pjz5$;8JC<`f3!Vao
z8lQRJ#KY#wewUD$ZnWye5q!(2v_8)gL=$q{EE4yL_EnkpU|QEvv0_G5FmlZWb}LFc
zZ=*1BU&{OJV`AY-k?H#Rc6mjrBl*aXc9;tyfs-anA3zY=+)o|`nC<F7ZJy7c?R?Sw
zup1h3$gHi%`C8>Hyg5;sbXe6B1-~XTccDouLAjtgc`QvO&xMvpgd`&Zda{)T{x{$n
zqDz628#7t$msifK*n@?7&RS)8HHa-4?-Oopv6i{k-G#LCbxiWt*O^kRWy()bPgOc0
z>i>7PI@cdovT2<##>qgrIo3+f!i&(p(5N|g&WNd&*VF?QAEar*e;h6T?W%DMt~BHc
zt22k9Ny@YPJ_&g1+G;DK;6ue?yi6NFJ5FMim55vYt?mb5zKC$LfF-iSZMceg(l(sT
zWE5&|q7V*<dw{*syg-YM;6s?e741TTWwHfFtH3~HsFcfMT;#J5t`2WhC4I<quhlQp
zQN+$AGjAO_#XkFtts9cRftfHym5dp?5Nj1d5F26%Ewjv`<G?X9_;i_oDeu1FDx1Za
zTnA~21}S+nvx*@Q;GuXT@wA+|clX!dR#Gh-hB1sCI^Um#0w@_CQd<9skk->!^e{lr
z9Vdm4oD(5m`6KR8fsNN0K;arZ@0TA7?sz2R1)zBgT&lnb--TEBYQ9xB38{Rk_!=AX
zLPs(m8%UG+PgqMNtp{OV-S`$)fY@hiPSB1J8{j3&Mn2U)h~2TI(oH)(;AjuN7C&VZ
z6c;23W8iQ?ZyyH7Mq)b5r9BG5oqCpGTtyLWWr=Iwt+cn^^(jLYB-zf}&;WQ3cAUO{
zF!<3c&Z}kE6`tM$AVC;IO3tF~gX+ftc#(~~ubpXf<<-{vg!*7wq4u{B*?QQt#Nh*D
z*MYfl3FM63jEud%uzMO=tB)qRu!0ktY$Jhj{AEn1s-&)DEag?eYVrJAO_R(pZ5i}b
zXArBJN3Mp=gzc9mX~d0o)LJp|#Qc2knT_!W)ayBN;dEPb1#FOnelThBIj|(4USE=d
zLcsMuv#+$r5!VTSg?c270!iy5UaMgfEBqMTZn-v>b&vY8qQEAx-7vVndyCv2vb<{W
zUfVE$c9=cT=u54RRy=aw2TqeJ!YYut*x?|ml_)wjNl%lsKTbi!0*LK2&3w$0ft#8p
zI48XFHAd&gLqmCp!iKdn2<q}Rz%UN0%?GteG4z&+A`62EBZ>Xm1?j5(ZTnxLVZn!9
zdEfi~I`hQAWZ+SRH!!Sy0ID)7ZavQzdmdh;KbC&s3NA>H1<HArIIv?XX3%q^I{1ZG
zi!85C;@{(L&n?T%t=<=l9?II5wj`(qiAx(pkjz*vL)*zDxTh5^j{F>OuFT3_)H>Oj
zt782Hu5QzStRY$oQVj+i9gvuksGI$(o5&3CGwO<biDUn~(&nD;AMNIPa~RTW#3ekP
zFsmQQY}5pr;>J|e|Fva3fS~{v0;|_?Cuj=Loz0w$8i7F7BdL>J2f-u1G1nbp6aS@(
z=eC@@1E1~fMW}v`sMb5rsUI=LO@1j19bTh_&o`n0e;TAb=$m5&Y$YT1W4~4V*-QN)
zhQ)Y&BQ?Tb+22MId^n<6!ug?WYQ{;c6aD<R$|ju^`m1v(!$qOPSzJ&Py&V1?*9<kq
zPlsKk4yysA_TA(b6`t2l5AmD~6!U)^e*%vV?ryu&No@xLfic1VaRCPTrwIb-czJm}
z4*Z^MtOh;#-SOm(Yudp3R*ia_in6|rJrDJ`IZaEq0y5HMKW|`n$l!gRYcq}Gbg=Cg
zW#@HLelyx2Zpa=QqaM8(3sS>bP0VOA_7C4#34>W}Ts|0;MwiTPSUIXWJv?usVy1s~
zT%iq%UUh^$@#2WJrEOUB=22h(Dc=5k!41}V{$sKjUNyWe_`#NXXe9}9c(+&_aVBW#
z$I@s1A94)r9fc#0!*#GrCELKMjdbdUC3}`E-z77=pC_D2E7tu%>z_ph+C%wTb8Zf(
zA{LJg`9^|D?Bh!;{5D#_8gx`CJl>#=+gdDqQ)Z4c?6LNh>(FmpWY%DOdY=VyCRbRI
zSXr*ISOZD-tjerT`LPQTao(mpDv0Xg>lZ_k+)yH2_?!k$J{}g&d{_eLx&6sgR#8!Z
z_BGOFhvUbB?^KG?^Ll^l^5z7VPy#6~e*Uf?u@ky@b&u`0%f?PlXWt_~N5TM%>FMCP
zmr7kxoen;2dpC9F?=`u*J=S3dP8C@H$6>Feyj;UiSuys|dCw-QihP=5@UIrH-4Gx`
z%W%rZh6Pbk!)TpuB`hYoL0n)`-H=&XoqkmM=03ywo%3#-Z)4PQ)V@k{%G;@%7u9}g
z5{g>dNJB}%Nkyn7g+U)|5`>z`Kh0Ub%vuQf&%8*O`$6;x3Bxj(ol=}xQUrD<GtKeQ
zHZ6~>1~tq{MVBu7&JWLr%?1t<f0DSB6bFC+YT1oXS1-SY*vqTpTW_CXNoRAws08%=
zR#D4p&sL5V*#i)LK$Nu}WjsZK@xj+t`e>vI5;l}>vN0*pa27;#rwM2oXJ=~Q6Q7Q7
z7=5OK_d=}!y+;hW6)<X#b9<CP5Bu&jF~W~!uWL2RZ7NUq1dTcl#vQ19FP4%K9T%o}
z^=%8HYPt;e!B}27*C;GVG9SbaK(|a$YbEfuSMco5KC65EEblqf%H-$IjL4y5GKitX
zMv|pRL`xV?OO|YFv2|th_L|Cc__$7Q{|Ld*zlR}*bFC%ww8z7C+Z{R|tOKC^PMWW;
zxd~mH?5aBJHOz*{qLPbHSpN7n{;vk~Khyvn<I>F-r(b1G?Z)C3o($%%I2g$WN-Z-2
zZR05WX~!|xCtIbll+4))+cf!C;$wn&pY|I9D+0!3Sw){V<MTv&c`Pu2jLw)ao+;SB
zLZp?T1Z0uypwD=v)Kg3RP<BEi)QnhpHUNbH0JyOLC>!#00L=zP?)Cy4hlL5-9}43{
zUuvzVY3x(6IhzQ+i0n5Ve~<VaS*B7nJKvZ5XER*evq1;g6=&7Mw>xFv!EjescyYKk
z(btijOe?@dJ@)gRNwl=7_ZC|n-Heav;fJu{<Oax4a%pi<T%YhNX+dM#j|?xV?mYUR
z98<z?cw1cNEG}yuY}czvZ<+jFm1hxN0xuf=SW9e?x0vILI0P=wzi<EB*QGJ);p+Fe
zCe%D0R2IxhC7+sRz)6ssXFdJ~`{w$hNNgah>r3<pGR(vH-1Lo-vMT^DYJ_<-BzMR8
zYQ@s!GQN(@t)PM$rOX?V*fD5fu(@b{VP>480IA%tR<aQi`%*PU9Nr@THu;0REJ94Q
zfX!=ncbslq($eCp%0~@h#=xLsro9~r0gQEdc<Hw-?<b!j)O}p`n&^HZH?1db0ZZg1
zWHWCR9$%#TGCoGlmSW*>h@mkVEa5{$186G}tVAG({ft(whGui&Bq-jABc|uonSen!
z(^J8ODyk{(bAO4q9kiTyOAEmJ)9dSquDuP|!_RQ`*0eqLRi>n<?74M7{u{G4a2Q)$
z|G)2qNbk$^rwpc2)0u$}O&bg-WXO~9mF_n(&_LYr$WDAE_?nXBH)6#vv9VR0uiFbK
zSkz5sXaDAuGAZ%S-KNcul35qJ&@>OIW8x+m%_k2)=lrNt<&9ON$t4Xa{nZ)k&7&#V
zk(Q-aust@ja}A|MO`CSOdvV*5P1u2%^T0R^1vGJf35kh5JxG;7ggx8{ItkgeVq^oJ
zPffhX;NthLd$8+0X|abLXBq73Jwk50=CiK_ZVJp`%!~Zv_12m-0q26^51U&zfgIhS
zy#(ulHlOwe+pp}Xu2?G{bzfg2tE^i4<W~l+PZD>2?qZs#plg+0>D-lE3b5z6z!(p(
z*^%-fL?nKB6%@;H7@O5U3=AU-Sy`buQr*80V=6j*<k+frC^5R#yM9-W@fNW@38c$#
zw)a{T(^=k@3XZdH!53Pvi2ny!8qyFJZu#_LNWn6ia;%-;y5kp%t5F-4Y`#{d`?@vs
z$9h6RhJ=JqkN6vJ8Zxw?Q3~Xek1uz2O$W~G(FF1+&@Qh1%9Ip_?4K5*(VX_00GGdb
z%Z9@2uN&+IAmDbfolE}+s)*fCZ&a3=jY|a04N#UaCx%sgAj1{1;CXsL>^Cs-7p^HK
zWBnT7apVz;a$)O2a!-eYk~56RXq-}e&>`X6==ZnE-FLZ-dktZ4ZG}(Z2Y`xQ#f;*T
z!}4P1IK4M=(#zU9#!HIK60MtO#FsyUj0xqkH=lU$G`Ph_E|J&X21qSsOP~~FDsLRp
zj=h)scZRr{L@$Mq>h`YbTb3IV&p%#bt4}gcI*~C7bm+GBC!1$UV)llG>p$<U&D&6V
zk=z!XX5LpofzGFdnEYm-AR$DGOGotPdi&1>ENZ`;z1Q;_fz;3Qt51N*^-poTdHa(k
zd>x_`cv-(@_%xaTr`ib@ay7P1o_Q~(1|Ka*TT$&<?v~<fgYK+i${!m&v$M{!I(2BM
zmkC2MVXBn}kDIz;w{cW&;)Jx!juJ(ToGp7y>;mt+*{7@!QAyS_07Qt)ArB4huktHp
zZkFuGY1%}QPXbFDrHg{fVpiAfM^Xgxk_3vsuM!rZ>NZ>X6egu_r~V2JJra(tcgjlG
z%eE}VzkBzqXAT__S2;AwKt5%yTEoQ22P6D`3;$v{{EjVUyMHs0cV2<4?rvaq{86XW
zK@oLJ*v=3_kmXdJH=;7{>|m;sSG4%t1ziGRL;)kVpD#p^Juju548j^5{}LgaSxst0
z0eem%vSNj`ufeUe(1hDPLA_h?4=>iif26LV>b8IG077{mX1dA1*9uZh!5(cM%dv?R
z9`}=(rrXnAK{1}r5UUTWZ!Q`qKf=*Xw0U4HhyCg)aqs&Xlvo{Zk!;x2Xmx0CtoN{J
z#}jZPqDu#PrYOLAL*A!uMd|xX-x@B)0HY!bZhb8X*Zk%W^N$-xXYh;)e9BPK4_h}+
z?%wS?c<9~HjMrwGnPd@QR3IVt8Xen*Q{OFqN{d@=TsnR3wBuq=J*n!k_BdWK&jRAz
zELI{dCMxPVV}s_0$c3{dCf2x{9Db|t4@dcLQ8C^nG-p)k!nu@x(e}LJI4$@aOZ&O!
zi*$yMPIc-$sw01ZJF%=SjvuT|Rag0TD^3p56KTmSelql>jf<C(oBg|c_sa99y62Ev
z6r`cIcTY{Q&ts?R)UhER>>>QnVpc3%bvmLgBwxcBXz3&;M!cbO|LZN^H<c^#MYKxl
z`J`A+kzZ|;?)?5j;Hz8OZOBF2Uqu`tx`cP5Pi_+V$Ne`n_**b7242EPVI6-X7p`rg
z;o}0mumA479atWiJHxTE5cZggkzt*8xH0OlaUs9FB-vQWLQ)qlf3-YaKD22H`pEEi
zzr&5mNpol=$fR7N>+{oG1orVoBT)luiBO~DKncaF1p?GiTXPo2A=45e2gvXEm}oCs
zC>2q7BIFm0el)OMs+#aM9)P2d4y3w^>pXW9+7^4TmAl{8-6MLFM{uziAdz^WG8y<(
zy7>A<EC4deCAmpPG^%?|L_`E4m!P2Y3`IRJZ^ytOKoO~HE5)tWyCXg<{5-Te2V9p5
zzRo^+3||x2GJ}sHgaiUf)FA7Z|4!z4LW!Y4bx5V8f4lOG)Fw!fQF~gc$?<SrPN3~h
zd5LY{{6c;6gDicfzJ69M9Vt>Vqhfp32oV}qzujpl=R*KQRw~JWBVirq1z;1H3M(Y!
z?EC`Jc1?74Rm}!?LgdTa=)a88K3v~kP&7@^=&1gy3s38D0I-B>LU<R=I#VXK4Y<xN
zns!Fn-B0WR)-{yV&)N;WDLlR~hXw2^Av1!A9L!SM0lsdjeB52hO7Wy^H6YS7;=<zd
zU83~^sTyu2&x7JU@!PokX2|-XeHO=-H7$NZABrzeu%q~<!#Tx!hdGvRYRY8{T^G;p
ze=V0cU<_-#R0LKXwU%U^BxMl4b%@1+%X&nQ<DG1dBmoLo8ZR}w>-?}?jH|mm`w@f)
z3wz$M^-m-8{yX$7u8#I_y+!^om>i0RV8nx?y(A&X0aJnOf)~I@mpGi%fh`5TGe&r4
z#ejp=ug7U(0Ofi!?#Sa8J@VJu%}SEt3`10MgOxo2ssJu^_4hN?Y3kU-CQSIl918z&
z)>Q1u$kIX4j3@d)g{xzLPtabKlg=|u`5e0oFSSg8&eAL^K9?L-p1Z_mewYJHG4?6r
zG#e+pq9vcKaru``-hqpHgIE6u;8Wxwp8iCDMA<OQ<*xWPj40}>%E8wxLKZ9=t5fV6
zAEm^*%V2|H)2gt;uKjiQ5q4L3!zMy59n8qKQ3on`y&^`Nwf->$g4pdJRq;QEZHBFI
zF+sgI2M?4=IhzuPlhV>4y^4nRDf5@&)9fiOGV2j11aq<0AotA{7p#}ew5GU<bxU~(
z-6!Z3m;fRQ5tv|5u|f$Cq=_@@>K;v+oc!|U>zB$I6ZW$pepe<<={_lgaSF>{KJ<?L
z>i?VA(y$L+3{|t7F%}4#sC-NSLN35T&Z!(mEoN)-FiBTZwW=`VGj7VA;iIT%paS)d
z6E&L?Bai-wxZ_PP2@~p9C}_Upk=vks0C})~U;Q$R2-tI%B^q%!^1PbWj)#CmfPwR%
ze#^M_s+r+nNMGQcADf$~bpt*<L1)GaY8;cmgtQ{;8v2RK!b~3s%<27CfN>>7=G8<2
z-V}Q}0IEgzUGZhq?AlOC!MOHWv*(4}GAjdJqD|3pY%Tvu%CR<|6JM%XVdhvr-);@?
zk4I|5@8H$u8)lfSo}E=m-X&pY>JWEOP!X&?IW}?)48>2byjVK8!Rq*3&<(alw|b9V
zec!JPS^aNf82)N$mu8`p#kjpqXI+F?SO>VJ@l@tVJOdS;xuRL3z2wCg7c}e^fi(yw
zxH<sfDIr2Jp0VRNLh+M|CP`o+b0XtJkG+h{PwdFzawL3Jy}>*FYId}WF~FRaAGojC
zW_=}caWjiB?DiP^GniDm&Yem@$582w6PX0e68snyF7gbn^8D@vx)Bb5Oe|`;dG9m2
z_g$HDa~POiH!w}MPhUK_XhnAc5mp@qVyCQPr-iDbtB1Ih0tbyP*Ns6QOBOs+!UFlv
z5Wh-P{&1t&{h0B}<Pd`*7ZJC*Gw+Y!h3qL0I|(id5=_57{66b0w94!1x6|zyQDeb#
zbNaKx_^d~)>_CXk3-avCGS#plYZ5UeDT_9*nBooT@Q8pvf5>WRyxWFa)Pn2(=F`La
z8k8Pkm3SgBWH7x(2cac#U%$vJoH=2m5L|{<&RD$SH+&^(%BRJ`S}{cr5^WNCwKW-J
z%5Z@n94Gr6S$M4@?;vb5T%GGf$yu*@w=oa<ykSXOB5>MB&o$)Mul_b|H)mCFTr@5#
z<o%`##98-9T6-Q0uT!E@@4=&-&t@%-52L)(rkj5AgZ*S~Jm9GX{6pae<p*74;K^>~
zP^B6sku}952Wtb^w@+~AeFScH20p?|6qvD2{AK@}mfIQ=4&QyfgG}<m_`|bIZ6%*v
zl4pzwU6PXG)i_7K@>o);h8*m8e5EhPU*P}^Plr5&-BlC!U_sFAKP;GWg$GQVDC6v`
z07X`)RAv@bPJp~;;#cnHCn+^*h+~McwAnI6v*9p_iIP&sP+;ar+TWBH!J$`uBYK7E
z80#Ans@l}T{nmhOYKq~RQJeCQ6*&{mbyQvhv2V9TUIY201|i{#J;Ccj1Tz1yZFy5n
zK70rlY`Gk4x&#QPR?mwbS#>frtA=mfgN`X)x?H+60vM=dM5qRA2EV#X503F~xfPjJ
zDH7ex?%)aM+?474382v94!^SZ9v`0gqUo!+wd(G>IY8hZD^!sil1;Df1v_;ea=ozD
zY>diMo9Qd<O}yhtJIAJ-<b6T%?~#dgV6X2DUr#dxTfZW_#K7mws8&Xnd920P;!ouh
zpjDqs-bK{{P}e0loj`s@h+>po32U{w`YV@y$anWZY%Ogv&Y%x(sa{7ADJit=nZ44I
z;ABhMCzW4`m+e%xV=5KB_lwCZDDQgPuw&Ev(`yGGeEvuW^(vApd$WkChL@&uRU@1j
z#84wvz6~gh&y8tW{2WRD=e4=Y+dD*%D#=W?*;7mU$^96vlRGeDuOcp3QW_NAC8l5t
zU=+-+8#{-D*b1`Z+&K1r5(I9si5HERU5)pTPunv5uCo!GywZn;%HtLem#v!q>=SHO
z=D2_Pw|Y{+Sw0~ytq<rbH7@?n4x(9m5F;)FH>p48*J;j6u1))%`|fjxyH2U8A7RNN
zdq4fOio1Rfratg9TrT?!BGpyXZH|<?U-W1(MJs>xd=FgcMDD_|3&fQSY5JkM>Yv3#
z?Q!pmZ?JLfXDXA%`f+k9<yd_~6eSqI7H47o#C$-R)1oqQ{Ss_P+OoLN6Z6lieK$3$
z%bo`#XEqufs_DDUdw1^di0<6Uz?)LWHuv2G@B$vbyjUKRgzh2z(IoWtadM)Rmr8E*
z(m*-DB&C}#Lf#lg=&a?9s^Nx=%egtQV`MzuvArYtMm6UxStu-zhLs{asuH-478dnr
z&RR@YNg=kk8y6)Lag^Pi&4<L+ckF4vZt>>4uVpD(F3J5#Assnu|1FT@GQob$;{U_R
z$H;?L8NZP6A+3N*cJ6Pp-?%RZC{kj!&8J!b9Yt&Bg8Vf}Hz_BQUu){(^tzu;-r8Sb
zX49HFC2SG!ehh!=eB3taEwObC!uCj@i_s-EFF%l%S3F}O3lVjCZpO}t%IG^oT2dH*
zo(K!f*Z>gFkdrJ&ys~nc*)pZA(arV~de}X$VW{6Q+b@%N&U+5U#3odKHsvkweyK8A
zW6NB*$4A}rW8bOPm!jf?mJ5`<)RD}_Rey%W^au%($Aaz)d;QW6)$8->$Dd8^lo_5o
zRFg-t^`9rbSI$DF$NpODb3Koa`KH0bSm7?5BRcZo14<eX#rP|1fyFz644$9v60l<t
zM&~>HwXpkg3m47*Jy%-};JH#}?dl>m7ziNM?<KcLYdhPWuU`$M%y7~x7+<~GF7F}h
znAA1^7_pu-S>V9Hz(E_nS9H)5uOxi$iw-^S+f%vECC58kxhZVk2l`L1L=XsqX7_ye
z(r<o`^<1tMe(WLZm+*^^tu82<X#W!H?^Kk$XIc~M`{r6n9HK9J1bthL!9!zZ;lK37
zT+?V@^_K2PX5aiXT(W!POX1`;#uSVod!Z-r;bAiU#<JpIaYJ)|i2hMV*K=kU<={uo
zN8_PPsXU5J1!Gk;w#!@{$?qtGGt5tRHqM<}!a|JC6db~d!kP|8qbgi4(Qi7eTEIU2
zC1%^+*%#%v4at(mz@I~K@&il9T>8CZXSvzIu4=>oZ0{VT>{(rUW-sBH!;Q<gw$GbS
z)Soglvwl^S<J$uvK46o?l&IHUJFTxoB!fYhl#;UcAPH7vebB#hi(GGDXFc4Hbv!mR
zQ;Nwg=fx$N_h*kHTsL>_e}8k!bjbe|E=!|D<_75bsNxy>4)@BL>AQ&4%~Y)T)X!)*
zE&(<XFRdV|&Zi`1!qhd>OEzki_}X=~_>uLn!^75}3LfmL2&DVYyLz^BN0MkqBg>y~
zd|Pbv`gg6DBpB4=8KPo2zoUfeZm5NVkZby3hr7&o`y!@x=d_68wI^@%@3zG4?gzfu
z06D8IRFNM>6#f<21E1zSlk)!6qbi}Jr*>T#?yXc}^8Zul_kgu^hJQ&duwTnWTuHGm
za~)gHQx7RFYry(4|FPxvp}{WA%U!vvM_itt&a2v0l7pzs`8p)*8QO5<Q@CuPvY+xw
z5L{#7;@ATxSo~q;+$&~mF$AZi(VEIfm9BjBXT-EyrYXC-{`;!uKi%_tz*Kyz7xz9)
zxRdYU4seJL@XGHiHFFxw^SEFFZh8}`KgD0f>3GLduNHdL-BB#nTyb7@Zc%RiPAEZ8
z)%AZ3TqXFXFQg3pH*03SjZ7MDtuh}Y-rMdy9vI*p8|JgJ#mqIVA_T;ns(gwnE57{Y
z{qloTvXRS{OS}JS$x!9m*rAFPsIGj}=4mDG8qcEV^;}j19yD7EV9nQbcUms~3{;PT
zM60@5LZ?~{EE{sl9sN<4k5<gPU{%V|>yLnV+#H^pxvB(xbC|quVH&V1S1l?Kl@RQS
zh@ox+xT`Uo1Q=Jzn@h7ayDu5?Bn(H<65_z}TSo&M{IRl1=4|kS4HU$=ycf(DuHV|F
z)S>Z!oA=r}@)-Bh{4V=5#~c~a2pf)-(ZQpQ?r3Wt0>kipA~zTHSWD>?3jHQWf58ln
zmjbILcjs1p+V;O+hU9XQ(F&3bOJt#R^<EJ-)99oli0dAHviq-LHyK2AQI*2R#zp{L
zPP}^1g85|88-x$bcb&?K7CdLKZ5ww50BVSTIb&6j-aTu9-)O;>7yK=5`<whPqmiT8
znSHqiPZGkU?_IeR2U=%oN&!7lq&)fV<OK;x=I-QSw`m`|7uy*Jg&VIu0Xndf>~HKv
zU0rM-+rlT4#rT*%Q+KP{Cc9^76X>TAYQ0@cjbNFhD!6_2Lkq336bhnemc%Kr7=RF~
zN8#Oi>4qTh_t9b;G{wQ!|FpWJF1X*$I{qBy>iMLC(_GVfj&l0i$j7GC-z%TBCNGl`
zZ-Fvi{Y!u@Uz3s0d_M+qUxRGtOQzpF*q%UPQ0TEAnsHCm7*#o%1Q?OF=1{5)N8S!8
z1N_v6LbWr^VbfReJBAca$ac|JYJ`6qGB69LK^_sovyc%dk$q;QL)OW@u=kJXKEd$^
zh*Xe0Al_Hn%-2baol&+NQlcX_!S>g$?~LHWY6`4{n$LPpgwg8hDuwmQs?2LTjm4Rh
zA(T$uo564wp%>}y=5A0r7Id-g?ME50P|8O;7uo_|r!Sr@%4KqOV<~L!2#&zk#5Gv8
zfrtBuU@X%2PP+9;gb>4X`%t(vF9A@DT8F(6dz()I7T$j(B^Q2#Pp#kPQD@XMl*)M0
zA3^q$S}Y>J#(i#~?h)q3#B7&ZQK;bt4)PIM3BWQhSX45fj)RXRYzi%&_KdN<G9IZb
z-#kPsmbV!~OUialvx-i4+KP+_i=ChkU?%Q(frpT=pMI$|wQ#8wXtRW8;4Y8Cy{Qd5
zUUp;o->s8mzg?s%<tkQkPZGFl^uyKT^%R+dCTLG&#W~4{QWUwGmO=DLR|9tD8qaZv
z+bY?QPXb{Ik%Va09PSB(FBIB(>p1PF#G4Nl+TJ0Q7@h2S8?MDoSD55?4cN;T0W&=1
zNqPvYjv`h98hJ)6*w?Lp{n8H2Pkj8sN~M}=Bv>@Rk<~pRHplrjct>e*`hCY>UQ|y3
z25<$D{>opG0aL@e81ZW>b9BeZGt04dt`A0M_tnnuPfqV9efGWG_jwoVWebb=`eV)d
zs&QuskkL$5jM5~eg6RB*Ot%uz#$M_7wr(e?NR7XG>eAP*B;W4mh~}Im&Bfa`FoQ-J
ze#mZ3^S-QMo&4O(TTds!rtkS|tdd#zK72i@w86Q&dHAyT^PY}#dhNBS1H4gKBi%pS
zUVg*cN!B8B;$}R;VQT~RItBBZ#3cVM>n<Pj&^1SJkxzc}8P;C&1~o*>p7il+?{$M5
z^l8yi4LW}UDGa(!D<<M5wh_OmGVvLubC=My>~|Sao*p-S;i4IW&3|68KYvbhUJDW^
zpWN<4hMzbL@v0SSw(GjjO)vg2KXDVqf^Q`wrJ3FSSn}KmJU_hc=$zkr#G`J{ClzIz
z;82mA^<rG_u**RD(+%UT-(hi(ZT8{mx&40R{mTq<z9ObEh91axc(PtNIm=}Y$QT|)
zAsTSTehQ!0tOd7)VRgUca^<s~b6t;bG#@^WY@cdPBx37s=X~8~JKVqWD<xlZF=P7E
z!KJ5G3h{Tz4}o+U_DPG#=nC1R&jHDX5$!}t>dn((u>-Zb&-+=X?(8mp3;ON?2ISrh
z?tZu@_WjM{)F68D6jw0f=A$0(eIe7J=sW%4rMb%z5l7O*{X2JoL60E&7plg;77Uud
ze2lus`GGH*SScIgN2=FT=kQ#@`*M*;<6N9QvgHp?k@%_Y4Tu$q)L{JgyJe+w?k?zi
zuyb|Vojd)WdA}}xw;tzp(*!7xb{7Y$k6{JV+laURdzHlXs^&X~dJcp9%pSWljM^D+
zx%Dz4BQkZOE-Ql{gIK)|DW7Bqh1$_!id@Y#USAOMVJWo{2<)9QlBRr<y(Wt`Pr6N&
z=IK1PJ-vrHEN<zm!N9{GRaQ07B`jj$HqeFoDZxb3TO)<f*+updGpoYb{dKlYn{i{0
zv7Ck*A!XSmOqEE6Q9&vLt|~O7F?igRgSk|&_Q<QO{(oG6{ei+W!|V93b+*IDjBRY4
z<Rugofxl>&mR}<Q<@oZpH;v2s-no6>3F()h_ZF0w5YD7~dTn-(0M+!<ZI90zZ(eo-
zg!NelJsB)1b*}t(m2{@ZfNSa4!q4>|KaG##nl3WLB}LpFBeS;b_tu<+oevv5Yz{U(
zJo-O649SN~<rmM~3~9esUV0_g{J17a9EL3FXir&*`OkA)NwqO_!B;sYRzi5DNkKz{
zX6WX2<u7W_%*>1kh&9Kolx!+diiY210{c`HIoZDN>#sdumD{mgYR~Y&X5IRzKCXb>
zU2D8P;G96!v4NSY`@<ozIr<$o_i5`$A3Ja>W}(6Z3s+`8TU{<FkN;BT*9N@PnO*Dv
z0x3ub3kZX&sC&e<IQZR#LY1|w9?lv0)cJ|Wa^?M0*-F9wz4Z4o`dJr_LB92j4&lX;
zl?3c%g;y3kLI=q~Xw(k7eU7cu&+p1h4qrg__rzbUQXz1r-9xSn{T8>q!=!hk6%C=H
zb=16d)c+n?EkcEZjHeH57}d8sK0b)Pw6w1@@Ms{t4WFa-Fr5LT2-)s(GcmKOa_!`G
z>r0NTbP}rf`6cf&el}dWP|V^00XDR^l!23dH#ch{Rjg2G(k1XIY7ibA=PN(|^6N9)
z*97DNXV(hL$trCUeGKAbl5N6rdP4N27@03m%Lt@`3o<9)WPXl_ZucoLv^&(#a5z|3
zqPUK2ybBhmZ19z)?r$>>4>B_tU%%sT>6@AhtGE9f_rdZuxMGWEKgxc(b@S$+M39@W
zDkbxBjV+O{hZtR4s^GMmN^e1hi3=xDw2682y*$RlhYjAOolPQ^XYB&5G<GWxSd7Xw
zFKwR!)(R!UPcXx(%tAk10YitWN$ID55pCXy7FLEB^jHo$3KpajZQvb!qdl}{CI)^^
z+1$CVLmtCb1n`>Ja3SY`9^uklS`~J`7`ZO(&VW3-$XR`r$!VMf&y5TyD2kP!c8nW#
z7FG2Wy>FF?%Y1t@<Hi%Rbc$=<NFGxVQo1I=yBVY*3fnUnFtGK+zlY~#t_pJAKYQYF
z$6SAZ<XmZf`}AA$@;9Pkb}PJ1Dr+ZCwD@o2q<nGX$0uJO%C0{_K^&ut_IK^(zi(Lb
zejavH)&mC4!-^g_mn;EmoEW(9u-jBM86~Aio=MjqUaeEs>y)m^#Eglbi>o@FIzHW_
zeH`_*EP2UPFjFf;BB4AXhznv4afdvUAs3e>Qq+VLTOW}7)vVo>ty|#AC{NJ8C^Xxb
zQWUH%DMJgN_GPx40PiDXs776R!b1R0cHrDzk$KdFW<{?ttTcyMYZ^(mpC$&KWLmYm
z8@TMFNk^?#^<4P`+rBRxTKuf;fb)DQLMxgG_0v;K{eO?(k$hO78A%U!x%cgM)ZOgd
zox}c*B{#NmuhZhFyBok~MIQ!Mlu>u}qm~DI&Mm3V&D-xDro@LAepsVX2ik)x1da0$
zGp|+}%~adaWm2+=%63U&%;Hhch0=#gyR(Ui1)d|GqU`KcU5`?$GaeZC9Yx*l9m}sa
z6M>%3cFUcURQ9+)ST3=_t9B0EB6d+UqMm72!@X|yjhJE#0J849N8|$d@v{&zLNeB6
zVkwMK3+e$UUXnEs)!hB#Twqp|(!q<0*Piebny1_b`O`QgI8|{moq|>c8R-{eMxJ;i
zX<Z>NzS-<lohdeVb?eg)k)mTg`mFT{|J0%2uX2P<7GECQbUW<~r@K4)?%f;@!yjh9
z2ZyK&Ab;stSfF}a_hB`U(~}jDS0+CRz7`=i=_+&vfprzhV|`dtxPtP}FLZ8F?~D~p
zhEuO8D62hdC{r;aI*vtUi^R~wI6R#DKSR!8EnWkMyO#0uoL<ZVK}~izvMH5{5?Kzq
zUXp#&L_^$a*t0?eG^J@+;loqe!49(YNDTPM;KAEc!s<bG5CufWxb*mWyw9^OWN~ky
zEagGJca<V13vG<)`dGHSN1;MWv*#69b5=u+oF)<FGlGdmHO4i6jbsHnGV;oQXfI3b
zQN?`3*z8yQ_s2(4@O4g>QR^%azx~4u@dA?c?Xt#5@Y-XgW>ZODcmV__$Di5)eGp2C
z@_Aj=xu=vpQ&Yd4f_c9vKx^gG9`vhp$H$yz^Wa~CcPMX6qbLLyg@EK>8{g_L9gc8b
z>zBaWL84NkDc{jhp<QihS2QhcY4fJ|L*3q~)hb43P>89%zU>*BO%e^m91{NBD6Ejd
zt9PjL=5RK(IVlX^+uW+svGcq*aV*^7Z?>FiR=U)d)%yUGQIdGk4tZf(R;n<oHu)}i
z<PFdBam0*QI*JWW_g9HoJ`eT)eR3xMJ>tCuP`U)22gA4o`z&rb8r;u=9iX?}SK*oN
zt58&XgSC9}MCXknulJu+odtW`Kj-4VS_gKCtP;)n+A=Ch?HqPpb4dbAOmZm*o(l86
z>-R{jYS>w~l3ArfjZD@LOz{K6FZn)`iVd&_t7r2kSfh*E4e!`RLB)9~CJtH1PS*vY
z2P}D(HZG6Qo!$+k)X&ohPT`t}5AWIfcMt{!Rsm0L`6Hy9tPLyKsFN@I<_SU$?2g!Y
z-Au5$-*A^c7{1RMcO%g*)zA2J<ix`p;Kx2F?nX`WPf0Rnia{wS=x6`&S_mVuzA)_<
zgFwd1@PTq;A8_=$9)AwBepC%oBAJ<o*d;x2gECK_9R5WFx#wE3<FBwNO5+)w!{2xo
z`Iwibu2<kBK=K>NhlI!4ji~**?X!}rKDOC+1-`I;kJ#0Y*kEw2G$zvWveFKLdh$R}
zEOqSI%V%YW)MtSOo34{L(1D8-&fbkmi6)I#%cEPEN0OY=Y?H3;aa<+7@%N)%&ujN{
zOcT$>PsGb}7uXZ=av~}RZm;yn2$Kn$Cr10Ih*dSth$HVlKO5pef6FSJ_ock#nn~8Z
zvEJ@?F>~6wgw^9GOGoZa6{8;|`P9!;Pgx)R^ZD0U=|C+@NFFte&#?gifp;<O-722I
z(I!*Rj86R*-Hk_NgQm<zRdL);ECvMJQ7DEC-u;$$(bD$RdgEEOppMNSG)JD-I?`{H
zpZbU_!U0vs=o5>Zt7O+P7gmk$1nXR3p8GYfqxhMY-jqlK-n@eR-UU@3btXf@$+DXF
z?sXeV!HEZ`BSO4+Q=|P@9_9Bh74@N5)`w3<S?D;T>~A-p{FZ6FOR#Yl_vvepGYDf_
z_xQA%dXYHBCd9nd!-juPoyy!I3q4upeYe!BjaI%ISF26vN?hXLI0j@{zogQ5)v&Ez
z2AaSB={3L4D|%H9f%~-NWI!gaSf-e4X}tND=g&y~=66CTEt;NAKl}HYw^#7h4mQ&|
z#5G&y8h=$Wv~I_?s#Ke3S9?dPSa<$&{XBqDh|l8@u|c&r_isC9<|9aht7K#1QBENq
zYNR}y6s|P|9To;;Fh%#JIq~_R<Lt6x^1hjWH$voyB@Z0>7MuL}HF#PSVdmVd2WLzE
z*Jr71V#a>0CNzE#E%#~%Qh&v-VH$2Q&<>;}ry*NQKB*@|kzFT$l{PEa{3A!v!>ND%
zawzggEcMU0fB;utpHIKm=ac@LL7b;8E&~?#+Z--Cx+9#Nj~n(3>~#Jp-E>@fOQiuj
zG1d}aH_N2&*Eq6)ex>o6<hF0mDPXmKCpG6A59T!nG|UZZLMmkgVp^H`0^hSd2?8Ys
zv<APPW?BY7&0%UFIP~{=TA!15b*=FGo573n!!zW2WOU{Q#-|<dd;MF=c;rm;*l}fc
zzl?8$z(&{8Ed~@&-o?H{ejJs>?(ROy@574`_QIE~vJJ%s1Nh%axu)5+%o7$~gsVg0
zz(KNhq~a4T5*6Ne*U)}j@8O)d#-2yZC;0KtuZ3@LhhGF(D1*0?+&fJ`CZqnvArfTP
za&RgC;4S&g0<<rqidbjtPf@y~wLH%LmexEIwpKz^V>Yoq4r($wZyenl4OvCE`McQf
z?Xl8wFgn^|M8^TNoLTvOB{FO>8&<JRhDGTyA~V}hx2{ZFbYn;(ZwJF4`auQW@xqu5
z;NuHB`nP77#e$k)@#U~cP6TsyIXwBVDS}7sA>mQUO6j<_>`JU$L^7}6&0g<EZk^>J
z2@wnKp~G4!#F%B(qu;KIv&S>9pF`c(8)v+>j`4w`z+%_sxmV<5#IDC_)%hn?F?nV-
zvQWsc4EkEGC-9psyHwfP4&m>^3q!9R;IUC0uuX69bb>YpF<cAqO~0p)DziuQR=^(d
zEced{lWO1tN@JUkxfI(Vc{ovADfk3ZM3&t}h>F7cH_hjw=Jw~3(!Am;W3!wNhR3xR
zvZMX%VU^qBR^Ny}C5)?ZSQ$zBSP;R~E>s@7QK(w;pq~#o>P#QZ3aCHj8WYj~r$sG+
z#I6vp;jB;&Cae-Hx0VF}<K*POxgI=Cp&v_|<E=#uR&FljCTvoLKL)4fC|&&YCKd>u
z`Atc`E)S7W@GBrLWwfYE9}m@d(d<jQbpO0gWznVm{mPp#=fk4Zr2*#h3-moFI$+3+
zB;O!B_%o}JT}yIyS~GtiQfShDet)2uMi0&@TdPay{mWml0F*o`i>Q|aj_Ua$@AlH(
z=hV;-QD<|r;D#v#j!{IIKOZ9B+cWy+c^(U&cUo}>hwRr5IP!pSh8Bz;;yqq@lB5@w
zZuNN^tZCZ@{jzK{vaE`-YC`?D)!a<W6=p0iY~Zs0Hs<3pgiTUubVxa~FZ8}uM5dnC
z&i+Uuyc`*q_9!`V@Ic`z%FKEf4yayvzTXHh9pjJ;tbUK-#F=f`I&C~9se9!WQmP?{
z#B6(tK=6ETbN2j>_`LL<-`rgouLpSLT;3L~jEocE1>W|Q7rjUrq<QDb4lct#g*(;)
zxF>PBL2p?d7Sq%XgCVatj{caT>gv3YF%rTei^RgKb<oxE&Z!Aod&kUgV74kYC#!)k
ziiM2~`24lGu5G~RP=H&=x-VT?IEXvNXC#M^?3qtKp=A-woNg#-*Iao5=@+c#{l7gr
zTNtBDfKUghPE+_-JNkT`^-TTOqNs6A-hM@DWLK%x{74<9Y;=8e&)Ms1p}?>cBFjcs
ztQc7A<mJ}q@Nwamb?Sq=NTaUs;y^aBCUS>Vth9*5D+bypGj{4d>k%Iy^^a-&35;bZ
zpS%0(hk#=A!9v-IM*vqA(zsLILSCJu+x<)|Vx*$wYh{vSAGVx1ahc1_-CmD<*}c~=
z3#1gPfzMjPz`>pof*E1_Vo&>QN-8S0;5$l+TQ1;GM2w6mV^hGg<49`8rNFQC@WUB*
zJzVh?Ndqqel@N=6A@VN9XE+d{TIO-#5rvLufB~pg4fM?kwz>!j*@~EcIdQ*hKDVX(
zrdG<BH+?rWeJ#9Go)fpJ=@Heg>ybzHj1mk9<G*z_yinsl(^S)6NVd6_yL>BnD3n-l
zaB(c~6IreUyM!!WQG-)gTug({&WPh5I5NJY8_Y^oQg#UjF_FQ!xRe627ua7)?8tC|
zQB_Q+Do9u_Sutemm<|msXGKi`y!-HMlC%2|-XJt2)tI+(T4@U!&;I#9QfZj~fD!o>
zhS+laJeJgBDp=qFu817ap~6dLi)}Z4?V|hmy&S#_6PG6Pt1a@ilB`*!gAZ=%$5Ba1
zxDmm&bJ-hgV^V5ZIMPg+jLs9!`0?~_ic4B9qi6&Zez;MNGKn9*R0Xv^n0RRXmXoPg
z9jHO_Up%C(6*_eng{rCxAt=g@tA{p^>QG@#b7$<lRlljPzA~VH{uEVuG{!nh?F0F<
zN2GYFJtaUp+jQ6cUh!1W&;2zgXE4NX6iw9{!jW3?zv&@E^6?rD(OdK>kdk2e62ypN
zhUc$YDb5+z<bc$iQ~L%TpHV!!8eTaOj@~2R5neuru5W_DA|dsIt-4baNV3I<5%M`{
z-)Jb%lxT|okG-$_>*9Oj6%bHL=@3x5yGt4*q@|@BL`qt^Q@UHa8>G9tk?!skfqOpq
z{oRN6FSz>>p1ACuGiTyGu`}k7@4J^UA3c|@RQT+ZyE#L(|D;i_^XJ<_IAQawnwSdX
zdgQXsy`-n)*2lXzq#>ta!8xx>i6<)w>%g>|wwr5+bNj2<67)&kOo@>`%iYKyQ-I?W
z-7YQp4rb^lNjh{kP!rwSYcy1Ovj}4sLT$wvYL#5ArmyZ!;kN9O;hqo#kqk{fL`CWo
zJcnb(upRPQ;BqZ)Rry&@?f{$c$<79G8Wqo10m06lrOG~eLdK@wQ#_H|lAb@QFiBf;
zd?DozoP|DtE>9LA&iF!w{-=bgPvH@g0(~kszaE=L|FRvgkB*93+_XIg?@AMdW>ksW
z?V9Ilp1Wt5XXUPA0K%#YZn;oh4|W#GUfHZ(a5%|qGk#QA3$x-`KB3Qyp=zue+T1NZ
z=shN};i_d#9jQ!QetsRmS_+MCM*}`L=sGsMUwrINJW82VYVYkNKX76Yy(No$>+X5l
zm*VqsA)mwcy4Ut}@||}CQEix|ktrE$7Urwv3l7gW)R!FE=>Nnn`$_DcZR0mBpc;ct
zS9--+NtmFxxL>eEHe&!GvQdCalt!Y@HaHi<+b|rcB`YRVD9^EUjA?b=R}bH&Mo!nJ
zRYpr37gFo5QyvR1xIw4od|$u82@67m)6g6s;%NTLiIV9b5PKOMVVY>%BQBd#Fc9J8
zbTQozee?3JqLVqmVHneI<13~_|Ku8xU4Cdt=DUsYZ*s#tQFr$mi7na4<w-M4S5*~~
z?n)iowz*48+ekcaj1gmYx~RUnd;PfY4wW57?zi8%*-9qp{WbYKt^WIAwbJ*GWaE>>
z$dts|09~przmQ#QLwq{-Qw#_+7_cB3=9-l@A&;HpMyX}CXZYEfpS^bN@{)Z$M9Xrl
zRrGv$1J*XpQ0iCIkC6P^mq&O`sE-N_elYdo;QKUQQDx3~>+bjNbnXpX9FFIn*YAI!
z?ctnnH+dqDxD!ZHdF8*=KfZZqv^=y^C!vX_yWG3!`t^Q4AzWD(QM`{sv!<rogsuGz
z_uH>Sx3Y^8;5e+i_l;t^e7UqPzVgc!*YDRI;|`#lNzL)oapkm^TmSiOk$ig<=)#G=
zt&aUOSo7PRyJ6Gv*3w~1K&fus28-Mo4!w&7PC!_V?LQ--_jI-cCKV~=#zb3IJi_Lp
z2I+BXTWcYWWuOTQPP!FJirq~$jC*TWaZ*^AL#u#;XXW)$ALfJj?PB2a_R%nw8JRGJ
zJ*jEfDXX@r=1?^52A%(7@~6HB|6-z7`*v<?1R>g&SbG^6_qUIN*8}c<M&Drg>_Ufj
z-EkSuh~M1IwG4r9Ue=n%6>Vox>g6o668HSI98mYq0uaUT3*D{SH;$rgn1?pQ6)c)A
zWoIvn3j0*jf^#veTjdb5>Aauzp*lxOLej|$*w-cLCA>OSVc~ns7I*IijIg#yS(e>e
zr93~sR@=s_=4E~H{aF7l&YEcx&ZM=)JGUt9gd0Ti^G`q3!bIiSP>fesc5H78KdPKq
zxV-faz8>8r8;qS{a(jr;ba}vo&f0O^yuaT#VhXP%bU{Cyi=J6Jzq7!9V>r``xU(7d
z7XKhK#^^3DVvhnak$C-<uGrE2h7H3y4hp`~*N1199Bw-G*d+hO*R$eWpoDl82`qD^
zt69Vbmc};dJ>-GS0M^giDA2a2K1v(R3XkmxnC8&IzI}D>j>uW0;3&rQ`liU-C*+e5
zTtC#%6!t{Z&h~13%V5Uo_5^u)lG?%F_rKEIk2?V1UdP8?JX^+L;EDDUQbVa#0E1x_
zt(3vrc|XT5E;}FW_c2QxHCflk0Xq&cdQ@Z(4^q&QAXq|1S=B7u<cm(U9%ziHbaomF
z{9@}>Pj@w&18j#D!72Mz!9K+hHrLG7biVlg%_==GUSVf5wpS5g_#7T8P<Qv-_?&Dp
z#6LwC>Fpns4Eg7hEHiYf0W1hETZ_Fw#8e3dzZ%uYS!tHA_|u$iYWe-+$J`JAT2$do
zb^8!rgiVa6W0?5$Lv>(=mnViRlw|<q_Z!(vrL5zyO))6aIPccaIZwhyisKuk>VC#`
z@WH4Xlza)$Y+m)zD_*-D{|w9E{tlliuL|ER-&s!*10-4~fJEBK3i4v=f{YM5s(&g@
z_AS7{dbl{nV+dh!uZrRP#D3;G;TaP&`?ZoKOFR4??Vcp|o(|KpWKu_U47%ks{uRaa
zo92TJuw(%&0|bhn6H~4v2)yZ?4W?IdeaDP96zc#%V<TE_>j#>A))9tBzdaer7pEdm
z4(V$XQ7%RUUlMnbSQRK?!?(eiDOf*Y2mRJ5Fk95rh3U?E5qoW<EHhHT%vwQ~a(v4k
zUd9>SM?E2yq1|j)+ttp#A=F~Bj|36&obVC3FRCmm$bDXCj)Yhc&pt$MYIO^TX=W~L
zwh(JeB0#iRKrW`l2|_tiv{yziqug2ip24ezF}q==<)JYI;JV&@NPzFws@s@__pZJv
z93ekJ%D1gzmg4D4urLN=9AuOke^bpi#7C&BhS*+#Qh;Kud3=;I$@BYuTa#!cNh$ze
z%Qo}@k^v|er#kJH#O~XMPbt5llk!xW#Z}{st?{A4QR?ovL~yQ|{gZYDRu!@!X=N3d
z?t=H~Z4b&Iw9A<u1yQnCYsrU&{mDGaMqBOLMC>tSUH2QE-=c-Xoef@8=TOpKk96ak
zbyXn*kKZq~nD4(&f`b9m1*Y2_vWzQB-5%DvRa&pRwTFaEkq{l#%L@)zv=CD#ZvlIq
zUg*QkFyGX?IUyhBIohvRF`79slnOJBS@jFl-uQFbB9_xZ7H&(4MgKr%RPvW^Yuzj5
z{3Y<3Or;`$n+A~^<X#kXAzE}<nqfTZj2OLwDCU^~3H8`pt#0t?a3&$6JR)xC?STQC
z_V?kpHkU)ktu&G7-G4hz_HT|_gY&Kxk`+Ok;&N2KA|_%hNFhBB&hOKdM|G%2P-()^
zh9e%r_RW5!Tby>w{uiW9e=Dnn<`x%t`mOHD<*+`2&{6$$fnG3g=7jT2H-wvv3|NF8
zZwE*nz6aL7bxj>UaC=cucBF8G=gvj4?8-1mPdA;~-31A_f{1#!g*zFxw43KAL<fO+
zT_(KT>MI-i`8%e;qp$oH$u?ns!uh?_t9}2#eg!kFY||4^ovD6{OZCD~{(}HyQ4niL
z$rw~AE*YZi3eJes0Sw5+DOS^Xg;RX<%u$Sm!y0zk>6X9zGUu9+{ruj)E8$e-^t#Y3
zVMLT6GkWXUjhNyUFJ45CP;ELIH%RPaIY#q+RNySCe&&-(K8z2q+LiYC1hnqIDot=)
znOHd(<FY|NU8mz8NezB*U|&bY8`{eFK~PBjJaPeqzE4%Fh<L9zCu_8Zmh$Pr@a6qC
zMDjek_id%LT!VgHZ_D(b^-rM%p*g&24CG1aFo{q2dX|_}(m}iMeW9HC%fwClar!zI
z?=~H%q^Z~H7CK;2iVW}kx7p1|*-9Z=7FeX1#w?DwDS;?8sf}x%Jn&Uwocg?J+v1Xu
zeap&MQ^NUBK0A1c<g%VO#$Purn2jaSdq0QCCl+7+0ON;jgB4J=1-L+PR51l<YADcE
zPI8q>lxvg{_OrW(2E!eUz4pWJ$y|4jI9ldxzYpCEM)y`J!NTDkhau`Fv%EMGHh||k
zBR|1(0GQ5EeV{F^mYsN`=F+!bJ)o8%TggukvVM{6YxpMFbt_>l1{=@&K0BV|V~Aev
zACt|wlEGXS({_SH(GQ5y1Kg`RwlPrzoF_9K;M6FCFPdXJQbKMyvtqUe9c&9cdjxsf
zk5=kM3{UvHB@xdw(H+YYMN3ltV#FQ&dGCFTPf$Ais^k3jVnx=5Q(J07aOVwg)~;bJ
zaHO*0k1>@1n4Zgu!2r#2wx*jsd^suT@IkG?Wu&Y!=lFs`DeNS4<wC9$#IdSuMvc_y
zS+(z8b(*z{TsKyXr&o|{&LDJJ(J?FO!t{Ud)_x`Q%9?VU(*5x<%XZ40!Xd1vES&4<
zXoLb-Ql7k8R^kpY{N`9s2=xB=d5^Flw&U#$d;rkER>n-i=A<fHk+VusbLtMSh3THc
zhy6X?SIQKF&)4s@%1@{s+%7e4>i*#VEqceAI6C2xiHV&~pj(1N39Gt2+w}VF+@dDa
zYg}Sk(A5xR%(A8)a?HdMv4A4wln|q-L5jnnyj?iNU+m}H9L^vM{n$ud+arwh>CXL#
z6xI~V{5OAhEya=i<A=7?uTeaVlB<1D|H+UqM23uUiZS#*y+!`(SRsspQBPJK_R|X3
z%y~6IBZ8($JrEL*(;Tn&b@<o}|BaOH?nA%Y!~JRtDcOo*wK;${o>V8*8F-We<t3|?
zhKxEy^s}`%6aBPgS-lA1^;(`^F_PejK{BzB`fOhqP#ZnVW2~abZccxsJ5BE{{LxHe
zolx}}p&QQ-T2`0td2`Mh58UVTpIaPFY9Ne4HmU?upZcFS_woH0f;C(}-Cz%F6UdPF
zl}n;JBB*eNWGu^K`=7EkE6mbv@qcW3^y&M|dDdIt#qsgG$8Ue_UM~(2j@k_;K@P=`
zDQJ4t&ztS12rk%YHeiG;9oe$S-iVWdEPLdj0_Vxk9SO)1H}It$Y%CWq#HNpKX{)c6
zdR*Oka8|eIX3D=&c4kl|cTidh47}ycudg%-;H1GR6+;pmxBQo3EwXqDUtP4i4#J%C
zfrb+Vbm%y~-%#KeSO{!qjW~1VIFr2woNJlYbM+3~{#Oei&F`iIpAqbIbr5$mcw*jb
z<1Z$U%<D%iOgCB;={B+mDys9~6YH#0w@*UbRf2>MLqLR8a;3Z(HTeqSe6H=9GYu!$
zqSk6%LebyWYG+T+e_Ob&&SZLT-<MWl0{3FwOZ{2u^u~1g>+Qkm5t13yB__Hc!h!ai
z%_37D3!E3(i$DH}<mr?5PfmybYA|=uj-|Vhh@_`Vs0@X8B!dm3rj^Q9xs0S?`{&kd
z#L@sqhX-#N#&G$UI%z6K+MKYix_1|UzB@qmhu57F0HdSzA%2fE<P#rsR@z@d0I(dp
z<q>GZNNpGNmCuWV`8%B&_}j@?wDlTm5q<T4HKsQ7Xm!u@Xm|9kyaOLQMUELvqithl
zVUif8wNBO<5A1VZFTLX5N`7?}d+oM1l+34@Nhk?PphHy2?bEtJ%t<Xo(Y5#`VfktB
zx1-(>Hk4ihA4*B7B30*6kFYsT3^36LoO|)jPf@{0HJ~{~_;^=Ui!rKOM~CnaCz+KX
zE*ALKb=xEI>j%Vv&l6Of;lJ4*U5@4+5GiR>PbHF9t-#owpGfmwp)hRY?c)JR)1KY(
zr+jM*6_4Y!OibU=z0Vz1ImZ)sP|vB}x%VBHeY+D?fTW(lGE>?AQFoY5Qm|{Lvp#!R
zcuG`KknW7~7S5mCAr`tPx@GU@mPu@PA80Ke%c6cciig)M-Zx_)${+Yw)fI@8l%?##
zImGlk(jvU50*)%kw(q^kOi#1>W|FMPxgFrz8}u+zfOxGifBpII60O#kc*n!doj_aM
z6XYSYm^W@~rA|v<T1K{Dw4Eni>GSloue9Uq&f-$nNPr20u+qg-06M+kYo70YK@OOA
zPr^(&wammzhk${SO~CCx$=yuIo3ww^8o9o#?oPkBu<GI`_eMGHyUjjM@t<cKHK3^$
zNA<-S^ge#-w36$`ol6%&ylh6vf;0$Rd5Hi$BG*qlB~Up!Fs7cd<DRb;<_g;D!<>?q
zHtx1mnr{6LM__9D!+i7*$e(Ufq<`4U2(SIHy;1Cv|Nexmee1oQoYyHLPTfT=&KrX7
zd*SksMGFN?NJHYMx<XmFlujYhd`7=O3=`yJB?;z2a?GJX1iq*`x=Oc2ttXC}VQX)=
zh0pTuV)gkfn-v#l&mXZ?H5f8bd!tSz)=wz08o!5u!JZX^pS1Ex$yJyHOwHihhnG^y
zcE|uPqUhy%!6(|MvN4^tXI|h*UVCqk5OItON0>cK^sxA)<pWs)mu2!;UsB%pt?Iu!
z6psmDNRZB_sa#xb$zO+P(O!|crXpg_-A_J!KrS5soz{MkMSuE!Mb`;{3@DxJ^9Q}-
zWz2Aq_{=<sh7nz-TDX)YPn$Dp%ZYw>%<T@NcWb>+J`J>ee1i^y3gYuQgv<&rO}e@C
zKc!Xr2?1Pd{b&)sz}4TJb6^@WZ7|I=B{Mq{f=dsd{1#7$1QQavOsx~3fs=yE!c0|A
zO<#jTiDoi^FTGi9kcnUTE*Cv)do;7;xo8iA$^UZ=UbgSgQ?AB`6a52~H!#?vl-=e~
zzr_Y9H}ib`NF6A%alltIQh3?A`yV<Joj=TRaOUnLAbkCN-}Bde{S(IC>hdCIcHP&t
zP1hSP_a5D^KAa5Cx|)sJ5B=3ujd4qEkVGE5HA3E+c0B9$7`wMeQm0`?<n(u*ykY|@
z?Hs3qooZ#oxdjP{u^?Fz#vAqZHy`O(4?Y~}Oij{B1*xPqdk?2*XD7`~G}=Iz;RVq!
ztpcQ?%H$&-Z^c}~+Cv{~{I3{w^Vmk-_KQ|~Wfc-30%ongKI7}xSN3Qcj_+n$d)S*(
zV6M!DzkCmeV(5@gmi=-U(`oc0L-_O7cEjXahji8T8kKlMb6}=D)_KZvl)oP18*cRJ
z(k_lXw)E*y7^<{lZ33XJ;F#<gJvo=@>WiO25U_;Ks(Kcm2XAML2Aze2#P>_I<-2ee
zewpEjgDiEb(~5Z&y)%`ewi(XTsykWypO6G6{9;jh2AY?~L;RQJJ4Wj8@a9w(Lwn0R
zClnN|PBMoFuF!g(x9rVwgx#C3x_)m?B)cHr(Yx{pBQ`kh61We}4i>+96Q$B4XNn!g
zo)2XXRn@b`RjPMv>y|=_-_DU(vAk^ZB_|1hHJ||HB-}zF>O@bB21>(U65k}}Ao*|=
z_ha7#PO<6-0(5@|7B{q0X_v@BN!Wj|&Cw`gFyyuI571kB)n3xrI_xz35r6C=CEPhY
z&SCCq{Z-lFbEdG5N6>l1S=K$5ny$;(0pp_4G=v0|-*3h|Eo9b*!^`y_YZF`+`J+^(
z3BLDn1TApe?91%kKqOh;6t<9>Fv~EU4>2fTMogDLVs*zT>{6n$o&vShyg+GV{9iR-
zs_w=u8`1D<7i@IUJ<;aTVsHi9ZN08<ks>R4kbtB9&QbdO9ZZ_@Ku7o0-TBqs8N~0D
zRZ{0fsbY35u6O;MLc_s6TVBwq%BGWv75TO9;O80vPUDY!OBo_<fqzb_@oR{9!o?pM
z0I&;EaTW)QU9BYy4A+r_-D$s~c2Y;6?)hfH$Qd$LCE?0OMBuuYXMU_v1o4K;UaM)H
zPSJJD8ss=o9&^%O2^U{?@Ts%9S}F)ryaQj$oh{h9txefEQHB2B=_d#S6-LXlkMBQE
z1!_qGodv#c$Hlrx^k6PsJ0<Tghs1p!vg#v=MxF-u^S&&e90vtg;(zLMffJxnHLV2m
z${hNt`9jwEk9D)0Qjkm9BVkP`t}f9qNS_(_UabAo?;}W%xI)6Wu!RM&RB~3l$};2n
zI6F-<ONKiv^knZ-_11akvIQwLuoEw?cf$zLd$sox2m^K;#wBsywwyCxD@V2Nc$?2N
zU%iHh=?RU3{dVNqr!6;U&Gt!K5t(&ya7=~_;&wxH46Z>G8z}EEA~4l`Ti|!uzWI6d
z<%U-4hdJZAgK}#|;vajTA!p(RUFD~>HCy?5LJ|>pxZXo<F0h7rVq>htY5C>brwj~_
zv~R9l6Ok)xUw7KlCc>q##i72x`-U_L=Tt`5Sc<8(+WYg>f4u1jaV`d=q136)NMa=C
z_Rc96Pfs^LPg?t!Gi)qO8SafcK;)8YjaJ$LOkALY4Js<D8V~WE9UCSnmS$Qi`^mMF
z*MIb-H21SJy6-pE39pYZL?oB~?o8^Ob1#&hj16z)hhR9cFD}Mt=#LcDoZCBJ8nCuS
z2bzk+14rDEs|DN)!F`+%cmR4>sI@6kkPJYkLmLq>;1B%-(*;m!2~6}-41caEULhUQ
z8WRoY!}`sWG~3ys%$q_W??xioKLQeML_rEjScBSnNf?7}wLFVMMx7sc!W*4~3c^!O
z;z`yCIq#713oNq}FmjFfAT&P(pL+rI!xw5Bw21i9X%!~`G(rZk9O-I>D+W6-YrQD{
z4rBUky@SulnXCTW2COQfSr=#z7r+Z<cT${f=>G#$jz&MQkcE<=KoVqLqOI9TXubyN
z^95pFmFS_yF;CkI(V4Gu!&tHGp5Ave+5;y0@A1wZ-rlYE4BvCl*5BsVU;fq3Wyh|}
zKphL%m;sTr0E*HC4CxvcIAX{sKE|{7qYb-CFD>1^qIMi%8aj|%_NJkfqlAq*MDhhG
z){Y3>ytODyEV-h5_@Xb#3HJy8Ss(&ZMUqb3^bPYGByk*MP*DP~aMN~U!k^xAHJ6!3
zgO$;bJC1~k1YMSYUN-;wQ+=R^IpcawSCrba?qf}0<-Y0K+;Xon$74tJO~2mY46X+M
zh870#dNtXQdA{{1V?Y)A66gH|LFSe2S-Pa{Wf%U!1DsJ0J&BgIL<855615{(M>3uo
zRIsV%x3IQe<D9s)H#je$d6Ap9BWSuZy5e`iti;vEax`zUj~u(T4mYO;?!q;fFhtm)
zDyqvbKn)pRm396TW_>;K?8#S<M@ITBJ^%)QV67()a$oFZM>9C0R9A?z?Ofv~?{Bgm
zCmSBr?7wBHN4(!#_||lH3(phQPrHcp{?m6@=ih8d(a8Nm3Y-AB%$WqjTL`kYM3r%T
zL}lASvJ|EfB^w2Wi1UHp5rzRr1sFqT5e<Xvh_+LbMo&%~+Mf@`zO9Iip5m!(LJY?o
zdR7QotWyH0c*1=g7ZqO03ZpUv8vX+xjQ=H`PNhjeG<Nu1sxoL91pZOD%F;1X8s7%O
zKNv|ck^EjyoC`noZL3vKr9|klNIq?CG^9MEg!k+9p8w+B!{zY(hYjy{JBuR~PP^g#
zc>X)?XoN9^*MQa6Fami+)7kvjpKHIt%iu}`5QKxyWE(aE7{V1fp3U}hga;<fU#6>d
z?ll7jJgZ9y^gPNjeZ?en_y{lfe3fmW{)?GvLv)+(X_a1)Eq|ziS)cxJ;sbfpw-!k=
zf>h;@86mMVuiGk}lEu|fhX+FJLHG4BTSJ2B`yFqkYuA+0rX6o2pN`x!If#Ds4poNs
z8z8Ke1)bN)P6$uSC%M2tC<OmpwtCx1;SLhpYESth^Jt*8Y>zFli!oEG-VG5*3iq)f
z#UcS5*80KeKaKPUlB%MSR!}8B7>{arc6sx9ae-O=3YY565zdvssz)gf`|y6$8U)Aj
zES7p=mhT7(u<uJLD(!%3T+4GdQRROv&rl75ag)N~qbigbYO5w>xYM51D6)cm9QTr%
z*E_17xzD;5S@<FsP#qI{Cx8=B>7#C-8Uizn7esltMa8fD4;9H!Q0c$c^X)6{xSfi^
zL*GYxmhi39m+XX9vQ2<#M2h0kNrDLu#WiZaElI|?2?D(WHJRlN&jmcjf|Qe4W`+_D
z4eTnsBu#K=VR5owgk+$ZglyFvdU1o<M*|7Lh_1zRtgo5n5kVS#juitch!62)1D1k~
zOEOaVKZd7?hlJ8~p|f^Fk6H_|ACFzMs|fVdQ`CMXkPXq9LTMN``&!>duGjH=btIO8
z6w!qhZcwE*)U<^~pm~C36ve)YVmc?Gp#G=GOhi@XHQrR|+IJDZ87|IAi9Un@9g{q9
zxNp*L;fpaqGwx=2>1suzUs5ec{{r3RXdgi36(%`jP#JgY5gUWXS$rvqwiHJ-hNvLT
zkXis@qP0wM<@L|48kPnb`+CU!_2;=2*3`_+Krb&)65VZ1)Pb2cCfRE>mwSNwD`8!Z
zCp&Jl3f-mZqc@BZH<jAavmsSBQpTQTxi57al{nk9oq!*~bAXLw821M627q*}+xYQq
z$j4<7t^}9CYqwf!GJZ4OxB>-MMJX1bjW@8Su_@;KM{i{HrXVZG0@1dbh=9?H-;_AT
zWt{W%*-<$_G(CI4^x-;pc3q+xq-b6^Az_AEN{ooyQAtH)%&F1Qk09Uwj)qo}ia0CZ
z7tTL3PA%E?RpHscwdg%XPH9F(hH_b}{90Oe7@US)xu>i;K3Xqd4|x{ew%+i_Nsh@N
zawl|6Bt{wso_qDp&sm|}RVM_V%Z#By_VWt25bj1@@ro@fB_zd`4@&_jKc?PQ*xRmX
zuw;%MAPJz$s>(R`D{D~oinF6GHF%Ijq2QCK5G4e8`ugZ7VKd|>g@if2<+TY!Hat;i
z%@k4rQzUR$m5%C!L85WvPJc%+ym(!D5MLh=vGH!$=b$vD!$GAf^0(JH6;+QbM|KQn
zw3pcnlpa6N0E-*RUw8h5wFo%RvUiUh*_@IUk&%{8?du_xh^gtI-crDz_mx9D3g7gr
zIMr+LjZv;(<mb!9?UwD-hD~)fa}@>!4h3&AZ?VX?M7RbzFEkieeyr*WTs=pb(?_?T
z6=6hSgo0i#OM8LHi-pkb$@G<w$AZzmRkbY(o&|MXfIZm1%Y)9F@HuqVwfa=dr-YaU
z2OeXs=wDL`3<?e^D&Lgv-<STPIX8MdpFZThemGlrOfVY@pvzkI@cSk`MJKK-g>uC#
z=1W3DW6z2%{hE=Qps8`A*Q_n?qDoN;-=l|-lA2I^_>}NZmRaR;K>)aVe<`%mF#Dlg
z+4yCz&);(U|3Q5mcHc_NafjN@n2`P;ET!yZ-w?!k;^Rrg3Ib*<5k8ts2^B8H?KgPZ
zj$=Iampm=k$&Xv@mo39>W~n!$)WxI*&6t&0s2LJ+#KaX^)UPl{17J#$3nFsNa|^n0
z>{&??0}X4mn3qFT36%#q^h#2TS!ijF_$RklvqMOc;=lTh#Am;OCn~JhSLODLF_z=p
zi-^s9jf#pIQ7i~m+8a+rJz}G*z}@3Ylaa==C*K=M#RT`OBQg-fzRG-KE?~C3k*QO#
zb!oJ$Hi~)T1(N^<UVmJpr#qsH=JnS{3b=6i)RZ~G`aSlrrHck1!Pay;S|qWHvmjl0
z(1CtOSp^kU4cog>Pl<DHb=RenBhy0+qp}e9EGs(%TYAc|&7eKq%Jo_<txK?cx~8q;
zw0b(Izmk<2r2Xfdg?4#CBdP(4NMu)IV#tLZ4y_fz6r~c@cwvye9dQ&8)2PQ>&8dLp
zPo@UDDBou`Mlm2&bo!HB=PQNbqg^{@6r0WV=`gy6FiTC{?ed29RbvQE^PTpb_03}k
z3ReCadC%4ZddK5M3wqOzyV0HqY-}ciQ$?JvhnPizxydOTSUkL}Z^g)EQw9^9ek=c_
z!#3YUXgnW*Kbuva%IoJVyw3%AkP`}g2Wy2=HR)A{7Q0-rpO*SY5<*)hxQcP>Jnfa_
zT*b=4;mMoT@qo+=+I7_-0`GB_eyO$IAAV3gctreM6X}F$53!SWGgC;{kJ^*Q<`?i>
z%i8_cghsOHOJdVym6*2aDvV<c@n1x82mfHAqH?K7;JnrlcGyrQjKRl6rlOW*$3wk|
zv`142hpy;H|Lw6ZtiC(_(EOFS3(mZ0?|fX>gSDb+;zuCqJ4U!|TghAx6%laNpujc~
z>ysnX-52TUvKcQMNRrfC*!rL9gcvs`yq@*Y@9>)X)bp{U)V71<y6iY<nWzNLAT*`u
z$OUsw_K`ftmaEKOH#sKN*~El`$ln>ZX`#FUGp#&qBOh_!O_`~e2JdwDYc5!?Q?r?_
z03IA~eX_OPb5CTdWQq}HTFgU(FXWb+_~;ToOnww9%zvCj847!8dJxGKxA0JhT@bc+
zQKVx1V8xwL^r2(L$ecFQ8gKyTqDC-fc&Hc*Dh=x&%+t=dJ5g_QjSr~e#2zFBm`*Do
z7{B;FT-b81U1z!aHm&^oyo=WbkGt??(_>OrX+7fgng-Xxh$%VN_G*SN(XuxFnW}>b
zAG6)~kJ(hW>8}<rlE-~C;xK?P#5TG?{VcT92=L-6Vwz09jl}8nIK5(413K=Lz3;um
z3}zSE<1t6d?34&mu55p01-)ZDFzvdqt-Y$$ZA(AF(cdwIm49uOmv(+*)17Lc*(D?j
zXl`&owStMu;<RbfLll+2Y1Oil;fvbT_|uLM9mlVv6=%S_#f5_Zw#e$*pk{FP^zrZ8
z1=%Ev?T;Iue3?>ve&L1}8{i*cicB*`_yKv(*~@j)R9ezUniWv#0V`fbDaN`u52_dI
zbj2BGpiUj6aZON+M%hMQ)D<Muz>}B25uibN3Efd<g@1_cx_v@TlZD1eiz1b1-Fl?n
zTBiA?a{hEa?d|8`GEdzZwIgK+xlwW95gbLkLzw*JcPV+5hRY){X;EhA4OE!LNiH>`
z_<LpFKg?BknGM@u7o3LOKb~pdgg;)kfX)_6Weq7m#LT}V1ozw4Y0EjawkjR&h9hnB
zdABoZgy)GTHqijF&ap}Dno5otJk3nNuxOVw-Pt5-`Xp|sTe0=B*jTrkI278xIuiZ2
zHd6wcT8Fndyz99@+ALgSdemziPHswTl&w7?7jiT+-PUy)F4K6X&*2}6n)5Iqd~%u%
z;S<MKJ<5n-C9&WyMkRy@wT4@T@o>wS7#_k$-d}EPyESc{h3666zS>h+5a3a}`w;(u
zR?)O~bt+kgPS3iUlzmhZI`dtt3ulG;s3a{*Xn-7xqo~LRbZ9{}1B$uq_rCaUD8<9*
zaVO<Jeif9Q?1Nkj47jnF!QPR=5;A-hRcp`mk>Wm*MlYQkFyuNHj9BjS7SE0c(b*0K
zY$@`s^GFCdsM=V_l_^$&#{&KcdEyUoWm^O=Oj}L40%3cGd@qbzmIY;JQs4fv*<Geo
zP5bZLs?vvDboWxx*f}xjRo)&yvSFTBF~4j<36hsw?6fbD4=lP4EcP;0MMUT|H$ftv
zf~{+aMx2304L7C;i3Ri|KJAnoPS10+a0<9ibv7`ih)pRDP&e4^Bm%8V;bs?QOuAx8
z5%T=uY{$(x&;}x=B6WA!ABKV;cCLs7qBlC7fcAM#xYl~|frG@|tXq+xlrmX=MXZjQ
zCwHd$lXd@y(H@qxv;_>9<^b;k%g%UU*<oOokT2f2lK)#*RHO#cobJ%xrE-9iVp`qn
zUERihSWwXLyLbDWlV&f){mFV_{}gRu4s;}wG$m8SBEr?{|4Jw;RMh{4J24`KW%J_P
z<9jF`n<Sp844%5ZOm%#bXzt@R^@t=}=9tO2<qSi=i{wN!C*cT#P|s?-6T3NGcYJx#
zT^-}(W&+qlGEa(NKuAQR4HAMZL!lLpx<mkcySx+eUjk(=@wLSR_50Oxw?*$&nJfN%
z<8B<IebmFwaT&V6p*lV0&Hnsod%fMC8Ons%s1e^3UChXjJPmtZVZmJ>i1RzJ>J2bn
zUz|rWG9vW{DesnfW)N!Zl^Z%K+0<$O@FB?=4|u1#Px3a^FdYwzOu|H>T?b}48$#FK
zKZ*c9h#^^7sFp>?Naw_2k4&ldQNvaL(9iPH!dm;->n`0Av>)4ilU%y(8Bu(mvL$Qn
zv>7PN@`j@v>TniKCa2Mgd$S^NV-gz;b&kBCK+n+75JKQkF>lVWp3EaZDpg#h`o~e~
zAy&Od%{SY89lR9mm_wtl*Hm#))Nr3)2#K2@GnGYS;%iY$kn2~-`@x*a51M^5^Vo0N
z4z%nJGYX-jrJ=d~)%;2yX<}j`M2ti%YwTC^RTn!83yg=YC2d$j)YServ+DWTsn3N|
zCaieRq7C!afHozz8MeE6$1Fzt=7)<aF4n<5DUN>p2VruB*|{*>nA8OyUbLn_@wbnp
zE|qW7E-#Tzmbcw~y%VzIK1t8W-hVi09WOK3%AbQg30v4f$wnu8rB$+fEj?YkHT_`K
zW7ACxU%*0lL}buRxt!PTVj@W9%a;Q)AE=JZjlh!aGZ{%8B!6XF`x-s8JG)*DZq7-D
zydJ2^(16%(_9rE0`-|eEK}hs4I{c^%Nv!xLRd{ZOt~iNNP0HD6+KoO|W_@Mz5{J<v
zcC-EMqZlw|6|6t)9&p2Xj=3fWBr-h`KpTaEV;>;F_XMBm98IiyVf<C7QuHDC4O8A}
zWHlzFOV<;58tdA1Oev))CAv$uFT<^^%*r6fSk)Qj64C2fNov0JP#9k2WjqWGS2%+)
z{vI}OedH-HP$UR-MJZOdosL&Mz?=dAkT8xH!L-(B%pAHiNu2CJTZCn2Mx<Ldt)7vG
z-!zr1jJC$@;<UZQzdaLnui|F3`$%>n3*X+8$@RYWT;=A^LGdJv6id0?9#c5oHh}#G
z$k`eijm0Ud_U!0a_I9YW#Kikw<v0mH4?F7B>l6?&(elZ!1g!m_^pRgx(<B%ph&iwz
zGylFS3{FB9Ms=vLI#yV;OPXXYsE?Z=xdc^Bn1XI+lJ>rlxRiHVbuU7<1{I)KT^P0~
z>?g$x0Ej_|2oMb190e&HAG^v(l6(9SbsE@U-4aFpoO*kxzjCa7;#*67?!T{jG}-?`
zS##+yC2p=dzK!wEr_=%Ee-WEqEL0uimFh~-mWS-K3C_J1Bu2(8#1)Ar3}E%gR{X}g
za*Ur}1wwbE=W$MsyzB^S^<&kSqIt<RIv~+Ov_v62L$Z~`WE&bVC-z};<-}~bxi30y
zu-LcZ8Ovorf^%n{9Y#zKoS!;Iy6Qi;g!ZKN;QVCpm?E%9<y#Gg#s`Ee#-V=r#op%Q
zb)c!6ijdR@v_`bsQvQ+9TxrA6d#e+goAim5M2e>dC}h^CA5)E)R)ps1sK3!Gy@9%)
zPP#7REx^GDyET$_kgH&%cC3iA-IN}brs;N}9Ehi701zn54+`ML&y|C0GtB|K3nOz=
zlCu**X?S<3RLC%sh97{@5UHU~*$WCuxK-{Pr(7%F49REJ)23*5QBCSF@Ad4s2N|h~
zLV`+OO4U9-&+DJ^0y=*Kaa+{nQOTi(60<K-ThqIY0ONK<4=Zx|pPzKvbPERy15nd=
z%{xXQkJ}-F#-tW*G9y&bZ*8Yk!C7P?LN)9{co>%F3)Mh@Mt@A2bRP>@h6sBy1fwe=
z!d!hr<F%?&DPyJn(D6*Qq4=8-v0hDt)kZ#MdkSJe$NFh$&+C;_$FF(*slDn!2#Y1u
z#Kdog;rw)09bHXDOklvvg^W!oVzuN?vwVS=`q*#ei~AA3C^IUSb=84hYh|9Qke(8v
zPYMW}lKDphBFO2O6ktWndIXaG<e`?WGKtUk*L**tqw3u3@KZLxxC9j)r=!fyK-8Dz
zb%KXF#)_x!lvmMNr&-rK@b)6R)T6w$**v`*BcPW`VrN?aR|_ycIwmO;E0)?N-?c6!
zY_akpblZ30TS;=rZWzAG(6nQptG5xp93tic#5(p~K~RS@o8G_!m6R4&#kf8a=`<tD
zr|r`iFXyUh_nBuCmp|H_g2vZ88DD%5>R1I0S<5GjGA@$-`hhd}%~g@8$Q5tyV=$$T
zn9<)cUPkHw6@mOgZ$Pg_hrGIqctDzu9GQS6pA02E?Gxv5zCTp=&VVFDDa^THoCJl`
z&%*J#@Fny)ciNG(ubOrdNp%3nU|dp&5jY||l)Kwx4T&##ext@fYk@z}p;7bgsdT95
zUUlov@J;p=l+g2hM;a<TU+$~kt7K-)NqA&LTe1Zvi9H;S{up;LlWzt6y(zZ$G~Z{u
zhv;JBiYCex%sDS|bJX)<JRUrz>~mJM0W<kA0<Km$rPK<q?<vuXhdxFYfs>hnfI)u1
z1DBdHbwC|e{u#UrZa4#vGf9sH&2yx~M6#Fo6W^q0YW@!B^R?PKRvY9HqRn~e%Afvt
z$vI1NleYEd@8s|=qw^;!*Fb!LVCT~?@@4)XHdS#^T;dim(E7b8S*9>s8cZ%cFP(Nx
zMPj=vbD@F*0yIV7UO-zdy?*X@Ycyzh$nEa}JM2G6JfOsqgX(Vw<L{7qjE`HP9{qgj
zM6to_z*78~<FVQ>wVR`bfNvHquA5#>bNyF22!|VZL$;e~U!$+Ztg-I4%_x9_L;Xtk
z*-$TYgdeRj832d`LoYtAV4_BZP6tdMy~D-A)Oi6j%l{eXD~_|U7FR4Bm%j>v!g`;k
zC?1|YJKBltH5POCsD$v9>=eXB>@(<%nP3w_wtONk@)L34M7#R)eiCZfSEK*(PG+oP
zg=p*;769CH_KA3s0OcJkrLa&VzL@+XWctJapep01;Jn5xLN#)QfnNXU=lHvZ2fL&g
zqJ8OLq(^{S9}UVCi&yIpbzz!vpXpI)+Dponop;+-EjkVw>q!tOV(}ed{+b_+HoR4g
z=}BD56_MGSAw}Txm7wv}qHMcVQlwdp@PtC~1G63zw<vJGUmvT1I0-Qgf&U+_82<$x
zXnj*3sKVj021R}9E7!*pgjZpb&tfYPUg0p0Gp0wt4^HYON(ivjAR)>!&_F@lEx<~b
z0Q=4npY@-PmZL6K=*c;unNu#Rhj+|A*WoQ^(`5RTl~P2fgsvvq0B&{)@%$b|)s17O
z+t!{w(zcnbtHmVGA<bqk&7tR2kj_(}O2>QI%M#;vU?Rr4{zGt86+fC2-ZN0v&T<sb
z{ONt4A=D)XQ5>htB&HKg3AA!%yP~2koMT9+M!5(pj1*~SaB9jHsDJvGx;k|>)P}zZ
zBgOmzNe!%3ZRSM!hx_1$NLM^5;6kXyNtlZaFeBJ#jWGw*L~(F3Tgw<J1HQCMDp!~j
zXGHC={<9(QDYrouP~MqqJ7&Wj10?g8aN_<Xd*xl9$~dW=ihfg+@~Z95%>~`AmZPc+
zP0s$7m>mmX8;}5!^O*v<-px2FU%eBZhHJprSc&V`q8IR+PkyM35)=<j^O;HV>cD`N
zTOdO|j-=&zgIYOcQn?%nInQ1{xmKDA*bCbplEo0+qd?-Gqd*M(75K&XvfzV>H_;d8
z2*=2a@)8cJCru>ZOH#i8IY?|YD9QNowD&mK&Y!xi3aD~8148_{Jhb_uvC56prL{>h
z(m%}+3t})Y#K^M7SCJbn9FhQNeaTuD2ze<xL`V^Lw!^w~C60J)i(kCKLAhjcYh>L3
zyBEP5FHs{r-YN={JhA%u>Xm~nq1vz-;irYFTnPi1pniGoGEXAq@Fy+x6zZavHIHT|
z4}~V=uJW-x$*g}~xq^d*+0i~hpp2tWN$uY;5g?S46w&_qNf{~wWoQA=A1Muu!kgDR
z;7<5oxQp^#B6zEf+g7FNJhXk({?%lrI2DcRr!oEww674AgM$J)D;OyO9PMqF=GzMD
z^SM<Hp^p(2F(8?19W4-LUHj3i&I`*4T3TA8#EQXsFRXenddRQNjs*czC=L(Hqf!2!
z@A(B1Uc-9*+hg?khUJ&tlc}P$S;ii`7lRY4mcBiB!a7Koy(>&eh#f6&H8)Q!hc{Dv
zl=Wn3`3Qd#y`mG-tjucrc8*-6Y=n*)fsB<{{l1}NT>qpBUCIDC>Ra&nN)zZzh(3UO
z**hMB<lR0tLX%_0Jq3Y=5yi5`O~e{NvZ^M)2aAjHf?9%F+>B+3PR^pB%yh}A{hbam
zvT0{5BO~Xr8m_Zj6)!0ga_zy6Trf2CHX#4s9o6nUyH56%M1sp9w?ZIjL!=SeF{)^Y
z;(yGW9KsQiln7(}$c8${1L#aK<djp-@)`Ibo~uX=HxOeD0X~FTo`^DaPO)r#UAagV
z!ed^V1brtZG&CKU4iU3JuV>HTFDj~=eXAKsOYv#cL}}Cz-M15=LOEA5x=>%?<67N{
z42l$lBLnNiS<ZG~K|y;ZPHP-Sav$Q1kEWA%f{~qztI__<$c$*Ibm#{8+YE%&0xZzm
zjIiGkW0#}=dG7`y56}U(sSjJK{O;~1*%Gdo4&93qU#Kxf>?ZGp2$e8OM(G*o3g}pW
z*d>0x{d`o7oPAV?J+c>1y2Jtpo|O23gnZ&hpL%bqvJeYOzQ4J)2z$E+!BD{MmiqCc
z7`~j7EJ+Qk{@nrt{gW<%HK9)J#nSN@AV-7CURO1pGu5xCSaXjiqj3^^U3AH4(H+ZL
z@f<387<jKC{>pmROX$8UOiFHtw>=+oVtZ6<Vr^?pkL3XwbM5h%6~p7dbRq>4O^Zxs
zdaxqA2q8kzu~t=(r4@y(46WWa_hfrwa(W1pqZW5oxKkUJzu4VMbxR$Uffj;JF0{(X
zAcXVr)Q#s+J_}-GoLsBMZOZDODDxxb{if&s8YlUja}%OPUtWEj-U(+0IyXz7Bn$-s
z$cjHyRt51(I>!;iDAb1emkk=HMsdI)05|JGUPqL|YA_=mY<V=dC=*`=i*DWdJ)E|(
zF7Z76`Z-RKuM8TqVz62t!J=-T-J(MHG<|Fu-)?@~C$JBrq7p!>ic>vv$1$Z)F@pqY
zVZvP_+Hgt9Z6H1@mp=x0Uh*U|0c9id6dEEX{v81DBprDLRKq-23Kfg4AP<Tv{+8>0
zUy$G~_~43^k#ri(mvJvnX(?j%$LHdBF$2b}UBvZ%r?3metB>RdmJ~@h!@z3u1Q+RQ
zQ|PMDY<1aUN@YbB<$cKVaK;&x5x!h}5}Pj)5P7`<p}M@Y#)C8VDAUeYSpFf&6evi(
zwzMVA>ZTtSaSQq*#9{+CcEf#jj~wlsM0lzjb(jssN#;8pNWjL}lN><Zm*R|R5riXO
zK#WTT2w?i(aTV6D`zo?sRiWE|LnBnfx1lY3apHK)^qdt<82@>Q<OqP<k)x3+sqI`j
zW<)9@({-k@h*>;ZPhtfmi=GN%jq~)qy^it>3b2?dsey3G(qu+GHy~3Q)-7J0P2Z>p
zjQ@NX?@dD{uEKKGj}FK9Np@~V!l^F|F7OZTQ`iQqNz(>$g7O`inLM<(TYhd@Oa=JE
zPVKF7;dF|Z0dvm1HciCkP<IsOnhhNl%mp}~Fw2om$YRy=e5W814!P(*J*{OaZ$qcF
zRXv=|&)MK5Ek?{b01s>q(J@(saId|t-HJ<fNkrZvL%0TVtqW3x^<pV;kE~9GY!(?b
z<1&3e6JnJY;5o%7h^9r7w0H)LodiiSFW`?j`d4=TWcQ7ku6tq8Vi=?OzR%6jfJ+Cs
zAs$+*-lXU|H=}f<4`&!hd7>@$bEs+coSYo?hG<*8kX0-G+#+6{+(h77%FD{Ay(=E`
zcuSA;;Y(}D9M#o<$rqWn+%Yg(wZZ2ctLIdekfn8YnvPgw8Dt?Id#QWGFbkbTcU$pH
zI%9X$J2z9}jfxnkpMr8OAhClTdi~IuR#-qha{jSsYa!b-$ky^;PbShQMn15utPs`O
z?uTtUkIpYbeOa^IhD4PjBis~HvTQw8yf0c~vM;?nSU>ii<8Ln*#ho`2l)h=@5|1@Z
z(j2Roh<vU?yZ`<+m=tv`QTaRwJ{-ldE(v|bG?z&JSvCu8I0A?*%YQn{-2Ej4j3H2q
zu^1=)qP8{yDzRkVZ93?MTn^Uwil3%trwBw*v(KZ4D=Se~BuCcL*T%XCbxmF{V#JVC
z_$9EWV{%K!FyUCZrQ?aBnkF*ig;EaKR1P>Tfeyf^Ct>U5gns72ZFN&$ia!CY(B<?h
zoEF*1#z}+(DZYUJe6kvsf)-gSRg>z&m+nuMY}MvZX{0aWkW`1-5KS1elvs9;lPnsX
zG-<31UaCol4zhXWx2Em0Kp!#93g)U1YJV6Z^l@og!VwWPJ($l(!2!M5#$+^VszWDM
zl65W%26P5;O89SL0E|je%B_W*ctz6=lB$8U<4WLWi12-Kwr*dUmE$`Q>zZU%eC5<;
zfb5}+HzqCQOe36Q7CYzMDNP65>yi0(RwxjZ%p!#SmHi2DtVC1MPyPbn)U6CUBvwWJ
z5Iox1222c?+2C#rx2!kb(Uog^%(8o}%zzp>p+p=9CTrZ*LT+?~hT1s(z|jZ~b)q_e
z0$WnS9E6Oz1uT$tP6$y-u~^qUcGYhtN5F&V=?pjX8YxhQ9~wvT3I-LwxYZItjw%8f
z66rt!Dh}ilvEX6jKK}Hf_`JhTEZmf0Wx}>C$G;AOSTU7McR72s_KK;ReR-RH$`&u7
zRq>+n2{3^EJ~rqN^Lq~`BC;`#G3UX^hyg2N{L(47Umqrw>HU*fTVQh|W8uQ|_+j?I
zkyVD?RAHBPw`!7P@)U5+4_FZzB{ALQDqelli)4gQ@hid3`O4alqda2UFx}OghxP#P
zCj4WLhws3OCV%8XHqAyI$U<|E^)%xuH0h32&Saqbk5Fo5y*R3L)W8=Rws-B`R@KiF
z1%U~sUa3ThUM0GpxKp=d<){N4k(4h?a{YGIMOaX9Vp7FpZU2ah9`U+B@^~uXM2%-8
zg^c!4;h{Um1p6%bI^&INFuA!Y%UbRDGA}io@Sk){%{z!MpKeE}k2Eqx#lZ+{?YD8-
z+|09!Gc<UBN@ebU1~1z~N9Aw#Y5B3?4$U%up6Kxc>Wy2A@zeUOWP1l9EC_5t;1Z1F
zwpyQLduthpf7C`Lqdoki8N8&8NVk9<ayO3T_DgRT?MAE^iW;c$@7$ey73v`V%A|&v
zhI&MZt=8envZhn2Q=tsBIckKIB%Yd^UXIL)Bw%-tu+ma`?r8J|732#5ep5pJ$5TlP
zktED;>VerY&+Wv?ST1*}Mt~C-;}6nv!3Tk9*XKcQ%FG;UJnQ?p8DvP*&T10MPH2U=
zAj503`=Q+fTV_Ueh~`gyRvvx^8wgmj)QsFiLxh1?1L(g9CdQcnHB~yNpFG3}Na(vP
zkNZa1>!W6EVPb=>`@8R(9`bvj^&LwZlt~bM&w=EZ|562>p%etHVC+0?p4S`lnAaK|
z&3!%yBW+r3U~X{u^gO#kjJ}fd)*h<MJQGW3E{qn8k7t|HNou8EBM1)W(IMFd7eeeB
zxO(e=H^EHKZQ3g<J!9en1z>fV-AV?onte4bPZCpQsN!!9R&*&kq#bO`VwOQ?YJ!cJ
zEL4qZV~EOtT(~4a{<=S~)M&D4fTXmaZ$~A2LoKexaMjhkG@n5Rs!h7ek9m(BHK>yF
zRFG@~>R6i3byykYi4NeG%`RjGetYr)e9msg81oXn!$#6tPK4ThyQHzs_&Of)yJ~ot
ztoj44G<xFa7l?IFAQ7_kY>1&}FUw8Qff2?7;DZ_qj*RdF?I4H#79LaF$Ij=8%^)lF
zbTL2q5HQ`XCCY<#e0rv;$tK4?Ky+zxd@^X`zk7)FpPm<~3K(VSfcc4iA&Mo1w}m|Q
zfA@g(_Q|*Y@3F+h%sIdeKT#6$D}?Hw`}^;1eULl;zyH-64-C)$eaip;GeDUCCrR?H
zDiL2t45@I9l<P_`l=e#bBO`?Z?ErFua|dVOSl*nqwwuW~s9IS##Z*uNbv|(!pq|Ht
z@I>PtNcdq(84=<mcL}zEkPC>_B3*Lo6kvhGc4Zds7yATYlR`rzdpi8@E@-42l*$&~
z$i=``Q|eS;RaD|uHl*lW5d3^PJ|JU<7OI>e({^?vDTC2!K|);P*DuIG<5ZOt5rd+y
z^Pg|s42iLDZuIuPreI>ZjL+|7>WBX0)P+94rMB7DnrdG$xA{+HLA<Nqv+D!6@Ge|K
zlsW|u7d?VHx&?&t;WQ{02$7>8gro;C$`Qc_0t~>iP=qPcA5T8RiS}bueG`lqv3a@~
z(uoJ+bfF&(@gRI^;4*<*-pkHLOR#8BOxDr2dek>Q${g_=Sa~e?E;toRfVe)&b78@g
zH4QR3_+1G^uu0DdhnZTSK|~k|f+RHxox049Oh5q@4nX6AM>rUk-qgTlJv5Um3grmp
ze8(x}y*RC?KVhczMd#ZSk>kiA`-nbYY3zYB{@S4P(F2Bhn<Ix2;F_9b_+zKxaw<l+
za6w@ZR00>E;ujhOq;llm{z`3v?Q9YhB<s9zL{!xJx;4gk2uMy&R}EqixNEKt$?664
z+UItP86=r?Lk)4Jr^O{7ayy$UfdJQGr=_MfI6JzjEDvR~K5AL}*fvvB)p1GMLnZyE
zSM~0yD5FKtF;qa=hQ@0hah64v7p7R9k$yhb)s<Ue8k`j7{uO`~QMqH^u;Bi951D1U
z=CXW8C?zE&EiG*YW(1tf&osvtmNx%!YU<{@Ug(%<zG91wjh$_A=dWF`K`eEiDlEG?
zNE5`aDAwdgD$7DG=cRVCA<I&h3Or=lCZP?5p!((*W<~ycA&Vzk{SJie-@=$iPnG-R
zWY$KE4<Vp9<1BTmkr)|<KIsF9LyY{I4GpmPN<%}l;C6SI`}gn8Sk{8W^7p47FqUP(
zx`)6RN#m{SiQp2L$-K5OQOW++aRrJH{Jg)eA323gG$56j{;pAf0rsdI98ZeqXlMF?
zui~acCmOS8D~yp*b2`jPkb+1l7!GpqsF3qh(?L>^CZEjmu;<T><LQUgoCMnUD&bY+
z5=7-amINgGGWkIUa;UWYnbLIHjIrY;08Ly+Y@5LZ6`PsQ4yU1CUA@5^otE|Duiy8n
zeqTHD9o4O)B-=tptZdc%wj_~2LwWmtFvVf156oH+27?$a!3|;G7M1BgcH>;hkV@D8
z$R^e?R2m-=2Zcwj6XPI)D&09rGBk(<StKN+_*+W{sX-l*;c-)zcYcAyswbt##v&<_
zs!=$#1|}YjQ1Qz~`aqV+`ti1d|F1A)Ow_ql%Y{_#;9Ar5l1qL;kV@kk=1BZYUVi=&
z$PnA5Khw2mQ>AV(B;sLUa9OS8@BZuooDNqSUF~8SkX=8(>nF~|f}14%z6-<$=zZoC
zqTm<Lbai$n0qh}Ctd?6X2`EYeT_aJR9rHD59pi>O9p4@p^CPx?sa>@1ueo$ify^R$
z<_cz)Xw`l(<b4iDKK^4)tW$|jsqV=X*x+%4ST7Lsy^W>1kt5&VhxUCHNFIDHSIPEw
zNtVA}dO+Hf?$;DU<RPF}IRgvghS|vT-K==Gs=okxPd%MoTO)o3h8`PhdVJX)PI>nC
zriUNAE$t8tKsM6VU9AfL#!aAg(~b>oUJR2-$-Qs>+05%5&cH9(hZJ@5o&vY?$@ZV?
z%GZ6}xAC|$zgBnk#eqWE(1d^_url%yTQIts{@K!Lm_zc}O@BYaiQQ`#?jGS)-jV!*
zJrNumqTla-Q@lTYc?8Hj@&2k`r8%<UxcQALq8>Rp`S8v0=l<?$OI7n)rZA97rqVVG
z34ZsY`wH?d-M{Uo;C$;@^F+PD{xmU$xsDpoWq_d7T~2G%gQ<NoPBQWst@*`oWbADy
zU2Gcs5$BX4`cA~4a1;sRL8&<`xj@#}e1G^AGM@ar1`IK(8bR!+R*YD0vXjj%C_r~d
z@ycyUYW@Le;`he#UD*1>`x^Jh2XuG);fU>9T<vWI3_qgj`O2eQDR~?}3b!6ndERvg
zE|hUrg5SS%4<AqJ2wAemO#V-E-yPLt_N^UzFEflHMHn3fl_r98h>C!brZUtBjDjH2
zq!S<kb<|Os1nHQJh#<Z9K*mB7L#UxiFOg<|&_m$v7yaF`?p@zs_gm{`AuA-|<($3G
z-urp>c|A|Qj*^&|@Pv`k%lZ5(h>1KlDt|3s1?Z(Ca?|P~%mB7RfO~cKXvN|ekooo>
z5w{ws7H}EvVG$PjI9G(Ikt>a_H0&~NDhh5k*bzCAJC$DnvJ1?4a5JAUx<N^s+85Bk
zw79|_vK?-5NH4d{Chx3(ZF$p2As{p?ccMNocaK2W4$6>EM*D<kMxnKdcc!tQ)<5a;
z8w(@r$B2}e4ffjyXYMTqS*;#d_;PQHH{=d8yVrUK23zjUeQlCms8#b{I)efQ_o_0d
z@2Dit=)1>dS8BD@XDMVqw-P_@d^>w>k-DCdZ@WqP<$hPe2&U4G{et~@^-?M$N_|V!
z3aF(q^4geZ_rJVy^`RKz(LZ|q)K>=%ut9_-91)N1Tj#>W?(s65Z&Sgwo0dQr+5OYd
zhk@q-OHTq&{dR;=JBAY$eidnZUfh69H=Yj-e2cD}JCRj7%D>isV67kbsJx~E(#^*s
zAJFS;G`l!}Q!I8^25V~C&t(OmZ=1P&*r#v?jizRGe}`m_+p7Io3gtUH<MJdR3;p5b
z`oa}Eo!ByV)#y-#r3>Ow9v&XC!%Ri5$s-M=dRbqWgA)IG`gB)WkJN25XD=l>_Q+$%
zcYjQk2p3x?_c(9K$;->Di%f_&PWK#?)sk=LF;IQyvcBN5#UcMLo#DP{us&yC(0ye9
zkv7^m^tzk|3VzEG=;a8WAao|Ph?yp#Q`YmXjKa{em*fBF`nkEe**4lue|&!B>7!NO
zirx;%Jb5)WJ!+qDxhBKb);6%Ps<L_}CbjJ@oNn*zb&8o}oZfUtF0jyjbbf^b^*ek@
zLk(doHa$KpdfCbD*h-Vi(Kv^tdg`jqw^F81<w{$&+hm8#0;`XXDCiLu>;e`tWFJV|
z{IIuc3LnL->CSf6Xye}A5cnZpL)PcEEWA+Nw6~Y`ijN{-fZQl`Y4~F5j{lp0cBo=8
z-r`S3J@nGM150f5x;x5pZ8n-6RZO`ZA>JjVzBUuah9x2#c)-7Zi>{uYlba^C5Aod-
z^y*Di7In?nnaW+z%ev#Ylulg{U*q<z9}Vo|_VnFe{MHkHBizvlM`Du}DLatdJPp!v
z4jY$UU#7Fe)|bLitDmx0cS9}g5j81LqfA(^-YeC#w*G?1&dHZ2I1NCIkV~KnZStwd
z_P(#TTAS6&LbkfA^-4Oo)iMexl+V%B)!V4i(NX!*xTzt~>5;Mmbw5;Xo?}FMr;xFU
ziGkomp}7<Mpf2P;{f>vjEhbD9%a`(Zeo9MIcFlm(nT_Ia&=zLCz78XGu#otm*0nhb
zTP=Swd7JjkqTI{L>r{7jgn>G4&@`Y*+%mVW4XG)L*MC5PjnI>U4R0j9Zai#s0Yq<v
zB=U{V@Oz)-`+L<dkUi%gk8qho&R(1HJZYe+^jLwiDG);53mGZmgna9o{U})7q9*xj
zfUXi08CI)<ye@9%x@&Z-R+@Qge9C)zhws{m=QBu4=t_5HgI<;`r(|V3-C%MN88-4s
zf!^uuX=xC+j>se;O)Oy&eZ?+)qlckcd)&IthsCw!J%&{2?+>QF<OVMOZ-b5f6iGp$
z1dd`ZUHBbB=l?n10oaAW8Wwktko>+4AR>!^4=4MmWUO=PBjhM_QwN1|N^Z-PSLym1
zXPHr4S+!WXs0P(5CME_^6&fha#!H%5;pf@D`?&g25a~Zt*Cwc;izM6pI;faW^VO|5
ztM3Js-2Oz)8eg5Du8C`h3f}bcib8-4@@Gm*OJ$u?xKb0CnwFJjAFnaRXc;~Doja;L
zNO7ippEtJt;Q{qpK(Lkx?kTEJ+X(6q2SkxN(ck4s#>U1=BUWo8`RYuh6osmMQyi+R
zB#7}g_|XAHsw3zT6-N_A6eJs#{HNG(Tp5$KzQ$UH@l437Ghz0}_8M?QuX8LI5uqvb
zZ)PT&U8St!7M7P=d_=+C0-Bq$Gtw=yYo?0(^o@-E2iW&pnCg4bhdztIKAM!LBS>+m
zFXV>JjGIv7#Iq?x{S=LZTZeHCAk3$SRP(g7LEeBFUK{{(->#{7mC}2#Qld)$YWWNa
z^wrv-Ii2~IdmYGDF$MuEE&j6+WRblZ1g1CP3Dk=>N%mLFq68K11=8`~+*e0k)<#e9
ze(Yl9kvvp0(IIxxp-h!zd$QG$5aed`ylbrwO&D!}U36gRgNL&F7rI@<^bCV#KGW&!
zR|oJr9VD%|<0eXO!?m{2%JSx!hXHN+A&20{l1SC&K{Ag;FeS^Q@yiu`9JN2Djwhz2
z85Xs(hTqR~@anVSVbkvDMfWdm^^){u#JI&dIsd>-!5Gwr0T8fcp7~hj@yTZ$<=)aR
z0i13Fm2S)rOQ(RP@BJ4`LxUnifTc#}PTFlMnJVhm$5AY^*=6k&`vau|mg@z~yN$hi
zJli)paH#DQYC|$yeMSvb^Pdt=4k{3a6wXv9%$IgCc|rGTc`wSvw<SJ?Ob+Pnwx`zn
zHqR6&$&s9b|BG?Lp*>wT=2dQEw*4QHnx8%=;oniIfN{IAx0_3m-Quklm@S^u-8U-G
zx3nFgPC2G!PKlh2C#+Ex{Rj3~p?oZ5-xwr+U&f=2K3+2~>tmAPZ8;M3a%#0y=1ISX
zvkOW}cK!pxY<#@hou%$#C^RH(g^aAsZ|`;pZ)Her2J|Q^g#0PEQ*jOMOOBbyrji7l
zO9pUGui;vt;sibUz~yBeVfvx)bS&NfF^5GZYrTF4at!sE#l4@9wDGCv`$BNqA7aPb
zUK(oO#bg=`A$-$OZW|@f8KNaH#X#ABxwH*SV*YOAS<PdDBh0EY;cWQL{U8fab1H1Y
z+wCS&QDpf;e3($^DaB3d=ztz)Ueofo-1oed8`2LvVIf+-A=-S%zuviIg&8O;>GTfc
z@4gKI-wvv@Bf-3CzY)wK$$=&eJp}C*8<8f5d9E@O^e;=pQOPPSO2nHtClHOltg`Vm
zL(=QMx@EuVcRNa1l^h>1d&9?%AKAV7aHXTv6Z0Lp*8X#^g-3vu{pNzJ$KH{T2N_{r
z>XPQ#Bz-N|?_22;1d)O9);|8r7!qKNzUAQmY>fH-NKbeFL;XWyg(>c{X&o2-{=o9b
zu}UEZmlletG@3%+=kDHxq84=nY6RQxmu&%a+5sN{>)>b(mhSr9cgbYx5bxUtCFim+
zQNmO^A;dTxh@_kzZpy*R?^be))oPWMsx@}X+sr_jBp+IxAH?#nk%I_HnbC!`Eg#pg
z)22s|{FB?)##gfHc7T{2#Ioa@?hz>e1GYAK$6yVw`i~ouZ{cvUi*OvBoyE;nj{-W~
zs=P;ihH16*$Vzd_-IBXr>|Olsn$Z*Cnt*cga@@uHm)@E0)oT4!E9t@R%UwGk8lF3T
zHZk{f(b=LiLj@SCf|nkG$@puJ9@Jjexqbd8x!_dOeTocw$;-H(M3Mr3d3*fD6V!<d
zx@V4c+R)dxt>2i<!FVZn3x6Bs#CFEUC*&_e^5i3wKH*%w`YUDy0HK4$h)5GFhj-|d
zd~JTmUve#9*|p(zaCaKR=h7C(h}hUu7-jcS-7xwV75qwp$SIJ7Np{jXJSFx4TkPES
z=PsVPbSrLJySRd)Xn1N?R+NaOY0lqRGY+<wPDW>&<=bM2mcC;rkh`gW#`g*zbrBGf
zcQkEO9vd5DxasHwZQsW)@|5tT;)p?$HmOQKDOQWD9jdwoh03iyM9HB&Kc+D}?|TAs
z?%cTx6O$de8bl93HB}iFb%{fD(@Qf$M&{;SOd|ooh&6i$hlvH5Y7cSNU+Q5AbY=@E
zEB-dtyytG&T}Q{s->FBr`&DE}i9xDCVfSQ1)WZaE+fk@nA-gO)sz=;Xb8@Z{f48*k
z&7`EJV&WBCqC|RtFv&Q@xA0jP*wp>()Kg{z%Hb_+I$F3{5+@i@D}gQ2Wl`4&jr3=S
z#6>J9v&mfVDHGO}Ib49+>huR<5(tiOlk6*UiWjj5io<Z<ePZS4!Yx(J1GvL};kNwv
zUi2^@3ia0sQJ<UXPlS9gOGWE|w#79wvsiO;a}i8oVc}^>Nh1e`bWz=yzkm8o3pqii
ziEwtcPK<m~TH0U_uhiZNoP!{bUm=>BQ_MWl%nav%GDv!G@Ls3BN{fm1>|&Da*G0~J
zSiq`Ax^X5^g!@$+J-%SEa@6aVYG;_Ljw=2(KVC>_b}tI`ky!6RQP`974k89@%3mO%
zqJj<Ll{UU|<w`Ji<VH=Ha4{*K=~n8C*BY1|sXqth$C=8E=}6)F*x1;7g?u)iD})y+
zjf>-teDT7g3}5(KjO26Cef0A#b1muz{a`V_aJ#2sFLYnif72#yLIs-6NZcs7catJ;
zfB3Bb(mh&?jK=-*=k4gEwnuw-J%Bj9ef#$Ep2HG3B=N_Mt?ljD;z0~=cHsD!dmnz$
z(>GUJCG;u2E9^=8MB8L%zF&qmS=)KK|6LhipU9}FWWa5yuU=jF@OWQKYpS8q_3Mds
z5n_u=KPDdplYeFN?~~tcGQAnK@d(tI*=zazDeB%yQJ+oUpa1hJrhnUym!I5v!29E=
z;Im!-+aKn-&3(OwtFVNxsm^?{UYe__b+EikZ$_-F-Ca<<GVW7f9*&rCj+;87(lFPc
z!ZNMN9AZK*>#29k$`n}Q!s1k~A1lPW<=~dpcpK^2y-%fL*}BJDP^cq!q=LM+9K<>~
zIbG4yn_BcmpqA5@l55pRmbnB0^X-^8S-a4+RVLb-gNJn|GS>g@;ivlg2$89taubCf
z4@2E<JF#m0IZbQI*3I|-u8};rbIz~l>8~A?GWt9`xTua=*$&z_nTSc8Xt+HJwO2%U
zdjEN^)tl{v2irNn8yjy|<NNk@2$LiudYJa$p^WA8$=dvnmq^F>kH~^S)vO=za)~cr
z(u6zZ$9#19(d%&LEVLN;<?dV+#QtIte3V&!Z{UHyx)mk<X8g?$k3IP#4yaBCf^h_E
z4j+qJQm9iO=1UOmoc%FdpY1Hv(2S2w@Yqbg|FT>9xBFXm_+~7waac93(nLd!J11E;
zHu02UvdDbh8REnGk!RjKiJzuKkMr>1qDpFI74`C%v6qrJp`K4(_Vs9<2}7?P49;XT
z={m&62ZV)~QK&$6xZR8ftY*kpReT3l-efCio0ynjN5u3Ko{QjTs(4qH=>b^@34*t6
zY)+!lRs!;lUl*<iL9H{Fu9zc=>L<LNoAVeM9UTX#8N$!cf6>{>+WK{`KU<GRQo{mR
zFT8GO7#Sa*zHQIpuPp}XslEp9JIo!z{(lpRKkJcH&ENsfYztB7eQQU;X+Z`Vn_<vE
z-m1@d-4W6NvA}hx+Sx@lD>XG7LItXECM)!)N)fgKi2mC-jhrE$rRkdxF?S)oB3Fu$
zv3@^Zd`VE{T@kMTpG^?VcUs?GyJBE~(8)I=_0bbR2F5%sxO}|t<j$Q-rBH5YYHISI
zyZgdlos3Adj&b_dD=l#&hteyE9xHQWgb-epsv*S{E}??+t^D-q)7SO&L*cVoC4;*|
zgwd9CbxsDnx}r)r;`8%NUs7nTtl!E5Rd-_&=4pgIk|QtQ+5WX3(2@)>ScFYpUrnK}
z058cMXZ2Q;V{t}CkwygQC*%{jz7xzWvLCCWy`ZRw-#WOn!mk*E!R#`ZmX&R3ZEc<F
zEPD2AcSKYaV`xDn_Wk>dEgc<42-;jK&|J&*zO_<zg9@$PDxhbIL&h(dUG77>4qm!E
zPnZd^y7Wd^S*Yf<c351YL%SEY2K0`3jr&dvc8rygp)G$mtedhrTTini;}3hNagXv^
zAU$5ZkOpRSb(vj##9IHQ@}U`@v3ihmb;47DM$uPt)kC4I*?#=%C;T_gi&KUxj(#_9
zz6!V5y?gij2@#R4vl6QMwX!PTO~J=gp(}71XS+9cJL-Ji&~D3*pJ6*74N`1k%`6pq
zb}K`|>u)2(4DL23Vao0eTx|69^z=L}Au;KRbzo#%gDdk32>4~1T)Xzlz9reSy!pLz
zWMt$@?_wn{J38mmYhmN88>yx4OfL>zj&ml;I;BT=ZF+|qXEI+cs=dA4VRdOn88>Tl
zd18;~3)kFRRl8{kO8Io_uUnhcL#vPIk<~3y^k3#>($1e5Hk)MY=24|<^Mh*hr#HaF
zz(;p4zNcbwCaRdTuwZRllJyX~h#uS-lbkGSXld#E#kn}#23F)}fE2+Z`A~78a`3e<
z7`CefDgX=lc1yp^!YnRlafegO%ur3Hvc}`RG7jxoV8V0NtVL{hMq1k0ZQHgPSL_pz
zztn0)qtTLcavCYH<VIo9C#R;?ZQZ<yXXjR34ULDX85wehDf)#)>?UpD`I0Os4FVWj
zMuwRhqqEqBT$qLDEX;Swt8iXTxa#WY$WAaaWq~(G?%KNfm8UW9Q$?>b=a!+Fn6`A2
zcqlZn=YRbpE<T>t@KQz3($X>|-J>ke;f6oR$i3{6bzn&3`z<;<;4?Eble4lK{8wjc
zMssU~COoPlWClW_ug)`{rQ8v&fx?+CAuUax{8HK?Cnsl$?`v(f2^Cbj#mL0kjnc_|
zcHO02b)8^R5jRq2C+~2qEc^2jj;zzKBe60z{3uji=q^i-pJ_aU1qE`3Sge4Dnz(Vs
z*Ewx=-}>6(-}Czg75`{6%e|%9sBD1d7KGb8-hcFNCA2k!x_iOQwx45eio^n&N~PmN
ze#c_tW996L;Wk2qr2(M{z1XvF=*pZ$7K@#f_3|aRaXx9ZUYa`T+5PJB$J`cA^x|YE
zPk|-ixl-t+X8LM}3GH*umfm&y#^+esny%{RH##4BwG@X={LWQ67P{cN3c_BQc^Xgl
zKcc!@+&eqzI2_9zjI^unXy+X^v(vAWwVtFG+<#$L{T#}Oh>B2CU<_{H5$_9-cPnrl
z{T$1f9j=S;q^m7%&Ga4$Q5c`CmyQ+znayl$X&<Xyz=d)QV@~nOeY}^i+1d}57Aas2
zhiQKIaL4Nf;_l+^Oz0u#6X&2su60|$LNi|&K#m8{)Xw&fgCB2Addz<iUb~9J;V#qz
zJWT>8W6=R~P1^XZ?DWcNO?UVFUINDn(+s;R$q4hk>GE6CC5|Ha>bg*vHn9L2Oe%4E
zw`!8Ar07Xs=eOasNSe<OO#AflE{NL#`xak`i6!mX*J0{i)@sr9Z=kcHNt~mubF;C_
zuoN@xm^3ol3LrA%<3e?HvVl68?z_j0K&YJW>FVmrSxoH#G*I9+as`6WP)SJ%Dd%&x
z%mrq_IeGasAZLS9f1Zxxp7C3ld|RXzV;cN2i{v|=YFST(tpU^ea<8e~<|*i*#N^~^
zb15mQsQQudeA-i~u9_w$+rFpFlbDp4cw8?*@tQPsHSe^zxRHrTJ2`pF%jb5v^|GgY
z@fS!Imq?GoHs&;%=c9nyQUg%IPr|)inro#2oHcu5X4#+kfV%a~(4LDB!K;LT+Nq$q
zOY_B4Zl-79)syfV#nR<c_iiIO&H(hLo|agIKdb5(8}D++o+|t@<>dgFd1Ww2+5_m8
z^9W?r6a#Wml)JmThNZ#rpek31)2BBapc!;{42oYP#9n6J=g+!_czGRqDL8(0f1eWC
z4vFDT&D`1_L++Bo9Tq~st_XitzvkyZeXFzrvL^31I0P+~hnqrY&s0n;_h%a64)yk#
zs9Q2#$IVO(JSUIvCVz;pjTpEqcFL>9^>d7j?^I!XYjbnrhbIRd*rPOG=xKWlyx+!M
zX{R#&`sud<P%9-kZP_A#JB)1_-!q%RQXV$Dw^7FAbf=G}-hjp^V-|w?wp8WiGfDJ)
z3SSN~iZ<U5oc~<<;>9l&pk%1*HvAW0fmQ}188S<z9}thCZ+2F<IuF#1!yar!)Q-)Y
z@N3IVVp2jvf<Cp_`i2MPI1@a;?Tp-1KQx0B>@pzZ1%P_mAa~CURuPj@$OQ8z2ZTf2
z_uMCi+eJUO=W39Ygb@<7@_wr{@e_E)$)PkOLn5s^Ei=;yvZRPTcV#|K#XBo6KR+@Y
z1RA<~#?nIWP=0+~+<n#>84Qr_$wGfRNsTcJTrBf;d9`K@_K6i?c|_mjyyo#2-98a7
zyQhpXn5Xk>=rKnt7H$7}6z8=omVvH|vimt!3n2noG5wm?p-eX=D7lMZ#7(kVi=4U>
zi!wv$C|wL@EVnn;x(1<O*rM2|C|k+Jg<FxP3mSRmg?A$%ryVTFC@eGrKjafgGS-g^
z9otK$m;HT*!?i^)D&F5($!l{d&%1l2T4#t7;F&j)*rGXQTw_Pa9Z9mX=jp?=&Brg2
zEG{O`8=M`EmCHFBf-LF*fcO}@rbJB!5_Hdkf;^}DFPu|Qz*1ct99p1C&W)vz$E$d)
zg1`Pb|4BT%APROm(?N+k71(2Q6_eK1)@DCH-u8(`+c)v(OvTue?I^F6|21GYQ@PF=
zE5~9(1Qhh)KhUm>#=D4EOlDX3Eh2?GEiH}cF{9Vj+1bhPo)qMza(Y-9Bg<^&#18b*
zeb_K@!PAetURm))2p(OP9o*L7H4q$abJfZ!`@w_FNW{j=J6(5j%Jg*m^5shgY!2K3
z1ZK<y3|N((dj4{4zD{kynyW7fvqik67Et#=DN&zr4{-4O%0xC6Ee)g=z{wR1CcJ)x
zSU=KyJK7@omBy2CXwx=AM^C4%we`=6quK50<kev@Vp3U|@(7bDE-9%8?A5s#3&lN7
z!G&aw#TwnX;aFe!CM882xavl(JuNG12%oadHi54<Z{8U9n9ghDotBa^hPojXu<8mJ
z*_7)!ky&i+lDFWaio}^Rp5EZ<3fq3(aew>3J07s5T0b#wK2Z+(r5=kt-VU_5b3K*9
z4L%rABO@cgJFS2d?R&Hmh{QygbZTDS_2w#tztMAz8m`W#8dYW|e?BYeR`=Yl`Jc~n
zNG|vjD4fU}aOpTSJL91y^0!S%Sd_IYckUEDaYAz)qS^xrO-B*UDBakT^RkZ>-N#}r
zB+tO$&%)ptdAI9*C4^Oda=>|V2rItr6&eKK9MF}0<KGuvoudTZs}5?>d#P*$YkSf|
zz~u}`S4O6#r*krPh-}^%(Ahd745;q=fP!E^nSjy8fVZynKRhtkYG{CT5N3T;ujuF)
zdU+LngSLaDdBJ5BugS(e(v*Wu@<0-dEG#nf^3ERm<(HG-4lwcZ^0IlEN2)$qJyXWI
zx}(g>yBC+T|EO}(;NYzue>M$II$4z{Oaw<aL)Y36a=gRZ>T($%f{_^?56)Ea1njg{
z=qD&P4M~Q3g$30J`RM|tG|#sUfgPU<uzKr;W=@KW>$Id8v=dM0cmR(8e|vMBONOfI
zRE5F_t3$%+Sy}o(dKxp#^5s0fUM=l$9vL3~I#VOu4lq4ss3vRz)?Jvl`~Bw|IcY#A
zN;P7;e)0xtydh;>lkT|b*Ikocg*wH3gz}^y>5#FdQLh`&sd28Dd|O&-W|^(fAo+|0
zgy`y_Lx)bEJlR+y93aRVMV+s$&9B^-q#98xhe8pLA_V{VAbUB(&I%f{HRVcNuZ2l7
zhkzSW(ZD?hdqhy@_ljUi7z9Xd*~bHWG8>QntGfTu8L&9af^3;E6s|3FEl<4+-;r|@
zQ|^KaY#=JpMb7R+1>S!w{vQp48;{q!{?#(j9^C!E-6!}l!2b^&i2q|(9$!b1q;Ob4
Syhjo;GK{AFrL2p8Joq<XQjih=

literal 308785
zcmd?RcT`jD*EWb^1-taB^o|Hf7ey%nqyz{M0*Le?y?6CNr3EPgsY-_cq4zG*r3FHe
zF1;8dUHW&&=l8zfd^59Vt(ou7IcvE%z{xrH+4tVpzV@{@hH7ajU!`E6AR{BY3Q>6q
zB_liUO-4p;cIg86iSgNC6Y%B23pM4Z;43)xpVS9|FPEKF4Bg1cnBS59orzl9up}e9
zLk4;JSl2t{=alV-RQ&+r;f(K+-1cNlonZHy52e8$&6%#fJxgwO@5jQe&tF{5*U+Nx
zXoN`Qzi~=E`|e}bXF8$Dw(e(oj?jPjterWjo%e5<%1BM#JHPh)P3ohcl8<E7Mkm{T
z$vxbv6RsM88BG<%XOI3843O8OYV<3v+*{kl{#xFbTi+KpFRDEIcdcY(SD38_$^Uas
z)^P7!!+*~2WnJ<<^Y8g{m7Bl+b9u`Lw%POl{xtZmVhiPeE`AagO#7eHw^V0@|8shU
z_bkaYq=OFn|Jxh4`0Q3yH$HoOgXhv&xuf50Z&<DYg9L|vlvp9z7vWGQpc8NThK1xz
z!D?ju|Jzp27rgyD-ZK9ki?2%cOklmq=>gu}!9l$e{;c^oW%K9j0V_}H_NM))J2yVR
zyG%y*`*oHry8c(48tc*Q#Y(?DM?F4j!bG(u>+y>6P8WOB6<^@&WMpsSN=L^ZU;E5M
z5wKHWiVHd}KGnN;odr?-vIsV8dW<t&)eqXy7am*z*Y-~*Pj7PtEX&M9a~$_JABC^&
z)H&AtxX{<vSM+^Zj_4Eht6@tc#LPSWEm#2AEZe=`)D2!U0oBucO?#;#lc+3>+)S^g
zt<<`Cq>NBY7wMK$0ee%q0f)OqjK5T|#$Hp|`i*OPs>9o7Bd>s8x3oW>Jwt>D2dt`j
zET?*{t>%7$F?I6wLVeQg&CDLp{OOQzpYp&9Rv$;s{JGjUFmT<u0(^{dOQv~Crl(B3
zyu6%b6P&{Xz4&ab7%nsFA=<0?aH{tuYv3QVfbC3H=2;ukc3nw+m0<H{x4Kyn@3C`$
zZ#L=pC_aI*%)cJ^`%T?ezYHhj)4@ot`yrv)G|vnKjl)<zCqvdr$WM|v(Kx$;s?%@Q
z4>M<ur)G=;8-h*)3Db>!!<{5EJpXf5c2<s&;AzCp&K{GMm`Kn3(BHd-9o$87YKwfD
zJI!0cZ}+ERd1ECEZ|s%I0#*jW5~^n~ngMG%$qzNaJ;>UedgE-y-upH-`MbI=blv2H
z+@HM(7G6l~a|+z5`+?!0H{T8s_BY$#U_ISnb=oZ<eG0+R@c8;{76)W|iWsO*vex$C
zlk2O?@~5>7k#EzQTO@nS7K>?VXvAb>{L<e6BrGu_`pq0Z82R;<(b3?<*U?d<(L@ua
zYFYl41P`-pBEN(2@B-nW{oVt@%n6}<T0F@(=L}g3%xedy@02$+HZ~U4xwW-5UTKfF
zx<Y#W7DbC$P6o&3W6$GJr(+s3K{B!$p43VCS@j4Jir3HgMY8iqOLh4Ah<v(}#ZUdy
z{x|D**5;F)Vm!&3wx=h@cGVMnVG5RpJ@ERYX8Yns&tP^CRgy9UProOG!b*N2?@tW7
zBqVErUw~))7&tu|NN5h6Ih@XwAvWlAeDpYEl6JQ#e+OQ{d0KXN%-nZs-NdWIv(L({
zeSURywU6Nz_;7G|*CYMkf#Cg(P7bOj9A(UFCTzsS<mH>cw~_YB>14_2#Oeuw#s7Ef
zVlg)t*Vyx`q-VapA~tIbhr>^N*{koK{9rx#VRh{yuvNohheT64`NeC66CLn^3CH@E
z`y0v+H>3MlX#PHp^uh_vM~l6tL5uvRhSdc5qc-`Ky+GpL13wjT#qyNQ%)UmuOT*Gu
z(C>>upF!Z;>~Fl2Kl(u_A^?6*M*Hg3g$9Wk|E*2D>Cqgm$9hrqTFWbP5FIgbmWhG8
z<)2pDSWikE>sE6fU%d@hlydZ^Q+~85;BdRhDL@TuY>X~dNSgQD<;K%vBEG14x^8m!
z3G2a!-Jjn@^U_HMrMuq&x4*cvq^P|@UF;pg;<FHaweqw}`C$ezfavGDJ*3PD(Ud=~
zz|_oz^)aylYZ&?sCr@V){g-88vSv<?XYjf5$F|gpE**h*gj;0bX6z@Pu#7$ya49hO
zS77iuOhM35XV6F`-o&q?%Jk0{+HK{aZRO2?)3Xr(0%niA_v&Ygf{yV)xRsTaIzr&0
zexRcK$wt&=FH+#9P&*BhfBg7yP&07zcEDEh%m}mBWI8eE^ibOWPYD+n*Y-HR3FhQK
z&$zSRe6pUDn##Pb1wLr<0f5fx%1~zAI;P%zCSY%MV&<f3X?fXWe=%V_A!t8gC%Jh)
zx$alp`UN?4@F21$HlH~eNGGI2c=C~&*uSUXUXuTZgh*c88Xm6y7o^PSgyj3T{97^^
zWbqz)`MM(nPrdlX&+kN?n~1^#(HxLC%kRO;V&31PQ6JfJ-I$bdc6wk}6t%DTN|G=8
zOmB8rKw`RFm>%%8<=9oi(XfJ;EZG47@WILz;v;tA*|t5)6={bl@!-<{g8<&yCSCL8
zA@5`|G_Hi}r7eHE3F)B2*@;7-+El8;x;wrBEtzLa^ltusF&t;fYhQ%G8wY+n6ZC1l
zjxX$Wi_!7(w^C);?jVV8(GT$+3Cr&*<E1i|Po}68TZvC{2W|>QJTIqKER|i}lwF?r
zaT&9|DHZS~|C5?OOuo~@Sm<m@3zK2$dR<Rh#LT+`+A|t*nsW_gExtTM|7EEEcYK0`
zT~O;Hz09FO^C-ETi>&@y@_$;d11s)?={G2RI$Gc!7;7G~S@~5@93{Ul`?<YckuX)S
z?-a0ZwDVS?!OW+#uW#i)T;}=jOWqfMroUss0*723ZZeHuq(40QX)Qn837m=gC2PR8
zw%ox=CMjL))D8q_n9-%}HT`GdRTosnu6$gt^Cj4u78p4->IEI-2BAqI{p1*46z^D1
zcmQ(6Uc<87!J2+Ey7_pb83#h!(%5Ga<$JWzxswetSi|0b7*>hLjo;VCcJQXfQ2{^A
zedifx@|a~j^Wa5}c{6f*lbIPj?(>lnYvindf-`txM3h8DMGuaa<%RJc9;e59>tm(n
z%a#A(V!=gkg1s&o+VAWX2N}n>Z}mxh+V!u5<A5)FLa=+yr^j2P|6y<~qY5oD{{YyJ
z;fUrS0DQ(hkeGCS!?8+26rc6~GL!w?S!M+Q;4xXNtG2b6db{NB-wh$ySj~q%(+4C-
z0LK^{Yz>)p@plJRU!nKL&$6M>9ND($44(M5Nw?Mdq~<^2&4ZJZljs4E%ZuSAR|Ozu
zbUj-iU6I4h7-+t(Zv0}fvZ!}%f9Sm{_*AC^ilr{q-2Ui0m*Ruu6Ut(n-|&P{oooGj
zNucAo$1Tl_q^BlOVXN)}V9N@mtBTjr^z1D|1o{gF=(~}EH<UZgV5c@Z`ex%Vk7A=h
z^-HB#-_xr?dQHUMz2nE%&3m6CQ;<?+x7-%oLF}Ctk*IO7+L_+ZZwI?^4&3&um>iz4
z#bQlkf~Qab30J`3AXo@&VQy|7lf|x#Ec~8>Ux%z+tjoo;$R#ut?2mOm%yF!)Uvrc3
z{HYbg(aZnMw~?g)Jz{#-xBxe@>JIy7H35C+5|v^!2QSh*+g5t^=2aFq3(lpk)sO7_
z`%TRV6CTJdLwmzjQW}|!c?`BECntvk(iz2STCVhKC=-AyAO=!S;2n>5@WuOx>Io3k
zy=R=#md}Y|T~|I>CT_~I{<gV59Fgi`T9v8Arb%JF)>6+YYTr}do@DCYN};rdODS*4
zCWu%CXOdnkm@@p3_-^pepFgg06Hne<D;i%?mOqH*Sj80w?);y;po6Y)o|*W8BLH~4
z3}N4-IQQ^Zy=Dx|!POJ<KW%ib;mMjY(KjJN<lvWAa{<8EW|JmMAHW<w{rSorw_Izo
z-=lA5Wwq3>```XgXJ5p;eMiXBg=xJJB$CbfG69cG`R@Emh*RSZY|48+lK<f0W*7Ux
zSMG<UYwIR1?bHWzw9PdQB12oLYBP~*p7mv;01hEBAo^lBq;l~Mbbp$POG`z>8LqAs
zsG@cGdfl?79Ql#+XQFI=%MJ|x`6KDJ@DERzZZVz{lB!|t!``L#y<4iRqT=aQlhlXG
z5hCDr<Sf)eR*_0i=I7^Q4j2zB;YHxn_8*gypw$zLZt&yzKcS(QA<(28y9deSEJLIB
z#Kp5simR8n`Vwv}b**u6b5nPUxII0Vy8uo(clMT6+@*Vi{K#ha$1Gk_8rWal)@g2r
zUkf5s;@6NezXqqgW{M7p&8C!;l;ibY#UN30HSYI0nRt*_ZVml+*gk5Jd9F>hn&$V*
zr}a2DC&#^4uzDrl$g%eGe{I4Q7V-aCU6j72C3}nD=XGz0wdLj29fMZ347;}v>*5m3
zFUUy)4=Bp!DO&QYO@gFX@K|!-&g#cfYwiR?J;;K8JOP#86LbYkQ;6$ebpw<3lEW@L
zo5F~1Q|@K4&$tZz>Y}K)o~+|2aHJ1ej<B_SrNUZ`94iD5e-SXCw^FR0rs&S2dL?qH
zw*?E}eve7_Tie)wy2omOJ+b1$(?!0;Am;OuX?$PY#Ur?NBr1Ig)a(=p4N!cyWD9ny
z$vv1Eer3fspdzE>e;&zhUpNj)e^R6inVy+E*erE!$vjW?d>&M=&o7D3A><!Uj>d+E
zdXD%%!^<Yd@;|d{@=e4IM;b)25N*(FYm1_lt=H%P18=!j1!X2!gvn;U)C<}_R8q=J
z&qWhgABn<N=$}1JQ&Q*=2bB={9^q`u4Y0D&S1@#G7p|)A6j9@+&@NHu$~F@1iSDF%
zVW}wfIU1ng6%%+Ie8qAtCnQJnvX|zf6jpb^Ha*!sJ;8ML=2^1eUqEHuazlC)QaRRJ
zk=b-`IgCS3RylgvPNaY2{bjIOE!hPLs5MKN-VDw+q+l(*zyyX)rlxgp0E~toun9T^
z+q;X3`7Oq$=CZELoV9uq%N!tl5Ou2jhWt|2lH{Hd+PvKuMsjkplFA#Thf&BXV?=sX
z>mBQ}corEmKWO>^wtJZp!9F9dy0t^JQw@VQvF|g<oFRLADX5C<`K4!ih)Cr$Wc9H4
z1czb1qlrz!oNdOJv?PE<SiK$N$a5I1LKwcb+_p5qUTj;e=Yl<1@LdJcj@k@(=cTg0
zTB@&QaN#;^2x+i2+ff%C+D$5^eWpYq_wQ!*(%E{74I-0@pA@NgZ$0>22}@eHQl%_d
zO=@0J0<Y<yD6o1q_^y~S<@V<Hq`5C^&qZ+=rZxhb3v|5wg$yp3Frv+juAcCO3FEdi
zfUXOU=s8=w0S`oa<Qr8DbVZ7<Xaqv3=dWL{Ol~aB(&LfWN=Zh?0`!P`Steygz}Jo^
zO4HI_a>7@Z%E~sGGJBQI8q4Vo@mtx4$!cf;7v`|tGx!|lBzCLTt~~@>7H1y8#h1pF
z5lF^V_t#96K=HW3J0RlQU4^z%I?Mdh8vfpG(J(!_3%3jZ%)ImHAvX`W4Zpg6z%xC6
zk1Z$!*5wj`ZYl_#=OX*7s081&iV!p05c7F)Tim`V;>W@{;EZw|D&jNRebz<{WG%nv
zi|NS_PcQwYvSc+CU?;&H!STN($T&xLsTF|Wz9@I&@6DWBz2C5aZ4KcsBB4ThLP0Eb
z*(jg0aik)~YV8%<z!l&3e=mG~o8-42RKuZ~LSLmv`zi<fEG-tLm2A`UXzGB!*SX4$
zShTbLCD34vTz6;B2j8_Q+ID9jI>~-&$V*_jx=_FL)eRPzH3EK+T}dIQFa9Y^_V3LR
z5t8#4gYQ<*ElZDf$I-xKK;&?C(`ANs;Vc$bZDjDR*%ra<|K^C~P#`?L9eg3I&n_fp
zyj2N9O_g@(n8d+VDLH+79(RKJ3LHXL{zi(6316R@;;)0`JUw!K7dQ{szf#t0#?${*
z$a<@_!f-y#tpim%RiU*Y6Xw77ugCzxN`>S$|8Vg&btve&H7bl^rgIXhT$wY$L%rWZ
zNx}};ILW_3omA*MZn8yPqx24*zqwf(hgY@x8UpYXc!=?XTUQ)E|9i2H&tH@Oq=)D~
zds-HQtP&-dsxC?$6LYimaBdZT0)u+<(Q9WG>Hl6{WANAT6o||MZuZC!49&bkYT<`U
zSZiq66;bwAd@K@nf3I^EVJDe<Nm{9#{gst+R$)n5jI6Tj74iQ#5ETgu6u(%r13UwV
zkkn-T|2Hj<MDR#Q@YdaaHBKMiPCCW3XkuB;Wisat(h|ZSnZS3-pfKnxgCYyJDA|Qh
zcmY6FQUQPc;h)byHDS%0Agqy!#)tdenvN^tfO-CEI&_r#-9ipWAVz>ZvNKxUr03+_
z#<)=TD`)5Yy%ZY1L}Z0%B~}3f>ycEV9vHG{AMPL0dS!n>4rs6su9D)hw4y|MM73hD
z57(=Pf>9V}Mi1hCs(jvfALT<vJN@LkT$X0g!EkOsvq10Fozw3FfjLn@NAvvS@_^fu
z>QxV7<KCLvr^A#l_hg9yXRhS773m%(+ycSv8lk;HkMhZ%Pep7^sTWm@gmWq2&I<ho
zX4FX8CvUmYb^SEiSN@NvrRK5L1KV!;RMgg-DfSg@v;Mm<^tf$4e6~l$Zd;Yv*4mCM
zM2MAAT;`U;Q2_D?v2h%fv)z5sVxChfQ<MU;&#7m6csFr;Wa&7VOo$J&zE)WrP(%{a
z8W%W)hXB>i9|-v6#<(_NSLr_SvN~Tttr4heLwyo6h4-#_Y_DhsfZ856A0>|=31fAH
zs)>GkQ?fP^P@h6Gp7g1UdYIy7YwJH2Rwk`4`sUp16~U@i=`Aq8HqVj?Jp{g9g8B|6
zoNff=9TuGm8P*>WUi=8&-i{L6i3%$0H#!x5IkTsfsH0&}XPXah7enQP<l?jbQ&SC_
zayouBemcJu6X1w5Ds_JJefT>1FzU2gX=aa`!gRN4>u~E#4Uciq@lM^z!FmEKtCXun
zTL7_knT1+0ozba5oApo2#rS4<QPIqSppz}%)5Do!5(Vl+G7s+(CI*bad#tL+;D&Zr
z|7uzlMQiKeTLkwo0*6P|$&<2gi4PC<5}A>~GHX-TFNx-gr7cJ72rB3_nd17Xcl%z!
ztIWi@;!HxIT%G7@;9k_o$>9LzQnaS!d{t7=?&ECDqk;0nGi0*@2LI1xQ;Tz$TY(j{
zOGH|5QVwU)j)$h$^L7i*t^fhcnM6Utd!ppi`{n($V+=*TQ5-rJ496WW?|WLl?6sTi
z5`l_#<9bIieue@}TFXTFi6GXZ<A+rOBM0NBLPpN6=bpQXF<j+@ut7*t8Tut@`ytrt
zL4Vpn#n$4ix~X7m!YA9dRQAd~F5$QlL~#wBcE%a9s=Qz1UaE|fAG$sr);2tmhgEIS
z>if%YBC<pDpmFFw1)q+-UO$d>a9Pr8iqT~XQurP#ujRb6E-}JF*57e-)WBiu>kG%!
zA6TgXaWp22`;-4>!%jfcn~T@iKxJ9*P`?1@{@pS`P~MYCxoJRfhPxKiOMG(d;}RqC
zcBY%Rf4Moa#C_pnKDOe+a#4VtJ*Q6-_smw+x1eyH<x8K~`8N(%WGau6M6CLnUy|J+
z!CE|uaJpP>$I5h)Fno6UOMDEI)Vc$YYZWn;aB2+@JUQe4T1aYo4`dU!fV69T0EFgV
zy-z^NUf)I)`&;gp5)T;|8I88&gvGU~c*8P4b*E;AEOdXktWwESDd;(3vXT!gwkyg^
z^@Q8Fj^K8?Z$?Jq`g{F8X{IfUC~LGXcdg0}=v2a)x9^cTn8n$U1q=3YrIy<fE%+yl
zAdQ91t`eK0-$kSeX}Cj<sVeLXbN=u%vkn%9r;e{d!Rp{<$77V{n$T9e4KT|eKctJ(
z7j*V$KoySXRKPmN^K%Mb<=D5LXSoX`x(h(}e;)42?WHzwH2ijAIa5Yo)?~#B{Rhd#
z-*ncCu<q0Ub*+*QzgbvrygG#8V$+mSCReo1^SemK=&?z&9$V?IuS3<oiO-#$1I6wq
zqNT83;`9#2tLR|z*Hk?!C;#;wgTTJnADh`iVQh;&7};D{y@=KR3v#ywsbmr5oF}%@
z_kDd!a~)&s>S%bJBMVW|hUxRE>Y3>Up<3vV3NDD0eMRA5on;lnc-3L|7WQ$5FZXn{
zsr->E+fp%>^p*@{g`{L$mA7K)q(kg9oiQ-))bU?r<bW8OiO=-LEy3zL&k~r7{lt?w
zLWKR2=iUL4)XCgZta(dGP>?fe-NKGD#6*nVmSf@h?Uq^@pYY9ScI8or;q@Jk9r7t1
zO?T2xH!cb~4H->u#*2S)5@YyV;sVvzzN#_qAG16TNPN&h!hO7l{<(&4(M|WRfY+{@
zDzEojQ5OnVZI8wwB!=ddrF+!)%_G<bhMu+$?+)4|H&>p>q)03OVe-+qaW473^4sWn
zN&bh*bCNmk=Yz#bj!8ppd}{A{+I*TbH#TqW;I}|~qAwS(YZ+sJj0Qv?;n}%5J2vPF
zMknntv|;b1YVIEI$HVXRGt0ljPs^36iKG9Bz<(}|H^%|mXL8>-I_T}*nPAbxq@)-Q
zX8Az>K89N5C>9h4;d}XV1ITZacCQK&9E;byxrd=C{DOr{<{FdT3%0Uz6|`i|F7r)r
zvayVab^D)hg1B^&5zSM2#Y0^|2gZbC3~)_F;qNe9(@VWZ3nRUq*>C*z$%(jmiR1LJ
zwQcmk#O>4Od&jKXzl8(UrwZdluogI-b~iD6w|hfxQ7&k*fO>zkXYBu`-haL&9*fZI
z_u6sllm;bVTAIOoaeSXH+;6F$ah4%7(f6}`PC=cg@-x80acQCNoxxHecK`tp{FYp+
z6nnau3{bx}SAJC)-MHUuGokzot2fcHT}IGr^v^-t=kO;nUtW-6G}2C{ffbEYOCOz_
z>TiT=27SVA?5yBUlYjZ_d|FW%&=&T?cura1@X#20-W&~vU^nZCA(`ooYY;5HX}5AX
zjkx1mM{GLefE4>pt@nd0k(^tB9gch|FE>5h%<?kzl6U>VDngU=?ZC2c*p}PG_Tc!*
zhB|a?i-wMz?4Ei}%4IU66q#BhbJhbj>hJ-+pryMg7oA^Yg-3G5(U;ekj}DL`u1uMQ
zbx*>HN4LK(nF*d81v008!fQXXap~ko)rfoe!P1gJ(TG!-&1aBYBNkgxEvKlTk(4LJ
z<0?sn1g$0GzW)9-d>{c&b?(*D($Y3yATXb_b%0#iR#sM|Je@qBs8o2Yme1glQ{Xuw
zzN)MW&6MP?80^JqP@wMnZdkaFE~b@zSyPeGW9J=FqwH3BC6bhPkVJpE-ynwqsOBN1
z@oFh!OjC~fqwv4tZUGMNWxcKz@)sC&GKZdf89+eK#s^t3@v_}TcJ`*R!Bn9xH$RFt
z+aGVkv8_uP5E&Kaw8m5#<7=2-$OFQ7d*OuuduDf7?s4SP0kId%4%rM4T93A&Q)SUk
z);PS8mhi`sNu++h-_&8{R#_?qQ)cI~U!5<y+1Fe2sdkocBddj%H1h+ZcF>!=jSUw;
zLBUeddW}!)9dvpH$)c>KBL9oj^I>abyrz2Ex4^Msqht6${=cOjd(;1tdX|(#m$6=%
zfHih>c5*@>F?l$<)Zukw0U(h91$aLN7L=x`uC5Y;71JyK&h-2Jh0L4FBo9V<KnX0A
z;!+PW91s{2k$Kg^HPXG$U9VpG&DCa43I6S`3Wb>j-M$*F+4`7DP>}l0+V}|$>fKK7
zsIm#Y@ZUO4a$$aR_xOTEv#wFR<=TG<IaN!hGGsMf?LPUyP~k2Z4R~Hj43Q4^L|>Hs
zJUl0#jO^c8fapLocQ;I<Aw+p1AL(_-Z?%c-iI*;Y-NDCS2*`bh%Qia1X-)j}h}Wk*
zKmAZ=yz;juAIqUAA*jt7*SMQX5ulF?4wEN?@rBJr%$VAzwu1ghfy}R(C!~JrujbQ0
zK>-09;&dZu?j&z7Ni_T5u8#C8o02W#{Uz`U!uYBrVSHgzlBXYTSSLJmNhV*Y;Lue<
z>jj^1QJP(rZ%#CiGqbT2y<l@wC5*BWF(9x*8~<}+iDy$%;|v*M67ZYZv+GtNCjpq;
zHiEjLwjIYdRjrb(_S$!=3{R^?CBeS8%MuHMMcWGsn(rp^m+ER2@Lg_9i`ja@m8!i^
zdD*cTrwEhpM@<;N-(I8sf)d14v{Qhtyl4}b>Fy^n!yf~egO`El`2DI~PBAgJvs0~}
zbPac`U5uw3r)}9E?xW!J$wyc3_8>ax^!$|@zwrqb+A&`8j`FzAY@fX;rz~T461E&6
zl7v3E-{VmcYu7Nh+OzRHB?vUvs_yr9f#vfR;fBLYuT|YwOdGOf&z<J_X@{O$x2Ya1
zR|^fx8^j#!AAWo|rtV&X%1@*kmt(->xm9QWTF-TpV-_k(nu^~&p52q!PVzflCgyt`
z8;Rdc^%Hy!L=_ia4Mc&7`NpsNopjV?e7%vzlO|!5;xP&dQANy?8oF|FwIA|sTJ!Zh
zf5QLf3>h74vfwNktrmN~_zb^=g}CvKrr$K(YFYFnHJ6-uf_nD_PgGS5kmE85>hFy5
zX?$z3FD#omguzRJQr&jh#(A8JE6scxyQ;$esyIQ|UsJoTCjC?AdY}F7cc<=@)o(yQ
zMz{j=izr!3r$EA?>FYdWk9b3{yZRSoI_w&kl%NX`m+9hOO3+V_pl%q$j@zbdz|keJ
zCUdb{bt8jcwCr<Xgo;7$#ck(34)d(iF8%uCJd`g5K5?rrt9iWoChgi%h=VP)H&RIf
z7(U%f*<Y$7FEQS(y|zQatB=7OGjF3jHT)CHZoMOlIy{9Ow4eIJ0)Rx&?j+_eeoGqu
z({dbUbUc=u9A5@QUqmW+vC~EnLQh|Qpr!?+mR9bpi36nZHlvo{wyQ_Ef}-uB400{H
zU^Oh)IP;Z%!e`nxZhqB8#nadAz6t}{1s}>hU?~FbqH^C=-18E9+_FIF<q{b)xt6Gu
z<N1PmIeS<ADTzVb>Rlh5LD-;|c=`Gr@pEjBpY%wonC+MeB@jYZ6P4HoL>L&2_Wgy0
zJ+e~Q%7z|^?o_UCr}c;^>CK=jk5R4m#uuxrOot_0BVS)V-wA|2VX#uEslt)wfgSfe
z6q=_<NkKoI$irWFO^DNO`~J@Tf`VKH1(PexcoCo<e1Y#cu^^01;$4lqjm(j`AvjTh
zuR_^Lm^y~bn){O1vqM#zmoOqwIcH0ah6x77QDH_(*>+rF`9b35!B4|VAao>Lm=PZr
zHPp5(LRPuSNzI^&6>O7u`-{plMnK^U?8DXnqHT1eYuLMexKtyfR|!_!CNtBkx5xZU
zH$XcOB<xal=C*iOS8}wqQ9j$i%I1dFZE8*S^nz}?po^0Yao4I2l6B9t<t)jd50G$3
zsbhV%q)y+S?w0=P$A{Apan@yq%c-azzb>Z@7ca`-``jyH(Q`8BKN#un2gHfqB47Pu
zY)8-?NOzh<+F$#w(p@zsOu%r{`6@D<Y$0jv10qR<-XL3`X*KmL+}sI$SXg=ElD!`;
zp>XI241E;W_g&SHnP{<QmF+n~-iLH>u)GIm9!S0lWpGK~HxW{X>EMR1cYt^=I`plJ
zUh$R~{;5*Z3O8;Uy{Ek)X1KGnpio$4S_r>yhKcQt;!BE%K--^dD9X~3;jQpXoyrWr
z^SCp6#6Z*LZ3NuKXZmn`J$j}R{c4pl2ps=OW>R_a=*o|a{eQM9Je!nSG5Tqx%%J3Q
z*9*|s*VNNxWE@r)*2!Jxb3e?cF;VeB;6=tozsp~c9*0y0@uJxI%@4dFk^f*YbWs)7
zY+178pr7-oC0sB&m4<OFxW`7!9VGY47zHzwZN!fsQcCqo{+vGPZ~9KwVxk98{JiDR
z%!F=9jEOXjt#x)>#X;3@Eh?<+)OY<dB6AXs8nwe4<kj$a<R#k5x{AYG7Q6B)U_hbp
zml*!0^4WKG<$Lmprg6Yt-=@*KQGbGiMeenR#pc!a6p1rr!B^kQDuew_m)o1)>vG$k
zTUEJDtsm}kdbWxq1ev)JR>16G+A6&`NnxK5mCf@5H2pm}%*UW*UmR`7`DVn!%Atm(
zzJJa;SjOs^cF={S6@)nEgg9cpz<rQIDl4{ra)V9G9+p#8oBlgezZkkjj?1`^B0Xva
z9K?&Yukfa|{16{oJ&_puhz@pr{YYik2q50<Y%5@5vxC>u3#QD|JtHrm_c7zEmcD;d
zy;ifY2P&aH&3KI;?(H56AFKu){nNqaSqNtX%INcQ@XVho%~zh7zzyy3Q<-h>g7%uh
z4UHUFQi1Hv-$<pjD!hN>+jS=;fOW8TL#Zk%{OXJ=D#3GVf~3~<p@Yrb%c{0t4Cw!*
zLuUu{<ebHfuj^6Dir3li6~9XZ9DUyeBTx-jb>$2J5I}hccbnhBw-)#9eK{h=%%~5q
zvH}>qzKmcT4PaAwEHz0OU-ZH8?c{uPu2KN}^dv`q@EfCiI?%aXeJT0)V^uwym6RuK
zyW!7L|40Eztv|P+0=}{nrN5cMIyp+N44Njt_&k^oyu^zF9JSyDxs0L6iu-*YqK1aG
z8bR#KpHo3+fs_VDcY@E5jZs|kF3lc!6KI1|4HLJIQawJ%lvz7L@2I^p1CBLe*Hk4(
zJnX*^HU#v2pod;(cIXCa*)qS&ZGj=v6K2y?0+PSYvd!oGu;#fJDhpEMa#bg9+4Uoo
z!$b?76g|~M=`c)f6>K6h8{(2&FoV%T1@G!!u=(DV2hF3FBp}QK?n+aa2y)1m-KgZX
zR|P=8rJhgk$S9XV$&K}FJBZ@uLSuBJ_a<wVWCT6(Q28a$l^tPU5)1#+8zN{4Tv*!$
z#A^<0&Bg*nw&yJoOJsNkJMwt)eD(!R9Wo3HND3U1lX$baSv$0HQ%%{;5acVH4!9yW
ztSte?4I{v%l+fNtKKmm^pa9zEu<QAkH~4ZrQ>#S|A?u0(jpgrPe;@?92Fo~T%KAo#
z!GI7a`aGrz=kJ5eEaa((qw=*A(SgU;S)uEG)W7%j$!90%0vc4``DJ+bUfxBdKOUbB
z!@xh-vANdj+#tKBOR~B*a%j;!J*Hu#)e5Zvw-a+-tb-;Q337%+WTj@d((B(jOIu#+
zC}dyc8Toe2supOzmBR+3aK*ITP;{=bIraID6$9>AaV3!D;gIw~9?J-Aac40WpK);p
zAya`;m|b?Qv~B(uI2g_9VAsq>ZI;C?Z#zX5mihwqKE*m4bxll%8KEj>yrX}vCT$))
zR}T+!V|KJe`-bkmEJ*QO)=5cF-Gm}V!erTJJSxU|aVwS#*QdlK@~#>yQx~(|lZaKg
zO|5}CL)P}D89wF;9s<CP$`=VRRw;!a4AJ2`*z|R%HT99o{Eh`@g~ZidFiHp>=WItS
z3->av%4igaA+=?G$hWn316AU17o4hE2y#Ih$x+b28>y5w4#70)Py-7etAzp&0Yby(
zSHcJOR|Wa#;wkgix_WKUc25acN&!kG?{<o~{xVWA4G5<pw;%uf6AHZ=-l(oTo(Q*1
z=is+w$_!h!DVyUJ%*y)Ek+BnH&X<-cDNI#d+6@NCjHu5DS7mMFHM{7r3yo+{P>=S=
zm%2L3ZdSaymEEG4kD>yGk*lOzA2Lr}L{-#xJPG0M;miW0T`lyQ*g?UpC5L$-5McZ}
z^H}PDBPrmAKr>RbKQ3MgQ;77u55d~3bbDkK#$Bt{0ewb@qsNXe$EPwdj`aMtI~EFx
zspBizw15juRhS~}WH;SwP<B9h1q2Q$LLrxVU)ZA|*1jn@zeK6~?Gqy;7#D7P7ffpu
zZy%AvmzGl_N1+9eBSj%5xd63rP(3dv*Z2{1SjMI1c=-__m;&tXx^8D}998zxoVuZ7
zrJukESl$W)y`f&?BU?kh{=*$_+G}FsG(L6yY}lt2n^|_Ut{zz^&_aW?d%5Gz)vJ*M
z<-WKAYPdT<J~}U*Gc&3h-iRRS+;&z;`G@EW14Y=%(KQ2Ze0iUNA#Y>|vI;jUv^C0K
zh)!!ueN7ge-exL9HZ})1#{5=j8eARFCE#uV=VGp4qL=)@^`ezA4W{NoLZHJuv7RRF
zf@u(-uEVBvtLS)`rhY>wqZpzP&tM3_9ap)gk9kH5xgqt?Am(26DS*7<_}-eAH$e}O
zVn%$EiI@Lp9RWMp<90xZs?cG6`HAIMhfn^SB6Z3}bv2iCYuej#za5k>Cgo$Xe7ve2
z!%)bOJO&a5Z7duySg>_R%>$vW>VCWu5OloqP^(%pfGP>GMp<y+o@naL0L=?vlu}Bs
zB-9YzZkWHCRJ?D()!j87SY&K|P`Ow(iB6@l%f%=?sWv6nCFJvDh<84Z08>TiR4Th%
zPJSbRdOi}jY8XOgneD9^t{Q_9M}l8BWsLUoTD>bX=&_5*%<(%Li^o1i4uv^a(=SVp
zX;hB&p{QW#_FxO=O!A5HxIqiSLHA12x*j(U{(_8>aT&cKhbj4aiWgw%3_t@N`-06F
z*>FU|plj$2=UP4l&Q15a@jf!s6Aoy7c&a)~>6Q72Iv~f;G;1nvD@F0zFsBjKLm72`
zcj~%h)P2R;lODv32i0<R*=uSr>^*zLc=?c#Pk5I?&cBd-h7MG$Uv_)%KgDRn;4h00
zlhA)$BlVIJC>l}xUe{9%a54-akx9phsJa$_Rxk$O`mq}NLRyYpQ&^d<O~mfY;%=9S
zoPrawAP~cU=gCV#Jh0HlZ)g@t{s*+%h|HI$N23Nl7;XN3k1U<8afMjz&)w);f_P@F
zXO6;gY(^}iQ(JpMI&s)4J9?cVQM3@&P02q%=X0$94otjM;~+Muxm+hbu8vc2eWoi5
z!XS&Mf-WH`uO{ieY~Lt32y#iKMP5#3(08V(t5qV`+$mu_f<x3G`4v2i2Fm8qxfLa<
znsF9+=aHb3L{ndjtkCI078>W$f}w}^Bn`6!3N?Xe*W%Kx(2gc~JcV=69tYk5$#JP$
zdCYZTl>F>q7Qz%XPeLjgE|)JRxi(D{bL+>!F!Nn52k9?`g~eHXTvIE1g$($CYE}*)
z->V7?qSDISXSYk@6^euKTG{~Zhk#ca;bMtRC9WFpj1m#^Hk~SYsphX@5IryReQ&Ik
zX=fIefQB@CsxIhVgSmiCnOcXU=6pDRH;QVlcv8-cx4)9gdW|}pi_O#z-2}JABt~l`
z-~v3#6pFG&nwg2f=1C9VZGen=iPX#qk&2=5O}r+qQ;XE*=jE91GANJ~rcOu<!{T?7
za(k7Z(06u?Loxi~e%YWcRXNBXmq4odJ!bTTg%?sUc^y^@OH+_ml4+G<FD8Ag*#;O@
zmlx<<>!qIvfeNIep~BZ!WOPU)C&UC(SaLtd^p>5NyO=&DAiKJ9lJ((VK=lY9N*hj6
z{R&rRx^?5SW%1hWUJLZAiP=xxrqsb)2$(b3nnJjx>SltSiaV<>ZTJZtha=@u6CQUa
zKv#hCNJ#MY1{%oH_-sx=FUB7vk2hKPc_==*0@$1~=;qP}SHc=6`T8m8>Gk}rpD?1t
zhY{hWIP>t*Dm{?eb>l)1?5HX&;cEDMN@n5$LpT#oRx^A`;cqvO?R+MZW=<WHUz^OC
zAtpIiP#d;Qsv4zeVxB<JNt=W5D|`3CRCFwg;j*z0QadRDu1Wz2ML)28R0!>CHE2y&
z@!1!p*>V8MH!qcD97d_TbJ!`yEzBd|IHqG;wEIlI-_Cs03Yv$aVTh8YhrC+fZi5$H
z3F}nQ1eJ(rx5syt&)}&bRhZPXUfb*BZCA`!LgWc>8rqhvb-QVDG{a&_xtq0cUR3)^
zKwWA#(GWBtYtbV0*`HO7^?KBx(1|pvo32u`%;5+E(`8*Fa}PC8yw;#rRYYM6bj%J`
zNrlpzOv)jZe5(1sr0l-dD#?M#2R<NgtQrXGaq|Ka_)^bw)PS=VSRP1z{GIGCi}%sz
zkO+EXaeBLfr;1I(Gm{pj2?eNaufkns+p;-Z^xAq)sLCs~V`I<@wT!G89P&ha!q+xS
zZ5CP1AW3r{`<Mf-%OMz0b~6SN3QKx`s#bEcp)^Y`J`XiCqxMDaYACs?pCU7J=M>(-
zOXnJN7o<Q`XZxh@rxgqa2)%!xM3H)O!GT*5Q&7W$>;0}pLv{~AYO>vCw#@~paz&-U
z^hNrdOj(&-g)8$q9cU5@C%qR7%n(ojb+Old=3Z&v?UtCp-`}vIF%(G8k;s^gIQy(Y
zH!x@bS~8o^aT#sAy=Ro2)z<_Oalo1W_9bf3n9Pkq9ZNKyzEWm1XeH~kufO|Nuk_>@
z?0{%`9RaAFw5|N?4wEg1vp7SY_2fS!spW`f9ue-64$NrJ%%BeQmz#QDIoMm_Cl=S-
zO8FHT&7{m%Xk2ZqJr+1|TiA#a(dh1{*C~^Tz~|lPrkAR7_=>k#=oZJm%2qct)BMS=
zmA6)pKF_aD%k9HEgJN4Fo+W#G1+=;yVyvJmEl@O=pAoHmgwxSvLt~+cl}GH+{Fw=k
zUV5NXpfMNxn(ZoqmuGcYpIP<=T-f+_x&yDCoF+t@ue*o|Ql6PkBb_Llt1IUoZ$Cvy
zbr;v!!>LnPa%;RsbVe5vh>|a}&_p`{_eJS}RCnO)7VSpY`~h#)!1byE0WPfB^O9YJ
z?uK!DEeR{?8D%he=P_W!#S>~{k$TWrR556?P*g+NBW_j8MFe9UiP;sbyS~T{evFbr
zX1gSy*kGoibP#9rXR>ClP*;M#vYy87_52+$;POM57Q7?i-O4hFRrn*4z+fI-&>#S~
z)7tZ49SPxzAdzSnX5(5p?9+Cgh1#S%FoLpFwl*V8=cCj#Y-)Vpct>}OnR+xY71NMf
z!ct@1SPTB(pvCvV-o7uXz)RCrz4_WNdYXC6J({|tJa71lm0fo3y9O8*K+cms-CBZE
zTGe2$cwojS5v5n_D~hqyQpim6BHfGe&~`&!^N2cIUVZFWmAus?;lz_E9YVzjv{<X%
z%LC_~SNeXQo2<q8Z(M$$hV^A>X~&1jy3N`1<5f^C(Zp@?89@1S0do<ViH$(bGZ2`r
z>TRUnPL|(WUtS8qm1^+Kcz|pICeprC-<K<ptloSRxO&mys;xFU0mWAYP)Tn_Za$Qx
zg4}VAkX2R<bFIxG*v&PJ7_9J3ws6_oYTNM&fo#tN(VH0p!)Ye7wqk2U`5Viy(}(!s
z4ab~8aXYT;z=!x)6l%Q&y|0uIc@BLM1Nc2&T`?&F0cfp@33`oQ?q)ksUC^&G*^6|K
z+eDO4LZJ<Kby$U0Sw`@!t6<`bG!vLSYte#N{X3j@IxwAj(Jy{_Cm=|Ez-@Hqq+ZQy
zX4-l+F<R;DSGF6Q&_^m3-@lwOYdIs_Dsq8>F*in)spJ`t!#tYf<GG2_$lHkLuXtl$
zQ*~$w#6LPiuJHVF*3kW=#OcFd&Bp=*@$&tIw5H)|iG05t;<qhyQqtndLE5s{;r3Dg
zehP8xIKIO?Rhi;34dq*(uXg-wG0#5KSLTJH7B;dIE=Id!FA>cGFrl1R%IM~qQ=i`N
zqH-e`e4?i?rx#LopVxZ12``A4->k689i*yT-J<1xETLVvb1S|re&Eif*RvlO7j7O4
zyzonxntpcOzpt=rHa$Vn!t0WEqm!qsajITKZ_W;Ru;;aPrr2>Ucu19avRj+F(Rk?G
zBve%*bj-l!!SdUJ^7GyXw(TkOSy#&5<2=88f7}6=X!Y~<tn9P)q%@h#ueq165_j<&
zi`K0)!#*jQF3D*%mM@xumtNnw#Da0or9b!h_YLoJ`drir9n^*#w(Yk#&v*DdSc*4%
z$!j-||5)8jw&SvI{?lpQrSC{i)yp()Z6(yP%CEK0kNad%SwveP==kQ}v+eGrQ<?-&
zPW`jL?$*~(Rbd_ycdtlpU9_*FEafM6${MS`DAT#eH_p9lb!92VhI!jseE+y^^tR|3
zfA5llTUkQcXejYdqmTxlPz3)2EAd8wjA{vY&vZf528ZgsE~>K-OIaTQ`X!M}?C*`f
z*MyLqDkOXAxmR~0IZQ59QMMr@>T~-Amw9+q*jj5BQ#3N@EW)O+P#I+!UKl)V_&9={
z@Ui3v7kFvO2!$3#(mOkRv*)Tb(>ChAh7EZ0?Se^Z<)Fhi{uHwFJm=R?w{y<B-)<Mv
z`>J{mEf79_TkVO!`B5KV!+)q=2l_5E=G7<S<86;OA60o~tYq7n&N5A=d8E-QbJB8)
z8*+F%d?eIA7KohhskrbNm8aDWo_B)@<aQR*9_MMdmpMLWx-rVPc#|GmLMb$MZ)xo5
z9DFCp6Fn*0P_})$U!3bc-7Gl1UVG8@YqqT-N2}kf%<t50mhjTXdl9N_4P_NN-B}eC
zx4U=0luw394r|ulFEaeZc#vpSG@-j_w{B=dsW2Pjlyq>A(~VP*iP$<Ez4_PliZyLl
zUf=O<p_m^tv7w3qR+++<q|E3}u|3abcJs*aamiiy&Q4z%BVZ!JadqZgk@Y)X8Orbo
zst$U1DO(h0zeL+8((ATwXoeWl(bq|@p?v&)kwA!x!#A;zQo40s!Eg3&s9tLbNQrkz
zj)ofcWCq;zD4Fpz8%ttjJ5x|()V3=S!$j;rq}tUbzTgd4(3~a(!C()2_b|WIaUuPC
zm`zk$`AFC*XIQ190#OLhk?hvDgmBBEW~D=Ss5)1moL|Cp=sWwr%r@l4^3;XCfEj9B
zgTCPXh>O?$UJ>#UckMb8fdfIO?;y;E^teSa|7XOr6$goQNpnQMhrX&x<Btt<G2^M2
zpLC^B`1j$L&p6#U*Gm3?S8%w;#rLC#;8nbpIOA4QAWbwn2XB-wM>+pK+xkQM718$G
zYPF^DA)UUf(OYRI86TiAiebI;6wBwg#^+u+@~vjyoYEb_PW)L8ic+N?*`&`fcbfZr
zuCy#|tDaLP+YABkC&D85B7eWuK)=^57j)H?j@XfvM@v@6xs)Y1dw!U53-)hL*^5x^
zE%dZme{WHq*7`(x`ptuUfmp0_h2Np-ngEoe=U$NjC7}0a!D=_oJ$ClyRCTTU<eHJJ
zmBgj{ERLi0o%4wFK*a-tWn0dRK`eI526VmcE`61Tik7n6e!N8nWkxoIdlE?8GwOM4
znGr)rcB{|(EE|^3jB1Or!UP%wgMP(}?v2rthxKg;mzS7Y*`hA<Zu7Q54PH{{ApQ-V
zVBBc>9^aD-wqaEF_yl8npo~#BUhB5n71+LVE@T${#?Do_xM^^k7ef0U#Q_}*qoO2;
z^bbSM5p=Q8n5U5O;c(e4Bl)l01*~L5dScOIq12qM+@^OwDQEo5jZr)Pt_8E3SzjbI
z)0t6jJRBE#)!50s`>*Mw!Bvw!SC=7}68ikgkk+t?3Kr*kQU!v`%G5F;Rxp`}H!6>~
z^{#Ba&qa2)OJCpqa2@fIxzdfpE8kUvJ=8Nzb>hbF3jl(ODOJM`)5xF88Q}jZ>G$i%
zHV-h-wz)2LJy7oKvY-)V-(qQwSiA|FNWV7SyGg&P22ZMS9r|9J%QKl6S&?rBww)@n
zqpiCQR`vH(&h%dVS#i@3Z!R3kS4GhWmuQ-It^r<s$9v?TU<u&eXLO)<LOP{2nbi8P
zh`!A>gBjVa-#mMZQ%gaNnhPKsyT%T*;ay~6j>Th3=5gPLP4$md9?M5cj;eLBhi;`U
zrphm*5>M^FT6y*|O!lAWA|85vTa3c`4D69R2Uag^X0gS`EUN!boT6Bk?K~9G&q)i(
z?`DU*l+_kmvAj7BVCznev}i?aSZ=)NSMoLCd&PpE#^KIBcJDd1MHK}&JA45MR^Ric
zAF}mwykg=*b@>rR=*;2rgv-f+ecR~YG!>~9y5yTLmZpiAN)P<u<s9p=l+hMGW&J%a
z;GcJ~$NzXTl#U}m+CMgv4H2;+=0|q+U3DYyRM@j9`3>ZWHHcw8j?%2lDNBoFyHK^i
zt|UZ#Fdb{jW!@g2C`o0P80Glqa{`~0W%&fxxdyO3t!A4m(}EWPVnT}yVmiYhwkz2b
zXST~%;vci)my$dMoOBxtJ#Lq&-0u<;>b_BDkxgBy$DpLxH>9Cz##A4ov`SArh-`Lw
zUgKBZe({8NXruc1^nu6ExC`^L_RQ^Bc4$+0s6{&S)_7@3r_FMreXoLq={iT>;d@IR
z&T<_w325qJ;fU1xJ!dn7i>1{{PMF#5=AL@oXB0<-T3a@x9ozr98sj0qZ2nMyz`+)E
zuDoaZ!Ta5&%N6f@p01xcIs%FH^(-j-2VwEkPA5*Xhew(>CN?|p4L6QjvHF`%ngO)>
zUvC~H&fCavjoWvA;AwULMW^Q}{GNIt`q8CUyDUj@mESBTlY-sm-A3^`*y|_6h8_76
z(YQ3Ck@Ch!{n)|npo@4J4T1BaF;pEHsX=-{ERSzo@>`4iuUP<)-fAiS1-4N&>NbHF
z)@Df`_JUiSU`Ao%WuwT|nGtKxkIIGa&=}9C$ZxK7tU5-$%KW`3NXR3s>N%qKk6c^*
ziM@)nec==GH8-LeY=X;~=Ipxcs2|&1o`}pcq+`bW)ZF{X9s_vj)64fp=Vjfo?_UTX
zHO>U%$z!o4gV)w&AslB~9nEf*2v0T|%7^Ev4fHFN*AH6kd5q`}+|v=LWW%W2!y#WE
zE6PuVa7A&d`M7HyUw%Wfoky>3MbM_vpVw;50lf+5(u5~Z#QUGgf0>crV;6w+-{O5N
z5&8iuTEs;8GEwisJ8P*ZNM=su?Q7$u8L@m1tlxTCtenc9ig$-MyR0*uTljRqi8T3R
z{O}0F+9@C(x|bG6b2Fs1jO1!}Z=apMQ*#})<IpGW_Oi79HsrB{USY+Y^If{hJ0Y{W
z&55J_wg_$bOa+TpU)B=DG7+D4yYU~LW2&7nH|6&^2~+o?2D#WG+3NDMs4v4I|LmpZ
zRtFxZ;my5-H|E@$2NoBYxPAMyZdT&izIaFF<_l4KXN|3gdmZDYqo8MbAI%N1RCqWD
zUa1esDSH;8MgsvX6Us7$&3Fzs6hiWQs<mZQNhVcYM{PLJCf#n7xqg4u@@9FRtxGhg
zm;b7LPBI_JbA0o)lDvlfY|^&F=Cx(WsV`ba1NkPd_&pNuI9Kd7{OL3hOa8_9vk%`D
z_jSROrTymK>pMYL)_<sQzetxOjENSBW!}1A7;)<?XRBe?d=8;)7vbi~`y=%F>0s-R
zU(Y%rJ<QXa_Ccwpe+Giy*4~tT@NIWo=RC=RKW>sNsLbfr7vj?GXl7s&<(8&$SvvGZ
zttDV4j^AIiHn{9y-*dmC4~=mU7F!pM^msN<tC1lo`d(^ElYYtTp>@T}9Cc;r#u6%S
zT)|TCxIOz2KCsfWg)n}oXe%&k4FFv>#^IYdghR`(6|*NMq3X9M?ndZhhw#HjF!VRP
zTy)drA}HQ?lBF43zF_z@+Vh)vTijD|&cb=@@72`T1-Bz(`uy~vT8slI+xA_B+q~^}
z^t&e9izCnJg8#c9sw~#;hY!PLPnx$yA+19WoCL10`KkAv@91swrmuono^JX1hhfp_
zc;=bTLA`$kYjQwRHQ`8lLX9|=bt?`~!`6G6yjNvw<JixE7wqt<v;79=wq<EROKHmP
z-QKIn`L(T>*03FZ)H>v)W7Bz1Xyx4UWc_cB&<zpeyY#Dd39<D>qZ^lK^?_GocUOUL
zoh~7s){COTO|~&b{=;-u>w4QACqNWvMasuBir8ZSgD~hfpYywP5K-xUj$%MTMkza6
z?3&<h$J8?cew_&ih;QaXntWIlHXC#pH_Fm1>4uky(LXm9N;$Oqn)N$MvCnLB!@}!1
zH6G<tDTtxO3`bIk*)9%3S3;`1i9UAOAzNu%^GUt)NoFTuEMG56ekzkHu6<BMw-jbh
zcrz32d4nG!z$uKUk@wBI@yzCv;!mo#Aytwy=NcLtRZsO{_5i%kwK6{2*f4Fa_dAsh
zda*9?=j%}Mf&q8CZ<I)h4xQb4N&iq7we>~%R0Een+M!VP&8OIduf=SLfPqdh9e&2j
z=|K7g=NN(xM*DUVR%3h(Y{a#9zeB0dd1+5rB_HE#`!-C<oH88I9LnP{r7?)akqWkM
zQGUUvUCyuNqN$h&Uzr+<C3h`tnWFhxvs_0oIU_Wj=aU;HTyl<P<$_o}jtoZ~6bP~j
zT;jGP|3qhp?AW#=T_PyMshy=Yk{^w?4#B?JahnC?xSUNPr*4CDD^0p*SVu2dNc`xj
zcwufv*DtQn(ytKPxz@6_>zN@rq`Cif>2>DTKPcIUC0>NQ!4$Zycw+=1lkv9+{u%X@
znC!}BQet@Xk+8M5bMV-Q`dthB-PX^_$TETI7sx&6g~Y66C4tI}fcjSpD_@S8bL%?3
zXJ1)1vp1td-JK;l;<c#1?_X+U0Jb~1C*a4gz9dzZw!D&Wc7D%mbQn695IR239?(*4
zov+<2tcoqMl&L9FdGI5DC#@hh$$aXg37^kuXz}8=j>hv1xbO0>)-f*MB>5(E$%3qM
z&&^yB>7D2LkP>&{rEG!{XOZrZ2HWp@`GUCzmwmE$BIi|Pq<-Ihd&4;^)!7zWK71dP
zWy45z!W((JO@sU2*nADY;tbon0~Lrl@-ovBxGim_a;#%hl>zuu?GBQ&Ut+n*W;=yn
zH;!{_xb0TRm{~MGEW^o9h%;kwDZE`iDp&LPG$mp>fQehxC`)xtZp4q==5IDAxdbxV
zWV?cd(U|_P&2!DMJTx6pIxn>rzs4Y9{}(MONRkiQ^>NN~a}zhK59R&|CWWnxI{fiG
zyn`$nS6`^kfB(q%K!&)mf|c>HT9zRGS$`R~^(En0^GdK-Ov+kzgj%m#a?-NHV`06g
z3WI7?`>x2!w5cAlXZEkKz8RMyw!SKSP;q&VB6yVEx+LrN@;6V!ju+D-??T3MV-8S6
znv4rdeOC+Rpt~N6{m>pZ+!)uHV06g$mbCagk}247Q-9OqW^BOiDc!R2dsO44qmaA^
z%SvV)y3Pajax7BI4}W>u&y#{&ZJ^APwNs??p%d{_Qb!A$`@g#${ekEF|DfqA!<ua0
zz9LErh;*lPcMH-fJz&y}G}0j=Al=<9-6bU;qnR`e7|lSs8{V7ef4raeX~*teabCZ=
z&b#}QAoiHo#-|>7`j`456$=t<rgkO{XNTmL;b!-V{>`HYb9Exst(*=aj)O&aRWF*q
z%}T6`8t{y6x(sa{B<ej;u#mL;B*s_7i3IK#qtN#2+v{RAUrIFxKI-aFzh)X?(xY9Y
zo8CXy#NIxL&8@|ij`!%2650UJ63wn`eT8hpdR_MR#djPKB8(2iF_&uBRGu!{`QcZa
z9D1M}l}*SHF+#n$R3oNnL%z-CLoF<<`-2U6T<nwBagxfQd$OPdMTrT430pSyw!Hg$
zNEc4Gr~I@kuwuoJH-Anu)MVte?%H*H{49F@<*cqaN>d_2Z^-QX5fMhEv}q=yC@bFl
zP(P19TD45@W+R0vx)ZJIVewNm&5=}%=dQrB8qR3eU-x4j|CLp|U|uHF0~+;jbn#j<
zaX%L>rz<LG!|=kquT4K19WINM>+Sy8sQsn>2}hw@gM}u7E`@>5t75x|8Pml3g@JWH
zMjL%w(Fc}ZHyv($YuO@*(9ecod2V@HTl+LRv4{JVJ5Dw5J)ijFKWKgXt+S~v*I}l)
z??rjb*`AU4XIMj232(N7J)q9KOWJgkSY31h1erWuEEHew@(pPY6GM^c%L_*jW&FE~
z&2R;KFYo!~3(<3U0XaaGpfth?JOlQPUyfAbFGpe7;4z`*L`{QP%mU4L)qLx|cvh_Y
z&y?B_FLpIsenLKgSIrcdPa1g%#R$$T9onQC1ElZ>Lailzx%;Rq`*_!(;`$h5^wN{P
zeP87IPN0j}I-Kc59!FsiEb^{s)lGsn9*{Tchq&2ai))}#A8FKj&SJ+u{+SKp@em!m
z+h2~>^FluqGjQ!yWHF5m-i-l<cjf56{U{@#4hSTwc{_jp{e~CJ_`ECH9HGk3=p-|b
zV(D+GbHkPK&xkEclCM*&Yx}EQMyGY$DLFlJdZ#)V>?}jVG{$bF6=oq|cLR4|<JI8R
zuJ!!+NmCE5NCQ#3V-mkfzT~XJM6W>w@G3}w#>&Y@yH+IFL`TSU)A4)tv&kWKt?2$&
zOp`Ex!iIP?RO<`G;4A6<ndxz|&~YtGp|}!v$53K;0|;VCIwLhjUN&0e;pF?eT94k?
zyRIq>>J<&9H)%*cqf;h%m5A%CZ!DAPHv8Lu;z0@xu3$3VJ$7D<b%2+L+CFIn#yJ~<
z60(mR4@3z$H(0)3J_h+gxA&J7ZWta6aGiHQm?Fe9&`r@8Zkc+APD=NY=wBA`D@PjZ
zCz;P@2pwk!bk`c)I7st}9SKgCa`c|7{!WT<DLu#Amv$~B4}Ic@RDxPk@p#A{<50Ze
zO3NRxoAxbh5%?^|lwgbSH0$OYIoNS0I=n}a0Cw<F_g3$4UXN%C#$C!fYZ>iMW5bYz
z6Vm?tQ`Srcl$o1ea7INwu@I?F{;v{#QLQ|+HGO=NsI0;d<ccpE!ePcbA{z7{2l_;a
zi@k$9J#gTs#al0AI>H3*3jy-NFuyQl&DMD8rhz6s@^7o%*Fq!&5MpWx`7d#0Bz&%E
z|M);#>LOW&)BJmC9i%FdOKzE9zdD$|C<xe>&Dre|Cyi1yt#^EX+e{pbEBfn6%u}(A
zJk8uM5fHo6sMvS|NQ0XSDl{(<$)~d}rB5J%meFswO;7xmnt<|7{hwMVr9fvm^k%Q3
zGiubTEA3z&+O=PuHJhxn*ic;-xz5u@pHjjXhMT!<MW?uOX?Qsi$6<9t`0yge?;nFr
z$JO(*1N@3g^-bh<`i`rGLc)Ly3Qn*e&(_jo?4(r}>)8YLQ^iRFWPt8ZsMLT=XV$(l
zRJc(DF#BZ2=P)k<7$B&DcK>90%4iq3QXIAx1spcm-kcn;+M@5LP7~ORd9l8caw^NI
zn+P`=Z6<CHsW{!kOV``?`dssPDW)E;O2Wnb8s^o6NM^0WJV_lAuDG=30Oib{#s3h+
z#Umj|S`BVBvC1R1e6xF+o+ag4kx&B&WxItaItBf4xr`bAMl!q()uo&Y$tP9mCSkP(
zI&_L_$JV>@g&Sv7tF<Pwzb<1<nPPC1gly=DlaifnoGDDvSSnMKbpE7C5;PrCa%1ez
zrP~p}T#5sBTj{>9Sc|Oow3xPWl9lhV^)fsiN$=fb{(4<r_Ox{oTEj9v!5QZk*}^Wp
zS35OK$bye6zR-G?uC2A0`pt;S8+Pp6=>5f%bfAyNp#5lcHS0C0`{+?9aj#qYuuX^!
zR%^O;zM4JkyHC38o-{iC2zLd_4~{726I}5DXsxCa@jj@g@h=|CEUP3-KDvoc?iX6W
zFjw_wrF2Hgid^pO-7G=^SREHfi)iT&dd&KtrQAoOze`ixS<c`en~VNdt-j9Vjn?p~
zlP3eQagX40><hB1@c{}ASVZ+T?z<w30>)zsnBw0u_qkLXb{_K%yeY`>l?G&7HPSU(
zS}_q%o{2aqWZK=;pI$LuUeboFRs;3@LDtCR5I1v1I6SH~kDMj>l*o#l$80`x7L|JY
zU3gOV+F(&8oaI})3B@>|k`2|(kxCOONeU#LO*lnU$^u`lWV>hWv=2_e0d(of3C_w7
zPkJ)W;-?Bafa*o(0mWofv^a&IZ6fnhnR!^%L<k-wG_G<4uh%5nx|s>xm{mY7-~YbU
z!|C(bMc1m*;uAEj?0?r~a_!?!IBw{VddU4BTxxf+A{W&XSiZt=cJllJqEpu#^r=hX
znkCq!(#bE~>Hr4HpMYz?tV-GTl^!w@zh!z90AFbJYkzyGse48@nW<uPN1lSybP5$G
zlrqH29PvW$$*wEHEJ^d?O_!nXQ6=|90Xim=TPPK7`Sk~x5C2w(t@x^%;k5$1^&G=(
z==HP6AJ-cbDaUbEok10sI^u`c8G81TBlB!zIdJ`&Hcrvj_*$4MATuo~*K*C8$5BPl
z0_7!#FYv^{hRIFSZvjdc(Bcq>RQUN-Ct7W@&TcDjsh2MkZGcRlN(C}YGjW6aC1p&@
zdRJFmhfaWH!zt118BW*a@g#_A7Hi$GW4YPM9~fCOhw^=*-3+_X{a`?H^E`Z$-+CZE
zttL3F1agp3W>GR({3H>hUu$j#e)G5`(OJUR;?lDgB%yfwDPUK!qw+sU@DorjMR}dc
z-u81yC>**8y9sBZ_>&<r_Q8ti`3uOE)kzk)(S$hAR3bT!h@Vz3rL3bd+lUZznKNVt
zLxN%hdixg7el+wj1rGSIdSFvno_kL@G}|;URG)Jje)+W9qS2pN5bxjhWxcBzW?x0%
zT0G`#o_aF(rq8O0u)-ZT`SklJ3-t72`M#VLY6QEyr>XXGzYS`?mUtk@i44H;_{Z`5
z70c_kU++EpHd|?jki~U?Z;g=+axT0?cwoT2D^zBN=CTt<0{sOT#*^#i%)VcK9<jA-
z+^qE<n=c|L-$z~-C9My)U*Bnqv<l4%JwY5tc}Fj*R?qsVM2jT7beOFzGTF*l5s$$S
z5{2uQvake2unjAHN)D}7rG)F02hZGS3+X(e2Pd*!-lVpSQ%i(~{in$$=D3QcbyDp3
ztV>Ts>jJI$bGMntxB2OdrzB+vf4=_-c6opXu##6hvzAG#lJT{Ubr{8aQmhmcFn?Nj
zi4f}fJN47wjgABFWsER}Ypn#%!KJq1-X<fKCfnvrgv}+1Fzx}<o47(2wg?#>89|ri
z$}~&bAcFZ11`AScq$XMJ*&r!_D8UI;2jZIZ8ajoOzhiN7@yomRx8ZMjoA`b)w>B-D
z&5qVZjvi;8AAkGRgc+Y&X)~z7EI1kGsd+SY`$*EE`@K!zE&SvR<v|V(MQ>y1OdRl;
z(GOR*Db32OVqv$pljV^S2>YoxI#^|DnCNS5H8caBRHOAROGkJ6#8paOdRB2pQ*|&F
zr<2f<2~v3V2Fzk-*rF3R7oTvykl6-bK6$wGBmNon0rf`_l>%;E#kHZfw)*V({oNm<
zl)^;dS`<OV%NdyN`CL8hB&=L>%yBF|&`0n4`&il0s^Cpg%gPj*m!FkcPC%=>KX#@>
z8~A`WY=>mW@I#;9L{CK(_Abc;EEN9hT6y?wAC-0{@nY*=>L@$~I$~XqPXCI^lExNq
zBRiLZVSrrzx8m7?zY2I-XS#YGI)3-iw<|02rdqxoYS@}+F?BWhS;?S-v`YpUhy4J;
zAH%9M@LucGQT3oh3zwlLPRZXHWig3mOm-R6V6TyN6)UcVsM~zdo7S4{^?IvY!k6+6
zUqoh-aQg6NOqO)rRJh<0RB*!o>Umh=2}H;=(#sezQ1A!eEPVM2p+$M`wYPQZN&JQ~
zAbBsMGjLC->*X7I){>e;H1;Th#4l0{isBtYn|G?+oG~r@6V}-MZCqIcntoRdJNy18
zV<;C?N4|Y1o%nA(QYpMhIYh6c?#5Th!I}E95$JI;U}8ex55Qd-`f62s#-w;S?4p}!
zl3ip()LAXw=vv_y==B}@BaOu}^;@ggXDGbV(`=lN5G920?V?p!`XH(<wJ2}t!<TP#
zw@<r5t1W-7qQzTvhR^8w658%yK1(_WXUs$@Dp{~fy)VYI>*ZIgDw5V{cYwy-;>(Qo
zHv^l-k5Pl9pMjqcOsp+`_!0Cy*uCt@;>*`dB(djR(&A#nlskz8gUW`)8on&Gv&|B{
zh4DVsG~5}Vu^48cXggalemT{P>x=g%ac)R(@{Y}QIDI^4#XZyL1L)z;44u~~^Gv)f
z`wxDVDuvko2$TQLSXSVS#mcG--n#p_2GgpCg>Ap%a71B`vhcIy#G+vmTngZYkuov~
zjw^)~B5UI!i0!MxzJH)_>eeTgqoMYf=LzHAN>7o+2h6TvuRr7N-jhe@{e5wegL};0
z+-7K0>)9LG$60~(ce@;qa&~_a5|QGMCgIC|ltEo5;=ek~GXK2;XH5HU{7Y<QY2i+<
z6)WGd2%mQEnJ<cb+kr>_GDz*Q&Brk4nW>oB&IyX%uRF=;96y1syP3m+NmZiaH|0K(
z5BdzNJ4u-lW`jQ)8Z>1%)a9SUyqR~yXq?wGM``*hxV0ciqk|+0fw&*_W16Dp>>v{7
zpR^u#5Eqq$2jIf1$^e<;HVK)>4FB&ZDu=6mOW{#oqZM^@wMf{o=}1g1?yb_Kzv-)u
z?iTuh<vNaa0?J1m-8snQ=cdRsm<<9tTos#A>)N7XM8eWWli$!2I1*sTZm-dFaRzf$
zvJHEUD3-UrV`>2GAx-Ph^K6D>Zih<+mx=_=QdZq}>1J7;MgzMYR(5%%l?(xU5eIgD
z$MYQ5tjl^&Gt_<~<?@z{dFHy~Yu=V82enThRSBy`=AIHT>-x3EMfj86N&2rBBP0Pg
zJbI4n8!xVBz;035C19ulc+T)<<?)B@ZYy^wAiD{TtI0c(ti201j%C872ROk-ZDCFs
z&X{lV-ZppJ;~MGFY1Tp~1bo9-KzV<P*>#(eY6ER|+tb2I>=TUX)--_r<g{K-olO>v
zcawz*3|4kMejyLoSw7ul-)fb(cwvo*Luw)n8EsuHP1@qqkYa5uLiZ1_*yu*-x~5Oa
zxjuA6^SFF`#tZ?xr=~}6Xo|OyJcqVuXX27z^m(@WzNd*#;L*4GB`5Yy9ot>NS_Xvc
zc=<X5A|N&J<A;|$@E4A;PPB4AjjZ=30zz=mQ7q#V>-jwc1|}wLHhS{j&1n3CfOdmv
z_vr$D!Vq^krkn^`mh}GaITq&#t?EAhmkKCv9U#{T8mo)IYBvB@J)%K*IwPTS9@^!k
zwhLU?=aBf>OahEMGWMsIU39D@8++oEWSt5JG>>Y{{8C|Z{E#K7$#BXjX2Y*qXwmMt
zV5Zh~o^@*mSo7wH5A+6{ApLDqN1CIf85bVhG_qUVL!BJ&0Q=s60su&Li=ARyBfw75
zVI4H<ZfS^Q>D8d{*7N7z>beT@!1So*(S7(4b@`|y?!A`L!dpsY`KCih)<D>WzJ_8#
z{wOW$o4nMmtDiOXN-4%8hg`s_P$C=CBsB*X9HM>6=39KryAO(2l*}XF@8CX>=c|;7
zJQmR3PKLxu$8Qv$4tU2nXKN|m1uqxe{q1VK^=>aX9S+Fo7I?(%oq2FoLF?8RFww=B
zvBGDYKf8Ht=ojn`&afeX)p7y}iBu4J=IKa70e7^*CZpS=okwK`H|DAXutJ?k0UUiH
z<=URISXWTllwGa~<WGEEe7d)2R!69c5~Nw1X~r<QCo-zO#CW_tk)KckW`@7xuE>}^
z1~fzq<Nzk*O&|A`3;b7`k5Bi52pcbtL<11bxH*2eAAB!?nUFj(4i>3i`N$MiP<H^?
zSX==7tin32i+V;!-vQMNjeJ&9t{hi?t;uNmT3n)q?{rzxPVFf!?w*&cwnffQJUltX
z%YqA@z$haEpb%5lGu(XLL}LPoy7ll1`*im{sP=ENl*oWYN%LLE#uv$J6bT=wcGG!i
zKsw^|F<!oB+<hcBWn3L=l{~}s0|6i6OsS1tPJ~M-h(Uff;IJMHK5c%u(tPG-NeG&M
zKt1TYbhbacKZ@>pFA*aoQAT8rqX0NtUtvP*k<1~LKlC<*;1S!4XH=#>4uB~rS@l!g
z6MOX_+wBZV4np~%8ScI^t1E&==&tmt*7UERTiMC7d}b&B^Gh++SS|keJUYhJ?F7h|
zf$V!#^8mi`ZnKxPjNs`kaDaGi!Khb~zXc5mj>RZ85xm-^)<LJokJA8bse*}**8M!?
zm44KRoH0aOZ;Fn!AT+^L5y;0X%SNx&xbMbt%<jwr(Av&>DltxDACDaiPO}dKK#pSn
zv54>|vRkgK&^=jh59TX*i(#46ePx;Yr=|#GL!Y_*?oV^+?~W?YXf=C$aN3OZ?#{b#
zEi_xZR$i`1r4J@#YU~!bQig{aU4K4tb5p029gRl}V577>*wmbhb%`(CAV=hH4btwD
zPR}a-)}Sy0zpdy)Z`ulBNl4#hc)$#T@~?EfUghn7?7gf>+C1x8>0sRqNAMSekRxD#
z{@K#7hxe*&Ur0<@!F)o?*&BM%^O2-KO%fn*BX=j1@Y&;p^O(F@EG*AzT{Mn{K3d0x
zP!3PeNY|;E^}dIUP2C~D*B|Zu&u|y~ak`Qz*Rgi`=hhPEXsZW@rCX7p8=Szr4T&#e
zwQ#KhQZm2-)EsWsp?;z&G@_&DlRqm;m0h(n`op_d6p|l9vAOTFax8DgjsEN#{+1Ei
z9B<|h7#4`lx)Gh>#iW>Ejr<Q;5dmC+j9HIPAqjAi{Tu?OwXNlf|8$nz!~oGLAZ5Xl
zX09X9p_7Eu0ARx<$Nm=_`^n(+slqbi=r690Qx=D<!j-L}r318sY2Cy2%((ROG?Jx<
zV6;ypa`_wf3xdcVTPk02x_&3fW}o4AG0fhM$e<S&ZAxjMtU2@@4OVXs^asSlw>7bC
z2L+@ii;V8uBC~Ip34}NA=A8V&bwA0McJEU=7{ZVV<y4^4c04M8jN3#D#1FDTL-L8{
zmIq=huU6QrFFmDyxEt{(@;<61Q7Ent<_86EY&FuW$KRQ#gaJ7rR0_1l^U-Lb_1OG*
z$yucWgPTiAx_CfSor0`(bUBlY%qA;rI#ImR`+n*3!GN~=T&WO<p9QP;UG#A7&lz~1
z7#b}A9uttIQJ8LcgUfq~8zh=f&hoF`Xt=MtZJuewFycq?Yi;*0V=WDt6y{`yA^a9k
zlF%v+z$QKd>(IY=PTyM5)rrz@S4<>$)ttlR>c?etRH1R~(j#T>(i0g(6)oR4@WCAJ
z=qy6?^JPR><{zD})Z{STl&+l=7Uz^2cxOx2+dF9J19HJlPTvjEN0u1CgNV38)gim9
zKX`|}oA*;p)wAJL@Xc{t^&NxnakU@m^;I8A?FOm2n8lfo#4u#OS`HOT6$RRrkG!0M
zBov`jOVYFmmort+YcpvSliBwRcq7Y|$Cy?5-(|;?hgvz@Ho}Ap0kaML&7mEgMtqX{
zh=a|1wq5eV@k8vak1UYf-?n<w%laR8s+Jq`0kqs11YbQ&lQaTH#Z!31Ox10@%`ix5
z9K5^{XRfd3?K7K!u@AET0O)hAW@M%D4K$6_sd+haBeuO2oqQ{NW2e9ogRzcoFE(76
zDw@2KfsPWC{r_2jXV?D<P0Rm+)2;J63{M~eeFUTI%-xig9D>c6$HmhG4Bkv~uyHo*
z)&I;Eb=>%jm5=HgPVE_-)#L!mnO2n+@Vd^g;!Qr+<H6grX2?5l#VYWY9&IqwpBD+K
z)<FFgX09Vr+}Po)4hzRG-%qu%02a62?!n=t0ck&AYL?<VH)JikgAAOk>guKL=2G&F
z?y`Km?`P7gW!v7hceGl=M4*-?M1;x11@Th7?fjx$25~>N*&hc3MhQEQ)-(N9trQ4I
zu|MCBCrMEGuR#IhFy=LP;WLyyf^<T6#Y{r7eVNSRlVWIAqB6_qwX^1P*M@}H_RXb4
z%6pGXH&Id#kkK`6Nb~g#qyoNMSz@h&+YZyO-H;bq8)}%hscHKwKfj16a>&$Bf{*d5
z6wuck7Au}C%xX#sPm_|36Rmo&*)vxe7IjbTdRR>d;NhHH)esZ;G608(NYUy<3r&2y
z4o&UYx+o(PQ==JRTEfuowN~4^)KhU3Kv(ZZ{EE)?MjO*0*OnK3QDD$}2K7DZc<fJU
zTukJZ@)N^&VeyGD4UySB(}Zu%HWyU|uMJL9>w8%-v7c-|3cc00%z|z@`dp3YjL$PL
zdbiql#x1g5V6*I}HaheS!Yh+1s|jJ=fGcH>aJo0u?lPTQEHzZL7oD|C$0!Kr#g*1L
zco|vKE?|Ep&Eb{mxc;U3^x)9qGUs8wIB${sM}yLF^da4X`sXpe{x*TvY%G`yuDyW4
z>R@lfG&t@sRIB=@!VO&c_MwZRD@)1fZb|e_6!F7q%;as><Eg8m&oI7cI`gSD$3}IO
znZD3_Dr7nhm%0qVB^+0y1&|i7yuafB@sJrpU$&F`M`Vw4GY>KNS=RFC9Q|Cd<<n^Y
zCPpzx2<IIZgbfN#5KL)b3w3p^JjiI&OehL2H7hHA@Q4L;FMBd6`B!PVS&WMNpj`-S
zGsQkR-c!N2H^`X=XH5fITbjwQYIz_`v;}kaDpSW5wn0SiPY8#txpDOQK$8?yG|;yi
zE%Mkf_BqEA7`yYEz?oW%xk-#6Uf;h^$naFJSi*Z%ZHVzhCWA71mc$xQMVs|%>N%}<
z1(zJ3JW5S{NU#N1B)R?v9j;GIf*2faEXFSLYl*fh1Dc2yFje2=)KEQxKJ0i3<R!=_
zTLOjetR@LpxYo2Wp(atL8T_{+n~~|GhH*!z$Y+qi*s|w7OTc}?FfqUr2k=i}CYc2#
zTBkgp`^k3xn6TU-82#_a-MTI)x$oz@zSs>c;QgYB&wMa?;2I+n>U}A)b<(i*ucAj9
zO~z&}Re;)eUlh{aniR<pLhQx5V^uSyAlPL9pzylMexY0VV_MI?hVS;Wuc}>2Py(nw
z%8AeztoN?bVlEuCU7VBUjzx_O69!bYBK&&3$qED?pzV_yokF)e4@niQbRB#u#(9il
z?k61ABforR=jGVb0u=b?yP|S8Y!n8zt`3dC<sVvZ_4RilBNq<8vO?eLeOqKXoEhV)
zB)9AM-ss>`pe^?V32#;cqgpNPOa!MFC=U#0qv`LZYsCbBOdkO767e?T+Bd!N(wX7X
zXfWqLExaJ?Fpban={kj+SeA_iyq##iw65#0$3+cQ1pYoT^DbyAnJyls%F(sBRg}${
zBE%~aMR7Mx7mMR1!!NW1SZw%@I(KDeB?fi#+7+1AWEj3$^`h|Hf7xe4pzri+tV%<Y
zaV6SC_ysGI;E~k9`A>2sedo#gJ=LHI%SMYvAb8pUKAiNa$oRPTlD`($T(PsN_gL59
zVa!p6T&Vw!(Epj=F}A}^IF;9AfGh&)OF3R!B6J{@(>3$zxj}GA1^Mn|@oe3BHUR9{
z5ScslDwM6e%M!Lt?~@^}Ui!5&N}y}VW6)JChuH(mBZm9siB;pl&0w{3_&ej5M{rH+
zh}vk@7)BMIXMMjcsTA6ERp2UkKnT2$_cKFylxLKk(W9&A!>xq!Lv4=aTnb9*`~zC?
z;T5Ux-qoyAAup>1mj)<PV1%zB;YY%KYQL_5sB^^!x|gT7nF|D>ps;u#wUT%A89&6v
zrnKr(E*hEP7R@dzb$w~}5G(wN4wuQ;r|GJwEnZ<IwHl&|*J{^j>PLH+0)j3Y2ZHN<
zwX!W82rN;>q-rNmcq6}k*VyOTzhn5zaZ5IjJZGNCh9R6G#NVi#{~02q3+osC-!ZN#
zOpDDQg>-R9hZ%=@BuXstj4$^=>ednGSoUI#u0YmDU5~E|-tq$R^@+@N+c6=xcjqy@
zbZil??G-hRWilk09(n6w{I#v)ue4OnRJirB2}|17S^MfWIT@FOn1s7gBgg^xk3+{Z
zW*YGe%=`}Dbv{K@VDRY_o$XDC4EkmoT93D_xa`Qg=Ou?Gp5O1)$$xA9Oc<hI0kV-X
zH@X=$cYLho+|BtDRZTO26=8HQrv<$q`ZOydbkR3Ac}PV${h2M@o$9^jI7^0W-}Tla
zQv*W;@;qFK?&3kz7P~Fjq?o+C*8=j`Ct-Y&6p&%PE@Up&?;c~43=PCdt&B(T12+kx
za(CJgl_jOC_U-BvaI#mWxS}1fHF=W@HMJPJ;r**<?5wA9(yMg%nl{4mG;Pms^1;s4
zp@n;zkqIDQ7485ZUPzO)<Wco-ywAjWFQ8pG1LSU9=s#+3mF~gM<tCeK{LOx*zu&)a
zzA7GG1o1x_68j8uF|U_4(5m3&^H1sEpam{v1pbA$<@`u8))xq}DEL8o*qkJ<xUM(F
z|36)qmI3&5LY2_~RH^(d#ECS()RwgP{oE?8nx_Pj8QRcF4EsBNV3Rhwp*OnyYcg^y
zqt%?bA!c_NhvXg+A-3O*2d=VzeUEc!Z&op)@?zjSmGA{7na``Q@SMl~y}N-Nh=X{*
z0^ikx^H{*S1kHG~2@n;1vk*)jL;vLM5>=dzG!_)mVeAXRqKM$UABP1a=pb$H$T|&|
zcGn4R{t)hqz+37*v)MKgrqfoAMfrQdeTo~`C!}csF}`^X^b_6B9Mza9287;g35+M4
zlZNRd(uln!#vVcPd77mbNwQ1du|rRs`((_L7xYHi)xY8Q{EviIIkFmy&srq5yI$UW
z{esY+HF%Ky*C0Eo760OoQXlnFr;;}_#3M}UcS_#u0*UPXPIMzg_ax?vrBuK`ri=U2
zvIX0Lu<}1Iawzk?wNb7k+(N$)b_lL?cv{l5jEiY6yQv!!`)j7O@b{9F_j^pUsFS}r
z<O9xZCVBB~(<w@fI5lz*t0|{7S5D(xmM%EWCSPRa6hsv9tilxtnJr*^5J<WWIf_#3
zotA-Ibv!1|RNp_7lel*UQ&PHQ`Cr%#z)vr?r^(sXc>907(WCQnaP5XD^OR!N59QNH
zazxjR9gisfQ55NP&Q1c@WFh%*iX(EC9Qj8J#||sk;(2CFIT%{+%y372FwU8V42eRO
zs$H@>Msm;Js{bg^XoIN?iw+<&0GA8gX9W@S`qu6kA4K8`<LJ7Wz)4b0fJj!C9i^ut
zCm(_$N>@=)!W&)0`c92*m`y3~6OWb(>IrB<0ngPwZ8}-X0BE5pG^0S26cVwHKCfck
zkA8R`)m=Lz(3Vr^**YN*k7On$)l=aD!Vc1MYA0QuV*-&tHP{aX1=*h{t|7~l-}vD(
zQw+cvmL|BX*=7U+9R;*8xq~tBUxmUrV?g`w+6CFURR1c`R9p(CmyXBbMWXj^0{we{
zdjTzir8S*GD8IrAv*fEHS!NI+OFi6Gc&Ap6SPvr}PPUi8dpbU8`Gnm}EewYA>$#$p
z+sovUQ~qjS3C?<%Qp{IBVFKov7OsnTXN|36GIjRx^RB$CfMO0;{+L$%)&yAeX^DUZ
z=`wtqp`EN?;WRcmJQauy7-ie<zNy>HeX<r_^^*BdF??6}6(CSQ>wsZZbxR<{wnRVx
zAFRkY?=j#4E;>`kMtyi&X@YH)bF9jTeGfU9b(>WA@WSD3EV;pVDv{(`d={!qpz|h8
z;-B0fsDlJ?_HhHC=*L<zvp#3j8jGM0i&*@xOl!I!ys3F@JN`@e@bVp<X>zFVMNRQt
zt>bk0>de$qb@{*Z8wIs8k3%8EfnRY4&@lJhuj*6F=%Ha20TRD3nom7`y%sskQHaQ5
z_c#89M|kE^n`6Ku^d@bshqK@NYEe?b28oQ#b~%+adFFJ&a;(n1P_gE-Xwf-<&!}d3
zt48PimSr!SUU*G=R-9t;kQa>7#L3g0DC53_1Fb<A)Z^cNLR;E)@NFJf!bjXNrGzJ{
z9k2>ju#(fGH6S@!7&I%>ywp%sb;A!%R#xh_x=^IE|L9;NnlKb6A9$QCStaE2z`AxP
z)T9)F7C5<thj`J|)s~-QQ>9Gm@9Hx)trzu7wv}$2i57t)6|X?FftE1j<eG;n`BAU`
zXM$5%N?o>H<Yk2iF1v~y`ug^}OWscgEv?o>o8jukl&L#Fl8<E_L@dWq%q#uhV^0At
zB%rL?!Jkg^WCnA6&hiEJuP{KyaRC%A?k`<9=?w7@Z(E-`J_*gr4^0ixG=4R)`ub{R
zbXgmo$IUs7RZA--hNKmt@7n~$74gFo+kd6Gfb<h<zQ+PG^C>KzY4Elg9p0A4Gr$WZ
z)|=}x$qO?N82V+E{`UY4VMva2xM8BbnFsV09=u`kzW!Y_NtG<K$f?={UmcjeOj1{T
zjRQ7?810gytSnet@#N{}S#hhNq>v<fHM$}+yi()Jzk3|7*C&9}lyov(L6f_hwJqWm
zb#BRmFQO{`ECLS|l|HosQzhSPy+-QU;le+s{0tZWSc;^|!Ga2F=*7rKx@`ZlkLb4R
z^Vzyzjq5Y&rhSe!<rO85`rF)|MhQ2}ck`ph(TQxRmeNIXh@M57K+Vqu_}5B@o>s}5
z-;ozM$Qn(L<o%4Zm@a)u&b^$o^3RK2+e-(1#t_R2_1bS0R3tLLjQHJn?(6Q@=f5K;
zChuv%<LE((7KnEQF&XsSOq~vtEZ(-3jC+L?=x<Fptme)~0SK`lHC!BGo%~bDcZH*$
zhHFEQ$}sQglXweOK|LL>J0tS>92dee+)y_a)qDN^t>_XJNC>JEU(;pV!X-d-2#l0_
za!3@?iFB+@w|5mFDFa=;=w#4|w$eGsbm%6G$yrHzC2hvh+wDjT7^3n@$fMs2j8_!J
z8|qG|;OQ%looT|mwEwnQk#2g+^?Iuo1J`bVOr>}O+blyHlcFL&M%1NKJ#Pg`=s5-3
z0@|#;oLMmKu*Qk>^NXE!od1SS`s>Wg+co{&i7bkw{G=$eIUqggWBnYF&+pbySU|b5
zG99nElxVSjy_BJZ%XZWm6%?Nz70S(iG~qo~YUr-J{N6%L1f$pLQn%e1^xUa9Cw{<P
z)S2bio}HSz*L`k&U3Mi6_4qf4x42tbC~lxhm9zMY7@a<+JbuXyQ_voDFzSDgA!?ET
zn>0+C2Fh4pz!q3^?7OYiEotL<T2ox`(qw$=&^@uo6Ew_sugqOeJ^Ssg*ARRH*!97-
zGWv)Y9m*PQzw>Q|6<($FyqFw{wHuU;JK7QJZTWn<jkNEsz($t<Q`dtIcIGxk6VC92
zw7!drkBaV8TkPobW-y7hd8gg;?<XU@`Zq8OyNaUtmwNMnX@>|_m1l2o<Xv8)6hSch
z1iI#|rpCW-UBB@pMWFrQm!dt`iUg*7pe@pYX6Kzf#2@uJ^~z}|K9JuSIn1#nAtVwA
z(fl;(jp#Pr8T238y`?WYZ+D)HWs)qTkHX-cadD4bf6;vE{2WrWS*b91FmaN$gpa~9
ztfT8{qro$RcM5ZMNX;vH|3fDio;{s;Py!|^mZxT%Xv*dP`|Inydb13!()52Nf(+-W
zvJv;*{H|!y;>VGqnhlpMURyaSj|Hw1OeZ&jUCm79K9T4^c>XoKWMFijGA5tPPBIV6
zZ3<+4W3JCV!aIw2?fOEzVT_P4*EKOoml#T=$y$DK-=xQAnY3BsqBT}fM5Vq+;GiqJ
z@(tlM{gFO?TyauGEbQyvn_yafbpD82*L+8Ix+quQ)!fxP#`TfhFLPHosY%h$gUXf~
zo!sxa@_55rS)RpRzJio<26|XZ!xMt-*uZZO-~k|SDHN~of5k>rT|Z!<o8>EwdgfjU
zIA?};tbRseybeDA(YT)L((YE)KVOxf(1Id<Fl`gXP@_$kxl^R|B#WS>;idi+-@0gD
zUE?&H7ldbOfiv`l^=sx;1{%!EGmQOKYJl$Nuc(m;Wr>mGV<6Mi$g&<SK6z{&!(Ui6
zBAzDlR0hwJHqPqDeUgQb4o^9Gkqt^`S~8|a-;U(apaZM>m2G}gHtZE@^in})TS5nN
zCGso+DayD!qvj`hXi-U9O(E{oxKE)xJvxZ*xdQ2U4R5yStf!;1Tv`*x?wbare+KL6
z5RkBfRN3#Vucwvo_oROR`l|CSD2ZMhKSO3{{-Kw&it_tuWjPUbOHHzRx(?Yi68PEr
zN)Mx>Bdb0FTCcn=>PVrp$ND<7=z~(M)JFp~hB&h1<jY7G{s_;VsQ>;$#A%@2X~>dH
zZ-u)fMti9g4u}P};<WOk4_>ZOVLVUAcfuKrCT_;z>DrhM?(BQ*w%q%FvMAL0rytrr
zzSdXtMFn)^J@(f??Euz<h-<HY)xg{Cv6KDUy5w(_G$tMvz|8^mD&npERGA-5o|Q>%
zv`?GYt23;A1XzNrkoMJyLQAW8%LIIVn(w*1*-93@6P_f5Eu16S7kRSTu`8?xkdosF
zKWPtgzZ*LoVZTU`reI2xA~TInwgV(ox^Fseqncr*qDjZO2D5ApZ&RdM@G9RgsL%~V
z>aEE%&=bGDH`WmnE`6^!nD2~CJ-^4840($jsrxD>_c$9TH{*AH?IP&+_0Ume5CsF^
zYl&ZSR+87lTuoBr7{ZT-YU&<-%FFy6k8cFiy7e@Cj#&8x-6ms2ZsC1W?`u2nGvVBi
z-~U4F!C6ztRN4SjB?L-R90HeeCCa(4(Z)_5D9q|rN7M~3t7q!<tYu|-<~wtR^ob6N
z!=s|N*7QT`SShcrfErLvULd-fT#~m`d3N#aJ$g6e><G&+9H`CY5T6jQaO^-WmxlLn
zxC$zfMYzIdDj&d1^vCBuFvM}mGT{Qhd5Hd04n<$ElJqy~jgAJdDoeE58Gs}-BP*<Y
z;JJmlDT*^9myTazM17DKK0XXBn_mwHB-D^BKeP_qx^-^eZr{gIey0mqxsT9#@eOFN
z*U$nC@ENzH9E(mAE%HEPJ}<$p!D_QS@~Lo5ww=c7DC`Zz2`I0C<?Aq8@{$Ff4{`D&
zh&0_0L!i3F^V%@j=~{(McvqYvk8cg~&uHq)Yh@fRjHmBK{J>#S+0{BMs`JaS!d-tc
z6NDSf9EM{YTM=nxLAJy&8`nhxBsvc!{<i#XHE~ZNlt(4nLT6gtTKua<7gV`cN^S?7
z-_kTjtOA6u1-4$ZkhTa89fb357lluFJB++;C204kzx|_*0tu=ObdaO^tne@&AnUz3
zt6u)X*CXMkr~fDXHp^VtL`RQ6g;Z#Ie7~m@xI<k$$THmSd=^8X#lvP(kylIcH7Lm<
zmL)_l6HD&tfL@Yp&zGTrK5J9zeVC}z-Zn$&DNWA&SF*<r+qY6*@x$W8pUk=9;dQ)&
z`E{XBXHIiGd>Ps88rg>4^U_d~-Qw7h%Lr&IEB5L*H`mQ8;T1%ig7q(wlO1S|P&KB1
z`CKMvPl_opXt&qF^D^YOqO8Cq%m~y5p>~;k;FgO?@y{&88I=`S1S#8V-5GB7={*vE
z0~cE3YdZ9%U+J2?--kRQMHc(imGk{?q95AuMQFZLY3~ayh9!P;n##Se0MfZYdUITX
zqGh~3hZ$0S;a+58M*4Id;CdMzQP(s#kR?ZB$n}|DP8@WqlU3<^>|Zl1dcCtAhSjk=
zCRgS>N=~_49`rv>sjd5;_j4^jbkPV%`0nv})R(6Tq-V((jKCaG<fy1T=M0E#2|d*$
zI?8{tJ<U@bB8~A#mPRejV4+zFA(SOPmpwNgkAsA<&c*&`qR}ez8u~C$yLOP3THZCf
zNBQ7^`8eF=GYch*GHVhrdHjcXqZx$zau&hom7Y-Ebl)nEuaAzU7{O4(83>ciL)P$S
zHMRtB6JIa5etx{gt5-*0b@Lo+S4JvM(L2n%vLH<-Sz-UWaai+LSblA`#WZ?%uU4HK
z7CsCV3P$@J8#(d4Q-@5)zH}}Dy&Z@q){Xw$7o+QKwk6Zcj<->;fwfC8!R(X%5f2tt
z*<07UPliS`Woc<FytgR$4}QcUA|_q|J7l(5(J;m;mW5r7EMlu$#uhtY9xubEYp^xv
zzJ4X9t1JBzViD>qu&Q?dLYCq2nd`bBm^c^Tj)z`o(lL0mH_1ZB2RTq?f1d4iQ1r$c
z<9gq@XmrU_u{^1wMsJTB(1ZkkJ%7%I95YddOV7n_{9&bAynA30vfm|hefaobw$${n
zA<+xICyzg+gxvgY<nTp*n2wqBI!zg{A4LgFndB4q+LG8jpY6oftDt@f!Y)FSXHu|<
zkP~M5?8d9{y8)9IUeB4Wz0RZwZ8d?QsC^NdRU=-ah1}q@aN7$qVl-LtErb%dYFV2G
zN-O@-pBTp_6(s=(&c2??`s&@vVEx-yrG+^xRjQ;Rt>W%=+aHX<LaOG0`!<80^|+0+
z<;PW#W<}B`@4st_(VE&=E3H%9-G)92|F=Y}g7xgvW~-u9<5$Rt4iz?ft6LRO+b`$(
z=e#CcejntUzy%{;?69cb-(ib=T<i!w3aRfplBl_kGqTWOaZds)sSq!bPGa+LWyF~#
zfSngXM5-5jho7KcCLO&y{)!}hOHZJ`J>mzzO->TkHF8&KUv<y;G}wMhjI&C)(Uh1)
z_lp^PaJ<+2_i&qbw<WQ927RRJ$!J5*5*bZ(Y(GnAzue+Do?Jc>_!(0D7o@Z<R-hGU
z4VtQ?ZOF0ZrWYE=XSG!-eksH3WnZ3~(xi_0kBC|uz1v5YcuBfOq<Y?fgk<-K0>}Qs
zwF>#ULbQ2#^-f9pE=CaA$F{K#(DBeFK?@5}3i-m^97R=k&14m0A05}?oTbgeT5%^c
zxba72yWV_fUX>sx3Q8mZJU=LMmuni%Z;LvQEvF=am0Z}ERC47h@1C<&TE&)c!~Z+n
zd6f2%QXur9Ed4#8iWf!tv^@`t@&EXXcO+Z>vcH!Y*~D;a;Ra_<k!dB%EEz5|!><B_
zYFmy${iP5`=ly2??4zN85zg7YY7*bC>T<N!S?vryXmFHYyWtT9k<~)#q9-l?D<Bk4
zl<y?*&3lR2!ZH&ZvZaly%72ZgSAQ+5#d#e125MXGo3pnsPIn!TZX6#CI**O+yriv9
z*92uLDRcy`q5kX8Q-RKhNAopqpopb1e6~O%ltb&#8i(>KO2x@*#GOJ~@|yyqX!34}
z^4D|(HWv)RKo1iW_bKzG@sIkF_EfWhCLEieC1_{kNl!z(_L)Y5UhV_7A3qOSa>#@r
z_^UK<4|Dgo*!;9S((Kmju+I^T6_-=*Vy(mQkSJI7s6Gfvewy9a#Dm@aE1!Htzv5@m
zSL^lF2u-%tG|}!tsrQ<$Ba6kCp?-+wyKyRcPUHP_xO1S)I;VK9#X`eFU*d(#(-EOm
z;v6`8sn~?z8|$S?IO7RB545qF=OB#?dfSmG>3)~`MaRdx)YJNIqN1r?2nbTy13f`b
zd+~N(6a+mUN%xER-RsWJ-w`Ax(rIXrKuxsc6eUsjO^@~EfOaGRdqNDeA2g}k+2<`(
z#?Lu06K|ra%RRBk0p0kg3JZOf5o}p6Kc`1B(b>*{JrewTwAG|Oqekyf%sx+ycU=oo
zT4DQ%fy6^o%PkIn#;)(~XwDX8Wq`o?&>K`Y&B0rK7uCyz+l$8%eNIQ#Q^<r0#bHY^
zOE0J9Sv@kTaVnFz;O*;9hZB~}pVe=)hCHtf#eFVB^hzU$C<Q@${RE*Z%)^9OEKMbR
zrO9WS!MNXGUJfIqKu%zU3aDmS1rv3=SnEjo%;ACN8o~rvW7r~irTMJO{-3E=04LpN
z#_6LlmeS>vIAsC2c6ypMY7@5TRF|;`zajprV3Rf#?HZKGbI%Y{?<uh6yK>WR;$Z_K
zj&Y`-r<Y-#-?komUDx8r4^)%R9w0Wd2;#M-78LvI!3Mg}lvzj05n;UE7>jolp6ycO
z{6q}(v=Ed~ocDeosE1l3X4k^yILFGXfeL7f1W|~f$$JK2Lv>X_R`MBxreA2}Y=S_`
zV#i?v5ycvHIv}Cae_(v0@v{Ok(Xv+K#!3=#oLiLp_uDzmW)El4<FhxS|CKZj0d=Nu
z>FnpSty?2heOJGy^?ErI@WP}~C2d};Pc-QjbR>?_G~J8u(jhvNPsRn>q6`88&!G={
zX3Gz|7Xfhrt`AVt3bR<INJm#&HA-4CW@=|~<E3>O*+?bVjpFb1y@{^gL%^G&jaK@0
z9G6{tmsbi>9>u%b&p+Mu3_r37q=cv)RtnCm2#Q!|G)CDmh|d=DWyrY9(v`gD9>NzO
z2+4>DF)X|K9rkUa35$sBCB@aoC_S2e>0;0vBd<?6zuBP;TKgK~8qG7H-dt43q6Q*I
zp1Se$1)B`{$X;t^?D187(HA)!Y$%kU(j0!_E`Ai;8@_uQay&Evowi3@oTr(zbF-3O
zP8A3|ZI1G=q+?(+4gnHrwJ-1{um8OxAo|3t`pu!<Fp+Mw#roBV|KPl#Vse9seKJIo
zKO6Xq5xsYT%gYfEz?arQ1%EBC5AO&ra*sY8d{vUOl6tu*j8c@BVGSqIZ4W44x#P5o
zwV`S^Yl8y-#II}PrlR|KfpQ#U;CZ0eIuT`Z;>^iES;RKze?`kGOgH0+g4|JkK-;GK
zrZ@fT%?m(NX*B`J8=wLaV5Dcin85Z$p*eh8c1`O@swK<geH}&-;O87hgI}mWF^pd?
zTq9OZWL{?9r1OnN-I|FpYh613ApGG7py(FkEln$GXh0N!Z5BXQLodg$HOe*<53f5Y
zja|R&%&p^4ABDOHr1R*FB8;wbfB+7Yib2^)kN%j+4w0_?-hgv{G0r#5j_*nd=^fg@
z87Af2?@W=L3<0_?8v9Y<p!S=STIo-K)J{`rE<|MV={TwpkuIa%iuJtFU#XJ5nN$hQ
zt7I!)IPAkB*Lx>4_@k*{M3Mf*fI2wGic3N`XGEMF<yhYt<KH!|0FMhf-uLP~*Xtef
zhSz839bB0%FyaEjnB63Q-Fioc%pLrR@m4>JfYkmZ4+2XZ`RJl|my;4Zs*r@F(cq+N
zQ!u>@PR=}NT%SH43o1!tS!b-2!)OsNzuxngv0xkPbtI%*<@5hn8{_5mV1~&J^k7;M
z>*{LWkk%z9rlIgVKBDHVURX!VBiB6BJRh)IPEzQL^7@dEY<1wwe5ujbPzTN-cB%BV
zKfQVgdOFFaJ2RyN1#rD&aq>rrmiCI36_qxlU!^N@4t-&@@Q#knqVABmC>VJ>#%|=;
zs|~h8LLKSa><We7ynT$C5L8A5N(3@?gnlgZAmgkx^}yS}FVBO4D%6n`)ND<b!T>O4
z6Lm_oc8$NIcy_I(>OLg9UN&OuYh91al#B=PL7!@WGOS0zzS~E}SxbEVhHqQ>&H(Wm
zga>^>XrSJd7V`p74#WZX&k;Ek0uft6xEkf&(pnAHR1n^{M|hyFT4U@?;(O+a<-A>T
zXZd;loDwN4QB5TmObkU}D>|^lP@`rVclBDD3LOlEs!CoV#6puS<~<kVwwIOg?6N|D
z^S5#AzIwI6WyDs_{*y?p8ccapu8*0Qy{3&<*>>1RcHXab5J_-crA4?>h>>j)#yzy|
zL=TJ`+@=B>M9E0OwWa)}4Fj=xo_Ve+wJwUpsL{u%>%j&i55<12=EHWiltsT@hYz9;
zB<&M8vCkqDG!Xv<HTJmjclhj|$j3I|g9cQE7PrsP*VPUoa$6^A%)=KWoK>*sXJ3`e
zn~<AQTz;ic_}m-vA6UN;cKY|+EYqaQ@vtXDD>{m(mf(pNc%*t3g^8x{vVSy)!Lj$b
zKb;C8{pw{83|G4=-X$Kog`%hyPJ}3>Tj#aU@i5nLz=Q7Bi!X#uO0zGKA{_6k_59DE
z)Ve5YXLF$)mw+G0!^5ZBPw%?EW36C@q^G7q3>^)Ve1<?_;;7uD0|bAv91&2h7$4oQ
zPlA8t7SOC3H=S%i^X}hx6p;*40=8;ncBV=K;-^s_7=gr`4@NIU@hwt|fiuRXpm_tX
zTVuXFy?ztdD-RMUU5KlNMtwam*ho0jnpR2Y%ZGRRCdZ4M1qJlOkje2i_RGWi_qQEq
zXPp_D*zJ6!PeQ?10RU1-eZOicXs`io0DJUeX1=-JM^1Lk)Gl28u5bZVl(}csie7EX
z0ESRn4+psCOpCiI2QoPuXcrKzfPQ6Sf>xe73p3QorJ<^wa;=tzA^=HD44BW!e%cU=
zLCN2w6(a+#zU@>p0xn<YPcFM+I%;I`j^?$YLRr+gJ-WiDSd{--Fy&^b#O?Cm<6#tW
z*(}Vf6v~HTx5Ji5fCY#|g!CqLd&OcC5nSO2va&fwu{2q|PD^%s8%*_)82V6WqAmhh
zkWj|9YQZlyqQj7J=u~Y{Su;@~aBWzR1PXa82B-kQ(})VGuSALJj@b1_dG%H}Pw`WI
z1!K%ae4p(2zNa)|i%DHm7dYYRi(07IQ1sF0-ScZ>i}`Yrc6T!6WaPsgy2`)q?y=&o
zNZK0_{h`0d<T`VLRt3iIqq;_qZn|hov3^dz(VguI+y}ZC3v>9rwR-V)$=vI}%&q!<
zq2GH?02+W2hUfN{kz3_0W_RzRAVH&;%U@<&(z(5Ns7zc{Uv3R^3~tg)_oFKlN^$WN
zU=9q6_D&Q|M(@+6&2k(sA&xHA26^+rGf@lv?yzsQ$8kg;J++me-|6Bah0*jI9y=Sy
zOu#Eqz44?<N^#a>SiBvgHcOoNdn!r}te8|!z`}@F^>Q-Q0p;FyBe!9XodheHpa^-=
zE9w}~`6ks1;}tA~K*Hj_WyT$+7s-ptXidN3!Aez75@G!lm<|f%3I+=P>=s|+aRq+*
zmr_(Dg2wfe!hC)TN>AgoOy}q8qSv%!rfJEMEOKV|Dc5OFyJ^4~Xr3gH1Q4GY3fxJc
zCs&6SkdI!U;I3jVIu=}(jd<-kYMDoA{hMebU1q<9+s69y%9_4x*}}q>zLTJDSl311
zW7OewkI|Y~`RrlaBODTX)rWnrHWG1)c_1iH|1#R$`5N8@)`K2QaEdWYw{e#Xf%Oex
z143T?`5VtbmuNsC)Ij$x{Z-9+;G8w4xb(}uJ^DXboW(E}LJNKE1GFl#869iyemU?~
z;V3Wz1Ns0s)f$HI?Z5Xl1(*a;>#z!-CmHkP*7Ccu(N*CkHKFn8{;CO5!e|E;iht8>
z{N=5}_H!KM1>KcvEYKFWJq`o81tn51R#VRuC%+j0#|iFr_RLV;0@sLY*lV~0%j?sZ
zf>-y{mCn88xE^L6<JKts341a^#ScsnWL++6YFrsptQo((%Z{o>J2j<@MbBImBqpuV
z)`yP5%6Xbfh($1zsGJ;}+K>20_=3Q7T{U}q`_BK+#&k1YI+9m`SSAW45`hL^$dmxu
z=<!r5({GN!GxoJMD6tDgI~uPS^OnOJ&$8$}n21&Ewlfe~z<`?A*TNHC*adGnS+sIh
zr%i{X%!mM6+fcLdw9?KBV7aH(OgqIV<iqKH=_9uNq@nl_FXb0@<X_s3ASeJ#1w5cD
z97g+Gm&*464EM@h<qLzidjl#yX6oyHD|bf+%10HkXWV@kBe~iD-2t%rOL-<(@b}Oa
zsrP864kp0X^bOv&2HXz7$`ct5Nm#n2MTZpic@tTt_fFI_*O2M#&*`BC;w=(*8i6JJ
z!EjIv9+H<<m6sK^ei^bpI$R5JReO1FFFtiA&Y|ib(U~!^u{#3w<Cz@At~Vf=d8-iD
z?)Oz`eITh&o^N%(LzAq6HKpXs?ub2E!jQ9tejLl-*u)wVU=23+U+eem_(K5D!6@*V
zzytamm)@Y{l(z2CQ79{?BrD`UY05uLQDdXk*ieLBg}F-QwZSJpg8?+^8e7q}@(}1j
z3NjL<;JDvf;8Y85J69DjKFy6DZ^Q7uB{|z4#~y3(Z>}-EfK6X56|Fj1Fn;>_x9JRv
zd-P|4!p1nkHf{HI4KFx@ZxPH32C>YMeMwb)9m=AXkBm~Z`me!KL#T5HoI?}k)duB>
z!dL|JgVVI;s61WU3+wC++!(Zhz_--OWE%^dNY^b9**RLbYqkGsX8G<-qGPn*V(%XB
z<LvAmpCz@0hF)=!jN%}0x2T7ccJdofP^AAmy%{V*<7rO^cFa>q<2c&VcRQZGd4HF^
zi>vyMOs`LMwAr6|R+$%8wkB%1`JzHygT6N^YY0IBVfWg^UH1Vjdn<I?BPo+_P|G0(
zs}nM8e1l0SXMa6&Z6Y-;k2M^U<@LcHlNuSgkKDlsNa;cnBNSed&7E>sO8xCkAj*#?
zIx`ikJ=?2AoKR8#>I6)Gv>qJv_nzHllWCJL25#fm8<9kg9{e9oR~Z%6_q7!P2~i}a
zBn0X14rvf1q`QY6P+E|d?k;KR8oEJ-1_9}2NGa)t_xSs-_pVudqUO%I=j{E|#&vsq
zol?LR>25yDN;WZY)HQb$7?l!hi<b)B;iF@ga(Xb4;QfdfvSws{;1s(&m~;LK=SntM
z{l5auE1le;G#qi=0^(^blIsIar%*QEgOX!U><x`x29Kh`J6^SQo%)nZSq(T~un<pL
z97-A(L!Zrk^H<85Vc<%DD$2{`Aj|t+JT`rjq9IGvpz8IFB;&$SnwhTVv)+^p6BbY@
z%HsFO2!LG)_6htOlA4Q|n2L;pX>mNE#>w;sS~i3FmiZQ~^P#`bt+H=_lPN7KqQh*l
zC=#$f9bm_LLYlw3e=(@h*nUbZ^fo^`YiC~gf81hAT#Khgb&M-Fz?Yj9JareaIy#G)
zRketMPkaIRo<LwKOfIt6EWF~07lEml36>ELpt`#~2fR{qQ9Z=Jr`3)0@jhp8NyLg9
z&q?TTA6*(>k!YU01$}O->^3b3x}4%~<a7YVcz>-KV4p~n!Zm{u>=Yt4YfT|SksUHh
z7fT-{CNxPvVAR!J=;(eDQ@G`r7EZn#eqKU%Q-{H9Wj$NNqus!qEUP8iwiIQr%fq8y
zhg*Gv#T#B(T_iT;sVm;y<&k!gi`H)=*iyTshO=VXPwo_tk1a-ukHU{#a8#N*C^XX(
zh>8-QaY~ua-|P*MP&=N#qIIBsX-klrcOjyY5r?-iVw!O>Jh2QO)GnY-(ZGq8ne7Oz
z@sEdloF@fQ**R;Loq$8$Bb%)}Lz25ZLcgd*hMYe(8y|TZl|NWvxa~O;V+bDG2(#r4
zy5&>)D4hvyuE<gHL~y^%LTpD>YL%|F5e-6iY)jobBOXTweI1#Tk>kip7pGMw-hX}d
z6f`0VZAs4`OH<S9^d63TB*MquZ+wZ;6dWyv5VAaiCzz|~uDtT)jVav-?H=b~*{`BH
zEIwii-w^IjuN6`X2l(kaqf{~S;UZIC?Yfb<_A42PPSQS|!ys313J~P4ZnnJr7)#>(
z?u+{KbXoBj24qKP$SX4}*3!jRG@Djh>;^u?AK|nCVwtjDkFrrLZ+^KAKCSbKj54^V
zANlC8)OI+0UvhgLcxb#8cu1jg{9|vz^otc?6XtM=cqFzTwh=%zOs6Vm-RCJx^eE88
zX;t#%t7>ff5`&Bxb{1YHCDQpGJw^zVEoEwcf|<4E5+wy?NP%dlY_8Y4x|U8cFsgdY
z$}u_nsQ)A>IPSC|7^>?jh0siGL?xK=s49U5Hn2Ui4Xb>5pzEh^;41s2^XomX!BQXj
z?B;X3({69bw&foYDB+y%L5ldmnf5!;2~du(D3ee)8pk8h_0oPLKlTiBA-T$_8?uop
zpuWjBHgNx~vS|eoZXU#+=k$;DHJ)mE20#pOvz%{ex)nQBIHBuToxycgSL1n`7Z>s_
znpk)0Cu=&9u}BCbV^VO;FAaR)A7z_Nr&GOdl#CfaHC!c+xaT0hQI__i6JGg_aqDwK
zUnB4VbT_3?U&&63a(-di9Z(cOcC32xs<jBU4y}$Qs~w)nixr~)sGUL!_;RXCH?BqM
z8;;n9@GP&1=C>XIif}+n{$&lRyzRKKN`zv`KE`y`4cl+Jt_g%9me$GOeith?%1vV%
z;FcG&dlCj8SB`3Dg3$Aijc(g0>(}>seb$+1oyNySGUk5!s9MWE7kz0Gve@M6)UzEq
z+L-9a%uJ61q$JUh9ts~v`<0Ubp<}!&#}+3~1<CsT!2s;u<?<?DNsK&fHy>9szcXL(
z!R1*mr;XH>Z~G@0Co>r694=f4GX(%X*3x{g;AGA{NPD--DVsSvoi~&7E*>*?uZnON
zbk)?5>+`ih)leqSQx?Ap)BtR&GQ`PaS^C54i<8rIvyP{s=~9T~kFI-ea`QIpX~8tS
z8Z-pMZsue1<8r#RHV9$Qe9L5;@)4%O8>FJR=C;8TuIwaB+w&)CPexD_zfu9mM(pb4
zOvM8lVZF~$o2el81F6s2gij2~hQskZ`3?Jr)LM1e^p}-8TFj9O*M6AU7@vkgtR+3k
zU47PrlJ(eJcJ14jexJu8M@zoHS~{9d6!_K8wWt)lm$^vInk8!$eR~jrQbEq5B0+%^
zkSjIjU3|Gx#8Q#DH|GA;|H9onswM8{VN87^C$XfwT<p(AP=>g({c8|T{xyF-f`U#C
zkXhYYqEy<a{rVD$`^e@Ko7NEY3VWB=SC@9V%|6=%hl+AAJYapcTEvhB8EqoS<?rXW
zQ7Bhhb+vb?*?g783AFWzr<q6wrDbrDz=^&$xlh5b={tZDj*f<up}sIuIRrC_ZWRnD
zaOFgdJ5_Qpg?Hb$h_A(gB;2r6s+jbqHxta0(Yl>^u;1^Ea%H!vLx&}5Cf}>zI4y?i
zIVfV8TlR%yZg%}J)+4%xNJEfV+~i4;KSoL@*rJhYB;xA(KF7&Nb{}t2o6q6y@hDGG
zkMz$sPG$t^ev1WtBnLEKh`IJCuJsl_|NS~b82xcwoloF0*%#(Q)SVd4-U&Y66JPPS
zF;v-gnFCtbk7@81)uA~h+B{{uJ?_y|-(PZ#kegS&ez?D&{52z1Kof#@Is1Ij@OC%8
z{iLQ+=#ki|(yyz&te+1uq8$x;6oCy23$1YnDtH|YT#Cg3!>+zzuF(n}zy;<hfWA-?
zx2a?Nt^-v$h7Q(XD0@Z;^gmlcc3wT5<JDGkfAw~sfJhs#DOBa8$&O2^@@8u9H?&!~
zeevljNxN_uWya|*Nkb_r=VxMmn%~XmAM0mN)yOB?$Tomb3hLujP{FHzD61`4J{0ZA
zb+7BIX2RgS_}kJZbKS9H>TUlva#4UH5M;+}tsk`K&fHOVX>fI8hO^!pmp45tp%$#A
z$RoS%neW<>ZeRJCYFr;QJs!{7OH*VkHI;DUwSIA*WZJVR$bYvfYW`u-$YWJkyR4n?
z;h22wGz1IqA&W~PHL26zepaSf;JM!3on@QZ(kRJ#G)mKg=QBc)BQz&NP>bA%Xoi#^
zU5k8)u-hoTPz$+wh-0q(=T2I?w86o6-kWnJuXxAv-E9|4RR4R^1Uk`*!x-04uJ?yJ
zvSF~W5cf>c{qAo-c*F+X#B<+#2~*-CpX;1jRXN`-GVsVUdh9!F59k5V!t3w{Ma>XQ
z&U}60YnnrQNcAz+KjASFs~yb0-5o)cg!&Xu-pSeihn<n@Y$ZqLbe%w{E8cufC^mAT
zM;R^^vJ~P;^I+;6>TQg_y{ay1kx7kJV7J`H4AWZgsrJv{jt#kgJin2-m7bVfLt~|v
zw~v(hla#eKleQvA4OER84MwBdD!G`TT#}dLAjC_E?>4$49x7JJmN-n6edHVY`4uK$
zH1dqhnR121app>emc3Df8ZBCL>BgrTn|?4WAS#~|pSTPrT=?z|2EkGM0{*NpY+ZOC
zR%JzDPJ8?MJKR(9G^3x))y*?O>Msm4ra&!_XOEzd|C(Z#0k1uymiUktTQZTVfapk@
zNM#JRGgvAiAseY!7VQw#(!Op4*yzk)meH+i8_B>CSpow#L38Q-*>cyz)F^h_3GH^)
zwb709T&c*e?qmMxGQrv`24D{{Um~-cwsG&$h*8xY9jOL8Z7A&Qj98h3jEd$CHVmdz
z-~0~jh-frU2BKA5p3s0@?e1ao?^T(s;fzb9ntaOye;OW<(X(N&gmZ+dJK0S|L;6cd
zE?Dc%RC{dFGE5Z?rvy{E1YF;YvQ6y7uR5ZM=}$8egcW*8j-u}F?#7$TX(uY*e8?1#
zr<cxs`RM%wTh?gS@yjn)ZZ?)(2djQjOg<SezO1pvFxxMj-C(N5wcjq@#??IF3kLHS
zvGPy8pV=#DaXLqCE0VR7Gvu;O3c@Y2bXAFA=+|&hc+qyH=BhtRE_rcHyKV4*{D_i#
z>}sT4_R@cOuybtQp{07r0y7WW%DI<skPk<*O1L(Cn+EN1NxpHSdqFjSsLPKX^VHuM
z&tELujdnJyjh}P3Z5rO8NITqheQ7t$xW#txo{tj0d%RTtk<^n#Reow}>Jb~9Ep7NU
z6mCdXT2`hq&0RA22SPrewzYK8XI<O$W&a<%;2AQYvYtXa9V_%wU18b2)_M*p&~xiP
z0@I3qn>w~O(qc->qV`K(M@dw=Ic31j&jvv#rIDWurku-Z?^dnp6j0VG$`4jUk>|#9
zNHE_OAnOXhk7g^!Lg>^~XXXcARa+SFJ28Jy{Xncd>-JnD$h#O?{-uTPuSl9+8d3QO
zhy`Djj9GH&($PG2r>yMAJ7=;{;Pt%WF4gPs=&Su+^*BZ%=g?;v%V~J&x9Udha3q{>
zR2*`dks;N5CL*>W1axwp#m3RJ&o~An&t!UO0WUI3`e&Y{FHZP9Le-wd`dneqqnD$Y
z8+_wA3?A6=BiHI}<dvm&mS*P4)%Hlm->wgD64s6IW%JCE8Rd6pGbgPmfmG^PcZ=9)
zi4Vw5g3Y^Uz2PI{jzBjy;taY9D<Kf#SHaA@>_}-`Ci_>UvYsm*?hIwN9NcHW4|zZ;
zMt8h5`@nM1aCvly7`V74eQF;8g=nXIJ$qe7-RP&18ErkI{|ZN&qk>FcOjOPG^F(i`
zr~A~a&s?nKTP@ryJNY2yEBSXOwhm=&5~wz&>-86T34`p{<nn;DDFHp)sVBvIpiw;Z
z5g?QSmT%b1fmrE#xipH$h>%SX8xP$$6dYR(R2dlQf>y$YJR-$i9$c$p{eT8m<KoIy
zQKqd4?LsnZ;pV|`H~@R?q_g=(cwB2VOjBfp*ZWvzHTeW*>9}q5PWb+EWcSwp#r?<o
zcXJjO26}^Nkj~libw%ZxQck}Y0BJA%%vJj$z5Sf^rO=f#O)>i9=q-h{cg8{e1KXnz
zE5cU7+GfV5^||v^37;!#tUgo=N7QvpvXwJ|GqAdZUuFBopn1UK`MJaKn&>^t@zr^i
z$)~E5=9Y7x2T1wdj!$314b}^0{J_h`#-vLxSW8+v5bX3NW&I$P%5;y%1Jdbp&7h3H
zPLBqMLNzoHu1TT!jPrc~wW$qcA(zz;WaP_HhryRw3>ydukrJ@h<rhA(nCTfg7=Nf<
zOcfmh>2B23%=W)PY~UxpX#t{Tq0)xN<KN<I#NE=_{k7DO7!iD;Lw&U%Azp&M(<0{A
zdm4X5Dc*U`pW|-!BQ{_$hvn{-zk2)v$p4`L*fsUE;N!JDaX_&XNjrPIH8<X7UmO{=
ziVf1d6a0>s&P#sHna(%Ys-Q3;W(;}a%Rd*CZ=t6xevMx{&wCRS`;RZ~@DMMcNutG{
z$i4ktCq%OfVoNuU`%$CCt(MZc@QMm@Y3>wNs4>Wqp5|}lpY7Q}FSc(%h}xghO(S!9
zlHRagviN{A7=C`msLfL9?}f|TWY%!d^Wd9xRmEVkly9^&Bv#$h`kVoxm~GrCi&ON9
zwxR0kOppAFutUa`%csQpvPCLKyQL4rQIBl@D%ynM_!%(F>c3(sH&>1i*?%Ya_>?l5
z;YuBOt+vIE&X9+`_6Q)SaCaoTqKxmuBHmYCS>6mq`bMAQW4uZNy=s1??-bTABXad7
z_(mIC^itw&@#q)xnA_4P?TAKay@Lp%8G0x%he!rn>Ux0|1|ZDt_uH)XSHBAHkN+u$
zyo|1Qy#{NuOk8(ww24L!7qwPTk(yJS7?N75W5B1I8)@NVo|2fzxslJh+J`*obA|Yw
zbxjyB`6@R3B<*mkjPRcN?51tMPW%awXHw)-jO||vP4DM?DgY-jz3yTFX7^tC!j&#K
z-nO$I*si>8wkuvoaVRL=*N__>t+jjS{Z9(imn7dZE}4l5#%C1Nnz}Zcsn+TSIV99f
zPjMEJrq~CrfQW`N8F{F!^KC1D+erd9uNR_i<eZ|gagl(CvRg67Y$VTC!Iy(Zo2VDq
z6hc703TS!beE~MNBnK$!40WdlTTX@-zP!Eo;#NwZsGAl12`B=Ml5Jqfge!kw!>9U#
zn!iUl`)CTA=nHQEJuVM;mhJ>&Rem0U>$dP2b#xAz+3h^0_wNg66c8!Xuj!3vwU8fE
zQYs^p1h(09LBq5s;q2*uXuOQ|aK0*H0*qP)A#@g_;H4)Z?H-DkDU|T(*O0XWZ`>a1
zv+3B-gP4-Vz>eah_z`DlJo^CdbIpB?pB2I%+4{JG85MF?tshY424KCo9)-D@?l$9(
zxB*_b0Z<S<ptU^)>PK_CqVQ~4d{tQ3n~8C3q~8&y$(BRl3xs%=y)8ls_-DRsHF|vz
z|4JByPATie+!tti)%-B}rrvQWM@xSzpK;oKXP`I5=fL5}nf#hjQOS$!ak4trQ^sz_
zU44UDx>pt$yHDZ<oZRP(-ig_LGYxxn_>*rL?!ttXZ!-$?_Tt3=gL&=#z)Gu%W}tV7
z?YFY2@^Y?>FI0Cq1}J~Gg6RuA>|n#kbP|$}TlemJf=R+qX;2D1&<}p7kD^53#ST-g
zmMT@fW|$A{DZ*hQt~`wV9rlj_d9)>R=JUP9eSIJ4Z*aX&tl{;O<*1i_AV;maS?oIf
zpg-Qbz|EK$q0Y)5`Uib``AYRTZJs@>^}c9cbZ4)%^G5UpdT#KbM6i6M*)NDP_0*!)
zbP~DX<L-g5=-%Hs4W)9O&c=(cCR~p^PdHCm@5W8ZR79E;&vv4^>u<LVyWIal@6^Xu
zKosiEBAc)Rh`s^00-#i$s^of&W;H3G2gF#McXgevV|)TSzvNys-<@*y>=>k}xu02u
zX<f($6qj1+>EO`tO%7`9(uw$>-6#ysiiaNAiK|rl36s#3nL_bqql;Yzo!>?|wFrdP
z>ZwtOHexan%j}baV0HZc<r)?cG{WK*Dyd41MAD_AqTAutz!%y>qntMGCXbk*JMdv{
zlKjT><Nvh)MJ8p@3+enEH0;^<o$<DB#FQ>L6u)LPQckBD7i<@}Vz$^F8Ls$~@E;1t
zRC_{3bc1OsZ1M8sqZtCLG8glKMoc=;B);&O^RP~Vme%l@`#W4@he*cj6!<j{iHGMD
zfi}94PF>0`a-rGK=KA~drST86??h_PpI)t8YLf5rH{N!UBc^`-(a_X{2uYwBEcVSB
z5?e&%*1aYdQxF%OxHQr}%m;jH0bVV4DRFPZITJ(JS*CXmz0+`rcQCH799rT&#GiE8
z>=#q}kLkY;mWS_6?G#Q<%9jO8A)}Teu5BMDCnw9ToA)EP>de!%&}_|c=CzQ0<6qDr
z8T*>xJSM=Q>eG5f6S|^Msba#2G4LHN&ZrVMP!i^kFI2neKfkE|(Y0-BfpgEjVm0IP
z?}6u^r6$vlNhJ=vyR`a6^0m7Mt>=|f&6yWhqRQ<8E41FbqU(n~7p#A^?}!0cH;5;N
zaiL72BTe|Fz%*WxR^^oQ>}N#o97>y$4wTIySp>UXl^N@)1X~$%5-QQwHMs#qwg9=p
z-u#raAF=~HMrhG4D0k}B%L=2CTaxI~fBFZJ5GDabIp<zxoKftQJex@h^dPX(o=RJ%
z_ftl~(x6GY(mej``1ts7#5?c(2k-J;bP7(+k>U0}+Y4Mgxs`_Vmd{OV@`jo3NAN@P
zWY^Y=-B(SSd>;ILh6Nq&8IB(uZ<h-1F(%*~CBlPnHR{;?$buwUYZAWcu#3ma8;ql_
zw-+xBueQoM=D{BsKkijx{9EU>_)Cs@*`*vv^s-Fl649<95(OyTiky9T5@P#8K-&E|
zUgSddtVcZPQRq)L1ES{y&ku7??16CVz`8ur35KUwoE+h4X^<Q6-6^v;vg&25)(qDl
zb9RLx)Fpwk6yC&E?*ZOJ7j!g>i0?eyTc7>VtIX_l1SQev=yT|(s8uD$%#wF%NGI-f
zhU~A1soZiZKEF|ZWsb#LYHUmzCXg7?lZ@pXz3UbvY0^qN$=>wXq<jjp(;#f)hUj`r
zzN|bugy&ysR-eZ1KpMwj(QS~Xo*o}0Uf#el&=()RoGno^Ar~Xz=wRF+$XS|7?|BzK
z?p7z5x|mszgm%B3mviLk3i$x?HE9q7GOmSR4SjYB-R#Z<^A2{*`6tF~NWJ%H-Dw8<
zaRV<L8`zLdyme?}T3|6%Y?)4Ej~$mE8{UX%X%j@eQ46Aq$<v8R`8wf8rJ$W6wjmsR
zf?pAis&`<%kjg_EkMCcxW~$z{8Z2^Rc(lXojZU8na<+Q9_yk@oES#zJl!hMOwYWrm
zUcJV4+l8&e^Q1A{Ae>>^{Wid7?nyv4*qBKE=Z8fOby5%!A*$~jiq}`k<4jY-fzNod
zyb%YRko{Ll=X<DMX8AmdOp+PlJRh``DWsvF;&@C7^7HdINKhL2yp-u1A?|2D-}ja&
zb#>bItmdq4M9A<CoH*R^YTwWKVBSA@eTQ9nHdRptJ(FqVzI+W1zgKUcadfhUJzu*j
z>S^e(5k%>km`mDvdbnS-6>PW><^cJ3x$~6C|33aKm4lo$vr84YvJVYOWH(uDTVbmp
z1p%zf3XN=inSvLpXbpTrGk@Nu1CsU3jT|^fqUEF;MHr2mMUTB2Z29!!hZvOi$#j19
zp8boHv!YBElrbj@5Ti}7q6a^z&ENkCeQ*x&wDp!)CHeVTiw==5=w9X$hle4_u_lm~
zJtGOzSh>@Qqn9h7NjQrN{|iCap$E$QqjeosEAf4#nx^b(-v|k3_q^F1y61#J{*B@a
zzR*lX1%t(dzxtXym$IYI4k|VJbR77+l=37p+lOV;Md~eF1G!K5zMeeZdeW^I0}*G9
zQM4Hx=owma5S%tIiDMhUj~j=-M?>)Iaml`PbI<LrW@jQ%=G7#f8N97c@`-%dE_;YA
z_w;&qYm4B%|7&=Iavmh?5{gWT<Pfo2i(T9Y%j0@XX!6n+lDl1!08B$oP{jc{zd{a-
ztX|CS;;K`Q2o3fyH$4<!|5N4G+cDoe!?jP-7s+wTD_Od*Y;0{B*_~Kkn>)B*@8cdp
z*~>Mb1<`{8Kb^Y~O<~FkgGU265=QM0uPj}uW)I(z4v}0JnZ%bACJNj`DFVN{DP}L9
zthGEvMpVPyI$NO7SSsXWN%jh|Fu00N#OF+HJED+)VF^x;7gs&Z|1MUZYR)e1k}dC+
z_e-;p`xfYv9yotudauNU8x)U=M5C@w4Ijh|c8P4Ht_Og?j#l*Y$tAYnvx5<`>9<%;
zchU1eaEiDr#}po?{nk?bs<p6?phvTYZ1LZYG=A&_V^A5y92|FxObW_}Y7vrEiy)^5
zPYM+7!izV{pi<?g_ohR?1{(NeXTOk_?FA<4y%KWw5(r4X66~YSk3g9DD`tPh3rghq
zpxfK<sqb^0J?j|P+`F%O=mI*FxKfmya13kQsa}%p!WDE`>XaZ!jLnD*7?-UpZaOAY
zcyGgBSq;a*;9fS+^7QmG)G>)($^oHLF;td*W|xNUrG7KKlVU=TtOu#8goGbNXI5uU
z<o8^PWupi#21I3%{JqYnjO$01!{fumYHkg9<VMbCp!3{fI0fYLf5ky+%nTf@rQXed
zHlCVr<YiJZxM57-^NupH6aBj^{z&h$Cg|+S7p9(As_*J6j$49P6=Cf?QKWZdZ5x+a
zRmKM|vGqm`R$b5UKijvSTgKFck63rQ$ejFf5idrDdwOVhQGE2ew~{MBo)~yJ)s9Js
z1b2_rV87v+1PwkFwv`tJ+Ao{^g|g-5arc3@a1k?kiq8W_3FIahWe;~H*FuK+w~y;D
z+M@;p6scaPIWtS$-rdk@B^D%@>Il|lRbP$E(G=qi5z9cHWNdwlc8`%w$m5C>4AT<3
z7^mAh$k{a8aqI*!C~$ZCVdC_YQ_kVtX7r?>9nLqz=`Us;$^X3^?@au65u$!M6NF}6
zC2V=vMx%%AM`hi4BADbug9V<+BO05nfCXyK<7fRn%6XLi#NC?cb%&wn!*&gnbNr`O
ztvN1YQhpb(Q*g;r%nTZbJ-jGpqvx7!h{lr#1fqS{U-0XHe<YFi7e)&+D=8)N!4L^}
zDQP4oguKxKrVcKGEu)Gg$MPf<YLLGjEfFi<z%|q;FfiiOA|jg@fq=s-^mG#8h<Bf;
zq`S+aSeEB-MuZ-*t4&{imd^jGRPYM>2jf%)Z!TA)HX=FkIVH-xl>EMsAX!hNx9w$_
z?TQs7rUS+?gDtx>F<!mlZ<b8&(u5uchgwcmPAfDz94Bi;HlGB^&HfrR@zp8FvD%in
z^^pl&vogO-Y<-?JM`ZK_Nc!_BCUD9SMYDm9i3+YQ9D4!V)S4U^={j=O<aATZG3veF
zqHomURb0Eavyx8WH(N2mds}8E5_;Y=2r2RzXeYk!ctVNxwia<qk!y`~vc&RWh7euV
zWO=|lskX2{Izi>Z817A{hLFl&4lcS6gA9gs^sQ_3G<Uxv{F4Fbv;dH*j3V#cYy{F`
z27mUoZz^{_@xxx~w>^8!fY@=x@j^p;Ir{QqIk>)Uzi_7INO|mhhaE@xje~`vf^rV&
z&0Y8&dt}CdyDCi<AN~hHWmDlsZ;v+nOj%-*A9h2t_Z~UFV#iUH2%y~%Rif}P6l<Va
zEoY|p8%F`$nI7wEt}YpBR3m@eC@?4J5?j>vc$i4OFP=HDste6yPO$xV<~-i<T5eaq
zOEC-)>>2H|=5sq!Hl28l9YNfy01U{spL1S<Fr)jwR{dv?;he~6xSqm+_u;7X>e>sZ
zBWItu^_3d(8Jna!Ns3=osXW+zCwF-CpzkBz<xl7K;@6E~yewE`nQbn=yBb{XHn}m)
z>oxS+JZXdl-%pW0bQt(3{JRF@tWcH`2_yXjSFiI{fn!#i?%b1pu=WGDHoq#fX}Rg!
z*+asBXwGsIZFM`cLw1ehMlG&LR`R5UOkS~1xD4ovlAI2tpK-PF{*GmYOO!Os;5$02
zp5(`c8Zc<{L^jt18Pl596gIl6$9|hpKXl|;itD4pvs(`>Q@Z|2qE@Lw2WqV3kMw2k
z7dBzoevG-#2-hyo;CWP&I5aC|6S4Jh?p&tL$TpaY&}l~cbfl+3Sgd1HwK@x2XkGGU
zDKrxR<~D;wGI1zNKz7kP*8w<<<%pbV4PN7D`nM%PAE+(eTOzP_{aE|^2PBiqaTfHs
zm>6VgU3@$VIn^~IWUzS|(xceXCB(|{3o(HqEKdi!CGXgG9stO%nR9LK$JyQ~ygc#O
zj1(0pp4riE+xJRJII=>QA|@$P)+8uH#$Xc(g^@4P|9Fvea%TmZ*|!E$zJU=944RWe
zLJE?S5cS-uHWzB>#J|r{zWdmAVR(CbL*RZG^3rvaJH*dBXWQ~cmSpLRA^F~oh||*W
zgP6mg*Q1>1Dtiwecr1*BHF5cP3(3Ffi1hnOq{4qCmp{WLl+6>9Zx9;DVmMgJ<X*^}
zUeJ9N{vuD_M1SFz;7-Bb#ae-w&9>dYB~~SCaLp2W_;3=Rb&rx+aIA5s&~mo_K=0W8
zV?rXQfv=CBS5})RbGd<iBTR-d?;FWn7~EL+)VXJe@d4Gf;+~DrIV_F5^2Ej`WNG#A
z%g2f@v%)gPG%_W`yT&yqBcH#-GukVXMtmYoN?dMiHq!AlSf-!(Lv1;oD<0!``Njw3
zuHNSvWVcXc`PA@^-(YFss@TB!e+%#AzwHgL_H04_)pG*wq@b%5-2*f86phHi-0(-F
ztjvwEpLk6pH!QNDB95v$k%{L;?3jo6<Ve%B-1VD5-UYLL%P*Onumm##-aOOk+D|cl
zb0JIlqs-D=Sc-~|h4~J6A=GLreZ}h)MAi5R`SgjN6N2FSiP(LKs>%1J<$s=7QC3o}
zp!HSAeBLGrSC=zb9NM7{tw%{`!5Vp*_?95v?lYIB^z55rLs2y}WLHA&i^R?Bm<Me;
z_g$EY-qTW--zCIbWbmi*<jL<D84$Avbk{{@HMv*EW1+ih-4qF5zgJ<Y!l{^vC?<}0
zotm70QP@Q$v$a6H;3_u`pZ6_R5rH1r1|7bsOgJX#M<DcWl69ido+9uPQ;;i0slDLh
z`L5=R4aHqEgJ^LjBJzzLTKP|oGM>McTy3m=O=zM4W&QX}6)->!Q3@LHlZ1&>Y@{%_
zT1Cy*CzZ>i&?Qbi%jppk!d}4y$(C!{tF^|5h%%q6>6z9eWtD>+UfsjZhhdTa(kdpB
z9t)YcK38wgsqTz_b~@vVG?B4RUFuN1UXqF42<IisS*;)`=@_Bh>>HH4y5|UC9gO(U
z$*+TN;TB9~h3DQGD9p*4iL7Lr>AQvLhoT-LB09wOiq_8wKnby)&9pn`{(KXRJ`BOp
zdo7X*S@^F`s?C4M^&yIVUgC2uO3{^zbT~Fedl_+f?H*45_dS;^no0j0i6@x{y4Uq}
z`Y1U|)$8-?s`jqn4j2z8UX^4f2JtQwL*KJ<+^;^QOpDxceriA7J-qo$p7|h75MgTI
z>BPJyUH?@PgJ25mPZ1}lYQtkxn-8AkGy60xu*m6HiOQQlw2fD|K54@~r)qhFG78gr
z;~U$`cT9rI_okzZ=*N>QUMp_qDeCN-5_00@=0lwi#OwF<FmQQYf*g=R%$|nO7|dL~
zEasqaOeT_G!>E!cmhRdm@GSIffb_v(#u|bwUSBI9s{Bd{VTnT22NI`@vh;k*gFi}X
zKl<6@pcKJTR4IX9cmO#bFWl)pnq!kSUBLG&%=q7WdxhcE9X~$oov9tWMtqYwnydV1
zbzKB+__e9t_DIpX7XtY!@)aA|gj&X2$ah=yi=KuPYn(c+^2ygGUF}ibLT=sytUIdJ
zgyH8ULx}WTs4tNE#^r-~iBh=jhM{z&I$K)WY@K3{s$70QuL)2B;4Cnrc^o<wxN!E%
z)C+i>JA&B^^%mT46x+Y?Z<qG=Tscv-Z<UluhD)EX>-=hPOYbcSa%Sk^14|RLJG}d2
zcS{`;i2_7uKI4=Aw`3!Uj+@K7)(0|g^Ecg|%9)a`_jxcaa<Z#db*OQ?T3)De;+O(W
zeh3MFv;KQiVMdafrc;%~n};`9y5-mil9uzRhbPsu!<ODbhI?L|KFRHo7~k_v^oTiV
zR5&s{|2=GP!CF&Z<eTfm&y{d?nEmyB!<5$;o=@{JR;9os(rLdEg!am@evAtL+Gd4D
z$;Zqr(IHch5vhUYT7czjN)I#R9HpDuLLd@0=JGcGG)OBPeR$h+;@gbby*l7%oE+O~
zvqz;C8`d4a8<*NG)$Kxi)f84^1@_b{>5kxrO&%cd&Wr1-v2n$vu5UYGW#fN%lb3b#
z^jV~6)F<xV9+WoN-4id!jhZb1kv>^r5EAswWplZ`V{Ju)ix6TT+TX<|(l#&t&2{9Y
zRgqm+S!k!`{dC97y4Ow|xM`=REuU`4Yg(IL%a};6OddIia&mN03ya(gWNjRjf|HtR
z0;he%C-`f;U;hBxAFK0bbX8Q2{FL9&*k9Kjy)FV~z`10#K$5ivUwV+byn9GkyLO09
zlu7<1>%J<uqU#g8xTt^b>2phy%ZPCIRCR_~C2J5g8yIeen7q;4J5%Lx>KS({HSC4c
zRL#`2X8rfR;o10Zx(0OexO5Sgg*Pj&T(qJN+iD)k)tW77ZPw<>BC9SCh9M<DEr~M7
zkwH+*53=qxV`x1rGatQ_AvhkmM;Fzr7Uh2Ee+M2e?nY-TdfE6DGxJ$;GrXcrICSW|
zsVU}&cI1MenOTZzx<~(5alNd}a*Qv?eTah09AlCN-nL0v!Q_yc2u=)-qIEsx+<+CM
zF_7GFUMjB8=`H%j>`1TqD?ydf`n^x1ATKG)zecezPUw`VP~OozI{Ohb&5iK;Ogd}Y
z2Eht?oE`|om+#)8tT4?Q2c=UmKAMXKyDN9>kv7%LzB$*FGc=mR1}Veczv}tBM~_+D
zaPrY8E&A+TxDrClE#n5_=oDB(Qi}W~5E9>=j`+wW(PRTx7CA4{76VS;DtW1`<?yF`
z2E<B?y)D&zFCw7Z5^CCDbW!>)jUI6xDGL@t?$e$Nu9WNV;4zXR?>)lVYeR3OS*FZO
zqLH(2hAr2e>AQI8i9rU$yjmV|Oqsh@e9Ou+yP$Q>%=8c*;=XB(AMgNstc&MP3f+#x
z3Jk_XGTLY_B<gjr6f1?u?O>!?<dQbfBQ;}n^v%&C)JH{aG*eLkGONj9xmxO%RoITR
zd^q$(8?pX@Aul-_MFh(MxtIt*ATxQ@<_2xcR+)aj99^q^3!oe(MU4NF(;;;DS1$$!
zzVPi11eId7Eh$d`9fozk603Pd2{;qi;IPimtTh*f_naELHlBa7!4PaElBN{1`Vufs
zTfBUxkK1AI>_>5^Kd=bY?R}aE{<(PbjH73OuUwfC4T)aZAp(n+7T^6EHS5bp8X{F%
z>~ALd-T3H<@WuUVT$G&UZI;hk`&K<I(FFmfJYsvZ|Ju!1rB~@|Mx@0XHN21xWbquM
z_Owb^q`%ZVQMHzqm?xEsW2YkyP_Ap|p<SP-Mkuv>bxIG@)zd<?r3k1m$X)uQ$w!X8
zL>X?LymNtv_}$Q<sl;GEWQ{}5C)y|bagVdX(BNXTA?FwLdJy@8Sd6T6<JvKu<q17X
z)qhZep(xDqB#{t<T;fxYNae3Ms>n}Q*o5qWf2ihCXAyCk%6sPAWAPG&E`R;0K#<v&
zj%!Dry8U-6OXHx1T}a{O>U`5;G$wfgwOxvMbCj!B{gK*s{m|^wQVRDE>dotGb+!Sn
zG!kfbI-PhqewaY$H1<PztGy6>0JW(ohr8yB9Yk+s6Sg)|YTTYBvd*%w|90;}O#=?~
zT#B}K-!!nbH8K}2%+EKp;+aNyGxbzyFem|i&rXOE@li7AI!S4u_Fd0k85srDvTgd#
z9%H{3hj{Vm^mL59AQh~Sh7$RC5>_`8hTYQBf&|MsJ`9wIU{3`=Q!&Z+!q`834!2Hj
z5Nq<*1|WhyHCK_Lk0Fh2U?j_+Iv95eaB=Kk+7B1w0@uoo{glWoSe#8o78<$|Ut102
zX`u^`Kg5=a!<Q&>8K1aoI<;k0Yg;i8_SJ%j2Gpx3B=6=cj`Br8eB9k1WoIa;wg(6W
zmIqi3?ud#HMD(#r-V*#cgHccW4M@=XF4_;c$cSpT9c&oOQ#kS^rN<{WJGZyV;Q|o*
z%G>X(O0t^kNVq@#XDKi8+k|{BU(NbsLe+IF$`4j@tZJk~#EEZ}K0G6LTW)<(N~~Of
z7<LzpQv|A(>b+8*r?r?>FeNnmQELDpm<x~eUZ&{v_6Gx7vVGSZ5|sazfREQDwwsNf
zv+I&$3)_0pH=>=#2g6v10VmT^E3Z#5gKg24$FwvpO$jw0tq#hVJTu}o>oJ=k3oVQ|
zJx4=TwxnQFJW%xx08d{+q$i_7#~Z)3VQ4>$<C&g(A-R<4*EzTQgF&B=`s1j-mIwD(
z0v5-=mTPAosM~M)AN-be^w!`FdA3OMaJ?(V?iQz4gdp32jeyb_Os)o^(*jv8+d}kd
zIDTRN#@Y%QLav-9X~~pp;b_3oW{S;5i0bI_5fvZIX6BDox&8X6@T9Zrm}BLS{sg|}
zkuDE4{a&Q`dSx4sUmLNQzNqAfc_gwv2g`=~YSv-Y>BQ9k17>$K&lQ)BC1M!F3#b#I
zc?{mdoHYI13+Xxp5tbUdZ}Vwl{J?%O1*23z&ZzzC2#GhodrD~h{o|*nx`c$zPcgeY
z7ual=z#^J{-w#`H!+(_o^k>l>Y~UE;lulV&8poCpqxW@&dX)Z(ZGsLF_im_+xbUu;
znY%;ctGQC)?h@Y)<#-cFD@b+8dNgF}`E?9Ul{A7T#yR>X-{c}P)T<`mTcxi}!8Wh+
za-hB+5y_L8a~@zfoa}(XT7-*!=&xV~(}0n4xc$<a*bIXWFutw6gr7Kk$=nOc>WZ3)
zsF%kO{5Q`~>F?At(9?B(Zbu{XSI?_EaluVvAW7y;gzjifJ`c!qc`nWr)||#|>zX4t
zqnW6#-~!A;Xo<TG0dhpSxI9yVmLhH-zApDHLL6eEu}GJ&cdi<#(Az5YPLomIQ`Dc=
zKKT=8;AdYrioxQa%GnSpKQPx`d-VGUz1L^kt{5z~Sadjn4h%4d5i0C+5A?!>jVV|G
zbj)ty@>Q^-3F|Ql`At^e;VsWHqOF5&D0%0-cm1WtJ)3)i9^*J%jab|~=hb~0FX7R}
zYm?Oljl%VfVLx7%(BOcf-t4@Ue3jebERD#7BfhQvxb4)G--&6!7mutNboCAld696g
zWLwD8To2)Me%dNahwgdDWnZ0;7?)t2tJR#PXJk)W!w}l0pTBJ$YM%qLDgRK$@HBmW
zL4cKjSvn;FEv9Bc1`b&b!Ohw?Hm)@~;1W@{A0-Aa5Fd3ceg4a1&h@lO&qc}{0pF~Z
zlR|u9`aB6Ts))?kC%iRuVFb@O&VEs6LNmp_$H^*bKHqZx5$!VO7MTb#Qy*^^q7TAz
z@{CIQ{)#y2q}D!uA_i>RUIzp1JzK1I5bqYBjHm(6bgtN|bCORNhQ_(Rw_&t&^fTLN
z)+Eg@coI-4w+s9v<p$uK2HKGX+_nDuRBdm?)A1N0K;JXIb|=M3%8Uh^aR7ruZQQ)2
zlVa@5DQ*0MQxi<G<!;wZ@kQDhcIG>gU=BfIjd=Y<_#zQh+F%Sd-pWg}UGz>RDsApe
zLCkklD0i@vp)DIwLV~^=@(*<hH7?(_49(3CF;|x`@&dp)Vw|1X=G=pr9!a$)U4@X8
z7aj>h@h0Az^mlZD@H#lCaSG^8n40A9axCDh5=>5MfSt0JuzLb`FG$Euvl=<$j`*y|
zFe;ttn4RUMCtd%JiYAS$|4Lk<V%l5)*XzovXBaZ<&6=HM4I-DCt=4{cCQ?Y^4IMBu
zH{$GFjDv6@-8LR3lHTH@y)*JVCe##*=3FApOw2Exd7&go5x}f8OK0*E-oWLxYSq(U
zOsN)&;MklTu@1(jN&6DU%T38A2I7=T;8h&Hi<#fc?*!)6IV+g$^&PA4IxbArtyJkj
zRN=moAHm;EzRC}e!DO^xt!3&2cwFgXgKi)krMDrx(Q`f{-hhE$%gy$$`mT$OTnAqz
z&x#P*%Lb968FFg9**>Dp)J6L)Iml&18s}b<+poaNrG}6<_vG^|bP*eHYK-P1S)^FH
z<n}uK%9^o|!@F_S#Lc}|o<(aCqO0e5tDH|U83;MOqrl=ZtrB;%L(KDyEk8#pNT7*B
zVg@E1_2TV3N7t9wjD!2`!yDq#<=<T*cPIE_`OBD=ea!#<hAMpHuT9HXAIkV5N?@uX
zXs<+)4jel&y{v)Xps%NC@d142&H824$x<-HB$<H+BO*Q)3Q5rj>NaOG!u*NFZ|HfQ
zVm9Vh7xsu!!1FMUzOw~=L7&(oWlc}i-D<)*n!X?bv6*MmT+;vP!Nw8x@0fXb$c2i8
z`Z0DVsGNV(mcZz2^d2Bz(uLY+nEFRe`<OJABqoYm%>J<guJcey6{7ZoBDbyUqI^jO
zxfq7)*YJFP-S$iD!qm4P(_I7%hy~C|E08FWfo*_p+QyFl+jCuf#1~bX%Z2PM!#}(q
zlG;d=joL{n*ry8><hNeor(#R!v10!f3C)0`I??<{*s_9L?X|{Y`o}};t_I|KEV5%%
zSG<OBy?qWHRyLVHMMjeMN$tI5D_GvCVgB)+3WJ?F*+8@A0k?247tTMz(-AVcXFIH`
zR{Vc001F#EGvh9(aT|o4)u)6{la&~Dbo(2e(o31PUaN)9{f8Nxy#sV_%Lia5SCXl*
zVyd^mQtg6$T1e(jFMU?>c9Xpz#4H)5DFp8RlFV1)*Gxz4MCeBZKMQk(ogmAp8q2dK
zA8&+o`D`CPLajRw27E)(FF92Oi;SWn09>DYdXCfIXwbFO8Jmv|=Iad0fri{Yo^yak
zM=l;)_Ho)GUx800n;w}w%L}zgO%*<V10peyAE1#vyYm(RIm$!JSU-=cXnuMQ99X*_
zHef#pjnIhnnYdW(6DNKalb1DH`Sas@J&RDYpu?*T;wYmrJ{}o*_ngZCxqgsNQAA4E
zVj}p$O!e<(!}@gtr5aKhHGsAukQDWxMT0In&2R-729#FCxZTSp7#;}Hr~?Y~ojeR>
zCfwOQS$ieKvRdj_jVeBKRl%T?%{2;{&`>y>q^(V{fop(>^_fA}88qUB??IE}MngC|
zdRj(+Cg}2QX)&tP0CI2ab^<d|oqI)2xpq^CoK9{g`<U>C2>ujRduY2xgslW@MBJ*(
zJ1(A<yy7o#zM3P(7$W7)IXbIdhG0fAZ`T=?J`6L2fygoiQkzwLadY95x$JMb;UJ!*
zJu6d+8pt4RbWW~389~<;x8-SL+3HoUv3Jwnqm9oS6r>EPCiv;R(h_noN%($cQnhk1
zB60_tY0Gm*z(!!jEWC~`_7%IDE?^xhE<>Fc0~AG*>6xpRILm~#p6-|~M;S0Q^7~Vm
zd5ZkoTCbv$c{8ODMC*S-;v6GR)Oj3%)!@*bC2G8;7Gfui7zeJp((2vnSuN_O>99D_
zVOxDv4FQXU+z+vE>D~kB*SSd=@AXDAOLLk(#nUdN_U>dzV&as=bY<t8sZ;04mo~Et
zDcCe}S2PGJUv1v8Nc^<9)1?LVUJQd(DTSIJWP=|)Nb)C^4J$R=*Ext$0HG)Pu4E>z
zQipK6txi9>Wt6@XVGfM+&zz;Q$y?!=FL@6t-diil4+e_6{n2@Y{5PQ2fK~zV!}fdJ
zkI(R`FIF{YBL0fbmEWoq{?~FW|A#&GCa5Uw3$?qt_JHGZA2_q$n14^gP*NCmgQ!@Q
zR#e%joFtSzM;bL?mBI))%CtccZ&6NEb!^{#&s#ZW%XRTxZeibYgKX1=!h2qFk|5nF
z#tHK2a4lrIKTirutHD4t11+~14$WGhHf<bUM@tCOhyiUjEE7VOOYG82j~Gx*-Ch{N
z?PpU9&s>pgV?|}P4t^84_2ML8(Lmw?1@~K<ifp{t;CAGPYZceQD{=Wppjidy^`ZfM
zY~0b4pbyq@EmlsWDMN&b(E3e)$cFjOxv@Ow^!j%2R~>_(w$Lfk9d~BuEfDuqAH8}J
ze1dw&-Kq|<E)t;Y0^XYGQIP`%7&UMxteTYRSxA6?BQW=Cm)B$p`;rk2BAoCQ+FhFH
z6%e@;=?VgJ19@?vsBL7o2gRCqvDT}oiz-{QnqPpwd`E?VdT$<t=1RTeeYxjyOt%@5
zyk)Jw;s~Kjb4$;fITZdB$1%VyGZ>^>6Er-5UvVh>XI;pETn+7S|A{_40ddX8W9mSd
zO2wS4fg73lSh;08U!KBu4jB+W!L@2`CH^m}B0L&gi19<#$LSERhb)(;+{zH)MadXz
zInkKCiwW9$9`c{4m6?{IUH|lCQAezs5mSu@11&D`Z`rsb1FOQ11Dx-Yr(3vuPyFM{
zBuqgR4v-{6f@TE|RUinhgDdf_qwy*gF_|z<?v}dDiSepw5NY|dj;Q6|dLhTCg>sFk
zY||>2kTNsZp5!qQL<?I9^?kzAYj@M5Lg4~=1i+KH&R9qTN2!xW-C^#{4^dMLDaaL<
z?;jg*YT{VXP5tc40ghd_FBOAH$_(iae|O7$CHpy(7$kXZ>`cvkFJddwe<Dp~7<9ox
zM=di<8As`wV#lp;v=L99ISuDI@f{Lvq^fu3s|*}8ow`S(91J~6zycJsR`x~r_)pK&
zLVojq8Q(p?$Y<@z^YWCBgSkI`<+q4O#_6t!It|K5@5LVU<uje~8|j6x#)E%g6vW+I
z$G3hX@QsR}aX~-#Gc~zX>V?dyJOR^Wc+YanQNtOX3vG<vwk2NC`cfAT_Gv7i#H0fL
z)36|WV*P|e-bB}A8<D1uj{7hZV<rW}UsM=2a7Df~#l1Zm&kfAWr5S%}|K3KG2Jm&^
zx`b%?ku6Bmd@RpR+*7T;!y@yc>4$}<HHA^W??{cC+3ifw;S7n*^pAOvG3@wQa9Lh(
zoO{^ibw7PC8lgfJ2jnB3mNBoV^6}v)IP^dC#TfDDe<YzEl1a!*qV1h>5R<<$OKo<p
zVR`jT(oW-ho|y`XQeq}_!^Fals^9!mWF|>Lx_S^~CugN)#MOQ(%!?eUH$OiOO{97<
zJ4@dCWYe!(>K!)w{a5Un9SH1Ssv?B6NR>1pD9U#In;xQNH4kRHi7w%GP)6(*EK1+r
zu^VS->G>Hez>NPIt(QbpV&djG(2@)*bpbsMSO@tpTB_zLRdBX)8%n5uKeyg$<YCt%
zrEb9fby%If_et%!MkXS`Schl{xWWy1NLfj6m>YS@wO;{DI;A{txBLuGpeel8PXe=r
z);@@?FCoVzbJ1RcHbIL#@N9-1YYIgvk-+qcgR)JXXDVa2pNBdBqkoC{Y=~P<IB~)^
zUd?m%gOsO@hFk@i6Lo*Qn7*WgQM^y&!Gr<Z&#_x9R_?6S@_+wm_@4aDkHfDQvLp(T
zqMBte33$gFlg`AfBcHdzB!RpJ*ZTMQ)5P-tU3p=NSkBMuOT_w5wrOmn0sv$F>p<}G
z!gKGHW3K`CW)nuG)H1U{?<vzRCDI-^O@xS;Zk)VB2F0{ZuQ%()n>7BA*p@FkQ0FPf
z9(ot2P9;sXG~f`nphuZsjBhnydPJg`D4}3LY~Io~pPeRGClI=e+X;i&mJTWtM&xQ^
z#x*7?GK9q={ng9*$?ch=j+X;8ytKQS&$lgc2vr3b>}3UCi2PLDD%0ciGGczs`fiM#
zQ9|#qH7g!<>z&@N&^z+ToYG!Vgd7NMFARNRtj%?h{feF@r$JY6G>_wSJUQaQeyFLQ
z8gk<L@4dAh{MTPR-+I!k;2#@$*d}D6LW1l&j@qPCWknjI4m9x%E~jDPr$bUikCbU6
zOn+hz4w)E7-4!3ZJxy6NZ6*Vbhn=^`ImlcP(g;pE<Ym}AfLIr$hR3>`EXN$b=&1>g
zUU`S&IaXG$Yp-}w&h(kSE{R2Q$Za@kSpb}5`V_z}7Y_v8Gbklp|CT`_E>yT?ZPTEl
zA9yf_HE37&p*;;ww(AT{?-GNq1Z)oA4QobYABbhBR0WD8F#2I&#JkK6L;P332b1UQ
zntATyX-{R2^Dt>(eMD$Y1{r69B7yg9EhZ4T81RN@e@>tY0O(xt=g317X2*k1YHj8f
z3TVE7lFXCMt*Hg3Q*8mASP0nAD83!Ue(==$dgmP)uX2iq1hAbw>w)&0%~7X~15uyL
z_x#s(2KUFy#@BzM9PgfbCw50}0_eaBzdh^1f%JFNGZYhh8!D7VfF)U~&h`rI{|jc(
zwn#AFeM~ml(2)a8FC~@tD5%ZOf=H7M4|s(8>zco7eXT;e5vyKQW5T~nsrH)0Y-ri}
z@!fYi;+%Jvd0ccf+eLhDo7$JSw26B&wBMl}!H*&BHTK0|#Gm_qW-3C^L5QTW;3mA(
z5X9wN-L}$PY=T)`CU<uWCjA<;0N074tr7g_2YMnI9w$|I(ao;7KgoqNGac1cAR)?y
zm2!gJ9}Ku)BtqFUj;So9_8+LW-4p2Lx45&%Bc?GKdgnR%pJsFQmmD5hS+e>jrSFMI
z?a6PJ57d!|SL23T2#0&-c-|Mno0Dru?f35=__KC=1atL(7azbkrI8!3>0W#p;yMo5
z&kNC{70)mcJ9)l$nbOgm19%nouG?jc)fKh`3(4?rPb!hM!kYbtsWuK;HuI`K)vgV&
zy?KvJi55wS71@z5F2P`0D9-wk4*z8JL{CjSsza5ozoFXg<slttZ;vA3T^28!*>RY`
z@P!rrp;w1*wR??;$TOSJZS=nXT1$~&>uB^Qp1!L=UD?wC^ra#uXw6WyyGGcgjat*>
zc1FT6|G$JPRvyIc9p_8rVX;-kU##g?tg#}F91KEy#N*I->(~S{1H;<4t%#ht@+BU1
z_V#GU&{>lZ<{2_~D^@Klt02`0px4Yn8@PV+R4yx6k%G1EiTN&vdS1%!;*jYZFXr%%
zXxR*GKM)1Mb3#`Kxw;%H>Qx~g?W7WLWqolHOz>4gl3hT>uh|*I`Gz~)wJi=ZN9E+L
z{GscfnGx}_(W%kedZ$2NNnlOr>=1GBduakU7NnvhC&EqjR^}RXd|bwuIps}nPN~n*
zx%?)sFrqALl0MK416a%rQaLiApmy#_H06&KH0Ous3-@ffFw^;-zmB^l2)1#6c+q_t
zYFhmB^XN)Z0ZjO$^;4AFjG)oCT>Z$C*np)fH6a(R`ay%+=b*+XW^iKJXU&uI_PD|c
z!fkx(@W62*aIeJrg8X44Oo?E-AaxwRCLr)a6do*QP8L_Zb}l64f*cZ6j`t#5p|=!j
zMf1uGJ;50y+t2W!7O9iY?@?F%+FzsS#0@+)-yfkO!l6mwCLqC+;b=OI@Rzfr6wAf~
zOv}|sR9PC`2p00wlYcxv-6?J>Pd8I&VqwX4-Is-k-+7@@<|KhR56^sE&4ubMF_&Iz
z(g^_~vB^bt)MoHZ?#rh2H3HjRVHboC0Hx>vA1wS&sr0Ww>cv{wG+Gn=sf&@T`ZU?5
zqBX~{O8$YJJbDqyr}Yx0i<&?p`$<)f)h>zQzv2lMB(4nL({3OtK2m;V5AxaI^#QKN
zthYj&IlOcZC%El5<Z(WOhHqPM5$Y_Co-cWI8F!;0q0rY`G(R4_b`c;_u$djeudpRt
zNS&$@q;TItFAsNGnkl;S_m4Hl<zv7K=+KC9G*zvt_KgfuPCu+v`x$F`*crg*ZtnN}
z7kk(Kc}sz;iHlb;N(AnZjj||i3+y9cvQ(_39(jZm%*WWe;Iu^LJB<Pqbna$eIah9@
z6TR{&l(HPxwP}YlOrLUzR{!;;m|j$R;?WWvwib^s)kiC$z!ldEH$_Gn2v9k-aQpwq
zXg~Qsnyx#Z>hJ$I>`nIGdna46H<3+*tjwG2qU<f}+E;es8uywZS=T1yij25MBI_Du
z{!X9o<M;l%N4dG@eO|BU`~n-Yv6I~iS$Wa5rmdqQ_DI-BX+RGUy75#msNR_0WW&kG
zawge%apz-ISu@O(n2KSbyZ_v?{gq5;XC?qd&l*+qMi2)QdNR{sBF{iUI^5oSK1lo=
z*7X3?d**UcIzEnJA8c9;GM}^zlumf%@K$mVqgpF<-({ajrFNa`cPmCPp_i#GY;`xR
zm?xMsr=~EqG+^cA4N%M0{eD4#SX&zQ?eL_h&9f`Pl+0V6RrP)Q$zJF6#17wfKZs2`
zt3)cr=5Fs5d-20(BGRY<Mj;BfM#NLe`zN1-EQWHN+L5%oQ53dCrk=bK<mNj4Z7Hi4
z?Ca&mDuIRT0f^_O6~zIB?0^8vB3=`OW-y4!Sy@prB%T$-<|7k^yFWC)+*=IaH-((u
zXKJ~|<qO{rB&@qwJIki)DjUaVWMissq9P#F)wbPto$H{`n*_nwGtl7QxzcZllBxVq
z(d=|Ezs{&JPFhJ;5vG7pf!M?l5m+dcw;lnTn|gDDNfAvbo$v={pgVHHe}z5P8S7s7
zq4A{hb``a-LB||G`I7aK2#ExTw@NQX1zseLg0?GX?9;&7^a#Dk%wg12bCy;1x455c
zxwRv1q9i7wsfVt#iHim3g({p&{kuUAy0VC*L7(wgKUbH*Ov_EeFkc0rj6dAiN>IS!
zZQs@?X;uFC0YHQn(vq2RJQ2@Xj37l|QdY0rzSswF9rdlhpH(LEbND)!Cm7b@)$jiS
z|Gf`x(8UT=A2)q?4l}OMK}+BXVcPcF>^Fv2DlwR%cYZ{-lrP$@V<tXby-KQ90%{O>
z9J``x(d8GX7F$Ix-w4y^t<dZ~zT&w0v5@*sBjPRiEJ^x0Ko17A@s5+ZN$dG+%iE>p
zt1Ww?Kds{f-`I)fTD&29le%}@_D>t%)5^bw^0bP-!N^9yQF}MV0#rXZzUm;G9mhVG
z*?dHYZ_8Ri&$fS)N`ODTl}y=|-sJE8J$S&I&_rI8#<vG0xQpryi@`4#U6FB*ac6n`
zrGj6u(gr^!Sjc4p^tT~ju;b)n``k7UkUJR7dKl7CeC9j3hn4RQn&>Qdev8_9nhIiy
zKpsA`GeCbPE^Yt2Tf{`+tJ8-)lt{tb<R`dp0vvs5yM`s=OQlil#zZu?RCR`jsk9&7
za%plcE!`@Y;PcP%INZ$j3%P+=%?Wu4YyGK{_2ZXj3YTwvJ$Pp<6v}{9(<<R)QV{k<
z0K8k1eVz;8eJT`7ns@V7m1A0Oa=gQg?EU6xhw2Ovk_t3jE(V^TNI5<V<GWLD?`F%H
zNlVj{w;L%}q?o(zu0Y9QhE&S=#ueY>XDjmws08G!?=^f=)v+ZQ6&f)KDj<I<xUpE9
z#5WmHYS!DYGHfminzM~n5#nTGcch8p8;r%hZGVhSr^7jEjZ~5pzHL2A$tKN?6{y>L
z5&Wt}?IZ2+YJ~msR&i!gl0@E=DZcsQ@^=RZp<!O2uVOXOd8QBd+NP4mUS8Bn3UYDq
z-2Q;2<!Z^VOD1yp+*y<DIuH~RWqjbsO4s;ZRmc}*`Rh#_gI~r}{h|QDXIFBlFylmP
zvoNC`<^f^9)(KuxLX`6(KUXtk4GUe2fGR(f*SFP;f|F>kR28{?`xMPpTv@JhBU1nH
zq_rX&HPvrS$uiyI8iPR;e@RasEAYE~+F3j$?(t(B!J8JLQZbIRia+E(`okM`pun^t
zk79iP^O@3PBQ9Wgu4DOR5i@spgfcrB^TOGedxx6Q<F}LjG<!A}q}|<NKdD3NGErEx
z`&g&*=sjVnonuU>i4|KaF$jpJ!(&L~g&`G&dS%+OwHSmjtx7Truo)68Kf5eGlGl6f
zF@2dWtIKZ#*Z<h{LH6km1m#30pHT1bV;2}zMD6pJBiYv*ewRwaGPY+uVh0=B;gs#x
z`*(b<n8lX=@{uc5w6_kSelP!6-Ov6LyL@6HxBH3&<EW+K#ez@Vo$PK<i#huMnKZk&
zw7^>^kT!m8h&wI+mOscnvuJ(P+y4#-WAa$?^_myw$J}ek*UWydoYl&X)_}o&r)<1b
zUiqb|`2pu|dkS9^y;IYS&Q@)mJD<?VvHQ@_6ha)bNCx5ZLzcFEF=8A3Icu^%rJ3@W
zxJ~bopP>w>iOakfn*quY<d%)(lm}tkuL8Df&Nz)bnje4pV`(n<+IFBS$7VXF^cZJy
zIH)X10s>T1Q{thbmA?NJf+LAW|Il9F7r;p|!z?0naXqk<kNO2&-+bJ4Z|8tzLUAE<
zd&$NxuzfKg)Xm0HRyscI7UA0{`hL}01?o>U5NVv0Jd}SUe%v;Wh^oIJq~>1jqfm&K
z$K&SIPpTaJ-S~ZRxAxbsk67Q0zy%EBpZX<*qR?&VBBfc@rI1;R>88@=4szL{?y_d{
zT2U!^zZ)0o7UadX>1*XD9>{D)iVmqrUU?3aWaQSSq>WJMjDwU=5HM~tR8IDi)RC_?
zddgWdZY1~DNXmt#J#A-0ezGR%gg-ws9_x^NSxVOL9jV?RH|54c`*ul<max`{Nzm!Q
zt{xU|#zFhk8uBv&L!k=&=|1;7eDeP2Bbv`0;7kEVb^TD*VG~lkA+Xr{Q5<c?5PJ#=
zM>xi&2|L0qld#n#bwBE}IOE*Fg1{Q&Ot1Xn@ls*v_y_h9^f+j@c)pIdo1ix7xzreW
zv8C-Fl-$e;;-}D3o@)~4r8v&_rauz))SGt%S_4BEa-OsNvz(EaVgvN$fMJpH2`Xpl
z!^55K27STaWE}w9CXBjdy*rMu)Wv(%zu2}}<|bDlWBJ(%pK5o)+ImlX9K<SR%v2@~
zw~M{{RDua?E12<IqqM(oDW2d|KrRP2QC*CYL}pAqY=%y2jP0>}9-4Z{*!5ge!32zq
zl;gpgxxBjXDwzW-R)XmK6S<!bc1niXUeNB4Ia-q=pT%c>N$u5~XtZnF6VidY6_BTi
zUrO}=Adyu-Ndrl>8qJ<c-D_H;K>B0lHcJaRy=z_=73yv_Jk;dJt0DfZ68}t;#Lud?
z8dzq$_*H=Q&HLEjJt?#r>(uBqROX#du$fLMd93loZOBkrw!lzVhudZtjg)zq1be<x
zVSR`;H!Kk{)Wz0w8xpIq>Yd=JEU;X2JfBvO2hY`$8fVH*OGw982c-_dLl)*sQDddx
zt)jahk0<JVE;lbi?a!GyLe8ouUfKGD6e|B^3Qt@9_Gh`?n^XkX$8(J&J*ISBuyRpN
zETL6AX>OFW*1+RauxUfCkhfTm^^XIp1{Gn3iad18&l0rmeu}9_z&W86&Um-614yTV
z#gx?i9qx_@D~oTk^wK5{fqcW+8;BTF<TQ!UaSsYN0UCbE1e)vtqNY4_{!GmGj2N_R
z*^hVwv`R^k$6dZz8YJ3@S&CO(<1u6Z&JooKKASB`r<<pxp1BB|H2*5s+;c7Ox80FC
zbMHE&XP4c@oPlxsp~+jVbd?(Wkjtq_@ElB5w)_41Ggw^y<nC-iRzfH;CfW8;bBbNf
z+@MsiHpLL$OZrnoE}c1RgyL$@<QE&gx!r*y$No#_k!XhwYyQq4T#V8}-o1s~nFaf=
znY_X-aN@-7>8W~U?N&vyTuUFHb!OU_Nr?Uj9J+y(>qPWjfnL?Gp>fgg8tP>TQgk_R
zVB$+nSpyxP(u&L1JYsRvA`0#L#DfpvZJr{58>Fgu#RA40GBwYX$kB-DdEl?kwJ0|J
zh5@?*qyog}eA|7sFZWZn$k{3Qx@NVN07AFYcC2m<L3m0m`T1Y)6ZM4}IV(0ncF6D0
zdL;@eMmI2L-{MhqIcoHLPqJWx)~|*#O^290a1M5UR1eeqTCqNk&>(0<FfQ)(7vtbQ
z&>1jPUiAx=!&D*;!GRXS*IQDkqA#ap3Jwz4=y}92urdW-V1}v_pFGznqmGZz`7LMF
z4C>;i5?XV^e`1>ji12h$0wW@w!K}+WO(pP|z-T6>z*c}dHX8F@_!~=K9~Cd!p`bU8
zJ#H3wHcTMJY<Ab*FXJ+tdmE*LnCW6Aki(P$Sn(10)@Y^k(x}rEJ#QnALpPovq6sAt
zZT~2Wl!@PD^py$D9=TsdZmBpc=0bN!RhgrxRyzN@xW9{N*r~m0&IWLhxtQa&TG8Zd
zY<EMcdX8rhtFCqdmhD5wfrb?DzxQsA*eR10sqn7)GI-OU8xQ{7)=E%7<{_G1p4t^9
zM*gv-^RD0lzfDk)QsOUFy|9VD`(eE8`@gTyHCsdaM634pIiJ3A-^9bj+jebOqrYE?
z`xfz!emUJgUcV+g{#%6GBOe@2fz0(t$reeDwCr&a<1|l4_m(tk<6I6(c(58EgV>_J
zPftbpKjO9;p^+T}5oOe*tn|-DQWvf@mOfo-1kf(4{>b6^-hn*seF=2&^m47Qv1FUL
z+f}RWVn=gqeX@HXwXaP9KKDz*NcKrI!5sy6?zX!g+FafIlV4GjXB75n?n5^Da#^<v
zk0GqSc#|ErWm7AXH4Le=B2E6FpIEW6d8I5V$2`VjcAD>_5)}`tW~l!Q@Vu1V!dXg5
z<qnO^(Qs;0s%{IL7q25lM^kdEyZLrW>1%TW4NyUrhz5p|%Ce0N%nedRqV-MLgGHTP
zZmkgW!(AQzPlv3+93|aBy>rHTPO1#^<;Ir#8);=l+MEKcnh*K}g-f8xDg-){){<~~
zp1jL2{jz(;(C3m3o*M~5rBS8My`TnadwcHgs%VsCC7hXzPQ;_i#fRdqu2kSAX{5TO
zK;(^=CX7@x3F}G6nH3rqzP3%a3K<<APrBz|EJh$3B(#!7rt)3`k<(@x5m|DHe@D1=
zy<5-w#eN50H*EpRpn_93Tqn3YoSh2Rq?oDW&LJyAY7%^AP2oOtCEkDjWAO1;vhOt|
zxQ;qrulNtB(wo3UK0Eqs#GKU+Lha{Va6e!)B|jri0U1-rN?SlHlI%}jI6&5|Spiu>
zcna6SD;rTrX!K}HZ%H{&;y)8~(#d3J`5o4JhmQv9RtSf!{IVYIkQfylDUnvJhsD-(
zMMVa?g?*CJO`GzouP1mCp7R9mN)c<)I9<Y%kpiJEq2C?%c!pu%Eq+}~XQO+kpgYK|
zP>on+(vlYLY{GyS&zVXSPxY17a3oJX_T>WGN5!WmOuWt%+3?tshHqx^dQOpU*R7uw
z;3T@bn(@3g8bbQjzK^UL$+nr}%}MU`oA0e-kL=XTaMCEgq7F)=V&`Kz4zopkF4%|B
zCeka|9TjF&zj1_lguAn>XwPON1&7H~uIY}CPy|f*8f$UGBiBXbyqVci3p@i|>gw_r
z4^sn&-pbTnM`njzTs4LFJeT83L}_!h%J`^O?N6CB#(>xEuNKcuN+zq;ljQ<wZOUCr
zR{efq#*u>_-#}pKnwI5z|2XFi>8$3(ZpXU+Nak8FJHh8ZgrxWE?Oyw3*J0slc<%Kn
z{xvOMsPb!KcQb;Z9F3Re&FZ>}GE3~n=_<l)0rTP_EY^h9O!D}WA`F!YBinMJz4c)p
zc`><gyP&WnY=N1`?D(_Yai-?OIF-XemW6K*uWXuW1#o;k%Zy$G{W}h&P$7f=*8;qF
ziTK~yuqLvzpNK1CHzw6G8?j>0TzVv8HR048I@!DM1vw>AZR0)u43P2Q57Cz=fNM#=
zyE|A44GCpMAw~tMIL#wWbsMR^qBEWiS&$P?X(MJ5&DUMi71Q|@wML-F!m8B|iCWx*
zV-vxqZaG42RYqa*C6Py0C<*JiPT4>^`u*+Wza<4fR7%V#Z^j5=qJD>sA{t(9H<pN}
zO!w61dQgc}Sr3#n*)miDL#s>qI?Ropd!^t-haxyVfN2s=_gqgMf9yMjH+i8KNVkxG
zk0gVA?U<7xVN5@6X|Rzqy*_uq^Ed(nnUf+h84EDN<`&+EN6*Uzgi@Z;(iZc)Di8J&
z=?~Jf!aq1XE8D0cB102LMkT{GhGk#<{No1S?99jm>0ud?nF#ngWv;m_T|!6MYxmy*
zjFINW+&hhKLtI0J8uYPAS?}l~DH4Wdd391tkN2K#9|?*1FRjjo$HmWltyno1BhAT*
z2~7m9)_U}vZHiYwbXFq4t?LV+$J}({FpGdI_Q=v5!mCu{hb*xuSvxsJ9(Qhr5Cbok
z@0awO?;t7eUXb#4?|bpy`qnR|e%sQZXX_ibtD)CS=ymSx?(q#Z$F?or_ugp`eMz=x
zBejq(!z^MWDdJoGQfXVswb~8jpGPH@;%u&LG;UmVtkjSg-&u+^9yBfi6SUpqa0vRb
zlTxK~J5_AtH|01rsnJNAMwJ@+vb<7MI%*<t6P6V-Dd@WY{S7|Ft-qGl<Ko6^#=~Wj
z<f&GK-3|^xUP?|F+*R-Yj?6&f&#6F;GE}=Oa(TiDalDx19bO-Ldb6jg<9tT!06wy8
zG-(Y$<Jv+qa^i_hYIT!+l*=%kC<*IF_pUHE-}{*rDH(Cw)Bo-<2;)uGVnt-RiVnYx
zxWSlx``_6j%0M<g3B9X9p9_+`t6Y2ay-tnA7{@OP0LBy+NG@mV`oq}b25iJMd|5z8
zoH-CoKA`c(D(enOjF_luz>=RWy<q}W!>KjE+ArJzjDM^FQKQd#tr&O_HL)IvzFibT
zLs?DPBZt3#f(Nw7Ee89)d>Wtm<2X5KQ6=dZcKUS8+y1V$ROH&^AMXz|p|>??bu_Cs
zl9Q9F#U@p+WdY>V6q(L6P5Y8x&<$P*g_#n@xKlj!a0=QhTl;0*<E^Y2Pf_(Q>WN9T
zu&@q&ajk*d8qL(5>K{Az(0Ab^29GI;H72cVjTB|Z1r7B^#&n|7BgXW(CLm}1x>L0U
zquDVK-+ShIwT0+1Yfx~vUQ%x9oDJ+a4O9oa`b?;J7opij?<+mW0<bRt8=*xJ@`2q1
zr)aW?_t>-MbBSK{G`Ar;<>CeyA^^PYOQqHdSVU}g$C#e&&5hEucgBs`-i5PxF(6FD
zNH`6X%PMidD;d;qkB0)b-Lka<w!bx&<nV#vKFx@)>yodl-HM_uWB*r_F_RXn-l1$o
zw(RcMvyT0>G|ie4iv1t25t7VeHqH2X9}cKdD{0);hYCIy+}9gEmq+*o_ihFqhGSf1
z#Zyr70hl;Qdc91N08=6zM@5Mz%U(g|Q(Z0sM;B$NaK5kp2nylJqOom7&qhQa|3wx(
z6^HJ&{tzF<HR;<ZwPyV;^NE#}VZx7RJ}kh6v~E1io@KcyVfaWIs+cn9*xy~U&ZI@N
z+>lk8ss4eYsR8y(5ajfv41ar=FcnqO8phcsY1oFKO_=o7_wY9K7$1g?DLRUo6GZ$B
z`{%L}Dmd-v%*y!Xx*Vw$y>ikRk~6Vgz-OKwMm67Bm-GmsO(oE6sN7?gpumT#pfrXF
zRfFxzXMA>HEwmvy#pjrGxckIfDR+y>(Vz7)Wj?0vG^4h$ytK3wlM$sI%e{PxG3`3;
z>iCN3`a7SBjo!ZAj6GvHHJtdP>xY87N1k>+Nvo}WWrlHM8TAp96A}(zh5Ost>kV=C
z@39I0MM1i=`HHzpn_%+K^%G*If;6rA%8f@my#HSO4j;{0EhJZ@b=IxD!d^p>%9qMh
ziJag=jtfRYS7@y}*!kQ|tPF`pY8pBk(v;GA+sS}p3+~=Oe~>=0OFn+d0Qa|5sFt&A
zE}#}kCM<_kmUEbVsbL<;I?xx4^dsZp0d1q0GEOwe>7-+W4g~x+b*2AzvnXC#9?TWq
zZ%AjR+aQFT-gGb>_JvS4$iykXAn8ul8IwuT12biN>bgdA0MCx)L`cQKXB`JQ5t8=m
zf@O!2xTqT(Iro6dvwL-`Ji$_j7KPL8a*y}^+gnxd`i*DE;Hsr?2n;h-QgIK=cCrbR
zhdP^uSXLchaee{tJ_6(yM`Fa?AZrF%uR*S9)rXA^4mlbT^Pk1xL9e)!ENp)TK@t9@
zsOiLYKROh^36$gHEHz$|WIRl1nS3-VG*;{tU!blNf7>Ow7?JYZ7CDppZIAEB(ZGpx
zp!%Rko%k%if?qGe{EpCD)pzQHAz@BR>#IMly4xCx3A`s|Yg6td#i|omWZ~ASL|LXC
zE81z0%XYjjc~j`n=ixy(Yp7)9In=ZfU<|0@jGkNR%e&Wr-VTe{>%1yqQZU*ZxRN~B
zR*q5>NK^g6%#WK!t)80YBEpo2eAo1Oet#aSO5xF6b7DCQB)K`y3QYO?bbb97^33=-
zf?nNq%8WTRs$dyGLrkJPm%>V9t4ZLH9a1N%OMcz{J48CZDR$hDAVb|=WB0jHo&gPH
z$eu#d@hDSofSBwDEdPQhnPWfL(9I*PyjqogWJRp)82H`~*ol$j%uRE)=HSP#HXZYw
zw~=Ry>=l_W*f)|;F-WJBy_0ZYzMJ{B+lEOX9=b#B_=3@xTF+ZwFeIK6dTTqvLopRF
z(OiF<`w^gURzHH?PRe2Whhj>$gPPM<!VVVj6v^UUqr-KM$FFZ^eBTUHlA;PSynlWt
zXT<(9nb^C^7Q3^}{BB%8CIgkzs`laB?}Ef$ty#ne@WE#~{A>|v6Ms&XkwcX&QeGat
zcnxr4PS2@|v>(1VH89`U@Q@m1Jj<a+;T;Rz1lo0%%kK}V^WoEp)MnfD(mA5fLO*@p
zRw@v-A)~wZJ(qWugKIV@^+w}s;}7DWR}d?J!YWh>HRzP^YL%<5O4{rZFP2Pk-tO&u
zOVJT%yGl?#9vZ2&9lFO=Tj$8^U0~)$xw`icsF1%u_u`h%z(7Q@z7liqM~NzW-aJ!$
z3#UnWLGUx7Y*}5!9=|wxjSBhXZvtubv>f?*m#bfCTdtbKBuWuJMGy6&f3-Zwg;%xF
z;v;6s*lXobrBCpSDyzImH;J@d&0_YEC>fQYkXw0)sD-E^dF0o(1F%a6iX}2Asr687
zZY{+^863No7LwE6_MUVg5pOeM@Rdm;H42f2dk|b*nRW{#8C)S&PV}GGyR|9qqMGdR
zjC`Gl?P}isCg<E8bvjEc^G_*(aJaCWzpSWpW1(kbUXV6-<7iXVy)hw6oR6~<>XN%r
zCbe@yla?vW()~EGq=5quG%W^u>!3;q^Km0BIien%{Rr`52457<jqa`7{s`ohqyTBw
z|9nDgX{sYX%#}qMYa%}!6HwZD^z<$-UJC4PDP8?+D&j&W@`ZK%gx~$sa*eljdM(FJ
z5Mjk_*__f4V!o4MLb3hhVH#@*z6`IBmj0ymNsNRYLDFpjHK_=D4Gu7lM_N+p^#8W%
z-4(yZwt#|HTTXE8j@i`2hSeBzKJ90ugIRPX*w9x}Dv8JnF{NZ1P{e278$ofDJu!(H
zgesDi3uKQ6la1>F^!eLpDPP~>fYpeUOa+Z=R&m3XLye}B5=fXi-4HqN{TctBmWB9G
z)=I9+BP4CD>6jXf-QLH({whYn8wM0@PCRY0Q#C-Q+q)Kzpxq0W-Cz`eg%*f8f<4ej
zIM7Ap`%k`A-hxwRQ5$yI5Sc?!vWkDV117;Ah8&ixi)1nvQV}z$arLmheiWX)B6e9>
zQ*?RNwj8>RRMx?khT}FgX4YNtM}{8++WiVNkJQr-v3PIjqRnoWfSf|)w#Ukyf|Xj{
z@Xx&QkOChM7@a#U5mG5*u#GAHm62FIvQ5%nOXOUO8xOxlgH?}w0CXbNCFmC0l@?1^
zV&~Cqhv|7WWit}Sp<`rBlfJT(VinxUo!Jb?SM*$xolNMYXt^CYU2+7`hdJ(EIZ%=B
za~U33Bx2on@*x;NG$m9C1hM<?w7KHxBHEELEiDMXQ6?BgbklUAoFGZ3Y^Hg5yX4>h
zkiGkP|D6}0o3;R5K5#m{jiyRBHOfa~=C<@GTYBD_F7@2`Y2Z3^G0<bwS@r94D96E=
z`k7lht`A{;pl}9X@cVw0@o81^p|71^&Wf6n-5s~K-kjUhI%+Io6B-5!jYz3_gBuBm
z+Hj!96u8>qX8-8!0hwysxxA^|tg@CgHE>szj{I4u56pbQrV&sfAG%lOcKYeD58n9!
zXcRBrky?*8YvrnAqi6rZ{5*wEr;&SRrK2+r9nu_1X@RC{K6GnZc?%L6{5;TaHDL3v
zjmb89Bd*^|bu_>la<$Y=>eDp3hdO;1?4`J?`datDWq^O*I^awq^zvsLLoFJZAq9TC
zcasycu4pQK&mvQucWS8S?JG6~z8<`wvC8^>XO@I^gkG>Y6X;iH5Rc=rj_LB5ldq8h
z`)**m^7wK%R_jmT@!o+}tQJ7fN)}t<moFUD3zQXM$##A7lB;NLf{e6~!+Y88@eqok
zFS4EQS#(6ZV@`IO6dkRjOAtMTXhhsOvEE$sQ(cHD{q7jaazn;*NRF@1(z~a=1KJ!`
zwFL(U!*bxu+vH!(S-&5sqdFJ=_O13%`v8q~bW)dHi7#1gEo@d1x&Vt&Z8g=yOO-QC
zM4L3FEM8$X3^+hJI>MhC&SSoU{S`FlYb}*z@^^uGcT5rZ8&?}8Z5vtPk>CIJh&ynB
z22ibSW-6eqZ1%O;{5Ocu#*{S4nr|vj?vSn0x!h77lbz|1)cxGsjaUz2xUT@mnxgLN
zs~yZ&dk`*a>Hk~e2Qt8faK8$S7U1YgLM?Sr@tQi_?KARw*PpNZ$mzO<t_s6-rQ4Hd
z2(igV*ejv%^vET%lMS%f%bs1>KiQwLf8G3YcMlu&x^lbdT8}xqJH8z~&EJEEt&x5@
z$u%8mo_;L6S|k3C`99_ZBy5@qGp&d-3wk!0v;ohWXc%3r5hwm=JnTKTLD(ZRc32i|
zVjz~jbL0q3-aKh^xvnCe`QXni%$WJwdArI$&8v6g!|4-d5<#8G5L2qvS(uBpLPeG{
z@|53>W0I5lfL6qWVBbN+j7iJmve)0-OLfIhj8@%55-yCKZ^#Y2<@4`T^W_@;`2FV_
z`|C@UQJ31>S<$ucx2kAK?j%z?9JsvlZRr2C{xv$-OGO*Z(i(RbUKx1^TN6^ikn@N%
z*QE4)w44CRDRAoebx5W>B(Nq%MNPsjG(j$PNmE})>pl42&R(i&sUAj7#qCS}%=#*`
zx#_8Q<xlLGHkxkM$qCpv@&K50=LqwVTLRvu1fSc{$b@5RB#tmMmM=&^by^ik1GJ`+
zBQUUlUy$i}*T=(Ba$e&IfOL=xe-|IVmSP2Gcwb1%t&Dp+*Cm|dF5G%|RPE|RW`a7=
z1C#uKrqAI=ccWhu_tNhvfusr_Ryu;|hf*=g`hE_XzRsET$}g|t3aMBHaAYcJ#$#=N
zni7Y^uRU35wQ2l{H>Lf=)Au=<)6iwK3P&*5%61WilY}Y0bezQ;3p;IQD)~6i(hig1
zZ0n)-TZZOCG_MP0A64*#Ua|)*=b=5<IFxOORGd@`A}Bys573pOEq5oFpy+6-dpvTM
zQVVuceew$r;XHXwQVu!%z<hPq=;-r)?sqKng1eV{5|?>|I*_Q`NlA>EMwHXx-#qT#
zKp`6Fftn%a1C1@Crmbb7z_ST!+L9;FMZk@Nk@BqJ=V(<Xj}D3C4qf%9@P252(Z;2|
z&(}WSy+L$`m+}N}A`l&GqO36)<+*BisI6_TDy~bjU-R(0<wT32-maK6Zx%=TErEB<
zF!gArQy~uq8@5&KlNPp&t+&DVJ$n9<skS{SWkY^Xgi&Ju_ZnuUhGP~n5rHYH@pb99
z-n^5HL4H{GrL1MiP08szaT>KdZ{Yi6e*@x>V~)42ba^E@S#)VNI&mGn2}KhA*sZ)~
z#;>~~%nK#1iM7-Q!fmq{#4Ga~23b!h=_|@PjMI@*D}?#*?O+m4eR!3biFXO^bVa}0
zka0j-;p&-QQHD?aD_PN-695phtNGcVFdsvT0Smr1ks}Xaz0Bz|2_9*oLV)x9u7#o{
zrzIWpM{uX<2(1(dFC&^v(a6(p20A6&m)g&HdxG5yxn@<^y_2|1i1r*zngDau%bhtY
z!l>sgya*5f@1*4KKKg`l(yHNRO8x!7h7J{xs@IEXCLc&WuYNg7{`^=Pl_5Cft-Ol<
z`rNz`#DRQJtre}&gfqrc)#u_j-~+Mv#!}7tl(<7XU4xtFAm&hodz@3F3};zZ(CRCc
zn@H7%8%KwG@>U0AR4CDqCqKBagD|*93rl;`ma!cCKsHKaS&)eIhz!oPGn6i3aTe^%
zwFV^L$u&zX=IjPWpusdaIZeWjcc)Ex8+W{4_*`c@`n(md22_iP!sSoGw7@ybB{>0b
zVlnz@W8ALM<wa4(-NQXIkc9LIXQsYBJ+Nn<W1+!|&+9$g6v>=zvkpwr64fLP)Fg*F
z#+;PZ;V#L58X{@K{v1dcmiL~pK5KAnsgn?08z3_=U=<SF%*~8ZBzD}=VX<wq|Je0_
z!_C%rD|6h0G1g|WpwarFjh4OMp=E|fx;k+f7JV!A9ostj03Vxc_}eHZAs$%~Qoi#3
z;8<D*rw#1+E5=P(YkEWC)P6|;8^e)@1?G*=-*%(p{FvW`tW(yGlsL9LHiu?!o_xS9
zn)~HcfB}4L=ChQ9m}B9M&WwofkJOaIGM9fVR{zB7cb|Lf)(<!mzC_FAFK!%@39cvx
zu%MK=MZUeT990#@O#|3u`map|M-Spo%eO!IlzR*4UVd3Cqq?=Jc*0yc5ghbm-U=A@
zK!}}%=GL8j3*@|SD7rR9*(vkWg=-K0Op*!6&PevxraVkmnEi)v4GhrjDanlMhV9jt
zxQU=b6pIH(<N=90{bBuy!`yF`MeFpRa~@0L#sGjzimZ0hIVlWhZn{d%y`Hk~Ubq^*
zlj=2fuQvRuvh5UIgv+7*VsOEykc*Ni^>dLQ2=3*AMp(u!)7XAW?^KE6)4#=lKZ2$r
z#JBrVsHNN8-Z|0P_}OMtVPW=r0kcT3AD63$8F|a|x|B!Y7giAC{t+G0NU?L`^y{4C
zdwAvtS6!_vlG8gfm83r(!Y&H<a60(fPK4I%{|1+Yc8wS!*N}hj?|ENRbcL-9a6Aro
zd*fZR+32RWtXi=(kgkDUSs4Den6a?g%STsYoZi?tYY|=CiO}<4*((CXVPgTUms=vu
zmv3()6{>^*>6JY$P!{r_j3=0T=tW;6Q6jR$e^fBlr1(N0$rE2|)q`nkSOlN1+`-!F
zeiO?bf$n*~zNc`UQ(Dq|t3GdKUPDLE@2>+v3#{eD3c|sM0+16^M%E@GU}4Anl6B6+
zi#>=O9My0FXA1HGG^!}g9yx)i6wEC7^VWjc_Qbqd6hk~MhqR+Zyy>^!Y?YvaCH%W?
z&i(i#lYXJ<(bY(eks-mv;5o)6We<Lv5$g|aQrcTQ=_uJH1-PsLDdV4>{+lL5HN;~R
zX(SW9vem(x?^iFU6HN6EwGWou`;U?7VT=cLorrGl)wfX%?Y+(;WA2eRmKH{P+?E~<
z4&Xcx=ouR;L+A_XDVwh~lzFl&g-D0w37g2Sv?+LKb3aY%`$*40@~RT#xd`X)f0~ZU
zR}fSE^h-94U_~Gab4m`~Ik{(n+PXWyjbqEqt8Oy&kqRT6Oj(xpR4FyciK|<_6=urp
z8WW0*DShue_t{&_QN)kixbM`~T1%iC@t!|FW?r7rY3_G_O=ZIL)N@e5Mvv#3{P><;
zIHvUplJ1e5tesS5^`D3k9$nTZkyrl+H?XKb+h*Vu`y^FiOdm_aRUL9HKbb@|rUQ^s
zMf)EjHsE*NtDd18pvEll2Kt3wD_}hN@V52#P-BpL;Et8MtgUn-nJ5`d#3oeyuWvY|
zG|v5`&*!f#2_4JbYMe}(EX5vj5^RboX#8WX0d4(cPQKQFF$H?u8(%h18oAMg+3ipX
zndoR(u%{QTQwm+2(tSvuY65fkc*bDyDfPJMZ`V(Y)1d3X@MN*a3O~d<AM7=;eXiDe
zB(QCzuAHS$oS4aB#}OuvLCd-{IfUobD<}N<)fwT=LX<%10ChpqaYX;90gfv(>MU0@
z+JSRKB*;qL=AEzSv06M9)BjBMbAXtmP%3|NX%iW|r*jbW4Hdc@IH?7Go}PVz=}t6@
zuYDO1D>IX>-$?T5Pc}mbwm<fI@%H7$<NM0N*%P+sxL2DkS3={W7wlFX@y<Cv=LWe>
z@Ky?-UO%ma>~KuM`vOd2vHvD9ZKS6#{s`2{t14M-7$28B0S>0!vW<<#l7{_Z*<enK
zW+{%5ujg;eKVB9%ey|fBa*DrU_gJr0soK3z71bTTeWxD%i1E)+K&p)NRBTse69*ON
z*^TH%X{e5~PUaejC!8nhxXDKA>k(g4Bsirlq017fEp9jQXEY#e7s9kDAxbcDqrEd{
z@sopjtbT2Ye|P8WcWgq;u#f%=mQsxYJd({RkPmai_bg_oyvxpTqtagVi+cepvGca9
zr93xGX7;Y#qINuM$BAR~F>KqFCb$f^g(rpZ>#!;>#<==JXb<M)ndw9ZH1wPpcf}Y3
zLI|^G?~0~8f+q5e%R92zc7Z!uE83y7aWmzs7^5Dmz{wdUD@d{EZIsM}^CWU35e}C-
z+a2>uTBd|Dh?bmriSM0pz2y$2bSp`f0VyOof*L$FA=ZeQa0=2Dh%6)dcOCr_)lb{M
z-Ti7mZ@>9I4fvHjcbcVv8_rD6OF9E>AM5=+G9R~c2H<Y1Y*L#<Rhr2|rr7L+dHHc#
z5u`IkP@;6)H-%Sv$_A|p)xhnsvezk|<X*UWQd7a^VaL$JLO5OSoWN(Y68vp1w&U9@
z`0w9#U%VUM(?$SGa{4yNCGFSS(Ezrx=I58<cgcGOw}I-%TN$vsN{~{SIGSG5yHz`W
zxON!DL{$BZp1u6tKobTKInEV&#cQZchAWMfmu*(|)F}rEdst|%v%zFg60uB6Ff}ig
zD9P$B@;`{!Nj1mvgL{%|IK7hf_nNq}Cr<Yo(6zi#hI+?IR_{Dnz_s(sx~OvixfRXc
z_Qy*4+`{r`@9=6ReMD0-Xa~`p>k$w_=~})LG>4c=9W;MMjh5HN<Yv-x;QDyG*fMgz
zJPwOzr?+FX%}XXt(+nCYX$CoSQ$YpD4(anlFp0GmVk_D-5vOth6rY~*h5Rh~)<=Eo
zQ2X#5Zm>1OnKfcM6^vw-jU;z8gn3|3xN~P?3d52TxPvAdC7DHfO$|!8d)2Y1x~pTp
zzQ?SW!7C7#?(n1I>j-;>ORp*T9>?$i$+TGwb@!x#4>Q0Dde_#mC;uBlVgIlIEhxqe
z(ld!5ZJ)4?fk5!Lb5{))9yX>Rj>*3}v++%*gY~A+%58RWP*tLD+Vg21SOG%k>lNZA
ziGbr&+&jyaN1$Nk5tGs<KN(l9(8h-bTpDHE<!==66zSOz22D2v)M#hR082rdMho0~
zsyfd#=ojmHwa;8wjhQCI)3tfpBu&Um*<qS`Z=f7n_slt*{9EcEvYPRzKkSBd29^)x
z_i~$PqkgQ)Zp<{1@I?03xskGh*R|3t%{(RGOpWu5zq2z!lb!EQ?~oVy$A*g0m**)=
z$7o>8BMP*1-e6fdNHFTti98J{t+KkZLO>7TKrtOM(3BCe<rv=tdzy7u$uLswsL^d4
zn5woRKp-Ph+6-+^{6P{(X4if~{<o)HS*uR|<!WfH&lS5ph&5U`3U#AgDG~3nv`*#B
zjBp0**&|g>b&6w`^3pA<+S>J(m+#k}`ae39YN@;jqW<`tM_gzdX33p7P{HH6y~jGq
zx0mKe=SJ`42M>y|^5)gC&<Hce(mec@pfh7KE8^@SU~~SZDJ;~)sbk<wtCmp!$6}y3
z=y#dtoVFiuc=SdWH+Rc*<qp`=KLflI*nr9;2`mX$H@i+d$Q^&XuqsVjifHgVA@6$%
zKhBq0OM8&OZep`Jw>@CXe4mL433Hb|6j$-Jrbicrcqd~4mkHE!-#fF#v}=+prFR#8
zCJpFtsdy=xov+|5q48SI7XO7Ek*t2<Ua}m}nS5@}2!*dFA2)jJ?*60j=*6pgqAXdf
zT>Boswp0@a7##j{5lU7-l{>N!88Tt2lAY_=u)nNBc3)^KPj<AW0l89Jmz&$;K6Z0A
zhQ6uAu%1P$kz%j7s4`Dxv+s%tkKKfqo3gRdm#k^jBvk$4#s$<daJwk1ugB8!-(_o!
zh@h1(!x^$~PNBtemk&4B)AOdv5>TzK2__S21aHzKE*&BsFX9V(jPx<!6mvbPq(d&I
z=34fL^5OCkcW-$T;y4YXI+vef3d8%H@{ZfoYd2Ca&y2e3dY}8|p4y*Jzg+)a#BLZ$
zumXw{zcHZ}{agQ`h4we)-?U)3hH9E^^Gz3Wr2*d9Ib%s36}{M9vN2v$ES)kti*yC*
z9&RDNxf&CTFUJ3`1!x}4r*Pkwy0ir3jD-YYOcS>w1^!Aas3`TCv(o$fHDJ85`H=5Z
zV0)#N9@nneGgNMgn_`oH5ZhrBLi-JtdCZ1!PrzfOve{3UPZpM~PSlJ~>5MTKIfW|H
zUhE4wo!HmYZ}Sx|IH2hdWqFMq7COZpCfjlc+-~w|#H&3~6oztq0$xXN!;x6xRgK%R
zBTxvVWTJwdeo2JS^evmee`|etBDGtkR+h&UT`wC3(A7ze!CaPl_r7=1uh`ILtv5>Q
zKhw(7EjXa}$TjDXX&L0Q-WDiwx|_vGkU%>T?LdWi;V}{zw$-)qXf|wRVXy7^1=B@8
za(Uh7(}|4S*e4+=11366is%OfzGMk#t|*NgaMHhxs0i9yn4Z5uc_sb}_IR;}LZy5x
zC)PyGA(Rc!h5VMIlDrmaHFv8L9*bEjS~!MDt8yVnS{uHlL9`;0OSbykMXchPNFW_4
z5p3==rE2|RVymPf7!T~S1teeIG-|yuJafvN7tlmVEuAe&w%%4G<XX;QLsO?=)TxZr
zK192Wi%%9%dku;Qz#OO~^raMDjfI@-cxVni#G7Rt=nMuD(J^*&exCO>110{|tDm&E
zve^=_E^0-8&LYLSY@7FPdynObA3RaAKn>ktQqRxbxGGs6Z{#&jpPeo84-EW|s%zYb
znPqFV8qy}0OOkE(Xu6c}1lMRc0mTiVih*p(%ISUbK&XmXL@=q3*VB%+gNunP85uwD
zaqSmu#H&fL@--3^5tiN6OoZ(Q`jeKsqrYMwJ>1US@H>~gK3nJT@mdG{fdRF0c6k1O
zmP(2xQambY2C`Oa#D$+3aTilk4CIwFe`!HxD~rv^Q=kJmWUl*DjheV;BBrMx!I=oy
zR&1>+2ToNw$)h3nby+)fCR`)g@ksE4gRu?z)6wZ`y2N&)>HOs%Nc$_Bq^`eR*watv
zKFFrdkt_&m+^>TEe>aKMe>X`wNK`%t7IT=({#RK9v|d@i^{&0R0BQEDUB@sHLrcm=
zFfJa>Y#F!Ij-hTO<QDu17uf+Wcn8wGnwJaoFF?rf|2h87b4p90qf-;X?^08GEHXW8
z0+)OvR#$%HsAHs=(+rwlcxOsAS9F|w$Nii|9PXypj~8vrv+^Yz+hVWW2XE~5CkRsk
zyMyIWr<H*-i))s{Tu_Jz0k;l!T)CrhtYj;4N?e!A+hALqYL>7=Ov6L%;2TzR!ih(X
zQfI4GHU$=JTN1JVr>QsH6kq0KDvHDk<4~PlFQF>{CEMoyQ@;fXpLFcpGzh;U1-Rs9
zI}%)oclBwCI?wwz6WYmBXK62vnz~Te^;lv{rKfNC&&a%`?x9rHSafv?a5t6QQne6n
zEU|iOCG3eXCAv!I5vxh?TiE<)s1@&M&B&Wo?Dm)*0)Ez4q)kl)Suw}9#>9!YOlX$q
zs%Vcm$6Cgpg{1WqU|HGf5D;*!MB{sV>=`GX53X~F+BoIy-UiuJ>~EN+^S)!DzhS&q
z@vZ1k!$jYtx+I{3-R!b(<Y`qj1fmRq?+;C~-HxNapDF@E1rASAOfP_sYk2go^=KL)
zJBJCb#6{0dV^@(qK_Q!p%H;gk1LDSFqt~xaLWTLNvEZwt$Q47(0U*eQg0BEg0W(nK
zW>Tyu<46Y?Pn0fENZQk!U*!Ie{wX0R4`N;4@*iqHH>Iz+Cv3paNOr5v^6{S9Q8sPF
za0ptmKR~Pm0n_jT_}oY-Yh4nC&ZXToZLK{Z6ukV8k$I&(51V4Ng-|-9+e{2V(RvDc
zt{ItGd3sQPW*{hbbhiex5a=0HBEKVwMeQpDZi21{VH>^4sO`jqTR4`OKbqHmFZYft
z?0*&fX^ow`lA@6QsWtR>cVi$v>q)`CKnE))2A*HJ9-s(zkD2bB)BW}|WT6mr#MemN
zkOl+XgH;%XRK1ZpEZfDE{W{xGIMk<5)0UBE=UPQN?tcfs{J#T`G9_CfG>LTwZ2T=t
zo8-?E9wR8f;jt&qc9EMOUs}EL0t?h+qTH*|;B2=c@!+t(50#BQHzZ#1$nuk1;xxWZ
zy>}|VkyYNnZbq?Xjqh=pQ)oA;DESq&Qs6}4UTYv&_EujI1cgTqWVrXs&Ul`(%Hd^E
z+p_Wq+YHQ-L4}zTp|?_9&0|;NEp<Yy9Sha;z6G;Wz-h}_!Ipt3=|300?_%C>$|9y2
zzh4I_j@xrr>;~HBx8VUWe!9cwJh}v&0y!4Y$nurr<D<E>kej`x#&yi=W}#Ou8Xs^m
zrVv|gL%m)X-nr|=>#Ops3cFBLD(fX@*YSLjmLUJL3FFBrdNq$H=2n>$PmTG1`Lcq$
zER({PS##$IMx!_)-I^?ONaw?RWWMgRuKUBEbx9fCde391abw26o8QQ4giHUsqyQr*
z9*>t^aI1P-{0g?jih!uKA72LdW^G@;?N+{)JK8!YXS+Kb`0ZlpI%-kxnjttZ;x!0k
zybMk}W{)0_qvuW^!9;--i(FV}QYiK!mSZ|yM&@n;+PIMe&0!)Qcg)5=5YREaqF~p&
zw%PaXf$veZnZQV_z?0m%`>vYqM++0?Dy=$&vmE(<wy=|y6Tm3o&hW{*Kuc-#J-S2E
zaO>5B4)y=m&X!aMPKHX#WktBGt*mAhkX44zNSH~d^g5G{BfqD9``r&BGU24-U-gMa
zlc@sEC9*YF=JziCZnG0GO)obT`Z>H6_YqD82GblUw0$H7P<M%?afP2Mnqb;`oHu27
zXFjYwdhS4v{yHt_Nw9Z9MNKSt=ZrsHTQVhLuJ^C6*VX#TL`UQ2_KVtUiaJil?SI7i
z!>L5&Kl6>saiH@hKLf9ws)x;1j0yAS%t2)?=st4MDLm70p0Bhv3U2r74{V@(dU@sC
zr3+<I!sb9R@h!}+(+2bJ=6c!lpT-1mknfZE&R_RWi)RriB8OLP`6>C6`{lk=-zvA`
z7MQ0PdwXAt{eTr<18r2MjU>Cn;PRChB5jD7l?k<#f<|iSO<LCwb4^@ozr&-99f<)R
z#7l}pS;zXuvetUFqGyd%0L9CUwQ>*MY*mkYTe_{7WHRPn73?ok-XdSC9^1wCPMozv
zgW%0R%$Tu(RnLWYZP$)bX}g5tv$$~(IX<vEOFlggYjvm(zeh;Pz;Wgz6n^4w-2%1a
z1ara`ZU&<Nj%^J(7;4bE1q1==CV)1Cm|F?nNvOq}_}fW#e}Ah@O4m{vYJ<dkwNDd}
zjW2c?*y(>!eK9{%^wdrvlLD9prlQgxU=$(T5xoJ&FiHZGnZvoli!_Ck1I+RLrvveH
zUA0jZHIAh>Wv!c;X^WY6r(}7>lvF(`vN-<9UwU88YS7xApk6C)25R4|tF8ROa+ZBX
z1^nn4uB;i<_}te!UEPaUG)G0(!AswM7Dv7FznAf$9P%ZV$EGu<Mf#D+rC7^*?$k9k
zJ}euchruu_U_An8_W&_Y@P1$FwVIRkP^61FTiDzKqV-AZ@LIofh@*&=)JlU<__e~o
zKRz0C*_mh<35|Ym+h_!hjt3i>CsXCm{{tVxW$_v~Q&k7l%G7Wkc_<pj=s?lIQ<maF
zB6U7bUxq1HHo;KQLi>@VAGHbp3%j%OdeO^13I1P_E5j;n;WE<Zf3Y9|J$Ed4v;Kk3
z8}SIcnlJJ1no^wapttQ6goil4{nGF}RXfPX2hy)yF;Y14h%*_&x~-@LZ&>2mU$8#+
zzM5P!I@l|?TqT|ucyS@!a{u%ZNzLYNhl>z6MqLY127sStsU#rR(ik~+g=)t%6<~|s
z`)3L)E$tSztA4=C{yDMRF*<!>E4uv0IV|)@tM#r&8=$GB&U|OjQvN-ubcLV4A2c{;
z@*k>0>fgtk`VyP-u8sc6A{I7@(AKFf%e#{E{s<<~9Mgg<sIx~sStHdESHEU$o1$Tn
z+t8~#9p{|c6eF+boiA2iK=udd#B7n+M7);gzS1ba!|R6A4V7${)5pu31=yOv@Y$u7
z$t0IF5gVwCIO^N$ZebdI8fh~MwObqzZIzqFzVE5Cq8?U73DP2|B?xVjFgZj4aZa}b
z2r&lfJbH+oG^x^D*w$89`|{=PL_qTe#Jg)+Y0<~1ML0L@=;epAqE8c@r)?9j_By+;
zN?psRHkQ0uTvpHI<DZyUn@zz<oUVH^`qTA#Guy{rfclXH3eEM-A8qg?OCq}>Q)apt
z5XMtE>-THM!@}gP?hvH7t+QpV^=pT#!hpO{mD0T2=L~)gRzke0wHYog`0w~ri;+^R
ziOB3<Ks9Q}9JLU7zzS{JybuR9wvy%i$>m>FWPevCND)bsg%TS$Gi&KZ9>XkB^)`|*
zZ5IYicr8?zve#i$8On|e%8A?)o@;%&B|i7(dfp*h`?HLP8zq6j$Z~Vb`PPhfNCOEg
zx_W%wn`-4Tfiy8!26@=0DeZJY9k^2A+L2#va%uL5_Lb@X#dCo1pb_F3K%<i`Q6+_R
z*6>oa0`N#m7!?Ha7N`~rHaPM%(i6KbE=|cP1NSuSui2;B{`w__FE2ZUGc3{V;YP4O
zG!z`E9=K-&cQvBcJYh?(E~Va6YEP_h#YTa@`->h<HePqwhK`0QPycY@mz$OLiS5#N
zeY_Yy%6YZk^?49yroEBhLTUmU;CJ+yOv;18gu;JZ{b!;r3K!ho>PrVb02+59w8|;?
zx{eFg9czA!YQER_81IM`3C00hN8|tvcq^6%Xx{|xB}Qp^iKn=(*y#xeTShx2(f4qt
z_6*6^$`vl-?u2L7RzP($o65A*>HTfD-_}M?$LyPUt=IeDi2$DHo14rYg>fYf)Tq+V
zb+*F4c4VH}*H5q1_^+R&e-_GL?z=wk37_bX-z=#9w##G7Tt-gXrj*>ISd#H#u29CI
z_ghHRYf}O538qHO*N)`gk*>m}@mSXLHdf8RLn}s);8)hj_J1WC;1Toxms-T4#VUlb
z6o0I!sQpV69b5a+{r8`l&Iq_&PY7*sStT0wz1n+zQ^v2k{or;xW(0M3=rA<ip+cad
zoTeWCz-HnK;b6ibhzSZn`h6G-_0fIx)0P0^<}=DzYb026haU4Fs3ukzk0MAQ2a4yU
zG%^p8)={C-*oW_ys#!Eld@Jbu>ZNu>Jkau#k=#j-nKnV0x5et<<t+m$!e8Hs3$P1$
zGH#e$G!#Kt_{mC`0Um`i<kaSh_GkSomlP4+mc~D{Zk!O1QkYwDN8v19RXbH0);mR^
z)6H-8@_3Fc*=c5ug;x7UY_toQrwsSFyFdPrY4;Qc9LOp<X$+@;oMoj`9RV9~+m4$l
z?Qiui2@0$3ojlu*cSP0<)0a^$nx?6zs}pIeQjmWRRq+xTkCaGJG_N_PsyPM0_j+11
z6P^n`kHxh?=PAVW*`aLr9n@iABtI3m384DinK$tT#)-qc!iV@yp*y0gpI23iCF~B(
zvO|&(aXL5}Ea{DWK(26S<>fc};Lr1p(fPlPFGrV&4f=fiRPu~2_}llcSWjZbKme<(
zvNlCj|Elri{#oB@NBG)lk3zgQ5o~GDYQzadBS#0w?w;L7c?Op9AzWJZ_@HYIj^+$_
zTU+Jo^_<3h8TdEK1ArWZXmgl9)NON$pE#d$pxnizzec$y281#L2R*j8L|R(_tXSnz
z60zkjb~ays<$t^*AU#X<F^xN!8&vouFY2})xox-ICsY~iHH9=WaYcU)HF97+ABnkW
zeQ-)`Q`suRLYs$~N=N;=4>`4q$#pcrRrpXUNzh?JAWHTqlge|065qX#J#|ags7ei`
z!Ivh>92@Pr6m)g4pSrJ?zudhytvu3n=xpMU8z|P`xAHxx>Kq7L2J#@bz?jV9?4WcI
zrF>Zhsv9bRfit?9pXTs(qyIv|)}y_a^Bp$FZHp_)#ntf)r4=RZcZigV<_?Lfzw_8j
zzWdkz)fkDf&eJ>^QAiP{BAJf6EGJy;R=e=&JlzqKTI5z`8bk2|gc+lYd@lVYM;lK^
zottNR9uaI5QG<msW#rjlW9+Cvpbn<0kt;pj?=lnO4x|@@PRDULe`X&3QM&Up`FJ7E
zoS809U6+$*Fw3!1DDLS#nRmz27&U@#JDR}xQ#Sb$)6YR?H1~M<C%+vNQwnjUUrMkO
z!}{dyXlo>qQN9fA-padr=B|bZf|Q<u<^x6^t#X8g8Ji)5On)=z;P&G+o0K)1UunG!
z1?}Vn8^$v<B9gs_XYRc@1ynLq;vF&Ly_K_8va8y0Y;AIqP;aKpd&{43)gBBjE{vI<
zLq=R!QNLo{)oJxm<rP@^B_ET|sA>D&m5h$}#P<EfHY_n!B}_uzNqV4z51I~0GW_9C
z&B8eE8j7AfghONQ{R7^yPU9zzq9jvUMggPdd^k75e!Y93(?%7f_TVu$ZnD3v&B((Z
zP<b>2w;F>m)F$&;Hjg%8)-UDu=656pIVEYN!=G&eXx7qF^e9sMYRW;yIU*rnD;>PS
z&OMa>kEyqei}HJ;y%j-98l+RYL%IZM5G16fyGOc}?(UQZ>28n~kj5dT2N)2fJJ0s_
zKj-;8yyVp|-1pvBto2=}O`iLsah+$6kZX6pbKh5KWw-!t2L3(mks%T!5c9KvDb!b=
z$CO=YLWC!(e}_88F+=OqV3IuLY=NOlazFRLuobX8I^Y-omY*>aCp%P)5+rR6+l}`B
zJM&hbRA40%0(z#Mh9wSq=ZPG$koA$9O48a5WNcIxsXcOs_J@c26ek<o;o?v2?h-#j
z$;oJf-8T?zt3n=^&_8)@Qbmy2!P*@0#qMrAfU3#kymVhJW-(m(eVU<wOld6NuclWK
ztfbK3iXYwQIi%<oxO>z-FGA|<7YF#AbyM;lD?$cB%*2HWC3E$YzCLIx-pr(k4s?>A
zmzGI3T3#=u+gdbAR9CQ`Xosu`P}ZSQG)b@}X8Av6KJs<n!yoXiMc1b+w6H8Wx+ROq
z2hkK@#9GDoqW|)Bu5pX+LmWM2Kk=unK`VN^0^(ZRDr<8(iw3XWONllo&%C1nQ;EsE
zcJ`M64kUN>VB$7Zv>hO?EcA=MoZSELEyH29CtkoW%~w{QWQdK6+A(Nj*M29HG_Pad
za!#rJPQ-!dbAe81v6qT~m1T%3ag%+X6U4dIkOoGUMiouww4#!KO${w^|0Yi@_o}r`
zrN&2l;!wN<Vm<#ci00FRc5$P*{{mGo@B<isBeLMamYE|#u5vq53r(>9zL>bCX^@9~
zb}xrYgQTWVcyu<yIS6j%`2!gSKFs=G-rQxpTA{?5*=ejPFY3Y^%o$0B+WCdatR-Z%
zt6rG4)%L<#1?}heMf~^+$g=1>FMeAY+<Z4B?>-+}?Hi~kAIC6p7-tsq?*DTh{#A)o
z<?p=$tqTg6`4R0YBg`EXe=sCBM|n#4+*<p|I4^Q=UTV<0*gA+6yrnH=hDasv<5Bou
zq`Zsm29qQvA=Sa?bbz}ar1r46+757BKz#gb+0F%w7pAK(&=Gq>6eygxwnwv?q)$0z
zBInYx>wee%iuSZubJizef04k^HW7u7wMgev73*xB_MT`IjGXn^@=fVD**yw4NQQ!q
zM+PJU+TikUGg@B>o3oy?*Cr9&Zr}c+Vojd})1^>feb7@8|FC(SDPfHOnMMJ~p8D>L
zq~8Z#YIIBuup3D5F{du`C%PZ2CdLHv^FK@mM<ZbUQ{j|Lj2}MZ;{IiG%vNqEgo!H1
zZ#3T-T&fQgNen6iXEHWeT_k-)QGYw5TVKjLE*ON8bJ?klE1TeH-zsVgrgTlI&u>d*
zmEL(`g@XK_6_Ap)C`f}t591T&f_l1bX8?G31#KA`?=J}PJ_$4WTx^O+^Ro9^?WuO%
zoNq|(6_j$j+5f_ktL4G57f<YIH|G`KKYLRYH=4a5yWW@o0Pc$E7o=TV<nrPlL>gZr
z=JyqC->Ax0&|=13)=Y)(4(4H#S{V1Yqz2K<dixuG^A-=+wwX|C$NS=5CYm_SSt<Em
z-82sm{fM&+gP%Ty^`(nRFn&V$D`%5#Cl2?w(-)~@-wJTho+YDBY<)Jj2gZ;h2dO;K
z#6>kd*}qy$XJs2FRMfZu=z+4M?@i8ckSJOeUU#+UidRYw9Y09UOyGn2ytx=792bR}
zm=<^~f<CjVo#BNg8+CKe4)T=i;cUr+K}d_BEm3NQZ9$aCOB9@yM0slfA@}Q?2;LJ{
z?s#OH<Q>~VL)KgE^*aI09<k*us$Ti{Nwji*e)xi(6(;&ShKZ?Y89m0JEhpmsDW1->
z#!#2WT$x9#JKmVbi^G1cNLh-xj$8NIj+;tG_`IP_!|96|luUsWx!T#3qEMKSb}+5l
zv^*`;`$UerobI>iL<we!95oSbM5D;|MF)LC0;ug;jd6Q*vl#xMC%u5uA}fw%2g6So
ziwG6l!l_j7PX_@^Z4PRhA8cx1ZD8}_UcfF%;Ly{C?I}lTcEM!KYurK;Ix?>bi1iP&
z(PRb>$6yg(E(XNS_^-1uBP|?zPip+Wd02>oV)yj+D<=`~<WDN%YssQ&$><z8bcM>v
z(@?K!w(#=V3HjEB@BAyW4J*z7SR-;i9a^#Vr!c3>J>T3wuPT-$C3=0jMp67jkL|~h
z8y^EbH$oD)h#v6y_1@RZZDh#`_m#^ceW=u=&cvw=Dsr<TMEGrjT0c^o6C9q~oL_0F
zD$FN5^^EvQExx{<N-N)Z8uVqoO_dy;{Kv{-Z5oLUIz7*yJekCqooS063gq-EZg7&&
z@46}^;h{I1UTc7SG+wd&b&^&A7X+JzEEmC}t>^DF)SU7tR~g*ADd~+2a(R2W4~*oQ
z$=_8TuNZ&aF<Ut(*%?PNcwF`m<@p(Y>6g4Y$T4~Q18WT_O!h8nWvBBya@(pTZLc>^
zIa#7wAD+GE|3aL&ra->pp*Hk1Oyo~ZDi}EwK`_6Rkg|YI0;YDFJ|a?cMd-LN99#Gm
zg6XUh(ca{jyh0H)@7x&%gygdah0l&Jem=oSl&t?1ty%W{lF*}o)uf$JodhUvI!~UU
zH|rcS{oyp}T8<^&;?n!r;{{{PHfdx1@wo)bGwwp;z(GRd-ga2S5M*c;Z<IFRLNnIE
zXQp~2A*xM+nUWEmTvF_u6YM4zz(~Y0`SL`-g+LA`V7I9~K-ymCH3|IQl<t{-YyE5t
z*e|N)wQ?({(zvh*!Nbk2elI`ti<`j${n2+1nx6{{iEr+rua0aAKL`~(WRjECCV~7J
zIa(00T=`<C9_>)X(wj>!I}X>^JV`7*Ib_i<W6J;HmD1xdW#1SUh1bvHMn`zE<4vxP
zIBU4Zj$7-ziXbzopJz!iA8$PF8WQ+~Op_~5d%S285*k@8rik7gl?5h@DF<N#LP3q*
z)yQ8=3KernWT%P%WJhB%$__z>_y9OR&tOsB?7FjuhOSKd=3Pynb0)cg81v3h+}azW
zswlv(;vQ5STki$1VxwIvyDKs}Rc$-AUyTg79)tz%;wKl_0x&C5Mc|qep=4P}8^Km8
zArHs3yTq-EN(g0-@D4A`lmobaD1|Wwk1cDbhObWKe-C|kyjeVmu1-tJ5Jn$sHMo28
z@qXI42Nar>4nGT<$h4j#5h219c$8j&d$>ofp3iZoAMFT#F!b|86>PjP-vPe}s3iM3
zxKN~O!j7EY6QTVL`^*y<q)e;gbB6wxh%iJ-9xRn*t^=W{hoZmDEj3SBycIX3oQ$z6
z<}nN2@9AQtbD#0Zb6~0iuZX55=JOKZ)8MuB>`zx#I>D~K5k94#S0Km$F5k19RbklL
ztLAMo%?!dOq0QR-+89sreFidY#rYlDBnme|UgY|SzH3~R%)->?s4CT#Aiv^a+G*Ks
z$H5=(<fo&>yXcmqyvMEC%VRgc=x*0@f2!zLLbYZ*TC5~}kZYqf{r@04shA*3`y_@w
zc8eK_qQO~?x-`mESLY%%1U8M$)Z@l4$a_=@>3Ms$g86w`7wvD2$G9sEebIUirxNo!
z2F<e>?ERygLSuhx58vsU94(Rh_*H(J7eM5^f!bvXEXCG;85`GYoa{D1T>MlxZ6hQK
z1Yw3p(1MgfRM-e?5kL_my^XW%>hU@P56$(7+G@>`hU{GM3zUv;`X^Q1)(VufS<|Yt
zB@KbmeoUi2q={6ONTrN-ND<3mCY&x?yY|<+c|4$3m@HQgPLXxiw*oE%QFSFnyxu*{
z^$+numg2|vcD{F%SzeFM!jCbHl(z*@k3SwU<S#cLXXHM)Ce?5{6%zPbdwPj(>5IGk
zt@4dAbv^!7ZF<k)8yARm_vAeCF~jQQInD5-<*MRw^6I<$+2&evFy{!728zg%ROXBA
z7>IRgC0$CuNYI2jM@rOmSfg@3WS7Vy&vaXf-=dq7i~vKpK7=*_bR%92VIBxLLR*It
zH1Fk@ZLfc-N*DbLDz;WMW#y+23GDskwsZ8^HD=7O-+RKXg-@ZCXRbr2>OI2tyPeWb
z78Fe=#8;nd&;M1ow5y!`vD{<Z*Euw|syl;mv5l5R0q9B4nKKH*ty;ls3~6EnY`V_s
z1@wzsP4_Gj9K@;~qH71PMR(-Lf;Qzrt$sP00vj6~dwifL3Q~oF(-8c0s{%-5(qcVN
zwTG`7kKBhHrSqEZidL~It`!{F0A0N(cp)ZRqp9V$UGspzVCoM{1tJ?p#6DI4Q1PZ;
zlYje!YU83~JMk82lzCGjYj@BCYk=8oMo1A~J|_6hdhJRgvji?G-~hN+nx9c_JQY_W
zSatFIj2!?nY7&{1q~5VaLvvJL#=Dueh37*sKZ>>q|Nehk0G#3_6iuh(OB8DP<joCE
zI^R)}w|bOP3+6TAVVdf$P;Y#MJ<(1jh-DcQID~b^>QEEC8_zumd*eu9^Gt7LWa$UK
zU0l9~*0;~vxtNT{s)gwMgp`#<VJ&yA9vL1D4qQ$>26vC`jECZsqgjwRo}B7+=VE2u
z_Wfv>Xw;9FMPR}|6;fLG$4hMR*_{U8mmxk1PTTfM8kCHccZS(xhvMb>s*jZxwdAnx
zve8QD8-O|4>0+-jGTw@(upu36VJKhwvU7mhffb;~finyJWSO^Hf<yTHQjZrn&8r%t
z8bw@KmC*V+_)t^<moqGEIhZ~nW?Gx3D2?sK?dfw<5q?;Nb}-~JbA|*aGU4FTQkR0Q
z7|syZKC|fLngnSgOk^A$P?Z76eQ*gmrl=yq^5ltX^C|5c>!bdq{Q>-F2GxKfw-@zc
zW6DbW@n?zH9*{YMuUmha8ci3pQ@$5+|Fnt32KH~KE0(OsEAh3$Uz_-KeG^8H%d3S~
z_N%huzNh8d3O$VKX^b0cC8OwiOxFr>vry+6yR=^t9kCiHj5-qpfoD^OaQa>6TidSG
zx!fB(AmqY|60wM11MwU17)yZmF{?H4S4|vBkf41d52~-So}mQKzmv*|3=!5_EuE#|
zG!o*fWUQ}YBTh2HHhVv4o{q<@++iR8jit_@tMZSzS!|}d)e6I3%CI^LaLCs_l*;-$
zq%l~}?<nw%JaRF=686U})RJE^PVZlw`$z6-^7sIr&a1AN2i6o;n^8qRxa|0Ye3vEs
zRt1Bp_4cvb>9~D9dR(0GNa(E>jBp(~)S|&o<X}iW()|odxsP#CEa87)ImIs?<A^qU
zW1ax_L$<k}ezjV_L2DT1%#7(+cN(AjRj@;%O0Dh`%t~)7txSqPT#=SGz?L(wpV1s{
zsmfTB0gphzoPnA5V_@Nc7(KC5%Q^4QaUTrA5bDoIm6<mG-aR%$0TVh;i&u29NTEfj
zq}|XwsUQAN=Ku7AGeyUqF}iQ@h}6qjlc$`V5Sc6RHBD^bSRf>?n8GaK-u!K`2`1WD
zw0vMQ%U+`f*T^gAY1&_d-s%H35FjCZnqkTBygA}Gd~IZVUUi<uVq5YXL!YGhu1e$B
z)*GqWGJ|x+=GPUrqzH028dP$e8hhtMi!FtG2l5`PmMoK{H<oV080lz-+l%S6K@T^U
zu{u8OiQ$lo3hs)c1ki`WdN+>F@CLk!vCD_9ei{#twQf!84$c{pQQ3iXELgf`is5fV
zlP{}ouHTD{N9t-ldW&z4*P4$;`>MCMghO8sHOrr{pW31SrA;lOQlU-9ogH*;nWc@V
z^_2q)J@7O7e=n+{+MGT~9uS{5UBlxDaQ-3*`ujurq~@R3&LIQmkhi9y<LKjOy`?Fa
zAamy(hpL?`>{iI<R4!pKLRZhtlA0(l^&9`d$6gO7T%?H_=lh^zh4*4}v6MAj*a&<l
zVbLPJj^7!zK)ZXacKT=20sh&t08h;a{jb(KBm!a%W(cYj{m{<f?UuAp`mU`txUFsB
zn|~N#i@ILZdF{W#o!IFTj!sm^dL(&b$(E%i!|4DWaZw1k+A9JLn8QCPRk6X5QAx`k
zCP#u)Zw+YW3!2mqYO(X);)pDQ<kfk1{Key5mPa4g$={J%mWFrmyM6D+*Kp`_hK~^|
z-k1>~9eDLQrFal`O+5orEi!@(9sAnxz|YDRz&|>Ej>nFfpw7l;(P*x#a}(}UnoAxe
zojv4it62cjky1H1DGo&=>9wgCQR@|8qAxm4It!ivZkU`#9~FjQ=c1QXnB}Y1b5-c$
zUqc$9u7~f=7delyP7`-?O5ehGI-S@K6Ql!JY?-vA`p0{XP*#>E975iwED8)J8*7s0
zEc{SDwjnT6I7#di|K5~))b99LAKK;iw8P^-`0D%2BY{>U?4rV=JQJ5i&f$)x!2i~_
zbeQyV8NO5Uh_*22$A5o#my2y(^KXdo<kbI|um+~*=!qczfZ?CCLU~1r19_NqzLGrl
z6leIzb*tqcd=mJWl;;jO8;=};e{)SQ{`yBhsILx!(T9#zo%YF+CS1%|R5hSvhHg*%
z#xh*|cIHy_j^1dzBlN)CSN#60Gq;b$;a`-%^*O+5G&V*#`wsau_UP)U6p0onnUd{u
zkb+1c)c}D60aClJm&q35UNg<^fCRKS{f@54ZBxM{8y+p+XPnpCkrJ;UpdA7pK~|GE
zrmk(DRKu=%s&Lv)S`h!UAvYOK`)O<9VMG-b5uH0oe7E7<t6|1S?NJ)P(d4Ito?@==
zkB)z%GsBKOglZ+3h6~Yj@a6GszPQx(Y3<($Xho{j-zx-q9h01+cSoMF_;XZ<+}Psn
ze+BkvBVwEQMZ{aCJ&{GP?RTZW8cxk{M4Ho)+RF(}L4SSZ0VHtn!tK#lJ9H75^%+_*
z3hZYlT?Zz)n>q4T5c7D+={<~vT0x{7g9`0#*?zE*jk6N54L7Sl{hBTP|9|T?jmlcN
z*gUUmOdO5p^<I-{k$hR3=0w8Mn#0!$Wm@rev*6_*f2ko9MKPejS^0Wd3X+%Wl;{6=
z;9Y?T_u@KLI9Y&<83WzEoRvP2FEa^VK)%|53<FI(0;OWapc;t^zaJ=v8eyfA_=~D~
z+7kiiZ;KII)m##P#rVV0`C{}@6SN_<q{igNFT_<mOx2t_UN2<<ai&o@9*F5YsXE`6
zTIw842QfLZs*+!$k#O1s+aZj0(Whi78G(|{O_@d&Z+qp@sv$YTJY$^dS${)quce+>
z8y3F?!DzCkWLk6FB02X(7MePKdn>Y(w-6dV?$!rLgstI3J5(J=M_AtC%IcF@WT~)W
z_x6D?aEl;@fm><_Q#`(e$FG=WX*pBO+*oAccT^q{bI>V8iu*>x9iuOwTt#1{Ae4wK
z3!uz{T1?+{R1-QQk572R1Wbk_G`Mu2x|<!ushqBFcK82MnR_*XGcIy=eWLOx?sHp7
z*!u86{0&VzmMFbi0Bdy3INs_hF*ZU2YGldGK5i1e3r)Dhk-z<Y$&T&nPp$3~6+!<a
z<D?k@YQcFAO3K!=WuKNuz2R41j*d<sV<lLMIl^~>E06X5FV`u#p*y0#Wc`TRZ;566
zZw#k8t{l@I)^fc(MjG9dC~qv&mR0I6I>phZ0*BWPQU0jh&BHIV?E0>dyU=~_1F<@F
z#tv&PI35ubS5Nu1{nv$=T*2g7r05Vs&-wIOivOiMUaX_o86>}x(E6e0@a<!&xQx%W
zEyZ4A(|@&--|#{0!(uuw%DQ~v#&|wWO3jL=^QU5b>cFpcOw#uACC2TUPhtv+jMI?N
zi_Rqmdpb<qRTE_+?C8Ac6(+S+VAMz_+Q<7Av4PS9jImb{f5{H}NQ}$aswq-679kqX
zyA5I)v=S;Elo(V&K!7I0_M)y#J|d~-Pd9#JOM`kVqgPnAUs!QjZ9^e3mt7|+*EbV)
zzJ>nv7G)0pL4OLwC0!1XLj3P23;1#+QNZfDv#1!F(->n?errLl1~Bd`0r$Z&ay{sJ
zT$n_uDT~i3m*$EwfSv+95EJTY(MYc%QBh?{iv;&Z%@t(zG}Sy6n`?J`C7cpe93waN
zjv)T%rxu<?Ev|sbVvtgt1R2CUjmsw>?!OIT3WwOD<8_KVSki3&f<?qG>!?*Dq@PG7
zEMTWK2xG0A%q((_m}Gdnq^8E8HJpOC9bz0;w;<ml5JtJPzei79`VMWLk@i$K1k(tn
zif|L2pD86E(kVa_ix1U{y6oxx)3pp~5v^{mpNvv`Hzgn<;qxpkXl;pcP&@V;ZLqbr
zA^~tVoWIV`WbVL%X-6(jh9ll;W2VHM27=4cVoDqAIoYd&I=um>xOEJ3uy4Gft2**@
zgUZ-3{`jVF;}<NTC`6(dJKxy9OlS6G<eG`N@syQVZTr{UptTlNJ;a=tBSbd+H|p5F
z+YBCFhp^J(uQO!gQ*Z4f={VR(lCZP(hJd2lGK6>{cUPUwf(QouybB+RLL86R323xw
zHsd$8D`+AS_p-wdm|l5Ec3iuAk}QUdzuXm72x2E?%FC6%@kDurD9<jEu-1%bJVRS3
zD;(txjO}#jLgMFe6$1CeIx+v?Lm${7i12wN=P%}6){bZTVG<0fVI-BgC$lH3MXQzw
zo21NQ=qWbQ)O+rDzOLs=v6{}0FXS?EmwWK>?rNZWaIijtO;uv?RfGxHHp@fK!geeZ
z<QPX5xD3zHS1+9)mn*hzr+7afw=Pxv?_??23s=1N_b8389Uq03S5H^PPFg0A9lSyp
zupZMV_5|7d|2SrawU;;`c}}b`JmSJCFL{XcjTVSrMt2?xn?G<oh#cTUE;o%`hR<^E
z<@p{nvA%d+4gT+tCX5*#Vd><~xd*vYbtnK$x#{U*`0kH>>B3w+iaq%~xQ1AmF_xw3
z7NQ~JrfmNfW^5Rw>1$TjJ0(3MJx^Djmgk-tAbK0b$pqAQ22D^T?{n}%ku$^kroaBd
z-eyLWBM#bVRD$RScVNIp4_AH?Tl66Ofp*ZVJLxYPq5e{wY=xaLJcu1WoUIDGK8!3~
z<HNddZ*?A>#oysN)=kEMdMQ2n?8}=e5nO!|3;o<#qCKvEkzlGjYt^<LFfj3RD%*v}
z2&90bzbvkz=uhDSO|)d(-@0Rs#p(8;7~3@IsJgVw{B1jN3r*7T!iaAWB1-zweAZdn
zWx1MP@3i?sQZd;S6olf93h<ge5&kQsB^%I`M}^VsML|p)#{oyYB(^N)2u(&=38id>
zJ59u3EsyQhRyN(e2KNtq9NZuygUUAinJ_rN@EHYe#}i;uE&!rEE52KtqVoz@nvPRO
z@QlctN-`W!C2!Vlws4=ROWdW<99(p)gQ4X^Z%qmd`wNr8phXY2ZC>P};#}zp1{_K4
z?P!yE`BtH@aA~4?jz$=?LMDRq8$kv!?XvciA_15pn48wdd?<n6Pb*L;Hnnq|j?@5Z
z>K>gGH@<(WENU~Hw>lOBjN&m(MR@Evq4f*V@-(HafaA<qCdmF)_Oxi-Q7LtukhB(m
zv)1hXJa=PH{6cH@HEiO#mOWd1!WiG#A?_%r67yc8b_GL|jG#x-RE;?5v?QcLn%gBR
zmSq%@w=1^kJBla$4Y9}#3xoEl6T8(H@ZsLg+8E;trQp%MrwNLY6L}fi=d7;_#*oWk
zcEB0VsM*#2$5cG1q9oRpJjWY0bTzB>VC{Av22w*0spC0E@aOovLHM-BBJL)fik<nX
zWgAyD9pHL1Hlly*<U=`XCd6ni!3h0;?U9TKC!ZMsC<g}yoK7Y6-4JEk@Ze2RJ;4bF
zS5UW%+ls%jONn)TI0yIDBcJ5a#}x6~?&)iCe~Ne)ioL6e<P-QKe+N6|ZPp*h8y;5a
zklGx%kmE^JicY4@jEoIS6)da{!d;Th&`0-}?rRQ^8NQ-VSbt0Whw$;-R5XTkweW7f
zoA$9y{7!r5zrpa`4O1hmpqa1(U>VqXqc2z+%yso<M<~z&o-8Ti$K#{)(jPX|fvs?6
zb{I>a3$LhVAR+!DHxIRA@3m2A@raS<T1Uy8O3NJEPZ|JV!Uiu>J9;!S`<balBg~3U
z%#b2@JLJ9y6Kbi`Kj2sD?@pLu{1VcYTeJMGMT8Sfdz{8=yJue{gf$M7ciblWQ+o7o
z)2(ga{`5cNfOYR7W3PmI?chs$#4&$Dy-kJ<cI1;H-8LDs3RD^ClPG|^LtAG1GUs5Y
zh&H^?nyn9(&T=kCEHmZ&Kv6TM+z;9>TUdrZQ-iaeizzJvIvFcz5}`|e*rKk3l1Q|x
zaU~u=!u=eO=)IO`5C0dtcvAG{ABVeuNGqg3($x3pTIjcm<j$ANr9@vDW3nOXxAvw%
z_-50ulBvRxz*^%#zkc~w30j8&8hfmo4g7kOJP<eU)QBxoa>&S(^X0fpL87x&MFC;=
zZaA88*8N5Px{)=wyyOpB?G5(uAa*+5P#(F$H@~TBl(p8;qfhsLk^r~@AzX&PS6BPb
z`yXQ6zt&GU%`Krd?I;u_?!F<1TPEVbKqBlsMmi&esQ1)cG1@K1c8_>jYFZT}q5sg#
zUrJX7H~*F$2Te*sK4hPrv;xb(1PeX&>XR&0i`$;zX&2W0fZ;^o+J@NhR`tfupEdC;
zpqW~+d8KCs1SQ#_^8rhW{NVyQn)o;CG%%`oFhH3b<bAs*_bZT`6zWZ8tiGksVXjZ?
z@b0a98^U%G-E-^!x#x!;VPxS^L>4$|h>x9>0>!y98XWiS3@v0i&d8K`ZD-y@v^k3*
znACJg@o0YELGrmyp)#KU4HIbPn%Yp^w|U#u$67eGd-I*uh@B>fX%G?O%(z}Z-;S1;
zdrz!bK?{bPP@blnVR<T3_~3;2#kF`<3;O3OD}!6k-TKv=tIKP$(%iM10v3bY%?XIp
z0e>~3fdmJJF{jqtC!yNH)?-zldyfGA?^ba5^-hH`oBERNZH3{(rOFZWo9;UvFc<#!
z-zSwaClmMe{b`9=P6f5V$<~+_7@C^qwj~*NGpVX@=N#u^0ty{S%7^Eqzq7-PDW2x2
zQcRJ3AY-$UNzcz^8gIvEB1?8FjNhzzErmPY<NL7(0Qhs6L((@}I05;0OEm<Wvoa9*
z(F;Ci7X=z9|1w~c$z4<8*ow9uP_-7AQv^)r$??3T&7wX!X;_*D0bKmpzvFKM`V)jd
zhWC)A=ELa;Ay73$DF!MB+y~+JTsCp$SX;L=K#igJvx_`Rh@*p_%*Yf?!;-(I%H3E8
z0V!;${6#7Sj0`LSR|1=29P*X_+<3RXR<f>;MH?OsMWX0Y#j$%u0w=bJ5J`>VmS#q>
z^^~#?s2s@F>8miPTYl$pX)%7P=V`k>`R^-|R=#ChIf==i1cP8~l<BY!L*$IqoGf|n
z`Zn}MJ+GLerNh(Cm5*w(KN#cRCBqVwNqkzohM7Kg==9In5(``%jk7K>sGZG(saCZJ
z$O-($y1;BM=G9vU^BbRB>A$Kth((R8Z7NII8Y!L9M2cMkBWX!OVXE0djY{gmLxnC$
z7;;QN7Be|m&l?7Kbn?br0Kfd=pC0T5-nfgY!x$-_(9~Nb13Dy?YWNwhE3O8_Th%lQ
zF8cc;JBB)l$EbvX{4EdbmbQ4c*%<jEL3Ub8FVt+QPZC97Sm%--5-lReo3uBYTq?5B
zNqNK4-GzMYbRR6P{1Iu@lCJ+5zz3hpT)p^_`1n4Nv~<DVXL`U+G%}XE5GA{^giXbS
zPH;|968WydxZVBHvcQV$7GVCVdOU%Zy4;p2JOG&dWP|e{6dXvBZ;CL|2`b!8z*%WR
zAhgVXiUJqap*^neWN4$c-44^yQniU~!i)p99gsk5;|L-J$2k)(J4d2LXmD$IW|9{3
zG_+r^kt&+^=9f;#jIq*3)5#VULwa7d*hqeAQbYJs*c4NL>C?2iu@*l5DDPCFpE-Dg
z$KkjqB2P<P7@=CMj4|mw4Hv)={$1cNjZ+?%Bv#&Q(pG!4uNmXfJG7?}x|Ggqrs}nd
zQkpHk02A5W`!9^}dg=Pdct9$qf(`5%G$W17%6z9N)QH^;1z<yG=YlaO!<6hwgcXqH
zQWjoVM9Lymj-k`~N2GCly(51*geMU3kQK1IA_u$5DpyW<_^aPm-rZ@AfDlNK-cW%U
z|Kh`&GB&8QVc4faP2LEI03jcXDH9{e!H`7xJ+mix51`4>3z>5u8eRQVMQmlD)jNk@
zT)`dIjDy7u;G4SEKzfW#;j<I7{AW-AM2VS|XF5TW-rElK)Jf<Kw?nxvwlM~gGMnT8
z6_RHTShmBG3;7NTNv|vMg0TCKO6GXHK#;MVLH6wa*jvkE8C?2Ap)#{}!uns~fQ~Jo
z51cocjRv$}nbFhIWOdjIwiW_bC6^$FPPa?@Ul5%;KD8&`tctsAq6r-!ZB-jh1;snT
zsI=mk!vXpQsi?MMtEaw+Ge!%GB}ZpenZSeEL(f%hK$G8CUl;5qTtl=a(*@|0i)KSA
z9eS_*dZEm|P`v(SmA`~O+cn^W@RZRlOL2|R0FmcV(fTy!)<wg16Q-$A=0~~YUQdZ4
zjMGJ&{FjK!8kGT4l2;4q4pJPdfYZed!a11YQQ9A)&o2oY+~X0(FY(4Na8njH+(e_l
z00AZ|aMtK(rW7@uafjmusLDdTi_`gI{w6nD=OcD%bK8>D57t2<c)zEzObz*An^QQR
z-M=2R?2d^wxHXNkz$kca$0`yfg{nrEu9<N+BQAeP_((r>e4hB*H%YY%j#c*<?f*sU
zobtvUlFHL5;8G<g!1@BHfb(e=ktUXa457Rr$N~3ww=Y1SDa5+pf+5Px4R5&cHEW|~
zJ>E<-EmSupdQ(ZIE`2ELaEMGR#GxWCd+N{O?XgGa$kqHf0t{)z(}Zq!AYpkc0kY-k
zRt_d40rr0qZ>`{fZ3Y$2YJM?IWX@V!_X3xb9kI3*|E8>0duIR>@h55Q_bEmEGPJU}
z(BSnu=A<d}XR|OtB@r(5DOkPtXyU}lf<b_5j&NiK#G0vLsBO}ES&^1BJIr?P+7}P;
za#}kSVqCxw-{Yv`Cew2LV46E))S}#Z)1~yG#XvR~NFt~s5$8_U@j7L)U;HrrLSv`4
zByFa0AyCKTl@I$DV5lp`U);FW2vP#;Ixg+tUD47^NS3!;K#tVyR%2$V>0FfybgZGf
z&5L#EvgOt-YhMn1MjH6!)`^QFU~2Ug15$EO;fVX9O($z&v%H*Y`!pw(cRMe|#FadZ
z+xw{0TjWbssa=MJCf@Q72Fi*Hh$Pf_9{6CVspgGdEPd1Tvk>j`QZXh}ammj2Qh3Rd
z)g;oG!p%+)+>|$no1jcXT#QrHR0zS=xecdD=5)lr9)ac`=VW_Y+0|kyz;~q*gwgvX
zwZPOK-Yf)mqPbq|3TPqe8%ESS^a<z6LAVIm`|Kn!w(rDMm$)#z70KIjPIt^;t{_a&
z>-y`Pr{QQ#j{kZXHbb9=Z|_EhBc$Z(w7JVfHz*{G(wG&<LF~7>6}m@)0KTjmiVaMM
zFC|*}rzXFJn$BP{*f;8b^uBUtBOM(LrE4b4a9{;;Owkw?fHDB3`<S&nOwC|^1ZaK*
z=Txu}{U^ZU&0BG-92T+c>AZoqs7TwIki_Rg$pO~uB7;49u#7<=vIsRj64SJoA<86N
za<q~_0ayBNhaD`G0QD{ky|1Cp&OK1P24bA_pvrEnhIlL-PZRb!*}}g}P&?z-fo~sY
zERbW1#;PKaz-lNCls|8Uzqoi39jv+58GAK|EDpc*qW0JtaX$&a^rZ~e`dQ3tIet06
zBXURLjwjFY(#8L#n%nCbnX=_$TEuwu9X7n`@x=kwgLBBPbF9w;$QK73QN|bb%=5&I
z0cp((;c~AAz^qX)`EH{UDZ19hq-Vje_u-FYZ7Lc_0lx{~teO33GsJjl1@|!5ceYQe
zOJPYU@x!}m#1g+~yf%AC))%jelZ=y!kn@TOdLtd-$Kb<<OQ-XE&nD>e++g0VB}=T_
zZI7F|5SIYFH(;`_@C-kC+XyvPmPrAn0kBF(3js0?99|Z>#BOj5aE)uDWkedPbX?jx
z`fv159o8Pp`F-=u=+r6(3g{xL@V*VVq#wTm1SP}Qki@E9*AM2Ai!}eD?v(APObyY4
z1$dfr@ge?nF&(hr&v!m2<?`kbUlBwN7khj_e(*2Ab9N??f}g@yLf@k~QmQnqu6l_R
zh7t$9c9C7B^PF?sIF6>E^83Rt+B5avBbCbnGBAdU%nRx>NRC<ZT<9t;@CJ&Edt1QM
zF-Dq*yLY-ankmPQf%n|HA3PKq19EFn$r(UbPJuv$_>e_AcwF*>F0*j<$9RjH@S)D$
znx-!{QAHYaPLeI#IxO1*sVZ=6U`IJ|Gt_j@5Tj;1>czm?JVB|-hw9oK?c+Dt8Q9c)
z2u4uRx38By_xJFtv$1&J=jn{<yn`-tMCcysyymmZ!Qv@Sw$Kk2vPqMEJqQH1=z&r4
z=3L`oB~%!XvWJ9~+Kjv6pC>E+VV~`FFs2)kJnb$6P&itVb$rtJ<P43INq{%EN#PcQ
zm|EKrTmY9Q!P8`c-=JO^01w9g0`w%TGq@NJ**mg1MCiYsujADELngSV?ltNbP&E+;
z=M$H&4RG%YE?y>BgAGew<qHXo#f?r&(Ma&>WB$t;p>9|=2I&x7WeYmt?RzXwPxZwu
z%E*%;K&`3JOBJmmbR#11zKyH0EZVd^)yS6!j}W1_XuZ+%oi$nfnOt6JWdUBOGl7^#
zNVMW;;Up1qSzVf(QDXs~d|{HESp*ovh$PI07uG!#@=$3yM$};S38<KpHKO)l-2&x-
z=|Va$(P8>8iSO*umL0QM<Uop6vk<LSUBkAsf>|%HdDeA=+`95;v+|?5cE`Z%a-*4R
zoa4QU=%Cgao^*~>l=87?qNLNo7G+QORbVVW?##74`(yuQ-36ieQ4^f!XMknZDFsX<
zK1mJ-;h7;ZArWCrxcPvwJ*SwY3U2}Hh8=&g0TN*4g{jJ_TF1Ttfw0rz1cWz`otuY{
zL*r{6f2<0>-_AGti~QVs`Wsm<B%Go#R(uvE#S@pv2=C5VLOUR7(~)EVltZFuAofk$
zW~e@syl~A--9%ppI*qbNzUT#(FEPE_;ZM|<CY9rWA>a#yPeMNW|7ihYSQA*IObvp~
z{%-g_a%IQf;|2LAT}!tNi=U5$&H4Gc&KLYzF%&)h>(lh7W`W=g7FJa2<JJ9jGNyWd
zytOSS*Mn9zyfCql3_oL8`YnfQhpp(dwZWE69%Q8E5_6QINLpy0eYs#2&CTIMpZ-<R
z&?&)E`-S&$_Ic<E-w&FXi#182ZHJEz-wbc^nmrqiLWOR>9fHy$&+g!nBBt7h?YO8~
zX7#TBcTVb$o5I7#w~pTb6Wmnis_EI7#z}s%gueexRr;K?_^93AIPN{kX!U{4{)ige
z%K#-*8oKnkl<N6|x*V1ePo=hY+LWVxi%>tBU96Cw5eapS;UARk-}?2R!xH2aau(%4
zz*Z}_lE9pbN}Ee)I!!EH{UEmzL8r#FJpSn-*H6h0eC2G}0w`0GDBG?zLPW)v=qZgD
zCVwQRoGcxFK)XOM|1kgCfD+3id5zsy6b?&L^L7J=@Xe#Rw<Cp_PDal8eb<(q`$Q3Q
zeM;4ecbj+GR8-;9*H-lsk>;q>oEFY$AE<+=l(4ah-p<=$qQ5eZdVxbFsEy%<h`B&z
zM{E4@3#SR6e8JSuGZ$C)N}crN-ybI~>>pt5=^qywU79VtL}r|GV`8eRIy_UnCq&`*
zIhUTf3`Ln8;~9M8^};-ME@~jFGhB7hn5pcYkvLJuC9T5AX3f{%Z1d@SuPgK7;uwj{
zYT_Au#)`zq*{-+R(PLN?t(C>FdZ=cA(u1ORd3stNM)K#}G@ABGQ>%tQmL}Gif(IGQ
zEfl0yjfm%@6;L9ixa`Sh2D7W8HMNa{2{f-xdj5=e_NL;8q;LCZ@hH%>4?<cv?c56B
za~2HiHKy`~RES@&a`a20BIJ~WOyqv6nrk=Y)SYW!p=Ob2rSy%^d^wkd`F?GBO4@7f
zG$Rua-VvTMj3W76+Ju6Ux@OMTP-o9Hh`w&I>)D8!;%yL6JNVYUP};L<@?zQykf7%h
z`<nPft=3A{W^YoVAox3aEc%zo`=l}DPd1TM!QVY6Iso_$y^a6Isgd1DgV@a?MB3Zr
zhRS2NX4ACU(&M{OXhH-u00ZJum7H)=nlJa7G&^V-F@iQuPU6?#S|p99u*P;fWzjKf
z0*>QQ-hM7-cea1`@n0_01s1PCt{>_Zv2&KT<l~|0x$QhoY<H*KM2~m$BH1gCtbZ9E
z+)uFXW13y`KUohwtr2!?^;@cG=d{Dmz%tlLupJ(QL}LWNMgp9o&<1)@xS8hlh=TB=
z%2UwqlL*IMt?a30B4CTuC-!dL4=T0yJpSA9n`io1*T}JtszTu)E?i>}yBRzG{QG!?
z4D(nE#K%FUHb+)C2^9Zh)?*dNw2FEluTNlgw56RB5ur?tkvQ|!*W(=ZEd_IgPB6~J
z8%*BXkP8g8xNe<qWJvdi`O80A`W;Oj!q!AWAE&T*Zr%NJS#qy%C|v7cc9FS$7g(p-
z3=(KosVm;guX7InrmxR<JBztkI}ILs*UJHtl;AhE)Q@YW4h#080BZ<uJAQCM*jpd3
zSm!z5Cd|1u@cTO|e)9vCYo>R%kM18-?YbT%bcOiGtb-TvX7#$ny840jVQ27JGXY{4
zcFXCn)rrL!N;6$yZnp7|8`SUzHvjaCYG35uj__Z@?vI>xr*T|S1phsTU0>rQIqk`c
zYBv?IQBbIldlL5KR&oxDD+suh8s+j`)$I!+nsS+%yip4zk5?+$0~ni}sK0$^jB*8+
zBdr*;RT?+BRPB3_--y=7o#}V8>6TQ0Dz;V1y+qa=hON!9QSa@@^y5iS$GQHFHB=bY
zd~bon{^x891LxI2uaoWKCsVcK#oBeNDK$*LS1M)QZ`Vg;yTlP%Zk7%=f-8ouC1-?U
zlWGig55`|v+h!wE&+l7}yYikwXS^m7_XNk0?T(_*-sKTwYs3=@qVzEC)}#moVJndg
z$yEJlSo;ObSWY*N8?wV;LIkVY52`r&B<K<+`xbuN+R0%Qo%PWAnjByO8nJoFhbHml
zQ{~RC+}bN?jZ+f$o74aM{dCmOcxYWE1YjTa>nfX*TUd@)hOd`$zc@h*D8cyBD9Kg!
zOr@F-C__79dyEqobKKyEKlMJ*rN0Ia?~tJo%iWgX@D$8@r8<5`@Vk*Lml|rrZ(iQ%
zCQ~w>8WZJWMu?8wx>BjaYyy;NJY4OpDuprZbhTy8^AhE&qHV|Bc`K92=3eftpVWMH
zuxY@qAdGOxm=waPByL01Bahb3iFAF+v|idE(yo(rJYe&2?V@1|2(FX+A_Au;btg*S
zM@0mYCZ^cJo)fP5@E4UMEGtcW0Y_xRG?%SNdIZz3G_JarcUgYpmlB~2i6$D4!-0d2
zt&71>?bZ7qncW`4r@r5=n0-F?jl{pkOtigW(%&g=SmCl|MK1C+@-VplGlBCm)baZ2
zkKtW-v(iz{K_3!0D3+y1izF0^H(GL15*iI6>M8b~%?+|qBaXYO$Jmg{TjwkEdw-EU
z`59uNMv6JQbLa*W?S7U`Yj;;J9$Rzndn>jtI%e2)V!4p07iBYAGkM0-)53@H)RUz1
zYx&?;x~x%?BVBXRO4Ox{0b4(RobJ>oq~ya?AC_{6$6X21V^#}3%d&p>6CFJ5ML`N4
z<pgZ}=%<NS>OHn7HK$#xvWp5iS3j_*^}Sb*_Y@CJui78!-S0iC{MT-Y&)d)ZFG6v!
z-m@J>uZ+fup2Jahh1`oV<i1>fNk9HcGHMaeY0uj3JlfJ!v0RXYgV0u%Zzl>;-kMML
z`C`EFDbVaR!8L3(a2zg-F4lD3zU$uBai!=_uukcs)E)7-b^Vx<MSVILFZ!Lizh(Yv
zT8*l0uj0NZ*XZQ(;mO6KETt2#wcoC<`1eM5cj|$0u<IB2ggeo)GduP7$~&e5pSe}I
z=;H(>efmm=Ba7JH=}Gmi^I5eI>09_AY30X%7mM~fdL(xKd{4`Fz`XH=3-1)E*rH6i
z?U#N+)ICsQ`&g;X@>(lI5CKbK&FAdBZbMF|=cpTpx$o|1yEwlHpAu!uVc<092tjJd
z!rOWtq43RNRf)3{Ua}8v)^M{{H~m*gc9*85@Rg-C-!(Pytr5<W?h11(h0$_ZUVerU
zc^DO9&f<@2yZOy(gB$g`smw<V?j^^!F5IgZx{tnhnU-C#m#mLsiw4%v)RiW;g^EM^
z_$c5|p4ivYC!H%)tDAb4))WKvY_ra41Hr_0=W6Vi>T0WNl19cN*{oHhZg4iZcLr3N
zI`Z;Zusel85vk5tW9N4JP^gSn6`UyCwx*S49YbwzY{}7|HQL9?HF-1Xp}%EDI@GAD
z!<y2xfiBLp<!4syQ~mD%jMK*HVDTBIRu3SYaNVJgP&w4vAJesnremEkKMyfdBb7-X
ze5Fkkdi*6?`2}MpUr3{5eIprNx1jZpvxV<Z_MSzfE0u)k?4>p4C=D1jj|@x}OYbqM
z4Rd}-EAlt978yi7G4JuVx!%gF{URBN%l%JqG8aP=PsYlM7^zcy&lj08^1NS-_;^WX
zT$SK0BQs2(^y?%VC4##JMxV%5a3oUd&F(IvKMWyPn6Z`DzSZkJH;*cL?aI_F-^&ej
z(~VXy6?1aZ>?n=5M_eyV{2gjna3(dzIl{SR_e51<Pj;klls|YGhB%4E+;pFuIc{C2
zKF|+mB2^+?X>~eo2>hlec>1&f(`}p0*)%;f3w#Nj6frEPZvy82c6&X#QMyKjR(Dp;
zJtO;23*q`Xy9a%OCV8~~j#Ltxk)V2baj?Ob4&fJf7F*DK7TDlO4Xo7%LlE;b8XmqM
zZXP(VD!^0yvo^;kUS$Y&`G4t*T>UPN(;brC&f5sFyTs#W%O48}XDzSKfd<b!!KJ2+
z#p0E1EI5yOV+7SU5Utv^1jy@-&~xS93RkkE=Wv*NyMe=Kdd=C`%nMv;etDuQYt{gr
zzw_yhM1%zQ!3lPL%ZHYCzgFKJkBVb9i$DDvz}XSy_n=4QckUmh@OoGE-8Xz$!_&^$
z34g)?jz@vv0}dLyXi|git>|hvWiXbL2GTUBP4EL*V`SQb&GluvMZ(Z0`$n>;J)S0o
z1sxUED%q+qj13p{i@mXag@~{4Ug&Bcy*qWmyDRc<`D55h4+pjVyP59(zmbnmbII;V
zVq!?g+pQvxWrDC{ZQFCF4cXg%?a%~LUNPLAbl;;EcVGDzlr+c5_t>3kL9X`QO+!;l
znTKp)%pn(af1<d7vv4rEV9AFg)PKz|uKP^oN7X&)rd{m&7uFp9VSfe=^Nv5-nG?^&
zoGugewD+R1R2(U9^dmRr+R6Yfp>UH|^;~<W5_J*W8JQQhc+%S%;Mm4Z<@@%^(jMmg
zS~b9!VtBdIK=Q8$KLeZfQ1+s}4rT68W#G_r)WFxuqZ0a9F1Vf~70+v?rV77(%~dFl
zF5t^?(5?BULXCJ6PNPVh9uvJ18929R0Yhk}#{gjttsxM+9rFZ(Zz$|X$34}s3DM&|
zpVhzJ{lOQ|bZ0clllZ@@{o*ZG&qVgrN4C#>B8;nuw$TVpwnJ#M(i^-@lGZ^yHgN~C
z2$@!l^qyS9#>rq<rNM}9UfQ@VxL6wX^Q5zK5+rMQc$7$GP4CMHJ~A|#DR9~@Tgwb}
zQU>E9G-r9dW`D3-5qJZdlTQ4Xm20Kk8bH*(1p1**=5fbo8Nz$^cb+E3B6BgeG0dYK
z;d^|I8ZWB1B8FtIeR(IqS6I^v1|5eP!n^K%aCehR-=WLXMDgpy!z(e(FbR|i2YeB+
zk%e?<)bA%fTt9G&qhVWt5w@_9z-cFa5>VoVnrj>}Q}1jYUvgloxfQc@D%Gp*%;0!b
zBFYw-L$w=l^>_AWXa+;ybkcs-phLMwWJ^VR-k&b%yeMEpDzN%ts=GP5>qA+>*>q{p
zJ68?VW)O87XQKbLF2#V^I@{EJ(TO)+q)~NE|7)f~gp``_#MH@Bj=0v*-F3iNQ5!pA
zi<`bPI0h@prN7BHlq52~v_GTuJD+zt|Jrl?eZW8lcnj)NB&yPQQ|iuLaKGTAM)Dol
zTChsz%XL1%vAX{)7juL5`4}?Gx$Ai1Y<OyU9pe8uM0tJlv2!5S?(WHoS3?2{_0%+v
zBSQP&<ox&gJtQ0AnoBBVvpmWQH&4dR-zs8NN@pHobH4ZSam@02#%kU&!G*atg>kI9
zk>?>XF<sN4MbfBN6~&(-_FyUc0GzW^j6m1p`F`2?r~mn7_QIiu!_I1YVrfX*x9hcW
zmE*g8JhOA^A(QVcfmlZ>k%qBf-|}nyc=F@3S^_we`32|XK{u;~$&!R`hfh-9;lM*q
z7~xJBUz5$7qjJDDNyEH_+`2K%X<j*Lcd1qtOUizZ3oeDS6qKhC0%)GHRXpAYmnJzk
zqKpK|jGY>Ft%j%lx)i9t^&uKMbv@GChL@q^S$mbj&Y0B{>Ww+4dyrg)<^&B?uXHJd
znL*P)PpXB*ww)Pr$+JsR8`Rje{CzVLNuw#Ri^$WCi`l~hUl9XK1vDYu1l?GSFN>kK
z8G_W`SMrPG2tdPhkg)cVXyR92Lhhk2dH2!Tt(C75!<^b=IW4&!55Uoe5L16XeuzSn
zw-zaKJ>z-zKH}d??{(~bApDRMZc}?gtto8Cr2UKk-(ySoY-sfBI@ZVi<Fa<Txfs$a
zO8yFWTr~{n^B7yeU;%%>&))$tH8gtH1_|7L_<?-WzXd2GTP%iJj(+6DQ*U$-*g4wq
z%{miAC-y0E{Xk?#G!Uv#IpyIdKjmk?P!GA>#6RO1@UFv*MxUksW*<o#@=cST&(FD%
z@oXXE-h()QX@cN%u-59?@6ICf4159jwaaDs)vJj+wk}}&f|pNxwHwZNj?b>FhG$T&
zzx^)YuWPx4PG`b%7oEIRI@uvh;L>QOFP!v3+oWDIQZB$9c#b;E-Iq8=cnOi2R{BaV
zS+JcNYFyOi#b-)-e!v!KkY2wmZY-)hXuv^-2q}g$Du7t0xma8M&#E~as7j<*DBP7b
zEA`!0X|r54=Oc^-ZeNCEv9vV0KY%>-*82p#`j>1-1^S7MU(R!MV6A)(Tu*b1rBABB
z8NlMx<uSJ;?SE6^Phz?v^20Ki11ktypo;t!y|1J3>PdK1{~%sLls+oS9+;Ndx!|5i
zt@|p4ejLT4AT`(jGo*T|Op`vFqO+&me$|5<<xVyi0xs7#DEn&Jg8VN{O|NFpA&Psa
z>iIl#2!PR&FrhcGB)}(xnw~;bLOchtV^#l0>}Z6L3Ll|eY`;WwsZWx!<rA)b(#~gH
z`q8MoTyTy>djbzrp%1)*S~A3jjmx4u`6AYK8uZV9G#FWK%i%BOG$vIRxPBSzGqx_Z
zUyF!A3)7cAte2AK6Us2i$3`;{D^;br$&BV6W7HzkOzE5cly}H<rXY%}+Cw9yk-XG(
z^7qd4)fGnWjfBwODc|aQGkD<LoUMOWumQ}6d6Fi?8wqE)!}?hx`V!Gxh5OlC=bE#D
z%dSdOCHDJo&AtX=h#jHOz>i*{w#0-iWw=CyRG(3EAi&a1&zd!K>(mxirevfK61fE`
z$R>XMTOPaM4p-EIuX5?r-OTOhlqu4xM5YztN9l)Ct%nz{oS_jLwy}?Nt9g&~;-8LH
zoEk4Nz9-P?e^W@pcKOtELA8~7$pZeSv`aNm3<iBKguI5vqYUe+B+w@;K9@<~7-59(
zo@X)0Fx_j^8<K#gR;3cOZt4y(w)Bm#KP?vSQ_~Go?zbnL1*r-?C}dhGdx<Y)Y5(xI
zqgIpu+D{Duvk{AMt&Hy#=cDiOeFyb8wGuSsqfQZVP$8!wsH%Gf%#S|&d8y1nE&f4g
z(L@9ZBJmww8h|BMXQkf}g42S1Xo>O#_gKtxgRUwy_qrhlBNx3hh($3oa}dcujL~!V
z>vl^m7<Kdfcnt)+*7=K2ybpqh3#^PFXh@j;60xa&kB#bA0jZpVf_~V;tl{wg_u`LD
zXP%dxXZe%`EN3>8eV_-FY1<!>_d%zfQB(+8vCJcUxH+|A5fSswAIaIqGp;J8)sB98
z;n;p@7@bm5(@fBp)eF_xYZc}FVSb9UR#UgnVe<Y6Z?vq=JDy3Rc%0qbxsmo2<yVpD
z8Ht?pnC+I=h<Dbosi3Xv;Wxzik1`!UhL8_mwOyjiRGQS66KP?}<41ImpQ#O;!iA2B
z?0gI_6!DT9TZ;|4x??fao4kk#fpoP^uhvjk1KaK*|L-qU+@hUg(6btKnySgNgcuh)
zQQsW4JxM+WSdgU*BEB}JaqpMVX*_OabHUJ*jRvunNtIpmh`<rBB#j0tRn@x)`4wyg
zeg+I-=eLg5Lr#Z+_~8)}lZhcp@;B+a3%vhsbrv5Ak2&m)%MHEmbG-={J8{FiS(W=)
zLOVT{4nZH3QLem@KIR)A8z!rd`;xhPHtc6=6owr0)h{|Va*MkygkB;O-tO+oIaSeB
zMiwkE03r0J%CyE9C6<G6y}`y3xA?*geoU1H+pB#E>EiqmqckQ=<~}$6o|ezGUn)pZ
za|fTlguV)XqaeXC<I+%+%Ve{s`rU$~$k6O(d4?2o)eE(lZz-NI6pyA`rqo+v4hL6(
zIr$b5xkDB=i%LtVKuDRH`-9hy-w^xv3RIn+*Hk|df7jG&3y&H~>^Je!u=>Qt`SN+X
zNqg$ak{3kx4GGo^eaH`jPJ$kERcjsE1RYY*T=5M5v{=6}@y&rg|1s!siQLbukDbG^
zkyyv7=LqbGA-eh4a@6rO&%Y3$y5J2|obQj5Vr~-;NPFgb5xz0>68F0G+&{2fo#xs$
zy#4)ohtR!)by5%he`tEks5siLX*3WB8eD?AyE_DTcMI+=L4pN$ch|uJ1b5fq76#Wr
zgS*4&`+3jzV-{=iV`f%&&%XAqs$Hcly$we4E7e64Gv46$z-8(Hk)_UB!axD+HF#DA
zN*ZLMMxckA!4>CUsxpp-j$4o~|I}vcw4x@U3=skj!CNu6+uCoH>q4<e^6$K&MP!?g
zUji;bB0-3#TC{hkA@Bc&^LJUwL<ScoQS)uM;n6>u{HPFv6D`6%9+x<3bFn5nv>t*`
zOsZ!aa;wQgpc__LIRbU<mUEFTE36ulYZqcip6C2WH`bh0@PRv0;)?7#v>w_Q3LsS`
z#hC94bM*ooKyqesG2wxUC-`c`MUf{G2RYX8hd)h}gjnIU|C}TS0@&;zgC{JOoUbV`
zQ05Fs8|ChxJLo`I=d3I{s5W~=SHA29-(XCNE|BjB$-nO2JQIm+)w-``E>;~NoUfm_
zR^->Q#x=*<t|0}d+#0x9q~wQPuIF)&I}06B2Wil?SCnLz=G-Gp0Y|WWq6}bgW?|6D
zw6ch>p&^ffEfOtV=#r=5(teJubvyZAn?FUy_)gZVSMzwWN8Azp<=fg9<s=z3N?-}h
z1sI2#8<}ThDJ!B-?ymu<PI2nb%WX~toE;9It{Hp#oARfq`WSVqe7^FfT)e6K8$5EC
zVCy3{J5w}rFzY%TdTevylV?Eq6XYGO;o1g<`1_en{r$UCl+%l_IUCH*Q-A1BthzwJ
zyU2O2RlCj-WfnuOAVKXX9Nsu0wCYPD&D8|Q(`iA2?}ihW%NWQt!Tsq0tyf>-mvX*$
z!>}6Oc$?#our<He&!Lk_{S4BdEsh@^SFgXAm%&8FOfLgne+g+~P}w446lGHPHMQRC
zWV(lo8X|X82mr4*k{3ya0OwjGun5%OwxeynSj(DMJcc30D}J^S{SE(G2_T|dm6Xe{
zwf~2*dp&z;Z;1BHAf>VFn%ai)i~!i{e&cp)ea^X*KJ39)I%|h5FekmXfZoL*t5t#>
z5a|Y%3M#iJ`c0(iu0AA{IJgtrkY9WYAbrVYNsv&;N^1$HKOcQ{xt=tSro9IRnd^0{
zX^L!~ZskplvfdqT7+O(~3s{2dLKP1?4yG3B8$a!(QUXKdB^zl?Q=UE~%SCry7F!Lz
zy+TS>y3};0&!>eSnf-PKy1t1R0_`nb$ND*WMe6tilg?k&OB^vqRrJ7tdXQ4`eBG6<
z*%2jvF>LCkxtw!%deWM>Ac(#cUBp`Bg#kOL+?FX8Q2zq)LaWgs)6|<19dEJl(6}qv
z_7gQ%lWTkB_b_my9IYQhW5{&nBvymQcnuO_?39lOD%xM(auJ^Z`{09z-vOu((g8Tj
z&pk-DH#{nM(@iduqL1y1x*9|9VAsh#li@bU7e|Bq#*@j_!^nSxxb1YpI;FAtpK!%Q
zqS?O9kIjaCwG+iv1<lJ-%)-x-Xfdj1JA4pDHupKI-RJ0_`qOItS4>5Bq|x@lF#lm5
zmC93@6v6jAmSi`e^&tQl*x+5-XDVai&X@j`2{z=p*ko`xAU9pRbM&REqCyD6uW@D<
z+?%N7Idc_Q?TYd*Fb8swf~nekR9wvFB-lHXSu2=;jyYEY2$)`%)KIhJ&wFQGRLEK~
zZ(|?xnwC-$(InJjN<|3yur7AdUCk`5X+V2E<zQdK=;z-Qv8d$$IbaGQ-NIW;%*k)b
zyCtW$VYyKcK||&k$>p!T$UA{*K{z&<;|KT?Q5Xz5Ugpy|dde-an-t<ww!P=J$)E#n
zHyj7#zXFeI0;9(SPKE~Mn!_9uSm8*)a^HHaXz_lZ-YU8Nc&)<5KPUzbAv{$SrOP_T
z>4TAC^)|B4n`jZBhEpxNW~`)?dgkgXo&Lk8JP|XK%=J_|c-ug!yH})pS6knIm5+G=
zU3msiT>Ff{A&r#;d>1FZg(q$uj{lPT9Ksz=*eeh;QA}KpaSB3QL^xqVJ`UZl;OQJD
z+BAI;c~@FHvr>9z{@`&>b#jKP`uf~6a(Ms#@dk9C_^14<qr*hiANQMm7K~BWSZ+N{
zh)jxcgNi$cq?yow0o$MN6E`YJj`#@k)3{<19g$6<ISDqaX%)ZDAH6BFUnw>QLtPZo
z%&XOcbO{Jr^@jDn&+!A2h&<itOVPm@s{=j*UPmRf0)3Ww9THgT<J^!HUf-XpHFW7P
zr>*r>o*Uh;OGURS!(A27t@-rBO%}85o<x1C?O)COkWs*{AeC<bzPX{=*>P6D;JaMR
z33yYCi`pDfv1t_4`}55)r4@ci5@<;C$C2*1+h-bv8%_nxNmC>=tK<GjDsKTWrhh!5
zLR4$@EqUK^SeJKiJ8!+cLqWb2wrEZ9`I#k)iKiowXjZm}1RVrejQd+Gu>g)UpOGMc
zzH9V5x?{x2k1rG_6|b;4=~bzpq^5<scv0%XtWcRyvRPBLuEpmbU$0B!s;V)*_WL4B
z3D7e|;_!aVwV1)`T{EB1leU4LZwA#xT^kZrQ>}cW!$$1W64|aZ>F>f<J5bg@`6agd
ztr6^|BSjFu0-F+1at@bev~z#*%9U3oeE$@UYRuEqzq$V8FARQe(jkfDmbkh4zCkfG
z$_p6~x!yfO<#;g+2p)ZZ3h(7c^^Nu3UvzuUeuuuzh4m~A1MJAK$Eib;TG3Kl7KbM*
z_9Bn~uT^#3X$}HijdF~oLtClP{mpMqt-_9J9+b0Jy`ewlw$J)i@T4s$n*<PW0;9W9
zx+;hG!_fe-2s)W_O|eX4k<09LlMLnbSwP74&c6PSO}JCC9*--IwT*|^pJnz!6<7}+
zq(1}1NZ#t9vjvlxCw?kl!*H=g0m7_5m1a^E=0`dHNK$!>DbHfixTUn470$U(MUuS-
zMh($oBMMmHQb6Uf{HHKudep;S5GYy3wuA7MkyE+G!hpCWWn5CNluFt{Vd?E*DkpnB
z4D6*wrON;G#)X1G*E%!q<ycIbigE0BGj$BmMHtK|bLy}_^$MH7Vp#`v5M@8vAgd>0
z;A6Jx10j#jh3OZ6e+<6?45KyMl(d>M3{dLKMjRR$|DP717+LbBe_l<H_3#{K!L_p~
zx!k(05N6oh)z4#F%eqe$50|t0mNJRst-Aul{`>qP5WkX4FKZ^4_7fwC9KBIEY|z1-
zbHG>f;(udh=lqvFex(W~<aD^@U;`9ZGUH$?6Cu6+H*%#FI`lHt9~DIaFC-Ulh@p=f
zj5>>0-W-V370x0Hm$_eHp8>v8xX@<BP9}LiZS6`Y?v5X2hLF4}>%JGiCy6$Xk}i~y
z{d)cDYKQ+C!f<{>g+a3OSJvhOY-YsdQm`blr&KTPyv*g4KHTT)L%+?H-cFaB`i5z-
zW9a<#cuX~5g=azb4`$!i7VeRPqZ;g_)vfwkHA^wfE<d8hkIbm%FH&`B@>SsqG*x?v
zFAH*cH58vR<J5VF8{Qar;c%g+;OcjQ7&1^VGdB1f2hfX7Rhg_j$y{I{Y}}HA-Ii_n
zZ2*G83-13owf<HQ!ql+(?+TT@CCw#zq>>sXx$YMq>wfjL)cSe^h}KYMcZGbzPc{jS
zklO$zF{Z5wRhLp$`lK)B`(nn`3COOKphB?^dQzKX27Uf`;-9&rQU<6*i_NF5vtb7m
z3R8FQ23c$L^9-i(_M%j4(ovmb69VHmvefmht3Zu2;}}2(7Yc8Uy)fvBi=*%JDbx4j
z3z%R=5u1o9BfET;OW1g&cW72OiCdj*!nlVMh9vS-gPtAFTIgX*RjvLTZF9TTK<9pS
z^w3^5mW7clkOboc&Rc<Q9VSEYh1}#{zw*2bn70OyH5hJg#T&gRzZ|xmAJBl`F<qH1
z$sFYo`lM(on-9j>+NZ=dP?q8RF$uJVWge#kLiRb{kww}*%EonEh@N#nthTv4>eqpc
zdl&(n!9A5y@ydF(fZ}mw8Vk5z91e->D=CC2aU*F+d~`5ypzVE^fm@pa`Ecx(vQHOS
zk;7r-bp{zfKgW&1$JvUqsNN*;u2f*Tr;;bKZQ1v5-mp*36Yw>xsO!fgvCSM`ri(l_
zZ@iC><ZtmlDk>a*!QlOaHS+!DUSLrX*slFz8O6TXD2<I$JJm3&sM<fHNSVF<h1s>X
z)*A0rUJ0il*`jx150pjitfh%~xxrS$goNo}q}Ft!<v)-`sT#B`wDFTUUhZ#P8gk<>
zZ9D?hIdr6&y9Q-Y%M%?-&b0UYqhlj`LeMw*X-Zddg9y(bL@2}R-!_b#b(VPz2ml0{
z*ofhf5}FEGxq8q3J0`K6+I}@n$k)<>bQ|stg!z@()~^<86eCK?N~%#_sQ;*=y4uJR
zp{a_U8hABHr5WN}PI))w&(ycS;hV4JH5*hDwv<sp$$)yjl;MNiazk*ZJiSddsx(Uf
z2Lc<xRI$K*0!0q~)nb8cNcWdCpD28mB$wqf5UQkeZ?xpClRXB}a3f+=lgbls6f_7|
z5@1ph%TVWkg_mjSeRvR{iBG;VdZ)cZ-#tTkdHMM$@+unEbNU-V(*ZH9@+lxL!8@DK
z$MFOr_7v#D<&1?EGtiv5x6%XaUlkgbWy^CV8DO__1DGx3YS-m!q2lBRL@g3wLLw?R
zBq0EqjAq<X_Sg=|-)+l(Cm9{<C|iAS+#23`eL=b(`674;Wf==<$vH&YFiS(d`$m0K
zlC)l7@+&{5CbKc-IE9CwKKCV8Psu?DIWne=#0CFid$C_u=r&YEDuB*7z@(Mi1{e--
z6eo;S)wYCUqO85|CJwe8I{(Kpuy)$PZ{+6YdfKl(KR=%@HEZ)mul#US)W^q#C#zjo
zo}!hBMFRMm2nld!6QR~8CfHQZMC&om<Iw(330{M9%A6t2a#%60L$vA!1Uwql3AT^z
zLF|K<^|@ng%=&XCard*`h0S$H0)MdL<Jl=MR?17$5iknBSY-JOSq?D9{a(*)p~V*1
zYm1|Fzd&iY`J_c&kWyg#F<+qE%eJc^Uw92-$Fx}W9pjroWqF$6He=!%UQ1c-{jl_V
z%~#+ur*IE`MP#ZK#i+7Mu>DbCo<jZLV_PZf;6e;eGqYikXcQhq|9-rDIEb}{bj#I!
zzHj73+U67V!S}px)A;whpf+$E`xPWx(?b$492IT-U3mDK;M;ir;Gvvv#_0v*yg9wf
z-#|Hi{vGpqv5cwd(*71BBx3Q9R$T}8AymMvy~4a9LW`m;ZZd`(9eAe5B&d!f+pi3G
zAO)PA{El8n1?D}C5g%j@Z9~(Nx<XC1KT49r!-s_5id`;|4utidR*)?O0W_pFpbu}z
zv%{d%J<sXNbK~&QKM$Bw@UrQlM9)(yE!HMYkoB@<vPzK#jf&b7Rno!(`x+$?+a6*+
zQ=mctIly)k1i-&{rq^51J`$yG-s(aCGqq#Ebq%${R_96gMgz8u<jTX}U1>D&H&K)B
z8l&Xfp{FPn@DnSC${Dn>EXjX|8(1%&eGBuuD}HP|{lsN+cQKH>ZNkDh5W=mNw#7rR
zrPDdL1`FWh7WqXd-J>n`Ga>wwK4&r9(5D8^D`Fu*c&jE{+eHK6vBqJ$VW_KtRV$0)
znVZc5!K}ZtY2svF_8Da9NLsR#;fC<EM|`yU%uERPzcs4<W&g)>M?*mgwxvPDjXfP(
z*=l4puyi?5T&ZZ)WNWZ05&-`M@q%Qz6s(kLivW|O62KG;sWHa$CVbl+|CC85knbQp
z_HgmN=56WyE^WH{@$vnm|G1HPNBRZF;;`?2!Onk<$boKwuRxd}AHNj9d3m2pL6LLo
z_yB8164|9D5yVYKwAbt;#v6mIM?5Rew;YpNJQkiCKhY+#tWCCrO7LUiy+5vNTsi!H
zQSQ9B+Vt51h!K&|fV)JIUghibjBgJ=YKVFMQahW?6@?i8$}bEQm``)W52dG(_~nu(
zfuJ;&T}%(wZs4xLOu@Rr7=8Tq^fi9nzx2=|YBf3RmfV{*QKW*X*kE0o!I#ez%nO7I
z3O8x9SAPaw{^hO_?Wou~6JxU>Qj`0(#5kqETmAnqZC0M_>};d7)o<I3I1ZE=&_N@g
zj{JPZi(u@$`)NEZCt}UIk}n?0d*P-VUDh(29x@BlWtdSpYGSTHsiK6+B+R2@8nVps
zuC}-jho^Ya;O)*4xV~yx0->LF@_F$1$C=z|3J7^3i$ZQPz-sGW-7nx<&VyZ5pe0!5
z9O=C+iwBw_w}C8H&Oc_~=_xhMT*xY&IEWmu;|NBqP;d;0&2L<X+TsACr;gtOd9;0d
zqGLt3FVWgG3VPD}g3(}Az4*F<G>X3!STX#Fh9k4~9cxA!YY5yOcw&gdAd>`o>ZW3L
zpjM`gge_;Q2pj1}d@F*2KrhTSJZ3w+obuczV1FRH$i9EW5fOZ0aK~(lVULtM8I}%c
zQzBo|*V(5gy@t(>ZxxT%Ut%Fjy%gk*=&qWlehaWD7I9eSrzfcX0M^tvP{JQ)RsZ`5
z1NJyKN0*>lwT6fbNI>fJSYxdgD%RL`E7zArgTgEtlCU8XmU+r7k)f(kD>M5Zdu`6I
zm-B_hI+sSC*cbb81HAg<J^KtN`!!?`?o+7Ls0_3tfL=%LuBc9~{QwA@HKxxMKZ9YS
zG{oV}^`wk!ckU@vp>cZ|aW&<JL5F8m+vTo=DQPYaZ5;RZSp0GM)c6r#1ZdGVjVlya
z#d~x)kSv4_V<^=mt|`ujAPvfttqWwIE*y{v==GAHX8k?GLEHZ(u8>*|{#Yy-+S>+#
z+39~J@y4P#V3W7igtVS_>ow*1pKs@F($F>*<jZ~&b|utETY>3B1(NcPA^dSBwvwdW
zvm2hmpPGb($ENK^!h273#{msds2i*|#-?=3jXVnml&G`YoGXpP@*!o<&ev?ukN-c2
zH6&IEGwpzb>^7-tqg)-PhJ8)$QTiF|132^{GG%hct$eCNssq!SHYl;2mNoi5bm_~B
z@+86xRs>y(S{vFQcqIG0$U6jg_}nt(15L5VyQY6(+h8k-HJKNnTo5bx9ylpEKJ`~c
zFaF+HLy!40MklI5WtuEi+l>4XBU6*GQDMbxATTjPR*IvS{$gzu+E0R$7}se{|D5Kw
zftmti;{ZRIK|*K}uf6geFO!GneIt>SHKD*erp>2Ikv@^<xPS?oap7<a0-7IaU&p6-
zXC|@JpIuhUoAR2(I2;J=B&&`9zR|;y2@VvQe>o>7N6K5XcI6%5v7TTz+Gdg0H&wMK
zCc=ja#(@ToODI6#XmYowtavi@S)`}I*b0d9<C(yZ#$uacky)C+yu)?rcW)>LED|?;
zi~o<OMmKu%oW<VIw9Xl6l?jo^pZf}Vv8l_T36eF-Y#}}_EQV43ZP6&q>O|D!!f0RM
zNSG@D2-&2}ARe0;Oa3**F2YZ7{=lLe0K);SaL0H*)nb%_3=VVc7oL=1AV1WO*@G94
z!<7stS38WZp`XK{i2nN4l=p|e-w8OTJ{AgEvdG3)15SH<_Tq*Tpt-8zhc2sSZwP+H
z9xPApa7G-$1S_yG;Yd(=q$BiHB$Ys486=T9lXYuz^MV^<%E0V^zin=rqmUX53WIpj
z4jb}y$f^GnX7P*Rtn9lOxn>Y~BIgu`gJ2|Lr}IOD)8oAnqsXYYE=L;dcLD$!f2^!#
z7K6vY8v*O%0wPdHLhXz-w{EYC9qTpTI6FYBl|w(DiwbmGx|j|~Kn;tlEbl+1iZZEO
z`pw8Q^0_JWG;sf^b-2CPZd|3gIJA&6PGdyXXYXTi9Jv=OcbM0WX3*tq?^pRz7outf
z=35R7x-Yv?Xlefl(WU49eq;viU;Zw;e}7ADd3kg2U1+)9MzYhvMS_^r0Cw1Bwyi%5
zC?YD;B;*^>QmAWtY9r=cYm#$hH@TUaWjbUv7R7Fsau?EGC2s~}`#gh!D*>4NPX3~B
z*xPIRF^F8%3)jn42qN~rJ%$xg_yY_}Oj2qf;%J>?p(U+s#zdi;R}nJ)8Ut>h_ZbM@
z65sk~GCFdOA9L-FA=;B%BzZ&09BX#M=;J0Xl6+DURknBy$8Bfn(5V&XO<79<O9s=W
zVU`Q%La>3gg^DCEbWIsO$<RdHKu2;@ih<DQI{`aE$V$=?F9xEZvw};CJiw5()*I+B
z{ro@K6s?Mwb6vh`VY;brFH3fZbn4z6=Tl4iFdIF_r?ygdwjA{c-lQJZJnv(`x5Dx1
z@L)`e#-!M6S^o?3eeyG-w^#(1s_*)-;<WK9Kom+h!2X_Kpl;gMHaf$IxXrmwG+6ti
zk75Ui%07EC1ZzOS!qegHVbGYa&&z5e`@UiqeGmK<&IKfg+8pyrM_uOuDMOd`eU?1i
zW2BJG*q}f0NR3z4==qsJ)>f2}{)0VlO(Z_@&kL4)4f(IDzW*ZMgRZ+$KC|%FAU8E}
zYO&f7#kvrJ<3uZuELAnGclD%f1bL{r!a){x%Nio07KA?0+)9+PeW1SOgWEk(ISWFd
zXdEX4K{hg}gUmxb-bJl~6JDlF0p>)vdq2V6fOS@<yNZjO8fHkZPy98b!v|O%#Q{|&
z7vYPW^8fT^|B>}r3_+fuEOfhiiL`i37^Puj{P+Y2->9hMhRwf=YeZCl*b9>XSkxtc
zg$KTwgd9Er5$BM&d--Zt3aa5hNY?>{022MdXvuCrsYx~9XPf7deozfKVC|{<V3Twg
zC6dz{FU*m6B}m?`6S%18ar@A@nN?IDl-pa27c`~S$5Ogf4$@@;)qTzjV4eWfP=G_6
zXI^!-QvyB^Sm2jD2KJTPIPB7gJQTXL)tFO9;+nBjrV?$2%z{(ybB=T9N%zNRs?mzd
z_U}j`Y9hcAN|o9-lAjs~QkkW7X<eZf`V)Wr9@7GhLcZ25ZPz{{XA@1rvD_`4Q3I0~
zCm1s=HgJ3(rsMjPiIAO*NB%qgSBH<WpRHv1$Y>n^)m@$rLjrlJDFMkBmS6>XNSm-t
z26vntW+-LC(F4(FQ%N@6F8R~K)at{!ysv!(p~lwbO`6S!H~A4|f}JAIf;E*_eY%<N
zP5M+yfy|Dn)6Ivlmadzo#`FNLqw4t1BOj}&o(S)Kp0CT~fv2Kl$pgp-A3ZJ2sGDMm
ze}1cHwX{)0<7H&Zj6+=1K!gvRgNb7p*GsJ3P(9p>0!*dkx9%n``Yn};C?kW4<y2OC
zOI#<arO~&TIjavMn{ph<NqDooSfFQjAYPV$po8`tm6JxPXBdd>6o}@2iV-2VGS}0u
zAqFdC_ddpz&ioG0USLlZP)}O<JF)uJh7kxzt^W4osSAY+kY-D3Y8>}0zybz;#D@of
zI#Ck<bccPeButS3<ueOrXYDMNBB!UkxYOLl&C@ac#u)!U(Bwez%p{iQLD)*SkC^uV
z{a_79VdOO(d>u8$+s!)i>wB#E7@4wjpGW@b6D%?5=jlx^`wmhfO29%|vuAjLWUX;2
z>3r_DG48Mx54BNXQ?}vy$xX{#NWwF?me)6$hN2DBvdp3nP?cBaueC2uMVx>IsqEg&
zPc*1%V9SZw&m6X8m<e#Oirh_1Q-~LXzgTer+yfgUez-D~X(qj{sNs!tx~viMtl{%?
z`f6<ZZMDTB9}6{btLTT{&$c|&nTR_1=T1IVhb}<duf3$(ciqPqO8&>F-R8NR9$MgI
zZ<W_IGP5c(nNWNAtCNCEon~_^xjSdo3S_X-;gSH+v+0I&jSsk9PQI^!05(^m&Ili;
zy6l!Md`C>g0_bp;Sf7)Sh&ZP5!6cb!PDdP0hFcAucEX-2{jPpGhhGqhBcR4(uQ{pp
z-u&(gc3)T5S$s@={#)Fb>x<el+?GdNrwsc_w8#W-&O}T{OkD+4sB4*xgpA0+jMTpl
z_)8b>&#2-h2p_Y!fatleGs+~w{=NXzI9R>~5pb*IzQ~)5Idhfx6WUvJD+4uMb1<>m
zJ1)%yu>HJ?iicFQ*>-bF;W%}f7-&V>sahCxLSw-$E;%bmfo^PA%=1abrF*Fm{A$x}
zx6{^bK-<%ior9w}KkQb^szbutqrr8PD<$s?<sK0=GqC5iKID%-D$b-tfl`H<0v7){
zP;XI8Af$dCfdeUOci%(0<yV5Q3%zY61vX5w-%dBh7U2L-Fjtp^9YfyJf6`ejq!F>f
zqG_zvasUzi3Plev?N9;Dw8)N+aHCz9#`D9o{MPmgV#L}2eVs`k-Q+Ax_xSYs9(_M4
zEcd&6T*KW9bT8~*^q%V*=a6!d;y_rqsHS2qnJgiuCliSkXn@3ghY60h#zsR!Mw--1
zbaq>tQQUo9Ohz+!M-7OedE0g(t3V-s%4@UljXi%jKAHI3F+$Z)xqruK60ngZ7x{Tc
z8IX+vcE}B^aW-ab%yG|1P$NwT3lP953fKeiW$`y^@j0TCb?lNR<J(l^r6scS){sLY
zR~2T2UN5*H&NG|_CQXXag8DpJ2W2K>;t+{m7l-A1f=_MjEQNw0GE$qFTrytfOFONX
zgbQg3RDdA8v-$7TwxA5UjHUvc5ew%tk~}@-x`vzCS^?<mn6;3)_2{0-7|_TBPFC;~
zu>%~XO3N(6Lo2(B!1z#W3N(bUi_&aGk64*~4D6<Y-~3~=LXIaSS*4wa#!<N&_kZ0T
z!^=Ao2{bY9;{5Ii3&%@n&-n0qd}52)ifQry63M>X^czatptqE+YyRA|Xo47kV#i?t
zZAMJ?Aq2^#ecym)+dvzTviVy!l92t`c+dU&es<moGFw!B=eMj^wkfs8i2_Rw)!{In
zj`zZeLZCnAmEwftl@(*s+wsve>B+`1JZljfq7Jzb4Olt+9DhyeKpOZvplb8S!njN3
za4&1!X5g%-Fq?M6r)C^doHO>r7IHZl=6>P)JhjfVe$e$rt3yy&&PCW+BRyrfxJeID
z5m-J=JfgKs5Ua0^Q$A(>vYX^jPu$KtqsBUc)xfWGm~&2j|N3&NIvnahR^{^Ld{!-Q
zpFj}#)YY_x8-;8KYn@d6yc{0X90(X?s9IRtSABJ5Qg+iM#*P6m^)f(`Aj51ESO@J-
z6LlW*w8AS!jVDup0oP>I+JDOU?gHTGHR}5kKl1Vr|IRCN?<O*GdC4{{c#XVy#kASa
z<2TP!NX)7HD*3e(R!{P(FrN<16i))%&z*xx_WR8<BgT-PG_b@4tWYfJzyp*v96m0s
zetc5WNEn;Ki6|O$(o&68L-xggd1EGNntNUER;Nv69Qo>Qw}SIuDavCIi;9Y*Vk5;d
zV7ataeh+7455i1(Ibnn=@UTUScSepvrg=X^A<mM9sG>u`!n6~#XXkZP+VPf6wT?9T
z2UPw0S94bRkJ*~lX6nHE$-8oT!zZ`tp3myMY2&tg(<wo$ni%gR1?W#3w@~*$Qqq9=
zNJjhVuWamp-8te;3?^1ggAMa$IXl={wh3Eu;S$oVUyY>Laje;!sWGCG%*bFh%0AD+
z;P@O<w;Eo?o@x5DuT`_>Qt6;#;wO$SISdWoD4J-6fh@FQLetF4;4ttHrv<#!Fz6Dj
zrSQ?4dQ%>e$ecs)8qTs2E1Jw!WXGO}Nj&3DW)JXHE0=UX<Jk?J<RYrfhmmsa)h_Cf
z);lj=5}>24NQP-wQP`A39Jxx%XK4COr!`QsbeqjO9^-wiEoNJW8Qwa+*a?*!M#mD0
z)}sOCBy#qr$k<@vFV@2+{q<eEW`@<C5%D0=7QBJON?K7^74HqZGc)kN(0i*m&~7Wd
zPcI({;^lf8klbvf3Na?NQ=j6kI=j@Jha@*Kd1#0en&4=KT{DXoEpcGaWpi2Tr?EO{
z28T2Y$h<`n=y<G3%eJh0Pk4VCU@_lhcu$%=p}$+?=M*<{<#Da$AmVM#j57P=N-Wep
zX*@-(%DNI!5sJ{v8hA##(2Byy)V^Da>l?$|k~8ajOqmtrG@I{P-i#@=itb~Feo6@%
zNk?w28BN2+@xzO5>cUF_Q;+-qm~(OWJ|Ziz;Jn5+>5rPLb43w@AHxe5y3Oe8%;K<n
zM$28&;C{ekdliVAEK$yORg|XQCCZL5!*P9b<tF&yN6nt3R1}k#Lq+?@E&jMXhmC|@
z8uc0*(RCY9!P@7ku&*_oW^EF~2Ejt%Hq&FPswiQk^ZGlG@*`tPAqoZW?Yrp^;kTf0
zk!PK4#p4LiqdAa#dOlMeY|RYT*(;&l*`xd*`hLA5t8DiZ?|6%EB;52f#Y*;u;IAut
zGJE;j8Xj>YW)pgaBL#Ul3j+!!&Ac+P((l`FX!k|{A%SCWgcJ>y7gk`9tN4zAW=Xgf
zsrb$jN2!_>Jo!c1*0q^ZHugkQK&g(Y&v;5kUDh~&!qU&^Da9QBpNX81%+akwmdltI
zLuo}J%fSqa8GaptqrjY)2?keXZiwq}>EWnuDvO@D53}?JtV~Hosf}b5@Tlu11#vJz
z_ExbBqOtnslB9n-)fMAknpY1b6^&=fS@tz^G)pom#3c?2U}z)$Xc$2iu{Jgm+8wRq
zz9pqnh=r<ll8?$8nk({LB7CM)Lsj=sz<-QUVm~z^g~2?{P#u5JsbYS<&J+JqrqGC4
zL<v&6stqO8q&AT;-c=UVS8=wt%}w|?i3Z}-^30hBu4eKe9m!VL)fj_pQt0P|I4OA1
z3JufKm8P8UU|j>}KG>A;J*)C?5rSs><$kV3gJAyty4zW4lD+?4!vs#d5R!M^y!#V@
zbR-k@cVC;j#*=9n_7rZ=$QaMC44ghfv&C%s_sF>fX(<8TIcq=qYFRFKa<>zCtyCQ~
zXRF^YB$8+hS8<Y~JbiyjxdYM1F=#pIqAFO2()-Onk<3k{;xNcD*QuTSmgX!C{*s0k
zA}c4PX6_)moW4s0u{QdQ!AQwB-LHhzj26P**90R*aPJN2T`{p3;-I>#nB@_WNU<%F
zP``|QkWikIL{t=#u^FQ4W|C>5K9C_t#r!!NWl88(KvphI*F*}%^rvHZ+k^C6^)9wz
zK!(F|QzL?<S{Jl1Er|hI=AJciq9|pnF6=DYp$IYWByv%ITj=C@S-T{dyV$xoURqG$
z97vG^7hw(rhFR*4Qk_qt9MU=Xqh9l=r2A#pvt(M2Fw%WV=92%q>=t3lk+H^ut$JFb
z107W}Zp~4IO&5m590_K$v)?9)f5mCe(jQkUX#H`O@64FJn)Z%}vvMd6M}6)2J%0Qv
z*z9}!^uH+SDWLkmY&Kg$x5*<i<Hyp#Cg-L^dPMYp$-xrJUJ%05N?lZ4^AZ|;la3(n
zNgj<vAD$B&t&Bdr5s5dPOfjVROt^ZeF%l_ds(|3R0FBX$jbnvuZV6pQW#n3mr;3DK
zp;U*-QSeRI1;KREoW?@7&qQ}tTnV)b6I6%B$iSN-BwLAnOKQ>EOxN0skoTNqD<BC^
zv6<vawLoe?&$;ih?2Wv)qXALKj*V$j|G{C&cKqiQQk}8g8BG&Ah&UnS-N<y^u_=OW
zS`k94G|G$9Y*9D^Q&&VJxujN*!j?|J+=`65hSz|v&+4{86Q!2rcbieCWM9c(5P1Ga
z*CN^_NEhBN{BJ+<FZv)$bDU2osy7P8UY)sURER*OCjw1E_o{5<db++FlnHbLm1Dd?
zcC<ye`F8UVSa2&oz*E<!--o?_6NZTUVe!257p?F>vM_r#7dy$MzFpEyJ#4_C$vU?Z
z{)2C<UGl$tFWt*iY9gAC!=$D$^IgsG39IN&PEtymL&n~V3ASUP#~Zz>m!54XV@OxX
ze2udmlsQ(pIFGPu)S{%&SGiS{XpU*vRY(~5)S7M5uaS^&PYGNKTN<il7hV<A%zq(`
z0*Q(4axsEG&vg_U&E&dESRKConw`gRf#&l1BqvjqiC7bg$CA%BAO9BUoBsc_03IcO
zS}lpuMi7E0iBgj2KkE)MIQ7>NjdkyS<)A(<IlQ_V(nOeA!l}BWq+b<Z>p`!g5N=6U
zac?jynry!01}Z=7^p`uEQVRk8%b3gyGC%)Z>Jm6&@h0k8*RDU?9B8gCX5q98CmAQT
zzbq@e(#E*&1|jgd7}{VJH@GErBq+t<(aFqQ;u8F7&N&SCq!cC7`qJ?Ju9CEY@}8_T
zBMA<}zKUwcPfqE~hD!8hFI7jsj9%$o2t`gkHeU95t}B%NqPDlUlV^VAGsQ{;$NTP5
z(1N_Kvo|O7H=ncwr_J2mQ{w$j%<Na2V|(6?d-f?v9{)6KwD2xGG!G>uNzYZdfdkKH
z1@6pF%9T=1t&}$<*G`ROHqgyWOr7tUknKBi8;*-Ps|lYO4r;$2?GeHol)w(a;)4FV
zjK_npXrWG3lnJ;HqPk}bHCHk|xSN_;Tk4u>)wa?clQM~d6E(`X8abh*;nT1UOVDmC
zsDAyZ|I3eVb_x>$Z2>KomfGJ~p`T&<vVy|wxYnm)1?MM%z4Ur)?c_`3`mmOH-xxv{
z4$fF`p2oN_r^Fi}Kcr+__2CdVyBD_$qSG}E*z{gYYs*M19Z*_@c~q;$qvxew+L0qf
zVjQQ%LB%%kSP-M;XB)V@ls{3jK;a~p%3`aSV`T7#vH5AHi@Z<uF_iHBn)(~~L77`Z
znQ>{VoMWN~Ycs6sJG|+TmvmAclL3L&ea6W=g-sX&o|QjEd|pfmi`{&T8t_IRy{+^x
z7gIXvK;?*m=Q`9gWn9gTU;eCO8Oml*cb?UWWb?Elg&)Xi$GwZ3F&W9N67@^6F2@~?
zdU-VusmvB8v27NTsscakH0<2WD?u`2-Pz(KmkjHXMshPTYyDx`S`km3^W|^#k(Sdo
zg6(B}$xRU;*!Wk7g9jMXcRCcNzL_+;NfrN2L6Q?wQb5BUk#16E(i#d!C~^)ZEmZ}T
z(<h7U)fT$z+9RwozAh=kev=+?`khb54ISQ$off7<PbCh;bsv1p>$ZztuO+5q(6=<E
zt#WVlwTjEzyH$hDEE{=yOQm^)kP=@mZ?J`Ga^}kzuNrVyld>Y^sLYbx7+q^9wXU<P
zZq4mef`Vfs{FU95P94(-UxYKKz9c4Fb)FM)qjp6xsV5Z%$x}hlKvGt}lWEl~+<Rgj
zA5P8<GYHdmc=eo^XD!j`BGl7%?T7iP?d}t>;4^IDrvs(r2{8FUWs`upotb`6VJAni
zov?wkmFyuUA_rhY^O4W9&-Kp%`}(CK@YBdstlx&|^7O-C&4Wgo^*JWl`F5jcyj?Y2
z7R838SkB_Z2GRD9-XS6x@wxJdi(!_M|MnF!XSZ~kEBlL>oJo}G_+n58j=@azW`Y-l
zooQvpAg4GC0#`A%SDlX%HlY23S46YZD=gyqSe|mXo?rHw9}sR!m?n?*EpI~dobtXB
zm7<f-q;lCP-piET*i)zko8j&+|4ByFXggFPI|7;aY7BOf#B7{;2xxh6#w$*;9LuFr
zb)MZGZ`D+A^g?CRLNMTDAG6&t{`Zq444Pf(WTiUu#;e8D`lF=l<+fbS(x}^3Ov^m~
z>%YtcU$hWpkqc#WVC%4IYTK{LsOUDx$l}UddwW&*Z9azs=v(fjiH29%805|vBZGq&
z5Y$Jzf5bS5_G#&CzH?R!2uC|=qZ+*{M4nKclif#cW^l|1;}%aw)Ojd4xd|{hy&`qy
zdODYoiCK&y{V{kDh*@UY`bjdG`2iyLhmZR(K@(GChw1q}k*g2Fv5Fm(kHx(6Oa6_f
z`3tp^zBN<3=mxW4dk&}4<A?KTqZiX1UTS3Vv!3_2%tgb+y$OzK7~|{xNf=SqhSN|M
zUoxAltHa{>DTTb6jx+x5$4RHFiviRI7s#;HJ*5)+<`cL@W<Eupxh=%A6GFuC8};+&
z%SdJ4pq8H3(>1qgAwof~guuW+t0pIxN;=*o&0^2mo#yzR&;{MqbjMfxGTcSy&0$pM
zGfxr4wmLuXotsq8z__PxST<ncUGzDfBjJ05+T@$mVAR|T-YE7owIb(pnOIo#6q|(z
z-$)TL-Q9wC7!vVbG4wFiW=h`TK+2u_hjF)<qZJ7TDT>2;2gDlVs@x*9l8R*($!+-P
zpyJ1&+3BmN(PJwshqEqh|0}*K+-}&{{T_){-TF96JAKaLlunw_wjhrx-9@jAd8c>Q
zZ^ykSn1Wsu>FvQU$L%7!qy4%+htqLHtE?#TeMkDUQ2xSCd}u%~<?O%VBiPc!r`BU@
zBDK&Q;#eu4W7n%zQyY~jjq-kTy`_1`?Ftxd2UEB;Hvlu4G~631QK|*JXpUGDCeT#`
z^Wn-*PnFt-&{a^;`A~QKV2*J#{8SA~F`1B?43%dlckwH5v`EO0RJ<uSBSx95RwCp8
zp1;o0&<g3}<j8^jvkVNWthddrz>EB}fQp>@rC|x^XgmuxsFMNDL9)>-nW(eskjSh(
z38)N6FP|_$%bG}hGcn%2?8cp`GjZ6fG#~L*$Lgaduj-Wv#dByGNzXIv-*K{nt3$TM
z_>cX@yA_9<|H$+m#Rnvg=e;mq{rP!^snT^o;-c(BH`L-6R;0}tM(-IA_4fRZk~oEu
z$Mk%J`$gY_4#M()SsxlGnOXJ6#`Ch?X%>==e4u0>RNZV>A54!&fW&4d%udqnp!!GV
z>CRy<J>W+5DZ2p|7kt*8*l?<R8T+|MzW9Kvk$kAQ*ui!lNfXhTiB<t_DLvhwH0W=J
zY}irbH(m}XusA3})t3D$FksM^Qd#pOZmQ0o&EV8D&ir%!(}I`HSXOL7`pFGt+nxGO
zhGV#2!-`T4lR?pG*reR}32fom4#{SorcwiCBC2Q2-UyAchb2)>pM8?fqY5H1LX~xX
zA%dC}lxcj#={)Fq!#l&K^drddoySecmE9fWmcdNQTPW{m)f4}Ik2NWA*}pAlb{erB
zw`<Hnb{;q=h)EDwgbuQHm`&#}+o)^C>ybCXalNid6vxaMM9Uri`+lp9K=K-I_Ea)J
zeY*M{XnpJeu)%SsCTX);6A}VaHeRNkllv|duv<$|u+wB_NCrq_c(UNMs;UQtPi#JA
z<ffnsX%ujyOXD0OX!81cblyLA{`h{Igv*LQi^z|~^m$IZ9)$vX>o|hJ|4E?gH?f`j
z@t~rX;0<4t&za>;zw$h!H_E9y|K*S02<2gvbJPp!TiD*JW7*@l5#>2fxreE##c(s8
zKBL^!jxocIv2{bqp_uClc_|Ik6MNEV#i!Wgjo>vumWK2!Q8@KR$2X*+5vQ`3yx(M<
zFD{#&^fE3UYkg1p!q)8l@%`)XW@!Ph+7Zl;>ZMZw1d($i)=&q2;jUU&*Xl+G0WrQ$
z8Qd^gxZM7|c*-v!$~6yPJ}#%53GtNa7`D56hw<)x>}*;0ATarIHO$kE&Wc$y9q@K-
zzt-wfnxbttiK#5u=d=0z&9UdzV{DI|XTysk@1V~6+H&J@3TT&K?=eT%SbUbpmzwNw
zYo0gV<|>-@!w4KdY*jn?9kZ=AEHnQ0p0sPd942TxuIcca&)&GHx>ME*>RxZDKl33m
z7AAO9^Mm{SFzdu&^f`fK!D}Fx#L`?h&i(P%vMv(eQq-3*fM_vyje|`qATGTrMUkt(
z3pd<K@PNtHNv<2pv6M?gb4yFNC~LSwsm!Qn!TZzkGt`XkPAT*+KLXX}j!=j2aA-2m
zwEi-dkTa9|t*RW5k0K|fc1I;(NA}S>TW*+o&4pyT1G27TdlplZXPtuT@Q7w4bWY-3
z?u?(;2vkgK4tD-p>8?6~YE^vDwG~F*zeY2Pkwh295?1~hc(XKZhY_9r!xC73)Si|d
zCofHBf%LzQCIwAIg=%+g3=jTC7B5zNToOy>Ai4^MM~p@8%%W3?Svjm~neAnqCeqT}
z-eHhr+LP_K$WN@F{0W|XftlSrVI~sWdy)JAJ}VKSs$znSf+yKxMV9NMZgvwKiQ(TS
zUW&B|V=yR%ZH~4X7TYMvlQNS1oXfyjSMTP~&|J1_uGWv;SXrn9y~NxhIteJkQXH+s
z^yC(e*RIl>)jvD!3e|u)trRS)^CM+Hp8__UJ0MPRRfOaGSz6zz><-nb;C=Be_u}nc
z{!KbxbQ+WOaMK6wx2@si9~MKnjFVWUqNDLB?^r@8*BS)Z=Ayr?QdWfIAQ|;q(C9F_
zSZxcj>g6fycb8y|21>kbO_6Ryk*q8?5|N5)hWJ-v&wa}0rvPF!EMHkFGYZ&{3d<^N
zbyUn|W%mn0*=jSKk1=sA)SBoEO8rSiBV$QCT^DBC0!ejR6B@?#rMWT?naE!AlYPTT
z7A8jdnMk!TA40jelWX+;FEv<#vTbxWQgkkebL^JcethA9Ib8*fQ*OCu52fR7p%tVi
z%Ij<K3O}_WGNKzzN24-Ysjqkq6U;QI3t1)C!2wiWFCH3xSS!iV{K6a8$ySWLb_jp?
zx#gre0-HJK#oJF_bf3A3L-fk&9>dL3?Z>($Ij{h5P!<DtJTV#iDTi<h8s(W3yzB9I
zbf8zEHxrl0$rLzYH@}FrbYEM!ee#T_<h%HC%F-3j#PWy?4>V)-#rSnyIA1C_O;)B9
z#)+DOBYtG4-7uzM3H}hA#54&fRdQw;pD<%|b0Tq~<}3;kn?`HtxglwLs6kD1I?rz)
ziN81OM`d~VVzFsbg`Mn1<+OBr|KA6d2tLxL3%&HvrTL7;ufa{Alt6RL*X{Jdxpdz@
z+k2mG&csY|7{dSIw}kTX9@U>w`5dDl#=*T0?q)*RhrfWv^39&}eE-t+yaD@+b>12O
z)YPMR6gAEK&|NV%xFBTTx+e~N3@X`7Ro64>QqPV1P9QD`<`$E8lG_UERaBxf38uSn
zeLmj8_JGe@LO6?{Q=k9$|3+(5KA_~-WVv1r;kp3@#BEj**JXsrlhBNmI>dM&?DN`>
z<7xwsWsCp)1yCiyi{6hpF2i^)TW}&<RSi?7$KdLim>B1IMV?Efo?2eRig)_Iwp{(J
zFviRE7H(&5eXzLBkXRnq^R9d0W2g;-74rq6#R-Bnp$V5}JH_{a7nF{41DU?@`KTCK
z-D%apGKtKSjcN!{(qEV>56x9CidFA#2O~EnT>MIrr+Wt+nS99>lIeY&F8M-XoaAe+
z1fzNR1;yO;R<LJyze{v<Z5sVX)P3njSAr}}r}{i$9{=v-jILS*bs&Xu6w#bfKD>z}
zdyo3)YTY5EvUy!ej<wzUxo>Qpkuj!TWQD)|B+0#fPi$qjE}e{1(3+^SOQvUxGtBj8
zPNXkBuqVr~rcr35C>4`L{L4@0V@GyNYf_nvQJTeaB>y0_DWC8W1B8VzoI)8KOTo!!
zv?G(CgJm%4B`}PUb($=7XVSU_zy}QD#AUY!bwywoE^Y_;+N#~U#$9x3#Lt5^d0p!0
zKs$UH%vM|-5@bil1kqZ{cl+aFs>>T*t^+k~lmyDDs|#h#%xoTI{kD0moE@_q4gL<R
zj>9t-1CKeW*D`ehc^CC^7QmOauih<7jXsjlEv<ryR4r93m46@amaq}+z;>%8YG7Be
zg02=VN-U1AJ(`_XV40f;*pxvL7CpA>5q(R{o;Y_7)K@5M$3711DSn%apNqlLvy;o?
z{A7QvayKb9VJ=WgOo`=3psd$m?(r)Z<}*hU?)4p)s(tJA{r=}O;?Yo%H<7i@8}ccG
zE?=z=D-DWws{_SM)cSNoF+;i^+jOIiMNM}_)FvATnoBYg2P^7=vEc2;znBb1B1|Eg
z;AuP3MI?oQhM>ozH#+e~h48ZTVjSFu)I%K*1eX^|1knH2(!iDmh0sucgdlC{r1%o7
zvIiFE8Q&-jNYvGdbZn99rw^eE{@}P}=|b%vR(GQOm(z(dDLR^-B9wjUF`?pr7Fto~
z5#%7QJWQs*MRAJ;f=z?%W`buU#?~ES7Zv`Te40{~Twk@T>P)3tnQ=ndC?-ipfj1#w
z&JX!}uET5X1%E#(pvrJ&V*B8mr8<$kzzsPK(Xc6DefuC=d09ezE=w(dAX<cHH$uG?
zLkiIutVsMhU-8xK1s((T8J`VHtKk5&SZ^|NTP5Npka|a);?9?_*^ByYH2&uTRK@bt
zNU<~zkG8P>^3r`AFL3@rr~Y2VgvHI4kVMh)ng?(7LB#kFS#XtHGiugyP(q*(yR_<j
zoRPL8g;p`i`o#MM8D(3+zW6=TDMuy0Ki@$4>I&7Ah0krnFXZQo(EHEaMqEc^G(FYM
zV-~dOfJ?U39Syg)v>u*kqOx3rm955+o(~6pJFVx-0MyJ(Gxi_W{PuKgC5J0=-!@Ie
z@yLfUUjJ}nB{Du7OW^&m9sW3`-QAC^Z4e0I^?jPp->)BM{<t-6^mY}niZC-c2=l?u
z&+qzMv^c-7fX9g!_~|?RLD^#Az>}<vywvBNeBapN>kZky<OT)*4@DYv4*k3tXO*D<
zR+PCxnJ#=cQdJ|wYPWg)vu=#Kxoa0wr|p(f8aAqksb#M9u0_V~+0t_-EM86Eb&^o$
zJzK$@Q=qbt?zo<^S^}LWtH|jx>+<*~1ohL06;J=P399o#5j#L)pL`l}lX#wJb0$=H
ztcG$|=X%3TTy7xd1TH+}E78mDD+O}|W=NHkKis)#<v}UnG%y((Q4C@P=_Mr9!OCAV
zV@_tACPP6^;XzzL@HK9%3}z|1UN&;F+;C}1y3W^S;7G%J{cB6pmzI=dHEGV^crvbX
zVIL-s_oE)A>2c2`zONN2dz#*CVAwyuFDSTC!K12{@rj*IeSW1|)zR7ox41Dx@gtuZ
zaw%8GKQ_83=M{K!n?7$jQNC)E&Ri9TQ2QS@^H^u^Bf4|&Rv&)1M6Y=dJGpMo5L|>v
z_OXH7y{Cs8VD)JJ2B*jkD-32y$$wZ3Lh|(AyC5J6e@M*!iTO-qN%BJeJoWr}m|}D;
zNx;mVha~MMHdQ>S>Pa@D4RGaXTo^@jY)gwmcxM@z_^N2Vk-o{ZHHYrJk>q`4=~G;-
z+PqbKx$W_vXr@(_F&QI+L(<(<75ToI#>gP&u>kE_vla8kn8yk#Y_>rDqMulVgWtd@
z<lxMQU4o2F$bGY?vzlD!3XQj3gVl;8nBPGt@98_gr$M8cwXRO~Tq^d1oqLFazC_UG
zhId2{9WIMam&ROtlcV_l$$(<*&9AiEnMzEg3z7L({YwRr*P-7~KV&*EN1RJ$S(i%<
z?v>BV68RUGf=A4W$`Eif_!VFgp7GJ)R@wBB=)3&}gKA9=G-S}|#1hp`DHR$`cm5vE
zS=l3T<zj8?KsVj`SbyK4P9MMD5NmuPw-pVcfECJ9)TH6W9P*;+ucNHPNHKtr%)w(y
zWuu?cl9}l<GX2N_=Zo4Ucdf!><Df5$W{fH+jJD8Wy9qY#p8_w3UBc+T&hW1sx}U89
z{w|v>cS+`VuF#KzAZ$~&0A`xz)qk_A*1rXOo&u6Kss_wyOq>QXdh*y3ay@4kn_-$c
zQ{|9MoL=K4Zz~>W6jN~FFZ8~8ta>WbZpHq5=_WA^S2xzSbU^&|OdBWUYnlrkwdy2w
ztn7~4)Zu>QI``jfVe~YYZuusR?$GxA7(B^;7sXKXVe91@HT`AH?|z-8E_|B#wjCjl
z*}ODqGy}>NOR~hKeXHcs!O3T}GjQuyO?wpZvfY0V6$=zAtTleG>^L`aem13x5$tVN
z)K<OYxVgWbNf)VtYTq+XrsK@)9KXJjygU7obM5q|nKSi}<8Z4@Qh49d)Bn~I5dF4+
zboeXdSfl6u7vW1l54-0FS-Y`#!*LLW{_<teG#mDEOcMpxsDe#%eo2)>u2-9{-nzVO
z<8104VMKPyIQPciNL1<%-v%x3u#94{3PKC;C1O6Y=JD<dg?pqV$wbTXja(SpsIKTL
zhx-aE3kUe-R#_ocb18FDRkNwLcYBooq+r_7036L@tUPHg4fI5#wD4D(YC)mDO)4>q
zRR+h%=Wt-BOEd1B0cf@4LC=(rPRKhy^&{VuuI_f{UaGF8hVtUN{|6neDlhkv(Hh%q
zsT7qctObw3-LdrI^JXP4=yXS4u2fg_pGrAuvuq3ux*TZf%;fq7EG^D9o&s_<Lr0CH
z>5{W8k?NuIXgRm9M#Z(at8Q`>=1KhNA}=F<_C<*WUxjq^)R<du_F@8c{G4*Ti-y7q
z5ec4LhYJ+!45a7obK&9~E+|QNP%<kVa=cfDs!k^r<$p6Okj%_u{>K(b2o&LgLgw*1
z2Vk?ib4IehVeT45fV)IMO%dHFS|o$nkPi1+JQKd!s?EMhFR!+joJ_8<STr1+WoU*^
zT*2yN8FvWLeK|uIP}RJJ=yXN2prpmBqe<7<FfwlBRGiuQWO3$n0lIy&_l2W4r}mtG
zfBRExX@)$jGPi2L1&tM3Xks(@ZU7{s5N46WOzSki_i)H*`H!LJvN5P5JjmP?!Vtd<
zj>E3j3%4y+?3nLM#*tK3!ypRk5K+EJc1j@RY`?ne{qb_}4xK5#!BG}g>3{&%5Nzaj
zHS4gIQiCFkFomerHPeU-0RM?4wXo+;b}!gUO0+a$g&OKK+YQyLtN8zD`pT#_yQb@6
z#ih8rySuwfpm=e23dOxhaR}~Ck>Fn3-QC^Y9lqSpyS^XS%CD?sCFjiS*?Z5-$PH0O
zK8+<di#Iz;@)s>#Nz+J$pku@b=dK1gbg9IyT40!?f*TJwUMT_7VYCd1xhsgQLp^4|
z@8O7Zr~QXMiq0t^1p-tXu^!fz6SHGUf@t|bj4c`hZ0C{~)N~4w%&o=X9M4HEMYABQ
zP^}ru_~Ig@wsxW%Gx7U;cu7)TMXQ&JL%MB{kO78I3vg=HWA<@r#NN?-N<zHWvE@d@
z?GzNk6@_|fSBT%}=%2}DnH}b`ck<LXA$31mc?pvgAh<8V{vSt>%)5U{-L&cVI=0#9
z`!*}Yp6a<*)1vQia5+6aoiNjs0~1Ewp3T#_JRQF%s+1u&E|Xq-n;dtFF_bD$+<wrP
zAFK1nPOyJNSJZJJ4XZE{!DJx)n^?Z~dO{IaU#9OW>>(Jx^&k)+`M&y^EpA9Io#cM+
z`IziHZC(z9N<h86@cm4I<@O8s5uMup`cS8Zl{zPsNhec--4c<_O|w+$wfr4<N|%EX
z2je2pm}}tJhDWbJaJ7Ijnv^VxYn)YzPHG@2lRNHj=hn?u2fr_0N>UQihD*(x8k9Kc
z-FA4_#4vY(Is{+^i9bT9hFon0Y0Ni6k}h~*t~XV~8b5FP)IOvbX)-BjqwTjQ3ceD}
zOf%p{vBVgxZ47i1&IHwH_9Z0P=WEpQF#>dMvAG>SPn=ibM$1zL$Ib3T!cE%nrgI)~
z$6A+Wc8n;D@`)x1w$A(LTFZ#2S88+wRS)nneKUeAO*cPTFdO{{`g&&Is+O*;-5!3(
zx{Z4#)-uB^TC`zJCphCW^4^)*>pJsS^wU~F!)sbPrw2ynX*DgeYNsZBcM}AAE$UJ*
z!O}5jPcY?o-WkR(cL!@2FDQM{A(!ehrX<^LLgiT+-{_SVk_+YI>H8VS|5(3SOAPAL
zkX<g@;nTA&N8wtB{wa}b{Pkt>fes-SiR)`JQh^KJZN~9U)HKniE8o0zC|w0@(b+0z
zexh#N3<?II3tL*H0mwBO?*_8!FC*CUO?PeFfJqR@(IkMx)p`6XX^izgTV-u@sWBwW
zKwZP7K8_-WyR(e#TS24qab8|+<89H95M`J5oE1CU@u-kk^}_zJ6>7S{V5*n7kdb$$
zP$nji_SXtI`&?}}wd*J*JitDIXh7hbG0+6wHW?c<V*#;}5VVmwo=pd%ODgNk2)p<V
zlTdCrA#7X}hc9Fy8Q!Y6lpC=S5cdf4yA%;X9_InrT1l6rvqK~t?LiOn+axp1*kTrC
zk}ZK@Yeb_R`|v(x!)9WyjQ#@Dm_@?ij{Q{~V)P?NCDZ1COPszV@>Ix0_jDz@^*n+G
z&z@?o*LC9yi>UtsS&co)+{)eSORC)_?R>EOgin()G!>5UX1`zd7Zq+UO(wG+Di+jG
zi?_bmREp$`l8fdQ|DoJ?b>Xg4-Y^o<d#@~j^|oEU=BZQe*^p^$p*CZ7=EOS^FDJZU
zVXpv=wU>6}JpY7*rIS;fvgL)0j&M$nxn^-9zr&Q)KF)2ku&{2C8Brh_+Q4v>^nn>i
z;R;G=GwJ0sp|AS~a=zbI@Jjt)Bm0B*DXyNG;S0Ba^eB(N{eo@H{&vZ;NaVBCKcT2O
z0at$e1)&eBy@gBWlc#-e$L+HArZ3e@F8rSWF9N;&NIq8cJn;Calp--j+@dp-)hI3~
ze6hZPz&E+YW(;x06M-AMnHcp#RU``{Mnv%MVWyu!LnCnJ7%FQ#Etki0h(;BW?`W!E
z1=BO+VKpV5;7$ZSw3(LEy8_KIQWNn}O%&;QY!{<^CI?OKN9mcM(`}t<aF|7-y0ajx
zm$!Za{=j(`hEmr%!t=K>Agb%4b2^mr|Bb|HjE+_m@zFMFv_W=BOmDAf?Yg7$7$d>i
z8(gt#|1dvR1G72M2wg#IUO%0;$H&-m>Nq|@<5b47{9d)zd7$6J3sZDmNUG(;Ybh7L
zktrc03uiJL+uas9f!bqsTWatR#=04+QKep6{d$;$=JdhJivzeZb|xCENjGTFZ3^E;
zjD5>s=Pk51SChsnPgf7<lTQGP8GiN`?|^(>F*>jXXEgn#H<9>>vdH&wEbSOeV<DI2
zo%Hy019gFe#EI`k_(ne$Y&`f`mSCfl&jn_?y$VZyq3_IjKvBwXufjJ>fjuZbtI!+&
z1cg>qgzx45Nu;%!7I8H1j&1G->56{3k~&T%#!F-zQGy7E)M%fB$tWx2LOVZ{K?`qq
zMK(WqFmk-pnL=0pzZU>396CrJ(C___>u#_xizlSh*1Xf!4e)po*pg*ax9Ke0Qbjb8
z)?-tH^i-LO`1{LUT}eRB**=mNJQS&2hafzc<+QeOus0stQwwlMIUl3bZ0RwZU#ZkE
zsG-B#*p(c_t7b+hw=t{e3b^Tz@bAtey}uK$5>+ACwE%ylz!}q26aYUDl2bPWw8{Hj
zu~FSsKuR|VgxsJS_lxOgqoiq+_2Q%X@G>M@8+;?v38r$Y*;W*vp?eXj9M-k-SjKBD
zO3$Me?6=mxx`~3;2gLYb*PccktDqCLp!1IJ%<tP#jxVw-60OY<cD{$#tB`=Sym-Bd
zZcRnUg|usrX<ve=<Y->8=a}lxs9pJZ*Z<K{|DjbD|L4(`hZS3h=Iu}dk3)gYmaPvk
z=e-LuMRhhqyS<xL5gBgxdEn)QAjC`aC;fG#Z@+55>W#Cdp7S!-wz_i8ouzC^El@IV
z!-T8-RaQ$s*yrdJ@uCiQq~P&(Crm`{ekPWkUrkB{m2<Ehq9GhUdE>>_qt(xis#Jxl
z*%lRKsWND6_Gz%-|LE)SitW}VM10Dh@jN?0;#BuRbwTuIi0KonoLSqsQn(m$h-`u9
z*e|$Xg<hX-cXW?4hyQ~yP0`{&V8)>dKBzy<Unl`wek7>d%d__<VXvsw_Xb=`XDFuv
zCg2G5;^_kq5$rfSsHPh8?Zu4XWw|PocHj_Y5P!}<X0gvBlyaG(SClU4ywyuI*mR&a
z+N(3ClrEkQa_I<svk~i=uo=<@qdYpwI2<QZ4WHNi$)C%OotDcBRRP&V25X!C!iGQI
zjrCiCwk;AeZD$X4?^ft^hyYa?AC2h{-)V}{v(~kWjBm9wh~@(_O~JM=Ql+VIY8vQl
zWQ<l~<6pjR3lriX;!iRaZG=4X8OemnB@mN0<E?2#V=+%TJUX%!yK>LQcb0){{bo+b
z8Y@w6s)tsFr52aOaJiIO*q<IZ%fs6Cs5zr1!HN%enWMm$nw10RIJ7W`g7MUdEV5&`
z6PLF)>Gb&005~i$h93VdM8?mL9476aQ!V!{<j=^gSpTq@){ZZO!`Ytyo;pM*E@zHi
z(Zt6H`UJlp{SR|4B^%L$0}VQP(eW_vlT0d019`tJMNbn|uV)Ic8mBl0$h+}k@GrjH
zlKvdzPJChVs071Ck1xTd3XK!r5+}8XI2_ofLA}B@sz96+C4oy1P?);yvUJ9QyC4=<
zJpb!<Yshi4V>=|kofKtezMNCJl(r$mI!cR~i^VXoB3(<W#8tz7;A5O$0<XTmrw|?j
zpSsf6r$Bm%n4GAusr9CF9#YEo{TGjcblhpk!ZibGL-o?Xy9TrAM4yZ{TQ)3eq=O$f
z*-BjUe}I^aKd*@*DZkh6U@nFN2E=lw@WG{G1F7+#nJ5Rh9|DXP&>0#aTgY@}4;EM{
z<l@Cy&sssH%~4#Y;zcD-I)v72EjRm;kJwHHAxd|gq>Qk31UnPXI6Fz_(|qTL2W!bT
zw|F{(X%cV{r!TcfoYG}~ysYT|bm8t+lzHq;zI)Ea3Wq1vtCXm0EIVpDk<dXm%x+3P
zY;@{2vyG7tOzc+~%N5BsV5px{azJb>4Xf$h($lYgyf>eJsh{xS(4C#%ItLTWa=VGB
zR~?tWml<hae)!(cXL?_>WO^Q~m3zdBeDGd+Ew);Ge11H3`X5|o3Aj(Z`5qUg2vhj_
zavz<a!x-7w*%|n*OHifTE`U5WU%$o)JdT5<H;g?wc3mj#_GiQE*TpViIe1BSu+jax
zzv6)<?wW6znc>H&hWdD|o{0E?HJEkZ`(9vJ#M+i@Z@Vd<se6_k0e0tBm+ejEG%V}!
zFgM?~<?o5`$deXP#r2p8f#KYUjvNDE%|AE97EwN1)HXx3h?Xo$IhW?TFr<;Ohm&;z
z8F!phudtb&Of%gLG$mhG$upWB@fel1++f-tW7Bn^D?6P4qW9JopNhepS00nxy6(EI
z+TMC5F}dIK9Xg;LVL?M`j~lo7b7mG6WILVk>u8dHPbl0wI7T#|Gnq9Zq9E-Iw}MLs
zLEpN@k{-PFg%&vmZO<nILa527%ZFDkG6^Qcsn*OcQhva~bGSDt=>0yfRCPLLT$1!u
z7R9bPh=b$}UHmVN-GH+o^pxPGO{lyM*pERmo*IM$P7xv(H8j89uDqm^JA?9snh4!D
zWAuy{Pk2<})Dbq$HtiRZTunl#NKL^wPn<?My0>|1PoF?boMswnpX#?~%0sdq-R3(D
z&D!EsO@{T_)hp)4-C?BgBs1dd|L`c7D@vtB;WVAV9U_gLsUB{@Ho8rCXF3s698Nha
zmGNjKy_65Vxqwd-bmNPnWzw1ZP|=t;?EV-{s~<gM&Q{tngSU<Nj%6i9myiXGrr3z_
z;==s!Ehld(VvJ%jC$$D^>amZf^v6$R3^8#oSvd55@UWfCT~t=nIB?whJ)Wvfmz=_3
zmjh>AOxH-|L>L*wH2$QZ&1o6}*jf$>GF2BoNy95jCsku)(rxaT>_~YhYv=K^`an_9
zk`{7XX@x^zu;#=-h$*DmtrkGSJEimh0Bfk;JyYZd%*)mfL@A1ScB?BF!Q>Q}Kc*<K
zN3XD!j_{3+Q_iP*RXt6O(zF)qZ-h4;!_|;qaqiI&AR7??r}TzwRxGN3agWiw#s`Tu
zhufS9|Mw74tMMgIl3eN=(C2mz^VM-`!uxR=Ng>ZHPuLO$*qNRQRz@6f2GE0&%yWX0
z6k+$c#Bqm9Y8@%0va?b?m_zhF{&*@(6&pGlyy@++``YDjCYdFpnGr>9fH-W=&O7&P
zPBI*GwGDaca?kHsy*|RU&*7vq4g{Rxtl=AlKY4g94}Uxj_jKRJ_ec*cb|yuyZ#n#D
z^dtZKR&r?0;*D43cVGT|;ohl0IB2(0AI?zK24tvg4#Hy4Bn>v2%ociE?R+ONeColm
zh-p2m9rfsEFkrj=1gp$?9R8o~ZmG;Eqs^5Hm{Mn(ojLoXvA>$%r(+eTu6htzd>UJ2
z*mC}U<Udk4zV^x4yI*Z|_IN)h1q>gJ#0beBToWjtcncELb9ZFFEjxB^c%g=5eLByq
z`ga%j4^k)?M4lef@?zh!7U9#6;DO!5oV#Gpxu7Lvn~n^yf$2Y^9XG!bj23%p-H@zU
zDx5xn36m33b;U6f`YJgF!<1^Z{Z|*}aoVD+!a0dZkUd*!0~krUB%A4MzK-$vUzUeO
zahlNMh24D1(JN83-kb@N5SWe{7Hs*RGC7klJ`Y)FUek$3!5GOpy@bsXI1i*rv!Vg_
z6wDCv*Wa_|5mWb-nt+m6vC?mJ6>yhN6a~m)e3O|Ef(u8BsoJ%Zr54ye`NK^{CA(s-
zQC>5?m2%^?#ZckdqSo4vJKhf1=lGo9T+`YpFc<o5csmDdMD<pd1`LxL>*QbAMVj$T
zlq6UIilXu|X%0DAz3DLpBgesG@q1Eu5^T>i+(nq|i2prJ>Tt|Wg4JYV>b)NX3Zkj>
zCt@Qvz)_9V+KQEdM#g&fm*Zs2T!}=09sWpN{n56zYR9{2<b~s5wm+H6@ca_gY0iYT
zJ72_qB5i5R41OM$&f-x3$$Msa;=!<uXk4)(G&cAGb|+{+g)Fe$vzqHnn_D~kZH6<~
zU2_{q-(C7U^}PBw-FaQyvPi=PUId_s10{F)z(@I(P(v4NMGK9rW8ecVbW0uYjXppp
z)yvoL4tWTG?-%thw2NjkIFu$d1c3Vthz$WZKiK~>n(PeQ{6RyYqokTF=Arr4)$P{j
zAwVY2rGQ8*l-B#5RIO}z6YhAD59xqVH~wJjWbvB;`^43HasGa)Xldwp`9_avhVYOP
zyZq;dK%9Cjo-z^l=&!W|Pcn?2qRv!P9k$=(Wb&M+-oD((n9UyMF_)JPH+>#q*KXo8
zYOd&}8rZf<xPSZbgBQSc5gEiPpip<qXrS<1l4B(TZ_K+;zOD&-W}+~P)6@^pL1?>A
zS+)F*H{})cVQa(U!GoIh;P9Z_Kc`G^*jAClwTjZOVQEsHzMAr5xZT%r`+G%dUeDPi
zzarK=sj=s90gLwzmQaU18)s~|{bZ3A$3l9c&*bLB>fi9&51RS9O(ntUJmO-?Qg}>D
zqDc7#W8TB+`Xe%E5XsqgWzMH%sDlTOP;|>Z!-;_ikKt2OCSU!S&tXHhc`td0Q7rD>
zMznwbyo$aAcjoJ>5I``?U-*5ph4Mn${XdHM=YLwTI9$hdwByywVrRCV1KYB);oA<*
z^A1iY?<HG`{!jSLkB279<4NJq=gC7`Lyniy!))&hj@4sJG&1Na|M$}tut@YDW`7>v
zd>hwU705#2S$D<wf@A1&J((Y=!R!Y+zA7Bft8?rYfVFV#wgF?=j{U`PH2Q7v8zir=
zF27>OmW`Y6=kFs;MTgs_ocPD<ov-~uB1Vr_3N_Gu;l&s<0bvYGHpV@W#f^UnlLI>1
zz-&LBwIm~*7o5>AJ_Wv23sBM)gBXWA-79(pOzDC)oUUBNr7YOSu|k|9od~$IOUATg
zB$J&6F?xc^p9T{#vl;U>)DhK9ifCW!V25<mhW}9Io-NN61WTK#XDW-&i)P~E><X!q
zVPnf%2hEugKkgI$j8W=t(oT9jo+N(d4ObtR<q(4_z+2p1eVScHm$$_?9!dC#XX%Z#
zR_!H@C;0IXM@`VAn&fXfSl7N-T`a_>42~z)Kp=ybrxN5|FwroWxY%dz^xC?lrIay8
zLUh(o=%ZK`@}Wch@s`4PY?X{Tw#_Fq6OMVVJspMVT?`4<pM|2-bmJPBB036LPOJ0x
zCfxM>5Wu(qO94EMKc#bg#{U!TH|zz!MUJL1%%kDeUy#r%x@NMkbVmH}loF=AH5g#D
zo$V+@X3fV9D}$4LocNCHktAK|47gG(tDkNnT!w^gV(#eG#Sp$@I@9`C$jnIM#yfW|
zi^ty6_lzo)Pci*JwpX&WIP;gYy_}FDuC??YVR$`s&CD6-YdyIXpJ`qGk$PIkBxF5B
z|A$xbSN51py1a{8)_$tfrPgDdYZAT={gs3b-3yjBxJfqrD!DRv45c#T{G^*c?J&fY
zeYKUfs+gAYoLuE&Mu3Z=;ezp`#*mpno{s9EMhT&nB%xj3og$0$?|4(%7xch<B5Tqj
ztiP^+)gh{8lxS7;hiEr@8a2sqYDKIp$>|&yZCxk4p%y}`U`cEF-AwnLN^iDh5?^Bf
z!>m6GxGOB*m;`d5GAhl&{HY22pPaeR|2ZUsrV#@TuU?)O{%u--o^)?Iz;<L?1a4bz
z!3p9Ky!Lh0A|l)Ln)tZ5fW#oyIZlk|>ao;1lZ^gVWyppuN-8ruIsXMS6&!ujFM>1j
z5K0quaOxSCRn0-Lw=8?LP=*J015lW|^hB*geOsPUnfJ^y`#7v>=c<taH{r|op;?28
zTA2<}&(HSeZps?1JZ@Lv)(H=q*y=zphMntSZmCI1Hd_0!jWV$wrwvo<?kt{NuNsY7
zl#_-<C!_OGr^|Q27NN&2#mCZyGx3!}1F4EjJNzu@L3E@&J3;Gs8;Z-?CTw(N6`SNF
z!9q0s>SNA}k+m)gci+HyZwHg+2BJWVIlT!9UzXjx?~bPC{vu#S?j$>ef~k#Sf<}?6
zGzh9l$>$n)>M_0HZ@S;1M_2Dw5NuXkAAinZIJlmaQxK2KbWkZN@gK6Fo(!LkxqMyo
zxdqRo0b5F3#<~8plDJ&BPw*XHUB~)oj<R&Vr6z6t_p*F+QrAPU`6S)(abM-~dfvn#
zC^!Z-t5NGOn)~=<{?}JPDc93-m_0f*H8ol2AFovm4qAzgKD&F30x!HDt^r!HA`%mj
zG->s1K?|OjWZIANz%|k=8_|jc)$}z#aSKJ2ke|15#7NB2RvdKv8p<j3V_h>|4$<PF
zH_Xwj=97u=u9`j=_2;)>UKk$sgbe@*VFj(8M=j6Qh=zov&;4Ky2mH)w=ZS4(tT=5I
znY=x_j(S|na~PL!zweuhCN+sfia;m3#)9dVEJ*t!W-5-}3&{m92e<PDh0Qt2qTOuy
zNLNg9gGZVig;uFkRASb3*7Vn+&!Nj?{h+1Cr-$--WlcGL89_<MW`oW5`IgtS<LtUe
zM}Pm#%}3jCN=8m7{5*0%(6jVVw!>KE!|b6_`eJB4`-aH9U~nKM5U5YE5^uy*muLxN
z2vVk(PGS+n!H2*o;1gz*g<t6^cn=DZ9zt|4Iy<RnIVnB(p3jAn>KLkMGb&wK<H6;M
zFORP_goDx^jw5tstK!G1E+ze{=YN-(q7o}F?Cmv1af(MO$ads&wHof+KO)Of#_6Bg
zmZ_ruh~81!9O=6oCaXW3+Y&~Rg>!9++L6@Q(#A9<?-lU*zLP_<Kvp4Mw}kYUn-~%i
zq-?WLmk!JGGYW=`kheSZi%`Iiia#UT4yg)%$;41jMZ^Dow6BOH<nEk4-EF?u6AOOm
zV5f+Pr8B9AFumNen3-NGv^eMWcN_HBoxCH+6kS7C-3o_{t6mRD$3SFv76dl3mU4ni
zXv%R>@}%5U;9Hnx=iO<sP75n?Kg8&6#J)3lhnW>3z^lin<6CvP?*;EAH*=qxZ(X!4
zL69sww%50JMK)Swsftph#FcOI!*=w`^C&fkFd#1S&zG5!d?J`Tpizco3#2UBk<X+z
zE{lk0z1JsI$|{n@({lS~@)+DrPeP@<$L5|iZk$$n(wZLb`K4FLd74}aNA<XrejMpt
zk{q*s9Ba&>T&4t^7!3;UVyN@4N|^YZ{8G9obFDr2&m`w@c})z2VLOT*Gb~SgJS^-Y
z=s=KlNJPc~#shfFX^w&uO4Ki~hq%(MZ$h&##^T7=tmq0s8T$Th9LeC-;H$?fJP-1C
z#IJE4|1H>~?2glXV3I}4z-GtgxmlQc(BZofxv{K$!!Je|9z^2JF*a=rbwdHoukyBR
z0~p5*+#0gnqxf`G=*Pouuq};nf@D5k`U&$LF*uOH7#a<xTvg87$e@iUWfDai!o~5z
zv(qA^h;%Crc1ErNzI;)38<8BAjvHU6jT*?a2^;AmGR?rI@2m@pUh!t1ienqfK5|?a
zDG@EwqOS{%CigEm9F<5bxQTap9>ukZzkvg3NPp-cA=dl!TnI@L`vRNDY=eQEs%<-0
zkR(GN!XlgR%+fm18o)Hl9Z{C;g%;>T!g8JQ()fKN7UKiZa1cOmM|9dA`lw_y$KX*#
z7PS+ml&}YZGj2Y{jdEBTi{5E<vB=)elt2q>5+Ml3=kRH@<9zk$?ytzW73a=esX7P*
z`chq8-F30vn!|NGz3!0W-hTSyZBONc(tXVt7W@qIiHJCDeE}=vs~zIru%z1W2RWod
zo;cG-bKmuBe>L9R+3q={-#eK3tfD@dAdxVu;tfi<qG^GL29>#CKF6gb=4^Q?pw2q+
zu`i2x9O;3rJ8qtwBh%>Km#-2lGM7al+ktZ^LH<{Sd?%a=+4xYPngd;hF;}^osK5<<
zeui5iBPS#|)iD*Bw<Wm=ofQqlt##MfEQt1}lU#@mM<OS=Sf|t60;I`k4~RXvL2-7{
zHW3(1Pg-u98Ap2>2lyu6eQm6S38y=p`lBMw0Kc}J$UbCsZXk~-aiN;CR?$YL6nVl^
zoC6+gk-d2eB-wkJ@3gRZ-%ub77YR*s{2;@hd_}pjM$2)zlI`3>RKWWxLmAvz4nKHO
zn4j~V)?5xs$^kin^nhixs*vrrj48hk3t_|m=)7%zI!9749N3i1*6BU#pUX^yY&v8k
zVNHlkM_WOa3`{79c-5k&rK9Dfg$MA`pZ|g4one_;Fv)yOCe*Tc=|ViGHIiWvJQUj=
zMq}t6k7OG0=>fcmhr&9VeiwiwvmOKr$#s3Cc_=hchva5JyH4tYmpLvoBFT|zuX!Cs
zR4<TCl546UF>wDJH+kP5dYo*x&^hz~8UD1<s7<y8ilWm8PRArUD~X`uCWZF8z)z=?
zdUPz*{MfYH)s-TIGl5^|`EZb|T%V&M)MVcj%(R6zOQQKc(WzMo4$SGsd1-S8Wy^c|
zr>VyetRip5U;Xctbssk%9`xn9Y0s{o?_kzUVs4(TSiXkzexsK{m#vC$6)tg$vt1FO
ztdOlvBiIpTUmL?Sr;@V%^We`=CsWB7cXF4lb#o<}@HHk)5*UhJK=dN`v$HW~<mt5#
zEoi)_nzK}AlHRRB!mVCh83~d5XYaL`Tq*MI%ya}+tKPT8dy{|hCe+AQ(j6P%j^?`8
zXzQ)vuI{<lXQT;gLA}pS|03TP<}EqFP#8{;n_d0SnMH9&k=|{L!x!02Jw@02xF1<@
z;;(XfYUZri<S?fg-4>;ISjldfxs<fqaL}~JxQU?-=<FcQ2|f!TTDXZlz}ZAQ&UP@y
zO_V{iUVPt$BDhOuG)8*py`0aP);icg&7XThSo4c1Oyv0=f&+Avd-O~9$j8?l+h7&F
zE8{|F?}@t4lBnIgnN)EC6jzq)STy=>eU8GM=L4St0*+Fo96{lokNHeLhct~K&Y*jJ
z^`$6hb6!CL3Ye&J4h%IFLM>V{c$kW*^8S<oEL#jEq4x&rKqG2+E2#+b?~(v}29Kti
zXnXzdg5V*fb<++_Eyu+f&EThFru7D^_crI#2fVk-FAUZ1>i3%GVHTK@bb7inHh4eL
zx-HXcKgr*=oy6P`CD(KD*_ppQgz<l7Nz-XRECVf$Z8qOKHh5sHd7H!RYoT|K_=2*9
z2YM>s;UE?~Lo!x=A--&j`0F-$GV*tI&ywth#e519aV&WW!91Ngl8d+t9K8A?bKTAb
z@pc?+F)rtPknf6kvZO}9{R!Z%jx^kR`CYZk(bS%$gaMqNC`YsX?Tl>-t#$<8@xC7X
zyFFE+Jy;tak}5~q>qDP1kc4QJk(u(K5N(Wy90B>SG?ln~PZ!SV?g&kMl`{QHOWx0z
zev5#Hpb_1hkywNQh)i4!A7Mha3$9c-&J3yX`}V{KPP<Ch<ohv?&jo2&8cn$E?`#Px
zKV4PPWcz<qy(o3;Zn*DsJPqOy|3MpjkM{45;rT$)QFCb5*Bgw#5Asu+jiAyMHR!Hp
z%_P=JoR|)mpx0AR!DoVUqlO@bljtVpfd*VZ?Jzw3y3ToOmGMh2ZehhV6<Y3xb@0Tw
zkI|D&G}9;GMpV(z8hFlSr{U$Jr>mzz#)=#d*ixLneB)am;>r=`vhd8;vGd|eLpK;M
zHya1v^x5#G)o^B29Rs9jRp-$dYL{>`eI0sxRlrmfX1nB%1q2fS9`$_4w(X7Jyqy5h
zEGaL&{aQg;OG{v~eKQ+hGm}u$4m;F0EA<KlOL<DD0B$kv1*(SsoiBLN?foA@fY0I;
zmF<3K;N!3s1?46{k5NJ-c&#vEexz^45dEd0`YWBibUpx)wG`W)GGHL8Y`;6qws;Eg
zot(UB!Fj{Y67Y|QihWZ_toUNJ=xbo`iN`|nfk!hd@s{%YGh6-I&&G6mAl_D?1R7<y
z(Cy4hnlYAM*oD7$`wzxwb4dxq_))0;Sb6LRNn&ZXtAA8P%crMEPYF9ng|mR2R5Pf>
z!l}D3Y-?;Pwu?_YLjz>V<q{n*BCdv)t`^`@9WQq<6DKQoFdLV~X;jCGBrjpj2y20U
zz&sP>I2e)Ehf%aUGQ-cptC*U8@YnBj=GoA4t1B)2{_`Di%y*L|^R^&KuD_&yjhEyb
zomipZ@|prBB*_I9Y*6)AEKy@3C5VA%)5Qhj3R6`pcKfBx@Me2UsBI&vcr)-|H_mld
zqe{Ff6sRbdxdBgL8+$#4N}(-<HlEg*)sG;4r|RVSGl#G>-kjP)>rCZVbVYiJJfd+`
zAv(_$BrF0Qp@#6<xBXqHkFy%GScij+a-QV926H2uWa?zo464|B3Wpzd)lC_3p%Cvo
z@cvhviDShGrp6pq-D0#<)0_ppJD&!4i-mb+d3g2-t*c7|+?Dj%Nzj+vf(Hi&|7&2*
z*nB&vx~t6mElpdq@N;hb=jGxjHIKta|G9f-q;ks}Rn=^vOYmZ+-RbA>qAx1=e(tE>
zY`n>nW@R1}i+E7Pb{_w(KX6kcxC~$Rm_DWZywPGY<Q0}e`i*yH_M1fD*Kwb>@1F8Y
z=NrDoXTMDRIgys*uAw=u@6q3(&l4V&4+#h!;SVymg;AfIq)zEhbjC(V>Lju%F#!>5
z_g(#w#dDk5rHBc#>TEj7z(NeV<mo7SiC4k<89RJc>(Q0M;+bTzJa}lvSX<&v#^b{n
zHFGd%eeXr<=-OnoCEx!J#~Ua6uZA+t$*v9(c5->T&*(3zE=nA(DVw+#s>!L`hNxV1
zZZGxD<r@kIgS-0Lt?}4N0UQ?knHNgd>L*J3sUh$BX&t=5sFxZCEmO2ATM1Lm4Jy~y
z9LkDlt)Ybl*gwB^(P%PR{#`F1tn|!eJMI&WuX<d}$UIBm!W^W=QMvfk+?srVt`T&t
zdpcN2qoEm2xy*Y_oIo-3*y&^mBem()JEFByNRxb%a8~_5F_>};C2Mw+$%hF%yfOOG
znLgM)Df!wZen+B|fYc9S%T8#NIo{91pWJZr%9FEaWcc2WQaMj=>Z@^ylq??W3;qnI
zI>R;#F5@2#icZ9~j3Gu^0y|3<-MYm(c6j{PjW-Sis~+5B<xDw-_%5422jcy}Yl6r6
zAhP;+rT|4HQTyHZL2c|Hvn!{~lwa|fcQbLm>wg^3pS(A<Jlh}V!I(4lLE%D}kNn@^
z=ROSTy7RF4<6i7F2I{-|hVy&XPu+9MSrjUwNaHA%Y|7}wRBjX|nouqQb4eOat7F_D
ziEpb6JOAGc@HUM2f#JVfcPVPnldw1v2W#+r8yv|~9nfT?J!+FY5GuXYESg3{=LqpV
zhOj{1`%`yYdz5o`Mio~^O)q&M`{{?B)R`Tuuq1clnZsZd!)SOOO;&=affFit+hr3*
zy!Gg?v0N$EdSAl$0^eQDBzZX^!9YsS9wJGRsgFURi$Qy3bg#ByZ)P-Q9H+QcX`UNb
zAUfP3O_n^<ABm$-^ixs|pmFK99zDx5I~k;(N~&-Nz&PnG-v<>`o3D9fs(!#ph}Kmy
z?-jNRZKdBqh{TsTgqh)!NtyhE(_-NO@K>mLWl+2b+VnRiN^$W7^n?|HVV6hyR%HT(
zGBToWAUQhZZ_#{fDc}t6Kv7kamdp6IXScLL!n{gJB?Z;;a+GrvA74qDV<_PnmENnG
z?dVL2g~ls<(R!9Wy#)dIKF{|*d^aaB{N;oa47lUy&8=Ts_km|%`JnoEpz(|A67~%H
z6!-XA{~6)#R5HEh#kJ(gaeVF{`-ai!a@^#XL^_<D`ne>)_mXZ19!cd@L%!xra2(kg
z`=(CPT(<hJ;1R0L^Eid2HIpcFXO2ij`7q2cL)qv2hjbb2=iAI4vt9g2R)YIYkZI?j
zDbFXGr#<Uae3eD+5l%XRrTT-bX;cGawCzAr(VXxXMtJCUN+W7Dt-^&jKus$KM;Q|h
zH-mnH3o&4flgo2cQIR5XtO#Pj>vByMo-yW`b4gxeV~nezW=sicJ~V_3nJQrgXg&Dr
z98F=|rdWqq3p0sAwYr>$qR14W^EITKUnUn?jx<e$Vj&XYoNz#==l7<l#}&Z@Unr}m
zN!tWJPXpbx820XrN0o8K5w)<7r3QX*fQB3-OoFw%^EqJx00U;G{7)c={VXGVc|8Ib
zljx)QHET`e_G@K^($s=?!TAgdv|>H=Ec!WN!OTw+)#|DY@E?4;r8`hPw>D?<WUmBu
z()GGrW^HAxW{a&0f9P(3A8!^L!N&%&f~KmozN+Ve5vPMI(jpT{<B&9OTQo1h1lPwn
z?}olvIcXi@oCY7*b8FodpE+uOnkYOWTD-qzH>2Qv4;9O6g5R*W+tmpc*ZVlRR=Ut*
z!c9!q+&jk!4Y6P<=Uc%k%<Iwp{p%xrEdo=HHcqsNh$fVZz*9*e*=ID5INIJjlwsi7
zxSu8lMTgHf>hkkVW!D2e5+c_;^zD83a^GHHQvzQ;JuknpV`CcorKPxpXr0&0SnJOp
zfV57aON;~hlux(&&2syq%k27P!LOaFrXdd+XxqppmotJ-uHEH~rbZlDF-+N!@a`U|
zjNm%!E*n4{=Ym;-v8wjSibv-UvLX3VuS&Ls8Xer22Zemm88|8qXITayFHS9I<#8WY
z)>99&4G(;CM+zETm4Og9ZF_$_+8MDBmzw;tNdNlZP&*{*08$j_gcWr8?h;m6=(LB&
z2{Gf6zuSNF5~Dz5fS^!xPVta9gYHNt=vj-fs!mS->m_Oc92oIUrBhx-1I5gjhq_Iy
zX&^Y91EAd+q2!^OE*yZ+L9h!h9o0aN6nY?SE(rOwg<2yh<B-5-_v3<tyXe6Tu8344
z$oP4$LEhR&2%m2FuSztpa7n5h#fy}Zzi!n)AZ1*TCS_VUV@6TGwsy`F>AHB1-lSBZ
zStTCORw=hFXzi8j^->L=E*Nv{?>Bin!<5GVQ1}mhj<GS4k8P5jVwmnzgU5fPjPG+-
zd8octr<+>d>#NOZI`&ILHsBwsKD%1S{>NmP!4+}GX?6}Nrxy0W);8}uvSZP6K2Ln;
zp=QAOVmqRMZc*DKMT(&k8pcIi+DqSjO{7Ozs>llsp^nJJ&abehp~i<(e-we`+_q1Z
z$`Ni(2muXgyQ;G{Y19uEV&5~hhx5+t=K(GkItbF&NW=CW^F!g+)zp+WJR0bUmWt|{
z8SRCi<A=(zo=!}bo`3|cWLmT{E3eKKgZw47^9zw5#(*qs=V}6qD+*&<9VHO-f+zY<
zYWSs1rRGW;HT<D^GI-2b5DqBsSWyOpD5i&(-u{>i3OdNxm~HmdL5Paua3)~zl`5l%
zEDmie@~q#jubEwA05%c^5-9qO7I!g16pqrbv;MxD&=NLQiN5#Pc=-6^&vYSpEg&gn
z8n@M{df_A^3k)3efEZ<4m}vk75JHSpbp&T6@aoV~hQiEUU~i7rL<7%4Jp_Hq7P>G#
z{~N9Uj}e9ib_E`UR7YsSOu3dyoVo(M>;Cz%a#1J7SgwTQ8veHLVQj3W$kruY@^@0%
zeB}GStD#B8+UZQ>?>61Mjx^5HW)(yXD{5W3nzJ#OD>^q`_;1u?$UXsU)pOv4<ccok
zv3~BlahHSlu_4l#4}G^MwW)C1Ao6R%wy05K1sJU2@Pq~v<X0o4+x}xyzjbOMR)T?U
z;@a7#aW}gNJhVR<%u?Rh(aDkGCJ(?$EhIO9Oei>_Y@_Rf)`>)VDjQI<0W;-<S_A3M
zR63*dz})vFVNeyk)oHUO-|mpbFhh!PoB5~{UmI&@M(+df`;pf__&dbNUompl+QeFR
zn0_V{$S}@z-K+_FEsq9{!3bW4gNN9Dkpw_^m)!!vH5NYNGjlXoXOwjCAA@esl(3Qy
zeRlq21XA0th@Th$6L1!JIz)?v=URI!?5l99kebw7Gy_x<-5U<#G@5n?>EKPdLS?)Q
zjJ#9P)`JoEAPe619K$mK{*15n`P8xZE}GQo06p+AKb{aOy3K~Q7xwj$rMK0MH({!q
z8z7@3B$>~|{VqH!>W#Xw!Y>X#(9G21ZOHSG=QzL$0-?IhC-Z8I6$0nw#x2N}R+Up7
zLox|Hs9XtUSJVL=Gr(nDGw$zJu(6KGXLey@G!U<q8F>lNganv(E~6g$-{^-o9L55b
zdm<k=kEjfZg3URA^``wJg&ljBD7l7K|Kvm|;xY*rkQ5{&y069hSDLAj-uu$(l3C+i
zfdS%r+>!X(=(V~WmDwdBZQz9ydEcs+tXzDw<G;Jf|J%K#1O^DLqk%Fz9vP}UHU_vG
zozJ(@jm<kAy*pF>RYEip2UlWuyta0F|NkrRX|J7SMH!=qjjw1^HAQgtitX3==-qY|
zxViVTFG4Yc@r8PrykI#al}<UqPBCG{yvu0vJoDpU(^UI$#?oRm%UWE-(Z`g3!P6P+
z@aGLrNv!`vkd)*arQe17)Y}7z43oah_3<Xn>yu-yiUID|{UBT#2{W#*ZH7xLYxac}
zik9y)a(!A6P#HPfFUyBQV$h-R>;6}M+!t-Xrw!k}9<QV{MSmHuC{Ph9dNf0Q$BK1-
zhzmtdPa(7Iu-GD;FYx#6k1W4N?>^W)+Q?&~axiD4k->^b^V!iJE>(+ej8PR6Kq;vw
zfoF)OTEvDW%vQ?bCaDaukj0m`00xNVl}3^1*-Az%Q;N!U7raNPm6leZ(&}Ytm<>$L
zpsQ-1B5I*&pF%B1-J<>usma`CC_;k<sLq9Qs@4~ejQUlkDM?H;hA`O-z7$$;Bg;Y#
zgMsa}SW8m<dITooc-+(iqrFwK49vSTgtTde){W*Y<?$MC{z$yE0UjZPSSd1k$S)u-
zpx(5FYXpz}c=)8Ryn`FD*M}k(uYs2rqaf?iRK4dFMb8IqrP3{+CWhcp-C=e43bk0&
z1;6X3SwR=Ch$bHFMofMQ6G56qKGO=n8-vVG_+uM<x!MtceiXz9x*Fc-as1=ccrk>O
zGpUt~O?o=Ap>}p|pqla20y;ej8hJp!#gIM)Y1Tzu%Q-TLCJbKqL!qAJ75C<?I0Fs(
z)j;w#$&sw%UVudwHNGeIbY*l+IV*$tmGQ8>@ifC0^|!c-iFOTp>2L!TS~l&<qLQ}N
zSH0*STjbuE>g?yPx|OF>BL*O2V?Gn41wbdp4imJ`)ig#U83?BF0&~AfTbZWW2m74t
z#cdCv(aUvC!EMVXL2g%hTE2)E%&^ll6}k0wH+ULUQ6~>_{SBi+7~FXY`E{MJ-MEAJ
zrq-Neylo$jFCXYcNRXJg5+5pj$B2x@4lg_5iG)=cXjW(e=gtz>5A#YcC$I><QkBRd
z4}A_iC`Z&0%K5Z5tz1kl+~QQcf+@8+xEDI<s`kv&1lh%g2m{jw_{JI1Tb=HJmvm9M
zsssHjD|9M8)&g0u=DIHJ8t(%f;f-;?zH@Pu$*btK?k&A=IU0uZ41s=@$ie|w{q9Hq
zV{Mg+;p6Ao7cM4;CR_NM@~+UMlBieQF=Sza3c|F4jw`XDN@qqiAImJHn4`kKFo4#z
zo<a(}=Vs}ZdXkG$4I~RWXh&hB!egOr6N7B?&svrT#Yz(BxUy;Nz_oQ`Ay>>4En4eH
zeNfyI?B*klw#B45vT_JnnD<uO3z&pH{!xl6j?fK`mL>I>_(|Zza2VG%(d}hWP=RW5
zRt%wYo*@f9GfcCX3qY)o4J!yWz0eS9I7*r@r$VgisXvo=LuH_XnGS~&N{DsJC3Kzu
zRGrhlh-ZtTVPh$^)$_Q9|0DGmk-A%>ZDB*4_Zl8-8ntI4s0O~`!kQd>)kJwC{2MJD
ztL52@I=S35q=cwM%)}|hEv-qR&#zD(Umf=qoB)}#*61mR5jRC|dU&=@2vh9IGwPP;
zVb#epTu5_hkAOJ-B*Wm+;2OM&Xayhj5SdtSSu{Oe(-AhB*NlyJi7>{a-j01$c?rB5
zr}<vR|87*6W!x*_=?zj&mfO)&$Un`GpgaTWKvH=ryFjjcY&iC|t<z&weMMTPN~CDN
z)gobcH{Af>{C?ICc7ZQ!l_9x$KEq6k3+Wb|5RxYP-F?PK2l-XvrDVS$6+U8qhYU^H
zt{Df)**BZ#K}^&T%391`l`(KnN;3cmv%ed#aTAdt6>U9L?;S}Gj}%hJZHU<E>)0Ry
zc)_X<JQSGjj&1|r0_g^04cvTDfVEr+<LGX>g$Da3+&gYMT4B7I3JrNVQNl_!XBixn
zh<G^@sZk!*i_3Qc@L>y<|0Yk8i!BQl6U5GcYT(vQ#iX7WX(WBVE7B`az*z+cffghY
zeln>-CMKFSNDG7kf(*)ziiGJ___c`^vyI@e>f~f8l4+0M9o-u#Ah(C%syu$x&je^v
zT89dAT39lykh?=*r+?$+<*U@T3mwbZ?LC$GqXN?Co84-aelnv^DVM3oTDiXR5^jKF
zD^#EH<Y;14n~lNxD}yD&ji^Zzs{I@$Wl9v{2IiJrCHgB%X`HMHQLD@eQBl9Ozoeu_
zLtD25!edZ#8&%iHmkUgGJTupE;8=qotZ8MT`YqJFNe&gdE-&{d=D~T(+0Mb?rW@5S
zPQ~vg;qKKJg+zF)$^W(KHbSdyhXJ3`+eGzdxoG%KSnW3koX^cmcUth_x!uoEpO<>2
zUxFA;U0%g_iC&RDtBlAx*1UXW+5}qKbZ{X?n~FVP4LL9^by<37q|n6IX4E90$XQBF
zpgxNpCt3nwxS18Cgw!6Y#l<Zi;)W9FNJ@7)IG1!MmQOQ)75492l5@xM_bh8Sxs~|B
zl+aa5#Oio(WEoe)&mHOB-*m$6@k-Da0J*>&FLAZQ-nm4h@6@_!06aR~iyveIF$q)U
zhHMmy-!ByqtQ0FS2a?1Pn3vJypoXIpUct>?S&X2K_WOTWKv-{<egBJBV|o{;c#Mu3
zMj_9r0AoJ;E|t@wNc(*D7}!O2<!E~O^K)u^GOJzfVGLsaCw>Rvn^zUDG0TC6OTH&}
zD;=34->1^Xm}hBO@}$SW27{!CFX6dZ@P%+ojGQwsmogPtyhlY51Dw3}+PKLyTyT!9
zTq)ALG#;)MMxOZ0+_gb78fQ7-I#$8Z09Hn%q(BSWCj9l@8b`$nOD+@{io{v8R(#kq
zdl_gzwIEoh2wmX|b^#w@=IEu_@a&mFadXB&SgMScyv!AGE^#>)_@!wCwUlQ;CQis&
z{vJLsSrJlW@uKt1s{FX<+bjRKFbi~Y_@Zf$8P~uIwGdopoow5Dau)aN=m;BH=q!a&
zE+oI@Jevyi+q27JW_W3F-mf0d*HDgk-$T(M`=9RrM@(?ZU1@#~NnHGsex>tAl}=>l
zY^lsA5i|3hw^&0ojafXNV==15oNWIn*G$N&Kq-<E?aXfQEGwqi5MDBs{NIMD!dDy`
z>aN!~Vmx?dEsn&Ae0&a&IhT%unv|Rf%8+Xq>D@<ixogLqx--g4StNcUA#WsSDVV>S
z&y?Xth~PIt>!YNQL8m@*5DGgK>bD1#@?rD->K~QKf<f5zZJ#J%?pz?t{~8QGr{w_i
zbF;8}YcEyQqiFi)fy{BRt47hT5z9s`s+D8}n72CQigmzTt1S={S=57{RN;Z7Yy9)O
z$hrSoY}iY)t`)<c3$5_Z4HC6VteD79J3X!}K<xN8S>e_2T#{RGplJ=Bs?BJP*l;c@
z?;$wFO({TnSqF5nE~+A7C>ZFi&HZlrCW{n=@f&>?x`~l<0zPPzeso;{gI>>h^&{8-
zA9$z|)kBrHRs0-W<Q$&um5XONr3nlTvM8jjmynmCK}l7Dve_M?mq3@T%m{D(Wf`e?
zhQ;5++*nd!s*MW@P^D)#MuUUKrlWkHscj<qffiKk<cbeL5YnwXYLOS7MDx}VBDZe^
z1Dp}V|FL4oQ^n*=%2ImeOJ6<VY*w)KluI5LB*8Jvhl;La|2#JQ4a_dIxZMFl7SD9i
zA~>t(7HWXD$}yZ_mRK}lRHF&$(zK1@JBmlW*cqa=-Xh?Ww_Rewjm&(KwHnj@c;{u{
z&9V{6sBSY_T@Pl<j($*fzb!|czZ{H!Ei$NA4^5p!-pN{b;1=z{{~j6Pn{R$EfBn;W
zJMMfxj(bL$4=@J9)_mzr0Snq>?I(`9SA;~EATxXjK}VTgyv<G)oQ3Y0l0BC}*i3B_
z0Ny4oVWRM#lf7!nIKQ5rZMlSfxqhf}2o{m4{l+X{53kMnd4Mu9*fk=p;tPmHmFGsh
z<Vm;f)h4h|g9gNb6$w%J&JS7*+GDU5<LNYiorWPnya`uY+?`SMP}U~<Y#sV6J)XCG
zHxm?K3C}1@sff@`nQeMN*XCYwU^QrClL5WE!l27!RB?Nvnb{<inuuvG#VhnFM@rYw
zBn!^y)<*;*6|&P&g)}^qy@d~bAJ5o!ma0A{D1DdhNign}JzD&uOavr{hN!~hRYZO~
zsN-#a48Q}AK9L3is<;vcct6jz?S8yv`}nB-^xDJv_u6`m@whnkdb0W2(^PAEoG@lF
z$o$XWa%esF;sb(M)^Ug1RL3qDJZ53rMsxNLR@M|xcvw-<+}BJBRX<2eM+OtB@?~01
zj#87#ZK$(<;pF9Q{nM2eqQ*wP`URO(Na70!j#Nd(T*j5_n(Kgce2D)7lif3XbwLJ<
z=Q*!XWvYS%cE&exk6)+NoSC?ZnS_aw@&LS?PBcYd%f+AL(g(eMj;RWb#evyK1pt@L
zo8Vo7mdWp;TOxA6G|A~n<i$y$@BmmA&ZX4{a!4UY9pdj%4O+63Tu=x2K|(}1Vi2tK
zq*X{0o}*PgPgk5YZ--i>QxhgxQt}}#s&naLYL50Cnn9G9d!}J7C}eH#6VH8a){KF{
z($(u-|0~0OIV%Zs`BFtZ*y{^XB3J<9?#E)y!BH{K*ZpNInmtjtr5+2WFHlR-VJ}@@
zFUlEH+DKn2j|tWSSbkuyg*UjHK_$bZAbu<pW3u16N}!`e2)97Zl~m)J5I3UDk2u*c
z$uu|jZ(E^%q=!cpnPH^{eD-Ohcf04K26+&ri>Twkx4hw36m>G=+Bk!`Yk#+Ei)>5l
z_5B*~hn0vkoyMyJjkKZ+@a<`RQ81X4|7~2u@J&Rk{>h@gOJ=dPua~Xp8|Dg{+|oLD
z*JaV7>Y>LlWIef@0pdWCu|BT+zL3Qv=I=@9`d^IEHkvYA-mTMoyDrhHGYC*_?vUIH
zIs8nRV@=|zyz<EFoJOiolRiA~;QEH6lquE@^XxbnTnVInW#kkeNP6^buS|(s#I|q$
zEUSrWjfr(qcA9-_*<%g>eCP|!ROygt-2u^lE$DZm2eoi+bt#U`$R^nSMn^X#1w^!b
z-CK#5C21@UE(HChkCSCxj4A;cPJS5lRJzGB!ubF48Iw50E0~F+t@0R;?){hfqau|^
z-K38F0d#wC6bzK8`}VsB9l_Ch+liL!xDU7zYGof@Nb{56P$)5ZfW4A2%!@7SX_V#~
z+{7pg9n-%MLhLxfFWry<Y+?(uoGq29ZMymc)oqc-Csg*-GZNgdxm8uyuy1s^6X8n*
z?SlL%+7NFjNT~G1yH(ittuME5D%1e~o9rcqAUzX-=lDdj%I?MkwJ+rA{rEl`ezeD+
zz(!z0GEdt!avRt%Oc`UEoRZ@C*t$7R5i8j9y74HyaUJ6>Snz6|wVS;UpCia-!djL_
zurZ2!>aWzKZOn)-IOoBfbK=8csL$}cBR}xT)!7}WPEwGPlba+N+?u1{bLvjMia2@X
zXGq6gn_loc9X3>g*dSYV0ZRa%Y<=$4NVq<kZJ5J>WPSLy?A0oP?T?$2;~`b%#rl8@
zNsXTk!yUJDxu;V88}7%o-?*<WUPjTj(*qpKu<-E23#O?_j3tkY-D49wwmtBR1%Tv@
zkM{lPCAT_w<fomgQ=wsAw7DVAcIdO2yoC<<^TDnMIEPH7?1zBSDhp0=F??t+7C@c!
zf$~oNIH|pK{vv**dU<+0fKCNd^dPA%vS@|K5xz6z71(P`(M`)xdJ>_!W9vyp8$~g}
zDOcF$?pPcSrE-)N3%bd^Zq<#rkh<WwpPpoWdvAhFX%vW1n#dsep!3@~_?$>L08e!V
z(?ZVXaEYYJL0RX2mZhBn^Mx0a??KdyF;JEXt?I#mT2_I<&-Z-8Lr1?0nr36yy!8+a
z2><HSu%}z!&p7I!as63Dtaj<HQsaJI%{t)szQFuSF1f9YMa$If6%U38QM&rdX20jE
z=7#35Mq&VU9bwL^7xzCFG6blXkfv=QyBRg})4+Dt8HrDFB@s}{Wo_2yTkO@n#=<l*
z^eaCb?7y1F+6m6Tp;$#B>*q&=wR&)dPzNhFE$SG-nEE0Z>MHi_eVmTQ+PL!SXmH)*
zh{p-KEZ+yNmrjCZ7C6RcD9(f*{`UQ*AceeX$>42x{BBm)`ejh*_85li0Z$wD|Iu_-
zL2-4#77h|zg2Uheg1ZNY;F^Kp?ykWtc#y$eg9mpD7G!YO;O_1Y_x$(Pt%4V*d6=0~
zy}NsTYi;RA8sOAq+(HRD7f0AQ`K_cf6^@syU?thXZj?p9@9}X(xq+xHZ~k)*>rv(x
zic|-L#6@NJ%z*g71uj!pa9y$uK4|0p)6S2i@S+s>5&iAG|D@l>i=Fj!YLBLi28a~h
z!hASCKLbC1>F*-pR5IcE0mA)4O3rX&HLXX_jc@O2#3ch4`P4s68^WXVj`SHczTx~G
zD!9y-F6`nTt*%uypbLk_%}0w9^PoTy<0aY=+PMrbyE~=#5;`ww9Ne9^)(V<jL>Lp&
zS2ZW)VF6t%lLWA)SC`=#Wg6<v)^EkC!ykogR6fj~PbG=8<=Oc(zkr07{m|FyE?~eJ
z25t{KL2F&I>@jJNEkjisPE%Kt9?%#C<W<ff^_0H1g$LhqywGS@K%5j8Q0EB3rGd!A
z>5p6~i{jFi-wcJRAAiG0kuU8?CrnEJT4>c9u`kFiNfrAY2Wpb>LGIsd1kf-^HJEd%
z67Klcctc{{sKX^>JJ|?HAb*eavd&HV!^~B!5i!fnf#8kvLxZ-%ttf)?Rr`Ul-4MxB
zig_k<MijgI&7o8yt!ZQmd_aZSmcE(#5n?_N2Z`spCUfz7!Kd*>sdv41Ql(6tNiTPU
z7q9+0BSV?$z)*Qaa}9>}9KDx&#OlwgKr+c?L(l^Cg{*H^iz`OHwMY#_n-63wMY#b@
zqFbXl56bZ6?=P(fd;R?8I+Wg@fvPw$&Xo_IM#`bkqirxV0|R*J!fXUm)w!2hxryuM
zV7Q!4A}*K2pbw|9%01QrDFJ`eyD}~4@83W1z6hs9&w2J<0{`M1=NGz$RQQypdQE+~
zQK{vq96XX2(RAOmdTho8@QHrJ)mBZE&bOyO?A`(w@8u2v{mJ?MJWM2tJv62L*z9>7
z^L2ghNWlGcTj<OabM5VAZ9L0;b#Q11<-f$MzT2`KV&P6M^JeFW&Zm9<ZT_v+aXrH;
ziE@_F%eocv4dQ6OPy)ysA^VLaOObycz>B4?UATd&Jcd=XkGL%$>6j>I+(dL;3gs-s
z9v=JvI}6A(PJ1<2#0{eJ+$GMLWZ#yXtoatlTg<K!+ODp>59Vz4+U{tcF{CZ@2hLWW
zm4sb6HU%)o4MR*^)vlJIKN=KV&Ms4pr}ozF`~y-H$F2s(U&;H9I>*jV-Yk||??LYp
ze|$X|q&aF+wxmL#vzF#mGi@+;)ph!nGLaZBMQhAIoe9FP7xx@1n5jc=)y9&g;A1b}
zAMZ#e2z!H=zig|c!`@%rmhFcL{7l96mI1{XgNcK$N)Pgb;=465n5n<xNLD;wj|bTh
z(S~tDWNKS+oYt(P8#_Am$E`8GQ(=r21yAYFqoMSe{~yl-VU;^qe4TF<B6cU@Iu4#^
zdBJ?W>W-PLc)o9I)MrI;tWK#t`iMUD)rwFfREe7P;oI^gO3eYu+s3;0gWH9tK1WM}
zsbds7mQ9w<{q${<^XGqqtLh3x%=zH0OxiNyTvJUDu*sYI{Q2wORsASjL#8Oy(Q3Bx
zy-<mTpf<aMmE($-oL0DEzm#dn(St+SE4rr#65=uO^00`=3)yZY&K}bv^D~S5Nya@n
z@2f!GE`;O|I#RGI(m0chB7*G7PE+R!;28gy(B%M25(Z7XMzxD!R#65WkGvvd;<06j
zGN|~=rsn^%0AoI0Y8Us0Bb)tx5x$tqH8-t%K|zNW9_uY4-GeEwNOh1OSp*L+_cw{l
zb(v1Kb)qO@BuF1i*v7z?gvsg3S%j2Mf)=^56{K;ve9wMQpJ)9;m+P|=oF7dMnSz#W
zJ%e)~dM7`{4($4q`|z$*mr&0k&q?8*)L*YI6ckSFOaLmvZ6m#<wC(NXOvD1$f!LrG
zDaq7Apa73v37xU@V85X=U)U+AFxdVhz#v>fez=v(z#TZ`QGPduUKVW5RR6+~!MN#o
z4)oR3Ces^_Q;nHe;93T!cuJZM{*K@7i=*ob&_5>sWdFt<Z<0gt+4U~TF?8>j{h86_
z;hPMIA+wAKIOEy?L6;q4cj~WuL|sIABk(lX_Rylf4K0goo%dxi9VXGP^szT&+%>$y
zpc7vX3;1U?W%<p|1?@m%x!teA_c}@0>DcF-4S6o$C%FUnLsKR_c4A^*9{hc$cF3~Y
zv%E1n;>_u*kw~JINH`eK<xNHv4w0Mmy=vT_wy%dk_2@F<H&F;jnaFhVEaIZ5K*jRH
zs}#+ne!Ze4S(!1axch92Al!Vi{i|Rm^vFWhZ46;%u8b=43Ue`A;QG+ngcJ!8<9bH9
ziWrfAE+wpR_g^(IPS96~(x1s|e8b3p+274p+O}ga7fAumnm2`RqKhO~w#qiWZla?g
zbKL5Q^DXy;MUG0em>YTX&8Q=D9Q99la!C4ZtrBL;#4Tyjcsg$1#WOEfJLR_`oXm)l
z%o#s0qfvYRVbAKFkjH(vz7c(-$X(!C&%Mc%Ed-QzeqlN+Rh<~Oay~Yky1kFLm?_=A
zXO=tDYXq;Uq8622kf{cbd%cfl`}Fo#e0N`x%)kPFWPiG?@}B!Nl>exRjeS)sblxpH
zH7p&-F8!XE8QS=87TB&}slf`MrST0+nWvq!&pL7Gey@!cNQSK)o(zs-7l4q9^n({L
zX%IZOO<JC(&$osCcV&dn^A`Eu4d5H&Y;-6~h?thAeMq@`W4pS6o(-b91?3!d9>pIy
zx{lq)ypr12Dgh~C8N>E2y4GZxteT)%G{(_EF{oEj95|Linkts5(FEffTfe$Y()%;Q
zS!O4KvUOzIV{Vhljs(NFJp53@4k{ym4&Ud^-rohHwXyI>Z8YQIM$oRp&UI48w^!J_
zT{H~a)ltl~mzevP-KT|R0SQ_#7w{;5xq{AS%JOW^YrJxYmG-kdz7~;QVUqjvl2`Nl
zw;y>|k4;AVY+(t~Yg)>)BL9RPtQie-GE1GA^RAPQAce&C^UBZF4z1}_=BDqKP@q$1
z@%<(2FZsONROwCocAZNu&F?aJm-+MdLMKv0?l=6Rn7-DO%w~Dh;>28D0kulJ5(S_Z
zJ0}0Re621H?I^OF>9EQuLa`GPX;q&^f`92oY(M}{p}j}+LVjvxuOp~AuhGgmjEgb2
zVDI!ecZx_Omn05r#FOL4&tTs=@Eue>$}Gd4QymWHB)cfAQDE>d4~V7CrACJb>Is{Q
z;Odi<y1OrDP+n1v#`pZ#Zn09*6KX>adUlMlOf&kIL=Ao7Zc+9GhD|G~#^2~6uE7D@
z?{8sVXT$xKj#-2&ni#zk{coU6=bXGRAk>FXZ@2d`IDa1VhEGX^24YsPVje08UU;O<
zb9#h{y%}LPeBeZYh49Y4d6&8U?{b>>c)r*u21PrPo@I^<gyWMhIuY;fNRC<%NMf#0
zZ^x-qs68rzX1|UtzdcJI3qBAK=t-qT?D@SCc>dNjd2v1aL)T?fzLk3uNJUH(2L?*2
z#Vqe?nFrOp?Kbya)iyB3aV0x~0^OXqi?|tA8B{!sL~&(M>?Yn#8J&czPQ_vCu0>+;
z&3#JW3PK6P)W9hCl|qNVgoS8Pxh&=Q-xiOj5*JVvo{D(e4;b4~mzpC~rgX@GY|vPl
zk2FQkmAq}^^=4{>WUF*qwp#Ty<x#%wZh-1PdA%|P-K{qqANPpzJ7@y$kuYvWh?Q7D
zBcSFUV~Rqtf$b~7qvD)Q+}z1ujjTdZ2zqUO&YlkZu8U>P-hxmyP9fzi9xZ(5;Lxg<
z_X?j;JuX{$qWz4DV33l(?)3i&3lY;fYvQIuA61DXzE3*Mc)&b5>(~xTlDO<R6_Jty
zdY|>z{cmloau0~Ke%aZp&4v%Jr%RseJ0cRiY5#p*`q&heUGV?r*1Tvd*~hn){SZ9$
z_tPaD$2))<f5Tph`slsF$&-$HDHpZpU<b{O_AQdlw|yr-dk_7y2%vyp9Kxgb)=}Fc
zcmpNfw>~>vL`a7A)>W}KM-n!{hLyRRmS1#oMf*Gr21zSogE-mSxz>As<<&+nYDe37
z#r&cE(DvP*V2jC3jxZB6zucNxBX|+OV$t`&xW&g^x=+AeA%g`1ybbMK1vsMq#+rlH
z1-R2SetiET)0BnieciO37%>(`7?m&>F|NV(soveSP?)VLM;}e+@u^@Vv)>IS(u~U}
z9%+f8vR0{C@$dNhOTH@p2C5`h*;e}WR=U)_ghA~Pt3U>VUje2xMPa8OzyC5ANvg1k
z?R@aGx{j!(35UZyWXr3iF;h3q(NstpxnG=<TCJPWTdrr)Ptpu6q=Relut<N%!Xrx7
zZooPSigP9V%QDzPm==S+ZKYS#d&);>S_iZK)o0ksy1uEZ1=}V>M%Jx~%F<xcNQ4nE
z7U=Tx@bc2j`?s&Qx}Qm=cn-?zcmKEQJN<K1X_2_DH;h*Gx;d^FzuXSygHMQg;kYbv
zpb>hWl4o)MuHrZ!&ffn>Q`|?>AI81aV5Z-C3!|U>!Yy!vE%hbSc5NgsXbpNe-yV>n
zCj!(;l7ke%`s6yni;b@W2zOV@(yQa$Q?$sD=(Z;K7|TjWNe&w|w^t8qm{nKze`&i>
z@l@16Ha<(LyS}R(k98^KVq?E_D$Xy*-?-#?UlE=xKJuWB?j_SQK^R|U-#QKAmqTbj
z8LWCdr#$FhLCcH0=|b)f(F^iO;c*wiA<?s9gTF6F&z;N`_-_bL3SU9#v&rR+1Pv*k
zJil^moKNtYJnilYo16hlmFLlr_ks>-r{RN$LtB)=;)S~I#3%M>734qR1z7TzLzAw7
zN##=XgQHjMeaN;SIEPR})6BqDQLfK#$L;R*g$r7%4@*zOYlFOV6}q$hGKW$yl*q#f
zlz~VeJv3{H3=xK9sx{%4ex#g#P$bM?yDjgsd`AVBEGFSbWsCD=x>kh>xlnzt(vfl1
zW0Sx=Rphp%X|tlz+7;}Q)8JK6={=P)=DzyUHvS}t_?f+icK-{Re8oE(IKbA{crj=r
z$rzz!EjdD6Iaxi?R$oM(q$HL;5Axh)UsRqpY$FtE@5-(&I>!!n#Q6k~nNac}(;L>b
zMOb!CqrFwn7OUjy0dvyfCTI&3w&C`6)I}o+TIkFmuXEe@k8}Ziq4=ATpbopj?Zx@>
z+E&21PHbKn>%VNx?+li|ZNxYogKKKW1iUl}BNithOlxk({chCmO|hw&4*rK>`?EFz
z@vTea-%L7gH^_H79`4vG(Q;Y&79+ndI-9FG1DjJnT}~GCzUD=?bNGxjDVcxs&ePur
zI8!=kHEr)Rvu4Vb-_r}c=S&&;CGwn2I=B+)bY#-mr5e+QTz8JAfkZ8pv^!H>pyK!o
zN7t(Lqv59;`g$XpGOa<}ioYc_g^ot;*cNz=g(GqA=8*Bh25dNp3Jfk{Kfhyva{EcI
z=ow|9zh+KX`EHi8WZ}%?_cUf|BK3|p&oJ~k;Llj}KeNqww$KFh52t|plFW4f#G+Ha
zk0banpZ?WIlT6L>1clU5NvAGcIhP6XGi8^WD+fLI78pklKwL|}P*<yv50)lueBd%H
zna|<l9&$xwmS^P%r(ehdK75(7$p4KjB*uFYYZUaV<Ci@;{)X|~|C;#w4Hv?f>%C2u
zo$ARP)GyNe+EpAoOE{~Y_xTa?ck*d_F#or{5LK&_z7`vkw3+mmlUQ@Qg#o;+MU}bG
z1_gE<lC3bF*HLZ)b=V%|8~^k-?pMijzMt&Lo%U_Xn4}M``XUcG%in|)3OJrm^*04p
zdp*X5)DRxiu41s;t_Sqq1^?~X$Rb&~!_Sj`O-<X_BFyI_CRy(qBfhWK|Hx=#yhi^u
ztgss=Ss%nc??_36B4(${(2&@5YpJ78Tu^P=P(<83`4ZAMD?O6UNF6|ZA;g0M9r=da
zC*B+?&Uu|`=K+wWFU$Ekj2PeZ?RKi5@2!pjAJ}#Qj-X~=p*gpzY0`!vZ6?nb7N&Fy
zoMCa0qnjFAf=B&F-p9`5S=-%+G(KwF2%LP733I?63{X#viUvo`OS_xxc+3oFzi1X0
zV2OYD$?R`GTFnt94?QrFJ@&4rB3Pb^5X7jLh`wP=Y&f<;=$(#*id7z@L|9tCRJ$4i
zDl~oMWc5r0A3q&3h=uQXr1}&3a^6^og#NuV6Q=I(l>(0Gq9hu$WwB)@JaU1IuQpwK
z<d(t}Uap=h&YubDjdN}#GpqMnB;`SJ!cLWWbIE=OxKIsUxNM4R>Z1fukZP%=&4lWX
z>$YrquCh8YD0keJ#PCO$I_H?8^`~j9vk4qRxSrA?1HPoOxU@v2HXY;R%?Ie?<L7rS
zE-oIcv&er63u)eEdmWbS*|okK>61NO@bQ}3JgC0$_PdOE6FPZ&;bzj4yzuwaz507W
znWHuMyM_r<E)R!%vslAFGr1)ek;1NPqpP3)IXuSYaGU(npmTpMHxDC+FQxTd&E!6Q
z$6!ld>1}8(taaXu`25`06mzXVA*>?}vv^IxjAn=s)&R#_VKyw(;orRkh^MvEx7F?j
zPLNBpXhEQNC6YeIVY}C>Tc`g{Z~mxow(Vls;fffsBW0o21>Mc4q+;V)<?C#+E)gi8
zAR7fXJ$JksyzMHxFOMzaobbsXfTNP4!qoebC;@J!ULt0CW7yN;`OUfi-l3)S8LecH
zkJ0?W@LEnizK4s7U?R;%!VErZqIl{ZOJbQyXmNbR(w^~c!^cxyg*4OnS|)x@{1IL_
zFM<V_nQ5!|7OB52Uq^L8d;*0tdyOsfsQ7kF!#xpKTh`v%{cCD%y9jpz;c)fXBRsHw
z!uxeeRAxiJ74pxq7LRVAb`FnDP+bZM`?*E2=FTKb?H?;e1&D2wr}A)&29jJs$REUR
z%MGlDr$&<>0uwq1+bv#y7_@F#OtiCpQD7#ND(p@Oq{Eeg>6zjv$=o}$JY;>T9evVb
zQr7EW)PRCLY4|&FZZXM?9o}%5Tv)x+g>(0cXhb~yx}}N^`pCk(?IDFxJ2NUyZkb!M
zMQ-pKB?iI2NFM>~jaK{B-LU3xw(`rNtAov1ekZK|8Dr)BlcTi+4&r8d_9AEaxQPuv
zuul)r7k9KCIc)9*X&?%W>j!WvL3QbhC$eptH<_>h)#zUp8afCVvudCT9~%?2{)4)N
zuFTG}=zmXDOP8&%3W*mQ{o!PnIs?<dA+zzPR?@D>niJZbd|xQ81ov+&s(7#(7Q;;(
zmcC>pS>rS*$JJS-Hm-FZ5Q~8Yt)v|m_AYjg{$^bh9sqazW3moah32pp=O3xgE<#KH
z{!IZar5N>8&V~G|8e_cj5+BIS&xku?{;4uG!U@vL%)2B*HJo@Jv_S$M8!dlP<Xs+{
zkB|feLd)kYPe>vpkj3bb1qV~^{9mjDp5BGi-xv4XT^?PX!E#kh<X?h?7d&tPnBu%}
z+mLo!g#-sJ>&$`kdLK%LQ2eR9l4JFy1tUPT*rmMVIO7}ylT<@uFwE~{J+fH+S6W3Y
z#Upz@S<wm!PTG+sz*No3x<WR{PzSlR+z?FfbD&1Vd-GAh@t?<Pj2@LEmA_5eRieFK
zkvd9&p{{lIG9BMc<USO(5WT^vWM9&5Fp)Ato8o(M_-F@h_1ptVw9W7@m5~AsZMGje
z6r)9CdwpyJHkS<4$c3ut^Pvm9(kPhzYN|FtjeRZk3I?qlB?}`qi*d@V_{TyIA~NJJ
zRUSvm@7NAkN46?hcca=<dhZVGeB#HwuZ3^X$UoKNN!o}6^*D&HKB~jQyeLfn`K^+p
zaOc3&9!2+OfN3FxB~c1)Wun-g@6r`;X@+9c|58bYl<!V9L?y|h6Uv?Ba#@<hW&?lp
zt5jKxM+nq7vpFSfPMHn8k%i#HT!qJg4<q(^J4{Q6Y6Y42C&#T=c~~wv+=0dJhvbH~
zzAIoGGxb*jW!`ewdUUs%qKvDGf5J#vz|?LNKe?X=HG1q*JuA3K$3%-P8BM+itYg4^
zZ%)wa!NN|05r9g4IvYzJO;K8kQgIXg<-B$6#eVg!<4*6UruD2NBj;ryXXVwMFmz;O
zMAu~rpQjQ{>(O?C?`7U+9|-?iUELMEgFP^-*a@g(6y$C!Pl^xIkm4Hm&&<(z)0i(u
zymfeG-g>*$d*06`rySBqi@gLg_AZh~BL|X)hv-n-PJ8_~zZ7o=gfvwbrE}FPe)^cl
zx+|L0_UY+}c(E#jHpkp?dibI?0<I2ka1ayQ(QbAF+D!XINM2i9{YQ~(B;j5-f_@{P
z3D7A<xkSAq%&E*PaX7<lm@1X6qjIX*cqJIU+)`x3J_w$0Uy7Rm)>Yu{FcsO`vEE~=
zP~dQ@>xr0kf|~9^+B$tmJ<U?p>^JlA9tz@=DY822*j2P5OPOQ|e<gk5ge}=92{8YK
z8^XIemln)tH!(ybXY(0?#AuTzicLG%coTqDlt2Tf*!q;E@BP#?qCJdrM05m_!r%Ra
zE)`--cPi?QmCaJs(7MD&N@kjmaS_sc2d6DNAfx5=^#@)o4mS7aYnJrtG#Ii&oy5KL
zo~C0fU3UA<W!6dY3q{P%BNXR5<6i5LCg0thi&oS#vpOe4;)HblbVm<i)4Z&=BW?9o
zFEOOrMPoFQ)#rWmtTo~;yDeqs^{Z8d67*5G?uR5-r5N<4aoz)(6rQBuVy!_m+ED%k
zP$1t&iC>9zGm$>BoR^c}guX!ax|S_>u}rVb!TXULyMiBM3z(eKM;~_z?=H|Yu%Pez
zrn!hUZe-yoW<Fq`SExq{1SRg8_(}~KLTIo<TP{fg*wCz|xqKP0YKWK(gTp==T7B9c
z3vwA5`CjFaAtDbiCR(@Pqc3dpU+vrb<c7Se*S4>s>|4>;Mxv7|eJo%DYod3#KPWbc
zUl9vWb~XCWgk(}PIrN*EB^rY{IA@X2hqCfqHDe~Uf6iGcCJV(;g&|uL@M<yuCi6{i
zv?59j;ufmd@o0v=!rfH24m%d3Z)^<QzzW0-XxnfrCQNcST>5>))rlWljJaI`r29S_
z6dP4|ma8-Ikb1UdGmb?w-Ztei{uv!=o(Oo5`s+irew=s?Qcqvh>`Q`Sic2KS7em<<
zhn7i`|2f@y4LyeO8|VBICqF=H8UYb!Pw`@E{D{#gaoZ>XHo$(2SEoDpLS`YwEvHlO
zD@@g$P2Sr<ef!=92cj5saG$Az6pbp$dt;?ITL89PtT_{G<(kIg?)vS#<z4ZQHYdB-
z6olb6FCH|1>hQ<rz%%qiHL&$8fz+u@m<#+Yp94Mv)aX0;C~h=oNvi98f>5Wz`=g!o
z7HUzgp2uE#U0xJu!b_mGR`hiy>#E_QvOuTAKnm-77pnmu2`BSBO=}xi@<bl^?1Ici
z2uYTFc>V{I^_gb)3bsBDmMm<yILa!Iw<esd#|-&ZfAf>C9m!A4?Jsf>rUBmLoo`d<
zgrp|pX`P$v#r{bUjW-gD%^pok`lPF)+3CKCox5l`;?FhGY(jRq-|vb698BOg9B$BD
z-^E(R#E}QS4@K@oMpyaIQ%p!nIUIGM6g5$$Vye84eU{?>siU3(gJLCGMJS@1g~J?F
z{jc^ZcpwF_5Vh!P1u2h1Dg0w`$rLtkGTqKhZCA5@j67uTKQa#QK68#6{8y0^c}qM_
zBTxkeEm@p@vz-XBH@l8r%l7nf|8(;~LQmi$TZCJ7ZA<C|ci*wh`L5>};1K;rP7IM6
zh*Yx{*FEU*;E&j_uo!jZ7w6sxkXDY9f_WPjI&e#H1T+%cy7jId7*vZl+BfGNp|;w`
zYuEO@ldYJmoE(oM*SB~Fp)<RnwN2pn&vuO*{JlC=3;WxCB`L*p?5ynQC~%V~dG^~K
z@AYHhCAG+Hi9gz3>-yfHr&r1x!AC{h>;j%-zuduwT&(-vfSA<mekE+@HUqHEk>EvD
zk@z!{qtN^SDukBTS-mbo9~65pj38TyHrE~3f*Y_i^2^XuB`!Fy?NXGFIL~c+I>JY)
zaC5nt`rNQ%19r)W-U=!l35lH!3Y1otGBs1ge!Leq6v8Y30uVQJS8qMTVF}gKRrn%d
zO`vW_2a5&3OU*UuFvhtw$+lJfeFhg54F)$&I$iXhUb<RXC5S9P#k*}k`^($$@iNHz
z$l9Sb?8MuAEaQ@0pGJcd#2kY#YR4(B%ho#qZ7r+E%lWvL_EUz9&mW0jsQPy${=;X5
zZvp1p?DI(BaIzw0{55Gf5Ao%-z&D?($Gk8*kJ0&LTRl$RBc33Z%hwpkS<cmNdN_}n
zG9x+GJ&deBW`P$|VO9`u<%HioD=J%1Vl7E^iwsAQy^m}uB|Iq!aZRf-c$3nOZ49&y
zRA8mi6mSCYdmhyH#(NZPvifns!Y6!&nwGmuSXMPd^X1PB28q5uLR*|4f3Yq-L$?|E
z+Rv3R-2QNhziD+E=%TXTJ`J$A)kG5DxBfS%em^gIl^*)M>1+p%Q<QG}l(aF~km*#o
zPeYbPCu$WWx;}U?kZ8d#GM7oJLBb$zZMQIwVWnU2iF3}`xNdvqw=<>XkfU>*p-`Eo
zW%JjXuLrn6X*^1U8h~Q7m(ubtiY}=LORhZRM)#}raOhcLh3tAw6DcrLd4-GUsqqzB
ztjiMcv6ZUGO^$QRpTrgxEzRv1NeaX-9fst&br=J-G%%wn5AweO+%yhb{D_a|sq)W}
zG8=zsRx2|<He=kJ;)4^M=m<Hk_b!4(WkkDuK+XjO>?sH&S)qM;QC@O=Ppn;mROj&R
zeCyAePR^Rxa?+Q3i+E2PGs@g$4UC3DV#lasiQHLyzl@ly==dh0k$l3#nQK^qcga5y
z-t76wPpto+PF(+_eJ~<`-qSwPNa>6zW`cc%Z}=vx45KJ6-nXay>=Y{I6DV!+b+-WL
zv~W>itk*PTDjnAtgaIRKkTE06w)U12*_G7~T6mzZE&u~uM%fm~M#6maN1=Up#{$Mt
zLIn3(EWs%=oB{SyJn!p)K=>T_>ILU1p&xFj0C||HM<A>3dd_*j!L!Z?HKNwA^=$(c
zwf$sXg&||4y)LM@14Ag#O024!=9W6U{@5mXq(In0I!Sc_Bu{mzHEs<|(qD+!0`)gR
z8pb-zHv1MV3cGO)Q5rEi0@!VRfsJFBciN4~*|8(N!Y7*8PuZ%*vmnalf#MI2-#@wW
z@H7xxWU3(u*v)Ag^4uJJ66TB>x7KH8(o-!Zm7C*m@N(cmqCDseR63px_XV)3AaOl+
z!(|q?3){bUEAHmQVRzOOkK&p$4fs1>aVM?yR!BZ8s+KpTFx8(-d-x^$ou$W!;P{*b
zoo&BvuPnD#jb=qE%{Ih+6FlEWFu31teVBvXEuVV}>vV2!OndmA!dLP42Pg%1oQ1U<
z1sL3yHyU*a4(nt;x-Zvxigdn482mcVAVk{>O@AT1^R}yyv>G02Hm%u6xF3@jTBgjr
zlyai_ul6K;mDMK#PB)%fkg8^F1-m<P!j!LjdM<Sp+Le{4z=h^`7OfxEHDOXi#wIeH
z@wjJPQEnJ3a3IB{1^3T$jQs=%sxNXXNm5oh+BP3W{-zK_C$Y06Ht-;<x!43(^s7tJ
zFy_ZY)q5WcOkH)sY7xE_Kd&6_Va!wBNJKil63kI8;(MNVrg*2O$l?&ct{>-t*S0+T
z#r3^$lRn07#$WL=S)-UWo7BjbaJ{q@<;*88e(;J{-=2o}+Uvar9SOf8stXU09{g->
zic#OL(Q5usUCvRD6VpQdMXTr6jOjN;A|dyP^J*bwNFCk3)MoRy0Q*4PYGL&J8=>&I
zWL;wO4B?kjb2U)5HY=YX4weU*ln#RAXtM}Vvg@fv5B!u0N4~G@a;~BAPq^X}`q`X@
zHcorN=r0?C>7_yz?S~LMOqDq|M<(<LTp;Ht@lC&hr}{sKgE8agKdG_Yi=O9clQjv|
z6NPN|k=QD|H&!%}`*-9Imj*p4sN)arPc8?-wg2tRLZB>Nt&s$zpj=?@E7*~agAh2n
zjF%X|tC&lsL6I5?YKf~Z=TTx0r`kyeCaV9&q!tqva~5<KWRv~9h?}-PoYM!3dHKf>
zV(yHz$Lq$AF*hTI5L9wda=Q8c#_)$=RRJXi#fmJ@>G5x4WkyGa(0^3df<E;8|B;G`
zXegtrZeOTSEO~JLWE+F!#vm&m^OO1`ik$3Lnw$f8FIL9B`=}V-KjIUQm3>3{9p##g
z3#O>3bBLAEpi@rr{OY*XI|QaJ+z3-!1Tv-SYQ}YkrBTGT!|gB^m`xkRU@InCi~48x
zIMU;!^=-2M(RFF_xTN6PG6QKN*PF3i!CdD^g{GR;paAMBW>m!(U~0OEC`9P~3}(Kj
zfrxS>JZeOZbbd&cg-eec2bg;`>#-Ocj+6LYhYZ&Ms^ZZ1FJK3pzlQAMd`qzt+$Qx&
zBqz|sA?7GvR7q?vV3AgadqVotmJ)u{iVBF7_yti7vVnVAq;x4c!^I1rKIt0?4VB%<
zg=~ye?>3G<OQ+9MxK|N4wG1AKTwSFC5g?pf5uEw)e-rGVY!n49|Ci3uvPBqCmgcT7
z3eXz;e-@x2icyjE!ZMgFq&AZb@DXyiG}`;iK~0YM`*~6710>piPINk+(oEb^2cT)4
z?q!0IdCC~=2pT_l1d2AZwXB0uR7!5RC}?tV_&ayn+)B3xT##2fK9HP9UZX<O;;3$I
zOprnfh@Y4giNB%nhF9M#x*bnP`|hP!l8}-pCTYL3faJak9GkR65+GzLA^vdL8#}qr
z6dZ?<MmQ^2eg7W~*d+iEf28DRcTeA2mAKWDY4Ruc)0Y}MJTRpzAEJo+?G}oOgwI*;
zGvMChp!5Mg&(Eo=CY^6l11kjBI-H`#i|#JxaBl(k0S4;}>8OiM4^ADYxo<&t?ayn~
z9Ho=^p(euzuF{vuLsQ{`J_aAl0vm47uj|;fwmiw&vjteceY4Z~*_Nte(^J4fxPHC(
z!zsCjS9V~6scH{bZd%bNqV^LbcY^p=H4yzTf}U(O?9H!qP&QtLp$HZ6pOk;s)6&)W
zX!ag|eSwl)0j1;xtprzHv=FWTDln~=mJKgqFyzz7FO}_vLSpY!9|iwuA#>(TeWE&k
zx`Uyg3B)uOp?KcuFbN9mxlq~;pwPaEtjZ=<>A9{cONQNtx4)(9dQ)dh8V*Eez9Wx>
z(v8Z@<9dp*AJ%0X{0`8N!Xth`4rj%hi{HcZtfWsWyoK+$J@6FX7`iFhYjTY7LYrNT
zY=}rGqYBju;7Bj6{gRt|ApcBQko;_S&C!v%)_#%TM@q=zM>cCHs!r6mh3xsYhS#-)
zTHH#`ZZZjVD+87Kn-Ma56ocfq?(n4a#Xk*l+o<IMb>I3n?7J<Gv6isynEDwNITJJO
z`<FRz@-0=lZ$Buxb%Ys2O-}7y<@S%UrJ7gJ9SJ%U#}$m>g;(&eu@a%Z|GG{Y0)>^+
zxm?T(F{8iyIU5YAqrIoidFZf-Z*|<ro*7fRD1Cc%<B557pKCv=gY*0p3%1wMo^c&D
z<yd$<S%Ci6D-GAkpPz0@A{wBG8-NLW!qK5XaTrWQ19E<D<ZHF-lIii&VNgr14w#cD
zZwpnIDYeE?G16k&vwps$gybn&^X|@s`#u83e|IBG0-RKFzNav;T+tu2??)%3brdP|
zJfvxpXHXbS_ypWwBo_bbi_cx9$7B)#J<*n&({Ew}_CwOA_?MG^koW0PL0PHh>l?BJ
zx;x?|x*rChOX>`Y6ikt$6A*8WP`}))Qe_lugQ?UNX`=d{hc>LAo7gZcwD6dN-3mr*
z4{?<7hB^$ref&|A8Z&d$fZYMpF<anUOSDmmCK)cB$-$@iHo&)14=cMXA1XN-QEC|+
z|FOP29|^Ab7*Zs+C$+DZY%F)$NF~dPp|56ynoHZ*NtHc9l$<U!YUpiyI-oEjRRnNY
z&z;T=2@SGPU8?Qm1$brgZ9BcHc$fT$hVbzYw=cZ6{JfQxQX0x=WEw24k21WQ(GR$G
zS+iXU#35AbtW=>Pj@B1R3hW8@BSP1!7AJg6nfDIgLGOPCyKKde0@;4mpXsfnmEJG?
z6a;k;-=NnUG0$_s4eB0(sFch1GD!u7Bc=8OEkX--!~F)OfUp8*>CPQ$+~AJA=fC{|
zWoN*6&UvZHi04$#8U82D1{m%b<G<s)p-7%z_zgaU=ixj7TVVW*xT~v1Rfi2L!j*!#
zD8_bX{5gUW>5m4^4{U7>W6s9BlwJhmyyg}bFq(RLE0YFy&ktAcNQAtuJy#lcyA_}9
zj;>nVuQx1id<}#iDf}-7UVtpWchzli7o&Ur7g{e-dvLJ#YIo?=#-XdeA{{SH7N|{?
ztK)GS{YYL_@W1A!e@z2k=p-PtN%F+jU%wRnC=v)0N&5=Wu?gQmLF*&chMI&7W(PE6
z#380aRao6S2xutt+cR?$=)U+FO7E(xaM(+5*Fkw!>#$+dD%hU(wNw{0*aW{o{h_)#
zawmbG`x~S-t$(WT>(xU-{VK6wqer+{O1L4exf-i8dmX>I;}8i@tNLBsuLukt5bK36
zhnHI}Exywhiz1Ej+HVGNI_%zo9K;4wzldVXB|43In;50qpe3vC19^rTXv6o@$d!O_
z-NwO$iNZCuHlW?*rJhjsvvuZ1c?#yfNNU4cDgd>|4j$V8g#kSuY!Yk;wh><R60!)Q
zNj|LO^?9udnOt5Gob=&e=~kp|O!#3`ILJ7>+NjK!IJ1}q$DegioE!>nX#jI&sU?WR
zhK`E6*sI~l{q)_Zf#IC?GG)2p&!JC#dLIxBcM|)`xHi`PBI2jF<aIpe(07fBVrCg8
zT(dW}wO|yI7`V`DAxOu^%Z^)ApoXBVjlH&DwNh6+NQz27`gd!B)-T-6*A9BOV;K|8
z{&?@bF}&)TALjYr;)f+0=NP-K)nl?tTf<PxBu&xhv8B70Ea1EBiS?raXM912>T6+u
zSsqtXZQ>l*QiAr9f5wS)gHgzidFT&C37Fsc9}YnqPele&Pr{gkDn%1B87P2Q3}3V|
zv%^%y5}`$$NYgr)i6_goSAy|<{uDr4Em*b3&PI&3?Osn*f3CC&&L%!fFEvilzQ7As
z*_QVi@6jFx3w+Pk9WsVMGq7iCSUoJqQ!YJ+-_~47?=Z*xkn^XekeFb*fHqd;kBMDz
z#E@9-hwk?gigKgkjD`k`K4OBVIW~l-!$IksJ?ZTgu^ATFW6|<*g|a?_V8-o@n@pq1
zM&Fl$A2R%|r*x{{yMKNy@w8>`*DS`h22Q4uO_{Tyt4Cu|b(`X;w%@rQWCm?PfNhG9
z$vcBG>&q01?`;!Y{IIRShcOc#blrc3>cq*2Kv<e#>sO-yqY|y0nd7tIV;gCt4G{6u
z5~WrlAOA!qD3Qrk2-`waz7h?vOXemr)xh*vI2citZ_H4Uo7e84mn-4|;IdH(U81pT
zo~5L)w*Y$sPbO%Vt|CoVLcgQ9%XoS({3dWmOqTOn(;6WqYUTzuW{B`oJC;$xxq%IO
z&<&2T!66O+wByi!U}tNgd@Me^PL644(jKlAY^vfjBkb44b$O%$o&Elf0%Gs)Ta3sx
z+O6Iu0+)c6+AyNep6$opm-Go^`&JD&GA?{_mrk%lDZhPl+8rUnAmMX_x{&#ti@no-
zTAUZIT5@Y9xJqbxrl`9AH&0o$Q*ntl;d|W)8@--1jLA{#@VEa`GSJmO*<kfa=s9v4
z)0lRA^meWF*5bRvi?|>%Ey7sDUJ(&5WgdD|X}T1(RWR$xj{!LyvNbwBP9{Q)WT|zg
zP;mnl;$bQqqNb5PCLZ5fD>u<fG>}FPK?9||hh}Ee^Y2o?eUEK0!7B8rP)bH+s5bWT
z+Sr1s5ieAig1s*m9wu`1gRJ6y%c2jbMljiSk`cE{u8sRBH~&=o%+wH8IN2iGXux`@
z>Q|mq=eJ4s^EiQygp_gVOcFIhO1@c*dTHnk@xqB}Avz4K8E(r|K|nuq9TsN|?@LzO
zc?U!kY~4e}NTg-<8$zjJFW#3BNPSM2LS%E{hkqnp^|bu^bR%*sA*0*seY`yh_WRgx
z6Rg(Qtfp#+z-7bk7dW&XcSJgQy3vac&To4!d|Q5!Acp#y<&M~oKl<lg@O6}qu^y|x
zl_)TbR)6Q7)a6W#Z-w>$6FSj172ylUlr!_R?xF9I7|$n`#h?D6+7o{hZk&tt%ix9i
zI>z_M>7{|5D{)ieOU&9cPx78Yzwo*WZ%2XsAq@=hjJbT1ELy*AHs#b%;AWx@*xc0@
z-|td5rYu}PKX|*1icWhw5cx;r;SS&cZI<U#CT4CI{4u&MaeKa5NNWi|FIzAZ1(CFo
z2YDgUr^@;S!^0ihAes&pgG0r&0B{mf8>Z^tQgdlc`We6S=wkCakMbYfHAow9Q9*Vz
zzQ!V$PJUAdKE7O?&_{a?a&Kv)m65pJpxV|b5hFk_ZO3g|0Yo&dAkCZSHVMq>5aLFt
z_C(w$<W?Bf93;R`(d_QBBvwo%>sV~b-%_CqF4jm_pb@ti^C0i{5v(>%B?DQ$`C)5g
zqkFHJZtYH^CLOz@SJJ3&slOVpBh%A*`w54K?;4S4?mK?{&bon{i-#veMc+ey2H*TU
zCg3L4l3cH8|NU<*ny!9hM)YHdnTl1dHf2;Hp7%Z2khy#(Le5H^V;sv)oT~iSMDpzN
z1B(q@(}>iP5#>0KwVl?e?)}g7_)y4S6P_EZ(`pab(WJ5jHw0<nN7dV}ngPTkotF>U
zjy0bbwK3d2M!CXJ@AY$=mvU4kQKAd|M`OF2d?vt(bUs99EHnfl`)qoi(x~75nYXI`
z-J|(tQw3rU7w;bG9Lrhy4rUdFdxD(`5BA0?4Z&KqAT&_B)>L>FSPN;G33gl?A;d^X
zHCXitGR!`Iod)!7g@EyZI_rFjR+I1>$cRB-%P&($#3^&KheqF>e1p`X+o0_xzWB^W
zdSCZ^fM}F<TyZ%<8YS2@JA@PRqjgs``g)Uea1%Hh(EzMLk#TV&{qIU~@?vG2s#T9v
zk>>Esx1Os+C%-XooejLZR$wS@3H=aQ^;lZ9=?-LkdaOOI_TTxg-RWI=WbZAzulc?M
zTulxh`c2H}@X{D$yH#Yqa1NQeS<N0r`)2)3Z=DPNhg!a8dvG~Xr@O`Y9pne2n^nHs
z{7;z+7f#vlu#K8+c9dIoCK0XXkM@6$KD3ORY@9aL7BBdOL!tU8HXA_roO9>jD(Pk4
zO=TqIH}pIbXEr^<q6_O)gU*O8aBUSSXUbzHQL@KU_I3J13Ore~F1LLI6FQf;0<8+i
zsO@3h%&nHKJ`#pKkcNz;RU(LSYpmbU)HTY<{-OZHu2wI*-#FT3#d|r>ZmQ63W*6e4
zLKyv(7pJX~Wh*X<7DzQtt7ssC!S6$;&wZ1%DFf(;G`MBy1{r$1a>IpWKQ#ZCH(z&p
z@);8EDaLv9NKsr5$u>>>53jpijD<IT`t!AqgMq7X&Zqe%#OG%J5oxvDI}!M)I0U$g
zN@a#DJsp8=*kQOqDSaHq;jo@1&%Nc&o&G10p4G>Zq$gJDO^{JJLB1{1MG31fuhrhC
z2;Hk0sj)8mFSN1C9K>~TjAS~D%?}(K3hKns4#@?%+ounSfZ}abvI3H8^-0X89V|_!
zokK+kF;`~n%|NhXZIHNa@R6Ek`mta>(&F%pe*NVCV`2O`xz&<26wTuh?`)m3=>-S8
zW>SNtf02dP&0S5d0~kKOUGaQ8I8-$e)8;8lkIBlY{)g03W@Cw8Ht(9rB3*sa%jvc9
z--x0oNbpb3_heHtEN^b6Zwig(3KF~U&q9|!52`669<xRK9(J#ADN^VIOdR46Qa$@M
zcFxi+j=4JpAt>NVj5IM_*9utvEQFhD^6mVf&04JeT`VVc)R*^nN`B?%*F$dw2Fb4y
zgnw-}efXCM;w*AmQ)PX5Q1>3ZYd$Lx5tzH$gH6}~0%!SO2>m@c{ytv!Z<>+Ve;A*h
zp<`BqM6Z{+D*f<-QwZw<i(^HA{qfZYX20D5<tqtgKg{qB4rF}8WA29(c@Y@>jQNqi
zf?VYaE+8(zrDUA+3VDv>8^@oIjF);-=R3(btTaI0W)V}xc{ew8$gd4=4{;}B74S@`
z{<cwh0YR%{`yyd!oN->RS_CZ*%I*|uG8*LXmfE+#>iUm^0b^rUXMopI98f<DsWi~!
zr9VZMINaERs*8!L{Yz8Z3B{Y)xamp6e4Jp;w#q+ezE!)vR>d=d=;l5T$BQ<LiopQe
zmG34w4dv!wD>R)(au`%}e*u+q<KXP=<g{R5`OD}Wc=zVg+rR29o7<^)ic(3$<3>d{
zRd`aJ>U;7t;U(TEsjiba?h)Gzt*&HzBQ4c!e|ns84pp&pr0K=8Dm0DSu;xH@#=KM|
zp34N2KJyPv75Wpe@Q}O6h3ZJM1<OjO$fHr7h7Wgz!+q1h%me@2;;-xPr4*N}5A3|t
z5+4gGsz&iv7j!J8$?B%r+5o9W_s{^e4U<W(K^*|B(q-TVJKTDI16{e?SQLXfcFe#P
z8{bmKja848GUsv}*DBD{%D3HAIPAFl-?=~hMHK!!HHB^Pc4MGS7+PLYagiYMkU&@H
zF<Yh&%@*>`o<EU;i<u|2y~^%6+iuA5j2#g6&W5Z?UP~RlM>}o%`fT<5`Bnb%o8Vj<
z!5RU{?e_PPmzDsrUrPQF-XfPl-}`A<Jdf!*V_x=ICtmyBP)}cL_iTyvEz#tFL1)@;
zQ<?AaZ>Kmu!LyXfdgMLp^3tDhgN}!OK70GTDM#Tl`UWUnP+wehP&Y-1N;BIc)Lc{`
zK9lHC6ZHx`=Ov4=VlXklPu=B6cc?5Gyh}{^t-p2#quE!fE{##+2r;nG91s{5Yvy!q
z5z6->a9&`8>XIjR0~IAoR=&@6sQ>)pXsj{`(QV(2{0R|Huo-Qx=Mv-v7ScY(YqQo_
zy%uJ9mhgDqOQ-I3Hh-*C7yz4$eoG1Mj%PiUECn@)R*fF-dGL7tqHlIBU)4Z9=`<ph
z*5GD59DNDif%gIk4k1p<OL<hV?kND<D!iQY)~4q*3NAEfxx4v=@Mgp~Ki!*f9l=O+
z_7`W|IS*v+94S4oDSPXgLVTYmGv0n^-~TjH#6~(zbx@&~uw!P`u?21uh^fT$dDNBS
zjQ^5dz!h58lRj~kDhrG8U7TR<v#&uubBME%dBTvRu&O5Id7&s=(o`mKSh*_|Mm?N0
zj!QL;81DJM48PDicYoBU<TpgIH$J2o<6EBOJAdRLkAP~9V;h5&O8^T17JI`z@v>IU
zW;{j?Pu-^9O3*B%xoLS(PVW$p!?CkVba$-9!}R`RgFJH2TU6hlZl0kPEFT88eq=zW
zvGNG`g(Os*EhtnTE5}C1U5tn)z8>!<)TMy!sLEPY2iLVO73WK)MO?@8E#FQ-_dK*A
zp1OKF2^9et+xF@LAtP*S=Pg3$nG8xl2$6ZN0Bu-ZDne-+NGQ-&$B{We>avni)#rD0
z;;W!Q2r;HABhpA0+wBWAXQs0mWPp0vA3({1mZ#*!ZyQc-W&C;F<VMP;MtB83he%hF
z!ki)($Da$uIB8uaG5S|iwV+PSm~<Fl(>}jGpO73%E%WCXV*wud#6`TyTv~UME5DNN
z_r6*zdwsi8KWKST$I#^K8<@CFtW7K>$--H+ax@Eq(|Yq&V&?(dq3Awxz0()ZLgtvN
z9e`rO2f-BbSZ+z3i7eIJ$`*P8QktjABoawLTOku!641THgNgSt<}s=~e0iPlscp1I
z!AIR&N-yg;+hdg!=kTKgotyV)@{*hKi>kypj@y^EhnyLYgs4FBS>Lo)JqnCaXWx_Z
znyBZciNF)Sp!{~Fk~#l;2WcTQu&>H2gP46nE=dye4(z@AH^M#Yr070pVL)r{siJ77
zfR@{KnF`Cq<r6(ZZM(hbc|bH=Ju&T?wKMkmwyh(41vfyN$g@NH`oZ@V%WLh%@5?gq
zI_JhjCLTEZ4D><kpeZF%Qt#6jjXv%YR!d1p8B1))2jJsN1OQ%5XJ?hkh)hzY3g(`n
ztYP1*pKz*Br+sAmG{{AWA#<5l8lAPCGBFYvFf@8=Q$AHW+npE_j3a)|!VtN@5$S4O
zy<u4eJlvP7clw9ZzWWoD9dA#aZ=P$F%){wR4($nV7MDFfd*S4MbA6>xSs9IQo!;kT
z7CS;Ym_AMyl&^fAIN7YoLbA9rGtYM83wt}0f(8#7Yf{w`iqy5}p`Ev*KOKBACszz!
zGwu3O!;MNV$RDYqpAYVi?=x5TWjmgx*Tz~6Zpug!yopjm#fi!r>fvJk5}3->G+KoV
z?h`~z)NyqqRISVx7+BAJiU6vx4`|&R)`Yo7*Tq@W#WZ*H<O=x;g4)UWw1oAv#M{VP
z!^*!5ndmz|k0@$PF6Keq1rU*DkUh)I!ZaEYe!5>ST?2)NJv00KIe(*;OFh9965O9~
z95+IyZX~5!RAc4ZuXBQp*>!RB7<a~A;p1LKd_r`xurc;blj<r~igtRDc~~CTA9J>f
z2H!l-X&-VE)|A0#xT|Il6DMnf3sBZwJOaGW0Q~23=wz?v(v~9-*u+)7M?=9U;)HJ`
z_O1`LS;R}of7YYDP6d|1?BO@7<@cOT`6vIjEAF`C4LJX^6_QLF`*|$c{mDpRqy%i#
zwc1yh(cb~OEBl{gf8I>8{yt|4LMEA-g7xS_c*Y-iUS=(;;Sw>)P8GGG4<5n+Y=q@k
zdKlDYp3rJ=BE!b))6zZI@yI8kw%mhGU<eW8*<o8AL<tX>;7%v-jmb5@87!M0(X#V2
z&NtPy^s*A=ylkZ9dav}``8V<9y;s+TU)NUtr!Zn~Sz3+KUpavLf|>U@{r?J5<@(D&
z%=17@+XQh<WXh+*MP}XZ2|UMd-*Ou|=H%Z=e+KNd#9F?RbRU}YWRX<<Hgf?=Zk?6P
zwg7nrTvqB5(9qvzLVLz?ay7}4f$hZhkTE=Uz!zApors0rI~nn=7AM6LCT|@qcjkiN
z>4_$z5~5^KA#=Vq94y3$a~e%fvXUvpZB;G&j5FbpIu|%uGY^zKOlLjacN#WSSdIM|
zsgVT%J!}&d@M_oMBQrI0+pUfLO|UefZ!i)>?@hO9o(OE|rM=eOnD=>p@p(Bg@LDX0
z7Z{S}%U%_8IWYedj@qL#2e#So?Q%IRB!cgvng0ns<bNtB?keB_-kdXtu;#r1%KQy^
zB7C=r3VmOaYJD_Ak9qEZb&oCR!vl+x2aoF<v9lM1$Yy~Fck41R{^_~d+!Z{{$}KKu
zzJxxM_c!&3xMHchMp1gk)i@`ucManYb~j%A=)~6X0Ho{&J7=9^lNFh|g#*Zx^f}CL
zE%Z3YD&ONC&W|rtAL(*)42OPECsOBx0y`x~-an;65vd6H`pU1#z#SlKPe#`gcPkU@
z9z}c)>HdAHkk$hldy+a4y$_G#)mjWZoCJBz`gE6zwzTX|*c>|`raH)a>4tF}>(@mN
zI_@MM;12O)v#j<*K~()Qk;nVH%?umH#}N-wPHpJws~M1-AC+WC3EV#vM=J{}yDcwB
zk0c~p!Q9}w6(`^+_Xt*i!_{*cKpPW(E0C8t`wqmz^EofHpcNSMpj!ryG5UN$PnGut
zq+<*>s$JrgM@$j&b6@`_VetHzH7(MsJ$Ip-6HH`eWZT22Lmy#JUmt?xSI1%Irt`+8
zcF5U%^x3}-bzx#$NimYvK~M$vx0`DtIji*z??CBeY=8W=(1%MV)LG6uthLKL!Poa^
zo3|MxgBi4_QDo1|#@v2=XHPz2?+jG@5<630tOMCL4S6@snpPE_a@VBh1bB;b=zX|k
zYcA`@%gA3)<L7-DyWfsK*%l;H%fdF>6foro&E9fC7atj1I$vU9cNa(BkE@ZeWN#-#
zI&@_yZI4h3%Whkbh{}}z&<5xa??7OPFrdwRT}Aqow0m7`Y?j9-QZkrm<j=dw9cJ!%
zvx*?)e|w{kEu$S*XRXo)w$?C^%r0{>kAD|jMvLLL!aau}1Ut@+_c>0P>b%*7lIEH^
zWJ-Y{5|1tAp5y?!%#*@*R2M!>3AC%Uh&1Fv0$mAXZ`>&GHD__~EOq&$57*$B(e882
z74w&U9X=3-3}XKW0>zJZ_4d201`j67B6r;1O=5ah4435jP~eAO(p?fF%4<!B*yxU|
z-j(A7MKMC^$FAJVJkO9VO1@dkS2Dfnb;KuGV!Ef{P#=n)*Tz%Eklcg(wr<HKc<&sq
zY)4+*zOXu}>+2-z=7TwbOA`w0g<=s{L>o!pk8t}&e>ZX7nQ1w~f$H`YlX`#27DehR
z5raO+hTsMM3kq6EO`FM?n^XD$Y-TTny7aE-fi0Ry6rae57Euzwk7(KbhCuIx{(Ke1
zMoL6j+lmMI9{mqZUjftR|9!nNTm}q>4R?w=6nA%bx8d%?8SV^s814?meaLWkAMVck
ze1C8LNt-5Znx@Uu`#JaAbI-~4uv5KkJ{a0vo{j|5nk@an{MTyC>du~-`DO4xoE`AN
z%s!q0yYFB#mz;UR!JT>BvUUdNU<-$zK8G`hwpBIzd;iz}ooqXu(0aIE=8Nnia8B?!
z3iPJJc_4iBUik<(3D5!9ccwceMoQ=lJdmr<S-DxZJB?kp%v8!=Id$N`m{*6F7cAQ&
zSX$aeiEJas$8lmSK08=9i>NtflP|rEpgC#%)S8d(Ig1at#VWTM_i0Dg28t6;iPxoQ
zn7T^ZYcHZ8_AM8R>y+eYNcrrFnd)px?W^RNyaqPPDwoRqguI@jnSlb5D$FNVW{XbR
zx(k8q8$FBndhaV&>aShKdz`KQSCkVmKHph^Y_RD3{hM6Dw5jN%u&eZ;+oE9eL*ag4
zB<AL_#7?ajmqi5~G@XTGZ*9*s3`%Cf{a1l3ZA(2~M3sx~RS|_%z(RJl_A8GHW@psF
zPnoq;NE~DTg&?umyfM#!I;xhpY~hmvQzGp5{ad!plWCX=DrgCB=0VK~1?w1F^(xQ6
zqS=Y&+q+g+<!<pRsp6(@o|UDIW(IVC7W}C~1W;sfxw&Pf!EMy*)%bHPj)d9$>l|wK
zr5J1OL}9OJPQDt{4lP2O%t%V_PAtqLj)npoDq2eKfzY<Q$u10ak7WW483xDenW;Ef
z*B#E>JplZ~dU`U8axQThzmflxPh_a<YRPFFyq2bFJw#}e284iEp1fp}r2C@2Q+9T}
zQ2ZIZJnb4A(6M?6Iw$p+<J7;zr!j<}>&}I_14x_M(9JESx$E_`<Mq_uKabOlz>3&&
z$ynM7<d3!5$9V4-;NzM`#P_xaD!b-<@q>C8KEb(utbTr8@wXWy$$5Ms+Rg5IIkVVX
zf7C6o9r97T{CIZWk=$zD{hcoF{oX&k^%;_3G2?$AdTsX^MZi~YlHD`sh3uTn*A=-W
z@&B{{ea(DyIj_dT4<q+mbBX`F&U%-Je&v9Bm>_9Ea<Lt5kg|Dr+VY=Zoa55Qdir*t
zDxmy_Z2#)Zn$hD@C%Hj8;5#%BIG+RKUa_9~0)sF&z#1%bD8Q}1)^EgGT}2)Cs8y?*
z{6f>ZZ6rbeIM9I3EKV6<h}D5i=;vMYkwF))(r=G07_mj_KNu<MOZ!tJoD0&;H}jI#
z>mYDFY^x8Ww%3I)bebp!o`=x_5)V)4(a}S3ZN}5fla47|iFl%WO(pXwf`|=`^1Asb
zc~l5|7JIW7zAmPCd+Gi62w;&?6mcC1C!5GR?8@e0Eiv1zzIuZR$bvscv!_M)ki}6(
zcvvUdFf~Q#s@E<cI&llL;<KR#mPckjx%(v70z%c_C>Cl!3JmTp1NrHRWqCXaE_V7{
z1uRkk8iRzHmrBCSh)f)D=IJ_dwO-uma<D>HhJ&5vRFn^XUj10Tu!9e4N$SJIHX^5M
z!{dAX{k~F)nOWMs{kd<k_ZhCU=tLU30;U=aW)QI=3N6Z7LrdFt4R@=6?Qn+I2G_>G
zHc_FPt~XtdV>lFpf7Z5Er^*f59~~`2^6`*p%81MBXiG+;q8?)ovaQZvUu*3}%}mA3
z#pi#B;Y|vp(QRisPjWk+x1Rg0ANXdgChp2(?jVe$3CX7uj7}q<)+jE=2PC6W=c1=X
zX@IAd6EV}bDFg(cd%P|8QthBBO(0%aG8+ERP0brUpWSMLEbw`u&p&4;FRsBKnNd*j
zau+p5iLG1yS!Ds~hA#)GhX4WABV`gs*)^1lpFb|i%Pg0w;6F)EjYDZ;M7WhnjU^Z-
zQL|Ii!&-(As;C_(a1&O>K6blcsjD0~<5Tz4uvL(!FD?K<OME>}#v{^wqF)n|LnRiJ
z@msewsf-kH4_tL_n{LW)lKcB;E?GJ~<H<rc4e1K8JV7X#3bnOWn`Lkj#dwBZ25a-L
zP2Fm*a1Ygy-_%vcHGfadYq9Q9fTF=7<pt<7YteBBHLnbtFH%D$v2NORi;Ns^Z=2Z2
zo>qS?@$+;($PNam$PwdKpy2JpqCAW`of)`gxm;76l)}koiX~tz*Tp4eY7!`mrl)y4
ziyqamk%u^jT!<;&Qe-ulYT8wPjscIn6%+K?hvJ)L;ghS%pW@MuXlqYP-40Ugvt9v>
zX8UoEr(D{Ym=!3ts;B|gKLWwoN^O;=;s*WTcf$0t9bxCI1=j#B<J)xgl%}WU%z2~1
zUM)da&Sr)`{0hBM!!AG@j%RUJ4xl|}BUz|^G0Ez>Q#}C|zGZ?k^`gVm$|j|izdTzf
z`d{*Ry#}ARyboFF661M!<5*2e@QbxRuUrJTP~UG!0lv(81FD;1(g*MDkuMh;nfG+{
z7fhW39s~nhU3qWz>tTw#6n&rFE>!=*^Xb=C_hhF}rtV`9X8nAuT(-Y>MSbORek>j1
znGd?fF@D>tX4dc)j@L!PywzFnKA4xk41`<v_7nVz{Q&pQW5>{>!3f4^al4?1VS^;5
zNq1{v#;rb45*qOIWKX^gm7KCO@Ch{G4VdJF9uW5sWWNZcHopFu#_{+*iDkO182dGC
zUWUDh9aEb)xJai3IkIOs46-dF@5Zuk{c%B7T^vK!xWKb*E@@091TE0(x5v6HV8)Ww
zK)wL`Q57u3>P1Rz%NO=37mv2M9M%?8-j|>k{i9Y=2?=a{FsnR1T##k%QkNjTdG%?z
z02FmQ0;$hyESgHO-aP4?xN#oL#~-+<geReH=s}i-nv~nCq=1oF$i;?l7|`Ok9IyJT
zEruXu1v*vz8!gQv0)(q1VutD)>5}Q-GC`SSO$lyMdYS+GxWk@f<PdIjjrRISkdus2
zkFZ}_d1?|lq>1{mf5Y;Zjj+PhI9q3jA<anI&mg&ZPdX{71cK&$ycBBPmUL$GKE_2A
zC>oeS#3*#!e}r;cM^Kr-1+3g>y?-SM=}31j&WR_eX(Lxfgf<O%)}be-z@pH^OZjt;
zC~Yf=WD#!;a*;^;F!r+;&(b<xg6xKKo*~Um#jA5w&L~v6&XmmWz3u+06ZO>~3TG_U
zXR>nkv*YCTyS+Z7pm>erKUZ2g&NXPjLlWH0z5sSk0rM{Q2&lazYE@cQJ^~T$I%l2t
z`jCeKMiv2q*fsIL&%<P~KHnzn-~e9`3XLd@II^4}jI6IUxI4m=#pNF>BZQ?qIx-S%
zcBVB&Vt;C8m9&6yTAOz9@kumo!ofYJb4y<<;3TzlFWEQf(0WNF26h6NIjdGjkV_fx
zOhb!=0;s+n%^>rU%fBv6>b3l+Wzs`wwJVc0hf(^<-1Vhqc@mER-9)EHTgFOAdV-(`
zPGc%YW0e}hfz79UDNQ0dnVH)@{(K4zkNsE2JRoj|(n8acvYg-f_6Ockd}`>PpOC#N
zjpLr^9=F^TA_*50HV}8?!+7AqTY&EA(ffFO=Nvugv}T%oBuQ2@2#>13B?Cv3)mx9(
z+TG!bEJ31E#4f8YlJAU-2KjPg7Yssi5-r6;EubIitu%~MK{2qtBoFrXPj}7^@G4x^
z!`~$zID921MoUjCGj^*8{UT4Hk1N9H+7l!UPp*hMM=`Q@YM7T(zexAnz&yY9C)Huc
z+XCK%e=;y7XK}+CI9sVvcJyN3QP6+eSWM&-=X(E8j9d40+_A%zVB9MxLs{fEbaoP;
z?tl&1ci}BcOp&A+8BLJR#ofgw0kLoNX`p#U?3_mFae4m|V>~dWGRZ$yxy6#^x}83*
zF@4{7?NuRjm#FC!1CwOluZuN}HQGEp;8=ONo<APZkF3?lulu7~`w-17yq*kBN}Fim
zV-|=tsBHHK5Nvf$U4Mt0M0PvQum^i6JKUQ7nM-8u>%Exj!S?TZ_h-8A4)a(q^|msX
zALmU%f8W%TFlDu)MJ=c(pot9&dstofAJ5iU8)(*N_gt^N=lHbD_^8j%^9Zi;+I-*Z
zzIyNB9Xekievoz{yO;N+x*uTN9=s`k$(Y;N8C*Y5I+)6X6(Th*b7PvGMI^kc&ZWoi
z0KIRkK!{w|%&N|u*JOhkjq>v!dQt4Rs{gRBK3#;hK%QH#CytzyW>PJ*j1n75ZoDSd
zJ+v~$n8m{*I6|6a7j&tA*T$u!^lIYAOO(>njY#rwA9HyrGvPf~pk)0ruDH6T=oY5<
z1&KOJ7^UWFQM=@tLS)ny83-V!2dStHoOuayy5?C#D=J+MgrX7CBWL|tJ6h5nhL}sz
z-y9ZG(ETi8N*CH<yBoI-hw|C$?HVM_2^x>v$s!S!q%<DT<-`5ShYVb%L+TizVnXEP
z0fEz&)4hQX+>`-dpq2BI&Dy@yln~}mrc;(b*d2eTVK&U#TOF~J=lg1>i8Bi&#yj(8
z?o$FFPnld?fk&7?Ft{irtBG}IiE0eaGo@6PJW6UElZp|GQf)~ooMmbtm?tu+g~Lw?
zqX=pGdR_=_D(lgtEPoR>`H=Zwtci>L%w2*Zx5<s=NO$$C?1@37yJg|f6v(W!3MH&9
z;g~m3ZXt$n#bur)vq8Dr4K0L%^;-kqeOCN;=E;b*i%u1o^s8d@AAec}vc7?G(!BUe
z>^dU3s{SIsTgh_pf2*0d)8w)%kzWD;h{NLpRV||T!&aE*9-YC+%S+-t_e#$KQEm0_
zP{I;#n7$5Cc!}OKdss;lO&KJ4H@2W&*TZO6v1N}F)nIS@Eiqqyi6!x1=`E>}gj)$e
zrLEr>)s@{EF(>~l(tj@YB_<36AW8BzrPO97-5`nRAIDGb$Ego&Ty*ye)2nLQ{HlD6
zQWLR`rsR;(l#*_!y<rtO&4dCxew~1`Ht>>bahJIfD+x0rp#kHZWrp^Lm{y1`>u_dU
zJqX1>OE%mW=eWrpv@6Sa(9qZ)V5+cF`Z&Y4f==7Rj)7jws3qm3o{(U?>)5n#em#xR
zHHk#{M29bO|NUql^`iTv`AWcgW5Sw{xFR>7BkZRu>v%750p5eGs!t$26e;*fRlgHd
zl4>*9TpEs&M0`OW^s<e#B}*VofzU}c+Y`b8ZcBG6HmKnMmBuk&Px)J%T=4bV#Qtam
zaqeH(`S4wH`V(C@9PW|LMKs<#QM?Mz5Qge9zs-X1=R+w@Wm#2nS}I;;XGvGQd5``e
zX<3$3P*Bt>{~cqC42qf-UfEFKZjY^S$zfy+f{Y&`BiXI1i4ctKtRV+1UKq?{OCyCa
zS0FpnW+PRrDA(|StU@Ccr6}nLYtQer2p+?ApdrCWJtnfgr?J^LeB1=2c1lQ?<!O{{
z;mMX%^rd0eV@Dr8`0ke%=DfbzcA8&tzNd|jtw(*I05q4emBkh06j>EtX&gcE)wtCy
zf*HQ!h9dNhK^C8RZ^t?LX(78QYwVQP&)neq8K3Wawl~Y=E(yVEe?<$>dM^Y*%sMP*
zzH=-Iunv@)n^a|&eSdJxj@G-4D`bLRKlLedCG%e=Lff-t!E$c1uj%8kn##wQ5Y67!
zAsx_tP4;2-=l186R64>h=h=*(%ar#{#!@n|^fh)eUNf0f#f$h5ysI+iKFy>se!ehU
z%*}?)tae!%Mkt{k`qiog{B>=(q{<657UTkd#V)M|Lu0udyuo|jsVducx}bSYl}1Ro
za(7q9|4Lr9yM$WJ?5qWCzd$7MK|APXwQHH@FVKI)sY?Z58BWmhS?3uW#j%2NJiiQW
zgJv?D%LcY9Mz!OIOiygz+lTd+r%ki=31UB0rH5-C?(GgHB&WY=eUiw=lq9cI;uvge
zwxv>YsZV%>4A7xuW6pjOvqDy>RnWtu1r`S>AgYB$X8tpCKOkQofaM?Lxo(k&z%;KP
zSLA)NJDyMhNMEig*vLjyD}@JwymrA63q}<b7vyOZ+;!=Rsi8tuT6oyF*P>Gw0kz@7
zi49>C(Zjs4t9o5jLCP4CLKF`Fm`uo9IaGZmFm31n)31CNbps4C9UcP&V+E(S9E?W>
z8AFH;Kp(*AkWBL+@z{R#JdLgcrDNq3oMwt!daQpBs+a}B<E(U1_y|4!sDj%>BnlZ(
zq=TWjS*o;}m1YLfs8Ndw$GU4$($cKXuKhCwPF~Q|I;EAgow~%w?z&OA`H`&#NUBR0
za3cfhuZ@95=iA995wQ0_rCc(zHLqiD#^lj!f84ZUw*%bCkaJdcXE|Bq-JSW}%Z85h
z<-IZ4uq6)<(*4v~)Gs+-JWIs$#If(6AN}NwK+0U@VU=i}6C$pfT$5$mxt1k+o_+~d
zc~;^P4L{~Oh}9#CgnH8aCTcV_Xa(uwDsUy(EbyN^rF=12lrD9DtMks07n$k8gO7R<
z?$|%2e36KvD5AW>RBqcW)~_ms{X#?TeVj3r#{g|Y-znQ32f`0O_8Wv7YIdC@LZ~t<
z{T)Y16A}<XB`GsTMkfN^niV&RH>%oV-Eht}t8eTMVFHJ3(f$0wAXu40P?K_W_UYyN
z**};|XRX1uz&$GU)-R*HKwzg6&ObSMR3A2^J!~E^L3w6u4SrR1Zt0t@WER0=0=c90
z)oDI9b*dzodbn||&`UF{F=Z~V7>fqqqO(Na#Yx_c)d>Y|4u8hqhjV>;f#Z~}?fUSU
zzq=oYU-sO!Yh)H{qd`cY;GZgZh?hoKvnng9xf0Az5W-QooF6=J;=Kja>D(*~SqRbX
z4=BW=h1D8K^J69@1?aGn20PhVW8i<qqJ_jY<7|5JTH@uAou5hoq^%u$Lrbez24SJm
zju1OXXX@HuKg*-=VLeBwfk8DT=q&n|T2hD!y`^Hp3d--Z2K;jJ3B;cvze>v#(rEVE
z5oWMl^;vCP3xn}ueFqtGf6H~<#M^civr`;>@{jSkT>Gje$GtQHpKNpWFd9-wUS#g5
z7?%uN^Y!Zro{kS?t;z=2ofzgTuIHrzFIFO^$&c>D?`)dU0jriiG@Rn=jBXutgQP%3
z)b#P7F*HImkr|7o7o#ruElN1AU2Gf+yojHN&tAc9#VwUfPPIal<LeXnjPs>VH!<bo
zT1Oyzs3Z^R5$o2q=lDgU!qF3)xBlq5`LOJGDW-sb+2qg0OhJUIoG2}6#C>o6FEVo~
z4Pwb*uOh3Nd@e^pu@^Wq?oM*<&mX0G<~CkT8TPcoZl9bLaS?6d?>ms+$L{UC-*<MJ
zfz?RAJcQ_s`|0`I$~8+$I)vLPxyA~S51Ik%y`<Fix<9N0Xu^HO1`=_wOm&f4IlkIh
zfNu0C#hFQ&lYgiWSkT^avk{`(8s#}~=n9q=X5it+w-Nu%NJ+Vj$lXV($|;*7Oab15
zEF(uP{^a&?P>>?&>(^=)y-U0#?5v~*<*D(^EWzB>{<WWXaoU^F!xDW*+2T??qq*I<
z^)xx3GRJipXa3X0e`Z0$=G8`pfplX8RxlCV8I-h#WVj?vY#Z7BJ~D$T7a3RXGR)IZ
zL|u4?op9?*r}~<EH63T=oozP2(@)($(QVl#shE`Bz$R&g;2TR$M<at1f&H8?vW)Ss
zQb>K8Fy+JsAe4RHwp^GIcS8p3gp{o#rsX0Zl@(Dil{KF{M?}9m&WR2ScNw}KQZoAP
z3ELi|e~j!4^ea}HRwh^i8%n68gcK2aKxg+hpLN<g1YfI;4;hb);`ty4G$oOW8(u~x
zi0A%(IO}X=ZEX0j`YZ(>ykH#po_lW$ZUY!r#OGKg$Jov7j_9$%&4D@dc>m?J_A@GD
zNpQ9y_zpgpd@g;cog0;&TkNR(a}$TBFQTpnmvTa4R_1V`Y(M{B?6v*c2nDiHt|gj9
zJyhIs>gjA?{>wz@EeS*i_SMuB$Tyf*?axdgBHR7(BfSF(cO4HXo=adtoSb<c%Q$+<
zr0!THDKx~dXS9ZxE+>u?po`o%MPOU=uv+5_R0(9azGIu0@k?{V9D0b^;$Lqd+VmyN
z0F1JckT!k;bH#XPfCkay<VxH5d-+{uq1Z?+8@0HJc`wC<HZdK$#nagf9Rx7z@=Eg?
z^(&t=cX1vsQE=gI+MkUG_`()x2HC6j52v;_aOf(%!4&YlDe+uDAu?7R1lKUFUMMm&
zh!e=rE8biMK4u#iV`o~p8&&9#hmYmLBxC-W6M+rcQ{^<%;9D(x;h3-Es>4XD@Q8};
zW%EkAPM34)3&59seao06bLZ@opWmD6+?c#_-;m*ULv!%6RZ)u_2r$dbVdYS+&}9(M
zcP|kxjY`zNXDkKZ46Q6G)VcI^nZ05i9P<}7for%n|IT45v`%&368A<V4F4rPOCJgE
zqYK^B{|PY6Dh%jt2T<~55HdwRl5pYJhN%h^nW{{O-BL`3O~<UM7+)zLc@I5{Tczxn
zmz$8ajX$O~MBex$v?|C6ABUD{#{X_>>j*I}!0VzFWOyt-Y3^;uk9pbFy97VlO{D3|
z&$&-7xWa!_GeBq0y5H{YUSij!cCdp}_y)irXVhrLzb!wu35m_03%adM=a^Kkxw)_%
z>k?}@Kup*^G?&Rh+D>s9zjLGE%`19TxZnc0|5g?Lvo^Yh0?Nbo5&ziF=j=mMilPAX
z!^@H@d41>O5sZ$`QPA7<J61cv$NJkJk9LFcjY)RJt$uC2+UAy_gr`PQ+W5;MlyMhT
zPpDULODj#c_!2=uKCO<tu?k&De3AEhPwS0t7;9J6**ENbI!x5az^bNtXow1@_Qiqz
zkldwzzYN1EGf&G^Ba`C?7P&bM%TImNA(8vbG@{Y9{g7NOrHt(RR=4#sI;oNP<h%@)
z#6Y1cC10`joH%nl2xb8&$Z%yw`XAH0lMDTh2QLLDlb(0}{$!&WHx1YPyyi!ZSbJOd
zqDr&8ROjM@7c(L~R)FD~CA<HIoo(irv>U42es6LjbH_nuw_^L#cc4~&RGd+}#r&yC
zyt`3_S<A}P(yE(GhC2vA%PBzY9~B6H3~~vpL`)++b|fUvddJlP?(h5Z<TOLBp}nkR
znbmJi1=f_826hVz$t$PWgOBr!HMw-!3z+~5-~F;szhbMBiMHHCt9j*d(YXRVo+4Xy
z>-}RAc7y3~@O(nxUZLshG$Nlh?k>eLOS-y?nvM}@hzBo#mgDX7pmrfUa-$g+_3zLx
zwwZ_>v!QOs!*A*X2ZKlWz?NgGgt}!3FWy*CW`_$Q1e=;&G!1sE;*LCU$A+d;mlcFu
z1XIm+!y?Da4)lj7G|jJVmyBOR3>DVMuAsBy$cu`xJjuF@sHR7R#TKlSuHt(0(UXdb
zPVv&?jWzfT%Z^#m&u6&-I_PsYpT~qiJpaoUs{PA*?Au|i8@9lt)9&)ck_z8vM%1*@
zC)u32txB>|*oeuTW5?oD=Z-8EwaP4aWCH|ktde0pf;QqtXP_yn@{-inmcM5Oo{Z6w
zL)veLX>CNVOE{L)a#(Xgp1(S@9u<lF1TNy5ndF`)oF13P4>%BSuOPmoSl0GYK=TZY
zQ2uIbv8(Oegk?UE`e)8(y0?al8fZ{GoMxGu1AFl!rt8{|>YF_sDOR17feQpE5@a`p
zWWelodPN~{ZSspWJk(L|R%l2b+tfG+`0+R=)sB#5)Mo;jF$2D240dMQXcA!i9Gk{g
zgUAa*jGDFzf;T<yyUz1>&v+=ORXUvBP#-R4-Gg8hq;Xh2?&8je!ce6k-hF*_4!leK
zeZ*hPFXy(2bx5Z#CNNXL+&Nay^Q+3M(Co(t`Flhk3ixDEojzjs(WF+d?w(pp5#||r
znKrEvj}xWWtMhGoeV*?Un4gZ{pY<v_FJCjNdGEN^CoECfy%{zrpbex+71}P)ZH_vh
zPJ>JMBTsa}KEJQ8a*Rp%f7q7vLsW`Mh!qpW$wv@nlv^(1957Rg0|o#Hsu&ga(evo#
zyeg~0u+pewyEYxuLz-Dn4fj)g^`SX~uDv_GA?P#Ll6>i-HCcn%KZ3j2<g-49OvbLm
zZyj!IFb-?7-?C8B?sIdj6MBhZWJL*0um&J0^*Ekw6H)*XcPyE*v^+maP63_h*sZNx
zmva0q@E)&U-nTyJlvuIK2(0p-ub(3Bjb&6T{%<)IE^E=ItY{zx&aesFAs4HTg?c1v
z%aqr`Omhm)TJ@UaP2#ch<zL7<i}i#pkv43m&-i<vT>DU3)f2p-pW>7pd-7k?IpuA5
z8Urb5q++(+!7Fyx@UR!UR!ECW!Ex3&5!qb2K>MA5gT}pG8P2u)y)gd!*{4lklk!ct
z*8QzLlGuBvnu!f|RYWoq!bgkupEl>d*PeNW7JY2)iCt%UC;T)c^DJJTQuG&`pkbag
zt@A+^?8g&(aaHkK>0){VE!;RvSd!dSUS$vZIm{v>yDyX?Qu@|j_}FkTue$Zkcy%_B
z>94{-NdTS*aYPxvz0V$?k+umit~GF|%bJGIY@A*K&gX)15o}nyPDOkf0Xh%jk+e!U
zV2rL89p{yMcu=Lju?GLE{5d>+;WT*%o!x*%>y~?MTvA%*5W&1G;!|Y=nMQpe;6Bty
zD(}uJ@8&cu9Q0#tJdUq2{B?)94W#vtd(y9>2p`@WlFtV!^^k8p4^&wzH1z(Rc8-aR
z7|F;^CpBSx>;&z`4n;aC$9ifHEvp8OaZ{cTEmC1c>H&aN#kQj%X!H0I@pW|$tx~bp
z;#FVEGWH4^`zf>+w=IoKrST&TXs|8a(@b%Skqx6(fzHJhdp_igJByq(@}Q)87)5I>
zd@jBerAZFmps?MVnn^xB8jX;J=9B`E`!`j=rzD~IZ(X6J9Y#uNno%0)yTBB`LkO)+
z2>_6Cw3ja2x32!<y!ECr-=&pOiP$m(F~AU_k&B|shj+PY+<Ro#ljkT#e={6q#+1H(
zU1(^fh$h+drXC_<f*vOB$;{irg^i89*zV4#EkNSZwiQaCXnlP@uJAgma5>=)NwleI
z4K*2xaUA;Qeu3kEiIWO`TiJMX^SjmXPbP-f)Y4k|W!(1u_T2h<xAKw8>xLRIT-~=Z
zc*Z~Xe}8U0$oAIM*4{48u-k5j{A>3Z(^(g8)LxsvC70?YdcP!cp5(iC*zmpoYuCKg
z<u6d}cU#{2eAe!{g=EmOrc)pE4Seb`QQbl4_plZHUR}a*wz(8(VuKg!?rdyeN9w)+
zSao=aF%M^AiBQp7iXYf(0F$t@nvM(0dtLg)3v?ZMX*TST9Wbpl6PCu}%tfnlNALG>
zEa}(Pi{Vi;ldoE;*ICQO_|n1#H7Qo6#hlpOYDG-A8oxXVGo5+bk4doM`40A$#*kqn
zsQ@vTi9+9jAYBKwRANe~$tg*!yOa4b!y>{56&By8f>g*6Sx4J50zcDTr^j@3NQ4pz
zH^`8iKJ9!@cHm=q9~6RY|D;~*xULSQiEZ~{8qH?2%}S$NDy#N=SE9Ky-1l-+;;~-O
zeFhPXUyEdU9jx=`M7u4v9r92<dp%B@etaMx%Hd14w(!oe)BahxU6xUY1o~NL;!`s^
zZ7SM#kg|uY1_eg#5b(VVO@WL~I(ic<vArIda<c_=!bfP*Q^SJ&!O7WhAKAtw+~oF-
zYxl@VuiaV}P+G*$(dsRFy8lz(@J<f5z4t#cw)Kl^Rs9RTU;6T#Gh~eoHa8F1<Rk3m
zP2^WFGqo}g`px|EP9;oNMJoBKK0N{Eg6tQE<`l6Ng&+!;Wkhpp{zgHiOBqrM1mA%p
zyFt@Ai{gEQo9oR<J%zg>fz}FdM~>x6o5$zW7yRh?s%SJE1;o!WVzT0a`o)~$Oy!XJ
ztk&7|zMS^6x(O)-y}lZz{<ncxy;~S0U!1{~|IciFX25|YG%(z!A?vcEJMMY(=Q5cV
zi6iy^Rv~aWPvC~6i&na`;)CWE=eOobB-Z+HdMc&mTRO@20sn1^N$<bheivuFG$ZU#
z<f&m&|EC2QBc?-}CJr`pU$fA}P3WyD2miSph`046@^?SDe+YTt;b$Z$Y<|OeyCv)D
zJuABIQs_vtvGiGZzPMZU(Yileg0<NarP?v&bWb!IO!!Q{yh}}kFl2A?VB6KF=KoIg
z+dVc5kieYNaXpQ?|2-cH_6xvL#OIokx#W|GAb}1NX4OH$MZ$W<=dYz2ksi~){vv<p
z1Q3OQuoRL6ZcC||8bepELBf$<z4)4zff+iZXJJV2?6g95UM|&hT;me_@6aOakK7K5
z)6Vx#AmIE@8U)&Wj;oJRAw%=Z@-EI5A+MEjzLjx902Y!0Q1b}Tve7u++16Gs&cH9O
zAalr7B@faRQE<(bO+8GP3M;s<B&-_#_(y4irid%oxI<}}9#2kBU*veFdWIWVy0~+w
z9a>H9Jp?6RW!@kA0<-Y7J?wI>IyWLX*&$_ePG6SAlS;`I*uxKdOeuIz6gg%5f@Ax%
zjvE`@`6F3p{m+=a7m0e`M$RFcq|kUO`7%(LzVmI4*1{_w&_ay;3SykJP?80w$7JfJ
zohtZ5r8IPSlp+KPH1YPq>l2!CGL*)#>V=1x_`G!&@IF0nwc*Kr0gZt_Bb?TIrnMM!
z`Lwo1FT;JuKNujA8nhx~fTX|s=_4${KhsCsnDz(sY6Kc78CN)vyABi_Bc#ApwK`aW
z^N)F}O}NWcDjPE`4}u}N=!gIuaClYiPn{t5Ig%AFgjQ}t7&J1HLMjo2IFatp!Tr<a
zRr3V2LVQ`8s2Vz9%&^WeA(8Wd`qSS3d8PGVRr|xCgup1Un4qGfg2Q)K{kmMOgb5??
zwj%%;$v-|(<TODhdgzAp_@}ecX7xVlv(F0La<%|X3@_kGL&Ul2&?i>Cv+fPVW%8br
zOC`t&2>3iOg`pDdH~al-_UI#6?+z1qF``W#6^m^I3MPif1V)GV-O1K4`ofw(vrI#f
z1yM>ug1k0MJxnW-p+u1npYlyz!rW|4a*Ns~n1m5~iU#_`7Mcxe#7TL2Oz(a@HDVY;
zc8~$(?fW!kT;QUL=D(M#m?6RtNAiEmQ--y>MRfaDb>wuQ#wnUFB{uwjxlOasq5^R;
zEe?p?v$G&hc}R)1?oxr@9TjWRxdgV4wX7B91cFA}aMtd>vyG_@9!_6tPdrP*M>Tm&
z6FWdRRFhBh3|Q>`EWs?i{XQrxa3D(c=_9MFe*rI%E0M(UX;_=A`E}IK%rtRXvIIL(
zCHN}6XSQnZhYf=%KN8&+#IUz#g5R%;RA|N<iEW`HizfNo%%Etpx8U6EG5Sa)5<$s;
zMp*x&>EpHN*NPqk&tRW>B#zgYQrMk|St#KA9a_N~!~DZ>Fbp{_z%+Q48HGC=sQ5~z
zfz{F3?68ToA!TOF`C&nv(8@|uB04Lgy5-#^XIi_^0ov9@zTR|g#f#ao3|Rpl0ItFY
zGctO=^?|fAcPH#;7z`{Tl3a8ec~{muw#;qIyDw!07mxOZ@3ro{CJ>o!_*gO(Mv=4x
zeYU;=48|hoO|__@GuMgJd(mvb%XeSd;ozoxg@<ZOBW}T;?xr<&uWzuaU@~$^O43gp
zVAxJJ{~?g%Vh^JbP2h+Uh7Cg5{fYsoi@)pMIOhfr>TX$&Y|jaO9+q|ceqM0rhVxCO
zd~Y-_Jy2;IrjA$YZW}MO`VGyV`X&`vQiwJHwx<;mq)>cNMyHN!I$I)ui|}u}I5_KI
za9qds5PRdMyd~{(Vx&5SK5u^#xQxh%arC}}*|#hxGkJhBkagJO40suQ2f5ZK?Dezn
zCkOl4h5Kp*?JiHrqg?;JeK|cP+JAaG@DzGf+pY_%noRiS0dsV7o5{_(IXfCx(Qsl9
z+khcgwszQ_G^LoZ=w5bFnjWh>@&0bf3$&}h%AZ!*bSMb-pvXgv+OQQbDP)4aHkc~N
zZOY(KbLb#x9v-j>H_`_#Xv!$#Rx|>sQAn?uA%5It8f(z8-^^5i@ei|_T#FxVF(GXV
zMk<9-)LPO)ZS@iUZl&~ybUxmG%wdNY*tm)CyK>9HvXck*mJ1^lT_ut#27ghb%S-~m
zKeTeJLfUKpVq!^xSMkX{lM5Lg6;Z4;6tW}g+|5Z^1{m6#wmiP&<h!0=WK{Xn`erT^
z`S*V-e~%<n+2}_>@^3x5k2X#_VD!V}e`5ywh3*ar{JB+O=KRwS@O*@AQGtu@GiFS6
z+K8A^Ta=LA2wox@=s`<l20k8Dh1mq@K+d|qGxKw*%7~&yG?hr~7^5bn>hQ>eh*g<U
zo^Ri=XtEk-Oi!3KP{o65i}0l>Z9ZzLGf~UHwmRy+a|sx-BNod<Bz!j=0(xz$qj^Oq
zOs8*;cFPX7zk%=W;jU%N&`_CW9ra^m*hGQOnD(yKvuUtaeqy=CP;pu5BAYDO1_&(X
z+s@JZyF0IU`^^bxTQb&fC?N;l5I7U$u4~?|K83ek1$~DV1)pt<!cv`#|B2RNnxDrv
zUgXwa&)f|z?$_*6b?jP0c7|eEIXRp6IEbF6!u|t<P9xXv^#aeVXDvIQ`5#9VNSonj
z)`4DH3VeXD{yz=!1Jk`1AAU3yGiygsHG$J<$3`zwTzWRz!zj$h-6Av3xSi-fuY2L$
z*5VNr4Z#ugBE=J8tWNy#^KJ6)UikLBu8=}tsv|rgIzUB_(z#mGg{O={v6Gg+E%Ve`
z%oJ+KroS#P({N4_oSTEUe;jeNNY&B(sOl@#__#ax(K}5hJgY3bp89WUV&aLSMkwH|
zs4gj#N1vTUU>yPsJwV8;t~j)=dH){8C8%8Ubn~b<DPWPjAMY^}XC|}whBOP|NBQ4R
zh}IpY?Dd#+3tkF{f{0oUq+;)pn6H957k}|9UXsTLk~0FkUGw3Ue=~TmmfOTU3U|%Q
z-51cQd-7`Gd9RTDOwA;ll&6%MYkUs1xfK8N`feh1L9cPww0oGzl5w-PK^Oi)Et7{5
zPc>CA+4_Y627E=~X@me35YMzRW*L<Y6y?dPP;(j@?io5wdBmt6dagM(l(c{5KkXm)
zP@JcwgTwEqo0a{pni)&Eo_Ey*hF>-On#Z7O&+y9tw<u(+*fDiigNMw|=v^<*Qg8S6
zdQL0HHvWizGkX<aT^P_`Qj_hNo3ai7g&Y`=JA8XF5r+Tu`$6{P&a>x?$EGi^&zW5w
zrP}n@f$UIV+LUu%3U(cxJ;g!_YiA(e?GKxh9r_Z&V@UqK6e5z9cEZkg3!dJ`Z5z7o
z?&zhv^!R_(4l-|QgbCgNppF_M3?tB}y???{%Bj#zq3`B|f(=0;B(Rn<q-YmUf%<0|
zno<p-tS}EXNn|v#JGV*Wr!)1)OXclSh6ap<0=~$q(w`drWR7_mi#B5U3~<7w(Jwkk
zw&6Y_q5}ty&g{KPHA}fu<gG@<*j&Pjv~5~K3jmRP_RRB{-1zY78;dQYmcjDh`-dmj
z|3=ZdrtI`-6|{l%$Faf$XbyFWN|(t<5;8K&R-XFK>QX%}a~5UC_0>jPZ7IyWSGD{L
zefOfgW)K5nv=bZH3~6=+#29xL*fcXMMdA<@+G<KXDXYW1^eOlJc=;m`>O6wJjZkcF
zKI+TqB=IDP`=q#KXw2%jkgzQGoSGyeZ=87><+De(V$uV;ir~VmHt%majh!mJJkKyc
zjLv*)MK(;_zo7wC^iMqP+3>hi^K`GnzhhDLwjU{tbK*rlX&GtZCowz=ws^bF$VTBW
z_5ZH7)3wu$+s49R?lc<xsU~NI;z@!wPg*4U;duxO=a6=5YmVerPziOdwxPCxbv&24
z)QeELmu!qCfJ$*%Y{_cp#S^tr_;a+zZ#cq`TCy#<Zq<|eOw7U%#~$l!<Ffn<uoShP
zyiCoE<{F2&xp~d})szUTEww2T;5t#&_4Y*H9J}+2=xvK=pVM!L^C8*b5zB9z%5OVR
z^oV!kQT~Je+f{14`}sOjcHe*2$w|upWojMB*?VC`iwEUXdu#^4wfxsZ0*}-Yf&W=5
z$NJs8AS|ks$lJ!X1NOQ9ZHOMQS9_uWh>W^uxGpM;&?`LWWh7_U{%zm>;ig1j{J)QP
zI>>U@pYL1-N)c`LHGK%tQy$Us!}}2bU9EpV>r#&!uzT`A*8ZbWc)*=d^|XI#1keK{
z6xiTNEU?5hs>D<kgWI;>Y=hUER6i_hs)tfg@o=MpkRSgQeB<{n8w!g21Zm2=nkynf
zrKbe9Bmo3)b`53$Ig9}}ixfhSC<uX%1I$l`s*N@BeQijZGscZ3|J2E-QUy7PZ}QXV
z>_H9fY7@!2qPpf6`lKXylEWgHLG8f5?8$>pWf`o)DDD5y;k0q?Vcx_;YU8Ye5^_Iy
zr%XC%Zv?Fsw+%u4+#vtW0hl5AnM)#$_3Ye#+0U~5M#g(3H9BM9`9;Ig6F1gZ8!NUE
zd3645Z23NTS<Df8tC@}3D)cIU%iE1$bi(U*iyiI$=ugD#ALKJhuy;+uRT&Yje#H?9
zfd!{gWaT0$iIpSw5mZ*e4BsCAbjLSXJPizXl}n9W5tnEwF?_@q;+$b4+vBf>5>}`!
zL3iqNNwy|Vi{S$<9AK!OthYk!-qTsVZ*X<41-U}OS$kt`Ki^$Mad)o4a?HhY+7s<Z
zQ9&1n2VqBgX;&(hy}x`DZT1@yYcB10sqC+QivQcGYI!PN$QYb#Bt@pm`_?|m^j}#5
z>k$9AjF_}-FI$CFkEpK2Gr9#iI_=Dsi*TOtufQ`rd_#YXXcZS^O&}kf9z&Mqf>K1V
zM3qSigHg-)-ufQ0nE`D}kEOso46-ufZsRdE)6^GFI^qP<CCC^fzZ4tWW3(d0h(>}X
zk|<h@;smxL(`+PmDO-)&jxRl^3KFfPPr>edBujJfY31Bv|K8MxW%;_dE3zA!WYv8q
zVo--02Mjr#r~wH+J#wGZbrxz8`OH7p(}BD~VpS*;T?`k`xcYQVSyhh2_IYimzTxa1
z6EnpcV4TE_fkT-9^Plaam=eBBF=x&UpeHV-=+otWKXm?FP>E!4O%873*rkg^h^;JY
zU=>lely`OB=jE2ltJRERx7=AzMe2kifdQNvgM2R5W!910^K{-dqx=YcwzHr;Ey#Ea
z0G2F(lm@h?#h%HGO_4d;$Tl@0&ACPn12;%|$0_oeO$2EDcd%m9(kH*q6&%YP=haKc
zDaTp=gW5urF#=>jwsGaw<X^3V*D0OcUwME{Q!hcXo8J!;cLf<KF3xgR$a8{9x0nI;
z@q%Pd=X*w0lG5Xf1W7f3%m+#VEoR;zuOzPbF;)X?0kiELQJSv6sto#!$YV9qh=G+Y
zKyaPE#;)H`GKh3P+v@9{d@!yCAR}2u&dj5OmFM;oLyQSZE5@SlueC(a%pd0G=l}bo
zckXeN$Y*Uso^>7e8Q-=U^{;NvyWaiQU-mfxT?`Pj0sXOg&5gj|p%=$v{W=@2#p`%?
zB=O+%*;(M;`9oB6wBya85)dlH0N5E{fbXNC^TT!QJCpxIK^=FE)i9d}{%;42p7OwO
zU?%$D<=EzYy%Q)jF9`NNI|QCU%Qs4A@HiUz)B-3ja1ya%Ya@oHa|;{AE*Xo6<EAEL
zW)Ee&@{hL=INfH(*bVgLf0e;V()>0ckH*hbz80l2p`QcYBLST~kCqUt3xAQf&c|$*
z(H{UqwpP)7PvL2GSV%G0hPf&}AYvmDKwxy8Oq>)UNNb%(s%l(N{^BAGzm;H?Gu*cL
z<5t)U&OmNwxM=+7<5{^ch6r7JD_$vK)_oEqd6sIBtG?ZWeE%xjHvx+|PpC*t`871J
z6eD1NRD;XhY`FJi>=)mw`wG`*Fq<#jGU9Qb1Qq=SrYyY=K;ck%@19?}3Dm!(TZDaP
zIC;mut9ZR@y3@&SNi*?^>d2^VN3UUV@$2$VN1vG+&)K+}qOo~O6DLST)<eZ)ZQKkz
zFPchZFyg)XX%gG&6->_>0X7hDOz3qeOWctXBCE)Dub4`V((r*z+cIr7gjh(-)7cpd
z5|nJ;LjyRWvAfibw^V0S9V(>TmUkFMBSvGmArFEQyCS!cgI!S^`u6QyOz$1Yfx8fT
znK6L2sTHmR-$L<wUu;|!9Y-p5eo<&u9NDsTbjiiCWpj42z05AQx6cFPZ`@`CbNUgC
zY`k&|Tu$;n#;HPw=?lLcH>2xbA4i`a9QROfl3O%kf7jIgv0`hbXb6%W<i|)&3x@=a
z_5H{BTM$4|w3=7QXYXZd&wy@pKVWOpW}CL*GwC2LE_BowtLuzmVPHL_?e`Qx+alw4
z3;rf=ZIHClR8{$-Cmwf_E+7Ud4kLVliA;!4kLot-Fq@UpZ}2$EgTHKH5d|H4e05Xi
z2inFA&2nfUhyMM}6%Yw2^ISP|(`X$Y0L+jEDtz+Ff=ZwuB9By$5)#%VjCEuEG{Q%K
z^oD6Cusc~_(<<4;?)?ed(0H))fD43Dw2o7nvtIp{sM6x421z6II-f+qwvs4Tea?vB
zBRn^EVf76lvm@%1otb2d9$u-#z6dV=^xO(e<s^;A*2TNOkC*6Jw*73^kg|dgiW{ao
z`izvq^HWDha%8n&Wm2>SQ*_>9f7gJS$d^~m0I&_TOG!x?C^YTJ4+l#r<Kh?WpX{t9
zjVw~B)}&iw26DMX>zjb^Q}t>Rl2-jGg>-=5!aAN5%}yIp_fU7>`Vfc0MoP*vp3<L7
zd6>T=g`~1Q78unkOIFm#yO5A~>ds7pSlM%eOtAKO@q(09k~X~+Z^>^%C8%a@SjMjp
zXo1-|)#eANB29S~7TaMqHmHG8eR+YIulUt12qBB^`rY$(UJ6wxDm-cRw=1<)9{Uah
zSQ`KBPCC55-d|_mcZB_y%6MRV-Gl0f=1};LPX96@lJNWt03?R`0l;SDMG(;{yu|av
z5I_V)<WVE~-Mjna`~ZR=IvcbrksvWi3GN*`-~Jctc(K)o3>#z{)_iIRrvR6IU0(h8
z=&65`k8?TJ^)mKyGWiC2zB}FNJojaEUv(T3zuz}*iGv&aoQ?POYviH|s^$qQ*V_2v
z+~Ri~LzJ&J2mR7D<#_0{Bi72>Qd8Lf%>oIP@gvb<;9Jo=XQgBa0dddV2)x8M<4EA#
z;>|Zc6fe7r4Urw-$5L>vOC+r@EZd(Uw3fej0(vo-1?2-6H9V|TaZN8Rq56DF@d-Fu
z8stmKhjg1UAI<^+#K5@TXnh5>PpJ;=axGI-QUY|PVHpC`v1?akt2c`Vd4C3Y$4jFg
zhF8}bgmDZtl2RfV5sLD$=myD917tdEMMq1SPl`IZy3k32RoEMZHKMA>xjQR}31H$W
z!u)>lPOgTr;WXO}t0pU)b!F)4%qL{(8x2b%jf|RQDa`9_{ZqpJGY8dB3xeD{Qxm<3
zbjW0uh+PyB_$Lhr0T`sgxo^E~GTXdAuhyY2yGL_u#BRce(SpKJ=$mlM(x~=tz;5y@
zZe)|Hb0s4aZ|VLZA4(w_^6ft7niugHAP=`na`5Hw!O4^zNR`##o1SJFM)txCBn8WG
z(gPUE7{E4Xz^tn_3G;>pXTnGVvRL2v<7b3~Rp#V*4bw`K4DN1dVFO$KkDK&`c>Qah
z;0biAFbvN>Cg0@B^A}g>4o&fY@^(?9^+N!ui{7D;6`-`^T3s%AkrIOM&3K2|8h?j!
ziIG|kmqyykQs|9}7IyxC{RBhmh{R%|$6AQpaycfl^VgXp_;M*Q>&YB$Qe4$%<d#(p
zKf9RKE3`_-GFpsq{BV0xRspbGzGcNq3a+LvsUQzg@y1_VL$(r4c=m2~z8(!);VF>~
z_Tg^(S^C#`MI$iTuSZ+kSe%1vk^(yWx=HW-jptseMqb^ls0iJzNrkpj<-(11yk@E&
zVhq$kQxmR?=u9{D=}Bb9Z~);A!GAPKbJ}?TNN7ko-aSVj3gc{r3-;%Uw=vK4>OS8v
z+<j~QzZ@{otn8@<mjFaA@7o4@H~Uo`6{at>O^Obb$}~F4&D_ccMlK0yBK7q$p?e>E
zYiCQDUPO}L7lY(arO4|WBFpyM(GGuL#UirFCG6SB`pU~TyAtNvGi`7s7eporqPzXX
zh7D<`e6G=BQzzQ4-HyI-mZMh$PK^*|Ndz0Kk)5o}3O4wUuZqmy;Qq|}(n9jPe#Kdf
zfb;+D!8)R5J8d~&3vzAi8B6q3$8g<IG(6)m>OR_3y+1R~ytS*Df<yfe2Sl482<Y6A
zBs3oYt<m@~@lLm9i1Qi*(e_aw)_tvCZ|ge=XrfIF9!cD=!i=7tMyAj7#H7Md6TH>8
zGA2#A^TC8Z&&6n3t5iw(`%_0n&g2q&Qtp)E30lV_m)}vU-_M1vx7(MaoHv^C^75$I
zSo@bMkPHA-?mt{0bzLB({vmAvIB15aX8@PD?z|hkY~ygEcU)gz69hO2JOl|0|0h$h
zLk!n>e}4+jpI2auPf1O6hhP8Zqj#A}<#B($(M8k9ET^)5*VJt5{eBwt<$Uufhzf>r
zU4x1S8MPYu$ed=UHjE0C)|7dMV_LLTHu5X=)FuwoZnl)Is$Kh%d|c5O{YIs4+2LO|
zhb@ST2d{(&52N<^HIN=Yne`b?;sYZEOdi$J0`hN|h}y-H!vha8)KVe+|HQA-LUdQZ
zJ)jH_rhA$2yrVQe?G96aSHzt!$yHCttOng^ofO?jL}<}Ks~OZWQ3HB@MqD;`24Ir<
zvt?72lj>?VR6^4VM0Y8~|Mud3x%})8YZgi0Dd3ELW#hfxW8D&&PK9UA(A|x%2!R1v
zS-~gm3ep*~+-ZH?Wg{KQ9!$JO@oO?6z57+>SpiDjUB2?*bQd)qAWLgW>4H3_dfcY2
z$29DvKrD)IuHz57!V;qrZA7T%B)0--mydh(HYyd>DYGF+?eM`4NVVz>8FS|^)x#U_
z&BHD>bO?0hC6Gs)pH&`sa8Bbu9jVK2F=sJvYvabfP_5n+<u#!gNvjI!&%U0TFA&?k
zl#hX}ich}AybscC50U&mQBTZG5G@I4nEp;SEAB-lUAG3Lp4is0QEL*U6c!0ZNrr(I
zOy{R7xYQyC09|fr)XOr!SliQI0P7AMjqJLTarTapVMbe+(g6vDMFd@79bHqkAC>ZX
zk;o?O07J6iLmOb~PG(rQT~Np?@v)-FPWJ978(LE$4<$%Ae~wEZL9O+68xWS|ZX@95
zIgz)xp_Cn!$-w56zyFsOnPjHnd+@yU+@JAJ{Uj3;Jpn<O_>(YGom5a-kKZ6=)YfGU
z4zWOjBd!4I9T#|zaF7X!pyLnnya)22N4G~dUsFxUM+rgFaN^yR*|lTE((>c|L^lur
zla)EOHMO<5e5|$uzlGyQy9Mt1)y~P*`^=5ES4h{nyR-ToJPvBXgLyC&ZQ+m7yqKeq
zeC)@hPeR0ec6?KXUtJzYzsCw&r5AnJNKN__`>y4SSFQPIf?w>tZnCcA2SblV3JVjo
ziuRl!U-j#B*HpXO=BVqdBGR{7LyEr@g%SIBpczG;N*A*V8+U`mLQFwUjSBKJpCBXK
zZ>%?&ID^6(N}u=I>{6UIRU*M&trBI1+^%aL(UQusKI+N?QH)3C5z>d1|2}9P;>?o8
zUOLt6Z5tj+eU2JJMjX?vHslylPmD@qd6-TsKl?3~1cmBd#w#aSk2wXx$~-xcx3&tx
zEl8#%HJFYHs=sl6&HFB71PXEd7hUMEuP4lf@MG1ori7e4^K&hW9UpOpjNtR`%i$KP
z!&q<0!Z-i-?G3x5ue$KUI}w{hyz=?^gx<1AS@8~4AX=sq{K%z+@LyHIvTB*gBFebd
zwL_O~gPJ~WV9Iz|HQ|^B))Wk};?pUS?m#S$Y%}-&dZfwo;<T@Do0zBMug4{a?*rby
zUx)oE+ukTXl-Hq{Z4_kDlm&#<VFz^gO--PTFB<S2pf-P@q$8PWbg-;3x1s=s=pU!+
zWz~J|*hr-zxP(I#g#_^?1UV--H!P=rikck=QJ@r@CR$vABZF`t7a>VwX0SmYiCzyS
z76!O--d_%m?w>z&oVix>p4Z-;a5q~uCMD2ScO<|Odi;PR@^M&eT3P}^O~jNcB0DF-
zz(gaSQ5TEbbqW_T4o61CUz}F3Q1EMqYUSm5|6<O!E}s1kJ#^}BOe~uD4z+Vr06q%j
z`&bAhhUDU6;&XWi{FP>}RK-ALFTjRlaO~h%{NB%h%{NpmlT`E--fd1q`)7VKwc9-H
zA63%N%`84NEye1+mKKPOmH4%tfIj<9GD(Iqb|69!ak+H+K{|)5GP90xaV<Pfke4&_
z#Nw-!#UD3TyI)l`)t-J?E84rjo0M_6mPha+b*D0|`!4l+__pdFmT!sn!(!eUH0a;S
zO6uapCC`ZbKNDrVbv>jk27%ZHae?$5o#WaWUO-M+MPI6Pf=XIU$4^&k4orB}Q5eUL
zfk{+aX(+%p)g@rgilA-;&F6N1kT{+^SY#FGrHAv6{4{2u&o0Z_d+;`*^Uz+t(Q9On
zqorJvU?mMPjne^JwN!hNDd@bdrVJT>1|R%Q+ok+1M%YNTA$woxUVBl}C49SSKs?XX
zeXTcH_}r1xv+{^s^&yEF+;BpUKVdp~=RB^<X+lG@xJF$mW5=_0{G;1heq&D1Y25Vx
z(ezb8ac)h!8w>6(!QC}D1b3I<?gS6+1b25BAV_eA;O+!>26szvcQ|jpQ~$+WP{nXT
zt<|f$AG2MjRL~#p65ceM!u~qiht)UoBR0cNuiUJZsQ$;<4Mk&^Sj`9}nI(&@DPL|$
zgIgFuJQ&A|TR{Xnsd-dqD%F%3ZPE@74$|aZmC_m(!PyU6J6tf`FdHIIzL?b$oC7)x
z=8VV0@(GbX0;OF{rs-OKG*qNF=bxF-Jvl^i^)wRt@t8IIe_Q~DpW(^n@;!HUUW+`=
zG3M4nS?Jgu;K0k_PkjhfFXv%B*ot4o+gj(=J6tJC^BC}+_$@c7w9>KjVx{yCxMtiQ
zxA)dKHCquL$Sz~Vbgb`6S;Ew6a()2#%Y0p=|04ly8cc$4T&5;!-6AJsEW%3rtaW8-
z{Z}!b+-R4KZJbHXCp>$&l@=YJqLb=lG>xHT#?yQUtW@^{(>%o_yqbP$NVquE?5r|C
zKgGnRktjq(XQkn`OBRj`(mRR<DZp53HdSJRJAF*UJ#T*>`ZlHUXYC2q?M0G`BM?c<
zRY$qjjn<#y){NDVcqyim6#f9&TlvDxrJ3A5T2@=qxC|cWxd%%0N%4Hc`ih>bsDlwQ
zIW0R`MVrJ@J?{QY4vXHb!5Az<!uKacTYFjoSM+kL`|wZo^;m!*N@SG30f!M~Vse^f
zt63vpAYc*5RkSxeJ}*9`Akz5Tm?pFZ?PrkVUkjQh$YkW}AcLWCLP<cc-xn(CAi%p`
znQV^*{f)X%j+Z2Zuxh$%*PYwQ6Mm;HBhhwZ(nrxc@wyRvZ2aCe)@3h5w(m}_r^#->
zKqZNB&<iJ}TtgBGADuRg_J$_RW8^OHHgTk0rO;4DjBmc6X|b9D`Lw~`pFuW_pGjE|
z&ayx?JoykaekV`nqZr1;zlBc6^ZNP0m%!UE|Hw>V8hry#ApxjCkgmH%Zlbtr(qlhE
zrNb3#7AntB%>1L46YP#az-Qv!3*9tSCg_j+@0gmeoU3j%w^^>}M$F?L*IL#_L8{Te
zNbA)RV!=6_Cr@GKAM)*R8}1tJAbOh62@TF-_WX$h`6l%ws))h|&-`QM;=~A{ZrKyz
z2Uxq7Ww&t3o7_yXl}PA@ek#QwBhf0$-;EqI&-cG^Dphc2)%)E_g0w5-Qn#ZoA39Ew
z4Q~-tKJWE!r^Pyr<KMUrzA7*vzmn*;CU?JWpkHV$mu&<R#1v?(_kT6A3INN@utkQ-
z6y+^T)7(S}C+N{f387oAZ|apFc+r_xV9ku>)hx7*H?V3UK<$EMj*u(BDy7^CQ8W-D
zvR*G56IFlC_ss_<x4j1;!COR+_`yst%s0!d!2spT3=M*@;LA+MBDrX23ToqQ9|}?S
zXH2`nQSV`nC12#*Q)0n!krj*y!}0D1f1!spW2-Nh?Be)8!JU~`KGI;e{ZhoGeMxXK
z{g63G$;|R^l8WF$!y=BlGTADa+SScbIgAs9r*(={rUu2FM()I$nJ&B4Xd6Yc*#4U7
zF}b{cc%--7ELlG{F2m2&!Pc7a)gT2AF#V6@_G}VsENr^dYfvAA8DsdBN5Zd|C=Z+E
zg1qEtMesA1*gWv#^$!x+(XQeYTcpz!F$i8x4^6Fat5wm78Y&y<YSnc5cLl^>{l2P}
zHb^N9bO|%^jdP^=kyD=}dcm#57zm$4-l>Fq2nKnn3~H8i6J%4jy@p*I(te~_daGWj
zrY8ZQ4*tpSgsbl(dBh&H)lxRd0TuU{8yc^Hk;VTZj=_WhF$>%NZ#IIvavCLL;Yknz
zyz(!}(3UJKJq6PiE4b&hqaxTe3H5A|Fmi%g58@bbHU6FKiFQgFKHTO`AAPlKzx=+T
z@#-qbvzw*-I8^t+p5=RAGQ;2ut`xP<-z1qmLDUWPEwv~=S4+ONQAB?^s;eyKXMnM3
z&0*ObVs~_zn#EXnltNqvL;Pg6&Lv04x6HOAqG%#d(PGvC+v8_^>C7a75ZqIrYtz7!
zDkRr^;bp>K@=tYU@PdBU`q77UEX&HfU`hKn;$sjr#mDbAvi6^WE&ooCVH^U5ftmq1
z0Wlkw%Y}-!&oT1hD_USPpSz#`M2x}JryCt*oX$1(&mQ3d^DBJ-_WZTEfXG`BO=%wJ
zS~D7^$32ka8#qUU2Qtn&?p#s(Mvfd3kh(ngHH5q$&k3gves^->Hg?#$-rtb3tc&^(
z?}Pk5$@`w&Zz}03<9tUEnhN3BL+A_clZWJKp$U0t`loB0eshT+nn|L|tl{JDlts^c
z8Xe9OdGqeL<M^smxb%kew0Uipe9B81{?P-wB<M1EE^lK!a<Yv~$D}Kxy(rMq{B$3a
zRV3hLI55|<`+1$2nb{d2>|Bpqi=?!+@&icH+lL^~V-{$6)&<}%n%)7E)E^IX6nJz}
z*sY#mnByiT6_whqx0gfFK%>r0OldD6mq$yh$sPvq2_~zs*DM@P`(hn*2KJ!3dg8GX
zXS0OG3dWP!+`!LI@q^Ct?_*@N5lH3?)ypdK>_n;Vh@}Slw4<bg_Lv`zEJkf#^@i;-
zs~)&16NVB`3s9Nw+)67bXpCLGXm+>bDX~SAHm6DH)3FCf0>riuM+;?WaC9j6Rwg5b
z00LDdZIQdfxwSX~0TbL>bE<1c3I~K~dU1BdK(L5~f3ZG(bk7{}J*dQJ66`oi-n=&d
zO38Nij{E$4kNF3Oi6)CV$(QDXT8iT?Bifk{Lquz~1QTJ>xmv<r1@ZZQ<i6w#3kt3=
zh6PMM_`&*w&l|m@NQzX&*?5=mMdJE-8SIel@kwk)&I67DS0xA)I2StospDP&_iW3}
zSC)V6!YJ)rze4WaPLot9#}9B#werm_>hJ5B0{k53lIRfq2_hr?o}QE{S6V&d$7tTa
z-f5_376z=hu6`Q}{`SoHOL?jWOLj<QBc|<M1@S{fVh}+$mfL~?gU3coo3WsVCY6(h
z@Y-jelbVQm$}+^P3Ju3C)sJ>8=tvSgcbQeN4cHKEexoanzFmu{kw%oxkR=XodEFRs
zeY9E&K+FX+)Pgs9u~H(QWaqQ;IGcczpkdSzuANqlDsjJGZ2?~E;F;cCBshK9QS8?q
zmm%VmH(QG;iERnVv4sut^Uxxj&9`O$vrA=ZKzb3Mmu?}KYFd*sr<@lXl{#7C)3K)*
zEB(}-AR>J^fsnye9K@8(0qD~0tw~NxnQw4c^ePHfXj0b}-`zS`xHa{+q?R|Y1nr`N
z^NVCr_4I`(>W0AjS?;mjS#|?DCwY*9`zSO9lP|*LY8=BWgkt1b^n%i*y+tTCR7<?1
zyU4#A=S(J^mDMYVkotTitx)b+si}3zwsk9l$uIn$vbc7zw4Gg5G33R5?9DKTAwZ>A
zPRzuc@M|%bz`atEUs0>%vtoON<Z)pa%gHAf*O=4X5whj!NB{9#GD1GGz?50#XDa7H
zy$+r0I6cfRk>pTA1G?05^jar!(+U#c*txl?PEVkoDsJnGks$pw=a4h?vf(emkp<4}
zhV+0YjkBD_Fj{X)x&#ydC_yZSEb@@fY+o|tSfdsQW*1Ywy%D;vs9~F*?hV1~TS|Lb
z{HD(=`b=i)%EYSMEUw1lq4OAFT%ID5_!N!9?S&^;9qGs_n<*S_qmp)1vp;swNGxy(
zP78}&_mY2cqNV;FHQ+>0RE5SkKysH`4*B|4CF$ga;w<kRScy(E!x;!~gS5qer7s)x
zf-S8RJKHg5CS!B<8T`?Z^V#Iw2G1rTk6^ZUApn8Od?|sa_rFf5TD!1goO%mec(5$p
ziAG0ncKh;`Ox~gd$o~Os{(77AeveZnf(T88$Gv|Om(@9}e)_lbbpdc`z)Z9&;~&Zd
zY<oZcXW%x$lS5RN7X<*_Kh}JXt8*N??>BtE7>l0Ts^~X6KxzytYu_)>-<?ypK5iq%
zjfwOM_U(t+(U(h!$mkJH%nm&zE{Zd$^)ucMMPV4X^|11$BS5q@(S-1x2L*&VESd9h
zBn*`^sPk$*oH$BKMEFYZLcFk6%4t%uR1lt8jTV;e21N9TpTf3tu3_Tukw!~0rSg7p
zJG0C@51A9$<KY0W)@K2_J?zd{dFtiWC4=*(6oxLZs<F;%L?)QQJBlBkP5KzVtc5(A
zV>1sw3wsA<Ji_fJE@OrRvMQf8bpk)buv=u(em-_|6uOG=vxba4?ge|g6dkX;vqGrj
zInyPoQ(KiMQ!=`GCOo@|MbE2Ntv92(57M_hAJ@1x`XOT~{VA-3+Gx}(0#r$Pbln5Q
zU2ja9*SBBSKL*_~;g~As<~vv|u6s}|Grc3vaS4+0YFlg}X1ny5smx5GnCB4e0u^*f
zzBhdJ^kVaI@vC@CkY9CQ$OQb&`$L{$z<$c{T^!kq&}GIu(;ptTlG~kk*$k}CwfZvB
zs=wB?tD6~<YlD0;dr_`Iobh4Dh!8y=SJgzjB|*IUN<5D9*1?Uanj##phy1N=4{6@H
zwU%E)XVTpBv_f$y-vQxM0g;b2ji*Z9RN2?si{%J`HGfKT)f-vlI9$A_f%^i?M@@-)
zVuunUl@>ECUfy|$k$An|@hf4CwfV@MfQf5?EK>i5mER)e9CpaZ#-RS{SO76n<RQOX
zr@U8XvIKu&C;00|Y(BFnzz`d~(}O2~E=^z>iutMEptc~xUi+Ueir!&a4L#ByB+cxt
zcR9iG(R%6IRO1NmV}}>1xK2?I)t#`FFaJ;<SY5h^r5$Tf=LD#1OLLK|e+bGjpABYW
z0jJ!39Q`>xKZ>8B$(QvxyE-SpJe^laVbsk|&TI@KzL<W&+E0W3Trxfiq-U)X0LH8e
zq`M`Ndem|;xx(g{uj3}M1w;rtJ?7_`Kx{X7+Ps3Ku35L}1evjLctA3Z`jP;Jb>k6L
zcti`^QiwuF9)}KFsSpW6x!z2M#X|p<+m;%V{yoMJ&HwPYVP))QfFNQ|`6(<#PJo@1
zsxCGocR+$pgT!H9$KWs;BgCi>(XY1v9uSqj8dPVmz^NxQh7Dr!xb7$KY(;OL8f7@-
zog1XK;$QNKwTY>_A7Rccdr;+_zgv2M)LQsvM)(&e0d=*<!VmYA$I%waioXQn#%Rbr
z@rK>iM4a@wd(vBSA6iXUw~PWF^v}tsM?87({g24_)!w13^nq>al%{Tt@?a;%NtFyU
zI!;*QO*HgsBY7dIPb$n8cBisvyJM)0hiXp@$4|6S6Hb08#z?fruQ>=%&%@@$ILs)o
zL;&}6i=cPH6w255dbW#KD4#xKtg;p#53l8{t=gU8tv&GmyybdQ@Nrl4%^b*u@}BbD
zfZ@Y#s@~}*dt*QMI}~iaTd|*7DNWsp1CZ*!@BaYsnpM(aj#|HcrpEy~k>2x{$@7c9
zU%ImaPM_r%NVpFnfkql}1Cx~@6hJX|<|eoSx{Ki7WA}opu*ZNi(#13rjphTrJvG!k
zT5FaYfNBLC0XBd!6EaetFrf8XHVU=>nH@3_<v$|%r{U9q1;5F=PSo#=n=1{d`Z-$U
z;^Dh&75)n^8-|h`;vxBbdzl!2z0Tah;65UoGB``)nx;~@ZM|h$1t@^vx5Tz&g!)M_
z6HwkwDKKp6$pSaPU4sVNIN3Jit8LjR{|;XEot$;F8UaB=hWB|=^RG_6uKnx4Q}Fu;
zIxWlF4q>+#$>@T1;}~kpYlwCQf8B$p(0$Cg4#?Z!yB&C3;r)-tLH?8ec!DUZgF%?6
zR(jr;@pd)(3-_$!`0TAFnF@N#^C&Mnp`Bf+j5Z)P9{yEs@M@6XV4qz&KBFg`tszX;
z1QB1!+OcWx(NXyIO)o`Bf6UC70z)wkhL+dlk8~ICe3sX1hD?0l)+GF!<muyL%AF@@
zq~h_u`SmIP;q6_jv3+10|NV;xHxuB1N&F^~O#=caPt6t-%tuZpV*x19prjf)&F$|W
zYFG4O9Q)HhQW!Y*Gby+l6D!Wa)jLj^7N=Iq^%jTxvG80Dww7=a=^6&gtAy-&Ym_4d
z6L)QeI|@kEBh0}3{zcfL4Xgf_hf5Y1F!-8SXgl_yEye(PoK90{ARHw^TRq<0H5)wD
zq}TGaA+8VXLeS9-TWa&^6Z8mVfJ%Zqq+Z0gCja>NlJr+6Gy_W95*1*$lwQ(R+63p;
z*l4pvnvXM(UqC7c{im&j^OT?(MgQv1hyQzDmJO-RP{xgLmF`t}IH`i17@KFQ9ah;b
zscf$}o3-s0N`NQI1V=)X#(^@)VQKs3uu9`O<yJ%1w8VBmfYXb+U-{M7cn{=Kc__iZ
zn49^>Kf=h7sLhjxC-ou+>&pnY5$Sg4Vmo7u>XcI{SV)jBQmqFXHE4iEr=YuD4>WxQ
z>2ALpR3e?Q^2=jjk@tO|c?I9A@ZVvsm-P*@-JIuuNNuwR>pbWX|A|QHlU=!ywP~&b
z-7Ef;N&bxy+Y-2m-{x&Zd?BHAC`c1CO?^6Jg^NL16)M6W;~Ob3Qj4og19t`(v;8%H
zUXpixW%Hkv-@Ffu*)R9eIj>Jed&Yb#SCv!`2T*H&xX<PAP1~sI%~7P;j~R$sGpL=|
z4p@)aJgV3I!gDT8121_J4VDR12{D8plY5Hb`@r$~{eg5H@397vR<WzX>Ck~qV~<M>
zkV{z7rVnidsUW0+$0Fay>r&}cUG_lH=a*D6Z*vDDhB^y1KmA{SJT!OV-HE`cAfR!Z
z)5<HSvodhw$>7;Yl1`{Q%WYzlIw>mVrh7l%8o#^$qX0)-US3{zo%?p*`)1|mTbY^y
z=r%y5pZkLmPQe!VIL3B#4Z2D#bN(kGyUOU!l`jcd%>Jj2_?HZ`mG657wFMZ9m;=-p
z>(2ZCasGe;1JwgySV9|bcg@$Oyq!`0&v%$otiax6?$m?d>w+BgdSZR)4+1@vgW3u>
z;@SH=W8NLGI~Mh3=l?(rx^&QNpJL_{EU6*OrE%SQx8!MD0T`1u;jabNx*wQer?1z?
zE0G3*I)d)i5RLsHhsMM{9iV8^wee?h!sc^H@XctQwKO^#=D1QNbDJgX2b-w~Q!jXE
zeP_PNm0NzLgRip4hM9=$1T=K3{OSnQ$`8iSOi{F{(GHD;F1epz)q$)~RRq8wAcJ^F
z!J4iP_}l2sfW_3frhkZODxR#^Nqf_u-}pGy*g_7E*C(S1AwhW~+sN`~`F$zrT|ND7
zflAkZL_N|!{MiS}&$KN_zxz%hht^X?OmDEe>-@{Z#V^rdIvA9Yo`aVhiSuRd)^;TO
z*6pRT)^i?E$<IXuq29Y`YhS8^9?2~2<4boV7Wm{>bB?X#R`lv($@~GmVQ8`#ztUq0
z_#s{^E^bPY&k0zFo*wc8_4|&ZAa-)zt+ST?r^g+#T&`AMqy=o_y{8D#k%wTA$yaQa
zBP-x}N0UXGxKy6}Mx&!dt&dCxpRHZgYR9Q(`CmkSN(6)A=Thb&z4}5^J$D!nR=yvE
zmZzmp!7F-QK1IXZX(&uU!Wu`M7B+V8XNL3-+5Kbu6XaWRTr{bTss11bT+k^#e=D+M
zlCjNCnM13U@Gc!5;6250>kNrn`Vk-v#c}})TxeioiGdl^9CTL)*7{~Ro3YLe{W^SY
z$*$)HAR|-0<*D*esNhALe`IGs-^V<hWfm|t?BKnd3LVS_D`(o<tQbMsC|guzsY-xR
zSGK+Ka_9qRP6LWkcp=5}3UQFwtv9OSZ%77~9c``BsHmJ{!6qFbps#rx>NMVl@4g6}
zrJuA4CH;+l$>i2O@bMOMQMh6+=LV?`<*{L+kNv>pWiz-NeaWe^Nvehf8|CV%lpy&;
z$$&zKE#;(Rf&xt1=>~$EF^a7OENHNCzsa&~9SpvGV1KpL4tR!Vm<l{*b_kUhn2I@j
zPcT@0MFRO<)h5}4T8+eLZnVu)(_mN1hp^M(GvLSFU9H8f3o|O*^8(}eENW6_(?Xsz
zuz)hmlyzXy{U5e@sN_m@V>yiC%PL#Zmw5AQk@{9*E^nT7$H|dhLy`8~IA+a&)V$8`
z^WTsSvrS4Ksqd6DTfhc-be`}4sr6J7xI8&wSLz7aB|+MH)AC58y+d#(1^%HcI~mPF
z32tIvN-?!LwNT5Yoa-~x(1es9|Faq;+=(BAGTq`X@P~@}gd_3TC9W_3$JB!*aQ}sj
z59Kl?@aY`Z7X*Gfls=jB%B7xVwbo?Bb!^%G&~+K-bdeGGm{EziN$%VY0lobJ8S$v-
zBQwx+Kg4wZK5gEd1Z;yH{@b{%@vNrRZVvfJlj%73Ki^#d^rNNEUiri4uN&tMEe=3M
z`8OYRyxrEqZ@r!eivH82PCaJ*YjB!{x-g|%bRIVUPA(kZJ}eE34E>W)bzP;^R?P*p
zUJem^?-FnRwl}&avTIz~<r~GEglB+-&&ISgRz((m>i<-jjE-%IzjXRD?2y|I1Nts*
zaBrn`n9SC4o-zTvMDwjJi}s3oVyRIQVt&{pXo!}1j0emR=~%;yr+eu}+U%<n%UlPU
z9{Jx;*09%Oe^lpQCJ(3er|-LNW%iE-aZFcvRZ&mgu(0@1Zi1myz^I%KszU|5Gwr7q
zD=c2c24$-ZaIqQUc<qF2UPlssn`^_DlNruyl=&_v2}^=^yFvE5e~Y8nbYNM_4Pv?k
zSsI72hx7nZ<MWa=0<WF->VGMJT<%z3s5Y~_(DHGFN}ergXRB$ZetgTa#}`$&j{buE
zl>8g};SKw6agF&gHH6WvBv)G<rIu?Tf7ol8(E01oC}E6nDI?TtAN%T9R{4kH*ft=G
ztjoBr8ms<?hu%BSWBYsk{v}V>-6nzWWe)E;E9{BWkd8^=`Hh7Np&ibO;dxf{lp?>u
zAvIc!ELNzaB8xpk?6aW*Y}6n&mgUm4sw3F=qQ_O9cL6YvOcjFE+q#%DJY!+^zbYVR
z&B6w^cQ*<vb*naTPP|zu(zJNh!{?_0e8#E65U}G$n2Rg>(9(kT92bSNI)MTVWP?xs
zLgdr`cC)MiyI!I?*U|KMT!|W5B(g-QaB%~5m<$`2U-gY<jokU_EJtJFntmHpFjGO<
zDB8JyxPii*X5KdyI$nwI^+AtDy(4cWpe^ArF;hM-&TSoDx*J#xOVBjF#2oive|5<B
zUra);lg(c<*M#AaJG!pb(4N{Kb3K`m{ppnUNkJi+y0iI|SL%)=Ut-DHBU}X%va>%M
zyK`;q9ni{Hn{|D0Zb$*D@zLR#_cGvzDy~64_Kn7*4KFB6+fL_*sjc<kGuOt$vN_9X
zLMI%@d-Yt(vPpoi%Gw|@<MYFF<;J+HI(ypVI02clz)tuakoUnQ!8C`n;tiEcXsU0n
zhVUWF49upZETo+rqHIO~81l@dvM<-kLQMh1b8WHrV6@j_j@P6<xIGw3RUPZZQcPl`
z{0BD~`^wG_9h#yXl^Kq3)t&-UC0EBr7yXR){Bk{#+JoP~B}H>%J7n!k2w0PV1m)Tr
z*(I(;RK(r4j}5AfxV-nPWnm|hhb2cOh?LY@S_}}x$l>N1{|o-C7Yx_+E0rBD;$g#f
z=zOxIdYxEV?BnZSsq9{a!~TDF(y{CDz&(%&3BpitdvHAe506<3cs{Ls5B#Sd5(7QQ
zdv<*9^u52jYwq^%KW)^D>waDBw!A-E&mrA)ul4=d?s2*k5%>_HtIOZ5dxt<JMA{-$
zl+Lm>2YTbK^;xLY)@8BY0s4^be;)@YU*c{cp=Ih{>W0t|70I?Zd6;?hgr>w(*j#%W
zAIo~Rz(qRr2u}(PfQiy*0h}R66b;KVus#y?+<wp!z9v;diKGjp@RQ6cA+TM^Q~c3L
zfj}LM5t{)67#-ZwTl%s{%8M0oH!hh6a_OaUWzs+E7;~=v4F;@XhmslkKijso(%jLQ
zgH`xbZ=RKRU8(pL>`(1{vsb$U&gO6va!6maYCzrB?R=}&=4OvMk~rh{)W6nq9c$#$
zq!GXZcX^G82Q)400)rlvQ%)>LB4PfWwQgPqc&TBh)yT*=3x$Dwfs^+6b+JcF_$S0M
z>~V<{BT=8DS!|Ebh+}e1Wtz7xqPdQ*sP{(yY&lkbIH?M&Jkd>QP;=9Vl+4kQ_1$D7
zE{rGv-r~Jl*c=nU%+X`5+7f0)-m6mvv7SoVJF&`+QvXb|lo_l^NJQ<y)wq5CrbB1#
zU67Pd&Z|7tPQaQ8tISnXSEJRE(!P1~W+hpQT0QS<LnCE7)~I1=_P*n%A}<jh$i|g3
ziM+PcP5Lam{jFDl*s>1t5yBGm>pIgk-r0pc+DW=4cY2H)`FpIm?$DS>77$Lfa#`7>
zGSWORE=Ri9>^Gj9n2&myR#J!OV9<B8S9j~zMud_T%-V|aj0G6R8C2u7u(|^}>D(hA
zqPKBmYT8Yz)wLk&by&1GDYu=^|F7oS#+&dPqYvw%I8}O(xr|C0wOsl|UqnDo&iiG;
zWvs#)P7fA_(Y2qQ(}Us9bD`rUkKe_U4VDaQB=#lDbkV=>>BB-K#poc+-`pTdRL&oq
z`a{g==73$Vgtn{q$3IZyfk$NYU2YKXL+ju&p@-*}xydh-b{cyt$@X758W$-;i&fJ-
z^rlBqUk$L~lPk`Dqv9z5FkDt?o(*hzxgIz7Db~E364lk?*iEC*pRGl+iA5zRGxhe{
z&)w`oEy=DH^cnMGI4QUw2I}wPU`~rhAMt5rT#)E|uiirjB-X7ek85?r+Xgsn=f#IU
zS%CYz$B20w7H8+s=W!iR<znZ<S$7e}wsLU}?#j~!+po5Ghv5!qg>N^3Qm?_JGSAUW
z*%aVN0{XIE)SCVZ950e{JR4-_?`(bo<waOX(8fiLQr54?PhZ`FpA#;AUshQQ#{kTX
z#LcJb%^g6_ytmc+y%KT5t<b2}ZPlUK1!CxV8G_#tlqWdV6_JZ9j_{9l=5k!!eAV9l
z-0t@@2l_W%l!@Gycl_0={U-{dYtn@6K5bgf%*olUX+PW9C41-lx9RLDl!%BdiYzZ@
z03<~|j-9vK`2b<$dLH<0$Piy20zOQJ(cM1g|1SYvgG`IBrzLFu3@U+7bC}{6jkSSZ
z*aZK_1@Mm(eZLkRCXmQKd^t*Zqwlf7ynzj!du6sMo8(h8VTcgfrGd!16&BAdWdNe{
zB9J{$Dd>>HSJdrc`T<*EHleix!5Dy&?@WuVj`0D?zoaf-l3+}SrkFjvuP4o%qMS3;
zfAE2*X;qV^C&3SIR3rwOznqaCUO8M$wzo#_x8$<5FH4`^(LwZj+(lJ|Kubg#FSiGF
z7r9pU?|Zpu)79r~(e|Se!~d%Gmv*qw@wM>6Pwl~MVhC;hcfZ7HmPL)s>|k~)={R9o
zw^{3v=VWpemFsFf+19<n+s^XdQSA$!9PoSw1s)UVD2p)+Y$8gyL>SbN4U-0JiS_!D
zf%W6KCiD=bNxd|00=)_GqyX+v_9CC?`QNLARS=hAcu|sK^0?R+)X!q=rL)nK2~K0g
z2m#M7pQLMKR=|$m=g}H7U<|xHy|!gpF^m9VRy8cZVgd_sI94nDF~2m;mJ+d*6D3*$
z4m-g?iw=fsb*wz}%+ClVc2HzOy{9nk>s+fR7sO#Y*53|4Nm!_Fu$R7zaNAX)dwOrQ
zaV-N7l);7JXhN68Y()CW2X75o9EZ&(E48w=Cpbw$V4xXjzV{_EWk$l#%0vC?T>Ddm
zzKsav`o&3!h{e?EONm0Niff+De$_;)Z@CPvwWfgV!4gw$T5sgYkBtPx<Y7|e3{knO
zWM25btH$e*<O%`rhkP#zyd1y=&GaRDp}SdQ5CCihxloY$lhQ($1H^1Ir?H{&o}x>F
zxg^s203-w3J4Ek<0&|`UooBu!Z<D-q^!)V~%Y@Q@eK=o1uh5%UM#{oBM+>>RVig5=
zh^ax2RwKwt1U`<6m1+M4pFhU0+oLS%)UVa~(-vpgAGPajfrj&YK`O5Z@2-~#%GT+;
zJv)t}isz#luF;adRe78cmW$BCMTxLN93)`L8JS5BPd%!NR7mnbVqfOBKM*8;gzc?-
zU@`0*m&AB_$CkSp!J+(oH-BGj+_L9>g8YU0=s;OQho6Dpve2yT!EsJ!D2D%$P4x3z
z%#zlZaT-E&Pfm|i8u#<qma)F1G*}AQ+a+sHDoMg&&Zb2y`CyXO+^|;6mxHiV(LT=z
z7<jacobJaQ{F{7krN^~H5%?ZJxOXP=vaI6&SL1ODUi3G-=tHlJUo1O{_lSZ~pMud<
zGVl`si8eUfS{ji;0a<GQ7LQ*5jw+-(AKh;YMzeCk{TcNG%)cSvvJl-LAqdiOz|(Og
z#b);P>j}rlbA)3x8wF7aLBVv*>)*YCf;`4ps8I3?*CB<;TdzQ_07+W3c>iXXol^Gn
zc57chYYqVU7oTd(O~?=m#n2&k?ns##<q;f){j!+r0W2ed{5rWLK9~fGVvG~g7=~F+
z-NL*LCsPJ^(pCF-VIkeiBtH06pFIvOG+y^&Huw;$<*f1vPuz9Q6#C&HuRC5**WSaO
z4vZiaexp==e$^AZD$flENg&wIxkH1zq<|=WYV5(cs8ILHcA?pD+P5aZJErn*edp()
z@shtehDwqpzKqCj6oN652M?vkd|p5kU;zah(P_fdwtxwpyII8w{k&MlW+Rl3<0V0w
z%x2<AMF~mYU5~8*^co^v-|j)$r6WlFHPXNH!4s&SX2t;HU&AKz6i-t^$-QN3ENvh7
z7fs_C3t0KLgoou2AL<z;P@fKL$qoNCjIkQol~tG_!fUR;-~Mx|P{xzXSf;YZ$K$D@
zYNH_N))vuDgbA4CxNKoQhotSTL1@*&ZN481_6-jH@0Nz1x1M2C7m~=276Li(qg?~g
z)smPGA?15Za34Lb19EFac~RVN8&hDh*p05hEuoHl-`wgyo3aFB;I9*$T=o>OH(-Sw
zMrkEP#5J@IKgqBiwX`oFDjxx}wggo=y}j4lcQOm7Rhms+K{o;0Uk`$!^4Heyq#cQa
z0lnMWqhBF|aq{}<>VR9O3_5a=ATeezP9@5+cEx4!aZlD{t{T&4t~)OqnzA&;yCvs!
zp@p6A_}txg_Xu?G(RP~7Re9i4ddew68i1A<FU&}DJH5O2*%U6)6l+l@CW1T_nuWUW
zyY`M^`-fH<DK#XaH%QlN(ynjIY49fP{P)|#k%Xi5m%!^D9w(EG!`Ms>OIvFq2NL4W
zJciKFw0Up$P?5bbX@;>7&XX#6D?5r|0S04{55B{cHCfD5!5)kep@he6c$Pilwaz|6
z;0Vs?36y6dimXp04u{MrNTs793l|Ng=N(>nt;Q6puS&I#tp%l^fkONj0q<RbQmnz-
zIhvJ~!!eok5zhWI>4qmkMdbb@<#0?ua(*kzsfPAB_b-Z@tBcfW7S6r)MKfRN#R!`Z
zEEU-{cmT*O6lqulNSBVS`{~2Jd{nU)p@Wv7tyf@HZjyg~4<>tn61hrB_)0Jj{F2B+
zFM54_ec!O?n@hvuOE1|A6p+3u`ly5M4QpUFs1@}Qn|mRBSy`Fxi()EF&DWvO{$kT2
zl$g#Ci%h9b*N$&$AC(iTEJ%*`B)bZ*<dwhS+%D<Ak|IJ09%G_2w~yYXO=Om6&dKBY
z5@V*R&dT2kxFIsYaxElUpH*CKxsLi@Gl1ECMkrTo+b~h_hJ$}w`|?;E5*{Gfa~Qd<
z^u+i})oiOW72-WA`U+dM!Lw~d@xdX#u1}eonq&gQYjS&;DjLnMWdh(w47lPr?M6)~
z3S$GS!Lx;_2;cW-DxW~DMpm2m$^1k0nWuVNxOEhD1Yi$>J>IzO%w!HmUu|BObia$y
z7qF_kCyewuu5KeX?bz%0ie)pQaUUmhK<UDnrv!am2=X^SgmmYX3wZ7?VK=Bxs26c8
zVm&60xPIoj!n6tDqQ*-F(<9TUBxi?6Z)pgalTqYqju$wel-j)*DtWK3p>vyuY>Pj7
z(e>+Oz}AK*`<rw44i_!pEdj-66HtrzxQlh-;pq_ZxfLm>7456{{qGJ+vRsshC3}?8
zH<S=Q(>vIfC^u^oCm0pVwbpqguYJH7Z9~b$^_op&>i1Vb^mj0vwewOrF3xt{QvAlI
zCJ1F>L!C|E)ycQjoA%A-@3F=e&BV5_&-%??*!r*PL+lU9^8~+{S36m!a^C0DH}si5
z(sn*$AC+}RQ=P_|pN9OYxX#Z!=3fr}eH|%2Ty$nd<Pd&oIZ_596bLh`(XXp@5$I(C
ziZcnnMAq$4S>NiCmjjn9n@Q0L7oIQk4RW&fko#N>S9KoV1K@G%@zdPtH*&_36oMb(
zicTmULh$KAqJtJlQ2`p!-x8sO8ri=~)thE>`Ia+(`Blc_hOB@+JiuS{&3#kF05qgP
zvfPgy;$f7?sKB54tqm@#>6@Dp)53!~dBf+Op?M`#uqL-W>Co2X!jdiJxFQdL8;0c<
zVG3rItJD9n%*M{k?dxfh@6n{O0B3M*vMp($Penf;FEB>byd%~;t~^irD$o5KnXxxg
znDCgMA$UBq#yaUHocf&17c#2Ivjz8Me!5~y@cLsH2WSgX(1INqaE`roiAwm4aK8?4
zX_MIaw6s+}?VK@f^xSW}2VWZ>KB1K41fGO+5mO0ysLEevP5Fg`-XB5mTZaK>hZmC;
z?g{LD2$d}-UyK8<=mP7Xo&X)#x%YbUdw<{q`ak^`{FVmjY2F6vhL0QoLC2!zJSI@m
zKG&z><Q`e!NUI{A*d+qq7kdgGyyEdqh8Mtvfm$By>+9iwDjlU*Hz=$YRyh2Ea45Ms
zfiso`q?y>7A`0V2HmOg{7C37i@BlB&{%7?#!<{1=L7A*rz;id%ziFZ4x;{CF1_baT
zVH6OV<iuDIE`$|k^g3(~w-iY_-s1i{v^e-gnNSj3nqtQq%?B|@P(h*vU72Z%__8eQ
zS0DF^U{PbMnAr|r|3r4trSqJEb6`PdEdvVQ-im7LBa7zyUW^3kP^-JaShK^|bmv0n
z(r`-vyAkEBMDAZvt({)o8O*AIjUAli<4{wMr%ViqW@~YB>A6H}fX|lx(@-2MWfEIh
z3kB1tI=j{bWq=gS;i1t#C=(4_Md2%ce)=OzB8R+16P_DwK}=;+YvH0CVgToA0LNC#
zVc@K+=SWPAm~zXY%_lJakBmcAkz{eMGoW6ibH-#xlMof1AkDR^#*B(nt?AO;4O$;u
zBrLm!hc}o-4F===fQ#ZJ(VP68R>-X}l%l-(&9-1k7>`q9y`$w#G6J05&mtoP(6REk
zgBop2)bs_iN<sD>i$4~YsPcTUE|XeW0*55F(-hO7{id=!j8>mG^Vi$f0Z$)e@VDFd
zUbAkKB0oLo@(lYAqg#l*Udw+D?m55r)?|aVDdl+aZ;+eMv)nw}z{lrN8}Y~6tDxnp
zhconA@;mCqwjC>;s|z79Hr|0?6`qSkaTiVof&`)aaqRy9cUiuTkzNAmDeE?(`btdG
zILlJ3`e{$t$Th&{Fr;7{f5<43<QZQRn$Eb9H`2!4&4YK*4?n^sNODW=h&DS-^E=PH
z!uH?B20rZGaRPyIa^7&wT5Bx`Rl>(x1d4(b#uY<ZB#Ba<B(_$&`hm&pWAd-o*?RQ6
z)b0cM7I%M~s59F$VB9=Smoq!_9&z7pY`f)rpBP|g!Mn#v7FB5+Mo8@a_V9I`q}Q|d
zv{_NxI&?oWh>M+L*E`K?jv3WhR-*KHW^N;@{Fv$Oj|?xJGVT0hm<~nT|71!Ulw3Z&
zhetB%zZn5w!dA7DKNDcewT$@G+;v->&~GD6usz#r+5Pl1o6*(q^aRlN4%Y*#-F8;)
z<1dza^u~Zb1MKmk=M9Wt&vIHtekFXj2(5hg(VXFGzXCG|Bcwcf%FGy?i<7|Dt8XJA
z$i>QG&Rj@$RGe*~Yi2w>@hQ;hNG^GeLs?#gAeXCm1DvU-8}I2B-^4Y<L@utxq?UVB
z%#4C#4wm_KqJcz>Wo=gwOY<e1syeljR>3QWc_R$iE(ysO+ME8;(f$w}e}B}l8F2ZR
zm+Ol!Qv0Nako!|lx<QUYa#psT*Tdj@R^|IJ)8#et-0nK#Rl#Hi=-LQ+>Ia^%i_`Fm
z4j#Pnvw#L3!$$>ofK=?Y$<l9IlfwmJQXQ(RB&8r~p7mYKDG#`mIP^1q^md|bI}6)f
zkEj(pl4~*yh4Vj>v#L0bA^Xr`*Ac27zQ65VB}uB@2A5_|XG_Z@;!89WE^XHePW6|&
zpnu5}SE>7f*XR90Q&Bh^Y!!<G;PyZx=FakSh(|g;!@qP)<aww_A^XvT!Q>yj!w;06
z_P>}hFmxkY1PW|C02PlPL)LoZ5z=^m_?*o4^-t#gGknYD=6CD=ps<23J`>6|_9EHP
zMf)E5eyt1zc&;Z7lBsN_>z9DA&)h*qm5=>QZw)}2HTM5AXQsp6zq#-q8D~!y@xEmu
z5WM2m7kGp8(}8tXdP!YYzPYO#5`aI~lKe<vRJUC2lJ=C<XJ4t2CrtAjW^Ck(qI0N9
z(ZhnaM6ru2)+k+<stMiD4bye+XR7n;B&*_gS`8SsXKy}TK4J1bA&V(T$?E?leBa3o
zSM%SALZk#pQkSzbeB}+Vq!8<=!A*eyLzbK+*TuNTFqaH2S+kw*aa&BBB~3A1h@Cy5
zB!%IQm#TF1jZuRCiYAFoS2<Nsv~PBdhil)&Ft<96CNUXscASQ2!ia78{^UZj%mK?R
zC+N!Ie(#%=$6a|a;Ky`Zw}_HIa%sp*vi(&KAB>^O7WR@r_s;k{sMYaxyIW)SSZS=w
zs55DE^X*e;M~MGuK{3-NFy=<pn3YxlDxpw&S~W!BeDJTKI6=F;6~kXWZda7U?SvCL
zob<4Z;-x5Kt<t)5omz;(^1)Mvyl(=_Q3)>V{_k-0XKfr|5FyY?cL4gvP1@n(r`QY0
z-UyK!3ZgiLSRQVGM#7QbLKqekDm3-Q*ePt1ZU2vnBbUryOj6-*oQHM>HsL6Shb#Gb
z7|%5;f?9}~$GE8<SeFC_9a-Q=!x#u~^yklnU;xs_?nitSy3wGp9{k*+10c_)h~$R~
zp#t4PWTAw^5u%?;B5}-l>tBxw6l_a6W571UzEmEN*z&s!!7|B-p!_9iKOZgUOtEg@
zBDT*uJpky(jD=j7ODIvC0Ht>28=*Z*Rz0;;R;%ybF9+u%%)A7ynY+J!t`L2``fKA~
zzKL7``P6NVcn&=&8Ah)7Bq@HXt5YZTA`|n*ugsU=m>djShxRD8^_z44zEAmQh_v|y
zrzhY0oPPpX{+Ml9bYT)lA)D<T&>K(yS3ssp!zQ>yaz-A<caa*dz4-;Dv@2z0^X<b9
z$}SHk<Y(93rn=80ad%9Xa1%oDR@)yWaisVv>hXKG*#YuI3wf9jEB>u$tKR@lLJpzB
zOAQut%w&K+-ZLUa2P8ITwV`Zq09og|_%5QS=!PiXuV17_$=#0%sQ+YiW4P_ZW=~<g
z_#w)rvc+>Ko9mH)$u`uB?)9=?jX&4a)Z3rj;n}|3_v%i<L}ic$LMSZyg)l)3bxKK)
z>ro-MW`uOQd4@B*cn!<-Yc9}T=Yt81+n(<Vrh%-4WY4x8E{I#Hr8$`Ec=p*?HgL+V
zT<FE~P^b$_q<h(mOV`w<%!fY9N>?3Uha9z%6^-;SD@gnVQQR1yd8%Iiw}~yA`@u<Y
z4nVNih|G+uJW_P+b&qgkhd07Gv}(IRiyB5#HxQBf^73Vd(nCW?M~cG)M`<Od#zLai
z$r(7UC#yg^BvF1(BW|trWItQ%%&hpcygV2drtK7KVJu2KI##T3<n}_N%3M7nzd$c$
zX)`t}iUO^8E{W$u6uDab&9;Kra)v~CF^Mf)iR=!tm@$LA+7T8+;<{n_`gfy(SGssY
zG~3MGNrc-Th}CiGo*K;7>e|sTnKNJ*F+q0XwjNRwGW9hO7ap)9g4~>yWKATEWCrfo
zb*0CfgTGF#y(p-+-9|Sd7+%~i$L8FB{c6D!O1!Jby&JTC%Zbk1O!yOsy&ogRLjO+7
zkZYjHpLhehrDYVo@#r%*9;zadki-oUR*{K6U;ZXSCFDY0viHx57e8Pd-)niy=*|n>
z8j!MF0DjOJEIVz-HM@X2!G|rt;Y3kMDbMX)`?$g50F`&`;%B@==&zX(oOJ^(?W@HG
zHZGLW^>^Vv4bCazb5p2s58v3y!>bh*SFW2XYX%M24uOfNDMbTSxU4;#Tq9hOpXxh+
z6Dg*{__Pv!KqXa2pj5$tPrfTKxz~|&uE=A$(T<yu@7l{A;aoRJ(UmJV*gq;NrN`00
zeXrVMZCR{8ZA|QEA_P4B92T~iZ6h_N7hj2EghmUW$JcbaTAmkPDxU)##tyd?t9}cB
zsycGNZ;PB&ZwyFw4h@i(k@_&BHM;W>&wb`UozZ*v7eP9cY-#O&c%JJvoVh8qr<ui5
zE{`QEPQ(3QJpr`>`R69xpq!yk1mA&I*j$0r3yQOEJyQzi?9;S~G}kW-2bB7CadGjW
zPz!GIk&<%8!*SjrzcQ{HGP*0Ru5lm&aJ)j@QhU&d$-X)o-{m}x81p`0Q{;K|d9q_1
zVf(#$SjOvIAu#@GZIj*YkN?k(v6#j8Ol5bklF>5K7upXq&@7#Ca@I>m0hq@V&r83#
zIi}@s@>UbKSmJgya`>%%_zRP(n*0j#-QvGuk)*3gQ=R88iw2#L<E(J7vZ#@JrVNO)
z`)u&C2VA(%PjyBVKNB{E)7(<syIfrxMaepn^K?#3$KzhR56XRX>%ets`clWgXmgnJ
zi8=B!2GY;r=qD}eGhCIcN4*|(Dxn6$ft0XWzR5e}@USmkh7`*3_N_vU2-M8HeK~C(
zR9vnz7kM>=3^AE{{oNRnundT_Z^k&(5z>h<fv!5%e|mv<b#u7;xGO9pjcmzHFf7u`
z{Mh%=<bpiLBZe@Xplm6bOHz{rKiyYgIIA{njVr^X3jMcIzu@X7ILH1&Yf0WikLtyp
zSXfgecx$$RIu2)2cM@G*{!~XE#i4WXAmfGo_5FoScln`GxY;l=w3+#pNd$XSg}R47
zW;)yZ*GSmWk<v$N=N-T`hux}~oeCAVT4!dsF3&em$NpnAHMw?j+;~L&0;P(_kUV<@
zGx3jPy<#0ZNnhoww4uu*<ab_muHwg}%?~XB-;5F7c7o3j`bKI^%1NiEn3{%?v#4)}
znR{bGKO#3Krb$bgnluvrkPP)Gz~I}a18HYdtcC#zEG}pVf1&auO)zthadApY&!jsn
zKJA@h|DxZHnvzAdE4+W}6k@iei>AWf9+T%X(;*V^V=!lpI$Go*rkc3ZNeNHx)2q4+
zZC0&YinIw0Q7-yfC8obLW?Idju4bxJ711tqml9>?WnlMr`<6PumgMjW7-@yrL0GWK
zGQ0Sf47<~Oi*0Z`KR6aEJ(!M<V2z?yBdM*QOV9yq`6&@ivk~yZ58=j$WfY@J*ZdIr
z4T0%T@$>{sQm7ue?Yy1y<`{bOC4(dK%e=}fj_9q1=xs~WKaA$TL$t?|Y_sSxL`PBS
zZN#BUquYwS4@dv}j{XrRytV)~B_zNBVqf-fQp6pvE@(A3%$T71n3T)u`9+wvyx~VZ
z9<3PhiqBLvbt<nEv&G7@oDoTCRTE`$WktRoz4vyKfb}yWx|8OO>nS2@2S}@mdcBYR
zM1K}ISb(!N-wK+p?HOxGhe|Caip0uk>O{@Jf4|^=N)}^qQ<7D2Gf#9CN(J8pqrLgb
z${!Qpg?AhH@mwEnjEqU2^WQwS5aA8F_y(hwf>d(3e*!y(h);MWu-5Q2o{c|_{U#gS
zs}dN~|59B0dPQxV{S<jUq&|Q0&VT;>oH!VSeF+W9_vb+0e2>eu47TpRuXT*xLbZ+u
z2i!xhjx4y*p}=m;i^96iD?TAuzw%8CNKy?*ajApN3g4DIRTTSX;~UK;?~qv0nXKk|
zBGT5WV{b(o?Lcxm_zMi6TU^*iqT?KMNoZEIJCa(GFX^Y_5%EDz@>a83pKFh%@cv?p
zt_GMQu)d*!?nMK`9!`AK9zW6MS^K?Mz6^r9r}R$wbS?i_64Ahhfi2Q3y8wb9E_lt)
zPMCtp{Bh4}vNb@?l9|i)w>V=RpLE!z$mog+If?DM?sl{%QR(djj*4}0;hrusk6ocL
ziv1uIGLIAyy<U_VtyZj5U$bXGCIlz24`T72c9@Eetp=vGew;fRd{KVY^7Fn^4rqB#
z;a>Badd-v(XJze<y~Ow9ee83i2khWbz-?~&#p3!%=C(9nHk-F~1Zh!mZ*|yNnWg0f
z?8y$UN#FSKSYP@3eBRfL+D2s#jK#_f%-t{8M{S$>q&7DF7`hkVh**Eat;jojx@SAJ
zyo?Fez8i@j(jmnnpl$aveU|?VdPkj4>MH%@IK0`2C$F7@8OEa8vr=GY;C)q(#|X=w
zL;oSHe=y`WsnnqZ^s-}7akSEg<EO-;+4F?o9rmlzSpNBYci-&dJNeI{J<}YO@s3FJ
z_1CzsP3Lq2`uMuV&Bp9q2W)XaFcqjnu=5DaneoGM*E_^15n+DDHrv`uMVIJ}51NkD
zamF%A@Z6;MisN@|&1~)z#yDAui1PA(mU7v#L@~a~OO-J~=<#FOQmi#$nydOG+$)&#
zfy7>_Do&pD4(_@fU162LUOJ6DIaw0%Z=ppY>2xe~u3N4R9WyW0<gHu!-*$XX=`>S^
zVqH7wD(-$m<g-)if;?GePt?4<Cz9k-#W^;VMe#7ITm%bwhmnTDx(z}8D8T5P29E$|
z$d^$tvm9=|??3a884&>-&(Z{G;j(gg#71JBpI8yo_Bd>twsfyK`X1#*O>OTQGc(Pi
z8A`nbwkNBN(4w54N}YCO0tT1}Hg8IKFFOu}6TWmmE%=Juub&6-ZeD)e1RpnQ_<2wO
z^_G_Qxhn{vq!^R&Q<#;3o?X&_9w^k0$6=TFzt)ty03G0hHE~WDu&ha)839}nI%{>2
zR^4f)@nZTb^95pWIn&NQR<sVo{u$)~!&9$K4m<NaHa`>_^LwoxBe9>bX_tj6&xM$B
zM5a9@YP~-jz9Acbjbrp08@<1-CKet0^6mXU?+KxmfI@&WZ{6&@f286Cd>(%6zlUw=
zJ>iLRmQt5r1@v0`z;Vfa=De&RPD^(z&8m&4YAP~BI$r%aX%p_`kG%|fr~L2B&jx<~
z?A94uBj@evmkjv8Acfm@e$I!7!sb=Jx36dUas;W_qfc(6aL~fGHp#?(@cBpt$4jG(
zK=#u7_(pLNM%^?5U3cL{I>LrCXSA+jN8|nXMiE>A?W?w-2?0I`TdNTB(ScRORvhlP
z*~EPxJ}|16-PHA^cBs!`Ldx6nd+V`_ZeGbNZ}&*0$kVF9Q(vIeW1kacCG!wJ5!kGC
zJ8Hee$4g?fhmIIO0v8`1ZgTrI*%M6`1413gw`EXEjRMPM>m}o>vvKyn_j;Y;0(m#)
zHgQ%DuSSnOeKO(LehYn<?Ev_QAw;+mDfKY;r-M%4{u?{k8r0cmj5qS5JM-Ke4C}tN
z3ZHsy6EwPXw}c)s(%?mcLcR@J&c;U@IDe;!?>+<=d+ind9v;DmR`@Nd-3fuOFPn$S
zN*t`>`u*B9E-ROVyUo{z838E!<2C*y(QAV<-V?-X$evuh6`7T`Wch11DP*^*Gjh;)
zjAJdcd(x^gW>(3G%2;)42n0BI#pWfXZV*L9ELj1$x>^kA7j8qA#Mp-2wZ5^3rR#hZ
zB}7+44$hEjuoyXlO!hKt-{+!gG5v@pG@6&2S=aURN<2p8Nm^zes>yqWCN~L7=3CY2
zXk4ZVgya8Av*fU<0Clz6AfE9nWOGpc{zp6dG|`UChSW>lBMRC7aRH9n|8&17#gabZ
zK{FtfkXNFt8~mv57w8W1e>ae4YaZu;#>=6=hsRwXul!}5oGATM7WyZ1Rg<~Vw}#;4
zkQkWW71`aNv?4<rW{xG8^(!4B2JGq~`_kh-Z{*kats3mNJB$ff6XY@C&z9=9U1M*N
zc7i{BP4N5dE_@ZxpZ3Y-ShS2u<MV_ndw$1ozlkc*P}+8Im_XWPlb7!P^dgq591)TL
z_Mm4eKp0U$++Zlp+#w7u6Yll>JJ0Yxsy65$3iRLBKl9UCfc5au<n;XMto4vI6yqPc
z;^2RS!RqQp({-EE`<>Ha*YlZg1==%VjNZ?mZ~-N^cJ=OEc$`%z84*Oz7HK88@YKP5
z{KE$Ws~(nE(#~82>EWiPcmC^j4Wa6;_lhAN0sH<2dh+CmFe^~wG*EGTK<5NF)||9H
zb>!JD4&!D8Hyl3f4MjF4mF9(xmv$XFPjb-5HoIneeM41HqC-rCkG15blHyYH^QtZ~
z-jh_P49h8=mwvff6WzThs@+GHIX}>J|FFBowQbl@`>1}#NVc!_+I(4b^7`z!302#A
z14?xH{yzUGP%qDAjI8rh%ElT<OffRW5Kn{+G0$5{u{#x1=;eK46FnJO_wQJS*0@L`
zdxUy>nBSf|99wi5UBFjON!Gj3nW<vQ{c|u5*<GE@r+}n)xE2=*P5V%wQm0*hLT(O`
zmCH^NSDqa^hG2wv?5KHqaQ{D=&N8a1_v_k9NOv~~NE}e&P*MUChc0R9PU-GWk(ADZ
z@S~eUcXxL;(%tX&|BQDGzxc%g?!E7Ot+}o>rvzE~JkRKH5aws3pVoWn#{|}UsSDtI
zipRES(9pdPMC}0)m+I+yS}A`g>$!x%HTZV5iLg^sZS~sJvKXW~=1wu|4%bXM*k?18
zU7DBt7}EcFt)kYjcF5?IZE2nyAWU|)hgWGVf8he&W_6q+rUgaE?4x`=c!3t)oIRXx
zf8hp6voX5raiX#KBW%2}a<{#=&(s!_J#^qU3AH16y}=k*DW*x^>F00Dl5u2^QfvE*
z;zN}6eOgqU)C1S`Q>8y{0Id4d_(afd)!n+BfB0}8?<r5AHJqp91wjIzu}JED58B|*
zvz6o*te~dg$5w0>uNf>w-<>x7bd%vZxW;RMs2{zO2EyrUrK*-A_6b5pe%&g4w*dS<
z-}%w;gv{{j*Ah6sdP#fri(jOG0ri)1m8Aj9yf6xefQ;*YaU7A`E8178PYj_)E3&f)
zY=DBXE*i%_rO3b&rr~6$PDBxCr|fRa4*oJ7mLg-`vPqs)RK}wB_3`+^itM&86)A-R
zgni#>>nU7nIovYyW?zp|b=3TfTf4PHJ?32j^_O;;hjx5&`IBQ@R2ByB_M=t8wpp&H
zu!?u@4+Vvy<&fj9ReCx?eqFrZwNg*^v<8y&7Z9dqA>!j~9w@rMjfX~qOpVv;r;;Xd
z0)fAuu0dELQ3cz%JSAb0!Ol!$N+noyz>;%!T4%Eh)BZ|E#!fknoG|g63&M{(45hiO
z^|QBNi@#Ng#P`RK0?F`Hht~Q0@2YK>vVW{ZRe<yP+vgK?eSMvA-u*wmd5hxlPq)v=
zoEOIb#;OaSo0-ZwW&hi!i<oMOlQ4FbTGgV{OetZ^I!=y4mBU2^DACPvh9*iYh0gGs
zwCh!*#G65&lPkH-+nk_-66ouYR1H=v##jVJxNJ(+)Yu0N61j0!8f&Bo-X8mLt^}Y3
z#ziY3Do;%ErD^u;Hsn?%d77yJPm?Hg#&{Z{Vj0>sF@n*zghTW|KO%J;Rq-8zraPaP
zNVWHul+^oRTpy=KO2$!`(Ua-fbvGX3RpR@atg^F*XkXuTzQ#^hpM?dz_2ZFGP!7p(
zxPhXDFhxistYCbVfiEMf=M);!nqlzKJN>-Lr}uU2ZE$w``S|_Xo_B`&S?<YOizztr
zfLosvG7a{)hi`vq=aOW);_f~A9cTrd94{7vx_%N<CaHseik;UihwT%jUbaLesx{=H
zAMCEG+ZOllTMa1FZYu8+?<LKcc8dxcb4~#T`W0GNxO}tNQvlYEdGFi32fiHqnt1K%
zUvBU9ACpO5Cc3d6-|aEK`o0o4W?fo-U3L?_X-Tvgm4l<}gPuq0QJafL06)U{&4%bj
zZ#>SFhf_<$gC!VmV7Jax!<=%eL*d+5Uu#YF+)0vE6--OToOslCC|dFcX>5*aOKwSc
zWaZ&%Yc0i*<>?>|qW3h|*H-!m8)#J{=+d?#bc&A-**93C>tDxn>YT^cX}u=)74G*Z
zK`4ppRz8=#W7JrXlmB{}I{al*5f~`ndk+Zfzbt&BAL3|V`jDP_-M#f7HN^@(qAlP2
zOZRq6tB}Uul+^E%c@auySVrj3jJKxxjL+2ZOKtd1ox?{=@UM&2SLHiMwxZOgA~w=n
zO_PT4Pbh$MEKTye0GL+4^wgXU(BdD`banm!aPLWESEk`;M;&t2uc$$>(+u~-Afi>s
z<x>`wTe_h#ZO!VrDR+!7%Ad!3<3*W?F#2Dy+ly5~%qY;WC<87Azvb|;Sf&S}>E~h$
zat4^YilFaqN1^;^J?bpAWrhBU>miZ-qV3@^S@P)=i0IORm6-I1l4LYgjR}<*CN-iE
zP85h}IPOb2)<uNR!N7*j-9S%YYkU9+{av>NgcBVZ6Sm93ppvd`Irb&udlH?xFQ#<3
z#n}<QitKq*2WP<gH|g`@wufS`qk12Z9T2ACHs*PAL+kyILi%TncX5%dAUB`I{8iea
zCSFmaFW6t<#-=B;utYM#BO6F-o-vJvN=UKBvb=j+ZP9Oew2LZuhWNdl4LI=8LGUpU
z>C)8Lk++K;we)UlMR}!e3Kx^>{U8}<IgM^0O*F-J)U-g;@@vgxm_(k0j73FEXg!B0
zy#^UGhf#km6c4GJOD|oLOHtIpKoM7}!S!QHOcY5{h={>&8F;l+oR;PI$==W8aP;Jn
z*KhW^M;UeT5fZd6VN0h5y-qM3gzS>to~74ltTA8C()oBZe|&|=%0GX-+w8?Qk5LID
z9G?y+PU-4MvX<8PJC<_q@*ZF4iOlq5TQjD@==>ARt#FX3Mkk_dwX-IrAn)n>iv!bm
zy?eA-8?Fyw8H_P?h9$u=fpJnOy@^|sh1(Yuxs*uDTx*+BppN~xu`9>&T%H)8oCRX_
z67Ib3+b0AFxJjfqM|lyjbK8JhX-5YQ^P@jVQO?~tw+;Ly?|EO>PXj+<UhD9$7ToRo
z(z|!830_qmhelWKaxRe&A*F(`!A4Y=nxuU_29P#z!s3bL%S~XV{&6&!NO=2aPZi5_
zCR>cD2h)f@{6T(+YCs^A4tY`uSTyd7QY7Lh(AwBfyBBO!$1MyBT+3$zi9KQY=lwv1
zd2Zn)KruJxr^!Na8il?s)u0K&f{ZA<at*a`hOVv-g`VoWy{cy?wA%=E4oqyY{%7o+
znKJjff${S;6P9q_Sdey<(MTJOe>@Lj4~5oCHz+%<TCOa_JX<A`(sVU%TG!_jJU+0X
z(BBTzg@s>SinbX;NjS7GTU&wpo3_*~ltsrPWP9;FLONMOw=_0WuuGD?MG6F8k{?BL
zN1PxBH=pV?P>ibhP*4APqs^FDFw}Do_j5SiTnw%xqXCShn*G1#nMB-#UWYon3c#cv
zi;aIahHnCLqq8hD+<8*lkV<}2M=(BwfsBn@gO~5W%qJ<0EJ}b~7$>Uz-ZNS={W3nz
zKcU;7p%N=l3-AMo5HYx^?S=>R5|8~-zJh;OQYMO<1a`Cf8KT}9;?qLbGfCG*EE>_r
zin_WG$C#cT@fdPG(Y*8SLTG`TmqsC?MjE!%{HVaaB}2Crr&@FV)V4Y31Lo7d8*aZg
zvvQ5Imnr|=MPYU_iS(8YRgI%HF|flCbwtS2z5%@Qgo6~3?*jFr<we8W<J+YpzHdwB
zuo8vR9~q5=kJ={cFamv|SHmpQrx>NlN{vVjn^1Fjt2|VCeJFurGetG~ETBNv7&=oh
z?b}d*BcsDyN{HuJT0|Xd2pla`l}Ngj1v`10|C2}7s{FbX8^0%Pud;W1w~u$l&pvd$
zpEUY-{N|{^mP;?|@l4Lf$l{6d^`xx+Y_qifa<Ei*{p9bKlo@f+2P%uUAYtA6N1d_^
zLl;mU2_m`893<h@2e|PKX$?OpDZu_5`h+E@H|tVbyo2Vz6gX@7W^0Vrf&?i`xV_sA
zI58W*jC@OD2WCksoG3%32~C7r)9=NTMC3?EZ>7`Ea-!hMXIHQ+sM%-P8NtuX+DPcq
zG8R#7WnbT3*6QoMo=j?d#AN2~rPtgclIk3|Hh0geIkCza^Vu&Lm9)PRo8N1VQH&sa
za5`<mumRFOc-KJQ4_;2re))DzpU_c1VuDC385Xca@$Zh-Dg=JRN~7w>8&t}%T;Knn
z{xb}HE4k!D+FtygtJ<agsQ%@aAz`^CU@Gg0jhRoJlTc%is9~zzOh7=lDSeD^*Xr%!
z2S%oO@^98rVC4YkfZ!plwduMO3a{AmhyJv2Af>hiCT`()f5v6XfU^JBo{BQ5NqDQz
zs5Zeu`1+PlhN(qh+qyhI`H}R`_si|_+L&vD9v#+g%}cJ<Gq*_j|H8^rCgVkaSboUC
zVft**UU5d}B!|E<hj57pJ?$bR0m>5x+V}H!({dE$O;>Eup29>l>r+dJuu9$|<b+#h
zTZxD&hhe(6IS4k)(gGX&@j=TxdU+&*=i3EXt~Xg-E=stl=Dh)5rDr7&YuTt(19CZ3
zNB!`?@BKvp*lugm(0+SyBOOY~43wR<V$yC)kak(z3a*<J+EPKwaS;Q?njLfgg7{T;
z|1lL=_8$ED05kj!PedS#0tlh7I}~vxGb+NFl*cE_i%>FQK==eU>d_%xuEFTi<y`^+
ze-uT3jCpnT7Ulx`&osnpsX#HatYNBRw@k%c-uiUqBBA!3#0oe6m;OhT1EP=}(*~Ju
zfGNuJ@;GEV(^wLumYxpbonqx(11e*|A_{5qzEzMllmth*!%%Qj_LCWivxIwL$5&G`
zv~C>xCp~s3vvI+P6gja3<b6|8n(i3V`-_Tvwiu^4FE5ShBDWXmz?mgg_eLrYn1Rt>
z!9Jh^ol)ED(w_Wc_}^1OU2+8K5rQxKEws7%lD>O#f-oe-B~YtSag-($F>o?eA^kM8
z`A6-^!1(rPMTLfK&=qPIEK@yGSMK7ozN3zk`H%yHY32Ey<hhv&Rjgr&i4hCmY8iXl
zyBbTB+Tt4ug}C@nSfD7!{7`csem1^NCT(H_tZAVn?Z#!R$PRRHHofFF;u?Ew@#4@)
z4Ix;Hy@dul3u~<+HO^I3{lflUNZ#>pqvOD}wx)s8Igk3WhaMaQYT%yUk!ZHl-)N}1
zLOTxsb|C1f4QcPysOI~4KnH}tKQz2=+Z-W6KJ>3|;hh3$Qhh4{=fV#}0_S00+>lY*
zm{Ji*V$!6CC=VtbaLz|*RW4R<q;3xM)U>8-9z+-&|6ZiC(^n*>Y&`P>FYiMZ=2C1s
zXtNdy2$2a>+5P>%R~=fCsW{cO{DCZ5wn1Rp0S{JVHkf3PG`3y??9q7y-Th{9$?lxd
z);j^HW`W|nREINN0}f#lFU_&mZ|nH3y=O)6XYakQJ+=FfoKfJA@!s!B3JuQ5B8mG_
zF&!1&73aBYXgaCC7OY0&HC8o`Op+X8=@4bH-g4RE>fMvKgb0+8_h|xDs}Hgj-2a7r
zXl=q}%)kn7b)Y&)41uX`U=!~#MjP>$-)--P;T>NoqD|CCG+%iyT5{$PBG;jHHGrd}
zBP+*|sZ?}s(IV|-KcGJw!m1-w9C1c8R)WL>*ERUxPsNu64^C!YX)Qkd4N>AN8q=DX
z&NcX_9`>%2K+fdt&8Si6B(w5k%alwssnpSBx&~5$bZrYucH+|zdmk#;tSElEIBMCi
zOTwhP7LlU}x(DNhjgIDr$@d<0H6Pmz8$n#HY)PU`@M1Nozl@;->?P=Z{5YK9x7q&%
zH%X4e9l+yp9MANs6V*y1Cp7dT&G_R+q;x-$&e8Vl81!gp*VHB?b|Lc;v8+wQK#P5D
z=C$^H_%fIZyL$V{{03gq4`Y>-D6^6@44+ZzoY6awFk4n?e<0FO<aPc`nu!bB${^;u
z?-xb*b(QtFVB66vRn5ojeq%WDNp%{)<~9Wd6YYp<HvRC9iPN_xD-5klAaLPb>}BNF
zQVt~)_Lkgy30f-JeQt*KK)Neyqi)CF^e*1E?5f46vU@T>s8=PW3m7D2rnTKi+)p>M
zs%mF1L%+oHL}$NWDy-5W#<w;}m2BdxRW!`FbzSNQcTG-4;WO|z1@>Hoi&A-Awy8I4
z5)|Gy*}D0ZNtC+;+2%0y5i%jPbpEIzIQ`53t@N_w+`6DPQv8k5aa2yh8J^2F`jbK8
zETuF~1XzUZf<fZRnLby(R}D}yA^CRmMOp2Q%ysNgUc_w{6*f|Y4p~-GG?06S8moNP
zIrHu^lv*VHGBYM{G9JtSq62)avBEA5`zJS52g1Pm+$h9WGA*yI3^u}nH%uQfWELS`
zI4fCZ#Q@pbF+KiUl1dnS`E5Pk;Y1?a2r<<PTG|RyCc?_)s$yh)-F$-85OA&&y`x{`
z)q;E>y|aMTDv9&V87FSN1b2m&yrs*jhui<jlKbxZ%LK+S9Z$#WQF4+&iUJ2%I)Qv5
zRrc>+*Pr+mX66=bDYKlT_7;j0R+)xATW6?;x|jID=V<E(-=4+cz+GWeE9o*_5&9Yz
zhzyG*+(vK*N2NZtyi>(&g!xh_?r|K*mcOXg;274WI8Ad5Nbm`5$P=BTqyynY5_`4e
zNd4n*=_!tRfplhU5QP7H!!J}EgxTQAO`{xFYTxrvc)?8>|9NxTdhpz(=G18xXgSz>
zrf{Y}tN3G5PSp{*1u2l{JJk}pOxmg_P$+&TV=ZRYV625W^)oNvzE|v&7&IuFXi>*`
z{H<|r<~PsB&#bHU|4}j#T6XwLONFar>K)l=q9IxoJ~2Lj>oF5d;P;mTWCf9(fssKe
zkIrI(RRUk3*MwIxfBuB;?fcWk`-G_r{P{4xH4yL0_4bOWU_N0O1y!Su1>X{D<?O`g
zdz0ap5pAs65$UhZ3@-xh#o(GisFt6qkH88qF+oVJK`?HTDQ?9t3w)npP89k`=aEA2
zvanuLDP*<BLkd?)%O27;j^FF6uxk~!0On1&z5W;pe_Y{WXdla%bSU(EH`ouoZFZT<
zh(Cs$w+apu?;)-ekp7DM>O4?bHQm*cDDb?TWyM5)J6h00icEu0tB8v{Ao=kp3O#V<
z=jRxLcyuTqViS;vl*=+u<=_-V_4YllU#<VU!JvxFx+LvA<>eh7r;2@5xJXw&jfLe6
zI<%@AW~}X5lPZ$*y*3MMPHB?_Hs@0y<%f4_NPe2LHVNtE{fGrF?FNOVRvLa(lR|*j
zqC1pVg~J`R_07T|@u-*bW>3f1o6HZQ12VU%z=B=Q!4BYBY`)p#dUgveSV-1Ct9DkT
zVo!JS3~&~yo5_7^|M!$<usi3tqrHwH3oqH;Jjqi{?rOC9HrqtgQ{za7F*4<Z=n45G
z#Y-I2&cYgdA5<s^x*nccxMSil#{>@^qqANc(|n>GW2U~hu~EI-Z0uQeVbP?r`iPDe
z_vOc$m&d-%AOCy;&yKDgg8{YOKTYJB*30QvNPzNxdF=<2#koc#3jV#kW|&0A^Jm@m
zYlPXzoVf&j&z={7z(rrEM}}?5kJd{aQ~1ir<rcV5G1P$_`<^0sQ4mT)?lW@!S+@lv
zP)sm5FK_Ab&dgS3xF3DAY5o>u??!)BXb39*Ehm;l`m%aV4os^?Z$bI-EIeXDW;vsR
zmql3t=m^*0T<ZKvJaVG-@qS3OJ!*uyds6~w?^KEl|EEQV@(1SdApi+lKQ}%c_LvDj
zE_e$0fEqlsRRp{hNhK|ED3EJv)2WisWriwUL>Zsg041&un5TinN`S)-XBOohbm1KF
zgbpZB*%RxW%1Qm<!xE`65j0I)m^{lva96kF;3hCk<59N)339D|Z@X-YVbuZOTA2ky
zM}9+mzBdHu%?}8+6LauD>G-_xlO3kb!&Kn5^f?Sn%#JsAPJ6*#1(#m(gzxqmV$K>@
zVy%XtN<F=Wn|EB4{i4W=mGANYDPtpv>ewT~Qbt=wJj*BiYHgV=W=Zn6KMWK(STG3Z
zDXmc914R{jHz**~8s=^@Rt79-?md>V73{Mbd&lo=3;5ib*ZYoDq__HOMb^`zyYNSY
z?Y%6E*G1dT2$6m%s8J>*$>hc9mXUnJVOV50%>A0&`E&kz2h;#ke}QUV3<|oY^+}ct
zflVuw)0uhlG4KyKZ>7Qj|NUfy!AH+iC?-(CSevE}df2tzYjs<*eu)UUHW1IDNPd^9
zICzx$Qv>pTXKk#o<}t)O$W`#`ReQ3s&(kaaT@W9!@ZM>Tnf!@>Di%FxBvvMH(DRu-
z;xaHG5~jYJzE9UgXr@(}t4=KBX5W2p+bhaxTDw!KTSlE2m;Z5o5PZ}Xks`(|3LM?a
z?Sj=@QMFOEc^vPU=pwJfRI1~09inPiV1&oPoxDPoDUPk``t6`_BQ|*;`AmB&N=N%7
zLoW0DFzug!jX#%%p4Q&O;FKnhyy`LM%ASRQpd9xuNtRy%+Lko)9>RbzPo7+Y0Usj>
zQ885j|7$BjKwD|-Sb_t`t$X)Q=rp5zP3c;QBsTI^#dCmX_3G#y)WlK8z;8zIgGdlI
z%QW1m{D<lgD6D9_Hqo?uR^zhuQVi8d(^g^{ZUAbq10P{eZq<i?nwo><kOVh+EmPvy
zExioJNkNEfGQ%58_ThTQdPTkYB`mzY0QaNBdxMN!^q8NW=i{@X6?Hy?%WVVIw3s-l
z79eXTAN84h0`SZ~o`_2+3p6S-#vSw+^pN}twTi@VL4$K)qxZVZC<s>KJ=MEbUMJES
zP>C8&ky>tM%|axg64llMqvY#XQdViSah9n3UrR14VAEKzCwS_v`VFz7cmwuuuHCU^
zky+5-O#}V1o>=j$Gd04z`DHxCO=p&2j4Uangri&e;f+6?{zr^{=16Dtr;ja(oxK{I
zEEIv0fAeQtCjHcMd8@yl#Fou<#t+0SfV07mpnVV64Xv!bUt@@7qUeuiig*hup+pNA
zeiZYsZy}6#`yLSCVyRreoH}c>)q)%xB>dj<NVK)V?>i)!dR}+Z-`y!b_V(SEKKh37
z9Ro9qs7o5hdd|9DpQv)jH}2me>FjAplF!aRy+ATw>wKlgh1*WK$4XlR96z}UVlh$m
z&y8TD8IAPl`lm(R5%fOBplnEq1*a9}O-Zq<M>coTpa8RiwD1BwbFK4ZU%r?GYUNO2
zc_e|V3`bLbN$Ij`UdE`M&Ft>C7V9ddGhf|lg4M@}dhOk<NzAEaLkSZSUsqUW(T}@%
zP-anq=X+TBJ_C%LQWpgpk|kwh?#ZNdg4$b~3c()DIZahb9E|+@20!Qf?G&gN(7p9!
z){pz5Oe$l6*{T<%`-nGR+a}X|kCOhTvTv#hZEUYo#@8>!ic&#LJRfM_L1sJ>1n?9U
z%jlK64+cvvC9xVIixb?oL@-Z8(L3svsg;1-s;@lMZZbv-qb(x?E7d42lj#;)*qxrF
z1^&jWB*5A%KhI-awOJz{(F0-*Zi((x!i*S1m?-^}pj-M(|7y6->fdBn)?ToVFmufM
zbXU|6D?sclt?!kOs<6c`w3pt3Sm8<FIbf5Z2CwI!4EFKv$fY-LiO=i-(cMH7xL-|!
z4(Y5jpO%INAL{yy;?XJ?AnEl#Iqwc%Lt~Kb8W+g%GiwAHo396-tu!p(^^D*&XV7I+
zf;PXzYHxI@M*|X|u_CY9p%7Aj#yji1kX;Q&hYu3(0ilBdt+azRoILjn`|Zyc+9l!J
z+I`2z;PTIh<@Y{rXCMgG0Y_?co7EpfRGWEn^%`1Xxw^Ftz{9YC!}wuVcwfZeZd<MQ
zUKc{jysIo;h`*@i0i#I>fJ*fDN~zQ%GSmxy<!lPWRQ^~;wLNu+d^?=a!ejU@n0~5<
z4ql{cGEBolqJ;Z23NaqIe+(m*Na}_2*n034;BdKU-hZHZlQ&sa)7Sq^S2I6>Zt~~T
zb|Q2(&a>c2V(`x_(F42ixCTnn5su~#Fr(M_FBBZur_2A5HMtK^^XQ3<NRpB=Ip%_y
zFr>S7=YN#oRW2_tb>@(OPST2TAOkLC3*???ApmSLofPcRZOLG4r26Yy3^*#bT`h-6
zcDPOBMX9y*QQPZy{kBk#Ku0+|RygI};pOhGo<rs6?z~lW4JSy()=Tu%@OZtmjgra1
zrZQv8>#W!(*>m-EtA^^uJIK4lzc`LWUU*OYu6RqY@0SEzynUb6{>+@{=fz5_{X-&B
zJl=#WqjdL!zB%nmV^zy2`}C8Y67c(X!FKYtt^sqO>+uTd&oZkIFgoRcDdIz6H(ar&
z#?NkPpe^B>`(!RFV4aB1GA+j4N0Tg(Dy_}U**5`uXHU&kqxXA(TzxnnXqIrc29-%9
zIpYwd<E&m`GjIdaQ=wKm`^jN~i&-!EzpgeY;nX`4)xJ^3c7jynzOen*oUC({y4KbM
zlIni@-NAsING=df?}bd@G4GWq8teJ<eqX?xlj7DD{8~TK$+Ib8v0p*3m4Y29Gw_Cx
zT7^IhzRARD+&EHamiGO2E)N=EKus;)NY^V#DMzI9zx^wRiQR6TNMFoSt4I@+B{V-I
znnNiKz*be3r?h11Nu{zXr1htTKmrRfS)YsP{OEPglh+^M-7v$fqkQsM{^Lt_&P&8h
z9-xqqbv6WA>sZ*WHn-R(Od_!XHuV?>JShJ^3n0}q)0~>BA@FTi|MYeVaYd(EXfjG>
zD7-Q6`ove}*!Kyy^DpMSGr1Sm64vJw@o(^KK9vKrME~yC^M1!xL=x^4if<)<Rq9Wg
z<j{H@x{Bhybk4g6SHXt_)65ZumbazZSCl_+sl|^sCjngG=(R&hFy1jhLtrx%*5v4!
zf5g-;qa2UG7x{QV<Wpt$X>6tI%k<uF?`J9wkGZkf*PU|xy3q$=DRHOUp+et<Tin%W
zWlJ&=Gz&lR3hqhpdlA+vHAQG;i5V#x`2DF-liv4_(BJb-D~onLpZgYc2ZDVDWJ&*O
zHDKh^Sx8yt<WHs&`CO>Gjds2M>W^}@*}CWyvwkV1GQJ0PB1xO#NdQb~*Q$flM|tlc
zCOSI$&jYqC3J^4;a7_c$#By5}j$Gb%8=<$Vd4F&3@W-r&^WPq;Iy7p>sS7<{FLpAP
zvm`nLKa@!exEd^!;!>VED%YQFCCK};#;7Ln3o^h9kiqi4NjC<P`<uWh?ehs#ur_|_
zcI^0@@)x^ECCGXvujWw*_xi+2=MaE0TQn{$+M)F(8n!+E=zjz<_%2ghl7ysAUQ2&!
z?11?zl>aE(bRVUCS}tn^irs#bRF8b&g3WtykC+B}6cQ)IeC5l(!iZ@YA!rt2U088@
z8RsL-G0tcGbNQLdcO9t^;GchNh|QNMd~S-+_4pX9Qu7|rA=^Lq*V1yvc+i=D)s<Tq
z;nGf7Gh<Bv(!H+j2MPm33}E1K<P89mu*_|4zU8hrpC1MjA<;iCD-t(}ojvR_F5dDl
zd_oWG<na8mLCLgMfevaCtC<z+P^L9iBo?Yv!e)HcW~>Dq%eX@jkLNva>$d#p*h-?o
zl3F-u*I}J^g$WucTJ{xrRim|ATK@fsYfcR3biQ<m>dUH|inyT>vfAa7Ga(mR@Yc<1
z4o%QPinFD^rn=rs)^#L4=WV7};%kyl=Urjek&$&Hl6jB6t!9xOy9%$qUeuxq>VBZF
z;H|k7>V2$g`a;(uzrHxzu*f&3{IEM-oJlvrXjnj~f8%C1cS6}<(7u^?)af&_U2%0V
zJEB41hG7vLVS9qtxo3rRdU+7_zY@<=A0REvuc^*<_IwZQZ6wwk7<_+gwT`kiCQC_s
z1b+a*OT$MxS6Z=8eQxU~`H#PgsyU6U5oSs=Is7`xgPZyK#`fn>S-U|T_pbjO?(=my
zg~#K}z$e2kyI1Y)GY$myy4EnfxH|jvQ()t*OZsv*MiZhJzqX&5A1$`weJEVVU65>b
z0U04j^m*%0OL9sQ=4wJSAAc*vFTx+{9CIBW9n8Ht<)Z^=Y*ou)F4oJaTP_99wkQUo
z&oZrPv&qknssDBF!du&98l3;G$i@N1)ERlepoG#7!KR#3>jJQq@W60TAVc--#Q`EC
zGiOSBV{K}S9&z18#Fq%~@Od=^y{qK4q2+`6Q%%QDrLP2$zGqxcF9WBzSzgFSD-<;N
zDgW5JhN-g}8aIVZ0e6+33g^QT`3S`$zncTFbco|Ds>L+4_K1^&{LCwH0Qli(E230(
zO8y2*i8L`)9ycze>Gh*u(`O+Ta~J$rO(`oBR%j8EhIJ5s#5%8HO>Ts3&H%$<z1k+E
zXJ&Q-F0_o1dhdwlxii~Of(|IHQsAvRwEE`Fbz7g;?)jJ6`p!3hWy}eauXLR{wI;c8
zw}ye0$?YV`?HIOu9ZZ{yi%4+U&M{g?tg3m>23p-s!e?GAr(``R`LXqXXqb^G{?YPa
zw=x>pZO}K(+uoKP78NvYC5-z<xVQclWEx8{+m~u!!p>XcI1LZMWo~pIpso)BET(fa
z3L#r>$WrK0n1M4~-X(5qGlfrJ^crpSW-z0o60Ar}ZbeJw!CivxI!5RD-y)g~Q#3s$
z)=TbJ>2V4x%GF^MdVp@1J4S%Kk!asF=ZLVS0J;$vK(`BC%)XIUN<rPax#_X$dEvP-
zFVBzH;oRMK;T{6EQ)X#X@wl8lFB-4y#c5SNxmmSSKBO}G$HC)|+<#qJOCirfQ;smo
z!urC;qO=54UkHgSSzJ0X$Y*u5_#&8qAnF!(>vwsW4nR^YT)-4^VvGxR^Rkh1v6a9H
z@w1PS5_N(XcL6*nnD$rNBY?m<#gjEzOC$EtU&~S%&}O3QR+0Eur#?VpHjFr$h@gMH
zpBAR!jA(PbPmVpAWyMS2^jd?~pJmkQ85bt+n}v)FS7=}XU2d6`QCr0{Eo2K5#c!Bd
z%$9t}_dYLZPx#GW8!UwE3(AfI*R_P0dVzCgoml<P3@(C{%>H$XE1lO(I82tm^0SV-
zy^lNaeoBW8z}0+E4bZ~(bBsr+i%_x{C=Sb;yQ9aF0q~^`m;A!0Jh1ND<sppT&uzcg
zqQc<AapR)xgX@wMA@VixWQXrCF=&t%hsJdaSY#L<_<d8$gNHx74Xo0KQ366fW6kBQ
zHT|!4$+<qDx94wMZN+#B_*&0KZqrv^5=W6J6nmWwsowwe0$=@S?c~)g>!(hFQDfT`
zmZsN~r;`0vQj7`kX#kgY;5qPd{BhgO)+6?S6|nt`ju@kREY~>=-H`~Y6w*JSwMtd6
zBf?7zQE>il8^XfV{F!|!X3Tnz+(Ke5nlJFSU~;zjP1+;=m+B%cQ9S1CRdWtfzv?Ek
zIFI}WS9v9KVbtiemrNua$+^zY3fx4Cl|sx5gUquM1pcwZ^l@s`?B7O(`7gzqt)0C4
zyPkLc+`ZINZCs|EZY3);`v+@%En|J|oGfC(r=E_M+_Ob;!WF)8w$PZ2y8W;xvjq@D
ze%X{LLa*lW$CWp9O?3$V**Fkju-CH8HXFma0!M=&xrDRQyr}^)<t*M$N~*e8LnK0d
z_bvPlDb4V-7-{^e1T~^P;}0ojs}?M$vRgVxmNc+4>(~+pOp#_xomOrulC4Hv(C`D0
zh-RFFGSd+XG|`a4GiD_Tw6`LF{9ZVb{3my&+Ip#Q_#Qz==e5Mc2;`O{jr23hipOvf
zF?tPKiKv<#oi_cQ|88*+;h5IF%Jp+_IMRe9rh&$!j8K`hFGaV*i*;V<R;{-(!>Ve~
zp~Tty(v1^<Xi9nYy;cA1tT{^tPJR+~#_jQ9)u*ku-Xj=z+f<AExspyfE<WS80ZA-b
zsAG(xkc49T5fRw1)((4{Kp;s-m!s@&>)6^b_}J<7mkd?wd%O8N^;rfdy(50!9zaZ;
zfV?*)!|`o+YP(>#PQDD*w%dByG<eUs+pM#_vLVL1{)qINIlaTPtZVJ4+jOt}bP|5L
zaZ$J_#c$e5d0cfmSZe|*=ADg6X&Z<w{eyeH&ryr@`jC@(R3}eMc-mFEGQT!IO*BC1
zcD%aucHZc}V>WY!x7Cn#mx6`St4y?dZ6i4ois0()M3W){rMBHnRpO!hOssb{-IjlP
z4|hB`4tD=s#ngzxX*$?b7uL`!`zV{LYPGY+zazH^u{KVPE&<q^Ri`m~P*}oxH^559
zTt3Hv6ZrYw1lL`mjsWT`N5tfP@KiBG*Jfk0A{P&dWw!IX^Hd>tRc!6ED7}Sjz~oJU
zvp%^eZvPumvzoPAs+*Cs`RnBk(xuZEanrUde%lAHe*;OND>PaL@?7!9+*EskdaFX{
zCPy{+2Eh*Z!T51UqW_ngmrC&=P(cK|Q)|E19ra?R!)$iY|DeSGp8}0O6o-8_M#R9P
z7niBIu>@Gk2?oZUyXKF*k68KCC<L&p!;ltWnd|O@)c3RKzP?$GQ71=_;M@9#{fAh|
zMtUV`<OFqiqQot#^4H-^00ZG&AA0wivAScjx_1A^M_^Us(EqD+M~~0IVufKV<+Aem
z%YrDaGSG5_KsTPU4Plfg?;|;EU|<%M#B7e388uN-sY$i&fNSQJ_>6>+5@T=T%3oF&
zgFoiHBAkU3KcNHU1Ws1<<oYJto>TqyWW^V<(3vEy65fgMxBlJnnX`O=PYcS%wybOp
z1zFx<9j)6Z3>n~a9*RC1_<km*VM1}V6<4d&J~3-s$Bsc~!ASfqX;cXPbXLQ$F*z{s
zkw_WXV@W*#{AK4pF-3jAq>~v5V+qWif9@jL4FSg@;U)}4hfVTM)6t0|E0CN18|z>P
z+-tdI5Zx-Tgi@9LYzer#kg68qU_?3QT`x-siqE(U@#%(8hr^4gk%%HbFLTTi8nO6a
z%*AWv^Ui^OL&T(Q_II$6A+{{flRxqg%Iz%*f6Ku~=-9i*vhyef>qU!fh$#RLh?9kS
zTAv|bj{e(IjmKHM4<~(KYuf>ot!%K)zB&t!sO7}BuT6p1_QB3<Ss{-PyWB?Nv~fUf
z<TGKF^?c(XN1MPykO?SnDgvC?pwn9lW?m}tja#)BV1Rk7Gxdo)m_40wj4}cM0f;bP
z<JJ-tRdCOkM+ZMkAKN5`Kn<V-_9X<qYRIh&a1iR0wlgziSGTcs+h^ih<3P5x#all4
z5GTFIsRn(!$RZ_+BE$$MET!&fYCk+q6c~H@sd^OX)SF6L#DwBm$0)<sF44XQ&zS;-
zwrb{DL{h~)=@i#zSpq1p%^c)5p;0ml4!V_1m7^T-Ujfgp#X{pE%84*KQB+?nsz$MU
zc&c|@lNA~2=|f1wN7$p(#gm^tC_MaJ0Dl=S=$4)Yq2~?ivj+JCul584p4mu~%z|>|
zV`1>5LL0GUfw$kKH0QSY&?EKUqC<gpg<z%d<g7fQ*K@$lgxXi9k#6lmx*Xhu*~(<?
zIJc{oLd%*yy&H-1%GY-)0j~e<q)kZ2anA<<lsmuzM`XBr?_$Kx=Q92#l_^t)_~>rh
zv=`<BQ>hq+;zk<jXaQ$)Re18rd7A2EY2EuwyCZ<T08VDp5@Pk#sqb=JC<oGT1J#6<
z09uk=Wo@gDx29W;$~zI$VD*77A-FT%-`3XcHt)Mo?e_{H=J14WzPoO<j)w~1SHt@n
zSG^MOZ)u_Oeipvo)dc6In8zm&tE4qXsS24F)k;aNg0JjuUZPi-@P{QgeQPv<OWzXv
zBJana%%eAgDP&TNwV(EZdT^z=Q^Vz#Qx5(%TwPU!{f-N&Db;#T|6rVf0pBDlOu*0O
zeM<2R-(&>eRJ40dGADEMgBY?%83o0L;LdCa=@v`hBA2`i?yN+rfwL#Q15qzejo~JJ
zGnMJ2A3$jYZa(bVmr-(Xq5@N>57Bogr{Q%ml8t;_ofmvVwcoctx<v*7dh3g6-g<l*
zt?A9##1tX7<A^1v;msTIMC>2Ac$fb+Iz^In_uSV`(=SV1PnU+AZduyVU6A2w{WxqE
z3@s|NyQ>)OYUq~Ytb=1j!@>vibs&jb-S>ezajFTb%mZ~_zbz<}r(ha=09!OkMZC8N
zOc0+4LdV5M_Dk<5X}SI0cFc0WwS170d2n*@_FX&rkF5U;a9r;nmQO#fd1l_aXZ;}o
zeJ;9vnptE|6J<DmwJE*bYSDrU&tlq;uhq>1Z1bd|M<o2g^YiY}Mh>?}1~@yBF;jC0
zSu#M$qMupVvMB%7jXVh;q)k=%>MkuMS*PUS<wpwAQz5QCMltn9#wBjH;7#kcYzk2B
zmt2NZ?HpEy&+L+iG*(*?k%19DWfoso19|2`>~{b%5Fqmi4@1$|(;$HvYx9W);r$4e
zaT<F(@zIJuaw+Vq$WE;`;$#yhmSDTUn7WeW`yBDoFnIB^sTEF!RZG~PH~%OgJ6m8*
z%n%#oPlJHE7b}_@^zLn>%T7TIU%9#k(|{;wU7PTEx_`I(Cl>}CgIQeo++zmb!#o5o
z;%QP&km1zsVj0Uj%+CA5ceT>FfmHJdVeDz)7H`;#49!zujkj;HP`?uAwDBYV9)lkD
ztb{d2UsNBTJ+*)c8B;G5T17}7lx2Q2E2Y{YQ9zRTP7sC4h8_Sc+q1#xN8f6&3R{kH
z3`f49ON*H_?<lmfP3QT^lm4(@|Fe8Ch=s!SzUM*1D7YRR2`9ptboC)<-*AeDyG7w;
zXi0(}*oIl(%r<VS2<xBAgOkaYF+HMjb<L%a2?7{&0+2`ikNP1lDSni~U|b`L;|60O
z%kaOy`I*5zjmE7wRnu2<TP0OqHndxv*;gGkmAQl-kr8NtliDzgA@3DqqJgC-x)CX0
z;_Q)DX{D!dx;&EMcbw++^uO6JerPXQ=DPRm<C?-Z_njF79?nmDd@9YK8KX^X^4$$M
z_wJ+Fp)Np2#k+@0hfEW$0^w9kKPLJr?UuMGhg%9vf}nNHH5r#ZXky$F1W;$jjmWKd
zPLx|0BjAZgUA~p-6`^h>wP~f)ojdm_RU{pX|8(tc=`8z3ph}TeRd?oxfktJ<w-QX$
z^aKsF`<q$?a%N|(H~v%PZD)^O_uYC%l{gM#vQc#v(-FjpsdlO%_7;46@z{s~T8wu5
z;&#-Cx1%3waC>BPU8k``<i?TdIcvi>wMmYQT8`X-cgd^G*jqIx0bo2VsF|X-4rboE
zTcnf&%?~Z=@%%7z4ETMymXlA-Ns$aKet?nza>5#RR4qLw-EECBWlH%(6g)RabfwVp
zmJ6RRv@ZTjN8nuEhgIZzN(5(8sU*_X%`Iy{&LaG)VWQe=U!Y#W!J=syG*{&NY0DgB
z*o0exftI;O>{znfW;W`6J@hdDQ%1K=1%~rc8=4Rwor$~JxpC#v0rkHXbGc*hv*c)O
zSdQRbjJWx}Wq=I=yonfVHejU@P_=ezR=U`D_Cb-Jx!Jz%9IC{08TcRsY$+N=WRKuX
zMqOaIANmtqRK*U0Cr(N=uuz)RZUN>*<aXjd(_9#Zp_XQs1*K3zshDX3Nz#c-gd-6R
zONsoPd`gH~=nFqIXkALs7)DaL+<1Pvajqs?z`Ey&XjbQ_RjF>K*^Ke?O$-<9$sxJC
z=96jnKe}in>I<p|R0M9qWQYh2P+FR%W1`Z~*jP<OiKSz>@#LN3GX}8uEIMnAoclC;
zop5`yeH?xMBuZ=kOme@WXd9X+{SzktDmiUVKvYjc^^BQ$+>Aj(HxWeqw;)1;X}Nn=
z-&<#*RGfvoivm|wyYHYhedbpIYPRrb(I75has)UNk}$olqKF8p4(CT~9fv_pyMHLV
zf;J%Yrnzp1o7nWW<acD!dXB2I%KIR|L%}1g-Hc11?Uxab6p}PwQ1B!%>XrtOg5qyn
zbnyI_mZ`y?;MpG-FY~q>hacT*^<U^vFzKf!r!O(`wj>%w?fBm>$_IXjRI0S;pwT1t
zEwTgU)p@rIO>8Dz+p@Vc<|wdBbNX1+3am;U5p&l1B^nIe>x$&k-8z+Ti%!!5+#ukO
z`4K~!kZhyO(KawmM#Omn6!9nRpOEr<^G)*xuoHEnbLB9?Q|B<tF`};CTY@RmXwjeI
zGvH#IQp%K+>T^jY1vE5<MkN5F#3OeW15&^*MF#;<4kU<xD=;=lm+4|=2Au}@8oGH(
zaecVXYp<iq?RJroZifr_^A>wWa+_J(U3U7NA=Cs6|0UZV-bqGsdWm-y$q&o20)4&<
zIk0;~#O)w&x7$4ZiB}n`Nb)T(xbKm0(nhR+&}dkHB`*q1m<<&Omd2G6O&`7fvIIh%
z7S#rpAu+`3DgQEhA_W6mz1UI%DoeD7@#j|y0N6SN6@h@1YNp;C>wicr!DLSTmqy&k
zCC+)FYkoTd&G%Q=BHId^etGLtj+AiKoeS$w{p0wu=3~)6G=dyTs}=yr2kaj!c;SGA
zIN*Iq|NGbG?c2=$W;^bUBSDN&`y4&CK}Wg(x&#)k=iSqbyA1<-o7}4bd0BqU{TTeN
za8U^q^PUd*@llY!eaxWd4`4=WC;_cc|JAw1!wj}ygoqAwp#f#UlU-E0wd~j}+URXI
zAu}NZH!gs*>9_WJ9ZuEOUK3JZ_j_XK`p6vq+VqDd*qgXc3;)6ezrU{Pm8D?&h>_^B
zYCFc4{w9=`Cz%W$iHcS=;IR=UJkB}_6au}iPl7Xz_e)W=&y8OmCvpRxxSOM)4oOsH
z!l*LwP(#gjZ=8C}q(%urGj>)k?pyz<P)oy>>8B4=g(yLjdU_{L$TYv(Uipc=9?r&?
zl2<!j3Lvy1&;9kZL{zk~hI<|p#BOmFB=3wDm*7<kE?~sJgzW&rc+1B3B|UIToA1*#
zY6x^^6-p?n+`)*~-0<sC`IDIrjADqEuX1sI*jwe|Wg%f}ZhB?ZfAh!Jqn#rdE?i|i
z8v}MbkmJm+gx|{}++-1E@6%}NozJI?#WxWFm@zDXjcKPCnHn@R+rTlpWdMD$$HT#0
z6^FD-N|&<?Qwx=@P{djo1OxELb>aF8tc!;!0sk>%@)~tSqM&k=U9aiI@pw$UJLNb%
z+P()cGa%IEn?0{1;!LO7_A+Y9CMia-w<Y0f4c|%QB|4d}U+vEOHwpn%M-o|-Rly&c
zE4U%9qsX&l7+ZrEo2I=Bx)9>Kpa-e&41|T!=G=5_2T6wi8a)^_Nwh+GMxq<7j@;~N
zw~8WZrH$@qp$}zF+dBNfW%c2Qt^osLDervpxJRkIk76YUGps!A<YEIInms=LiMa-k
zmm#(jCoGLtR+Ro%{Cf{aoyz5?{tuazd4rQf=s9t~K#F71dhab@k)yx^I}Vr+!tO@h
zEPm35XEK*~tE|f3p_G<l+QbINs}(!hUBqE56i405CSZC9HULd6&p=|80hbFv1*X1j
zqp5}<$e)dSML(LU#I+-7b8UfWZI@sLzy4B5U3<`=0Njziy`nDBvOQPx`vs8rO#t?f
z5(3Wn5a2D7*fDNaL<xcB2o)AeBDLMn#RC7f<_X<bHBMr^o03GQ`1SUL<1SI;4r^B6
zvdS*)lEr7MKlE(PAn{1qFdqV0ONavlj$A3L{;0e}MncQgN4pY=Xb3!wYCvA;@9LI-
z-Rvv5km8t?bgu&2AT75Fg>!>-axWwPJ+C%>Z|lO)I2TgG5+e|gwW;oe=a{U6n8Ja1
zF0Phk2s)~;5E&!Hd$Y-t?3#a-?cX@KS!w(e`Vp;M5rwA7FJ){E|Mt7yyerhz*H|AO
zU;C#4F^bDJ%4eA%|F%|ikV1E2ufi;FRi&*y1<fD(U1ip@>o!n1hOL7c;M7hZhwv9K
zo)&-Wno7X~h+3^FHwH=qHFliJ3c?Y2OQ#}fWIGQ{)4|Eng<6>=$-TsCxQi&OGJdeB
zvqt52b1nOJ+0ANh2cv)6E_imrF0FdFND-mClixevvZKhOCGxUeO@UmW8{gA^VFZ{~
zzOvutk5;O<QZcng%tzR6Lt#4FhdJBE3>*7FI;{d7^@>=xrS4_H83_5MQPS^%iLla-
z_JA_58b|~)Z5|?{YNi)!W1Ro!6?;^A=1GY!q<Ty}MC)&73CNe(Y)#0zYt}!D#3!{y
zbZ$<+MeH=4nz{Y80gpa+<jlr-^{sh8KbifzF$^QTe|WN^*1xT*t8;mI^}b!X{am&^
z5`Fm`keE}Dfza&_)I2o73<hzEKj_e5;KKP~`s41uv2R7)+I07L!F&L03n1P_CAZg>
zIx5W!{`2RN_?@BTSE!EJYkOAVy1F?SO`S16TC=}X5pBnhKxpF-P(K#9cC4TE2~({B
z+lT_r7TuL<UrFDELE3@u_dP23y8qD&f{+|SzxR+Ng(5#n$MXVT{}!O|(=_`BE4DR;
zib2(#;~DtZi@e;Nz<#~{2b6t<=_II26theLt7C1f8a^A#vqT?XxiN`43h?*5<Gau0
zn+69cn_C(C<CKNv1o_xbDm@ap8Wa!NwN`B^i=SsD=z;Olg5Z;VxrAWCbh*-c$xqM&
zjHY-PGX+LeHK;6ObaFUt&Lo%A(m1>}BMmK9BJL}k=0(T>Qs&*w;oM4CTPE2U!ClLb
zVw6RpQYcnk5y!2)L2Qg%(jsaWq!R>*a)kmpf{Ykllelg2omDeR1Axj5Km@rNKWfh!
z(;fyQtCtW_J9t>hZPQ_d|9vMLVA7?KJxXI*l_E|^no$z}G}p?l3daO2<ho$Y-ut#q
z-l41}`G&hU`RWt5`<6Xhw2nVGE)#f`KXGdFS&508h^%{M+u>HsNAXUqt{FYm1*z7o
z-oLH$eGau(JH4dNYCW6oI^A_|YFAUM`c5nXrANnihoU4J=ip4m(BF>Ch9zW^An{z$
ztqy~=VFs-}&={bQ=f6eRb>QrNOZ1N5xwAs?R&GZ=x1YIdx4{sER$k0$I^_og8#`}X
z)mxP>jow$xFMxp&?qBdduvBo^Uu?02*IA8+RAbcBV<4VS-m;7o&&?1my$4v|7=J+d
zU5prt%f`EYq>Z@uNKclBTGVZ<HI5w`68RiwL?NeI%&xcgooa>_qQ=MJZH26BLy>6z
zmFKJm3#wn@t$68t%QUc(@p~uSM1m_Nq&PHg1Mu(rPw&+(mIqlI)cVZ8=d$M7oIA(y
zPcdZgd1(+N{!crINY@QGxvx1shK9cpxLtOXg58QwY7H$elK%6=Z*yD%0?=>UKBHx9
ze!>4I080Y|m3RUkeMYOLs?KJM3lgkA4*~o>*clVMdt-#D%>!o2T8vEaNT>v1q(&mQ
zR_LU(B??eS_Se6FxyBrK#!KaOR<h5UW7s3Slgi&NS~%S)FMV;OLmI2)`w|(eAnu#`
zehyzPj;ybDemo4(bu1qM)4PARZBQ!$cnx25Mm6_EW{qoSm~%a0m4Fg$b3@|!q{|M9
z_kmuT7@Hp2?<_Z9O|Qjb?8XrYPyW!_Xise+--{RgMzC_g{A(PbFoDl>TSxEDCkJ_B
zYH3eS=$ZmrY>9v_1`s~2+t!z}jk!^zME#uu>`ff6xc>Zq7GR!O-N}kZW@JCaQts?9
z{hSV+9mrIw$Y7MWOLoGoXCNsoHR9(Kn8)~C6#Uz;ZKO!}&k`(W`~{u_)FDgSx;dg$
z#tP(5X$R8)rVgJ)4_Gk)9|6h3x%joYPL0073Vof2-M$IahN<8|W=YU2Z}Xt(g>xh3
zuL~N>ooFvkU3y-96V*zg+4zyG>z@`3t7~qYQvjivq~~vUj<3#EY6l!<Jb>)qvaBa5
zhSD`?P}0M@H&qq2r5*X_{^UaV((ngS(Wu4zd1l{()A883jHFM@&LB-X)*08ReLfDZ
zLfJys;13H8McJ_OCVtmhyEOiD3ggo!)=u9`&ccg_k8SXo?%OCr?%REgo_rq4^NEBd
zby7?u-(t*WCFzub0oK7pm-}OkjF=LMrMf0lw6s`31`z|YT{6&q<xm8>OpE$@Tlc+P
zf|)8F9wD#VvKm)f6sB)Mbvvv{mC@)$k*grl%0Hi-J#u_@?5J#d(51T;j)5MX)8b1Z
zLS*#slB(tgO+_q3v}k1d<~TS0^1|Ras9N_04Qq~C2oy_sh)eq6xwp3=H%r~g*snIm
zY_6{%yDtmo8S?OLxftK@55yjX%@~fmdihF?(`diS%*30`N5*8oLdhs}j-BkYsb}DG
zdzvj_<tg;(wJvMDU(fs1>GS_?iOp}_#~q3*?**RDy7;$3JvTa#sG5tm1kbzi%AFu8
z&FIF7G;_9Q*8Fiks$SkDuE9c3e*jc;?T<NLR>bm_>Ke%g|8-M8;R(U~dsj<1>%e&>
zZ9)FrS^gbGr0IjjN~#COnc7<eY7#<%F5Eh_kqS2Rt#Ld%)nMwWcl7II8%jc}_XaPU
zH<CBPg=dd>z7tP4Y9`6k8ipVs1yaf2cPY}>|FS2O22o6>-Y}8d;v#d!OI*o_eVUu@
z<nQV_apN~N(sBQG$h$W=9;ycQ&(=nh5*zRuDF%*Hz#Jw~WPpv6&i^;Uk|rsDep=?^
z0WbQk3trq82mAd*iR3<Fe#jRF6iwR@C);?_j@kxt#Xoc<(V#j9IlhauR`At^5%747
zN&{#3p_Mwss@7307$NYYG=+U}UjTam04|LS_Gl=k`G?vX4E$1y-boVvnYa0{rH!=r
zfCvJ1rT0;Er9>eNML@5_x9H(+@I7zx=YcM%8h)?xPkdCnPnFzdfB{R1!}3Zv)^bh4
zq6e3}iKZo;jtw$Xkb=)?_lp1{eTC;!|9tqj-XjyNtx|fH^X@heVW&7h$L9bMcgL?a
zm9}QeMiDFjW9TZ$(qJEo(JdHB2xCxXzVQX&v!pq{%#~*CRbCt%Wtu5n83(d1>?#C!
z-nm|%ZWS%o(?sWMnsZ-9?5Jftnoea@NYMaov!y}jg06@F{xIEPUA5XG*4E%zr)dW?
zMR;UU@5yyo@Un?T*CS4=bx?x~#(T-y>~_k(V4i8{&1uFz)r}j+s9#nfCH6yp9DQSQ
zejEfQ`f#H!sm+cFY-Y-8mpFbM<Z0>}xUbfs3NlR>(-w34W+W};fBpTMB4`jP^U(Vv
zgR8S)0)D<f;4oyp*NQ@p6hJ!?26WutGgcywJQ(?ans{`MTW@U`IidTrBqRMFQ(R$!
zFhFFX#vlOhdfMn%JYesrrPG90-&wVin^{S;3ixkUj~S)Q%fXROz%9YT_`}1g^=nf*
zsSJBJxDIUOJOr111SlO$53mBCFW-e2(W@<_o=s$DScI;)Jx-tDv=v^^{vS`@0G8()
z_Pu3n*~XJ?V_|8lC)>8I)v|57WxJM_mu)TEdhh=4_kPDwM;-C5`@XL0JbyZ89J9rc
z#<zR$JRUUR)*rdlea_`pUFNLcj~onJ_7&X3A<KyO5i!U<=Ra3G&pN8^6xI@|UC>q>
zF(Y{a-)tpdA%1^e>Z}Mq$;bd|ZZipbb{m@+Z7Jf;-=IMEF7BTr+SiiONaT&|bM^q_
z#Y&HZl{DoR-K?TeV=c-%6|>^sdg4Z9;={IIX2~-~4TOU)hN{|Acl3*)!I#Geh=6%U
zs$AE_^OrW-B(j&Ea*$rvm$sklmZEd&Nk5B5KvVt38!V~*WFX2yN~ak+kAm2qtd2n)
zaV6F7ilB3d36xpX7_Lo*A~dD2AI(A4ZX&MEd1@Wp^fdKX2s7NEn6>OGq9=owLY?ta
z*~nllk)-rV*~g0<_Jo0xh<MJ*%dHM=RGRLWIvM~K(6OPe7l$he6of0Y{~07*SwSTU
zOYw6`A-X-x$V|`pI7LP@0nq1(%t<67>&2s4Ge7{?#f})&rkm-RZIR=7WJQMKOWas_
zUV2bHy06M$;Vy84(+i~6%Z)>7qJZRxx|cW&YIQ46o8-zcHxR8yQj8@XMR32o?eaPw
zs{MQ#z?^6_;2VDX^z9R3uL7IL>hULO>(X-a)^23Il0Ks=pY5}^=1ar57ZF9=;5a}5
zW-`6eXzjH!^uPTF^PjODRplIr5?b4QFkRCF)zO)ude!j*zn}iV?fuw5^tmtBnja(?
z70Of%=L|iEH$w|{O;LbeB~~rRoR!IA6p32}&|DW%6(w$TqDwp&oC}?wCHZ7<$JEFD
z(ANhT+0Y*vBx-e06{FSk6q@;A;REs-62|mnYEw}dKYJWJ_^6OD%0#OaI=-C`NKZyk
zNZ<)VamWiwe$#ZJ6iSGprs>%7@K?}WnzXUZwqWgVB4(iu+z{gfx|@Yzjusd0eMtx*
z()w)E%Wrt#@Wb*?G9H0d)*uu~siMz60glQ1)A3Bx4VW-&1-eRu2ouW?nmiv@HHtj)
zhy4NM$Kb?5d?<Q$Q5Dt^VB)p|T^_!$K7EVd?)?DAoIVh81W+Iqqo(4=&N`ffOUsV4
z{*cynNFr^kMK9~G(FxISpJv7Q(dlZ+x^-~g)D?Q<8oqnB{<lz?keYmr9njqlON%S<
z^}VhqC2G)VW{BS|i1TNBz6kqkR16#+(~p^4DeE3#JdwkDvs&uvx-)6k804aE*p)}n
z#(~K)RAFpPU?^y#8lOv^8U%~EdUiJ=#=fypvyVyYRFODFiTp%KvSG-<Bz~T)LDE?n
z<q^Uiv*=$loZ=R0&8R>!hYgCaP$jB#ezO|BI^xk?{-d@!?54*kE_KG1L@i}ZQX3Wc
zH3SW9`!`(w1k(%TtJ%G|kJ(i(Ke%&4m(sr~0Qr34$Tv`q!d8?!1e$vv2NNbZKNCuA
zUkz6YYAQIY;*<BU^W7@lg7)<@(W?g4>m?mJz|;T~yE4808U-b(Kg_Ha|HxqVhKRv4
zB1sP<?vARAyXcpcZ^s70NRh-SuvUI*^tS<S8^Ge4>FiGaD&S6t>^=dK#?ynGrqeC6
zxG8rmtHh+bOH<wEUCWpS?Ti|+YQcyHa}Y3v18TmVfZi_`wT3_bRK~-`2ovbII<!0~
zCX@%G0c0MDdwv2wph*clwJgI;1@r=#W!?+~oNPj}xYAvvU0Atj0OdOVSz5z>$V*^l
zKA8mpqe0mJXQX_;6~oHn3_F<=UN>O9Qz?*y7;2clO*i~r0Ae#g|Fz_9c2o(9(!DZB
zfTHRC_cK&8Totq~SyIRQ%dJ$kTJUn1l0u-A2o@7yLmcR~-1s_A(n+eI-zgi#<F5rf
z1w}R7`zcl8Y6L<ZulP@KZFYq%w`<+74|1EDJjBHPJho#f&|xB;bh$6J1c&iyRBhd7
z8z=q%IP|!%uQoa_T04A<l8U6;)S%d}QuLCLW)H}*RMdejT@@$8QLs<;aWC^T;fwI)
zO+|+Of%VUw9XvWajDgzDUC~*D6SMx0fFaIM>x5loT37+Gy|#hy$s`*?u=>Wr-xZ8=
z7@Ow<Syt##fJEr!f?I<|VB0@xKGMwUV_^oaNi~-~fgDOmz*usySIf&Z)WQXdZ!Qo8
z-O=CiSmmHE#wk(>LJ~E3>a?uZ2m3)TXbOm!+f#o>=(hfNu1?I0V&P9RTzMox$krP=
zSOX?E9MvEdjGP2pl{<%@&;YD1>$H_DAeCRSJfvDuKJlS6a0*c_!q8a_u-GTwpiA9I
zvgjfMh6^UkVT=E5{a2gW7K6Oc&wG(|UDFjTOd+rT<wtkU$)s%=lYE8({!VALQ|SQ#
z1>{glL$&-V2XiF_M93ukkFL&WQH^=Y;z=&kKgc72a4Sl9%p=2ItM4bY?3`@h_1Q;G
z0?2R9gF<fr=_df(<qzvf@Qy)$)CkPb#wXS8>y7mgCztWV0BD>2)CLwbH}f*#5V}`w
zKQuGCkN{yCAQQGQh$QAYc*pa8{>qk$L$uZ9Pm)y)gF?WKv;|bcoV>+|m{!ObN900u
zDd2rA|4U!jt<e5K-k&pSdYK1s4<O0Ny4bui$Ia^ZJK`IDO*B-(GTR>93q;{>N+R72
zC|_c%;dZd*UXvAuOVVTkbZ=F^rRA&`xX2K;ag<yKV9t$u)mPjQQ9_nRo!d4mzSRlN
zsp+{RJ6o1NxA4w00hvO!f(pGKO+=W9_B?AiF9NhJf;?$0A>bwOP}dhm)hFn9?Kg(p
z_GgDGKzg}Y4&|8I5@xrEwrmdI4>}l<Q^J6as#oD``B`4Xt^gGv?6-&Q?A2mn&?sJl
zO3LY1*+KI`lG5cjQ0}D_O#^?UwgJ<3Ra|k!-RKRz!#2TtP!qsB@kNFLv1P%6P7}OM
zXC_uLo`o*kKfW%Q6!COqI(D87t@H=kmh<f6Sg!vmD;I?rTyjmC-!1d|Pn+5{SwWNc
zn}wP%_q<bqt#$AY$>$6_p*ns}{xEo7c(1KA=NV;Tz3_)NOLu~23hvHo7y*c!T^y0v
z=M?KYPeRCK*V9sOWc?ZE^H?^MG^SOz%4M}n(Q;eoh!%_j=J~=G<bi`q3kT$3_Ntgr
zYX^41-Hq`OWmRdbV!KZNaO{8}MUWi>jm|;>?IQrSXjt3k&PYQ%76M8jV0N<zxxwkL
zqZ2)c4lS#o(RKuMLVUxq!rW|<A-a(kpKid`vkuo>m5!!s%zj4Y(k`LIpv0B3aB&lk
z=6~!U+pGM2t0gTv>}WSAQD2Gz0sz~!sz?wwD`op8RlG)eT{UaA0I1r?z`jY_O+hF@
z5M7`1bF(KU>FS8y8wwy^wzb+VkLLK<16B|3D|3=LvR<)E5|$R(#azV>-W5|OaIm0E
z)!_b5dekn`o(s3uVYfG$I`0XuxB>#4SDX+q#(jZ)XSefIx$vtRl0byPDQ4f7o(tJN
zE~XiCitZ5=cjq{O5b0~KeI9y*JAwlIWH-#ed5G^hfAQzujUjRX6dC_sMTMcWQY{N(
zOIiG}6tDACOr&HjOXUJiFc&?BZ2<NA`q>tvGNCK;0>}Ss@#srtx>d);ZA1u1iPAq6
zh{~jy#LDoawXmOTq^|c1#q2fClrbhSG%pQApvakuA`Lp5)ZJprenGkW{JId_4*$jX
zR9;r-8NJ-4E0_djC-H1&{gCWN#ei<w>tQOUSarY*F*iU3Ed~mW0%}Q}@miYo1QVqK
zJXIP@O{Ip5VHG742PVC^yNjUY3Tm$%ox?yLrL0OfJuHJ>*42OZ>=L`(jo$bwn5xCO
zAKG=pr1-X$EzY_3ftZ?=`qL^T9+Nytsd${eF;0@tx9e)`qG7eKiO~BY$mgXabVnW5
z!T!N}3V5Nm;GKcs;?It@Tn@HetFsZ*k>dESy`?|Yq-YTn7vETbROT?qf@^6YRJJ%+
zbKiU5EWN@Hr92A3&L24XE9txYM(kQ}?HnjuD9o2sdq0{sA~N2}%nj@o&;W{0iWtoF
z)Fxb>wOjbzx`rQ-HAcp2H4DvxT9xckk>4E430c}Ef>wB&nW<a;GO4#yhRDNx5y0rH
z2uW^4oC33|B1`G)1zkn{C;v{zy>n4qvKJO&ia{rSe)5!CXBNJ-@X~x>?@VT}!*bAH
z#KF(fA)E|>UhiCKk7_QDX!_KwyjD_0nej7GzaL;KZBP|tC5LCf^c%2=kxY&+=b3-j
z;G>WFVhorgGQW<_8sN2!e-T#U@kkv|8+k!I$jS>AeuTMy_##C$34~_U#NV`63sQ=C
zqPgc0S&PMe1+@n%n&TuT0TLt2!Hg6wUmJ??+jnP}N)SbEoY7J(=iGI87$~lNhK_qJ
z%C%D9{fsvRl*k%22Cpz-)4*mEr~_|(a{B680wpsR1402-iN*s>mkNPhZ3SQqH?F)r
zw3-A~S8hwf$;vUi)6bzkB`WU++9d$IsE&1P)J37`1~6U)9a0~YwvELb7VbfE2@3*S
z-HQ*M-!2ccsM2ibq%Ff})=J#|<p#T5^kb`QYUZ!nbghlJi*EiTekM8dxodmKTz?>Z
zebjwBs=9Nq;aNL}I-8oBI_e0Z(Sd1gh(RBqgHwUQJi7l4Wk~35d{FQ)d@RJX?clp6
z>oH(2fjfV-X0kCVkD@QjN2AC=g#|<?_j~=kKz<Q7KaCls{-+zsQ(AtB_?~(qSXitA
zK*c7)WD3TMIv<7u``z`ow<Vo_N92&UwucxVMxpCAov^CoynZzcY&0`uk^SLV@;g|=
z>_|CM(pT(5MlXc;Zrk~)Mo3+nYt<{Q5dVEs8}=qaMT5}Z<ixAMp4<ybneQw1L2}yI
za>~Z)IEQa-?8R&%Do~4^sG-6Lf~@JT((_l`b_+hv4F<shX(%ASCTlxxS!cy0Dq2z4
z2!!4TouNXhY7N)ui&s9(*#oAjr;08=LfezpRiwE;f_D?tozJMeZ>w+OcE=0}15eC@
zo89Y^l2WwZS6O3NZ|Num)su{XyREt)+p>zx0*Wt}kpq+|25?vaNi-FLF)~Jnk4${O
zi`fs`LWZ*+f&xoB%NLp#Pt95P9ta0$Da($dG<ZH^<}08}*#_6lgxQVTLTREKkCjaV
zKI``YnE%8PdD_F7lVavp<-XBT>>^1(b`;qr05<~4TguO&iNCjm?v?}hBn1pwYWSGX
zEUq1i1JGd&)(4>!{z`d!JgNXGVr#hikyRA`X?ILAjt5q{z=qaF%(9AjQq$Kqhd9>D
z5}IGogc&a5(WQAkTJ~BZGg`j>DV1$Ka<w2YGo$!;vU77GB1B*An6bvlU3VgC2)k?K
zVFa$>P4!~oz8wC6srC^|#)hC{H2SDG*EVmCyg?Maw<Kp4QLnG2zVb10OcI{#AhB8D
z_s=U14ST;(1&{ns#&n&|rQIeftfLHN<YTEi;ssD_Qopq`Q=VFf818u2a4h9cqomJJ
z2_(mFWkhz<E4CGYoF>x39ix3|VatMcJZCU~ybIW+oz+qdcnR!0L}o~9_LNmtNh4^4
z=2cg(SfkJmU6}kKUsatCI=EcVs(>C{6G3};&e_@eD`)6j#an3)#t|(6LEvB{eOu1d
zAdj)D$Gudj(z8CcZm}^M1VD!m<4VnwFl07lTaQN_8Jb8HH-4L~itS!%mZiTeOc72a
zgShy+`T8@%3yY}mHDV3G>e44t1NEdSnQn@KO+Qwkb&rAF<X_Rpk;oVya~y1#QV{Sr
z0@=NBc9pu-f^$~Jn1L5Mi49<8&2a9uFmo!jllCyK)-Txn3J?PD70p;kE`KV)4!uzx
z3`zxl25J^~^7U*gjhvU}Ju){bB79;<vHr1G7=>~|Z~?#7!dXbDpyKS2viV#o&P$n-
z^eZ1Oy1TZ5Vc%S+oD7)gy<ZILUq{g7I(9pZ5(Um{s<3Id0cpk6@lr~~3M)M@Z7(LB
zFwe%@NK%wrA)0yIzLDC_Z%d&|O!A94eN7CIQX2SGto->wS`>)Z1Ar4CA0w6BOEf?-
z<>cyG@ZNC2%5!~ip__J@_*uU7?1?~OCKjU0Tu_hcIAi_OYPka_CJS!mh|dXEGqot$
z^KHq)$y+!K1F&oEaAA{=8IZ524;n*{S6}ORJym$CSpc~ykS}iu7zDP17<2ZWs#77L
z3N=*fFUAD;+hOIbD;(%ozmIRWE304-wwu|1sSzt@{L`w6`4t?N|9uEZGRM2m`3;r=
zjN@<^<<aafB6Te01BHGYZhTvSY4_qsX~2q_99jTuGRci(EpPdsGnA^4cOghaFA$D2
zL1cRo85t1{DM1h2k>jX1xDZDOaX<w-u!JUEQikFusyT9S;{P?LxeOiBAAnG1qIONH
zi<zEI0sxi(kl2^kvZG6>;LvUt&t+VJ_P-Xf43HL7aqL^5atmqbn*WpfE`)plijst9
zJpsn<5yE}~A_|RBxtf&%6+H0CPWo}FOwpFutT~@P<^$GDo;bk_N6g;gdvgNQkGh6W
z!qpo|v*2-e{D#C$Ob2X0#?1NZ{IwL>`@}fyL~OFd0*@j~tFBBmAcT=q<kUu@Jc>4J
zeH2K33rPS>XmoODOzh-Q>;?n?!6oI-7i|XsR4NJswN7IVzDgEf6Z27}$teCM2l?1Y
z;;Tw(0h1J^B!Nf_Q_LGXixKwz*AiLZ*vVp_&_m1YZd|z^zxsnqy!Y2h9~6DjbBnI$
z0DWM)dDll}<=lqO&Fv|tvz=QS9egKBsY4z3e~>tF8v~mHOxPzzXMgKNG_$nvuSPWn
zhJgvAf0;PccT38RbBXLNC^x$w`ZRGRzvLn_PF;hvZ!LOC*^_{?eJ?=R1<8OWB^71q
zsWefuCNCbw%I1J@-N2IYC_ipi-Vd5Lv;b8K1iagyR{3`vj2mpP<4094N*y!4%nnB|
z4iB+Azc{1$99i*p1rD%2iS`M<-qG!EIJ^5u*@WBMB3S8W&sMCfCT$H?K$8~705T`6
zqBsgCl~n9<bvLVb2Bs>ZK?(qy6$g&P9~pip-(1>7n%>jONmTrGY;ctKk|;>KvT(e&
zu#rHnOua~RINq0#1?yK`u`vGn=1AA?mU{p~UHV?cNk?k-HB4Bligii!Cf2;y<n#sB
zGIl9m3<(7C(^Gmpo7laxs3e*q49r+%Ed9<&HE-ap6&Dm{gLmrwHk%>^c}hx&;Scm!
zte1xi!VedJ`b~x%*o?H)=LRt6HajlJhQ2Yh?|zS~C<o?F^A5dE!%P}kzf1wdKTx4b
z!k->d($BnD7hZhM)u29@>%o^*<{kvaTnykF!8WJkv}75)r;_do<4JdJoBiCMV-@sm
zMDD3=IafL63FMbl@xXj#eoY!WYg~BHR#5mWi4o|cb*NmwjleR*4nP+dE`3AGD|FYF
zJHwA9TS{EUyH)^q*{>b{EdZDraxFbjNU0=vJj%O%A4T?sKr*7{69Ql#k|lt{s{tj2
zsq75Rnc&48h2d*Ad=~sh*L>KD@{M0Nn4Bqtd%enRFbq8tW-8r>g+8+DD{nA{-8I(3
zL^NG*&UM&SS%ohnS8Aj)H4T6P;-{J@rD;P|)u05y${lM3#rY3LapGEx)Ro~{zEAW3
zyLtY~KM64*kXIiJl^u4V^QUd2sJd$CYi2hpEF}7MWbrbkMr#-f+2wnL%<(|<ln+9=
zmZ3(zo6X$%9(Gaeo47gn7JX#Hm7HaK{`O4J@B2c1()F?&(tN>NZWBghgWPUzLXzS5
zUxv1X@3n+^gNB&AScOX4W*6{DvEkyYxgOQ%l~v^NZp3sRUpF7XXVDD^y0O!I^gE`M
z;Z!6reQO2xy0otIUpute=uezW_f_+CaSm&iE;#WMgZAgxUX8p*!_z6sM1@4|ZgjdG
zEe*4@38zQy*6LvufHSRT2JT&cFw7o!CAK}V0I1&+mOKYI45ZKyGn{a|a)PMzK5lyd
zgCZ&a&Epq?&N`s}I(jaI+HwdMi?>T^!G04kZ{{k!^)U2Sx=^UfDfxsr79yKR9U;yw
z6qOYeH_ZFkvfE$5!gM1IK@%DXwMm_MhX@Qj_;Sz@R5uI|w%j@CQ1qNVgbiDnNT#%5
z#jjnWLV|CXUHY8Lqcb9ntm~2shyaif<Dbgzrru42coPE*m;t7WEo|EY@yAS#qeb0*
zI=^!<QbHeJS*HNQWrUg;RC}}X`SaKD`6dSdest3(-(I|lus2q>5(`FxI4UAtuLo~$
z2I3)4%)6Nh?ruIL1Y=tO(bMBL*j~=geQcJv6Xj9vCf1{DN4UPVoSnPb`67G>Cl%&M
zB6trN7iDjqDE>fbOCBN6Ep2Am^e8yKQO?XZw@<dkvT{`sNRc32tNy>`0VprF2$7xL
zmqTG9sX>aL1v##;ZSMJDG03)DEroV^vor2Q<q|+B%<%yPr4sChfh>aHOEtc|&cy`S
zeaj~A-bPg%KtUJ}qcoQnfPA|&+HmK}C5Yjqgj@epUVm|%1N$p9wkLF;izM_#U)07_
zHAxkF>MbV_0V$b~r-_VoJdL~mhgO0CWnO!o5D4A1uhC`IMLa0_=%ApD0B}tJyrlw*
zS==N8G%NrXIydNv9uO>UPLOEdH@`pv)WnB>DCY~o@b&gesxLIbO^m?t&cMH7N`eGh
zcUDaoa?BeFb@@gOeZ!rD`=Et*L{G&_L)FH?dvH>Lu#SvG*<CijP8H~@O1Ks^%oDt7
zXzQfT?SOBCUDW&??DasNwIZ+~sv!7AeDHQU(H>OR$^xu~V?*L?Sl@Hl1DCe?Woshe
zXM4ftO=RLFvW<Rh-5u>S5A2bVk$Me=wTmd%^L{2D`}IM5!~4I%f_=!(zmfl*xRlNx
z@jw2{e}3tDyK}!?G`;g)Q|-LOTzffnbp+z<r`UI6C5SeOU-EtX_V)I&qJ4p$C?qdL
zr)n^j%@oM-UF$7XECxQX{?xH<$c~39j#{B|B)r*OEO<kcN4-_-b69LM)cQM|&<7Y)
z6DdLLo#4?iZWt{vQ1v`q1;*&_oN&+I_@28#_zA{q{mz@%_deM7-h;_FjPKJpk4NBU
zYD2`VzBoA;hX$8tBtGuV`2Vy34s0L<(14qQEf`S0E{xKz5h_)){^C~X5)3iZQulE0
z_c~NY2RiG{n#3B;Ij5~3@ocICeVL8`y3Ho=_VfC$KX5#xRzTw<=YW87sZ5iixXFQ-
zk>dHK?&gQrRw@?l7h_6aQDKsHuYUN=c}qgsA5IV;94c=0Rjm@awl|2I<(Xm|fq5W&
z5s(-cfHYfRVE2P5{724W4f*c2g#haS_uui{j1P#34G|H&+_wEFr5T15AvSnGP#p{X
zNdRHj6`jg*+Yh>+V4rq;YZPovdJYYYC16$*vnHY$B|>NpJ3zSD&O|klMRk00mZka(
zNU5u_#_td<;v<tCqrL$-5u#0$2Xn|OaH@0i=vu|Y8s))2G^<k4#y6p2U5$*J@5L=1
ze?nzPFG)kkM&J@RB*|BHJ32V$G8V^MxT(!4!U9{@kCW!Ii+yZL==r&m*T6JhZTTFF
zE0xG}!|V%JH){I}!*oe$sF|vy0xyyN@2FimA@XS9b43&V<q&q`!_s4luj6&s{@Uhr
z25nc(nIqJB7uR~PxQRRr!pGcNj5*DR83=K@BqvMMk+G2AJ;kV|d_FR@&(HV1`UDkX
zfYCD)LW<laA}Ht`;H1UHf-%fZHjl}%OZbW=9olD5UiBhviHnuwb84s5Ck?g;=1fun
zy1i$N!G^Bw_k`DjzOY3Px6I(@oO7LHzkmUsd6P{B@7JS3RO~076ZUIAy;y*70Q{0y
zXQ3J2inT||>f?U^0c*T7AZg~BLy0gtqn*bRz+x~Gl-`0=?6;$dbp-4#R=b$$2ffF6
z!mROM`Rf-7GNBTT-4&~LUA|C5$D&#7&yU#V|6bo`07G?yP%bngq($-msIL0Dh4G=w
zCVvuqM5cPPujbBid)!W7dZx<rd^@(RuKg*WL=i3Y%ANf%YU;B38Y<`IFXy#|&VRdL
z=yh9{El_>%%GY(p2Ml~UTXbHQ5PFmse4LEF4DY%RZ@=p}TXoK632N$kIpKfUiEp#<
zeLN_32VNfnb?>1$IXVBiwDH~X?7OW$tqsBop8IJ}JRGWa4FjWNl-J%~@2+;@<t`Tt
zqfc7jYUR$mo|E6=sIuI)TAgQ(od4S$Hj#Z(CVXG7yyjiK=d-FVl6@bNOMiuRJ-Gt;
zyvw?$tpnREww}w&Zouz!wRr+&wDu);UQ_s7arwp%68m~hIJe8&nJUp%eHiBfHlmBn
zv5EI}{yyS@d^a)qH5c%zWIwfJi+=w`=6h@En-uM}Lk_$UyKB$!_aBAZ8w1fu=d-gc
za@qtJEw<P9!^1eiBnZ2*8)EG8<kS=?Ww2)Z3j~%zl&2s+2oxav8~t(~7O^}^dfY0q
z`o7w)sv`Q~S;o85$$qr(i+IrqV?#cQ{~`c=DwIcg<IMP|l?1<|{kac#$fL7n6r`bO
z^;G?K>O3mJX*#*FA%s<H+j|f~oUhbxBmc55IyY@Eu)`+)+AWHW9nE&go-#Ti|Kvt#
zvi_|huTFa;2_Y_k=9C8v7@*4f73EWjtD0sIm%@;7r>JQYI3pSw_gn0QFWt?b2_6b@
zEM}Xq>);Qq48b2wG^qR5ZDZU?zNB;(*9P(^X9HGio(|>no+2t%Q3?P6-dtOBTCc9?
z-B6?3%UT4Sizyl<KxQ*@lm>N+FXwKSjkME`YhBGh4f#aS^A+1$)xdz7Sbxubg|*g|
zoeGJgcH`1<5rNC^em)mrnBu0MK&&?ffRnXoNR|VZx*5@yy3h`GdY<_%=gfUjOrIY~
z6_U==?u7j%E5@8BX^f!0XD+UT4>(5K@f2uk*O|<}hCOI-f7V)reb4=-;g8}JecjXC
zyZ3wIBgoF)CSJcAc^({C!yj_y>3y4dh4rF;wIRr%I@AtMDdC2G$<upXF?By8ei2=}
zJoC-i>w2{xmU}iFT?(%{LG(;TiT?rK|II7cuD*Zr<6Vrk44(-iL4cP1Fph>G#g+tE
zRQC>16}lJ9>YXF#s#X|)SPqTO&dfS(3p99p(8IP2keZX$qb9wLgAtV6=m>G}wx&ax
zs%|_~zo-hYi8lghvd32FLS&Ar)W>-kKS}E>{BiTCYd)}>V^1iE1(fTyA9uXaosHEd
znb?1dIPP7C*LR+mdEM>%21y9MZnF#C-r)${$V*qKTXnVvITe5a#Yd-m?s$GU2a$ln
zK7+&|KXdZhB}&UY-i;&xRR!A5FHUoah_fM;L>O8tO~TDOBis*jdju9HN}8O4F|RPH
z7=K|4MW5*MUT<H$U2dw{a|92f%kg)c3SQx2VPil16uSROh*(ja<q;N7C3pE=={_$+
zzbyC-{&(CWzx+kE-|T@dZ-^K)6uj?7hB)uM>xc%v++u!P=wV3i@_nHLeA?A0+XOg?
zVou|de?QeUYx8|xxw8sw*JD}NFt9&-B=Wh}?>rA=iZRBW)9~mZ>bk)1`l6oH{&He#
z%Dtlb@2U@7?mQltgQ3;#hHCi&b@eVnggj>x+Qyviy0z?T+;@C>5hQ-QA2jDoZ@;Oq
z?VvdFm~UI_pOzCE>$qDRd9Net=Ms#E&fZ?n9B+>okKnes@J^5EhDpy`CcLAVyfzU(
zc<sCTP-VXyo61XH0AueI`D*W7JFha^uQJbeLaDMg0F~EH_Up9ou-x0V+|_SL<n}C?
z>@DZt!_|OO4<JEVXg`a<WoeIaFOTpng{k3NB3PTxx%xU6V9fOXkOy+3DUhHuxpU&7
z01ImltJN+R1(Z2KPCyUOpNK+CsQBN|GAL32x{P4RGZ`i6PbdEO+pqgamu?po{pE{J
z70fkq#yv^$EjQ2KSNI4}RuM7;J@GW-*ewj1S4~`R@V@5PXR-66i`FApU4Ol>v#q?C
z>`5kJRHoYWobIm{`5%8<A1PwP`hy>x*4dSCsHn?{;j1L#z`RfpC$s?{e%|SXClqL2
zb!#5=#+l%OqoD(169#Go9wLAVG3N9qpj@mWLfs}lOp85gvTydm7s2Qx@tn1Sh!B8c
z8_+dG(i<tI5!x=Q<fW7aHtn$0g73()UbN2CIDm+zF>e>|(H0$Xv_D61T1Nx`2fE6x
z*GY6-i;k-rga{-T=no079c1N30f8m}0A-@(AjjpdZn=)53S@S=<VoHw9*@sLj9)u&
zvW`tAkvBuDp2_0<mBwJQcAq`hITO{V2S!uWrvDAIC7f&pBU68i!0?~VRUHZ*ej-+x
z>ssp%Qk9ge;rfp>*ZM*dh}`xa(*G^H;|^*1i1w`EtYOZ^mT6jImNqHSYDD;3FgSw6
zkBL1#<_9vA2%{Ywln6W&HwuXzL3|JFdh&*7s3SECd`tkA$mj0R?jLP6#+tPt8NBb_
zPnDLLLbs!8<YSU-wV;uy&MM2}6L(kjr<<$}VCc{O!^<VdtIr5*vyB66blNM+%OnmV
zzfD8qIq`V^eNU!!xl}eL-cM$*117mg-1=*O+u}@%clT=>U#HxDoA38mTwmdXmNO`K
z4;QMBJLT8vQ}E~pnSUb<>iu#Tr8Op`n{9su-$FMsbN7fKiV}!g6~fzw&m4`8x~$aa
z2fSv;0P{m>sJXE$%a9O4DWIo)zELINSliKe3wP%V+1A*gLdr>IcLWuVDI?dDAvKZ*
za9-~d4p-(K(+bN^EJrkm5}83YOIMpiPv}TU9rA<^+4QOLfkLJkTZ)JcBlhw<g+}^a
zUwtJVCHXEov%#%bBOpjkny$lXZj0&R*I)wwxUG_)@@TOFNh<4Q{hm&hKW2-m$tq|N
zkB(v94BcW(MNIqPw2FnSX$_(?u}@e8U!%7;;_O|}qhB1vkEmczd8NqJ1q}@iuR<5H
zU$2YHbvis=>rR%N9JeF6Sh>0XyepI)Cj{=Q&8N5S60MgTFA}S~jD0QzlSeMn=G%j~
zquf`IbOzJcpG`X7Ud|?}gs4`}xqPp<tX5l{U+*f9YMqZ%Dmxz!p`UD<mfIi3Y>TR@
zqVa9p?fryATeS8Xs=OJb;?5uEyRzP^6A7h72rY*pigy%1XU2AaNX~~fYu|p=dH&)+
z&NodYw|e8M%KOCfL;JyZTMclIUHkS@m#<Lv+w;D>w0-Yqu2EH5x<WVLyT_w_uE`s1
z*E|2YdKcLZB=>81o-|yIsOoy(cwvqhlxsFcMSC9`-gi!jV{41mE%IZEj&grbIUYDp
zyI7ot`?DDKxq$%Rf~Xf7u`p*gT(S5`jR`MCT3?d6@2er5ABw-Ta<atdy@jRh^x4%>
zd3y#27vANL%0qc)Oa+ml*oY)Tf<%i)A~Oj@g-Q`Pev0$$89N&uC4O;%v%E<Ca>8Ot
zQt2Nb3dvusjs@Bat|H0f+CC_+?ubSL^HNT`{9!b^zt0xO5{$4ubbnNA{|7yyd8e;O
zat5|%DDu20M=H-1N>;&h>VXoPTR^ZP1(R8CuFoU2SeN&9D1Kz~>8sDS!S4$Kv5@3o
zQm*h}+D)sae~md@S#9s?{SM2!!a1ieXSj<=RSZ4c-LWyB)KvM=^$8KH{R1{Z|5Dh<
zoFD%(%vHv6wqf(XjKA<7WDpJ?>NU&?cEBU*7zF*TVS|!GCgte%IfPS4=&iZ3g>ohk
zmq~gB33#|f!$w>tt@x9w)!5%pA+C3={KeRI2M5_zw8CTy4!)+vwK>hHgk19-2f=5e
z?}%*CcH4F9=a|(s;r<mUk845)<|UE^g&~oPn@%<kZ;Qiq1DBKZO0a*sk3++!q#fSv
zKepG(r|Evez??g(CW*CfWK<v)YSNBg(}l^h8fhMsW)fo;-4G>*?ekNJ0XV@J+JkKO
zv;3<FP)fbYDZOnB+2tYZU4!P=AD<*ZuKOror=Qkzr|%CvWe*-I62gk9`Ql^~{)SHl
zRG^K@vGO4hl!lPStE?LQ8wiq3U_L3XvElkCLu*BYQn#v^#0)Gc&Kq~&wz9EDU@<sE
z6(rShR|#=dF?CrpPwaKB#GxT%71|{hU=u&^9|fhQl90}U?Jye+x_zLSmF)hBOKHT|
zM=&Pzsw=x3FfFVIu5CXGFq~!{k``>#@Co*BpWWm1&{txX4xAO-aWM&V-B$f`GlhhN
zPM^h0&h_u@H0fJ2`QyeF*+S_(2JYT^&gE}t@2|wo&YTv!`~Fb>9H49*+<+MztH7C?
zUIRsULdK5VbvNYN^`!kl=e9PPLGpB5zH0t`+AQ2}uFi`{in}YIQ*isT<rt;faj5*^
zc>`7`XWp~=-o(fV>W8LHi?RCl^Xpw#^jzBeNG+MPv(6m~rjH8dA^q>INmD^;tFgM4
z`)=X~uOrQM-mR<m4Fq-r*fe(Gap$#apKhvrUm$S9co^Hyd&ohA<Lu-2P-nne&41Vh
zb{B@$@!YR!OdgZrAI`nAC%pF9s9f!Ei$905@c%_vT$kk_rmIYw9~)Z^%QN0=iPK5=
zduxd1ZRF3G!eRYKJ&c=$I<>J^nLnHZGx;+|bnYUQLW)_C%y>I%S;`yiRfPtdj|iU{
z&OKgV*e5uPEOB3*tv@XmTQid6(w~y$ULdh=ZFu(t*O*_LtftxyuB?oVQ*4o6=4zu>
zS8iE7ZpKE4k7<gSY-<`1^)=(X$5Mib*|h3QJ<wEXn8CLg6mt4BQfT-{E@UoTA}$k7
zssF?RmPi6fCt+w(XOm{R6{ao?Z<|DXc-D+v$@Og6j0TN<Gf}m!*jJ0Px$US!V>3P(
z26$RIQl^&0_}&r|I(8`b3wN71>hkKx*I}G6>r+Zn@vA6nao+AOs9i;sgq@93>L5z6
zD563jH0EI+lA{%jP)r7i_10V#OT*QhOg>Fq)c)G4E!tvDaZULrcARm$IQ^-`cFFV?
zHk(4dx}rS}ZB|C((4&=t-V-?MgdDmt8iU8T<TrRykYNm0j;i*ak{puKlckCH*#5~a
zI+IPXf;o3n;+xY{UpA-7mk4Mw4x%4QF|%+1V9;+&!+B<uQ@%ZxpTxSN)mK6Bjv`dR
z#-Wm^9Tbi<?UP2$0oSVk7;$O%Q<%Jv1g2*ZREFjck}{kTvqX!>tH&fjl#u4vjID#F
zm@m>RM&GIU19ni5%JpmSA7~v+3**zYZeyoVm3P2c6-FoG<XMU`@R0x3jmi%By}O2*
zXzmLcl#(MSaD5FuZolvczPTv&whYPk8F3jH?4MY5?xO$T{$=TJLs)Q==vo1+Aw)|$
z#`Grv&RKy&9z@kbZAD8uB<w22tBmg)jmX7l4vsJt`=56fm$FU2_m6#=V@>MA<8>83
zD(en@dvRp!eq7lXbRBYLY}r90nGMzs3ukOKn_O?J90|k*=F<8>A`h#!-}f^OG+M8n
z2XT$2x1222nM#Q^wS=t;y>1H)*k_3Q?!^tbuGkOeJuS39{9C`va9`45b&c|Vx*Z!)
zCIAwl+PBl5*51(sMxLyFp|=MK!ZUK<9(1>vPGmjo8q{!idHtjB&1Dky_V&IS(S12o
zmAm}MDvOQhBhnkKXkFRsM0xa3AK8L#rbRI+@|Sy5GKbHSRo>YOp|0WqT4Iwhsm2@{
z{)E|?Gph}v5W_7=Bpnu=OI&j@oIUej_qLfKTprYf!U(Htb)jDK@s>uMhwSEfV;T8x
z7SyiE=i>ZS;)(pf?d3Pd{F-Ry_glZqnsaZ<^KGVe-ke+n4z%fzTIDES+|wgDqVpK+
z8LATt@bT+WcaG~j{>sPITE;3u%qM-qunI!tB&V=d=~DZE#-LZf!>ZweL(7XnFr>Fu
zhz?p%sIJ_q?L4flOrpFw%Z6{Z9WHjwfbR}QoNdawq!aR-0aaZOE9t!4l^p(~Ay}*_
z-KQtwh}nM5t#AGJoD1=OiV3BdYWHg^35w$XWXXxj6{rXhh8#?HZAknbM{p76$t`bC
zjUl6gPK*qprt7@Ox=~|Hl04joVKt|$)p+fm!S%G|Tx)Nl3srYJx*Olmg=@;Uof+au
z>3M7c)fcKyVk<wUP7UQZcTFaMuS#g*LhA-D!1@U>oHuyo`PO-C$o0CNhVIEqg<?2`
zW9mI?T(GmH4pE_avP2j!7^&l5iz$s)@KJ__Uq(+8LZ<Kz#r(J4UpT9em`sUZ;^2xr
zQGF8rB5|xzch`aR%i&c~rbZ90kVY+Po`RJwDe-!6OdukQp*Z9icqc9BLK;XlOjPlI
zJ27o#9sGDI@@N*Rwx$M-A^wSsdj$U$%Ux<XIGdNHuxsH7%Ev2C;J>5AoiL!f92^e3
z1um{|t1wdw_$a?e!Joih=<(8)geVr=nAlA#4}?5BBaS@<H|&mcP?U-LH(6PTzw7H=
zJG_pVMd;}0GL?$7iAm2^PVC<Eyh=YCb*kzd_hl=aw-7pkj`IqDlM)&%=M|_i-1hm_
zhq`~lVjmka@{~7FhUJO-#(_G7CSIQG2=*m2Z$uYco$1KQ$ye-pKO;uHA-|m?e*jLJ
zGj4<QZNmAPnVILiY79t7$p2dnp3UQf1dh1Al9{)A+}Hj8y563;<ko((%F=nri!e!3
z(;JW}O)Qs}wsK47NyS$yu7*hBlUbYH!?Se+#8s#%+narb_vT4|&2a9sa!N2Wu#Pfw
zzNPnLv71Pk@KgB{k^T1*`j)8w+>p(;A>|nd3-v^65#d$Eaz<XLnxaFg%SHp`p-qE^
z^gJ8l2I);O(ex8frerb0Y3F9|Pv21v%eMf5wIJxD{su=g{wL9hu8sJB9tvJY!%D@U
z6;MaNkThX06IR$Cw6Ew7dD6p*nav`we&~OnHO$c?hgwk4D2=NFML~~|<mu5d;tkbV
zGY?8mc4vA^;Z<-4@LGEZu%mK=?WjGmV3$arq6|`8Teoc7VQ!k9Kb%c6!90Cg<=fo3
z@Y&@m?Ye<oz|Xp8j<PeSrzzQ=3VY-*4&Kaj><bvr<(4+kbkLc0rFjN*CmoFJU4R`h
zC{!oml3qPS8*<Uj!^UG_D5(5B#`jh_39vt7F9<%>wJ=HakLa|^sse|^wiRUvhi4k;
zQyQr#bh@9vc()YQPI#0EsiYIF99RaApRqlZDp<(SnKj8o`16&G5xkv_9oH>d#Y1?m
zTpfarwO)Cq%jebluCeb=%cJ)!knX{;i~HKXYek3D{P>`t3<he?Ij>3K6r4Oz=0d$m
zq#;GAyyhC#4x!YL3?ht%zD2u9ed_mY;*LvamLLR9%V9=Q=;W7*5P`@akS8G*Hh+wm
zdgT3cx3(y)V<`QR6SKInYB-P0i<Uavm-GJ7s3a1qs1m5i5x3=kPMLTMYV&DrznKu(
z0~K2Egx8i~CAtp(a%NKrIFWyDaGiJ^oMa+09htTt=kp+)D%E4`=1%^n9g&)ftG5aD
zyh(3t1%pia1vwg4wL+u6VO~e3YJMBL>mEA$zK@Fgto`c#kAz^)zkk$stFG(MtqmHV
ze;ie|AsLR!`<`~ay^0UkB=Gm<W_O)f+d^&-zp-%qW*Zn)y&jihAQ--gvb1SCI5?mf
zbgU0}O#O{(Tia+pUGyMSP6Bt{&Q5B-8Hhcu`$<tz@oJVS*dm3#$TiO1gjILF=FiJe
zlK*GrXyzcoc<q9}W>9zEZ>N5#$&e-L)hAGXKuj`fHey<uvXTmo^XHx%_2xaH`u2o+
z!n~<RW#lWfv1ktNm`I#K<UA4H9t?6;zJtsBo6n+TSF(d40!LP%NRIHaVQ%;u+wuYf
zTEd^RUi5nk$gAp^-bc`lhWyu9zKk|ehdskDekfrK5Or$C?PF>~##la#!q~Q$>(Da7
zPyGTj^u8eAc6p5Fb?Hqbp?TyAfBa%)CXx2bkQ4-G7$d1g!GqIZ?;nt6ib{D9AgrN%
zx$#o4SBPgc?IV`ZOjS>IrIhnHk^3ChRY(8}QuGvtQ}X^?va6dO4fWn{Qx|!l=F6D%
z28MP=f6g!5mTRAL#PW7e&r~d>7ZGSWpUlANeN{=p>&_%mY%K({b_}q|`u_K5o8;lM
z%!ZeE%(c50aR#+xz)_$o)UvFN;cI5^&~#6jl(8wE)^eUtPL(|{Jx#*J`Y6oEbcrzq
zo#RtoIuW6LDSt1x+S4B)3RXcuBy-=K3@Le8u$aNslI-;_IEY_n>n8E=zK9fj&ccwS
z2yuph+p#*P4$NUc!v46S&RqF92}6?R-`BKO(@k|cPtGxSWeja`4%0~l!=43ZExW(G
znF-pO&dKx^X3#8{6M1?}XFi<URNwB<E*sM#M~K=fl@>H67hd)-22IP-Vq@+3cZl8)
zu6G!D*Pj;RjutE#^PVre1Q<+E7@IdBJ^<L<nO#4oDt+ttUFV&^2cnmM5=<@G?{&ZY
z=XK^!EeAzT1{Q8^Zj;sm6We*=j9c%o!*o?U+r)3&08{Dv8&>H4i{Q(F9OU`Jpdgmy
z*tpK)R5Ih$vC8;}&MoWOWm1#ftg@>8ZlbEgex_@!@0(B7<4iH5#_s>tJHGe5#FstT
zj5<T!>|@((er$q0zEkm8>_DyaY1lYfo3rb=m-yAr=i$8~dcNEyx@(T-Z;hCSR+S>4
zxJWjOJsr`sI(1xF*{^9dU{kC9H&V>-%p+6V`mgraac#}1y)47*vNk;l19zN%K47Ow
z{G-HVV#~In$7hnGS63Nlnxi*;`FVOZz_SR#W?`9<CCC{wIMUfx{qmFJ0<Cjn#U4gw
zp&4{3X@5a^`M;-&z+mD&5FA2fUYz5<fHY=QcH_jFp07R6KTnV<m?>?C*lE1d6df4m
zloW&azA}l#SoDr5D^GM%6=F^Tdy8pq<kt_g8Cx0BQGEqWk{`Y#bW*n@eVa#@Xca^J
z$fLO>VKQWSit4wU2g8*M!<l72r2V6_{k!fO8H@jgYzR+z4A-|$Pa0onSLL0eG}%<7
zx*vay9M+05O&w>!ryTaPKhJ1;%sUBB%12;We;rAMm-;FVw9FY}PGsMyW+dueWhknD
zAhrQ%>Qrw(ieb1IB~Kx&LeIA2r_c7n)>OaXfS5~@)?X1(c7utwa5Hewteb6@_FumW
z^GV_`@K}|!L?jTP8KxZf)%&N|#95%;?K1$WA~*rx_(m{-Y7ON_789%yHR1C>q|%>6
z@c1<qoQN{qtR(RCr(`LuKP}<>YUww|ND|c@H#icGLJXcdjist&xJsEMqflj_x#8HR
z0_}A-Tk$yc1lXj29i~&%;H3&auS!lkCE4WbL=-f;iBOP}Z-Ypo0{1(D56^r)dajDZ
z)2rj=-dp;LoZP)(wm{GX#}CzIC90t60sv4%WF+B*d;a7`)$3u^2e-AWtfn2GcQ%~q
zSN2mNlh?-|Q~=1nb2WG?@zHI7E$icNV3QVjYm;H>aPrBxTk|}ty>;Y|h=_Q9xRajF
zy0QuRT+SCKe6W5rb>38jSNZ})JKx|t?-!1Tu<pK}36wDhvwn_wJTG-j$gwR#j6dGN
zFQ13Y5vFWipLgu6>v^Q>yu#Mo@=MVh0EeBF;u;O~%J$b$x5N#U<Mta|V=KJW3|3U0
z=UIDey_Wds)Hg)$+m>Fej45YT7xgvz@$b*HQ_vlUMdfAHRTt>_vh4tyb}iwgt|YpY
zB`p?*9E+Ng74Ef~>dha`usnuE-WXk%#2$agkwpyCCPXw*F81fCB}wWPWAk$njm66#
zNY%|?!A$C0^Z~Ubl9?dk;;)2`rYq-$c~FP$!M|J`B2uHo=_D7X;hJ5C_t$8$y({eG
zJC><gj#G{&|Cy6Peu<-8#rGqfi%C(3>@@lFY65g&Chho}gLtZi*-hvIwmT4|u|pZB
z#!RekNqu$6lFy+^+l<@SVhG&Ddo@MJF*pkx8Zp!VUkfm8*dEUtab_8KYz=e0Yps&3
z4TFeWL0a^Wy7AM_f|@HX^@1MB4BlW|d0A1sMEawH1WvbcMu*YG_kqQ4d0#AlI{ztX
z5<#Jgxf*<^yFlJZ)f+nbe11E}!Oz%?`*Sn-w1vHYN+URUTI*>hnDrTbpj#F|VMQ!^
z&4vc#&3Qe`30YTV)U=%!)J9^PC_2>G7Q^kW{3iWPxVmdF8vHY5Ef7UZIHVB>WJsZ#
z6=It}iX9=TIz38UJ<;s$*q2dcf8l$@)0KCFZQ^#-<Kog?ddsUk0D#csIRNq`#wI1L
z##waiRYj651VDil94n1ag;;_X3Ml_jNW?Zl^A&sFa)|n<dq39>nkR&7bv@s(v9ih%
zjL*rGbX=rg`)=4A+P)w^uDu{Wt~;_ln7(X695e@~$Ht=ST!g89I4F2zXj<`~!!!Kk
zy6TKA=hlM`2s-xp_jxirPp2pyx_~y-?O{ZB{QjOhM&;CJ6H#|Vxyl3c13(b_w&=Ye
zd;b&lI(5u8JX`DNv)2F@HJIF;u0F4wy%2qv@VTwqpRZCWdnbB7>OI|_XaT>YUIW>M
z9ROA{UDt*r1TRr5yIvoIuQNx7*tNWGW`U)V&22y*?9jfg(O`j@tn0%gk8x2zB@hZM
zCB29*mSdl}R(>mHdwB4-$x)iNi}l85Kkq#3zJo#S8_-N2^z7_n?f@3uYXg!vXZd9O
z%_r&|mrn0`NH2CO$3-TNLi>bG=n<=L+iO~Iyg`(Qi%Ag#l!3W<y+LV%t%nz?lNDti
z)Dq|07ch5nuo&8*B+eW*uq+C2ji2|_wOxiZ<2oun%*TNS^qANpIE1;rg*0rw9uVQ3
zCy2;I3bR&db?Q7h>#*Mx8b?2=kNFmpN<V*JKeDq9WdLO|7nv>Zm^w?FKF5%7MkHC!
zDIF_jMT9lzyOIM(Odf6@RL#d&3xatEPM}hrtz&{5_&si(z4|$u*>5lk1LsKWn=5(R
z7mHD}KKKE-2DR-45(c}#f(;EZcqftXbmufJoU4n$C6@43G^vLJS4^m<WRHeE1*QAe
z{B)KfLYYI$aaS;x7=-=x#1m<^!i$OJw0xg1tzZyuS#FAxT<QChUcEGoF$$Mr8__T_
z>w7o8QA#S%pi@9)QOE_KwoYf!cMfr<*BC9mpir1})(z>wHEg<X>lWnjNv!&uVqFc~
zu3rc}zJ~R|32L@ya5V}#jv88;PkN175*^a;gGS5Ymm;#6PX@~?+^+7%WK;g0qeGR>
zfG1<2>_xg8N?OBGh2vc@l2ATjSy^QD$>r=0Iu(31Lut$bVk~H=c_EhLfP*Pv6fc_x
z%bTZv$ssE8hF}-nDnZ!+bN_btqWnjx&1eKMI{r@zC~;MQK`Ce8ql^1mo7k7tVM3b)
znlle{Ea2#VOup>v$R3Pr8(fh)VyQCvZf=}uhl`6#g_T-Gdeso@i6gO_W?88}7)AKr
zWm>IwzJyFPSub-|Nwr}B3;~C&_mSqq7Iv3gpXr(J+f&Q^Ryg+aX=U5#)+*o*K|LC;
z0jh5DUnc9Xw@qw<g5!XgV={jr7$8DNxX834TBLX7d@kcZXxaB;?l`ni%nPmAd+)Jj
zZ|<er?%gii3X!>Yx?h>-yj~_>)kHHe++0ZRkNOyb#$KN~qtBkMWW62R0WxJDxJDk4
zl*idBAC7nG7fdVDp7A{*i2&$V9uYX7(olh(3+Vei(@8zkK4Qw%v>sQtKL6F26p*QP
zZtepQbt{<f9w{*u5f1De&z$RP_ZJ@1(o<0LhS?-E#t<S3l+C)2_;EH7{JY#<Ww7D-
zYut{mFe3Blz!^JN-y2{_r&>L|XgBITuKxp@v$N;Xh$$+rc~}qSJ~S5-#QztVhRr!6
zFn%(k=puLW9Ls69UOxdm%QdoM<0_w$34+2n8xxMrq2Bovfq0ecZ%^_Xk@&bOW1XP#
zs#dGbk6DdZcsecHzL$-V#9LDx*@ab}LeO%|;Ckv-hdE|4>!XZY4t=ALh8%25EsD_k
zJz>5AAyby-5G=>(=$eFrtrAQr8zX*xfIb7Nx=pa(6)b>eXr<^9(ehG(T0E(tC0M|c
z&h1A$M-j58vMz?F1ZVbFubf=O|Lm^(keVW07uOL?ok46#WZ|}~<0jITc}^iam<5<=
zp`2PP&y%%xj`~j6WvP8NAYIz7fb1Pc!ABU8^SbbLIrE*2ZNf);L&=8Vvoy)b85^&h
zF$O~h6AzZ)!=YP)NMcKbIw@5esI2^w=CB)%=o`#n4&(IPT2ZbkY6_axxG>YneHsgK
ze41`YOl7&+&ePi5t=iH(`%OE)!Ai!OQl`PiZV2p1jOf3SLh=dP{d2Db(RiaZ&dgkG
z8T;~(?v1zmQ8D}6SmHoEHEmghk84$WZuLdLbh_Ts%Y5YN@7BMG<?(aF4P4vaF@q+c
zS~1*kCz{vw70?BqxR&cF85#@(x2(bG>{lnjyOs(2t1+&LClG!A+3V?<!?JbL#WEnw
zy}r5G<VH%?cENpn$bK901kio*xwjoEp`rQCTMi$9*bjCb=YcK6n6*fhVQdMotm;Tw
zw(WWeSpdUz&#LN<0hQC(^4cxudt;HHkRVOrv#aV8_|6(U13zUm71+>49|gnuxBMCF
z=&IT;OOA-on#20Zn$|q%d@rzlyOS6+!O}IrvV7M;H~<$}2ki5~PDI6gJt@7)&=Z37
za|HZuG0dNE@$h<_*&=(<zcelP<FNUn1KMGz?x_=pVqkj}W7^Z>D-h89SSFTcbi$<e
z3F~-!ZZV^`(V8|9;jP0^i0e4DC`)F!CQ<rkG%QL3{E&ucg2RYITTrnq6LVM%eQX=}
z&4pY`#-SR$#;|r!jjAUa4Jd?EQ0g3)iZKTg;we>6d|{6~M&=tIv|cd1%r{*Bw8fFI
zTAfZ@{M+OQ5Kxc*q3Nols(hlYAV>&EcQ?}A-HmigT>8==At2HXa_R2wZlsa!?vU=3
zejmU0-u%g0DC?U!XU^GspFL~l28+htt39;|iE%@Jov!9(T37q;p3S2Fs-V<~pJtp}
zePAucvBHY@W}U*N@5EEdJP?;jop6{TxfT;bnfKIw7yJAEijb+H1ntRkpUMl=Ll7Du
zMm`}``<rNMF>DgjV>xMDXUeXwpoi!gW1eSO+ps&9I*JFf{D2U<n#8G*c=g~Mu9>Nw
zAX%?#W^QiY(Y|Zbf7N#Xu4wvzfhC+zFS%6;Kk~yDOu>~N9oGKNQ>jj%a?_@fK=1`2
zm>?VH91Jfb;n;E!N?z`&p!}8!Kpmc=<c}TPN{t;D89|!0#*$GyMX@CkNIf_@%3VIm
zxPAn%G&|u1vv~!P^)i$rQ|_(hJB9NEAJ46@+na*TXeVVWe*DOMjJ2s}&mT9P-(+gf
z@0aN#oaQms#G5XX^;y(;%()X)t=A^`k$NO=-X+S%S-Otq1;S0YC-vsB_KhXK6ISAB
z90zGE<{HVreZ|yUx+O$=&7316nw+hIf@&Zq-bLIkaBe$i5yem<)aUdVtw|x^p7nD)
zC#6x)@ni3O+H0~b3IHQG0qhC&HF{4_$*{-!W~_le08qdFF4(dP(i*M^&@Sxw(4V)e
z9;_UNzmd3YaGHvebSTB)7m!na9vh%!iF*dq^!TG&-tCF`!?WPWMl6WGo)3oifhVt9
zo|0WM>uEJl%>18ZiT1Apch49<A_dT6R8=brTWY>wP9C`IW(r1QQd3hCpOO*cgeUu$
znWmklU7#Zmi^?_UJsR~d2BNb3YQ+7c_ph2-PBn2Oc9QM5Xr`+vp*9MiHYk+LLOu#h
zUT7k0VHs&Z00CTT<yX5oupp;$p<a?YheXR8=2yBeiJ+*nYtbC86}KTsvm$ANH2QT!
zq{~=7bD;UzgQXaBH6|sJg{=4BJgjsnq#of5Xi80)&27ybiFo7@KZK(IM`+MXAqlrz
zJGK_Kn{wvmuiT1fNA3O#Y>*+*qqM${->ciGCjzGgpY25{4@CL~M+9GLn0W7I&q3X~
zB>~?wuqi>r2p-+PkP=+N1avD{_1Na7X|fUDrp7`-mHx;K8aPyjd|L`drNWm-<eE_;
zQLLB7*YS#VEHI%Td`OYDG4i3$!=jd`HnwLcM?*Bj=HAndSWmhqC%GQ$(<bH}|9jM<
zRrC0?Zall-qk43b`)fQ`dW=T@igGE`M3MDIkw*jfeXBY2sXffU81p96ad}~?55Rp;
zQ<;qKTTxjMYm$H|p}RfG1%9Y5n7mr#+-6@3O}YU2X;V;<QsO3^_QweT%yXZ^WEXaj
zdC`H?#sGrkym1K*#ZL{%Q~(Ym!x2n_f6}h4jHm8U+t&33BYdtf$-B;q@P9WqFW{$L
z>sc`HqB+9L>1#n3H)FYtTIPekzH*K=H?|^Sdx^c3rKqhh>A0Ma6jfm+J}YQExC&%#
zU}22nWp1>ecYnO_CIKPwyB@nY1;<{%ch^!5bL<K-$yMa-&;lGm^VF6GX*CycX#?Qv
z0WAdXM;YEx0Dz<(m00yzCMC`A@+x$88ciDe(3|@3Q_3Lqi*Qd#Tnv@G#)V!`C{AjW
zRqy~MS^pB5s>qsZK<K8g)X!2Fy@#tp33+xIj&#~C&8?!)^EN6TRIsjbgr`in)9D;D
zWAaKMT>V4F#6KuuM#1Ex(!`C$;<<`an7;OT=L{dX($$b`4X|V5*KW&KUF(cKb>{A9
z<fWgvS%>-Yp@9s|snzgba*`1^(pf|a{25^Za?i1xT`G}lF6epm*M-tZN=>l0c4;XU
zZ0Ru59q{+|8g-3Pi(`|$UDvK98GQ*IgdXy2Cj%?%@cVrZ(d{veo2vtBT@1>%K0_o#
zO<}?$-k%!OM_fRsnp7#^7m|kSzqIp$kcEj0j3Bu=zr4+2qoruxh0y)L6wdg-8WAil
zHvKvPas-h6uF&FfY!rnX^e`0QD?aLx{V)xw%ig6kf%VK)pLQ*>0dcrDm=tYPRhXz<
zGn}R)75$wy39fdP>594b;HpnEv7M#xA?*K6lcM>{;Y|JW=xiha$U!4v7>tV2S-~I>
zSg*B3NQ}vSySd{K*ERm+OB#U@2PtFXS9ivil%6E^z)$~2r{H1|R$;nqc~L|?1XQoU
zDtvF@<>v;Jtplr<=`oHBNLGR6$%?&2STH}lJ%T&@Btf6eYX~M24QYS>(yY$^O~Sw<
zjr^J5huCy<F#uo^{)9|~B4Mm1K}uNwOnH!wgOTEwSk1uK+8pOK+j1zVQhAh8dg63#
zQ0UC$cQG1u{lU6VX5%K5A;oDU4EGG7T$ZD(rD=hvw8<4`Jm~X~Yx16raiRKSP`mqX
zChqz_FB>Yn^ya|5*W>p{_pN)v_U}(B+c3^p6Kfx+Uc+bO*+k1&e+uep#Ku(P0!c+M
zazm7zJ2nL_o-n+lt?wblO=|yz(?7s9k8a^IWapGnOk}jy@*gXEJCFQ1lh}V+58XCB
zxGNxoQ%)jb629cR;_3A#WNyGV^}eDBAATJNo)5yXbkEw1Z4}`1H7`m1RSOILjFL+g
z9J3{UWui?7+#UyC6<*1MGH%)wd!i9?>cWc;`+%#hB;yU!;Qb5SsBDwSi$X638LT(R
zQYnqbeL}Hgq{4JUF=xoH*rHaL!dyU)!u{c-un!iv+hquVJ#rQt?>r7zK496zrA~P)
z@OI^%I<3o2Y3vMZ+AJJbXK~jw79u;*fhc~9wktZIUx1&oYj4YSFbCD~PV-PfOUZE^
zmFPHe1FRRt_Vr$b^10nYl`sEk(|eiv5~VskrT*kVsD6GlpwKTT?Y9(Xv``z=0jN{N
z+AiZb%YHsIwb~OSU4O9b=dVjB4QaQGbVlBNRB}_+4UZPkVNnu;bq@_h)_}Zh5;7`d
zWp!fkQ-z?rS4D8dfqc@+th8Uni9sq1@-<Bpu3u_8l8U>S22I(U(99&^YW~9TZ(()A
zaZNGvcz+5zm?r@p+#^9#4!_2EQrmHu0fDJ3r%a%NZ<S8O<o*DGA<Qq@!_*khVdV_s
ztbRId)HpFD&tU==*!}fcxt*inc-TGnd5P|QJj=Qbs@(L*$4Xc3UmTF&-s-RdV5DC;
zwqM5eWMY&^6>rG#MHP}P8?N_AF8aN?%AkH(5IqYB@RxTWvNzUC0)<CJ1SXWaK)420
z^h?@kIrwX5d!!F2*JnQL_R4Q={(kyU?awk~vnIG4Q=BTb{5|hwLtaL$Qu6goHgJ>5
za5PJvCO2S~F?o_g^Qj>KcLSmP7Og0|%);Wi6-R2@L&jZo`@NwE{t@8hF3?%2=p86;
z-Z3~Z|Af$}o|&T535x~*O*W^)0K}ZUFJ%Q?zcVt>RrnrsGQ+@qCUY1rW1e?WXw*hj
zyZ&)e`wrWS3jq8hbf(sGFsxuyhL9Wq42O-4m@G|5`5?CUqv5~q`@afJLfe)c{z<PZ
zvl~ZlidU<i7XLOYKD^d9j=G(tZvJ-VZuTzg>Cg-zAvn!IYy#&D2x15VqC$Y$06?PC
zjD5#F_r&7UOVdw?Q6}<+QHtKgJ^6KP6mwSjCoLsC{VBu!Ep9fE?4V!J*JMmXp+VaI
zBEKwsF()FY<kn^UE$G6cOr>V;#qRTm<xXFku1iO*4LgH=L|+9veT1~0XN}o1&nes^
zoLJuDyMGF-8)8+obE9bA`5sA{y&(sXEQfR-5V7x@s<*abV_Tb_#_pJsSJElgeBe%e
zo-^s!eEJt9lW=cGU<4+R@ILb(VNTuZYUQ&rl@6oG!2`M&V#}aUcMc?MVk~;k-;P_O
zNZW(W2xZ@bo*{2XCMXQMPe1DS+BKRhFIcPYIbn@xL3J{*mbdU!Wi-D=4S-W7Ay-wD
zZpKVx^a*LbgZ#zB7l&az4wKniv;y)h`mV(bkomRO&HEiDLXW*~b>pe7zoVU7+_EAI
zaIAsg0+M_yO~Jv+z4tY9NAFsuS9<Z_MXXy6ect9k9kd|x?<!m~tu2}$JYjMn(;5d-
z`$<DUzm-yK=hf86w5q~B7lr$IDT0A4HzDS2Ikq_V)3_U6rOb~o`#c{R`rt}K9QoL1
z!~Y$;C;HJ+*)xVmGM-SzRG4BQhFI_%klNp}!oRp(=&83Ne1busZFr}x<_O?{iz-)u
z7I8K5XL{sDQE0#T;38zyn-^$L9J*98J_aG$5d3lMGRl~Z|BI)+V=FnP>Cz1sO$S38
zGlb29eQfpb*tS4pfc=g1c7Ybdt)MgLy8HP1`wVF1)Nb2edOgg7Zca;}jH(1L?`Iom
zQ~JX1A`O6xtm1F9X((l)1ng+!fMWOhFXM-BD>iTMS&+M>*#`+Jc;eM3);@O+g!nb&
zOz}pPihW!>3ohdvpru}9O;&<`wlPfgwD3Y~&{v-Z+Cihq0s<>)Pc-ApKNXgvI_TpN
zmnc@K83YbqJ>uFa;1xUIN@v5KGIY?^`xc`*Raf%#TCY`U`IvS(wI};<iSizB5jCnI
zV4dcFFg+dY*C<X?L}eyY<5g90WoSUzb-yk_zey8cOc01o<?*?M8%R@DquvZ__X&hX
zhj}c}jaWJVlJRX-&3){qa?g+Cclc7Gq2l%C)#9sLB_NVf-si1|lPK)3S7l*VY`_(R
z*M~KRq->8F(+~1$B-Udkel2T0F-qj`Z(ZSC``b{B^0-V^I>ZJ3%!I`p3qSM+Snj3{
zz;1KfEOZtM#UFw@ZtE+}0F-oKfqdJBLILS5Jb4x<6gb{wj7&>$D|LUtO-2~tNN79a
zBd2-O=w8d9kGFnVs_;eM!vLDyh~C7IU3-$pDIrhT9>_4odnXhnplGQ+`M)1p9L4N!
zVLNgCYbX(A3B~c}@%8!R*{cGwQYiy;F$mmWJNfTrPChMyBbgTN;xzWuZnUl{IJMY8
zP7-|Cc*K`AAf!+(LRA3;h-IGoLZM1!Sw<~@dg6`n198ba5&EW#_|g4-^e0e(9fl3J
zgmE)x8`F0c%;vH<z0^cI+i!Y}*{|0<`Qj-yZA;S&rbn>H4cd*IfSj2jf<JgSotL|q
z_^G|9ThPiKF7gWBg;nGbSS`tyUCCtc6x@}P&bcHe!GxjrDEv~nz`S{yq_Q5<d=r|1
z$W%KmC-W`kfne~z{S(%Z>GENa7n_%+JYjShU3ks~Cz(Si9;lD7=1sO}SB=YAlINKc
zaa+FlS;)_=+>IX~4kzKPOthw&ToFTxdQwk+uq!V=yA_#ck2SL3e;XENeZfimH6j22
zTK!8C;Wr4GNtoo`OesC`_BdU?T&szQV-ocHj99%pu5$v~BYy&zES_*Y7yUw5w#n_&
z;D5Wk<IoR2KVAQwIn2}=;nsd+@U)oujCV9wzI|fv!_Vw7D36PW*Hshsf)x=_#=tY;
zzT;ShQPYZM5>b#+8^yv}6PE?-=US*$Faom4Tzn$e0-!c6w(C_u#h@~m#mL40G~}gn
zDDAYyhV*hoaYU+D>>TMTY8MQb4S}dY<}oiwL>&nFYN;ZJ+8^XIW;mpEnaE~MjmxSg
zp%nksgE1Ql4bO(m@;>N&d591*bn=HjUH`ZZx<H7PQ^7GEVictUQi7>Y{GU(poLF%4
z=sUKe^fK^u8S}fP-??(B!gcM8tC#|R7V0Aywm5xFv}4^BdhG8Ub|3V8ZO?H$*j;c;
zQ|wRfS7+V@u4_2mVzLpL!L9^92{1|>c|RYk-_LKH_FvM)<>j^{3Nb!M8|5Xn8d}|Q
zK_7aghNY(5;=VtjHO?VmSE<5fY_<rfUKu3)FYZl#%Fz6)?$>}kp4>#}_#^cVALTT&
zzW%HGI0(@en~+r%uLzb+HqPh(G;?1k5in3($Uj`p`9b@eqq0xR0>lh1-6Eh2_!<Yu
zR{93{wOq;`u2%(IpN4I0*1un4S=>(jYn@l|+2+<6aV*~)h^VJXF;*y>j;d)Z-yhbS
zIsEjDM{R*|E(Y&M`TW|Bi!Y8!ARteeo$CnzKF)kbKK$ft_~Q^<DBL8=0fCJG_XJb%
z)!_iYw+oz&7%lArVeQ+|9Gc`ui?+&^!_O5hhb3mVztfU8Bc!>?Nr5KOnBKQh;-@9O
zZ$9aq6_(#u$|?DXY%I$so@_kpTTY6(IEo;pm}RFoQkPf`ppAgC6~C(<T9MI$EHG<d
zMqlq-gu-a*6YB@LhCT&UG-~|F&KN&931QFu0VPVnlU~J7f*`B`Ae%h#$zb(12M`+K
zc(5f9B=|b(DFJ=U7M-gSoWhMs94Y`Ms!V3@WUOd5cKHD$a?SxwK&WWQHEj9*AF9|9
zHzXkf@aPqf+xV3281z5$zxPjN*Af}R9z@p(!qmx?73U0WV;m;70MxMe&Gl-`|1PDi
zyy!1h1q2W^=+m2uj|z&}vqq>5`muKnwkFg9ayqnbmRZGx#c)*MMSEU(Hq|;3?8XOa
zbfC45ss81P>O9azywrIZ_VGi9iYKfO&kN4L3Ku3<CbOZ^3L+jOKQ2@|k8zQunlp))
zK^Xd|1a|>mUZ^~ss&F817P@$OFCsCkvNL+%y8xk9#R2NL++R39?QPn@l#-s7F!{xL
zAeSxWVVoV}euWBVPg1yC+2u~Gw`+T}2>AVkGB#(y{O=hUnOHOL>{3$5;cSSrd>?~y
zaJKkIKX^7UID@lt&-QRN3CM#bHL*hCEF&&zpF(I^6>d!~8+suhL7iXA7Fid!EgoBG
zjYGCL=5V0J$|rT59=n)gxUBNe(4}ntuWQ2qx;Ac#ZS8>SE?i&K;)HqC1=XxbvbhI%
z_8_TWEORK9a36$~fZ2l)cWkrq93tyE%@sA(L4NHjx>jsJYzPE?8OHQSbOGrvHB{BU
zgLX2^Q@D%{wu7eJoEb6^GQ);>q6m^@7*9N8avXB`vPHP!G;;5E2kn4q5XnJ~`y*Nr
znf_%2M&aIspzmwlnYN=>41;wfwIw?&x<buCSaR^z(yXgIakP_!`vEHrCq5?gb5*Ae
z%49d)d9t)NFUmKK_pMv@F=zcNF|EIlM^h68i67GjDlMPzEQpKsWfaD+fdcWJhxC^!
zOy4;`g1yur-fqN3$r=Pm{+YEg)7ixhtg?*MTrjs+Y0I0UITPYASJf*DydX6GKn#sB
z0LVGe+lZDest;$Q>W^P7Y^{$u>%BJYxFSG4&Af&X4U%#So8;hR(;%D8n;6Wd{>I^c
zN-|)FimkmXTfn6}5NywA#AD=(7{lMquoRGYvRlo4ll(a158zdORJpMU3O{(poRY{d
zhcSk3X&}b;Oz=+o!5*j|mC?h{WxKBXYa<=7dlJTQ$iiaNuA}4oq%<WNx+y|qGH;ql
zt5P<^{Kyg>&1uhSP^IP(Xf89D02~7Ir=%c&>na>PnXKsnL?nAIol9Z<5Wi~8Kk3_o
zemvPbG7t1zP7O-)z;?_4zi^szj`wF7Y3HO5G5){uAJ8cNr+)W&0vylb@d|kxs4QYD
ztx~n@V&*Jmkck3*rr9zUd8M580qS?4%czad33Ijlv+Q?ZClf~*5DG}cPvg|rJQ9o)
z`;4nBr=v_@KG-6~n9o^?sa<3CcbYPjhj4~MQ6WoxG_jqa7G>XCvgyn-u4Fs7Lmj{u
zUt$x|SFooLEAVZn!f1&pNT}7(y4b~7E@ydz*;qH0tm=mkM3}j*P5jF-xzirMcc{I7
z170y9uD?~{i`f*a4?1g7HkF$T-45vObUPZ~uJJeQ)Ri(*uz#M02k-sjhuwoXK1@Y&
zrn6*k5qYd?5y~#c9)qdjNkce)<j}=*lz~AD^GL<t!+fz0oik5&YzCL)_%@w6LgOMR
z{_F4p>vFEYf=OMgB>&$RfX){jv!)WTaoEA_oTxu=#`XG1a@wG3R0)hs!elsD4isug
z#?Y%QqGW)I7zt;}(~d@=$!*DETtTF(T3bVj=Cqy_xGUP(p^VGWprzz#3a>gfo#qzi
zWn&xrjpU=;spAYJ7wZ@15zn~=;oRmgz1|gQ#7~qRGnEyRC=kE>!HBqEX@I?N>#1Y3
zq0%;RN<zdyapj<*BAJrN!1QyrW-@Zm!gD&(4t}2wPX&Xe<)}29yQMSZ99I05gW29i
zY1B-a-J}dd^$OEF_@M108Ep$9%&~32vE4s?mbA<}`?pHt&mSHGUO_ennwW6rj(?lP
zV*Loz3P_QpUC~pfGHU=Q^lfEu+K=L3STCZ!yUu#dx95sx33wYi(6q{Oo8|*ZReG&j
zX~dmBr_yLs2U3u!;sn$cOcYe9K$HCp^x3)EYP`+yrNJhlGaAjz_QvvEZ#1FxQZ2Qj
z>d#!v1{i;((WyAmEEyL+8TK~q+W6t6JVsmSND!S3Fv*l;MRM6diG-9++EojCbjS;o
zsaU;NA7rU$I1QHP5djjqq*SF)w6vLO+WA=OgJN$--q14zbAL1k#%e(ESA{a<gvmb=
zs8m64JE)li&)#3}I}OKMycjypdP^@mr(fbsifVEEYtu{{E`6$pXz5^cB@?v?ZaKsX
z#DQC@Ki*G;AVpMcwjPim(J(qX@SUHpBNOWYdJn07u>e_&vJdAVaZH*fInwx@k9{~7
z@lTc2tG`2MFNC0NepuP>pgigaKz8yZ!0gM#u&Iwuz>eE+^yi`TKkA{?>x`cB=Ai<>
zLK@ZLa+pK;ukT}g<jo%BF=89E;@oU?;{H^~9(rQ~>cw;AqrOUiMoWLB4`{t|pN>>e
z-Eng_-xe>G76oq86GU8XTz})A;qZr<N0-m|jXGiIkYw}Cs%cLAbjBfWXXQlzsO%gt
zk8C7o<*B0Olv)__vySO*IiOMMrt?lsxpN3C$BI4VPh0nCIS%-xx#enjnEnHZ>fQh%
z^E6(|*s+P|Z(_PX#EPbW&eT%QruA747QFJT%so0`1MM*I<V_cZ@L1w{Z`Lm?q}o*i
z7tD3%alL1ov-lY{BI`gJP(FYWfGqa42#=2U8iyeO6;N{lwL*kj&W!p2%nurAaaqyD
z`itYFl)|E28(mR_>FA3%30+89E&@g*h!Ozv@V6*M5q%R#p#pjn4(>>(J`6zO1SKam
zcGOb{D?P9yNu}NcXwZg?uBe!M5?-7IAeYYBq^au;I3%BqC<MxfKg(ekc1azm){0};
z@MN+s5@l*?^un^=)X^r8$UAf&z8%+NGp{~6+F%e?uQ&Oy-wE$+!{pEID;!;x4J1J8
z<zuKfOWHwuNY_+sp-BX3CcXP0EL;C^US0pgN3K%2F)LO&$Xhqg2hptf40QR^Km14v
z$B#K0CP$2}#P=LG^2SZ6V8iN6)QlYBJF77zrE40tV%%UxMc60hlTQ0AZLulcKJ`^o
zYACa=h9@0My(YHbMb+*Uop$`znBBY`#c*@A$nBP+?;;l=Zxe$8L6EWG^*$^w;Oap(
z9p9yrMtGn`PlrR{F%ZQ1dmunRb{P@a$p1u2U_s;>3e{L>i?y&jRM$8o5#(QDVdt-1
zLZH97yILIIJ6tO{Jk(UYJ5bwuLh+ZUlr$^*E5dGE6h9GGTSiWowIlO=m<c>gyq#Tu
zmd_zR>YHHp%3itom)k&;ia7b30+d(17lPrH;Rx(XOVNE3K?d&??^FV<;BAoWh8dSd
z(!L}Sm`OF=EPnQVBI7rE%}^w&8o$Bq^tcNZP=x4mnftjHA(GPsU*opZ+tL)KKQkB{
z)Ok18O|gyUiO-r_9-!Nmy(J$}UkB>x*(N*D&P6VxV;&3C&8RRZ$q(`j63?qUw(50P
zHRr!05;WMz>AdwtPd$VsPKC#_E?n`>8eNl;u%FDw0{~Pb+di|Z!%ap0kbKlQFG>2z
zIupezFVVK^M8bYP6wgAF52eI6e&eP|mm8Kc1GrP_Ih5+&!Ad50R#dpD?MQeG*rFxK
z7j*`DE>u)hEY#C2N60map^^2_o&>DM_^LNp-}|#Y>NW{^ATk8qGYsp2Nx{#fxS*Q4
ze?;48qPGpwak{#p)N)A0P?234=*sX6eo7|yes^p?93HMbG@b-AV%DSwU1aFNp~@a)
z=F7v$)&ya4BV~bUg-jaj*;~9w_2?zA7IT&w>uBm;eY>a7sfFKK4XBD9#=gw{c7FSQ
zD21&%HTlF<7Hpf<p}*QAL$vjlibW}^`i-cJ`nCtT=S+}rDd(_Uuqe&Op)TBw7ps-T
zXgJyN8=b%0Vm_v(D7A2miGYj;v;jF?w+wsshOB<FESSiWIzl=nGa`r<DwjxcQ41dG
zlVobsZF1&zAivO%NACRg!oU&IvD4f|=+IGfwhOByRnszmVs_|{hv{t4h}2J~^URy8
zXdg4i<WjF6Wq$(>F@uZqWAVc>(%e=cg*+1q2YEq|y^U6gd744t;=~8YjH&}x7dqC&
z@Zj6xjS-+CYD-rN20|?PaJn$fX$Wzgc266r5#yE<&^gl)9KoU>qZGM^DdjBAM>wkQ
zlG2T3=knwDW0BCobWnKMux*J@giHJqtI9rPS8-evQZcPcsb5~w%PWm%!J@8HQ}h|k
zW+A$nG;nc@4ykd5!KK$eefLwh%ed;$80k1ULCPVo)7hx4*s3qcFL$5$a|Y|RL5ARi
zvJEmxmPm-yrnVxbKCx5dd1cqf`@5y_pAMG2DU(roD>4gn7SW#tzASpuFO&Qu_s)Iw
z52I$y=afig&bvt)CR=D~!!ahl<uoBCivBf)?v7s087rG0R>W9`p=k212HI=L94nzZ
z!H=F*lZB!6-(JUj4cIpHdskNsIeiLe48(dk1#*{17Z&gQr0ykYKWO$@ZgD*VkJcmj
zdDRlUI4@I#jY$9&A~+gu!ic-dBWtvG${2!)G6hGEEAkuX<XQ!!%$V21mBG@J(z6&<
zX_`kFd=%3*K^l??i=yP@iiN_k!)~`?)2RTNsmR1&rE{8FMYId5p?MaV;K!JQ5`B_5
zk{c4>q?eEyRRyR#c3aRphH8rWiu=_3R~k>1?;Eol)o_~=d2G%_Rt8X-xoYhts4ueg
zGr`|~L<<X_OJ6lvv){S9Y%Z%999eWQ@;J3**+E~;=vWSXFE|MWJTy5IK#>W__Q^;C
zOJVY>Zn##3|Jdt!hzCq8eh|<$Hw%a>0c%XQORc{PT>h8jWs{`o;jpxMFT*qH;e?dr
z{*UxRqdmzxpGz7azik0Crpb~%-C-4W<EHqkjzjj;NUWx?jCWGYI<1P~E>VOGVUt8W
zWt9)W*H6NY!<_ku0v-zH2j%@C4{W@}mUWqHWQkVeO=ogwK)>C5D2Xnl^rg4#4VrZ2
za~AR}M1QL5iobg$$q@QBZ2U}EGt=B@oq@JKz`phP?|$s}h4lN^V6xfp`h<*}9`_>I
z$UY4f)F3sVe*-oi=fj=bEq2z{+<pANB`!`Yr<%{bLry9URJ~gQCuEU<MEI!33e1}v
zl@Pm~B;s0q`KiQCBxKAe$vsVYo7amdKltT3flw8UXGhHeQK)J<q(_hSzH`V~RJ}W3
zZ1&LgMtMg728<jRm*G83$@7ls2%$oKHBE{hA@x|%H?y(g7Fq=I>~p=-sm8wav@ITS
z(@wF=+`PQeu_+4HqIBY2@f>_~@4v^%r@W4M45nWnp$Vj#1d`@;D%B?tJnvnNZiU0w
zkIN*f9MY<$Vy!W@gh5IjH01a>{}hQk>?K~1bfC?y1mAq_J<5=LOfxLd5QH_9eWI$e
zH?ifNki-3?#x?6bgT{Y^?n_VgsE1H^17rD~HYrS%FVq~MI2|Z4(%C)*C<oV=v`tF6
zs6uOz26jRbk~P;}>Q^YO50XR~_}g#i(ok$<*!~XuzC=(goigMCW3#BG{xK|DQ-Pi}
z&b&IjzCNNJ{Q(hWwGZQ}?VKoZPlh_zbj`K8fd^fHYM-e!ZAxQ<W}8fWGyp<F$Ad^l
zri2_R17bvq6Ko*dR2cs+6X}!-P?bHhAOR5E*X~b09F9NNoVcdm{eFFZ_w+RWDgd0j
zQ&nZ=JzNPs`FOYZG;YP_LwcztgwcQhINK4Gc8zps&A6v?ZK=O2!P#49VHK9qVbhlX
zy2|~%GeyLs5J<P`OYQ)({iV~Ce>3f@Ew(n=;gonNTkC1Y9nPoab@l1{nIW5TcyfXc
z|8#087GcJZyL$GqRf<PwBQ>tIe|FS|-s)U!hXEPeZ<kg?Jx0rW?r9gLtj7IGSx!l5
zHsXAV;y#F^;fD%x&LEz8z*Wm<(X1|A#R2n6i4`z;ZY}c-ds6=b6I#Y;wIR()fQq+$
zpzxSJNRRcT_ZPnWhKS){VPg`Os@@3!^jx<Jy|E!hL9!)ehCGkdp$54cnbT)d2a|U|
z+HQSkM?fhDxs{p%waQyDQ14d3@#3GG`|>}-&BeuZAA;+zW6}0|$S1$|KDDg(fFc<&
z4F)TUQMI~me<aq&E3|Un4~|~1{=+u)yslzhd3IBI${fcKe)KWUc=c3IAO%~nDOgFQ
zlvGAJok9l`7o_VRc8EhPbwdmG<<LQ)oR|gqIdT;+P)9?460I&Ksgm_3XmZ9sqvoDG
z9nkJ7Jozd=y@on3HBRLYWT{wKpER$y4IM2IrODk{{z<4X>hGfe*;OLVKY5%ix6foQ
zIfF9R5Q~q(VZv2G$Ep$+w}Js8!Ai67e->fSCbCdPu4SE*ErMm@T+c4tcQQBhUBCrq
z@C9_(m3$dOdNg!2*Kd8vj}m2+d1>C(p2St0!}%L{pYCz*0SKZ0yy@5ga!xFQSNHw3
zBrumgvIofbS$&|sL(}H}x^$!%zVnXLZSz-?#vhQU)050*`5mKSNszaV2Gf^o`aZ!g
z@`o?o$@w4mV2LJnsZ-I2dL+fZvZIo!YVIsmrV;&Ev22zje0{kRGu-`#rXS%%xwd0q
z{qdB+SKS%h!rtgh`E>isPkn|t@wj%2%j9)_)1yL69eeCoZW#@KTRR@V4hi4j2k7xI
zgpFLa^~{Z^nxtSeVzkf`W8{+YgnggPFTzrVFGCLq9W~{YNGM9O_I1LHf>`DjeL$25
zYe;~cq5l&)t&<RU%SwXhRHXRw)%XjVAbqoxCPi=s+Wa@=BG+{}mD0}QrKt$Dy3w$v
zIKlcN_-uc0@13{4TeBRuIa+pfMDVelUJzeU&XEc{W<Nf)Nssz;R=hdb{yd;GmpO>y
zj6ym7cj`*V;GjoZjM9)cO@86N%CCvld!gT)x7i@;ZfK;#XD_eo&udk?x_pDAZl~Zc
zE|eLV$GG7#cWYg+^RLvZBnIkOH8z=D%N6-V?k)ySv!Q~lHT%;N%e?7{#lGr#VHxDf
zOktk1IvTK|i*CR>tp!yThm#?k*<3Cr@-&mQ@m3Ge3J=d-&5wv3pMAnPsiF&SaQ2xW
z;{-$5@mTTA=RL(w9GBa^IxOfjlLi>&#j{LBu%`~1JDGV?B`O*~9fd)vriF~>EwYYT
zv8-;DsF31Tdn#gKrF2=V8_G&e@wA~ArZnEcc={);MkZ-n+$qvgf`?H2v7*&l5a}}<
z6Jp<1txCo<${HY&Ot<&iCU~dsz8UzYKbE2UA~*T~T?e4Z$>ZE>?i6o6Yzej%%3OJI
z(!AzT18Bza1==!}`Dh)&jsEl6;oEs}VX$F+8%?`y+Rz&#^-NQai-LIQy7G&|gaI*J
z3PC{?s-6AcebpX^2@iiKpXn#0OwaS+9vHnI&=1VMTBOc99~J)=?fSw28qJMGz(Mi>
zLFIF@9nT1jP;jgL=pQG(eZHuO+3vbP&%@Yk7UD@sIh&%zU#k8<wr?4(?#EFNB&%;C
zpG7xhMW$pJ4SaJdkJT=*EXqrr<=mjOWT9#3uQxf!RmmoYDB!VuzP=qO;8hK&1jxIh
zjGvAPo{@x>5=g-bgSbjZl_Uuo#t~*s=~a!%Du^7b@#Lq7n7A&4!5oT-&rQRMcI#yf
z(smr_#{|kW=w!WN`q8d7AugJznn8>|zTE}a-ZNDBjB>aS_YT~bK4eaXFmMZSh#L0g
zl5YB}I&zK?j;%(FywI&(CobUV2Oaa@1XQ5WiOBq1tSpa%rh@G^WrJ~gc=bJ_^@6ST
z?L_;BIAoU)AEzzP=x~AVUr(nhV#4?(#jMUFsuOKP*0<28ESHt5fZu`k9z}eV0pstV
z`|Cu(p$?Oxcs9HO$|Z6$*A<|Kuo9H)W)a8+=e6$Fe9;R)CA6dtVF}me65*Gn;|o*#
zmZ{eIkq3@jhNy3D?dNHZyjB+OC(eG6#l*7G2jF0tF+y5ZeYi{GpT3*!jHjffrS+ZX
zxt;&mWsjD|{OUZ)IS_%ukT}eqdDiiGN|idV=1VCXvYn(_3?ovzNEx=v9%V70JCV)>
zQbUs@Aku)}Bvp&)r^LqqhE6X#Rw^CU(3#By6!7e<7@)P_4`C&LgZE=tgc8zQk6LD|
zIjL!Av_=rfe%d8Z;R$tcb_gvhuGp#2CL(|$!%aX}FOre0uIq{tN14*7Fzn#GS!=YQ
zO~+BI>N*0lFRi0RAy(WJ-7I2DnB^G#OF=BNmVipSVCQq*AfmHTmv$s)f(SxNOo94&
zQ7A6CI5eRG%!!Mt?i|~3p^9#k+LR;V;mt!>Ma-;3wn|bncjK}xU-}s0xaD8oCaHpg
z0&_%bMo#P|pQ2G^DSjkPLB7!dl!*{d3?fB#C01-|V2Y-iZb?ne=7^bTcqK}oLM())
zPpZ>v=XP{RwRF~8=&>Wy{tuB@wRWDie#Y--@#H+kdcw&Fq{7pC>^>LkNs#``J=r=F
zSMzTJG6jzOoE?ufraA62y@hxEMrTiO-<dpLjSW3Bq?MJ&bM~B3brf(XV-+M*bjuxP
z&Zt0m)5t+690!#+2WL#G^xs3o<fQ*a)Pugu3;DV(OvLg=^z_zL7F!k2smcjZq4kqj
z<EBbJ-l}_j?&t~nE>fR`3odjC2&aL~hlnd{rDtSR<nB(tw;>&Vn3#q=M5$j(G>N>v
z?lj^z9ybjhWT8?*7BgQu7nF>?xyj9SEbPY26@RKxo>iV)LU%_zYW-v4n{jz*wbfej
z7%=xIHj)$`-tRBM0<VsP!CjSw$6IwCBXB^}whm0cSqNo%-!8Ie?kzp@acsE&^PLCO
z3AIMOE^_nkNi~&d?10_K6iK*hj|{(=_h_y?T2vi|Wch~51(sS&0TDO?c0ES@Bg%_F
zOBcPYHPX<CssJeaIrf((Ft*~4j?QXtIy|}HaxwOTA={VeA`R!@`h@cg5?7Fvbmddt
zV_jBE(>AdUts|3Tl+*o8tZo3j0=SQ&AMhuv*|cSp$sjOydJ@Byt_h;i0^{U_Ey+Jt
zGW968K_6PLGnu)m*)0K->|I|VzyKOp!W}Uz!vr*Hh&`BAh~Mm8TT{1uf(NIOQZU@W
z+q^`VINa-~KuGf2+p!H(mQH54ClY%c@8xWpxKPN2#`df^&RbSg=fi+cThYrXoL$b0
z4SBwRF<3%}7Jl?pGnFh>B|{QgI*-m6bNUc|MFMsG)mv`66<5JvYYR_+5E+6i9?iT@
z8I+)e+Q0EJ^vI&moj->9g0kO_HHcENSVIs|3{;=LqkS9M_-=Uj^rxSqC0A-VU!rO;
z2L_B&&c#ngEZM>hRIooPDi69A-c;3SHapQ--jiuv&sy>EtGpH8d;6mN#bRU2ZZlAO
zXaCZa?HXZ=RNdVDhg%N$W+^4E091Bg7F}>ROWOSzkT=pOJf`^FNxZKolgGm&D|j+2
zr_=JYe;-eUa28h!aG`x+=D{Qk?VE~1j{L)-;)+-1UaNUxfRgtI5oz{g>D9O#!LST}
zJ{mZpo^YWtobmRb%jeV*h3N#xdJ(<w$&|bX1lMPWE)Y+<UewV74fHf_!TC+;!uJ7g
ziw;k|bL$Pb`0#&EOY8ZYvlkJRswp#_5@j(wNSatwhOB7R3+fJ<J~8ay!~alRt3}px
zU?^&R8GRs6h;^<(1It@8uDQU>Xw$rbUIGgHOih?5p-JBE4D|Z7&TQQ=EJ--Uyqj-<
zjXrQ;9F@G%mFhUbT8lp412y--FMUXbxaFv6K$c(4a)9(#uNC4K(3j_L3#3TmajNX#
zUfipIO4&^N<Z~dD!b=K^u=EKLmUIim6#3<N^G~ZbJD4Y)XQQ_QFy6)aj&{DRIp1%x
zSBTxEcb!&;!1aq0Le=|;AOcEJSBR<u1L+s^2u@lsDlS9<gF2yDNo+9I#S`^UU*)rE
zaotWTIrKRc!#?Gs8*O#NiceM_D3(=lx^JUK083uhj1Am6kV@$o0fKJv&#ylDWI%E6
zz2g|xXVT%w6&dH$pAC-CX(%fPb}}}0_8^qc7qD}^JzM`|`f1D`jlnsejkrav37AF7
z&omnIL+wmxi7teJ(?w7-9+IF5JV?JE{TAJoqWtP8ai;rG?TZ{et0l-WFKdf_Xf1Yt
z?rvY=&ng_>vAq-r$bWsIXV~sZqs{%|Ap9LS&NvMkG>CHv(aSI<?FGHb(>SF=tl8b-
z<=T<xG1P?mj@9*K+^FBp;^Ptg;>Jh(*RJHr=QfM0`Up?&8E;u@<Gl1K1<sfT0HuS_
z{8Dw)wU2o~3;JDORT-mlMf4!QR%6}tcZUH>4l&@JZLOaW#Yw=0lAPgUkD#f4k8XG+
z=6kH@7+1Ls#P8v!If%%+g#Xo#yeSry$or3p>)CFHzT3p0%wur^pl$`!V?%3_H(EDc
z+d{Wq05(wQ3df_$vLw66iI=TdV6W#yrFTetRPjoY*81{e?&Up2qvN1YTgL16&4DKA
zOkWz9b}GsCtPs^<Urj`!@KQ=SQ%<<NpMxxY+VA1Lq*ahHQCPI;C+a1l#`DdKu-K<$
ziyo}!{IWVANOAfob?Aq%7GL4)^BBcPlgw>G>bo=`5PG1wr^j}Iyj6WyGdnw8pPBvM
z2G+d*B^X>yTrc2-GPe9O5t4K@F6={Bm{2PdhiOxGe~o28of~(ctBt9S&C8$2b>>xv
z(F4LD+a^+^AN;&T4~4a!P2;ej&P=E&v2VJQk<Hb3YIpHRB1smx3G&g-gC^f{r>Q?Y
za2%)gYw%=H@8>sF-sh!MX6mo6gl(&hgw7Eg-Sjc>9ou{_7~Q%B9oM@f%^#(C-9Am>
zkZic!J^?20_u>-8S;Rn;dM$<S9&<{5t;-KOqb)EQm7zBtS6ArMl8KQFR0u*aTQ>$!
zTLZKE>*WN*R2LB%_}^j~;MbY_eK}}pmK8P*Q#twum<j>lKbQzZhUH$Sb<WR1!>8_!
z;~XwB`=?F{KweH!QZMHi*Fu*kI@}I8VrmKZ)y^M?{aj7XO@~fqEiSIQMVA7twmg@=
zw*7n0{Ohzc`xt11s5o-sW`05d$E|w)4iTPU=Q~oilk>7y_mAsBd-P?Mv$UQlVq#jP
zwBI3XCJFf*b5k-1#y01Kq8p#%{vbB%fETpy7J#>hb_T65IJz-!7NQq;?m_B2p-gue
zwk<#Fe3lhbkJHrf9p#cza<4t1q?YNT6@wxgFj8W=bWv4NdHgaG0DVECMb*yRVjxid
z4tJKH_?7Pp`a1p0E7<Tp-Y~iGc|iE(B6<AHtuWEW4rskM<e=|D`4M%ice`Rx0N_Xr
z{&PleJfzp0{{ka@B#S?{(S3@~<Bl8j;^`9$H?U+=j8g=MV&G|)<MZ@S6$fMU()XLs
zA1F5>In&vcBdXwUSR@oiTk$@McUnKtES_R`2hJPXlct7W&)t;|5GEHd#J4h^_WYJI
zRv%b}Pfj*orlyM?!`L6fj#DCVi`>&z4ZGbaBV}rb!*M^=)=7c7;{Ccxl<g<Y#7Bj#
z41d_H6kPc$Pc+}q?{5;l3sL<PXwe4Q3SX-+MU&$##gPU=lNdxax@Ins$0>Q|z)n7!
zFsM>naI8wtqx})XSS~NRXBxuJ-0U42>kMNh@fh@bM2KXManCt^9t{sJunP?u3=7KX
zzk4(s?St!Z9eJtor08l+BwFvn2`Pl}*N9ed<mx#Lh#?Sa%6bWOPs54!N<Z@Fp0JU!
zq@cxt$#4jXq+;=t6P*m>2(9KIkE~y0&)T7*Oc*tQozqkvQ*3oPTUGykZvH(eiYX1p
zwmE4q@fo&d<r6U2P~D)T7)Y^}Lem`=Jivc!ML8odSinR=)vI8>aXJ_vp%;gVv_p`_
z`?iKVV}<N@-bOG3#WZbi-F)uclGD;)%$1GX?srq0cT>Rc_74xY-mrm7MGgxcR<7yG
z)zuI>kFXn!Xn-Nvm$mLN0hRy6^lOhyN{^+f)ZFy$u_Y!7$1VHE&IVnd{_3a?TE$E2
zA+fy&|4{!SihWQ9_&H5#sm>tkm<BW5v0pyePyE=eG*0WGz?LWuzNBB5evUL0Df-JC
zFdxG4_`0`_J-gR-8@<nHK5q-8r!}PoPU4F{rV?@ZpF6fjI)2tZLXVv?J8Bxn)IQ1!
z*eZrH8U)79HG})zSYI}Arl00A$9UYJY`}42qnV)N_spa}Hu~dJjmE;N8@6gH4nYhN
zE~?#tSv%EiGKNu6{FH{vJ|izoGi471@&-i)V+cPPCY|%-6&8@pzJBq7GOYL-_A5c#
zA3Gt6D+2ISZrXJsF0p~zy(JnJi+>;xf*ZS(wkc?Rfm_BmaClUc?ig23+LH5ypcZ0B
zQcYG_d^XREpHe_l6&3sgSW$XOmN9nFPcE78-rJ;^I`Q<6{J$1p&xid*OW2=3+!?3S
zXGiLL$G^|&&hUQ@wP@KQx5QcJYAik*pQ$StPG-06k754ACa3QZ&>p5FzU4C6@``1D
zLwqdo+^2qf$IDT5T*fQlxOjP`O)K$7zB!QwBgCzHU5!4Cojr{i(}f$}4{2c=)KOX-
zz^nC`CQXIZh=G;kfQL{lX(SQMXJ-M&Re2_36I@%Gy4Y=csRjt;Qq^y~X<20NzXL=i
zU!B);`kULY^?!hmv^>ssKV8T;@NpUfRAY6=A%;5yQb^lk-Jfjr{JE9+-}=hs0}1PO
zOC{RV63+gQBjjrX?@xW<>qt)b^^yFc00$&vVS1PIz#Hiu(Wu(bpCUzLMy`;I#2Fps
z+GL;Cqn=m0^<6X}s4;sN2*-?-sB%ivt=LKm!SjtYMVkfFOXuHw_Ni&H(XJ-A)>Oyf
zUZh6!?(2Qe@<PhOZ#X>lr6O8g#$|w?W(>=el59OtE&?8GQ(7ZA@qD2LS(GobDX`Tk
zgD0tJX+5mltx=_h(wP6$=Ly{H&q&B&v|+wjPgphh&PFlzRQHhBKYi?&^~5XY`5wE(
z|18$B_FZsnV>)x+iV)`{TN-C&4*JG{q=UkdNYZVvi`)ULth5Oiyf7WIz>WoAKxSIz
zfgn)iuVho*Bz$6`PqKAnQ?~&NyYf*{$xp58Bc_GBqKYT^5B6Hw<x~oclFDd=+4;*M
z9NWfzrM+pK{9aVW_b3<ur8(BUBlyO`IXPv(D}N9D@bhh8H5!oJyg5|@V^)AgU)yHY
zdh|@r<J*9%{MSo}S(@>fK7iAr)TtKa#A)m8xO}P4$QfP=^IYjG)eX(Z0$WS(>Jj;&
zsxb1@nTS)!(Pg~T3Hh5BYaUcw*=x*-Z65Q;NAdEmr6x#JdP+lO<|A>2--QZ0R)G;2
zSI~u07?m@+6-Z~<0A%|Ns~$%0le#z^0&rxD=<3+Bm1JQ7{S$`Q4u$ipKw_eYiH-_1
z0w}M^2;_3@SY7{0N-X6`!&@aetLWR&Oe#T#<1G%Z(}{<@yDyKC%uGFHP6-fN&f}CY
zyd%mwa&hb0nVp)-eY%#iO5(4y6Iy;2m8GbK32^;l-=X1@5fJ56J`z41G!$1WNChEt
zCRP~T=iX)_?%XOWjgc{^Qgl#!eEKrq+EnU-(v3_TVk53Ci;9b6R4#%r(O*TQ=~eYa
z@MQ3u=pN|CT&Yh`+ffN;KkTg1UdS-Y%lI?Kb;KLw&=0-*k@O2xyJT>&Of0_dT*4rW
zBo?NeSv?H&UzRyp$GlDrL}jQCLlve2+QFKp3-cie`n+mcO_lp)OOM|FgkPJg6yHGw
z2LB$O8S~sf<d4nB5hJc{;499KwH#V|;<B5>if~0rzux(X`M<>EJx^%Q2=)M$mVVb*
zI{)5mQ7bU^yjTr_16(40H%G7!Cx@@bqJ4@A65yH2h@xia0QCTc`KPEl52HNsJst|M
z8c4SAMs=@8!*3=jmV#{)o)95wWNN=^B7&o*Jb>0fAJdmUn+ROj08h4`7bG~EG^x~m
zAPBWQkA3lG!4m!uD^^u}T3g>LGXB}&w10_>xS)55bY9r=i#G{fy5g)_-Bz4*zc?y7
z*tjko_yqTx6|yf^)pVyMc2}J;slfx}rz8l?K6O3q;<nV?vCJkEZO;q%{&1nEtx(hW
zdfkgV44lT+?nWFtG`U2#ihYaW<q#3-FSVeu9&r#^wb0YoBf2iU=}SsEG5d`v!&~#T
z)|1{UFk4RJe$_4I>M@VbpSKW30^?X<NV%`<rw_ypnpy;)X~XzZ9H0nz{oVK+3RM>f
z`6#KJ6UM1q*q#<s`1<)aZCG=rd07^j`^3h9K0LK2V@OA56#;~2OLJ@wSSN?LJ5ZW7
z*{aB?UcxHtshNQ7W2QYnrY&mi08hDNWSiSlY&Li+5)(%Qd?8SVxEtm5yr&>*)hw>l
zdD@8l2<EZ9oWEV^73oXY5S~|RL>oZ$Oi(o$q$GNa-Kai!!v|G?anO24pV>+UrabF9
zP(G84U-39>bRrNqXL}v0Wp-Z)06q;ZX2Q3jCV$(RhyMg26a4GO>F6vJezv+jT^-I4
z^qkPGH6O+2vRjcsg`pShOS9_dqiZe9Wr7<b?h{wdEeqUq?XDQarJ}UBg_X7f5q*?K
z0BE`Dk;I}DB^+r7^GOc<6TpRQG_kk;a=}z4PIswz)26-2vCdZ$S{3(+Rx&FYNv#Ds
zxaO#;f{+D`+NcY$Q@5_#MOEBZcFe}{KN3@}LcHlFKk4j_T#sVlnES8mMegOj0+&f0
zHjEuMAqQSTotij57l0fdgROxQynD&|C+V_461*C5I{c*e*gnFp!n)o!ZT~glR;Jao
zlVD;oJCD<;Q96sks%hOFPihK%D!iiB5b>P2mM|2!6jtCTh;mNWZktsvU^zQY5h?hf
zWdy6?t<SP0$a3&=ae{Li0&^Npyg&1)-wJbW7h_IEYQZ;^DkO?CPZz@-(U{Q;M-DbP
zhoTcv08YO+q@Fs8eXxp$GkCFRc+|URao0&7-^k_}*5&M_%kSajlIv<hJA4K{B~*$)
zv?~T4HZnm9LDq0Gw{V}$A!KA;o4-pNplEC8i|+kyru*iWs)6ZLDZOxDMtG={9x~4n
zggx8AO$1VOV*v{S0F%2MI#Evn3PvPN7r!i+ZD*&<9NpF-Pv7RUzl2&A(v#Pq$q$?e
z4n?hFd|Z$W1O}wERu*0u=YL0<u~!Y)ZE)GIwN0rFCc}aOa?IF=!rXRIQ10$9OXxJC
zI0T4Z%jN9KIv$Fe9XYuQ)6Y}ReCuWP3X(dU|7zSWdEKEJW<IWTG>!_w@gc<#mB7Ny
z8?0n{v$HA$AWZSic*7ltxOL@!b%6gi6dk-|b94bTfJWKz*2Jk12-9WfQx8i+d~Ok=
z9M=I=+LgWklwaut*PAYro;z0Vq@?QnDLS2zpW~sa08xDx=R}AabxvGGj>%`PnZjq9
zD9meY1AV%Mc^O(>2>WB0yjKDZJh(dCJ1bf86nCc#zkqzyl3!GBxMhk#cc(L@W&pHJ
zmYW_>Ke(A{D+YLU003i7;a@f^D$Cddc1=Ky@Cz$rP)Eq=k$SH<g5euu$#c;mG#Lvv
zdAbRbHknL)mm`q?oda4NVc;G-$3!USNiSQBWgIFGeNwtnx!JsFuLJ+)@eBDU!Lv^u
zbHXpsc1r^0>xm~Xn0>cjUrNpxvk1G@LFK)cp%Ccenk-<BM9y$hR93WvF*c?E4?Gvg
z<`Fzbwn!wY%vR*(s@xa1l5{GvNn9AWq)O`77f9rCcRiNut$KlpWq7&ouAr4={ta3W
zjCYn-M(24Y*ZmjhxYFRH`Sc$RJ)4XNU;yaAkAl~}s+%BFuyfq*<cBJ+r?XR3@a)yh
z=wNa|x@AEQCO}EJo&Fif$l*7*w!W}Bzwygy-3K%p5<IQ`)_Lh$bfRt>DttQREYVr~
z4n+^p<7g4V0O!iPVXHlZ%b$LOOOX_!MofIhGBqcI;7fsflGlVAQIBaLI~fbK>>o;*
z-U5t(H4+ra&A{3*l*5369r4|C!^DpN^}VIVOGpF#?bn{at-=zrQfcfeyE7Hg-e%E|
zQrxq=h&pt{sj1K!oU9#qJfL5IN-eL9Es)7nFY4&a_bBO5OSEL;l~)8kNVe~ed;4IR
zf}7C}cwZBaV1uVU5#Ro?w0g5KJlg7dnm_5SZ~aPG>piayW+$}a`Pq4fRCM}=&VitS
z-nw$MCEA<i-Mesz>1`AZojYP+rxv<xEg#BK5*A|#etwXYse>tXYV9)K4tUQyu3)}3
z>jfvuJKT`)g)n_#V{l}*0vppznVdr#ga<C<*gjrcmqR+~Rk_6M_d|!KLJ{^+Mi|O4
z{XEC9HEZ_RR4Cx{<)qzXsa3A3U}sXx2_UG@j(Ti6XS8p*wnmtkn$>qb#&&_dj9^}4
z?G4QA@UL9q9QhxjzA~!Hu8S6F1O(|u8l*XN8FUCrhje#Jhk$@I97<Zck?s)b1}W+8
z?uNV1`;B`qgY$zi7^j~7?7h~UbImC&6n2cE+<`@rvOfKEgwTAlhlRZptNmi5Y3YG&
zN^7c2rwQD`0y2H%%>u=dj6Ag6O7@#|quz8f)pVjSyypW1D*||;3K>mCR5rg`GX(W<
zjN{0Ko9IlM8MWB&KT~IltuPQ{UDn*|FDb2;cTsv2aclR#v&y`FckIQ2Y=l_Rr$k6?
z<L*XHJE{4ss*h2G1@c!N+FChaE@V>QQfWaCT!${macqc=Xyv~p2T!&BAoV$%n}2h>
zDsYF!LKyjo?t8p^MDry))yM3d{9PYDZ8r|FUo^Ibm=nR4`dIvUAivtQ+^`oOaPsq?
zYt&8;D=E{2o#k|eft_DnF;FD(By43OUoMp9{8omK&(8wiPzhY+U&YD;{GK?=@~`j8
zS>3<yT|IGSkogGtRvlg4B#I=S_^n*oOE%(KRj~xGxX>$CxpL7@20?!@@<aJKL3_Jz
zLmytNd>%SsiLJpx&~BX5U{~Mao2Px<tLYm+fbPVoE9gU>(#a9;pR<l8oBrFC*6&Ea
zuQEO_?;ypvkv^~{fnyTUR!!wn8eR&UK3So%k<07($RpA^e70NK#he3Q2A>^pLC65i
zykeipl}hJf;SYqNZ$OSD#9$Pf4tzC6x@Y@Ta|9U7cpM)h9TrbZ$<G$8o^Jd&7cURU
z2l$I`#<ZMZArF;_E#ZfUwz9c6b9HuG<VS8OR||Yiw2PmUYv*chI?jXvJLP^pCe?7-
z&#c+zBLci<V53+<N6k)o2mk&N9l4Kq{saEOxk@1!k2@Kg!Ef$YN_rsO!2c$l@5+1E
zi1Lmoed{E`&daEME1eb%W{RV(>%K84%xg2-NxF0`d04*|sd+c=u~!Nla)5Cf{O{%E
z<-vL3ku|-?4ZX-+Ct%xpw#nC)q^uwi_zKl+c09fNZt+UAyL+?jWh^Zkh`#XcxICC+
zikAicY7H=BG^$Xwe4l$m_rL35^PWd7|LCyG-`wq2DPI2&IHSpZ3)j<+S7$8I7C9Xc
zAT4cA_~)`*1bNj)0lb=>l8VcJb)5gHT%9QXU40829&kP^Kd-%v1KJ!Q&bwx>iXgpe
z4G&{@{4(5{Fqm``e_Bj*a<@}K)nZDtp@ya@R)YDw-R;L;1^xP+%7y`oJWe$ie1KDK
zK9{mbeFiMhBmS+muJD%r)5q8+u{u*0TY6*q#5&Y8<8vi-)8}rNGUich++;NS&pl1=
z{df74>(2DQP@K=#+Ri@bPX>9LFKY?EM#t%$ESjXK4~IyF>@QV>N?Rf*8INfR3@F0H
zCM@9^+CgkySl&?!{?o~K#Po1s%N8dO!QSi@!gb#0Tl=@1>Q@hTw&&@A3Utcm@HFL9
zjz4pmJmHzD-p9RJoX*)4nm`YY_>ytUmA85qnX}6bXuhZ`-kUxRaR#1LJ|9Rw+h9Xc
z$!6CIR(<PUP0lP=G7-G`M|;F{C|LB}VjcQCu?v=6$V^p3XRTRs7HSVG`<hg0GY~Ok
zp4V;H2J~;P_ReZ?ePU+!9!WayQ!@PJpO@}40oe*<jCOVvRbou}oB$uegnj9Ju+QB;
zr`vI|O{DT?tzLuG4?WLKl40@^mXnfADXSq6B+~hTZ^8NB70hT4_Mp7}jY33Dp4-%P
zSH86Mw_2n?&J(uG^Y~hJI3oVRutBc_E4{TTZM$3$i&8Zrnxa<Bb9M(ZlE`6JtT0>1
zu*A0e4XFrkmKN<&9vLW0KfwdYkU}VHxt#52oojIVFJ9?p-|Wpm1~6)J0^X`aLWHF}
zT5ld%gwC(0v=%RG>l}u7rpczBfCg#>_858GAF1ETOX0tocfz`BS+G3=c6w2uYPQ;J
zf~gZ=514^j^LZ}2&apt5S~~ZOMPT-1Ce!&NaW(g~nGYE1T$b$?M4M`KvfQa&9Ykdv
zrdhLR=}(8@O@-3q;oP(SG?}YHD#8uB;p1KUeY%o4*r-%QjGjQX^{4ZAXnd)=F<5d&
z&|$B{8|nCA*~|8%*6E<N?xcE~qVs&|Pr+Qz)tzn^d46*wYL}Zik!`+|J7-`o^(~GK
zM}h?5AE6j$loxx<6hIVq7o7iHsly%1>*X|=XZ|q!$jpPPt-u8zlWGxS8SP(WBNo5C
zWogbWTgNz({`6$!4jUfourIr9Cv^0pH(@_E_q(>geN<t)JLb&Rc2wh|Q~Y#_7vWyr
z`z$ZZS*wU%@m3QJlu8U{2B2fhn-*L9yZ7(Zt}X11^VTODoRyy(s6tnpRdz;;LiauU
zpP$gWAHKv1hg1Y1uts3}7aciVzKyA5X5xneEG&XSyxY-oL66s+gz9JU@@TlS57+!V
z`8g8OwLpRQFDmnD0u&5GYTEHb*_JZ5nQ?5)#gjug417B>)69kG9CBaT(;?s|?y|f%
za4}@$*UUMy?C`y1E_j?Zv+%iK{V}}$sjc|%!-UVlc!8e{I?0i~PBd8Y1w+m&pHM%C
z2!||D0U<Z;o7n~Zt@S_I%xxZ5y`e&GzuF!>Y#VpzhXwao9v_dM(g6g9+2a5d1;La)
zKNNj$$9p!+>cNKV-XpZwF0^!`(Ad}rh|AAkCRM%8vne-tez{{hkexZbxpMN_7Na!O
z*Z<0a0R*0+f0`CA*mR$6ry<<j-#!Peo@m!vuQcwJ!Hk2m^7DarScGH%$Sw1{KV^U7
zIs)-x^{_6z@ynik#@VnCmK7^-2&Sl8a3S@)I!se8TAj7}1vBnjDXAJ>1@2|sTd7Vj
zT93xfx$6%B-M1RbT)&+a2D(AG4M+F9rHw?rVI9|nE{puMZdmAo<7SeXVfad3itQ>?
zj!mH7z>BN$HoaLu(H+PPPw~Ni;cDtzI61}jS7G}#zcZ4e{Iu-zcP-21t(#_*n5r2?
zzZ5LD*uDy-v5{L1WFDOW&F6+;2v+(jp_kLy5$Eb9<&<4iQzVx26B|}RgqQJoi?B$u
zXE)$KVlmEkTQb&9DceGL97A28f29XM4a1|GW0pagcjrZVZ<9yz*rI-`nxoH=R0-8Z
ziNlpe#CY0m#Dqi#_s|G4Z8Lki^+U}Am^VJLbH&ttOSTYcik0&m(rimRTzX;)+`7m4
zBO+d`A>9mN%X<`B%CcXyQn&RNVEE=i#q3XDB^2xLNo$mbM}@c`kr;ZFGlM$})L1x%
zeY~H?cVD6SZ0yZDj_I8}EUG(N9b4OC3~$+4+O79Iy5#Wm3_ia8_g#<p2`3W1C+2P5
z5Gg;9reayNXR!Sf`sldB?7p8AJ|jRJvu!f?8OPr=4*JhpY(d@MUt@kVqLaF}lDm4c
zJ`yNlUgZFIekxvvcUq+OGJc!g1gZ|?QLRu{6^9CGv)fKpy-zy1ja7jMzkc2=)G7O1
zJcaoV7atGT|M=6;CVi3klyK;}7+&>A97dsxiVEBp<<k3M6)!`yXOLEptuv}CrN-yG
z+~q0ls;fa*w3GE*E#~x>m+zPFL_zJa7C?FbN=GM*8RIz&?el|4L!Dh+X}<kbDm(^}
zH$+TyLav9#0|~5NCt-TV2eUP(RH*)6i_02AKcw(-3JUhqM6uW+9J{%LTpO4Xb$2q4
zl|32599gMa!#Y&%`m(MK^EZR&Nr>hhh4b5yjqkSq%gxKp&kvrnZHfS~pYC|i1CxwC
z``m2YA1pnx0)Z;K_l_Lo0h9|pu&~WIrTcBEH>>+qBa%#kdySvHPP$<EJN)hH1}1KQ
zn@dHGe<Qea_s4z*JPm|5DbpIK5dlRVjb$|S%jP^L8S-li)hC&%a79L?x09}n{AfU&
zq1KCWz+;k#`6WX~nKR>3uW!lLP2j@qO|Ek&mG;`#@gBJpzT?ZyM6(<cQ={@5xWxGj
zsrjl6G4YDbBy(Y@;c7c^CO?M4lB|NZthG^G{`%*ZC-bwyRTWbT<yiJ6YRe;^{O^K=
zfaJ`wJ=05?*8K+wQ2yVBfzg!3B71vw0!?kX3&J<>zi!a~t^~hh=uQ8CKJ{xr^FjI<
zX%SnRc$FDfR+@QHdO$W}2&a5d@`MJkqu`yoZ{|9A9s^?+&}ruW4f`X29+#$t(0*ET
zWOeN1MMN((l%VaRcZ*PXF|hP$U3ry!LRg}R<vssIXlSp-B@fQjwAcO4&_X4{4s9qv
zU8lytY)Z?4kzcTskTjKso)dt}BYKacwMK5jh|Ef=Ll&WO5LuYewn(NXfWv<=ORjV|
zS}P0>K;NIv@CiI8x?Lbs7(BSh{S<pZf9k7_%T@5ntKIH5x{pzAKHzzsFjk!%ct>Lb
zf2sQi#N@j!x-$@dwr@uxy#m$%cb}FT$_8Nd0nAk$<~Kb;l5+{mPqQj(2-0bCao8<n
z)_!#-Fyj6K?&<c|Q$KJ~Z<n}wTD)ov<_v;GCL*@)ON_(pM2<o65Mt+ZmI8AZg*?D5
zp6BtA(?vWh_r7x(7P<Ot+jLMXF=D}Ul5-MnGgE1jn;-22kM_B*Z~-X?FT57VMNxvR
zS)GuqFWI8O4^#E2pp}^wr92rN&Ea2^Tt0|Pqv5lMh3l%HFg8LB5DKF)n)cjO&c|hb
z^P$;U_#pGUQ0=m=)m4<bKxj6?W;hABbfpVXUeKunvOrVsY=@auv1s-EF|FJ4&rE0w
zujAcZ(EMlYWxpuwv2hTSe8ZDww*O}MK_Uibx9^YRuVE%3Jr&oY>iXCdxVBcJqfuly
z+$S~Z`fnC@nYrCf3i2{9ryCga%)~nr>hQYaGY%J3{iq`Q51anO=nk!9cQ=>E{Xlq4
z!q`|%9YLKdL@riw{6~*&&hV`Qrv_<Gnk5+%Dp{Xm_xnb#;N&B+&CU`6TCvhNwJg6}
z-%F_w19QOv!@FeaCI*^I#jbfnwq^AzU$VXmnvhcLd6kCm9Gp4G{jHrKJ;L<8`cMAs
zU%g2!68*Xf5vr$nX`KMLCAMGy@35amd(f2@^q%dZicfvEcYqfzAyL@T{oH=0+^E+2
z_7?EN{jiigpSwEWl>aUt*an1w`=*J#{V|#h?I4}z==3zopbg~m{2UG&XzD0ED`4#c
zuszOLx!wC^25i4N9WXeuDZQQ^u12Q46<$b6>47u2`EKDTy=JK02=?EQC2|(=D_GQ~
zQ{+RadNZ}_@g~q&;OkYC*8P?qGvJaB;VIbp^J2j-%jNswwD=FMzq01^BsRa5HIr3*
z`{>-B3(><2trmLV??^e<4)74=C6T3Mi;yI8=pq=~HA3)$oU)s?r*P3lyPX(pms0`=
zImf_vKz|J?LezB5kA~h|+N!6BO3yt)Dbi)#OMG5<mj)9~j8OkmUQ!^-<E9a#8JiI4
zYS@!YT~z+LFD+OoVN}R0(dDF}?6ULXRNH!ZH-DFTc$b-rvs5CT1gJ4g6qVT53-bWB
z!GPk}xcHS}PMDt<AenEo$l;|{(FX>ko(#r}ELUEc5RhYl!7b^U+@CBueh{5hz9y;H
zRMPo;?uiICtTM?hodi04J0eKWBiEl*CH$J!?XseOI%-bRfLkhiWQi)cKy(O=#!-B2
z*-9<XBB6d#-AX<3SzT~ffMZdH*j9ksD591?>iO&RthW79EsLQ4L^)x+(~Nhan`DlL
zQm6e8&jdHgp1h`}6e$Y#Z!mulkb$bVy!7<2^o$YjdGiLV-3qAanI9e=dWbg6HyvM4
z%1!Av<T5OQZNcGUQhj7@a}?e(%;4$q;q7K8PZgd(#vYN8k>=&0^;i>%z&7KXq`np`
zfnpMM6D>0<=EagWQY&!wR{V<Lm{s^*BOx<7tHj2SQD?+Ejo=ZcxRAT3?R*mVb5$0|
zH_{7#0J+lT+vkK*vOy_0Mn<z0E6y!epVsVTU8^z=l~3@y-$QuoWe2w>N?(iTjZA;X
zK;eCaA6dM+=r|ADT8(T?TvqouiO!*Khnu@a&@Hd%Pbh=$e$UXS>nQxCn!-+D*iuU7
zono}i511+<wd5Q*DWGyyh}%%im?fRY?cy-%6cYR7B_F(%8D9So{)0MK4PGO>0#7&)
z=T*-7*a@~cf!neWlyMTHPT564yj}?q&4sS4_ikZb%77t-ljA<vR(=07XZ7qsy<83J
z<ca&qY51`&F{8(9>YkSlspfYV$u^TfoN(}B)=vie9rmUw+tV*dFLe9@DxygOXe3&R
ze*g_OZJcn|+=$+YXWJdo4EdKs*E+z49YeGUf7vwQVZ3;83ZwX8CgsbIAkI`M3c6Q>
z<^8nanNgTmwkxA#Rm(cyMo>YeAJ9SeotPX#8y$>$y6lJGpa#Fb`D}y*_m41bkY@;P
zI42t9g#jv|;UDuq%cSKoC<Zf?o9Rtf!sCl~4-Q&3Xf-(rB}{)GOLA-ab@E*firCbT
z&V<???iO4rK`Iyu=p%7q6|viq=h8slhZKGg_^Wu%jkRNsP?5rGvkG<uhEQr)(1z2^
zM&kI=Qk0TJR5r^1sy|KC%l$%~dsqf9h@I&Qey9hDF$fe8&Wh`W7KY;SQ~9##Jx=hm
z6o1>+@uV^Q$Aw$Db|?=-rX{Go?{~V;Ee)Izud<=<(mFN{>4s19Q6%)vlv*m6Vuy&Q
zulp5JnB*1&LOqsOWPXHT&T8ZcHZL3hq$Ol!RYt>G{A^Sgq7Ye|U~8(7T)HsclA1R3
ztxvbR%;mksF<EQHZj}4*RLYCn`WbsBQg(01z#TORLOl77Cu|@Q0okmDJD926=8tru
zZn%hWYQ%kCS+WaRL@&JcM?>04h!D2hMelkD=A6R1^-XbfeYIcwf3*NVHNevTBQBJW
zPnz}CKd?_XvNmWYy_<SFM<Aj%#m;_?@rJgjOEX|2duno4Yuh+kR~IoAg*wCX=f8zv
zCU^fN{(J(|Jk{TGiLLQVOx^*-rlKQ7)_3<rLJu2WCdYL1qHc61BSs(eq`~ac`#yUC
z{rcU^Z$H_Uu6LJ@e&BTpjyjNSX?Ps0B|1HcNBOo|?SJh_<-vA1Q{7W+4Eyj2ufm`M
znHTs&?de_#+BTm^!rG#<NY~~77!A$fZ0l)ux#=FcbwJ79aCg{*Wi@!@wIu}SbsRvs
z4st10e6EiAazvC8KRcqq!IIvvn16eMupjtiU`X!T?*0rWEb~=y5+Y9JeaUXcs&~Ii
z*zkC}H&gGh=5uv$0!x1fWyZ5(@QOz5SPf2&+olo=<9)irMIIq7vGjpoyXXHnn?Z8-
z3F*ic%6{Fg{aTk3rovWxg&!C9TfeuuD>mWdy0GX&mqCPpsz{iwZJ)eCUo2v77I(XK
zR#Jph|6|C=>Va_(ikYEH4Oft?Xe)HElc|ZfIl=c#R`@LQDa5Smf#S;kLfE!#z2$s5
zcn6&>0xAU=V`05x5t^xtCGQbn1=oB{_&q8kN~i~;rtXDI#71a^ufV>8WWa`(cb{CC
zgXJ?B+ZoXmu9@-FI@(o#I+LuooQogE2x^DB#bwJIDIDAFLsSSiI(QYoj3E#2`kwPx
zXor}3LJacqcQ!~65Pf3aH5bQ68zb8w$!aueDhO41W!_sq=AM$N70RgfzR1q&Cs=rk
z8gih8EF0tp?U@myBZ(P-7D=tk<ar9T;;^~x@ej+qL%kk7O6@W!zH@uRsyiD!r>QY6
z_FLT`WM7F;)M`*f>xrFI&bU?i+3PGa_6<AL_wmMhzCH$^fx)ntz>GUHGqZ+`gf2b?
zB=fqtw`2D<R4|u$tlMkJY|b~se{<!onw~UAJP6=G@BxGS+}vXAcAkUh;#Cwt6JY_~
z`M-YIW!xRLT<QAUuTipFmc-8VJ^B9rgx$kA;xScq4e~jCdySeF*nnAm)3g8PE<s-%
zb#b0b!@V$&v}1BEMS(K=QtyNgBFjNaSVL!3#zvHomEf@Mse#_$r(#vYInUvP5M}kW
zWf0-|6uLwo;S|m)YYHT~Jm;u*Ik%mzoJjvfNvoRTKb_LP*GMi+?JF}Wi9L>OFTRo7
zkrDQ}Pbt}3kTb@i_kZc5<PCp?DIDX+Z9HQf<)O{VKmDQ6cQjk@N>jFmDrxVK#QijS
zhAnc9Un)#{iGTQ!gbs@i$@S0@G70%C?BqsR#xMcrE9i`1M4l0!Zg3{w@5X0T6wCGT
zvd{xM@xHE@*k%J%t_f;hEZ9h)Dq;fBD(@vw+9OTnWr%MnyKtNv6RX?>P1PY1s0Be3
z6r+T0Dq>N_^afdpFJFgXM_YyQA{!xP_G>^7^}o`qomYH_m-&7i*F`NIE>V<q6^JA8
zDwo(w?FvN&fpGZ?+1C0i#OCuq2D*;z9NMOaF11RQeyf}vxv__a6M9b<dQ~n5T0VPq
zK4B&U2`lat-9#XQc(_>a!4fVV1UaTJYZ5$eM?HRrrJ`<OJuYKi_p7)5*_|xr>^>gQ
zdxU~6F!!wK55e9^l&>%xY)S~$Tf$I{XeBNI0hkNB?jnGjDb7zwd99a0K}laX?}!e|
zE4V_(<0HXbyeM-$Nd(ayu;$SVtd1>MPXR*47^k|iM9#7QbAX2dCHs{FU$cz6VFPZ2
zt;a}gP5WG+bKspS)ef#nfJH&Wq%px4f{JSS5?`-O*ga?N`$k}^8l-u6{Fm6I5qASJ
znQ!k8U^~?z;x;Y8RuF#`cUFLJW}>%5s+FNlA$fj7+F*YxtrE17fpdy$v5ylqec}Gr
z0L7J$H@$bWdyTmN;`0Jup9VvAEOC+Fs_{Q6+^wJ)jjG~!u9<;<g1F-!SPN?ha|9er
z83PBdACuKj?F#ApFlV(a{iG(@;cG*9_bo|l<$FFiXUX!MZ|8LVs0yJU#)L-|dFQ)c
z)l85&HS|E%BOp3|+wE|fXdL8X=w>fyA_dtralb3KJX)Oh#hhDr%<<E;><y<srDeX8
z#ZVW)_D9J#YAiIRpTZD_u7oP`s2DdG2o5vdJX3m}FbPlbW1H*cTfO+3AnRb7$)I2}
zoZ68*j08MyPm_YcbbT9|<&{E;#&siBA}ds3KQqU&q#Hc-n#cHFQRqR!_pj{g83Xx2
zW$l(xcRilf&{U5?Ip1}fZ~71Vy-->Ezepd*dK|ueh1_P}D7iCa!y<<2K;eke><a?5
z@4NKwyBt!-{93?_;syZDC}AkG)l|vMRx3=>1cKxuVR<cK6h|OmXe*StY#FQ=3K$v!
zwiFu58`NoC&x2=_PhizSVHqm0@iW*13^OXhod1VV43<DY$q6zo?v|c>V9{L^_9d|8
zb$5R?#J4oa?sI|;gcH5D7kgI?M*ykV++g<=@C8jx)AqCnm*B>eB`o#G3k1p_FT-9R
zR{S=A1^3Z%WnkO$DeGzi%RxcG<!~u8q5u6LdDC}Ot$YN<=Lr+=30FwcRL6KZg)8o{
z8A2rvysGT0-Kus5p3G+I%to%vAO#3&(IGa@*&?42%}!)xP?UH!>OAO;D{?ryup*_c
zeeWz4E96@HZ)R|NBS_q;oB+*ucJu=eP)u3A%c1pyHv)XTM^Yxc%0xd!mB<jg)ZDz#
zn9;cLV5BMVTBueYx}1sA`*2Zj*DU80lcr1F;}H4nza5FbkyuEZIAlr%A*2S1W_MjK
z>l}eZBc7f`@OoHVYxA#(s*h35U@r4(jCilSMgbvHWtSR=X(MriUm?GdI<v`u{NMww
zjt|cy_=Aow!SUy|X%PP#x4I();-htWXJyOGd4_s)j>)0Eh>ertJB_ies*(|-FEoY?
z87E^Ee5ZFqBL}$iNT}BRa$0(;0YYUoI|`76ff=sY$}fT&hYREs$%iM$^0E-NZWpk6
zUW~xL^fRDYjem3B@n$_7GW?cF0)$=>kGLVjjucQ8(8SAbh%4fLRVwTR#pQKz+by%&
zFWm^G&>qpt3NMhswh1y#UIJ%Bp^}qt*z-Pj2QVgU<^F6$O13W)pSd0gU+ckga~rbK
zbsFu-0C)$8vLzROx*5rDYojdX(M&kN9eDZvxb><f3IJ#)-F&Sv=<Lndxp*~N5g<-0
zj<4-z=)WKc3q{XaQ6u$Y)x*|1ThU`bEc;=Bt)FqL;I;l^0AT?|{7?61SbH_;;on5M
zvUR~CW~BtlbkD9q@)oaNoB*o-jq_Q{lE3iC*7)z~63q?}9T9o-O_6HBmeA_pK2g1R
zl3E6u&HGq8^UbrnNyY2S=wI}vIbst%#}z17MC?zo=A%0>S^<&~de83b-#n?Xo2_XC
zkGTXP9cFx=X)!{#YdVt%qv&|$wlkw=QZUx@amJ`%b)1sP&?f0{!x?u7R0^;of%)!*
zZgDN}`}UQbJ>Yd9qddP3kg3_o+@V-YOEt}ssWwrEtlVjv&dl^}V~I?iqadam7o}9S
zBt=_FD}S)Ij@dYnLGBlov5oFmf<j|FI&U+-hce6n6G(J?Grsp~e1E(xMKsaJ9TXtL
zn&Xs?i}r!bsIt6n=~CHa{012@u^;dxREkyM1k+x$=pG@CY<JF`5htakiR_mu=V$3X
zg`-k2CJr#Tsr4jiFQsZgAds^!t5Dj^Nz1fl;xsa36qFs8CWdv-S$NW(#(BQQ>_jWc
zIpUOjoOJu8rIXZ!48e0a_{OY3=4tPkNL7~7AYs#j&OkXiXGn)RvzaHeNnYak5C%a2
z&$7(uTai;n_fVdQXvuYDG`eEsJvUhn6{S>F=Hx|+vO9VY@bpHiBAv0(Nm<-+73E)e
z4N77wD%qFQD<}@Qdt&_ud`}aFeReO!A|C?!O}xy9Fz?mw#nQRXj!Gga-zKG+e&HFa
zN+0Ua|J;!zw;L?76+lczmNUIKcHk1VTmMAPl)EX^ly4zhH6z>(5fs3{z`*XQxC@6<
zl0`;FhU={iSICDv)orzfg@lCQiUiVdadSIo<uCIGA9PAJMOFaRZtkq^DzY~WVV``d
zX{Y!4`nn?J$63K-c9z4eIdd*aFOqM5NG<s2(yaUkdq<|TVB{rDJeKrb7ONaH1tA+N
z=U?GwT+WcL0S{Wuc(szk+vm8iOPIKdL=>SGo}DwVrKtH5hLlBWd6fUob7-7?dfwe6
zFf<r#DYh=RI={+1id;!%Xd_*kn__uy9n?FJ#EsB7B_@L!-x(TL$w18LDPN0QJEHYE
z-PmHXhbSy{&XEZbBd+=8FJX3Woa)7KeG5wDTI&DKzR9GykntgI2h#$n&`0T$0DSfp
zi#0v!iC|O$&FjR>M1R7M718D$hc1%7IX`wctdsYtu!*cVHLyp=gM@2Pnfm0r`$hZg
zC(SHovg7|mnktoG{^4qWR=I~oH=b9J9|q1;Bi=q4$1vW6w9+_n(p#~vw=#ZkJRV0`
z{!MU-yYub69d^ql!>yPk#V14q3_g+`Pj5KAhf2?X32uGgJmd%U-|&f3KZ@Jc@JF<c
z4)kiWA#3HW?LBmnpaIbXm8$Icijv*61J${b6k<ssV(<mis6Kt)O;3AXQWNgN<-((!
zTu3z*WUCcN`|n>LQ5oU{asv6(4@qgpNuq<i@GmSvb4&<TqGEJWsx@$`O42L`>(?e3
zk+sS$b&%?cl6~}-@bpc@L%Nc*NHwDjW64QGN=8YcPNr*QBb}c`JV{3^e9|^<Y~i@P
z9;=+0D33e^uH4N2$~!8>I;r`h*&r>@n&qWbl_vciU@hL2Xb2NQmZi`(4ON&fj<j73
zTjPJRJjEAzwx?tx^C4ztc8CQ5UuIiQ{L7E1j0-j0v*Nl}2ag-EmvMKunl}z!KFcQS
z4ejtr6*-?2Abm5!@f@@L1>ZvBYp^47Q%I^&(&4cNtPOwcTUOV;LcEav`>wnxX>@>w
z+HNIsH8?Bh6;DENo`?KJ=&~*Mv`&4BMV^O1CJq?#dPX61+0sGsG#$|7iEc_k{ct_`
z<INM6@uW$W0PF(v?Z;sq^O9h?9I^THYVF-%5HJyzy-3d6xj)cj=Q>FBn&N)%sEo)_
zH%x!6^pn<@<acEWPI(r)e!Zh{*sgN>rCRq?k|Da@Oze7h^xP_~KJU~zFV!isg__#<
z^h{#`=Mj@mr#$_Rv4mKqhM76~f=YITi92g2I$aO5?=*aDOKOtt@UA5P(mZP)>%zQJ
zB(0|UxC{nFO}US@v=f_lt=V^<wM33`-b^?noZ-r+)1-y+Y`E3xM(aRe@Eilz;L}tV
z0h8n|)xsb^X3^sGb1tQso--lhNQcw&xn!VY!Mcn9*qPpXd7m5M6o5@&fRv8Nw9F7c
z;I6;k((H3DRx)&{v$uZjRp@}vchv@^4g5?hE35C#Avr3OnP$O-KKq`TIk^nY;-%y(
ziI><fgk@5#KyNs32ajVTU6PCAcI_mTa!<oJhxb|rGMw}Q^C8C2eTl|uR)x?irm4q+
zJ2tiT?a9Pqq|^!9p(3~I_)VRwlIICZ<ZeB9bCGLgI{98-uR;{PsA0V4M{5}}Vq$^q
z6Hl{F7;n_$XgU08S*Cp^fbRBv*W2~6KJM<XTWFT8#uhw3pJ)r3JhSAl99_6^yPKFU
zdW|Y$|H_6*QqJO;u+e)8<Q9!w2dR%;bvzyYTs!D<yvg5H9i*3t<x+g-%cNfbo(x0V
z6t}xkpp8>0mw+teO0+MmWc_#S9ilSkvYecZdw2>N6_SJDG41#H<H!oqOZ3CT4+d^>
zXOi?CCe|2~J+(7&CPhQ;?TE~c4?p}MGmG)^JputHb{V$g1YuvjxjbPd6+vCAL#lT?
zf`mYH&a8fkgzR=+&`w!i?4~ul&Tk$X$~DdwPmPRuMe#ocZ~k=8(#b1HS&5m3K~W2X
zs1yxKP1KtTnwU9=Xe@FC4pgwE{qVaaJ{uv$yzUna2l>*@p_^94!OTFVhetcRO!+Pe
znJVl;oqh7|g6F%i4Db!H%|(62Y)O@?w8Pkv9kso1<8I0B#c+M+QhKxzG{es@*_E76
z#<-j5NLUNrE!h?#{|1eoy3Uu$d$iGmx}{-e5ULnJ@l+AzZT)x&5^XDu1o^hMqv0hj
zpA8gA)PCcY2<lvHuN|KRU21cz@+n?+Y21lzBzq<SUf9u~hND%|{-&^fw<>q7cVc_#
zQaaZ9q0PuhtCla@5;{dNRKhe&Twngg6KvCky4}v~pknNPqhJzMQ=x&bv_A*=n>b47
zZXWwHBA3jK@tGg=GyjDRoBWZ;`ak%AFIALl+3$WcyDa{ZIAISH@_8XCVXf^+m%?_7
zJ3Fb05f?i0xhTxF7{{&N36zPENdH}lK-p1L{K>bRyo7NFImsusvg~})=;`_O`Oczk
zs3SC!1Wa!?;Hg3{j1ow}n5&`^rWeV^*B#mpt@8JVe*6%OxO+qOxIhnD&f0%G)x~L9
zep=g~F!8NWWY4)U@%G|qVm1;!%Bo!;h1WnVGqcvws{PW241*IG5b2hv5%<f|1_Gr{
zyjkgA#vZg{`A#DL)Nr-2q7QfFN)DkP3fp44fqE2ZOnxU4j=*IBZ<=gN99$56Fl`m~
zJ!{qSjCAoB+0@L;FR+ed5!9a`6SC_v`|;5y9dxjOZ5ND6#+INLqx;BKFE{2~5^w(Z
zGh}keP5*UX;s)rwY)az}hT?2ZH8j!7e?KUrEqUL2{+K!Tt?+dttUkV(-tg^2xz&qe
z@(9x}r)|iL-wgla2nzh%-(vk)?F$-wb9Slnq}^Mxmexr!01p`tr_?Fc)kE`A97Iy1
z{DcM+teJR->A#60pis&6lenomERNa7sXT4>t4KZ(91xE1WiWY2Z#~qviNvlMl|UQa
zC@s9ba^$N}1Nfjs4tu0MGfVv*y)y`OpavJ2WoENxmat`!#PiK&_WOg2!?a0KW+`XO
zl2of<a3Ugu$Jp3lDh0_))MAOYl1n9q&VNJ~20J#gZ0I8tpje8pGGDr<&V1LHrgWHk
z8zQ+r_5m~I*XG_PMcJT+Wv1a58e`4L{CJYXlx@c=jYLm|#?FS!w)Hj7&=ZC3wqQkC
zwaM(eZg{YH>lzuHIS&RXa|$I}RFu7TpLj0;)vg`P6XuvPDye!R0bpp4L#okMT$HHF
zYkbkOq>*DF!H?eAu>kl&55O-(EB9k0ddOYp+hMYNfQNS9WWW*<1@`mkBAvL#YEe&l
zaXT(e|5!s~))Tm^jHj|#RfPNAb}@YH;_Y_9<9LYMb6?(9=kVT2EaayAqa2zRUv;hd
z&?g*yiCd?H4V}mPAo>kE&Ras{v|Quw#9u-;P^Y<>E&V<)W;dIY|0AC|J`gJSdlG8O
zLcmgp%z_rh(n9wUs`V=7{ip)Z<WFhxGUP5aRo1u(u?&Sg8AmyGK9(a9=N}`E%MF1i
zv}QgBllxv9e8fXy@Mv<cMa4FhADO2dQgJ&=?a0i0=@cssazK^TFS4f7Jx7d+`i(Zj
zsyu--3`$_}dw6pDBDaf(Lk*kJ4{kPuY5+5xLB>(c5`)o?olZ%^GKP98L80i=Tf@+{
z2eV6*IE0X&jPdv${*S#K>UVr@+8FN~R!qyuM*A?*%TZ}FE2WvrzjlodVZ<oKSq04d
z3}-lOi13lw4+gqel85f=A^kpd=`T~0<*O^P_;weURfHU>Yx{4e|0B-r9~vxcA=s(o
z<fag|-GQ!q3=g`$s1q159TI|82BsUm51%fyoMB`H3|Wj^!$=#~Vds-~=4wo8|3^(z
z!>)pnATsfC!k%Xe$1K^U$X&mv$C<syFy{SFBsUyk#44F?dFaTS4Bi9{(xQRk^}D@a
zD`g8hIo~RwL@d;NsOaxO{o47QQNe(vp;j8HMml;k;WMO;wSr@D$lboEr_4pNi@DM~
z7ctJ*l{qtMeGoB-r-r6@p|*sCgVf4}8)E$7<7F6bD4l9ldvA84CYRB7pHhV*L-O61
zKAwc93>iNkJEtphRvjdLw|9>OBc0@7AoKn&(7BVdosFm2#9>RLHKj1+8A)U2G-Sq-
zIvQ0-z7mg#x0L@R@@C+j{zoPK(pNY*<V2_Ss4uM04^)gc%kFq%vz`KwhWD{FX9&TO
zRFoK791cO2D=7%xu6Gju1&q0<70YdS0qi_RFoh+7`b^nj_<Y`$$qMl=!nYMdnvp=a
zgeCXmBW!e?p9KDQ0Z#l34R)WLjwOf1E`D`t47f3v73FK{U=gy2sh-k;3Zf_lve>7g
z>Pi^*-yZ13Z^KsD3W7mRL5Ge~`pFfj{$L{xkJnL8*ZjkS0%@97Gc)r4_@_1&Pga+_
zL1-QMp5w@Tn@i($IF?bEhqY_eWxZKRPm7TVMS2K;)c+1~FR0Lzn=6xPXP{ROcMGIV
za4Zb;RRep@#=_i03wlA~6f=)~^IUlb;s#DPCE&i%#^BWX-ZXUi7dxG-EF!Xy>t(F@
z0BEpvO2bVi4VjGuqQ%p5DBMfP8DlIuhSmDZrebV%ilr!ucc}*MyueVkP%;Q7q&7b+
zq+a&ylMd`DEhsn0d7tky^-^S{on`5=O~6hj=|apzoxal{b?wi8cUIvTMm#=YluCoM
zdj#e{amjuZfcm)VI{JPXVj7jCVNGd0n<a77`CKsd^;hU$^a!YWoHIwzAi-54f`knI
zr{NXLx7q8Q#9PAru^@__7~$m}GP8<5jh!q9hx+4kX3~!k-F|3j6plY_utLd>agc7_
zr{{v2!)0pywWfav_tei0(d{-+I!27@+ET^iE!5TQM~b%2hR^IakEWU<<jI7u8yUDi
zyAMCX{Y=zuETlh0M3Eaj2*XPoq?jU^`fpp5+N$ndUND+0QiZ9E@7LZu(!i8QfZjJR
z)ogTTDpvm`O0Q!^zNHUZ699gK?gsP{0UL01Vf)Y>@hTI+cclDg_2N6&+Vj;F*6E=$
z3w=NKjlZRf8_Cp|yzWc-z!H=H7H<wfEpG3r5wFvK#*^gqUoUoi$ZI0jPGQPqU__yq
zdfwQOIVs113?2N4MVtGxca>E}rca#+VQ7Mb)GS;DA&9vvq{B_fq%!4G)P(;pvFGif
zt>5(yx4GCbX?UK#S6WM1KFee%Kn^u6k*Utyl~cP6R6&Sx3YHU>jnNh>UkN-e6?$<e
z@S;1#W0VIU9B1`Tu;VOI95@6fa%d<yD@Q$k<jh3u)+)i5epeBT>W34p=~qKQg+vfm
zkuZ<P^C+emL;4N}PbiH;;+4kpc7YBU7b87nmc!mjBKHajVKUlD>Xz{)l>0-}b<G5g
z1YI;LVp?Z8kKzv*$ZUv-x)3OTAv8!4(p;ZklxMkph%s}2Tedn(Hy!ihPoP{s*bU_`
z99{TuyL+#Dlb;g#gzsH)bojBEzn;tdVi10sfirHPsv|hU7K)3*CQ>$*`QJ|Im6{Oq
z{S~T4lMAcWbeYJQvV($=gYCaDS)v2ulI=6JtNv>Ykv=X(u1TO^=x;TC7;RS^t>aS$
z9IIfG=@*(#CF9L0v92<gZgD=Eb8h^lY%=M}`BC0vX<3gX2^l|Zir53*mU$&iJcCzb
zY8)z;AvKu<s1&_MP%@8M97{hH|B1H=B;^vUjVjvTzL|J;smn7!oRz#sZ9J5o7a5FA
z<iuL;j>UM3V8(gfW@jxPrsr9iy)dsGqL+7u6!An=o)YO@d1VQ`d1thnju83fAJ6W(
z#ZCM!;I&|PSpKLUTcxg<Q3N|(+Tb$!yS~_K94<6jjdY?X%CbYRh_Fh}g6$XCmfwZ`
z$btUhs8WZOrjIq2T#T+7F+GQb5uR0Zj`B2)=ZlvjhYFg84wkt%MIrCon1x+W3c9x5
z`N?jA_B9%oh3cIBal+RwwT2Ww!jr<C5Di<a)e}e!F%RT%-6RkiEE#vcfVm?jG&<ac
zaDF~>W$t+NoMna@hLgh!zEjUR(ZSin)(w)?epF>e58UF6`^uY)oAgdx|G^&t!|K{$
zaZ+kH1OMG^G;3KeBc1%^T_B)Ki;xC5>cZUHzO@xv1A3-pX#9O()SgLDh3zLr*MrX;
zVS^ftN$*~Va3=I#!3AJ{t%(zp!IxnevY?Z&(IVGyL^otBrW>Lf<~h(Du<tS|$BEuN
zKDK_($J6N^B=rgMPl?;*?Axp7h$e~^HuNi;0VdU~1b4w&+sss>jYL~x*#0+?R&Wn7
zfBsSHH?<^vJf#JBfI3m$lp_y286WLL>zDZnD(nv$5sniwuD0p_2a06#V!MO;an|B9
zcXEVN3AP4Wj1p?v$>vVmM*nc#rThGIk^ZnIy)~p3J4f%OjFZw`Wtr+Ad1@!IWi(-Q
zGj8hiFD^!HUk?^Hk~|s=EGw1c+C!D0x>1&|h+lhhzc{0=jq}&F-OfX1UfK7<$;1&B
zPI*P=gt@>CYSFhc{eINr3G$>xnNnGpV5##&V@!Kh&S!~()oZ4`25-Cngpxpub(XTo
zt9HTEI5}xNz-IWtH|ep(g+5;RodEW}E&jVzgwnm_kW7;A-cyGE>v6RKr&Kq+H?$-=
z!Q0g#Fy%lIVQac=k`S}oF2}BdAi>AlNJGi^drqEvh@YJ#aVX8&&4~d)IW4?Z`crL!
zF12(G-8o`<94dODUnfzHUs40^o_4V1|I-3Uu4jUOpnZ?}El~RTd%c%#5+ZZ9rsY#m
z8W~6gHjbkIDWb_pcG~9tj}J*9XIX_Q*v7M~#Rw?w5S#Do?R$;nXjF(+yHs?FO6Yi{
z;*i<CZIy)6appH*w(8N3C#6kXfCA+=Y9BjuTYEmTZu<a(?aJz*FYB&*mYlAaGXuD8
zPT02IYy9ZNiNeFgY@GO8)k;Ig>@?E}EL9p9zPx8`u-18~j`-+@H@i*SaLY>W*Wa^Q
z+#nydqQbZvXTK(IY<#ZIbskeNS5vp^vfsmyJwJ`8q|e;d15uVkgRNfx??XR9;)-Q?
zxQi0d_T0$HO1zr+W6Ykv$_$#URV$YoWfpAIcdXHjxjznx8!Z%;nbrD1#?lqe_KyQc
zB2_I#TSRh<wOmNO>+xs7i|EOVud0*L5R@!$7cLnp4JEW65tvO{OapiR56RD5Asx<q
zMyfRW^{bTKc0>O=#MS^(boj3#Af8zr{ur9r{B6YNvI<4y@yjny@{H?F2DpEfCg^~f
z5|~nUR#5IHh!B~02cHh7v~L0DxGev70<Xq}bcH4fXX;;8E1ACvsTbXG+^g|CMxAJL
zLmP{30c7c#F6CtdutwvWslOMON@!?kVTF9tFG^YK!=pXFS@c>E^!rT+inHk;HF$8W
zHEE$;^yZViHjPcqJpB{)Ft6d!U{cOfa4q#_O$=2->w>uKx*D3r)aS-W^O6ed^F6Xp
z_2E<ZU}Oc$gmu-!JjBSHK3NfE+HrAndSxzPdIUm4_)_HXD)P1&009K60KRk5hu4_<
z{vFaGUvd&Jjv96(LlG;|8#b#aWKK-~43)($84zrk3;suPh&VPZva*gg4X({S$&%9i
z4wQZBi?X-Bu`Y>?3L5<nAxqtra=QM*-h!C`(36})Yp7tPdZE>5X;<1AZJHn*ax-k!
zO+eR@)zJ7S06$~GT$oBz6^jY!dALdSf{A-;?6mCzQ5)x7hj+IP%7d5PLGiWafxFvt
z41Xw2_yqhg$@fb%{iBNN-Ejq;S*_aPya@z}ZFIolJ6<StgnIl=_7Hcj{`F5N@@#2%
zbaXT%-G42WSK+3u5ylU1*|lx*{1R9{`Q<TLylLmVBc~@w@8n1Zr$?ZD1KnSygTpa8
zi5E8CEdGhQd8qS}73Fm+_3^7=mZ}OW6=>^QC3N9$U)7+vR_IQc5OB8j@ysO?$1bF{
ze#BvtVIG*_i5->aP8xHDQlsZ3623FEVv(H7s!h=^>rOG4v;H%c^ewZpN53yWo=6Mb
zS!Jrb(mSt5f5xbOH#^c|h<Cu4Jv))qBgDX!frf|VqjepESDi>NyBW(+L&IVh$zt2C
zhmvZt=lNqY!WOh`9~X6lOZER!z%R=7<tQMZO*<-|n_Wnq2U36$Q=v3F_^{%NHm;>U
z+=B;hYKbm&qTjI`V;^pIhF`SZ?$nt-E>vBi12@|P^7Exjl0S9}$GxfWeLy%9V%vSF
zP(h#>Zd-IB()Uq7pMH496q$;}L=oE!K19DcO$k$2wlzoh6cx$ZF1KX&F%(PG>%ga7
z=c}dQ(2=zV<u9*myYF$YrUjT1fE33YSBJjw#wRM~y}hY#rzXNn<~w+n1k;io)1XQQ
z0jF2S3vc^OKo0|eS_c<^l(zvAcxY*{+@OOhstrg4z@-8zpf|8f_0?nj6FTj<gUfFt
zE*)b}U?+U=>I-1vv(}krcWBTZ(AH~FUnIiIXrd*{L{9k+nk(6S#G7f?d_e&86NpM&
zlQP6ddDeEO=Ug4uBOHqZ7P&z#B|DoWHkb`yo@UnnB|+F*s{s6j2~>I^;wCNP^th5T
z6j1&o?Sf6UICxr?2X`S3CBJckI%=aPI!4PLZ)rn{cD6D*={O)bWxk@O>vHCYU#QeN
z{<9;koPtEWBsTaXx;93mF~Zlpmv3TRb)1>s=$%ln+Uxrap6m)9Hea*<bG+C(AsvY9
z`~UZ&IUw-*PM;^16jbKyS7cFCh<Y!VcWaC_P};m2Bhh>A#q>$ZFr`WJ=ECZwyJNHz
zdh?MF+|lB%;it~nQA+p0bAaZc=pr0`cwTVr{|Ghwxjr81M3Kx!a&cKr1qt$#eL3;L
z5#*ibrV+nSIqnW#dOunHg1;8Y$7z-HzPlr%ixBw0&4!qssSN(2_pnY`c)DO>0!2@Q
zzD&~D)?*r<N^+aXN5KugkpHfl%vRjf-W`yMNFNM(wD|O{S=G$mA7nh;Z}k`}L?Vpp
zoe~vYS1WaB2!X|Z6#D@>ghU(MPa%gDGdWB6nuIQ@aEH1$d!H~ZOxdQKKy6Y?Qk$u;
zhIV#zh_%~4uUG$*EH+bDXsDu4a3Ozoyv8a^E)L#;&@K>pd4;R}64T^1pTpLBPOS%K
zc9l@SedXuL>5kq=HZ@FAaQw~nkxf@@MjZgQEh<JYRR;WdB5(#QsSwiX<*weunA0Vz
z6>ZgVtHH~NoD#%)BXCUh{achLw#f%9o=J0BMGgskP^X3u4{@x;hdj(J){uV-veAFU
zWP!OFKwBX)y@|Jr;t&2$et$XNdz9?{;?O|T(2ghw%)h_j>Tyt@h8>q-<KMw?$pV>G
zNZPMC8ElthduaT){_T0V>I9999wRY^2;00D^J@|N<?4_7cW8D`R`r`q5-p6O@{40G
zlBw)vF;+C^pd$&fRgl3)q78L?>HiiL<Kr2A6*9FI(@m(V_cO)$6P}>6ih#%#GuDgh
z(x=}1VQ;3^Xo;_CzlYU$m+EVM{eIf4%Vh0VG;kdK$4RXSY#=m`kB^_#<qA8)nI75J
zd%l@+h~#BLG&@i1jWB`H`_8}>jZt%=<6t8wvvMS~^{Yv6ujNIKzKPyj_VIm7lP<MC
zA6H}m;Jt613t&VXH))o%iJuk5jn{*Oly_OkW$lI$nQ3@mywyV;ba|YZ6)k-eKN5cM
z;PUd(@PV(xXoW5~@-QLm$%NM+roSZ)?1Ii1YOkcfcfIJ%CU~7(k(Wps9@r`y64Ndi
z!(dT2x1TluH6Q4JaDq+O1sbdUF{@XDg2l7MR$S%9sGDsRfb#%$*QBKxle^D6OcbZ$
z6{S|R^99n_ke6a<nVFMF`aV}fOVzlWW{7o21z^}|DzT*Pp^t)c%%JBhy!LOel)53B
zzyx9g0c;qE+{JtcZT(I25jv(XVAXWRep`_?`x+Ep9_7vK;-q4xtOK<VJy|^;aj}*^
zyb%|*r-y^Dupi1t(D;P4!|r}>=$Z&+FF>X!mIL|Y%#-1D5SLmmQ)p_8<!>WoxocVb
z!}Tk|2QvHxYTs=J>*wk6%#$SrC@f51Vu6c<&aNHq7mPXKXNXc3-LUS1(w)`CuIp<T
zJlDLGb}Jk+CyNyj7^a9};5RE6oCz>dn-ZheqYnHDAfWvyeov0oaLMWd`AH!6OG2Pi
zoAOq&aWfY*$if{D$22T^DJ7zEp6E>yBSlNs4L?PY`-*EoUd`E4A>!3kLRF*dSzB#u
zQ2vNdz}q)}{T3M3m==!CDh{SAGcPinA<7f0HOoxzmeJ4WL*%yc+ZRQ0u(@R^XllOV
zdq0ya0vUY#(0ThOx2PdyPTaKc(OB-RtQ&1epS%9+&Bu@z2S}E;hcT8Nmovs?<5h}0
zABn6&;(}`lY+gY)fPTO<q3Gxn7H$07OiY$yJQHq~1O+u=MH(fRQdrMVBCB>DShB1A
zITjB<vDsZP^uz(jW@b!Qrh28s?>j5ah{m~*GVz%97l96wjF@Ca5tb>7^sh>0cLqaZ
zL$)bx<Xb_bxaho`+TYm?3dlZU(v6GgaPQ#5D?Z=#lM*@Kl(WNB^^$e6O7e5tG8n3O
zv)ET=+RqMT1CZI=SOOk@?Lsz^bJ)HY>GW4{gS|_11x-;O;8k$ev(*+EB+lqF&F*LS
z2Vj1C|Htr<0~|aZfA$+=i3~+P=1MJZrUJy!jsqUv&zA*A_}xOpRWifId?w|oRh1(>
zSE0%ExGkv)TE?S6>Ayp>n4PzoE1S*#C0L1xP;Lb%+vShc`UR%MAo8x<p12-29%qN&
zSoM|uOd*A;WaqH#!ZQRlxwvkli{!At7g^xY<WN5c>Lbs%0hf$@uCJw4XhPa(>Db{D
z4^oF$blx!l&@_{RFsPtyy-18Ud?Hr3;)Yb@5in6Jay4yj$w&$fTbKo@kj`2-RB}%;
z!~d4CMa{{iMUqR5BepV=cTv&;4k*eKznSka)V<Uu;G2zR@rE4$O8noR2khFC_H&$?
zvRRtcMbk>Lmj3J>=81`323zUMXp|~P@T6l8b_x{MrjF@#7ob`QhOrR_5<VL&<>N|F
zMIaelRp^>61Tj3ryI91p@4&!wo~P%94z@pp<Rt;W8Ka>lkAC_3$eS(0zMtaj%+-Ox
zdAP!K8lTX69p#EJ{T!R(|HBiFzkqTA7G5a|22!2@Js98d4KP9b+0zDwkW<?p{0W<v
zVb|!AfKlcrg}0LTx(qmd)Mlf>(*a;(3QH1|cA!|mXdyK?->8-kHyr;wo9Ku>$6X<J
zopWZuJoor^W@L<GNK91ET=~&$lR)&}l}mWdU^ht^9S-)OcgH;6%h=&a$IBK<vyEE2
zvF*cd>K)87v%R3iG4Jl0)EaAhpFxG9q+NW5PsLX+4i9tN*2e~v0j&+G&@IoD0#OQ1
z9UmL&*&+U{5DCd&y#A831>&KW%P+T($wwS_0y0+r7S0ie*9BlZ-%WA3Y&yAb&%aqV
zTNADM$o2U|z*oj(EQt>+RjCvhIlKzAVB&lXLk3wXoUvMqZMO%_CTAI(sRwLj`s~rE
zw`-62CrOV>M)u7G;~AD>BB%VtZJopqFBewO3lHu`x<0&ry}IJmS7YMJkWIoE#4K84
z@9=C~L5;h<Mc)EH8&lg%*%WH1kPf}Z@hGI4OmG1TY94$&u37#l%d>WgHg)aSdrUIz
z`UL=10>3BmfZct|T226C?SKN+p+C{wZSgE!3J+iXSKDzQyZu@OpW(qD?Qqx_uWieP
z8l2bCRu_=!Ii%#Tk=Z0w09#|<bxv|CVI<H@I0eg=P_lhG!GDDomsZ@)ob(~I+gv8$
zWN`6{;6j+d?em?~_YyfMDYYUsHjwOSoEA#r)Xuk$w_bWpE%-8ja^-`A7jr)qy|rsg
zzZO$8I19ik-|`)D(K4RA476-tKBVwA_(WJ2n}l@4M-a{=%Q??CwlSQ>3iNMleuY0w
zp`f9j;fx&{;=V|_yNS(s6m4rcOQ5bt*Js_B3D&sN^X;*F+zYFJd+-wnv0f(>Xjshu
zyH&DT&mF<4a)yYp*9`{IG@s3<TrZ9~_4%?cnaE-uU&-l>1Shrynp*ywk)IKa<sc=!
zh3|5gd(BdOyDxZGBr#$zSG2V07ddjNCMlv)vWu>%qvz$G|Cw;eAs!oS@@y1^_BXqt
zWg6o>v1*eVlZ2xKm_}(ZJUckgh*z=a-ns9L;DPKakxidlBLZ9EY{5_Iw=;uLNvskY
z{uq@91?&)tbTUjkC2rnuO!9r)^4%;d4wd+~r}yMx+ps|ZjHv`E&VKZ~`nLRdceK?(
z`N#o0LZh}QUC)O2#;2!Uv+@r3r6gDxEe~b&tK&amZhSW42t1PUSV$fG8s|S6JJ{_$
z7Cz)m%pLZBXu9g4DF3gkl%z;^vvh}mva~b;ONVqX(xP;y<kH>UDWG(BHwZ{~cfAka
z-@G$0!ynAfK5_5o-gC}97urGdYILjgeUA0j3^s|}26HS!a_eRe=E(*{vyaGkO#VRO
zrs$3Q71O;YTgWNc!8o)UtG!*3&YiLEtNp*4sJ0kNgLTam)=AFA6e0+6pe(tzhz&~%
z!ZgYD==*2HOY1%FQ8k<6)n6^w&vw0zt$il+nQ3gPya&Jq3N`=~3QGib)Gi=t+wDpg
z<g+kPpJ{N}Bx(FMzpqL1j2c%z4+?5fEYEBGr-JEm*OC3?#e51FM)8h?aXU>LUnglm
zejmS=l=N~M+g8l+d@uZ@fS&Lo^byC)0O5JHPe9)MV-CUoWa|UI6-Y-=ly&~4Lia(S
zOciH4UQ%m<!9{wtWv|=4_4${}*~RnU$z$g=fdTInjR!{AklSh%kUmEY;S5v1LBEbW
zBPtLxK!q7RN&Vq8o1i)PMIWmR$Yx(X0G<i!cojdoUq^b4y8Z10^w($Wyx?8|I(HB<
z_!Bq^=$bZA7K-xYEV#aEs$)oVcdDiuhk;};n&op?`)H((u4AdWk;+(QWr?S~xsk-u
zPrZa3##WCW*1xX|-*`RHJWc9{(<91;!Xw#S_l0%id@}wOpz0w*&uLA1I5CXj^Jcg!
zo7F@+hA=UNw_@Q^hWHK2(UJ1ek=X{pa~qgIALx1ugdn1gcfWxcpnQpeAayjbnBs&i
zZ`sfguX$Z9m)v~J#Xcia%yMw90u%QgbP(CJ4;I=hAP$l&e%+(}UZKm{lJ3F1R!*N*
z8FKna|0kp1VMBps{?bn>-L3+LS>NJYHu}hKUj8Ad<{rVrM*`XBeSC=C=a17HJ^%W%
z+ihr6(aK*iC444euH)_+wZ0xQtVwTP8>4&)T2ISXoe53OBgMt|DqOWUTGeCMeeet4
z-D{z`XEC8U?#*LOLz{}fEVg;zJZ*1Gw@qmt-gDv8=a)3rD{%Mu=PUNM7n>G`%{Yy-
z7f+R!acDI17Y~b5KEOz&RKN9!e{{q?g8VD1%JRQlLprtrOzGF==n_si0AJeZQ(XZl
z_2u!yN6lP6!DFNFn1(3x4ne>zh1>F`ZEUs(5YZhJpmdwZk<f~X``G&Xrg7y~%hSGn
z@YKIeTaKHTCmTM&5`cZIG7E3eR&2{aGl9UhXTNC9yFRI48;5>PNlc|DTg^46G8cNl
z^Q?Wk1Sqg6nr<!laYNI!7rBnY%VT-Nr+caexwOnnF4FVG1|X0sS=;SvVBnmf*0X8!
zbiQ!(SD;sd{#(64k@<|@!7Q45gwZ6O671`LPsst9&O)I3I*Qj|;O?<~PdvLydILcg
z&RLK4#N{Ym^FbYLQSX4P2$*F+h8JfTDC$*!%ExL8{tuwpP`c7o!MrXR*0}=tfieK?
z6kt;X92KvO4`UT@zz7^M7BF3)a^EUSR8OUex-sngo$x1`;!5~ufudL3x8{3frvzm$
zqE>6cX!ffw-JidhN&6IJ3J}cI;Y(V5vSk0xx%gk~coryl6Bq#c^Ow<h+6ZIkgPZ61
z-{%$z6-{l|E6vy3E$1$yn&*kKrj-6>91~1@V}y8bM6wRmlS?iF)J-(mtTkmn@?@``
z)vs>8U&6CRv#N<|j}P$PVq(f9^ydKit3qU~1PewUB1RjoTf7A6<BF%8sEI-co*tqv
z%Rh`(uBa4UFW+B3TpfS+y}flQts5F1t}^b8n5;V>-TtV?&MGi5s42(OsT+fztY^dj
zcT`J`=S@C!Q{Agq#^^!n2%3dKzHt(iSjd)D>L&@O=RY=2a#EU(`I??daKb;(u_l(@
zBBW<Cu-4VtQXSszucs?|T<R<}rc>wA@&(p%u!Msqd5=>mkF_BXVmZDmwz&XTO~Sx{
z_83@PZsruRb=)%PO%(pvwA|FC-K25Yr0Je(+9i*H;m$ZJ*o=g&2GH!&N2UOkMF65~
z<!`d_w}Rw--0q_eWz8A09{U8rgO07wV4V~3pA`RkD#-uo=sdsa36y>v&EDvoKO&q#
zI2Iw;0@eBXW(E@9i|9Ov$?1OeQ{dbxz(;SZs*Z?j7F|T+@)@&?Hg#uH!pDgE>Lf3d
zBi=K=(Tf)}8xJ3Y6~1~vc^u&^Qp`&4$|yCXLJewwGOkGo(_bNmi#RM>lR;VbleR1>
z+Kk2}mCnIOp&8u+GhPa%s~d#|rn(UApV+=K`jy|gjE*J3=qaa>jMs{bf}6iw-@b$Z
z{|ZVv<n`Ym3FWR9k9%ON+e!==w-*3cYG7^}eE1kGCS2+_7#Sgb+hYTXIDV<qQ<Vkl
zIQKA7Jw&A-7nv+pR1R-k1OLs(@Pt#p{uQnq4K~*miJ9Ajr`dP+*l{vph^3~~kUr(x
z!oPl%$zU9+9PH2NTr9rRwY03oBhxLou{|7pyh!c!@uNo}Xy#*AlQ2(+fsZ6*l^AR*
zK6i?tD6z0T^?GMeIoBnk%2#~Qn2x$z{&DG5EdZq93=i9qZa*wU_ojdDi`Ik7+cH0T
z+B`@4zP9$EPDkT-b!Vm3Px06dKDr>W=-wI{B#~~^wB^{T8=jWt-3etp_w&u!C%O#B
zMz<_tEHzls^&?9QE^6v^Wna@|-!^Gwsu>enU#qP^;BB>XcFWGdVQlB+uB+MmizMry
z#zzbBozC4j>h|we_Mta~a*L2CX*wvAErzYB3Kx(t<y+}ZFSeC9vrmJ&rQ>q*GVvJ9
zl7%Miy}OLC2+<rBGWN|$<Y&`igP&d!>TSe&OBLBRszK?pCeOw;3o+e3`Dj>*U_(5o
za36qFyuV<m&BJmfL=!_TMWA{=&_mSzLBTWY{QcjWrxIXfqQ<sCS>L9V_@V#XvH#U)
ztGui+f%8Y-!!&Q#O#bMvZmP#L8FS%yuw=}Si^_3X|74n~dK);xP=8`ii<ObM9aslV
zu}@4lCRkNZ{9}F>U|`N|0A?mK^VQWM_Q^)BR|<Z$vXw$?yf!`0RTD6s+hriY%hWK?
zM&qpVza}v$3NC+-oNwfPBKU$c|2!t;!|{Bta0kVdUe8!cCZ3Cv20(w(Dgns);n@0K
ze_|88mLhLay{55cfKb}p`l^%PyqR?R_620v3i6NHen#+6{!oc?NOlIIz|OcUn4JX8
zq`hy#T2`%r#SUG#pFxa$>S{AHi2MstU+K^=pyT2@|8mdY2M`XzZ8e%jWpeHr(5KhL
z!wq9$vNAg;`TIh%!&41rTn?K=|JC6iyHu##QfT{kXYEOa&3ZqKZ2nvO<R^HXO%>$}
zAET^BFq?1#fneg+iHw!q5*kNC?5Z_PhqX}_+Tb1do}JIYe;h89^<v!vEme4__e1D1
zpy0^X`77CfBk!ALvg`8jWP0*UU36bX?W2%e>-HY4eMxbp*lec|di<d3tG=Ur4*I=-
z1*sl?H3GTDJ4ADGVyoSpVlW3YHhO^#EmJwA<<ONh2PF>HWRwxZ92KbX^ld!(_%8W0
zm~U?Cw{fIi#|)gcxeiQJW=;hvH6jBTiXsbNt9|M)7VsQ|^6E%;ybLF|8B1L5cHRoj
zdcH0Q6gc;T^=sNatM-tciwja&_^UY|k9?MZ5VELg1v?*b&!a2qLsczc6R9ft*U`yY
zy$tL`34tSs5u82PjP2#0`}h)iNp%#k;Ak`7JLASx?}01%ncB3S>%{AS!N<Ovat~lf
zEP7uWhW55B-yc#{fWtBSHnHT5H55_e+4sxmm+E?5P@w$g)o}<8{alO3$5-rLw}ZEn
z!jF?b$$^Qw=kq_$kH>FT+uu-_%$9$NynaWvz4(5BS0>96i9nyHmL_uAUR0Y<7Mg_W
zAqKLP^c^F}4KM?9N~DVz5G6!~zCX5X4ZU{u416n2EX#4VP-{i?TU$t_YSd)eSWJ9J
zE}}HAG(Y*xEF{+()7FKz)oFzUYevt>x}H<Pa}qjoSJM}>vGDGPZ<2zO9r;%B{VZ2R
zX+bJ6{t%{vI(Da3F$%c9$a6%svw(=f9fT5_nV}`be=D0Va3kfeqkD0e7kab6TVd9r
z4Is1@;)aTRg8;zh($R1mzHY6zAw%pFlxx5w@KiYR)x3akD|te50%Cm9;ASte9OS&Q
z&fKngz&(30KaHZ85DpXL=##y(?B_3W*1DX81ybl<c&5K26iGoFWvi{TfO6u`^d{HY
z_v-@Rg^|3NP|ZTlR+p;}HA1V?uNmE+@`|=>3^-O2(LJa^<4Df6;^j%@0M?X+cJLpN
z{XLcWY4LuY;`wHGw!&o?%I09QsRpn`yR!s)l85{WB;s+)#caxN)KS_iIWDSUpEqdm
zraVNs&l1buIEhF0`57vYv7^tZi!N$Y7I!PVB`<8CzsaY$`iwC2-hIy}8^`rvKTtvz
z0znjv!|iG@F+9_H=ziie98Iy+sZtJlz@zBs!ZCktEE`!E+G^9)d0R%i<muk#2PNv3
zRCl0JLV$2c%gvm_HggSlS&6ojYk3HBi0ZFSg)tvKXzTF@>lQ~T3ayisep(!_Z_1!=
zPcTwh2byaamd5|?HS-0>%tXsMjT2H3#dKp)7>a)vp--DZZC*lLSGv)tM4#%gxqYf*
zZG!zD{W!OQy0w<+*orjLle!p>79ru)r@C02$FNRjl=X)W3=2o&)7&-<0=;|}l5frl
zQiuVZ@_fYy8}>GRa)bQ8SkO_C0<y4x^&jXkm<Wnwz*ZOKwgtPXOQ`@a+x%pBgLA&*
zudT+_@|SntJ*eqlYrF7mzhb(l2hj8Do!yYwG^}0hD|+myL=GmBZdYx|`Os7G5$JJ<
zsAa!u(-tkv#9{*Ax=aJEvj)%lk@nrhMpuy?8Ce!tF&2<cGS!;`qw<V$d%LFg^blnK
z%nZxGMzdIwo&ZTgiy(4TQF_Qwo%_+^z3uIZ;}gYar*zMmAEjsg6>cc=?Cx)C9+S4i
zzUx|0E81B7Tn+IYncb4QsKJJ^R$*J)A|C7p@`z<f`$%w(U<AC_Wc!o<{)n}>K-F8{
z;yvp597K5ZAap=P5iTkH+4@#U=<xy5;m4W2hQdy%sf+3uvLGEay+m;vIHN~^Gur$-
ztd8fnHGJFilJ8xl!^-flY^aC`&ATC9G5^|9dQd?RqNQcgEsX3iDz@7s;s3P&b-5c;
zl?LF+rQgX8E3zFAl-(M>pEt(C1;gT&e#P78CU7!cXsy@G82Mc0W)c<tjCZgIB#&m-
zA|_y<n){2oJ0mD)iPH0PETe!y)=1I8pt#4WbPQ6-RZcug7qFG`)A%iG<opck+%!rO
zf6iZ9ZQ{>w9gcKl_-j+`Ou)w9bnQZhHUF2lLdtKjF<#aPd~UV*wU{8-GH;#k`3MZ@
zUD^}HtP9R0XJzBL0?DE_=X^H53?BnDJeI?Q1VxOg>XpbPOi=3&>~HE-ZEjan!O|B_
zva^a+!KAHgTVGWjI&yAfm~`5IZD|R3C%Q-B(8KHvwG?=(^uOu9-6MAl_qI96@^9m@
z{m1+!eQ<?^t@2BT0klB0qA>xiV{SPE-46S;@D?SlufRUU_1l8x+lufc3OpY8eqAts
z+SNz3^ZDg*QjIY=o-FJkUXOY{asN7t*M3uDqp|}@P|e>*kEGgqTN{4gV-kFC?MUdX
zXuJjaOcttnm@ETAJ2bb5U$1?twgyg2a!eb~b^erDiipJf-}+ut+aAnwqqCbr?q7r5
zKXF6C2CUe|MtK6>ugb-~b284T%Uvy3nTEIhX`=Wm@KtEGK|N8F<sf)4&`kr|zc9Ht
z+))LbNx(6{Gpxjk<wo$8dY@U8PaUCnkCDSt-Q~%{tL?}*gKog>6>agz7w*yR2>6OE
z3FBmS%J~UYED`1tO*u3=qwaXKHr*U15<tXhsA*>uRh4c9TQVrLLgj<+RxLA7<fBtP
zYOjBZJzblZb7x^;VRsgy_W_w^WbU8mR%3*t02Cp#KUtBztXj%!|F70FyABknYk~#a
zC0Gq&<Io)yKAz3-VmYqt5$emv_k!}Oa81<~zfxb#EqfD`diavGJ|s-d9wX)u@myCn
z9lvRW9xHOOxz|>fhB*dW%KSRYL2!Z!xzUSJ=bhmGAkixKo57VDkjb_Q!}k515J;*j
zG(5IQ$Xr?gg39z?Ik(NIx4A;Xf6dL+YhL4}NvDmZ4%eF6Yr^mXo6}B8liw~&8-Ad|
zN9%ELNE|10b?SJfFSaU|pAt$&Mgb}r^>%f`_L{BHIxyz}(dL<2s;06@aelcTo2f1!
z;#50tkT69D$b{b1msVzLGwA89ZSBp@2$qm700q|am>^Y8WS6yT`_bd&1)hb4s{k7=
z_y>qK0Y==7KF5)~!PXLSqsMN$2uq)T!MIwF7hE&L#l|CxN=Ka`5>B+4+@Cz^v`jBF
z`Z=GhEI7&48+`iBb0dOJc~E_HKq2BX*nHT_mn8l<Ubyint#5gBv)HD#3I_#7O*y+&
zN{&MRMgX(ioF!twia?!P1XVO2qh#Gzw0s=-trLwq!Nxb>6-*r*z@u1AOG;YX8P9oT
zq9uv@i2&TfKyFh#wX0kBy=-0A)43j=>ns2K;14@+4iS?(BV(9hl;vKF3G}0hBK^jg
z!vOp~<FFE!ibgEmhNf`A6af79H0D)u#Ic^8U$+DRYp~3hOO`)M+((F8H*a_D9J}(}
zUWf+z81fczDN=)yO75MqYlMv)zG^A(JgRa}A_ohKT7g^K|0o_X6U)Y!3WO_FIZ~VC
zA8`x!Y0Ah-HfC|)gNnVcBo3--$R<$nEG?yF{R+ASHnIKR$xjPfd4@X|_Jw6_G3OJ*
z0xRYwkmswFf7`g|IyixIbj~diVYFA8q;Vz-9A8TIa7SmrB4fz3ju>7l!z(W#3H>i5
z>9<->3nSQ1etf<Dx`cC3uq%V^yF{qo6PQeY8V4vl_-Rg<5OM<~rMu$;z5Xt%w%}S@
zB9Us1CQ;zxE^sZE%7QVtx+nOzTk2xn)O(zaBGbyE%7&x%&5W+T<xFBMWeya3xN;lz
zF570uSEw8FkAnS-`dXiRytZl<g=KdXQibwS&xkK1-M{M+etwG<F9}-IdEHL*T31ol
zs?1wu9TxK#8^{h+HWbsxQ!j`_YVRywYsC%V<oMz7ZY)bYLGbBhEr^8u#ZSvaw`YSt
zGDt+`t&{PUVvjXcjfkp|6UgR3mKG**%p_4c$6H*a(jk6*-ifU!%~JVcf>ckg(U#ZT
zGm3M~^v`CWmk@n}TfQC7DB9E7Hp6TJ;WkLda<?GaM%NTKgOL@T{+_7!jX(fUHX6v$
z7&IMG1&E+1L8@DIfF2~gIW0KOzW)(Q^jMpFBxhztekje!#uKE04b&(X+~L@(RAW8q
zzl2iX`MPG#Er!ppgu86b5dS*J*wE5Rv9^8F)sjsHG=AhdI#gOkSa4xgu}#JEABRd#
zy&M@0L5#ayLibHoC9QxDNleFo$6~Au#q_B@TE+U$zZisQnm)Yv#+2+pH;RHMFF|gu
zt}Klj9_#yuMFw@aaE#FLDA&XzXMVn3y^D8}w>&xQ^`I;SNbM~MAly%u0`sFc0}tM#
z<h!Tg0pnV;z!H<=AMBUmt4G}uu3g-i&Bvv$>@16g&@j6THDFfPs(e=?s7FE)$nDm{
z=7}Wny#kvqe^XwvCBI@ivyk6)9=iu4#nVa9>mPuLNspA!Su{|m>$G0#6zD(CMLy20
ze|X>PQhF;}X-;hfWj0B)90-RYg)610v41Vz%2JTd=UveycI*w|l)7j9WT{-dSMNSp
z6v@Tyc>=q(HeuF%F1tR{&}b!MWRb}lNAl$!*dj)g%kf=*Bn>fD<TLwDpS>lw)$tC`
zpHyG5s2?_;-)r5uoxaI={+K!!BR@m_REX-+RIy{89P^mk7juvjyfXzh(Ff3kGPJ3K
zFT?~EUm1Dace}jfltm$gVv1woCFskaY3I3>jicg_8Q+OjaxsmG6l3^xC|z-hft1AT
zI&}}7Zc%_2ls@#o9NEq7M)c*ba*|wWk)tOYD@-oDO3k?x7e6Y@vD5l-jrgHQ^GHX5
zh(PUe+0#Z-ffu@vL-I!6e_Ak$C*ISH5HH!LjEOHV95|e+x!C3xkPCxgVmd%wZ%spa
zhKpI6k08eBCfU3M#K%^aqV*m4+wn8==Uvj0;&V64ev=2gkWxeS!}#Pq-5)Lt2vcx%
zr$WPs+dBU%NKRi4w8)3P$akNxER)a1BRykZEFt~?sv>X+{<8QBk@PZ9K=^9q+GQRX
zb(6rHwI|2P!uGHG`Fw5@K%NCDr6-26;zalAf-{L^s1;9_#8Zx|JrDa8Ra#D39SSK<
z>H@HOk%%lU6NB-P+bO4qFdb39CRrjJSOs<cO<F_;=`3Ph4^?1?6e}-UK@ban>8*Oy
zc3AmT#mCYrbOH^dUkPK$I(8xPDEJS?xI=XYP;(ozknWl;HS^GoBQYZIR_sYHA8prC
zcYm9lrlw;?x2H_%pl8Hz^Z_LN%2|zWE<+r%7A1}L<vz#6WdafB2oC3f3jNQ`!m&9m
z9$|CQx-ia?Kg|2pSJnwse_VtCDdQ7CnH?E_Z(Wj>CJv)0wcpudisN#qVxu%#a=Qhf
zM<vc(9fz;|b?us>!x~fW-QYoE{6K-4ZyBzd!ytDMgO6^fu1YjX#v3=rYZi^yf5k}R
zHR9(!=*?doi^MV(d-*yTValy%TcK;5TOg5$TImcYOo52e7+%k)|IP<e8)<K&yYA`h
zzuILzfCr@LivtJTNz-g|1ST-s_6Kk--&k$;L#Ao~7TW)~{=G5`%p`i={dw_u09NsW
zwK|<sdzbq~M;<tOae!&2`M)TZHYP?tHWNo~Zt^SgLb1v&_-Ex0MS-0tRpk3JGTchY
zjD&fhoDPmJm~+BdRXJqT?~_9>xQE5K%u=DBFC1p1;xDH-26)SM{!u6oD~-#QAX$G*
zc|AMQwd7LVq%O>i5{rP=xm1}P8+HyWlJuSdWGYrZ9U8v&Su(Ql7PWN`hkpC=)5ZC8
z>)G1qHkNk<@btjPAmjGB60PXUCl&`BB-z#N0L|CG{~FzNYg;PMqfwaq(~KOso74?Y
zi>@1;5FE7lOxdTF5CQvV6;OD7zTz(R1{$n=%Mf~m*&zhl08z~-xcWNk_#OeYfcy)L
zv_9ilXeeY7RU8<^iN}picC0Zx+<oRyLF4~RQpwLY*26J`v2N|yzQ>)WIZ7Cf?E)(+
z^Zt@{a<SB-0jCIH*PloVH}kT{8D~xY6TGr<Yq)z8tb*8GQ2ujeTPImTZhP@;^}H+a
zy7TRQ2KI!#1zZ<}FC{QW?bGc7>Pzzw2XlNUh>aA}d6hNl8~WzAzq@}E8O^87Gvs>P
z7nXPA(&Z!aO42HSpFABV&B%d|kA(PGZIX+)pna#+zgI~)9WSP#)tI)^vG^?bdcpTS
z03(s67O$bSXHBD<vW8k&##$(3@y?D+?c@b+s&ZqMFb!&me2gs(8w}^d_kA;ieWX7;
z_mRkm6$rO&lWtzPPah$rv>-IU)W)sb-@3p&gP`8hLy!IYirbvl2)(9+hxFD?ho?Xj
zsphW_3sV(i_yV($tvzp>MUtch-yLsTxC2_!YHX_(DWJ5HX{Ve-uTCwgaa-G=N~S9a
z`&GkpPnt_HhU$TjWL&Mv&4giT2_Qr<vlFUXO)c~z#7d3hs#h0Bl&^haa|1xktN~T%
z&F^J6M1j`%x>*pyCjMl^nJ++G^_|*C<0CCAMpiLbbczqq6qSNN+~MbZQNwj0K3`Xc
zqBrG&e*cx^pVP5l>v;pKyD<Ljd}9AMZ0pJGb!G15{MYy;Y|a*29~VghH)s^{v~|I%
ziHRtDV7g>MjFh2v>w;cPP`3>^pf+LTZP8t%^j0%b1PdrPz_7R+8QfhjDgA59-YT(*
zhbgJh1-*uel9-{rgM*$I#=49YO&^^vkz6d%wI&*Y@FDB_hnkyc2jkOGi+9yIWZ15-
zkhz=$%bozBY#a6av&D_&#)97a12x&OF7verGO<r@M`7{p?{v#MOHoUf8jK7at^zU_
zCz@~SnS4>$%n18hGQ!hc82?#Kg=}u|MbyYJi8TL<OnUA*=CtoEljvT(h^)Q1y$_-I
zm)5R$-ub9_T9YH3XVYVz_9Rl7@I2!|CZ#<)8G-9pN~;>0)Xjcc{rj)z_aDwG3q#P8
z@Pg)uY4ZmP14G_v`B}KPpJxdPk|%(p?VpSf+pyO5TgPVQj;v%2O*F(+8M#%N*!mx`
zz}y&G<LOQi-{sYF<1Ls}0l|!aLz;c9rH*rSdHi3oO;>tIa8a_<udUNn?EzP&9YG@k
zwXBkSTa|Aajgpl`nbrTR2vQ5=0Tv0cwsb!*rl3(!`JBq%{_r3;qrJ7YH3w|wQpN?(
zi#w&#`{(W98-;<Q)G{4dS&QEI&ivE3^<jTl%J==P1WF~f5{{v<>|-c_D_$=x=A>W5
z2O(Bf1bZkHBIGPp3o^YE8yn>LO-DnmXjBGbL|2J6F{h_x<qevJm=RNji=_W5H({Ty
z%_-LH5I5Ts0-9^eSrbz=xDEn1MU;2|Lol=uq*l_g8Z1)oIo11INIQWltJ2&kf-(bx
z>FU&0$<O|o+$Jv0NI*h0wgZwD*MCWCqzEAb{3vXHU;ZrOaPfA^B06*2G|v&^I}4HL
zX-aE5>-wl&%U-qH*}cnbN6L?&JKmXmTQl%9DgUKq%l1!O$py#lSB3%`TKczNKhp&X
zTz0G&S}s>tkW@jbH_|u#*B?_uq{lAs)Y}@y00JUPO<_6NC%aKO=6jT-J5^Hg?GK33
zIc+rk?fGGa$x44~w$>HfT?-D@d3(`wb5lY!5J8qUM5ciOjURF*&T5kqm($*TWd*nk
z&txDILr^Cm+-xi=ycu7Iqjnc=x!qqT-+Nuo1oPnjV1{G|d)!}D*{-zEyD>zZ=LQRY
z+L5URir-?mqyU4)+3ntz8@Zbg%NKQFQSL^IUJ@ib=ERDcM`W8kaF~plWbw$a*L7<I
zXNGtmi?Nb%os2OXqv3&SrtlW#oLhJgOOkARpxoAEH!R;BKN%hQ_k~=;l$Opo^0WJu
zR1yr+JRO*L5LVjgQT>jMZ;7E?FPTL>^Q~>7VPlHYoRu}HTw_=+6uakB2s@2-w81ys
zY=kLy85%$;*Doi>TJKNZC&vSNxtKpM(4KDM`*=NVCBjZ_=i0`P?r(!MxB+$`oiN?y
z%x{V73El4L4C58*aLSg~PyO@;{1t}u`@CQ2d_Xr)wsObe-0N6*?-}cyYk?HnjDscC
z_OJ34ijV5j?^upyWXnJx^3C&)XaXpi(WTrN7Z<Van>b%pvG2~{?&~G2{5?&1IIH?K
z@k1i&I*SW8B<yy2UVmM+A#Pv%(}7&RLfg3K2m`yo!r&_aJGe5N?ovPjYEIaM60!Jj
z%+L&8P59on&A@+=l#o$i;{BaFjPr$#-2j>(*Xp2u{E@&*y5N&dbO$`_bYIC;PW$VO
z$$<nu53--Dt)marLHo0nD>%T-?BgluNw&91EWW*&FbK=ikHs~uJLFLJ)!yLLIhRK5
z)J~@Y#w@nDRVv^bYcL|e9!7NobCc7tWMxqk+UioE5}6QOj@c6wu&7IPAZ0W=@Q0U4
zBDqkT+_5`>TY_+;KHV#pnd_6J5Q0;YYbi@4sx;~9bWJ!Ap;_{+=)hkad1lCkiUJJ>
zk9+rCLoObZ;OlR#sItIuq<aJ$M=r%oVR=07m_#WN0lsCoM~MDj0PWGALjNH~;R`(v
zBfI7UvLiUWphh8W^0DK1y!Ebo>OHwSA6&o0PE~sB(0dQ{Uc`V?B7DSU{%j%!K)OK=
zjh6yVe#?O78##X`AL)q{jBKj7ad0PVbv^Mqfdycml-9LU&CtB!!aBtaKVE}hes1&S
z8!VqI88AeNLj91{VN^wUq$dk}fAgVf<>KT~L|G1%NDF@euqF2Ch+(VVe`$FVm<d)r
z{A1`9&!oIAr2YWPITH^LGU(_pnL+!O^RHhKb>q4pCdY*yr;Nfj+ZKeu#n6X}B#pCK
z9N{*6G|ioDklZ?HfCvpknWPNnI|)d30DwF=C}sv2*y~P&AuOAvg(NG^@*%6ht;+S+
ztH*uf`f{yBy2%nQ+tU~ro@r&X+FlptSv=2;H6AhYDE$*G!>pS4QE2GzX!GU8B?T9^
zt9J%zbC6fv8Q;Uo{a;b`GzrwX<&eJGt&K|B3smP?c$Q>QJ!(lczh5H#Op|nH7HQjm
zU#<%cV1?56a$hx$yS-|Q@dXgpvSIJIjdzC*-zH~WC7&}2e-88Ih6mAGU}8@Gy+l`!
zfzd!2p+x|YOlvo-D$mSDn43J&77|{%kX$jxhvi_`)Wcmr<HFEV91cM0P9)@RA2@TC
z*1e8ga46=UBs<0hJLS0}%n5Zea;JFU1GZ(jz9L66Xqg_A(<O^q0af^dX8m_KI<X8r
zZKKjpI)!i6089J-a#rI*gDsSD@INcVzfV`c9PbZ*3D0)|{g?!e(m_UWpKRwWexUoU
zL<d@oO5`3fJXRKKEK+<BVzY#CSSFq<oFW#J3^<D9I_8I^4Ev-Ld`2@4=b;kEVmaq@
z@pyy>>@-o(>*g-6{vaLM@}_u?Fb(QlZzVtNvFDRJ7JsSgXaeqhrJ>lfeYhw1>tfVN
z_<qnX`fKbP*l(9%_|}2rgL^P>)@Z%+tF{*^?2Dodnu2AC$Pd0Nl)OxQX@CU-sFqo4
zVBC2y;-!D=gaZ}b!zMICxt*UF7!&)*mmkclHM_V|2LF2WsXB*><_r8uwlh*1L#NeA
ze%*qL+rx!-4?x{f@$t#@MTovO!|QuCLr+BuyDouscG($@&sJikz`v5Oi^1ZN$Zjtl
z(!zR)qR(VtVGxqi=-50{Fr#wy!aQvyKp3tt7X`tZG^Vf8%H6zs3pq&l^SnV<h-G7}
zBro@R+u<tqF~5Gs;bkEb;8ULfcvv@EcsPp}H8u)(ZfbaL$`Au}7bhK*d~^x&h9yXY
z3!>$wy1kew%6FC+R)Ribp_++%BbzBvXW`4F^l^Z<B!rF?!^IF7h&jOp;F-&tB{MS1
zA{p6ElVzIOz-%5tk9vD`Kn4#`vE(hjx1yrU8oFRSVWTMg^ZsYw28nP%Z>W_1`tP8O
z_5IUf01{2>%fa~(=HCw2SdRd}Dzf!F;=s27NQmK+rI)H?B&Pj^{qLYN&M4CE@8AFo
z%=njicnkR`roW-KMzDfA94dDmlZW3+xf0Re32=vdH?fcRaP`Sebl6$xLI`*AKzaw9
z>p={(gT~>S3iHb-jJ2D6t^;eAVfX4d)f1E;lb#85JYIceG=!XNlc>1Z83^H^culOR
zP2WIq0f5>gw3xVMBIk)n`_NOc@m(qj`>W)@oK+kCr)byRN_SY9YSxQ<F@Am6vVEeH
ziXau(-WERR7Vdlz6~4HfWTamRjKr~AfiBx}@KHZI+pM<Tj3{Kdm3qK0C-=tZV<Ho0
zb!m1}G8^W+xwL{*zh~PuRkK8(qRpA8<R9s)?fvqT#%2i%`v>M58L5D^sZ4`%qIa_G
zMY)+e=j@v1zi?bGYQipSO>_yZ%%?i13t9s78r&w2cU_vFBR>{oj8e|>!1b$k=0I6`
zWEVOrD{#(Azbbr)mg2k81%K$ZXG42|7!%%qx!y4qN_tOvxanjZsZELHsMv|@(jRDv
zNi?z~9UnH|coCT(r1^>~GqpKFPZ5xZBx~^6ejz77cN`A!?9tAT_GU-)(Ezup%L2>-
zAs$GkoH5+cQ*&aJCmJ&45vgv$L1)f7P6$QG?Rw>>Yi#|kd1}te#;j!RLRL<gxhM{n
zI~G>DS*$I;j`fI;>|<oA3y=+s?-2$rNq{;I2)Jp0%`d6jLesy2YrX&7I{31_9gUoi
z5Z!`@SaG$Fbi2PUK6a@!{;V+vymgNVjS(KY9`-^u90tj~T_O?=XW!H*j6}_i3PT+R
zD0P){3#m}`Oa|VIiPT(<TJcDrE#Sb12Wsw#4uG0H667_ZzG6HJvrnc9FwpsEpu=sT
zO9-Pjr*68*&OVZZO~9-Xew5rfC2zD;INErg1Pq70os4i7e4QRk2tD;HUXsXU;nn9F
zLSNR5ibL9zBVYj{Qf3Wk7|wU3spck8#uprN6C7(Zl&$1e5jPgSyurOfbw&H7de6t2
z+7#AoZBRqIM@2g)3_wd$&*P3Ch*%Ja2$hjBWRbCK7h9E?m~bdf;9f+Y`-r6seM5S@
zz0}Ar`KnA4%f;8;a3>EDDDI1zf*>3#%nV_mSeQCsc05f>#fs5b&+xj(nkTg;u-0Gv
zr!k$C(5$=|EC2U$JSG`EZB+Q}BHzuI({jZ0U#KU#+rb>L0<QiS;6~nl?Rk8iUUzdF
zl5v9~1o%kiOm%YO{kpMFXDSV#*(X}j{(RHZ5$u-0Juxv>a=bCR<%!=ee*0@XAvkd*
z!xJonvOZLq9)}JYZr~}6l=H^}w$bi(k!KYE^h;F<4wv$|v>Z@MS=$CJHROHXVAhTv
zwWcvC#;+{Hv%54InvUUI)TZk3t-^N+QDkNW&1i<zm4Xmo@W<JIi80NmId{v*(6Jtv
z#V`l67Ewaj8B9`h8pqQBy#15@-rdx*Z&oq9(Hvx&xG$$a2RsbUe=iS%U4?kC->|)q
z#{$a7rH#1Irjc<aiPXLUsyMBir`R~|o^-L>x^M8>M+4?yIMUZ!ye@w({YtjC2CpA&
z`rEP=kS<!}*2m)lq6p%aTr%dpECVnAqj8;0Q2J*QF-2T+13#`IW`<FdgXiIaUy1gU
zBuOg8-rc~`(t=A0tZIz|+F<c&NXHWmfEx%kR&HLNzWRBCXBLskwiyQrnwWRP&d~qt
zdgvGN653Pp8DGqQUG5iGMqU<=)p*u>J$wGWoB4J2GUTgXf~YE*B{fsvn92;mOg-az
zJp&Dt7kHN<6UbuK6LyR78c|k*t@^IQ9qISB#^Wh#d}hyRy$@fkzcNv`7PphsBVJc%
zszZhgORZXY;kqSCEu_ftH5FwM#=<>-mS#(?{ykpWocW3gh9a~4Zid<w<ppgP%p@QW
zUzcCyJu}9o2p1E&Tuv$ufZ!%bOuiwXJ~nG^AyE|5Ms-axD<>L50c4MCBB*l8A`rsr
z32W;L764-Zl3)pt9egjDXeE6@^XQ8Vzd+xvd9r$;MLnGLdn5OYlG0So)kmF`wn2g?
z@A-=zmx|NbG=kIgE%tX5hvN}&M@v`udg*S0(A&vvuxCqsB)ibF#{qPd>mzwhJnm8>
z2Q4ehaubjYS>B?-CNg6UfTi`gQ}8k{of*imb<RRuZMvmMW+39Bw1Zvoh*(q&EIt;*
z5`lU)p@c@;11uA*b&t%_NBsIGsXIn3b0!8m4<Qc;kpTg-d~@k{yxqF><AvamwaN30
z`=^aB<1V5v5binIacHRI3ts!56>yK>!-W7coZfZQ?CG9HYcl*B@-Laa_T>_d^#JY)
zaE~7sv7}rXXS8#1{q@vrR_j?SR&Cdh*8}!5{w+M%Sh)@{@^&BAZJ&HE^Ae{9PRiG1
zW1X5m@qEoP=O)8Lf)M@W&xiAXYDoDS4;s>AgzORNTsSxVHn`d9ROuQFgQ#fBF)QV6
z9Z=dZtMbrv%R=7Wx4R5dZ7cvJUkq>e0K;}yCP0@=BI9Oz!Z$VJOn-bfI*cZt^F)Iy
zbn$Io20(<!cVlwA+!z_9wXxch=(x+tQh>JeFCf?5SRmTnn*;#dL`ML?y}cqk-cBXl
z*vK<fr0<Z^uzXnl7~Fc>`mSj;iKG|ff$h6(V}MAcMWo{JCKtv3YXKkw-nTuXHqgHv
zUvGtav}8K1P-D1NL%+*EW8beIZ?B-+tc!ROj^eW*EH*&TT$bt%Kk=Lg#~Fj&OSFja
zNGb4X1L#h07Z-&NHM7o`gq^>sMiI|VoDgm#=PC1MFxKW}BuFz0jA0agosvzoL_kI<
zF;-~7LLj#iXF#|VdG~`u?fMkkhrQ%cYr!{`60s{imrN#Bsk8tH3EBh80Qs4cf9gMh
z<3D$&2X%Kq`9DWS)5B2IlfQsZOaXx~--<i8t;?WJ!|hhe>7$g2&!m01HITA}r)$MS
z=t~yrs*bDp2l{2j4O!A*EC@Qc$A=l8P?f~8r+Q`t(X!qRnbEtP{>yN*vwQQkS(fM<
zKkucr(8v3A_m)Ica1lPo@|=`Ru>1(O`jfAD*)w?g)V}mgCciwZ;b?OQ>WSd!lja3r
z{i+9?KnE#_$o=iP^pm7IAb8+1+oSS{Wr>ta_Q|cPx~USIXuZ`)yg8q31BEKGyf6Rv
zbMn`iQfS&?Ff?E6E@|_JuxT;*s$6@YMXYLJc%bcJVKMWcR9u#D_xd{D-hUPSK=$N$
zQKH!@=CjJJ@4dcA|LngaEO{+19I}_ezddJGHNboB)txpux(z>iV_f?V2&9N>E?M{)
zRBFauNg)UkQqh*9?yzueg(<ZvkOJ$bJI$w=lbBRydWjr)Bd2N26Rk=EMAbP)@MMuo
z5e*H&leCRJ1o5;yKft0JXw@Asf_O7z-siP5=1URm;&+O;VQBF6Olq6V0rUpc5SH)~
zh#Fvib1SqihU4Zt;Z2{p>xPczR~tXwRVl~@l+AHflCK<bs2nl)e*MQ+4ld*2t$G0K
z#+MuhS#`Z0FKtfO`6Ht&#nrVx*k6ve9x)p!Mr|a=<w~H69?0e5T0o=4ln51U5YiFy
z5Jrz5K^6L*eCe#C4JS-xe;*G3FK_w7I~SpuJA1$g9Z(oVgWCpMpX8h-YX~6YCkQy0
zPrN1Hg(z+vDG*vcM}7o&gO>?002vy&?r7mHHi$OqJwV*5U;@gOKG|W8S!>Jm=65m`
zO#x5mf93a&S6q(RO=z}KYMoCtM(=ESvJ*XGs5BHMrK$LNWtAZB7?jC02=k$PPDorL
z^k)JygFEJ+MZ2GrK#rw&$+6E|e)8NyJcmkD<4#Wx!f7u6%6@`SC;Xb85a&H=M45&u
znHUH;F7pU&FMVS!)5N+hCbX8PRApF448$p>WHIbY4RqHz&FBD^&_d13`@+-$bz_F7
zcD2&Np3jfMlQ@Y4rP#prW`N-5)P%D?Vxv+-0aS~YMg4U}<U_{MvYdUg0C(E+kGWAf
zJ!@d~GJx`o_Oj5b$BdQbZ3!0!fXj4|tMEpGaj;q(=7ISNV6LU{clH%Fh2W2O-h!h*
znEmaj|GN0HCiCW|SNL;szT(*q(Dc|}(SolKve+3|<-3GH@i>pMdfKpkCiqXND5|EU
z>3k0?tqm7m*(2yj>7v@5k>DG0J+3=kH##xr0bae+>DdgNC9ukMpZU&n>+@Ua2w<zK
z>I=V|2p_}wdx0vvjphPjdCu62?_ZnhQ2FQJ`pocd;Zgd<9RcCqvq53$>cjczJ*YUA
zQ=-?M*G5Z%M_AHVxA}>S`FYgfSXy?B1@8)4%|UH#0EbrJNK*@Kk5Qv(=!pC6M7nLi
z`cjV^_Sa?c1d7OHy9tH2{+P`l`MtT3@ooo8#&S&cV`{Fqc3~fDZk}Okf`xGufU5OR
z6dHgSiE<I~2{%CncSLgQzIy7b%u&d(mAQHOW7Ca!<MNW^i^;&K3AH}gB66^M`viY)
z0t_%bRv2)6PZO}0b*(vlayaD<NPA~|>!<55;+HgQUmF+8ua4m$Kqk-;V=lb`y>tXX
zNOSYP{oD<7biWt=dmiUx|3U{=AMa*cN7g!n8ZT!6+r-~13d#8|ZsdOeQ*cSy{t~@Q
zCcXQGSqETbG4)qO!MY%xzwA3#-&&3-_x^@je@Jx=UUm9rX*jElAr@e`r(;@i=ld!$
z*ngJwQNN9*vn!HvHHg9;Go$fTYW_O!T6r{{U4YHyc%}LVlOp@^fI|3q)B`j(|8{->
zT{^V*{#zz(6}@6F6wIiEjSBDc;r>BPC&D?s{|-&4ZQ_i&%)}=|9cO%5f%Rf;CUl;l
z4O~L>ot+N(bqkO)90w*~(4X`M`D@&Z{2v`h7Oe6%2LL;V^B;Eq9cc`&A;_}ED94Y3
z3K;~6u_fxQeSBTtaTsNaO5`;e5%S8(;oit**CC=Is$JO}$Ut%!W#N&t!@me;(eL=b
zF6blLB=)CNg7wcC7vlpIHvJuA=Ma|;AL$nYC!_8o=RVhBVCuz4ix8g&E&!H^5r(R;
z!<Y|^vQ$`(sWpx)J*-Oofqo9Vlu$vV#n2xQ#m$1@_C(<)yRIsq)9UEFR!n&jE0AZj
zBENvcKsBYI%Wt&O;_+vl>^O7f*b&$sU=NJ3!5f6F0qblxdq`|efMMSUV4<kc&!0cd
zd0FgJ3w%Ex03W<>DPzgf`2GfaD_N+9<ePOKuIkrumMm;!Xf$dI^aAli{`cJ!&695(
z-1tKMe~q_~Sv-WAN1xX~k*9!0j|u24hZD@{)+dm0!$rmiM@HD#Fh~#0NsmoS3j~r@
z=$1UeI#;?11^BSo%H$2wLp<CCi1whi75{`Be^b1JNfhS@FVM=TrCH=C_jd+1QUR^A
zlw=%$X^_+Wm%SV=MSz`ue)8*Ow&lBf;?6sWf}l3pcYvpdB1%sl-34O$J7knBKpI1f
ziU-jaq5OomG%P2lM~pBHzcYu^l}_7_0kw>%pq16Eo~UdYPsA^VXw7sufiuZ}Fi<6a
z1q5B>dAwn03WhxY3E6`IH0?`T@Q~&V&agGB*>}FoBR7B9)%oJ8<N^ati%MV!LJ<I_
zs+;ku%MN}27aD0hp=c*UUDqCZHys3LNu%z8-QAyFc7JekbHlfFbimpE=p;!@OZ&7R
zbglZkJL+8Z9aFn5jY08|<DZP^SR^zj>PTl=6!;LJsgC*qQB7(R1H^SLr{KsVJ*j6b
z%TNW1V|MrOxbdw>uiT<!#a~<8L+H=KZr{nR`_(9ZJh_eZA0AyF9#5A;EE%b?5PjOj
zyoZg9kuZ_8-xB>gsQ0s1bR!Pht$f2f9@aY-9E?)K6G5)Xy^MZEXbo>A$(rx>YO<cw
zLbSAcOb;}DclG@`Wl(D?(p{|>R-B6+I|kXt$wF>+J5L=-v0o_}(tjE`o*gBBUe+Vp
zDh)>sbG%RijK)Q61ePYz(d2k>-M#Glt_gw&_kVQ9f0#WgJl`Wz(5ouzx8^lcC9I5J
zMd`O+=-AWbR>Mdv{!@tm=|}pt;6{b5;BiukRA|4fH_TT|u;y-BTQYI@&A(}BNF-co
zfQU5isvUBdSqOs+_#JXp()7=^1qTuk2v9u~B%5W=&4h&j3T_?spyawLDs@=da3b>-
z$saR~pK{w)wJ^Jy`0^|*WuRF38_MrRZFnuwaT{ZLTP%XgmQ;3jR)(R+rE<l?L0CU!
z#FkdVp&yxO>E9)5fWbf&3>V@UFUgWr0TAQmN|>OTv=*weZI|?Imr+Ib1wi@^EGf7F
z)`xz@Cm?uP71em!L!#KY3>d54cF_y90}4gO8)Wb?Jrv4q=EiJRPdqM%v2Q%QrZ-}o
z2DI);^0+vc-u=?!(s0DlwCXV8(b5^?XpD^Us{s?z!|+?>V`eeX+<4gBLZ~JQ9_it>
zQxBq?$8ft%PrKfTb}<o&lo@GI6H%m1u3+T{Gt;QaVcyj~-T5fmDBtr1EcQhR&0AVb
zTNBcQEeo{9dM8My&*uSEkU<vtc(4aj#kPyEz|&pN`N&d)v({FT@FmQg=g7{Oy-YLO
zAE0(a;C;SDD52!SQ6Y?=IH<%!6U+T2img54@d7<<l&VV`bu8p)L9)oswK&@Ll}XCX
ziN*f%>UbY@IJXtLYMDc=7?O$*`ClUyGW5%{8hMxyS_Z*G4gsa-LR=IX-muy~Wjz9L
zPl;I;-kXDT=>{`sT7qrGco^P-02h2TcCAsj6Kg&eb3Ko-WaAn|LeNqT&sNoENi+A%
z*>l0sb3KNUkXOf?V#t8&bHXXG`zE>Fpv_$9oQMrr(zAZd&c=3@e*GtCch}MrV4>q*
zwtJ7Rjl2G>k$Uxefs<fZwZXjwScK;5{Py*fSbupxs?Hu&9c?B#>Rxm%mfQ@Cb<{#Q
zZDGW`_OYQ}ME@t3&Xp987Dl51*V;o7YMoMbv{eK;7gAEUKw_Ds?t2YWD7TRgb2c8@
zRhhY9Q}p?&!@w-X;o0|l*Cm(AUd)6b!G&SQFk>BNB!r)Ssp8FHIew|SB=x_HN1(md
z&>rD4PPg>!NB)OPU^Q*CEM=ns6IIJyAUFAB*<q&h-$9q#`VSxJuxkTUNdg0mOIg*?
z!s-dlQSf-Eeeg1sq!9ufRpDf^X@~h&{}eH8EA$M_(7gTD>ko-SfOQaxffA*~%=W2N
zrekWtpvr9xmGWw0Z0Y58?W!H}yqV;T`An|~s`>)F&KG%vU5S<>pU}4yz2cjn*R1sJ
z!>^-C!6NkE_|b;>0fpVhpOD0ap_#dlI;OJ{9rb4ZIHs6^n^{}Yby%3C5skJk66WrE
z`AtB8q;G>;Q)~7^pmoH2OLp)tzxP*%>oFM*v=sMqy%FDY0C%g_p%+|ZeK<z&xZ9!{
zn?no50)&cD8jfWBwMW87B*oj%roS#v*L3H`=3MRJY=VCD$5T-KgWzPzO3w1icge-?
z&doj@AGnN*PIJsV>tnnnb2TXtS;EEMI->;!0NZNJ&iuJhdp^wc#8-`z?wVU{?`(w5
zCud<XqQWQP`9nC87>KKkvao>7w{%O^*<4zzHMYRZm5%xHAky?$AqR~YrlqZC8HUF{
z_L`pGwtv@q@?^4ai;#}Si+_VA&R)l0Svj(nD*n+2mm?a%+P2&DoC{cQ=Gz3}VG1Q1
z_Rn`noqbyr^g8kzcd~t3VI_jB5(?gnb{mck=8YrwNV2~Zfu^l20%EH8Xl6xKRFB(3
zb>-oAnz0h>_OF-`T6p+OT)}^W*`V=gOS`BystZpWv0H3Zj~@-yR74}LGO5SMvOby<
z=<S(1v7ql3=<ODblijKK23&;T@$JyF9LB-l=9CoMmL$*CQH_=}yq8sX!m~jeF2c$+
z)4|T9m1Dv0;M9W<<D?-Nh`kP$I7M@X8KWz_o_Aqe?my1(4xnG!4+$R5xfk@_;Hj72
zDIk~@>{F;!X}PXFO>@3+6_E2w^~`4s%OPTPO)|99v@1+Gj~o;y*Ces8)lhns(?Qi;
zH8K&gsVtg>aj8@-^xmYYROOd>^#Mn79ntM-Igm+pUXoC{JZk&Y;&bo@UaR}U*RZe~
z_Jn6F%=hIz{8F@t+;zcbfawVpZYFJt*00wwb>0DH$jCj~I0i+ZU`4WRh%svZjtq=D
zVHi;3VKK1DwaHM4^Lv;1_zwgM8UkKiRLrC~OUj97t)gaH#8$-0Z4$hhSS9PGr7#f)
z8JX$BcI(x~>QP9>?8cw7tCh{^xE$Tz%_EzRyl^1E?@O<I@C-ft=7$}@fZFSrX>WDx
zN*)&e4}GH(^Vk2~=Xr+_vy*h$cL;#^#da0L_6K}PNy*EqSeJ6l{&P=i^oS&Ft%9$b
znEX@2nj}Q^?`gC{vNE)}RGhy*?M%7NC6k#bKApsDF)oI*8T|=wZ8=$~1gX3Kz9c?a
z`u$^YLO<2>nQQe}|5llcvmfM>N&I9n|J!e=(iUuN8apY60YSY>G6SYG$Shmf2lApC
z>(S8$&nR;;bnc3>5=i!{sFh(I+OA36A2i7<ofV0J_pGGTS2;b^31`D4{n~x|<?uBT
z3*hnEU8mJ}pk7}T{#5nMX1!2QIO^ziueS8zQn4>ZCq$AM4V|kqvl7_uXI>rp*E)5J
z)M9;%Ki4lQlp0w~GQdzyZlYKn86eM)t+GtZ@v&sbda#SVzu=%W)^{z3$>#2a@jl9P
z&mHiQFt=Nm4egQP2l7YJfAI@zxx><(@wyv@IuNhi7{7nI7tSM=sV_?Wh<VwM@y9{u
zv34B@m&dfQaC;w{?x(xnmlGhp7ue0flV4Ea0vHCI>rw=6KT0sXmPLiVs^?`QMC&oL
z&W$2cFe6>i_7yRNA@LnP@IQ-RKmRMxXpvqGT|ei~YH&MZeb{^urbk{#{M;jHbu)0}
z)o%9>?r-h!%J@iv;==XQom|8Ez%_SjbE=EZUwMhWB3sH-zpDC;o^1b2Y}ryjV1D<l
z6BkUW+L2lkMU-8Q^azgy0~k6XtIER%^n2?JwtXT>4P1&^nm_I0YZ*RDV{-khkMP)!
zP&Q-;Gtrz|C#wMihndBHiJz@$=bHpU^IL`r=BKSSjHf4&jAYy)e{3ijvZx{Jptzix
z3?25TLfXSnYLo%(XUvK`C$8Z$=@ZyI6M?*1VGqnVu|pfqpd4a$y0tJ~j4(89?5x;T
z8o{$)L<fQOhxmKiFuYGm`Q(#{#dz14#fW!^lhXJ>?&?uz^rogK(j6nKw;@xE5;4m^
zTs4MR?vm%{hX8L{Z(Z?xR?&@=epeu?|2>bxJaJ8z%B8(j8Xrban~EZq1HaudM+(Ws
z7v~O@4sUOa3d!xz=k@%lSoh4x-tXDxzsrHFEjaThmcbH7f(j51D`!fh$^Fj0IT}j2
zx#}XB6KaTL-Qu7|7SL^u3hKh&D#lxwnfOdy|30Tca3;>MeIhe6gQWJGG*uyj?~vvP
zV{s5{D0M8A7L`_+5;<2yvp3pMly4T@VQ`@`GfBEQE4wnHL#wGV?;q*3l5!cNx=Q#=
z6aidrRtC|G6yjuEfzW3lIRZDG|Lu3~2y~usY5O&&0Qrx1NsJYIs8XPJQ`8}^;nphO
z5=Tq?_D;j$?AD&QX8jc(B2iTq#{r)1B2wy@*iSIUrACa2i)wV0&K(F_3RmtkUPg~b
z7*Z9FP%CyAsS=tAa~K?P9@1_8bHy=Zhy<OpzYY`5;1DSOjsSKMYBO`>>pPsBUT4FU
zd?5TknyxY|s<!J&cS<uf$WYQSbi)wR;?UjQ0@5iCHH4%zg3^tYfYK%19nvCFQs2Sn
zdw<L?xo|l5eRiz1_S)NZk8xxy;NtoTP?bH=MUim<-OTEY=BmULTc?0OC0T{ydt!88
z4_%tanYYv?cYZb*icR0nI-Q2nryysE_4CVSFs0JV@KNHxJ&XMLsT(`qqx>f3bFNV*
zgo5GLDQQTHN@61CP`+0Q>VsR}qwaa{J)5)Y#5pmUp;~E?Fu;~&lMiP*)sUJs@Hq?W
za*a(n*8(gpno5Yer5N3VsSGSr&OEBDxS)xJ>|u$dlt!&vl<op%B$^l!P`W|WY3Sbt
zZUt~X-T}rXIl};!#+3rc8B#Rh86G-#2LNzlsunCj;VZ8|ggvAs(9#Gqyx@n3PI>9P
zSRee`_R7UxD*?Gy+}62p=~bz^2}%ko;_TT|Q2ch1O)x7H)*x4JWHBv=Prap2gQaMh
zjk7KfJ167F+=ZG{DVX0o5=?iDvvlgTuH&B7$Mfo`M(dZs#}jVbO?$ko7}D#5mG3;y
zV_Y04$f8V4e(_^GWOLA6F%N!NGTWrt2;ObHSiIi)Q>7Eft}Cb9$pb{WasABvZwm2c
zifWZK?(}=M);?}vRL=&ABM`eLTwsc+i@s^8ldpi5Lr)1!Y!NFt_4nA+Z%xnAEYIpJ
zVvYfxrgV>OW?2aEuf5%Xm}u0;uTYnwFF&Xk>L}D36;w%So92)ggQ9GMTh_Rxm>0qH
z7F8!TMzN^6l5;gNVjNDYlS&v#kHrEgq55&-=D1(}5g!vH-38^SM@Q58$2=|pd~=qy
zY9&DB{^-{c3J8+nJ3qZdA0f@h-EU4#PSv(!v@2J=!oL@0JIIPZ{5?5I;GYq@ELz}w
zj@tmO5eGvi#(r8yL8T=Y%@_|Ru@;LA5QhDDG5#(7U|mFz%{&FQXkm72*a<D|FBq%R
zH`0|iry%4He9{Z*$QPqCc9)()y=X1ZO>?VMpRprNHiYynXNR}L!*k$EqNqqZsru}j
z+moXKjO5ciBYkq+TAHKF8iLt2=Tp5<op;PnQ^#jl&;K2g2;JJbIDsWiyY8f#EE!c;
zgL~J4O4q^hon@uut%!Q+B185zxsc>ow*J;rv4<xiW_54pdFqPD%EO`9JDHE`h2~@R
zsT~b!3W%_@QQDUfsEBvFjz*HnA#6Y6@e;f};o;KHreweW2<nw{v@D7^cL!#gwCJnJ
zv@MlA<vOc5@o@N1F#BMJwE8bvY~6?HxmCHqEN&AF5%h!zOS<R2WLe7lPh-mrLU`kc
zA@F3iS8~9_P-$QNXCzBw^r*lPpdmS#`nNFes*6B!&)e~cGOqLC{>Beb_V`X*T&@Ig
zKYku}2tCpU?x5WBnw6FTXUVBJWQ9+j{&&w^4enK+2OU1-i<+R)zvm@TyC=!b`TG$B
ztbxWt1#WKP9sGRcrF@KscT(@^*9Ah~2W%DWApci4_6ZDF%yF<El=e};NRTqIw1hV3
z*y@e1eg8zBH1^N9A&i|%XFvW|62JHO|H3gEA3)IoQF@G}k;J<oLqnn(1ZZz}c!ri>
zV_{^1FnqYWE~k1NGuOuFUxfVNDEalkk2aDk%T^Z|W+z{7p>Y1AbCWb$3d3zs8@11*
zwd9iEjqj1_?a?<w3diUkA~a1bmW)OU(9_f>7=Rf8BCG)j-(Pt-Ib~HEHFPs7zX5HE
z3mQ;EX{OFM16`D60%T#eyw#Bth@=?{n|LyOn=m>uFa99{G9^v{xlj4blksxx+0&sm
zb~HMqXqcb!pP<g{Y&G7>fwRAEz-YGxh~L)m2wBSsMwoq};x)!8jly<5VhdjZVv&h5
z+xLc9lOe%bqn?%afcYj@I^gM_QN|rz-Yj`dzJNLU?TW1qoX?tSfciMOxK8s40%M}-
zg3=!a9*;j@>s{`9OtEGCBJfCh12cjiA64JnE<XO3s_z4tPGO%q)zf+Sklq9J`|`k?
z;LfAqo&3L905`=+fc+^+7E^|}=R&Vrm&=S1Z>}@V<Hpk1r-8U6ccYR+IR1oo@9e9=
z-~<mV*I<Yz7qB^ZpqhuI1h;Lcr2H(4Jl@bYD8m@z0Vnxd$MzJzN9e)+c$P34V^|L*
z0%63H;dZK)>h7XXwFCNT1=deX0K*HDQt%voHJF*Z%Cf2-;fW8Ix9ccvAAIqt$<lxn
z`WJfEuc#tb2W7DPhImH{se==SG(CS{X<~Mj6)ROh_<+LLUDM${0>o(AJabTeI>#AB
zk<bW;Ch8~=_-#*(<nvP3XWxZ3LR|EO)Kz6iHl!!Va2E|vPZzxUzNzpm!snWJZM>P&
zE7=>>^<am)TfK}d5oFjX4VvTX$jB_p$OC_y&2%BL1?5Of%7Pf8BKhof3AXzT0nZp4
z1?7N(bf_oHq4W)xC=j$ix&f9$?!brU!Oz|O6yEp9xKUU(AAdP$2H=&1PDv@Lv1qAV
z^GAry9nh53wY3c&9v<pbAGBX>_nyL7t(}YZcB!0Yuu2tTO>CfN@Z5)s6Q1h`pu{hl
zyxqyqx`sOA5a>P!cFc4`>OH@?OM|f8;(dtuYSZo<QI16n*1Z=ubcERXXWaX`Xje@^
zHKOe4P?F@hCxn2t9e{0>CkP(FY62O7ESW0JC<p!Qe1ABRaQYb#UcEh>>a+~E3*$CA
zB|=Zt?=1|IahWqn7@y%y0{hf6K4XuK1LXLF7;`nmYPzyJ4*8ucr;kyP9P<G(b2h4_
z^jAd6h}L6uFy`)n4d`Y<O$Ngc*nDYz@-)1;q#SE_NoP|4>k-wuXYF!OYAr~wQgVXh
z-sv(G7m;D@Q1c!!TU8QyPR~3efALa(e;2f8gX`h>X2k(a=_qGgN@*X55NB%+fTJQb
z^W#{A$&hdxtc&p6nct$wn1gynP~%}(4*7sxCAKB+XsGA*si9$7h$iCN;M>Z)C=iCa
zMHM59DLJT;xX0G~KKM3@>|Sk<cVd18-l2!0%g@v7u0dy6NLK($5H!>4&mGj_4&YK2
ztBQlad06pDl2W|1M<i6`j*TfNIxed+4kiAioMB3f4lUx}VyX6o)mLvfyyIA1q@V<F
zylzp4Bw0q-zU`miI`3KiteCc=Ihgd6M1VpUK*Q{xA&tn_v{U|2Apu&MxM8=VSL5*M
z(}O}S^MR6!U$(PBQHHy&Rw70a$rJjFO@N`^km}g4q8Kq5O;h~l#AVe%TiZ@!>D)4?
zOb-M$BR-w&*lVg+!_&=9jlCB(Tj8z$-XY`-{?`&Gpn8FRKPM(jcTFB+9KgM(o*ijg
zy2jNKzQ@f}?)UetI*;o6m%Jl9!d8)%jYK?#adRtxQ(xMwr+--&l@fD475bqas+0XU
zwI*W47V|POW<=Uwn_s%v<M<II&_Vy!j}|mr3kX!Cqq%%M(EfHg9*$y*40UwMI3`Pr
z_aFXfk~H&m3i54G4FZCOEVIK&N|Z=7R9!9J$Xk*X2{{0+0H*+BimT8Da3nL{qBOOf
zC}xX|fEzHd7PfM|anWUSRP1Dn=HzvsJ|*yq&>Y%;j>s`l7Y(t1x2{1;hIQI%A&I<t
zw?9~7^ZEjQY{@6p4Ovp>*Q!qr@h_rq?1A;nL8zu9{D~0`Lp&S-gxUQfeSqDL(%($u
zBkWUqT8n6vi2(s+`k3%^ck_W4*VP6p6bn?QMpt-}!j}haUfNoCU8pY|@*{x3CcN}g
zbQq3mII_K(!2h%W@NSs@%<1@GjDcpFMWXDLqQierAMn-s;2RI@;e~SBy2G4=X2IM$
zte;~GB|_bKFf(BcNyo#P>rbrQvnU2MskJSkI9>7po`I{}U|D8LPsE&`N6cRDK0GB$
zRZ}odW@${h<sCI|02S0N>ea$y9s$5UWq-;){4ba!uQlE`#!o?*(;{uYUA#?~hSsC8
zv5cl>2r7e#78n%fuSKWv57hLz=ODx^cHhGMO9`dvS)u``Y`D6Z$m54sr1MDneCEFs
zxO9RU8lsr_`%@LGw`*IN=MQTy0WM6B!Kpc}!5XMsXMm-g>8V;>7E&LRN$Zmn_cG<5
z7%JZ6{%8w`?W0gT%%>XoO@;DA!2psLh2~D8^KJ;eD&PNi_rR%Li2tu4qnayB&kE7-
zy?1+KM)6M*f`f}Az#%3!L!xan=EnL9>vKIe7!;PuvI-g)cE#z~gCP|{+ZV`lP#pYA
z?5LJPGFAM&THYn7p-;WVs`$*mPU!0!poSUcU7iB=SjE9;367D$40bP39P*j5*g;Ol
zqobp~V|JsdcA@Up@g?Eagnw{Gz`o_S)|$SS+5G*uttG@ZpBZ#gLx3Iw81oLxUb5%2
z)TbdW7gH@|_G}2z72Em5phF6b@~8<<;sOou_9=$;8VFvW&;J}V)vC4t+8v^$x0Vf*
zGNBP#>o7`6#u`5+D;8iQr8DKPm$E#DIIjUL2UA9Px;z1QMii*@l2E4#LJST$7GDE6
zqoIF0?XTDL7rN}J*iZaESi(hUhwt2bC4bCzTnGYbqf)1%$$7CcSw$LEM!MZZ>PT{o
zMHaIFEPCV@Sc!`vqG-QkK*Wwu@rR3vW1e?foXzo^gm>=a)TeJnp#}PH<2!A#dmQAr
z@T*~e%L<Ll>YPhnymm4jowyt9qYT=!sZ_gR9m!}MvwU!rgt{&^JW-ngO*iKyixd%~
z%BV4NX`znlcS|Qxh2xS;LztBrBTzYz74gv{%~(9kw=;u2?UBY``}8~U=krf&@ooI-
zXOL|!O&eGFu)Nq~xk|)#cRux2Fyn4Dsqgxve^dQ79$Xx=c_kuF1wbGt{~esAvvNZ*
zP433*s))hc0$MiOmaA9ueto_}DFl~Op7-8oIZj(-XkfOC3RDOKxZcuRmp<gJL4Ttn
zuuWBT<!d4T3}~NxDmlV{IcYS%Vp~%fa6Pddvx;o-m*>YycaGr1N2!k-P8nH;b0lzS
zhm{%h|B#7CxRoX&vTCj5xV+#SAxn0mWzS~Q48T~$R(70&p=i%IVsk2fzV>lr(pGyG
z#dMd=oM<FKnd~q$NN@XT<G~AP7bU2Hf*|WmSGa~1ouHzjFZllP&?GLGdUmBAu5~(m
zzmvZD#nJhjgLUNk4UV1U`)(d)++0%*V_kGU%+jb>cXLYnbO+tiY&S0EYza2c3pIJe
zi(1LrEt0joakAaP4^JwXEi8-XMYOiV_C(_xuHBG!A-4mjg*S0hV%youwo>G){!9t?
zOP$u61x>uGUTVI?_}-pTk;NZ@fjBreU<pI=7F^}L5dv9qaxjFT`>6|a1CP(%$jo`(
zFfRj@;$t=RnImuO#{&Qp6JWQ)#hV;is56OT$3zbMYO0z8RZ&JTkDZ%~yI!62Co8kP
zE%>=7{73vGd;7t`e?`Qm?mfi@3BmVBu;cQKtU|{wp7@vFK0Xp2Lpr0rzUQQG%C5=k
z-6Lej;bvdRk|Pj%{{6E21w`z$^U%GbsMK}LP0&zJ_OzBv9&&aai5|~Hnq6H|Ccu>0
zITA)V*+<}>VPkwjJMZVyNu}6LEO4rRc4H<`j*IinhNi_vAs-o&#VS$(clL_4o%`4R
zA9b=n#FiOu!Cnjx4cBK&hVAjpkn7$`DmDuiwsJZ88g+ioL**ex0Iq7Q+nvL^t1<ZU
z(PDU0a6@DG*I#!a-MfU6fZ4$1EX$lM({H+a7WPhpGkKOFn<A|`?`t>1R3X*DIfcuh
zT~Hsk;ecA5K!43B;I3f)B?-WcwG&8=D>ZV3fZF71?ZMEAb+a33#b2Rkh$m>co<OpZ
ztLB-ClCk+|6KlXcKT%#S$@xl+5{G&dR6#Qh+bCz^$3Ngqs^P@f7U~swKPn7t>~tui
za89cMXtUv&K*E84Vb1Qqw<U>KK9rt|Gj(k%Ympi3xK*-Hkh4iwbsgyN^lbRe?rQin
ztMfs`+DXaEpsDM_`{8$(wlt$Mgi!RzVj>P3M}`K?f{M-c240LWl=;LkjCNh7NV9xW
zhRALaB@o!=Gh+EE))6xprJHHw%`&eR*1RB+Hnv_A+GaNqTd4bX5ch{Nk4}(hX>PdJ
zp$jW2u`B~5kN@JLW8{S9F7T7j>s$VB&$eG~w!WUedfp+}5pTHAslku0@w}&+vX8Wz
z2pU;z6jb&L$z<$oRN^G?Bw)VtVZdAEVT1cV%%S10<=Co2!xNa1k@%6&<|zdj<qNCL
z=sm<T_vevou-M~JtoPvM{<&y)*V>iGegT*wPVr;TfzgDM&_L}WTlnD^+ACO*IxzZ#
z1nr|NtNl3=-Ox?BbMkjCV=Q{@cqk~EJt{-KW}yy1Zp=a2q1dR3Vz^<&LA|ez4J|64
z4My3RUImVB{RpcTne+_`J2AR*h3$QGvPUgi&@Rf<W62{!kLXTi`6Msxn?i)GR3fNw
z49uh0fX3s|0C<RJay#`MKZ$;*PrN3Kj6LkyA8N)45S~fngju%S$@TM2h+Pr;Oi9O^
z*ZGOs6Ltbm#tryE__;RUTDpRjXjvoj=ADqYeXYc4=^jQOGeZ)9|E!3V7z4Bt5<rxW
z$mgq<9!;qrQ@MZrVG@ecC?wZgb#0th<xf5kZrc5PrH|>e+;NTJ)$!Ir%<l6dw*0-X
zUuqF@+#CRM{)}T=yGu!olDZJk2*I^Xo*NJ66u+m+OtSD!;b%c1MtwgzJ@*4FriEg+
zNnw>1<dR;F+K1AcoEl>CX=vC0=^PBSsC>_HADYEWXul?zD3$;P=-Zi~{!Yo;AjYg*
zujlHxKRVhjy}zLZ_peS#sX2T4xmMDQVzY<R>mf{RFwr`O>UaHiIUSFb^{Z_cr>01r
zaBy&ZtVn9DuSflQyxcmrFvq$8II%4@xk??~<f!jnJs3JBypR4h$ho(>c#Roq1BMZ$
zP3Oy)0XAGJuqXrS&zFSpv!ky&I5Rfx_eoJX0^_}O;0<;lz7Tn6N{GtKP;APbc&7Fg
zipZZ5@IzY3&ATbOJHzAEQxhP(8y``fLu=;7fNn9x&8Z?<tq*Cr+$^0GF#MXu)c4NJ
zL}%kv^3M!%A%Y;YK|SRfo|OUmNUrAK7L82S;c2E?qvRblh10_{ke_6b(gz$3PRh+O
z<pGi1PNpn%BYTg-HsA<z;{Q8BVkl5zgg+MgJf;L#!dVsHRVLsp80FtZV}z+wGG_F?
za0dtzRK~X)b$Qb+_rFC>AH`Eo$;pn3?1o?;Bw~W9g}^_Htx@Xd4ZbPn?ZE~2=J=j-
zS~~AJQXHhHk~;up;=;ftIiEWi5M}WskRw$p#4aawi$<?-rn+%LD~gI}r3H1Z7BpdU
z`COguTN)oomLyU?3*S>q8MUAoRIyB36N+M^@t-Gl%dij%BtKz$3^r|^lYNptJNofb
zu|BOYNEEjGMw6J7)G4hbW^NBttKMw?Yg&)&zQb{P=;=JY%=T>4A{8S%<L!>d`7=4j
znSk**C3F#cmw`A-J}tnu%QjOBy{xg#%P~iVqFQo>bIMGSTdMGGN>TODNfFA@VpSq5
zp$nqj;w13s{qm;P1dFund9y2vZ9&1@bUb97E2e}U9%uRRH(i&US!E&?1>m*K)2pI^
z(U>8;&rEn(1l4zPPXDlI^WWe5#DU(Yg0iOKvCqN?pp5nx0m$p{_Ut9h=h>gCeH&%V
zQE!C!XicaA)RT8A4Im~@oMa`m%i-91(bl~#!qaf<@WwP7;%Z(YuZgggX8P&)C0NO4
zByu$tC&#|wDY-@=Mr14%QN@hj=H%8E%1uq}549^5Bw^5jWWGcu7BD9qTTU?LFhi}G
zPk()5onZ88$;khw9>7-?b|b)DUMzW}7+yLfUJ_QLv&D}H`V1z+>jT48?613kdp>Ks
z6m7!w#QP4^ub)lt_UnXZ0#4uQ`UQIKsio3XP`H5(OMu_6X)jsWl3n+wHajjUzd8|>
zo)CgT)&-@8dzL3)*Vh0ibAO&qYUrnUL!_MZxk)giOc5|0rWeLUjf4Jxvm{R^&KrDP
z*`sQG##jPW!OHzc5@rDLti)kq%VDm>VN*-Y_Y%im9_hUFO$dQg+BbH{l4@B5!heez
z{U2nGlQgz=-sYKz2s>ka3b4_PSQK06oAimA2_$vrF)-|zNxpIWa;B6prB=$J@N}p$
z2&Fetj#?EILX3bav&&>i0OMLY0%jeR{c+}>s5BSb?9(fdAmvZhgpqp}YLXoEqD2-R
z5>{{j08XhAMN}U^Ytu8l{X@ehENIi309fh(Cc@WK%6wPC4rWQG6aNJt9l2OXfjrqI
z)Ye`cCR<s6&;NjzMc%ih0|0;Bp0jHJb<>guPVh}r@DQMG0{Tj^n1Yz;>6em!@(J=A
zcJ@D4*#rgMm&c!*+-^vF{7_5H@E;~1y8+br=o$I2*KAO<i{Uh>=x5ut;SU)NWe@Rc
zhUhJ5Ab(G9Ef!5XOZ`3DqDY`mnOO4B@FWj-86lw2`3aML;oO5g{~Kf|DF}eGCPxG!
zk=&r>6bi9m@C~({<q*|Zz-Eyv+$fo9NR7)Xk=nx%Oa401Xwjn;ip7W5KAZ&ZlHULt
z%13NWcI#hf2^cd1TA$Af);fX*n5pz4`KJbFSmf8eU4cWELVP1$Tnr<9xHw&<U{||x
z!R^{7Nal|liO29F8>YJZc0WbMsuayp*fVa&0PRZ&C0T;)ugV00d=fQgr0J2G8Ikx4
z&$Q_8r%$R#!VH)*N+U2a8emCcL;el`1{YyF^Rrl1okQ=NjDDrXtJXnTCX0Mp9(}yr
z>$evG)Q?eYKZpltCT?vVxt^?EkNRGHH@*G7S=D*jezFa8G-CEnpQkqn9)1z{J>0C0
z*-G8$OWw?nefV=fnj?1IzIxNHb9~uHd5|Lp1nO5OPVV)&bwZpJ2UPD`SG#X}nm(L@
zTqXXpb?~0qXd+d}6S2#b&@AgFQ@)d{{@_pmtsz90m*pX0xbeM3-aUD^)zNqk9^$^8
zk^-(_T$0T)=~m-maUxh66YlM$(_`kO4|voiKY5pzMFsS%3y_&UG#vl|E{g1dMK`9o
z?3XHp$=D3kTYn*|DJISFm`?dmLKXZ7FYiJaAS+f|d&ARm3#8xWsF^&Bg{+@WOsJq@
zVl<ZbwF)(==?<!ygsyG**wC}o7d)w;S(Yt?_lQJ$TkF)1+OX~)jietP)vjlF#|KhV
zD*20iF3k2PAp~kIAai9m)I1x(JAx6hQ|!jOyNDE!efV*$snhzp=h_)TZ}#a^snWKG
zf~oH#gr6Z|%%A?$)K>#M;yG;n-njRp<T=~pssNYHJD-`^+4(wK0$2aZtk=L|ui!y+
z2U|-3AI2<teAylOOS0^PuD%413^P4&5AN|(Nfp1Emt5)V+DDUD(TZaZWq(ESmM@>&
zl4eGh5K0wl*+};5=TLBT%l2X!fyShhGk-I+@u^RN>I$l(WxXfB*^+Cp#yqW5N+D6x
zLrIGkiS&khjNCWCW%Sa9_sP+ZINOKmnXP*Hh*Ui+WuT13>E-Ao29JnX>}^H<<Ku1D
zybk^@F(zjDRQW5==+U49VW5K9s)LM?d5xYw-ozoj3|p23ObDOIRwMy7R#!ab3;Ude
zDjS<*qZuO4Gc4%H^a5zz8f1MM7}>KZzboB{0M8aM?5#8IYN^f7LIB#C8J!->o?7Mp
zl|6PP47ATZ*>QUu6yK9|>&^kPf&W!!YK|fGSob^-n8kdVft99BAN+*2%G6%GI6e05
z%<}ELS(f?%Fd@KP1aj=4fMSF#Xbml$f^I$3OgMasXEioaeFLnj%%aUFfc~zBA{vhN
z7J+)oZ0bM#&~?yxUX_0Qp#Yu|e%HD>dOviJH+dr#@rjf4CC>CtgRa|z<(cmL7b_0t
zA)aY*^?4t_2<Z)@UY=QaLveMyW|(elMO+Nf1N(KwR7l+tFyGuq4gLa)R{mJ@BOTyy
zW&ZNZ+nOxo-#z8Ve7vVClsei#${U<2l)yjt_*%m^p;n@lPD}Sq`EBK4NvS#m5oiuN
z&`?N5{l#8i!`F~Qet^#|;IPBIg9wm1zMhxOS$?6wq^G8b9A7~^HB%$b38UFoV`hl<
zXX|O><X`~w1kz$}V`OEdqM0>~AjQDuwxUSu)$NC;?^n(-G;Xe~0yt7%0a?roh!2mc
zPSPXm{HCc4JBcRuv{~&Xkt~9$Co~RjL!4L9Ln)&{UG5D(tA$VVG=XTyiXzt&KWw$W
zzt3&+xw^~_{^K}>fzgjC0&5L98M_xB(&1JB4coD1>Xj$T))o`-aB?aIhydW1v<uA)
z#O4#x!q(PIO<_g|Az<h{hl=Z=D&?uhY+bQZ7Hbi}u~S@l!gCiH910ESlS+8eS|kOL
z{Tsdrt@OtjG5rfcJ*{)=1-eGy5-b9W4UwJf!pXf~<8lP26|RBr9uZ@>B2#7CZxzX3
zp#<VbSLz>oxSE!_ANSGs_M!(|9InT?zF~S4@a6hoZ9H<AzAyA-8=Ice9#|QT-v~)v
z_UBex{#iS{%pnvX_eV2bx;C_H&0!`aq0);Dt7!yw&A*Xgo=nqnnG@17uQcBN#KrYI
zTeBYdrYI2KqlaF&Xu-WXw}w}Uez4cIw`*bi$B0wQrypI~QoGiMj>9u@6W!*feoS|N
zh^VBl-mwPfjT9#q2r1tP-@D3R4T~B9!aX3RbOYOXN@x<Mxz}ssf+2;xWL-T2Xj8Ip
zed5h+VANz3pceNcbY!%1C}A|Vqu?c19B}&9fy>a-yA0u9alHp~t6UTxDaq3V57lPp
z8&4(T8=u>Fr&rGnl_>y)_A>z2hf(kur!{<1h;^VZQ$ob)${o*1=%=|;SVf&Zvj9lt
z?=qvBhQw?&ux9}TJm<N=DES(zAL;K#$)(W+T(6SryfgnDLIBzIlM~Ow<JOa%aJt*I
zWD}_w_*U6c@WWjj&`pZjF9w{nzkmEJy0peU3dbOl=Q@{kw&lJq{4-$<Z9(TaZQJV!
zj2|?mv><)ofo$`t{p6I6`<^E!(y@jN{|_G!{TB>ULbyiKxhm4j{3CH}51ewL>19~I
zIRhsoadc)r9F$}(i9X}#QXW}@dzU8+v7~Nx({sf3+=aCO_-smu)vppME69!ZGuH?3
z54Y2Zd6rXaDUCbw`zL}eck>YI$0xokk)VH&2)TFfTO5OY4S01RxTTdP&Rz8`0{m1M
z?o9&OkOtD}{7U(_3W=YKTXzzb5=&yeMWp9AjpIt=Z~c|F-}?K1Pn{NIdzW}G_3r(T
zXZN=8@Go(mXW`9H!28uG*Wyy70e>f#(JpnBBIbnhgfgYrJp*|TVE!QD=qSB7NL@-;
zyfLZftnz{Mnxt#la-#Jic8gS)FemsZ;f>3`r+qW<3u$FHT|3sXsy)rIvL&iL;Ih_r
zRnn#7dgmj}?8#*dFw;^!^Lx~{Bl)1J^8+Ma-@sr)j_xlZfz&Mz-F3gx<Iv0KK5Wd8
z)}iE8$B3gk(UOFg9e>bK=%wdk<%9gviVD9TKw;y1^CPp(X%l3WMLK1bS37<gjQ8pG
z_mvJ#DJ%jQ9fy3`nA#LM+j4yx^1~0Mgk{N&-E3^>m64#mop1R?#(4JMVu8wDIs{=I
zxOR1G${f5W1|%@0fHxEc>^~ZZh#CfLZ|eV}7R9Y|3aOOiM~#P!pGx&`2z5klzU8u@
zn<X%Kb62pMA7){+W!1YL@#Fv)gVBqB>oBawWd4PyW;|vO7w=tQmo9M#;a0D?qGixo
z*Xs7fO4En>gs(P6zJJb)s)NXbz3%&3Y_)3YNcjy3u!J08csN4<2?j9tv`0l?Hp;Cu
zs7xy{fV+FHTgh_hO?9WcI6y^j(v9v`D08#^RE}LeYG{D`_XKd%^E~|YxpD*eh~A%V
zY;5Q_-8CD!wv$@o&KmpD2(P@}Je}R$-M#)5Ej8S_a{l}=(WkwCpOKRz;b__Jncy9N
zUSl4pJ3b;nP?gZK6+jm}xOU82L+QRPm%Fj%{eVM0!u&{66Gf;1s~<eOyt<7(D^j4{
zy7_%ovHfb6&48vW_g%hVZbjV<A7sWQ|9h&*Fszoz$H6Qg7=xy%)B6E5NHv7)(6A7H
z=KOQwsD17IBzdsk;Pt-{FD5J>y2SaJzWE42pt<z9x$!Jj$E6iCdUieZK*3vs*79Kz
z5P7i=G*>&p1>c52aP4}LCZ}VVHBpICk)3JG=@YdRcjKE5AACRJ9(QkcZC@WgdDsk<
zT3gkV9PK#gZGC2e%tk>T6~3H}p{EF<+?pzQvYto8;bAE<KA^8Lgzm_lQ8uAMlH_ry
zOeem5udYNVer8tvFDZ`E(9OR_@)$h@R$CE;09W^3`Q+8-mn*kB!YIGatAg*!<IYDV
z00OW+?UBsHnCV?=72o;wqmZ*a_xk(&jGQS7y)L56q5s$?^)0Nnj>hxUD^knc&&s(U
z(3M8IGl$A!buGZoekt~%d}jKa@eG@JT61DReR=r{PyCYThO~A9>DhVA`(eG7HyxuZ
z^Xze&yqp4;Rq<CdNs_A~qFWfMa`+b>aBoPHKEH+`Jn-*Y!@n=ijZX~487k%Zrkq-3
zaxHWJXF|hmcPk5$a%jPuy&I6qalLihxz!#Z`Q0K=czG((;9X53DH#Ke9;QGTOqQNO
zGwj@bbdcv`GSwG*KfFmy(p0CUT>a~AN}5%c@DmT^6BotE?H}r+KV(J;UTGieQvo){
zfZ{Y>Z@AG1j^HsQ5OL+}rI>5_@Y~LIn7*kzP7u#;|DbfT<~l<-2S}1YA=<LZ|JpP}
zGzVW_Z)Rix7<^-EY5};s=A*{N=Ce^Tzdo@QcfDT!U9tnS)xYT|%h}zcJNpdE^rg>p
z^wX|+PhV#=tc1{i8jm6E6_G8*OkJ~++U>#{aoD$eB?|=SZii3#79C`K6KQhG*)oD=
z>{w7?>WT_{oPBrcqQUYNE%UwIv;l`0ulTD7PSXRk2Y<>U6B_CwFKazGK+i7p1G;{J
z<Pf#@!VZ8=9}#k`D9`n0tn1%b|IWz|Xk=tr2){Pk41@hhA6AEl_N=0SLOiG>3Au>W
zs#gn#MMd@#7mBeYvt%Q3*@uWJCN$*oP;-|by_#_(shW(ahs(`0d&87;n?vXiUoAG{
z4jq)dbi(Hk?Rj2eWt5_-y+n@3_n7I;9EIuNkRHkN4&&Vj^J(J+*s=@kj0Gs}K1*LY
zA675<@0w!Ktpy?({Ew2GVr$Ze!Je1_qf)lchh=ZhKbscM9_0LPmk79UJN-^h_aK5R
z7XdLAtn=)vIipn|lKFLeM{0Dq>qLS~gql4YLxkM>Oi6LtRv_VN7?Ye#Ii4IBY|Q)>
zhs~Fl+bhKkm(TI+b)SCO!>quYF>uJY%rjSIDq$dTI7cF*y}=+uA4<0hW72u$uG=0F
z#6VcU*<M{pivY{8AnF}R5zJjk1YG!6D-UBUIem^k)7~8q<qa2C!?u+z>}5sTlvoT{
zNGOc-^f6(L*riMFgwz95fqes@g_=c<<%y2^BbO7>BXkr%ii;{kqJRvHJZB`UXs2e5
zz8U2|w#U2c(|K9O_)?iPgX7@NpQS2$S3IEN{7}3q@Z$csA=r`aW3HhzF?%RU9+#bS
zi#-FAHUdcBFDfwE&=l5TTzGPY^Sntsi!mP(1K+S~$IEShyz$){AdXtEdB6_~VOIVC
zI0Z?79%w|SYi{Sau1=s7Q6C#L%K*qgE7hGokCr`C`76q2kz#K<F_YnPzIY=sSvDv#
z5<WsY&&=rWfT&xVI#snN9|K9e05)4TJtpepd5BG2yxF(gJiyHdZ#$o#YBOqsa8JJb
zdz~OzG--wq45M(6_Z5N6S0g4_%z?b|KiG2iTzB1?c73B|DO?pXFS0r#r|ZK<W;8yp
zULj7do?VpwyAv8hxgc>A8a&+u^cvRW61FGS`WA2zvM1tc&Kf%!4}UaTPC9PX|F+US
zoYL*q4m}sUs3Wje*ZrKMuld(%2qTb>yxOuv#;ZvEC<0f1{xy-F0#xUNo>=?n^Ac7n
zT%4Y}WyA!;XXG1*s=NJwG$Ywb1$o~*f)IHV)NWJ<IJZnQwn|;%Pe`!NbS+|AQRnuB
zb3bVOoHrIIaCcsoyF-i|ML6@n03ho9QVWeH)h3d9N4Ckh#5E|-4(QO3n)gtze23KV
zBUQH8tjL9aYm1G$CSogWz(<W0uk*A}V*SJMYSKqpP~SKATU~-p`^}%H3jt6P1>KkD
zx&%FEZMMX`_|6A4@*>f?g!Fdu*t=%BnM~)N`MK0Hd@sOgOo3iAm0z*GTYa+YjV<TH
zM9nei?>f(x?mhf~wU?UNeju{kO&DeCW5U@JGw&~5&a6!Twn!Q(lLx|AG7$mr01Fu|
z%(C)A0bnVzfZX>Tn{T}~5xC6fukOX~657sw++@$m{=TLAYjgSOztdo(BaFS>`JWcx
zmm~^JyIWvU;^vj+j6uNRZeMN1<)JwZ?vQ(2<IiS={!Fc}5l0)Nd2cT7cHeLpyK^S6
zlSPY^0idIR@S*YUhwn&PVpuR=C!Z$bu(tw5TAVdERUKJgudj-ZN}@Xl{phaLj{$E@
z8}|zfd><~cXAVvdtO6US`A25LyVAmPSm|PXIM<~z`2*~#{8)*pR`#0XLxPj%r%ua0
z#Pn!@KBFG7#)kF!zgl`D(nI{|D@UbuTpfF9ChTz^qt#9iA{DF|oEKCoDn~!JN%g3$
z2=j6ZNmXM9c{%Y$MaZIAo@(b;pn$YRYz~O>5fOv6O&tgDUf8H{ca%X4;ik5B5<<6k
zVnnc5XsdS}j`r>Kau`&L04)?3n8H?MPEt~jm~dQ;m3+-H|9&-i@ZQQfY&7APTmGjj
z^FKI|Ka}8gZfLw@ZD4)c@c_k{=8LMrgx``gz}<fI<Xm>SEb3fr4-mH2Csz1TBiSHX
z7fCa!sKSO|W*Dmdcx3r)K~zy+J!#i!-o32UGy7(*DPJwf;D}RzdyTH7Be~(4wwM>#
zM<CB7eSiqEChF!{9fHaYfZ}yS1vdjd)}`*3*m{-=bw$k1`hvp<+W+)89Zt<@>Uil6
zyFpOb)*KSHO?rEM)cm%@T-`|)Jeptet#a>8A=fvS6X1}wqU`_9$H@pey}bDK$@01T
zR5$CcO>_!E;6?7tY7(msMO0rrlePZe`3&(tGy`Y15tbKqF9}imT#VAJ-!@!C)ZziH
zK>{-tbJikn6L&m}hSFxAC?DYazNV6~^$NFfdqw;0{l00gzhPOJ#*S2*)9Fkf8)^U0
zTGE+1>@cH<FaYL74^4-lg&_^)TOHP_RN&`}nwEY=`dpon#~;`hrUwh-$FLOK2QGOG
zsp)1jYYIp?($GmQtO?$q1k>$4UzKi$m1tvK3O`?uYB^bNd`n*AU7#qYg{$)J8xnFv
z-Ze@vwE74;D%YR9vGYpPR+iB8w?pRYonimm3XO5v|K>?fcTiSX`xjU5eP~~=g+{mE
z17=<)yiz{PE=@@3xl-pAuk?jXPK7YTp4zNQJ_ap8VDAfCurO6z>8Je&vz?@{ZQFM|
zxXOqvd%D*+#!2s4^cfO2tp_j(f>OO6$TeEy#KwZ35)u!_`sX!URmPK4L(0F3?r48^
zFxRI|#YC!mD-DseJbx~F^RQk2H=W{eTcIa-XKjmP;CFTaIcQf6P8R^OMk}7tR8}$n
znVM$e(qC_Gu33{wsZs%kGXXkD6KxXKl_^d2Njm+R8{#BMk?3K1Oz7kv`v)4oTF7oZ
zjnb!2qCRkcA+dY04U`jith_=b*OxClR$6!cY=K2wc;_(#2L1j*CTh<1u>7s&-lJ-P
zHoNnuZ^qg?qrC|3jN_e0zu#4PxIcVeYP;ZD@m;ARj7Df;Gh&TE3P__5vsECSEcV3=
zI^N%J^JH7mgFAoubUa-6hCNv->$}{}aWx;>IfEu}-hPUZRv9RcuG}?rxsQ8F=7{rR
z6m7Z6P3Ds0P5vG6{qHu@^@6zMax2oy%ALb(wYO7{trw<OWD|}nYTj#PhYLaSx~y~l
zBDK7C@<$Gj7{}QM`XIk6)|dt|%i*4TKXCx;LaMH>_Zk!;CnI|Txm@z3YjSS4eeY9X
zCyz+|AxF+PurijTMV%Gk%TSEMnxP_BJF(V#kQhQCg+CvQQcg3weDr?vWb0dKutZ@{
z<MY5IoDv=XNE>(94{qFIcNhp6H~>H<Z?4HF1*Y6{mzojrFE+Q=<YmXZ((M2Dg;IqY
z7#r&uC!iB>=DN5Lbr4X5z7Dz|?YLZ7RcO{(^=f~!&EICe8eBlDQnJrLuJ+LgGW$fN
zGA=6u0h49!+a?7}%Z#rK?oSb>y&n5mq)msfnN-IfC=E{KPJYY~9be;ef4QC|I-Z&(
z=l&U#CM<d7Ld=3$yw>H?h2nVqR=-v^VzH7wXw>(q@s$VN@qb||CAaGa_0dG&E}~(1
z&kJA>YB@!`ei9lsN#358QM<Mgg#i8Xj{D+FfKP^_d^w!`qJkWZ`n2H*pz13W%KT0o
z2ekl^%+w0hmdZYbD#uW~7uO7P*@ID#M4(PlA$7T({-BnM-{+=}G~>g9OmPg<7t^xM
zC~k_j68K;UFe%WZO`a!8W4mP<8D@Jz)_L@J*xAe9#ZQ-_P^6~m>Cj{eZa%iVEAt;|
zReaQ$hY8rr>tHzM;4a;^{o_@oW~Fos9QrlCGdfmZS4LN3br<s!hf(KufwVAM_sJlr
ztPBTsX(YBYD5LRk@k-Ps_W7m)MwR94-#CFGYekvl8^iZ!=QOVL;=9+wXbgR;VaEyB
z3%S?cZt+TOf?R8ns&uw~M>opC#&Y(ccRc{`04zE({f{x4|Mp-Vj7}{33sCl_J+4rF
z^DCTg3*h%}usX+T;XpG+8v@`B7yq8!+myAX0aQW|l`hB#R5LU+=Z^0DvnUP`YXB~d
zMD(=BMw$<(F<J59HM??I-dN?d*bJA};;Ih)uC^$2%KibnsU2`A`7v`U-DGB8l<^ru
zKGcly<x<_J3_4vWm+$o!(U>w=z=kF+&Lgmy922u}{P_}(UlEC16%Ulnt`A9h4Fymt
z0N7vFQ92YhCKxb_L$!W|e*W!t)-DcT@^7!w?^POtAUO5Hd$&H1-VPT9Jt{cXS4<Fn
zv7)0Y2)%p<EwV<9dGTuex)T#05U1uuI5|YADa6X%#Tt}iv843Yq}ivrK8W+T-+h!S
z38X9Zy`Oe+ig#V9+Fii-NB8)qoxc6fKXqVoDH=ie=Ath5kw7WrRDA-p`i~Ouw=<a>
zRc#17Y9aB;MgOtq6%ey8w2bkbY8PfVzk#{O44^CnCL1aAIOEu{ox8$#$458=()ih(
zGoZ+g-WH=dD&{xSwo_v(7oN?wESodeL_^`Rddxt%TwP4+xvn-i`ZfZp;-9f2U1eqs
z%dcR<M<!;m{}?-Rlz!MeUG<88L`APApQyHlH0Q_be~o{@&f~AHx1dVe`juzV&ci(9
z`$3DktD9yCu)%-u4^64A$#FZj8m!!lc(+4rftJC0c(Ma{-b`;Ix6O#|Cs{Po0<>Ek
z-bf5{3dOE8W|~skA{oX~4a6ewgv(Z-q~%_MWPuneV-Jcv#cBTcKF;^9ehG}-_8X;M
zz3R1;G8B_s{{kRS|G19B^QO+bqeokB(F@Pk<((#g`unGy_ntNG13td8t$^fUmpo{l
zTzz5JavpeNMV7edSGmA&OC2W0<}GWICTG^;W3h@O;iU(6hpd%;hIUeCbH<m$GZxPr
z7_tl)bOZ;w``r)TB<NAqv=D7(7q*@{43-(ob6ko$(1|+mFWnjJ1dzV#L8w#lxi6Kc
z_<(BjiHKAN2CEq-(_}diy6&U$F$k<AV%-H^DX`{*R;GL4n)6XQO-_X3pZ<4vkALHb
z%i&3GTn_#3r5<*h)uZm0zhfjEDWkuTwoedqU$*?jUBGox(4ySX<Xwg)B6myYG5_%l
zCqZapF4WL}HmjxOOF}K`f()n1@__Nk2K+1M$1&O1hfo|TF=aT?>U;Qc>{2VLL&UVl
zWW+tpG5$*ip>Mw42)KRvJ{ljqoKti>Tm0T+5SUh^mAqXiIE&<d6AI9#fI3&)<TUox
zuiE$RxW(!;DiUcxhKQE2hm3Qi18<o_l#&XGJg&*z^Nq`3DRBcXQR%hUD|f-8abiI^
znGr4i5$^FO`Sx3g0R#C=>|>MRGB#ymu+2U~F@-RAHMr|8xC+Hs+=9;6^u{o~RGFBh
z+=jI79Mf+p@?8o~MieH4oqQ}tD*#Pb*AVffqC45Mk}TAw*~9CG1sQ$Rd%X}LEs8%?
z`2jJ8-->>(2=J{SOWgVcz2|=$>XteZ>mb9tP$6NtMz&D)vB=5RDm>|x6&YeQz?9Fw
z99++WMKWU_Ra`tCRi+3PZYn2_f~p7!AV;MZvvXnGM48yn&BOzkB@k8~{0BBlH(IEv
zNm9+*8&l``uiEZ<eHhVRS}P+fgRUOOYyi!ZkmOomEQ@mY>iyAb(Bz^9Kqk#Jb)+_R
zoL6Y<H?GXPy2EP3r-rB0%cUGMv(!13BpmuS@VuTAf{ENpO|_|_2{*IIhM3~8ncus(
zJZ0~?p5)zaKtV{;+f)_(9Uz#z$$cQ3l(?<-c`>zh>=^uRn`Twrgi4QZ#EQ58b#8Me
zS{+5%+PuRwF$#NE+EB(C6h7fmYFb@J$^Q}u+D!#y&+N&!MBS#h|DM<UPt8s%(<U>&
zq6~EuAtBjggHei?nc6*G#t3vo+(^vAF?peDS>dO3`;mZl#|su*>=%{!?_a9TG8xoZ
z@Jx^_Bu2zSvA^aOJ+aNOm#=46)zjaPRKV=>;xg-#?_Lj*sl+o}By!aMzM(l;W!iO-
zr(UGtsz%`AC(<BYlj&G8aSAv8cN0gS-&X;}6~X4x>(H8En2;jtOs?Rvob8le5k@-W
zM^&{-yk2PzEl8d~8Vn7LR7Q-df`}rB99WP4oZ;UW#~HX6UcP#fFm74|Uj${hg-IBi
zhO3Is$tY`yOh_XDpI+^hzxg4)F#V`h8%-EI5*QTRE9C@HrGRKv&?5n5@2>hYfa3lC
z?61%MnKK%LMKwIO;wW7c3mzHXjyN}}a}+JEjXyU^ux%CE7k_=N;c@(a^}N)X+xz)W
z-5yScDaSyZc`dCa5tM;vw1B1vxjbAGn?w_aNh(n2CNAkrNfxzd0Oca&#1lZ40i&JW
zog&*SFKlt_-ho#=X6YpN(=UXdZ;a2pul)`&sIe-c9ap|h{udd18&8n{)R@3BTh364
z9(6U{dxal!{x!O@em(_}MST4%lx5e;!iI_{DTUb`Y6EI-vvQO(UvLmF+%DyjF*;Sm
zck<_YqMQfz9I?vhb!pSKS|Ta8j%sL=r4=@Frop)*b~LvXvw86E^4HVOAl&RF{i&F`
z16V}wx;Rzm6KP8+w=X+9%=X`J2Wzfe`W>2~O=q%*KYX7uxqXTXtWpYZU;kH&=-3N>
zhEcy%D?-Yr%H{Vuiw!a_jf|DX0{pyRMw#1#UwjQ!U6FOZlN4II5&r1pJ3iO@5fHaI
zQ%(+oYDu8IyObO#FLeJXH=s`!c@bjDfS{tv(%g0O-jO7mS%*qv)@0FC5lnb!e%AU!
zc%J8pNsYIYy~4y5%Y83#W~H1s@d>}mwnaXCIctJcc@57X?@$qm-Rh`323KFb-~S$A
zVf--rkBe{GxXVUo+a4&dKB0Syg-n)CxtZQse2KLDQo=K%MIifZA~C}t2IX(6wQ1Gb
zxW9~Vif1!5|7^^hASwS#>M>e%WEqdGAKe3)MPv#v@&Ya>QYgtV0dSL^eHPG4<OAaY
zC6U}6F<Jz%VzG>}GMHrn65$qsU8^$Af!lH)MX&`5*SJ?weE5~C@p`_T=Ihec`z_NT
zskoEBcn;k6Z(W0I|D)}S2I`q60py7`ph6&T`AYGRmU|vJu23KzqPY}!bTYo|#^Pyu
zZ7Di0r@`OCV{L!gx!$#VkK5|#yx&gj4Emf=qE&Hvbh9#~5G^HB2t&`Kd*MFxr^)j!
ziq5J;u*Zvwc{1u|mMi&ffhm<Nn?A_F<^U-xlZb;!u2W!2PB{<N`WRfk7B05hH8||q
zveO|YX0B$yvedTMaJ$*qM2!9dun6%Pn@azL#=3<0Uw3%8{#a4bDn1!urq^ca8Pp``
zt1}=+C&joS&HdSy7F_4CY~OI6u{TQu2CNfw;$jA2im5EKQ7FrZ>(K4BYUa<J#HuEb
z*bIsz?<;2)w|iX`3Eg_)?;F)tF#Y{}dF-y!WM|e%LQ<%jb&I3DyhK_#uHe}nwj<4f
zf&zz@lk+oyN%uL<5U4VMaFCG#6<uTfhNU&5;=?VgQnzuJ@}!e3<(ogRlI)|mJ_(DL
z1RnppyHCdqe7iaILm-3^+7sRcTyC6h5Zb>7u0(H~j|Jb{`Ejogz7E};y0NP5l5n8L
ziM0CsHs4g+UX?>n{$10l>&dN%!Sp`Xo$wH8rNp10=Z1et=-wO$mk4>IW&E()Og}L!
z52s0Hc$`JpC&KJ|e{m@?K0PeZE8G!Dk(&WyX!btyJx$FxiP)A<%OL%oH(Kf`*&u+%
z1dX(;rRB}|>-JxIv5^FZQBR6W9D#|}#7J|6sgwwn>daSL3#T<LbUXW9<nwysZolIa
zw(^A=O~0+YsM^Zzy(*AZ4%Lp86}#Zv1%Y%q69o82BH~_rZ#fES@&7tey>aoKXh9Sq
zUDH5Q(?Y7nwS<%FM|ZSsi1`VrvE{ABZc|Or`X!%fgxWW3+dNasFD?0oc4>7U;{1IO
zApw*UYNR8kckdv(U@n*~R*IFMZml@g>{ea7AA0W9G85g;q>t<bl7BhAz3-q(s;BLI
zKze)~cO{NDq>0fGM$}YF$0SY3A$sMIIc~Ok7JOzuYj^ePG)MGeYc39vsj6yqdYjR{
z(dl$k?&eMD1g)1z_I+u3P%8Ev*sj>|y@<mq;W2I1%ZlG}rom#SOTbv$9tlo(SVPiI
z_CU(6tBc2qm244evAfzd|FQnSfQk}V1QZLQndsQH-w!ZnLJPgx)!~B%+BDTmRRAE?
zpy{#d&|9L|pQiWCic5Su{r@Z8f6;?=S{WJWR@nhMoMZ1>Y(Xu3ULvnl>*?iHL*(lE
z#9>;r>*1Qt2;kE%78I{4ij{2t1mYiQF3|-46zk$bK@#Qp>|nT_n0>PvQ*mKJ1*Jf`
z+Lq;G@}K#spM<w(bk%#4C&;#R`?lA#2n<P8j~=?@&iw?i*!$`lQjZ{UmlMq@dfhN{
zYpw23&+?rp0YZ3v;d@7w&7Wjckya-LU1>I>J9_d#XG+rK9~-KiW{R7Sl9^uxJ#5yU
z{f7t2)>|lZ=#^V%0wY2eM#uI6tQLZw=Q>R$)KSRLS?qZ_boXCHeK!YY%#B1}?RL+L
zy3_yMrn02RV{VtN>vYU0i*8#xmMYfE7G08HAAL3dYg%gO^VpsBv*g4wff6EB=J;Rz
z{-{aB*^A2aO(q`Ah89?am=l_{0u)kMXuv`xw{9(%l9qk}VaWt$xFy%*QjGxPY?^1N
zJ>BnrRla`TV@s4XF{1}3PK*jOago8~OtkOUjcyvbqvhr|-CpS0ozp!0SdYL!<h-f9
z&_0gw@zhVLZXiY2_o+z~5K+3Jg%f8#b+Z5N<nVn<ooAIBc1Vf5ml)=bX()n!!A^F6
zY>b*4NWNLtML?cFCh}nUz<TW)7}A^eu1sN*7$=AMD`T^-NMR}}6sg@vM;AT9Op_ud
zBFNlcp<2k}(3m^jaclTUO4;oW^HaF#<zq$xuJVXCA|%7VKTiNi0l>sJahJ$bHpdnl
z!PzCRY9%T(Cl@pd5JwNf#C#5xV8;@hYCgb|_y=(Fry1<ZtN()|r$yCxeAB9<=PMc-
zl5-<9u+VfzrA)!=&I~gd5xAQekL}}!H~CulXt2Ol<UpA4NO?A7Hv8UqC>n(=jDo@f
z%TP2{5&VPf*@T->>nj!f3sq#a3JAS+hf(<B_se<l=20NThJ}=6$UAu+(@<WJqjGFC
z2&1wiCZ(4hB40jhX*oxBi(i5OrBD?o7t7q-+l=x^T<oFlJjO&G&b|b3y%}N*27(KZ
zH1y0t2=L#0o)zs2^TA|*xUD*uR*B2fEHPOLTJ-G(#r?>VdUM!dQv|GW%ctTdY-*wL
zd@Vw4O2~qEn^}-+2rz*<=ZS+5z9YyShrJEueoA{z#PiPi<->nbI2l+F7_S~C#3Kdv
zGi%5rQ%)hQF9?A`Z48M<c<7VjlNy=xEnabLbkCD8pbqIrN>^u7W_$io>9Z;*&9)4O
zF_rm$6sEwCncvRNi*|NG%_HTADmHgB`brJ$LYovUA{4LI8!uNbnIv?ixw#R^5m<AZ
z{SnIvJeVATyoq=p#*8wI6m)n4pnT$B8rjIRQWXm-@TqxxjD8wObbwvMWvK<r+;|K$
zWvd`Vl%U)iS0&mT83eRq?(wpQ+2PEZPL8Z<EI>T~%-no<d1%OunPR0&^*3@ozL=<I
zFh9a<)+uM+KZKX+iRF6ebT$tDn~2l4b){L8TdM8B<)3@YUm<ytzU<pTXFaCUn|~hW
zVoyjR+K|$aR7?x_t3)CrpulLaMf*5zmZ3;;>=y@z8_MIC|Fv`2NyLYl>UO<|R4t&3
z1Pty|*ShDaW=ps>r1{Em3nyx`UjpjSBl-2EtGxml`~C(NT||i&#X8SSmpe{F&U14*
z(J^2^DuSlTJ3GT4YE6?*%!wx$Sp(kFl;s0+YKsEha0OVJ+U{Ab4>8b-m0M9A<&m)g
zu_OHVaNo~=Cpei!0I&~Y&KZqrGrh&67v)<Jw((vww?~exD5A{ZtC}${Fyw<0*He?`
z=#V5&-yb2)XE>ut(I-!p?5JwZ`5S7gBFm(kE!=zu@BK;Jh82(6#)JQn`$L7f8}K;U
z8VYk%ELiLmHbvj$r{Jpe=o<58M1aj<3m1F_!uL%|Cz|jFQq$W{{8S(RJGLBY?8g72
z>8hiu?4B+P(%mU3-QB2kr<W9@Q&Q;=DQW5MmhJ|n8|jdc2I+42&gK1mYw;Ir-Fu&N
zo|!#+_RN`6XUxk(%_)Meai}}WF<t4zDTU~3tT!M0?)NkdG9{cWMIt%jO{3whMn~pv
zfRCyhTJrgYUt?9w3wY3>6CxUo#WY8<x?DY%E!4TPi5T*&@T|2Zv#cfFD6t?X9pWLb
zSkQ%NF%Vbg63O}#CVcw-x)83%*R5t8>%VBX2F(v!Ch|*Rvq0Ty$Nk$2sjN)SFR%<R
z&Q-?4aOIHQ8&@7`Q)w`5tfXRl#+X>-kZ$r!eplD1OOhf^bB1?t>Y{wdmfNRK%Q;Xw
ze7-{;b54FSZ$rKQR+qY*G}P~F5x)r)(l=aOd`9D>kAdJ@=ow^UO}uUdX5W{;UOFw3
zKKT8{@tj1JOK8+0^}lC2kHrNwJqCXj#G22{k*I!-!t&Ev@_c+Iad~O}h|lE024#_u
zw4||Rkh%01_FgMHo_5JMSnc2xOg;XK(ELJb5G}B^ZI-;_?rNj3Tu2l-+NnnerW;Dp
z?Br6Gs<0QLs@ER*+t1D(53SM=E}LjR_q=}pq_y4tw;8lAP#MD=q^+7Pec?En-@Yc5
ziH2!vG}Hn`5y~Foo5l}q<DX6)GMR|js?1W2wFr|P6f99&7t~PTd~um0<?_eBwX{8x
zx=gzu)XXNLU2O0^RDGUUiZJ;cAjpsn3l%clbBkV}bR~V1nCv*dGxS)m{x@^~toHH&
z=H~ZDjKggi*}q2%))s9KV#0x~|NgVyH$`NpoguXf)yd%MM*7-3106=|9sFeZt$!LL
zSV7O6wJPmOGuh*76(Y^=1qXx7mWJ|}GG0|iddBD=@EmxIC|7nr11DpXUnAv@e=)o3
zV|s@8%qjd^?LtNTJ1|GVy^Oq$bl)o9k`A;rj}0>YixM_<{SW%M68>|XX<v4j2yff%
zc=%UvmNbZQuLa*gw5R35Ke1~@76yfQBsigC$djrhe(C21M}*%p(1#nlX+4jK!Ubiy
zFxdQEa5n*#wQhTyQ@p5;&YBZv=IjNF@GMJ`0-j}I#dH)=uNB1MK_fhTIFTRo6LvuC
zni`rVg#astnTpl;YgKEWq^q4>qKErOdFM_K`;F~c4~d1`i`9(Ghm%JVjWgE+v;VNh
zq_M69KUv=pvH5nc)gAWy!CQ{Vp|CN<C%o+MgYd@d_FOV~M>HmCdE=UltuxyX)!G!_
z)X?kMR!@&luX3EcQ|kBcP2Z}g&W=TqbP38_-!d(CtBky9@6wSk)mys@80B!<(uY^W
zwwYD@G{*G3770c#PIc*==<Gq){+*JSl~4+=7n3P>=RX-7_~;GQZUvTH?%yojS^oVR
zQS|+*6-Kub9SS0zCocar1S!))!P`YGXCPV=o)XTIUiNt~F2yS#C2@{x;L}<~(-5+O
z!yxa@z_HDGdR#c<Y0gfatl(EQ(cjq#gMP8_cz4+sqRG;a^?-&5u6Ub6iM?$a8sF9G
zeNB7Yvw}|pdD3J~&fSl{+*0_!Q4O}=5UK7;9k-MeED&+^%iKPQ(4oDW7j*A-oeq!w
zs^yFORn>HEHq5dcnbMkD|Mt+*vsb7Pec>xv2}KGQTnd{woapS+{F&-szw~7}h)-6_
z9}%1n@2z|AT**&<xJHXD-|;?hG{#TSXXYV$U;T19U7b><Sa3KLfmtR}s3IN_7G8K^
z|8*mLbSs1P+SIn#tf&;v!nnbG)MM6d*GD=rdaWMYA+4u)$=Bv%=3|u0hg-UI44#1X
zuUiikwbtIje`a0<-GXD}cxf7Avd`xoUT4C|Qx9)Cchxy7kjgyKVxi_IPsRabC$hZX
z@bSe?hg%N?esw1OVy}fVuao^+MI9Rq&M|7aIm@(<owhG(%a~#H@_&E1AeZe?^274r
zpI*0SU8^swS%{u|zSXmsqAbF&-HsPg*+j@t$*#s|+G*HRY%LwHm)!PI(R=5i9w$j%
z?4OoRW5|fLPn;69DeuhLK+yAl5=es9E+53M%_k2Zv8y|IZpKhiNDTUXPkp|<;*reH
z`VG&VDwbB!N~W%TsY2FFLA&5^9&_gN3#!?*(8-&U?sj5rJuXIaRN7P1>H0CVK(705
zebP4TGB&KM%U2#7FU{x1ThDWM7C7?$zZYP?_MGu!t9!^2X$*=%p7#>=`6p@;>t}3@
z*L-s|TXsqh2aYn2Iqi=BehGnT>yN8<wYQ00ogQwP1l7Dh3Yq;K^{VUf83l`s<eAJk
z&l7y0sO-Xf;<Ez6Cn%s*+KFu0w@b*$X2Vcw9g^T9;3*+OoIneNt{{h;OjBe2@4%?|
z+^h|0{Kv1FqwSae&LmB}bpf|-{o~6=1b>!(B@`GScZ@0E-Vam>G6u&=V6)Ss){jJL
zffR5H4`lf<x@D{CLQB0)?I$TV`pwt=b!o^%dJkqg<=1$E;es&xG?{6nUVf%m!ONys
zNso>;B(E}K$|rUf2f@kCtsH_oMi=E4+%Ghkq@=M5pHWi3knQ^yC5vhNC)hJi1kMvY
z`UeM)JbMZ0y&!yMH~(lm!geTR+TbCWTY6Q;_DFTm{172OBkD0=Pw+R_@>F&f!S+L^
z;$vD``IUfFT$U!G+!`hr8B}BRxntG(DQKm=_q1AZdkNz=UQw=VpQ*_?^Q|zbSENl7
zWd|>?S(9`61bHhH2Z7tDTjK}&NM$-ny36R}y3pow=-N7gbQH_=SZjMfWlWX%Dsi-o
zniZ@PxZrO9k<pk4H(9SU(5z1nj1Vm%wzQ|@8*NJ{XIt_UI>IyJGP1HIX)f~?ATi&J
zVEKL#9{6~KAl<P@EqVCYKQ!+ruIDIOyWCbl%%}SIp_tL*IyZTS%R0VJm_=u*1VYgH
zDd6-*|4zt=;^QMy0WzHrwG65zi{~E4Wzy!L&7WHb-h=zU#<nOh&vJYo$|v=2xF$Qk
z>1>bcK7Deoi{QhFFvgebH{IizMuOlH9+yaq?hr^WCSvw$fL=0N7F~djRm_c?4!|Bh
zduh0dLg8?CwIW7QE(Zz3%R!FCo$UAm#9SK8HKGod?szi=YE&(vIQ|~5sXBd`Pg58U
zg93I&8w;A1hRzqUD`nbqgWKxuqdD&@kw}*vcMmzqA_=IXzBs91-~m8D@SrPtI>7<d
zcFM$dymuGM9(5(3HXU|c4*r@W5gdBz&3E_?-fqM4iQRZS<()#iP?!*|`Mjov(lain
zBvR?ai6#@5i^W7-xq75hooor%Pb$L0MGc!P@Aiwbp|(KmujBkxayegs{#rWx&AlX1
zK|J>B8||@SwEOXwfAkrEG>CqIa$vyJSL>xko>y7oxpL2M&7a0pm+SLo<SqfGP&&G#
zie6x;Mw7Ve--;Jh?FI#wg1}H^+4dFt6$iVCi;Uz^Fp~sS4b%srX{RcYv?*Do`}(}i
zxr%Be9*#rbNWZ_z>mrss2b+WC&r)6ojEApOC@TJDro-9nc5zclVj}`R+@piMrbQZW
z5(lQ6f}+bZIEzb!e!*m`OX{%O@m-`?Ab<#=agcg=ABFnn(&s#^XHusL4JP2HV30!u
zcl3Ikkma}k#s7>FW#L!%&Zf5xcIvrly{(G>1=<HsG$5#bH}b6S6&;Y6(V-~{`1-Fe
zcNUxu0ljCM5c;$<?F>O<FYfeXj}eVD?&vc|MO>L!OhVXUnaXsaqZq4L68@T(;i5-d
ztTK`dAVfg$zyrxm1ve<2blyjUzroMmQpVWQOXa2FNWLebQUBFa*cNW|5RHBpXh+K;
zraDDxMgRM>m*grMb4{s}0~=YHN39{zupiP}!1CX7s@|c3=lm9%i_gfo>zW(<0~AK{
zDSfd^Ht_Dp*`xfy5!UJPVe`qYtPij<Z%E~6t)5K_I__IYeeH?*K=8~mvoOrmv=CaE
zn;8bWkJffV$yQ*1et*T}CzRpzsnl{(VuwP_s!Ad%+WNGyM96A5{}y3vAh+5itRzHd
z_4RkS2m2l9`!uJ<nksyM+8KG%mUH=Qi_dG<7~O^G-C9}cazD#f%MM4~8Rl&ZABO?l
zCnv9_4gddaRQ{@~<5bxHOTW%f3F0btdI!2jMtgMp8@%U<3U7gsOrSV3JBbR0fHT{;
z^;=$=2N+p)hR!a8Y|~H%IEM`!*Zd7HP3+fsokg}dl$i$mY_K!LLsN<wo!bM#8HQtl
z;?XQnKZK!8sIkx<4WM0|;SDik$Z6YsP4Bw@>QcXTBDuwLhHV<;&W1zh5*g*n?<r5&
z%c$r1p)10#h3kU5R^t|v-TmgJ4ubSck~1sam+C$<hYznVcHuKpyvS|~%0yekyNZlA
zs+?Hs-``9u)UmJEbz`}(gcHU_jvCoN;it&`HFTLUAPsjq|MD6M{R~-3+G<G<!EgNN
zL;Y(x#8?Wn7n(MO)lQ6tTP7W4JF$PWHQATju4-;>?gD2fQWrZ+h*ctKI^P;Q{Dq5(
zP8Ev+wWV<;RG<w&xH^CNuGlMA>x0sJkRGS*S``{Yj4a|?PBCWfer}H$&aS90O|#dW
z1C4`9)(U)f_o4@^La1`X3M|?944_Skgc#IX$M~9c|5K~rF=azXl)YoO?+7yP&=^gw
z1H|MXyW3@~9Rqg_&dm2R8t%ewi#lk1J|7<*?abGY4YE*EUlk8}X9)4N-r&OG<g&S<
zfW|(mx9KfSOZG+t9__bmG9+ZHuR;5o2L$CnzkeI=ChUKCD_3OsFEC9GMh@9oiB{$r
zpgLCh43U%8GS9($a0$6>{6HDxT-?dLOm7pj;<^a7mcQ>kZ`<z>!^c+CFRBYZzsL~1
ze9ZRC!2`2PbE@EXUo!-QBL%Q^&PI;b-=LM?Y)y+*m8wdtsv;sz-N@IkMeym7I>*P^
zcyiRZ=zG_k-#Q*m`E&)g-#i;%DeA59IdSt`P4$tsCGGj@5I>+*=S(${BxBkApz@!Q
z3(ciwov!&Px-D-p*tWi`uENMM!k=#zs9EC~vJnj0e{ISlXS0%D{2-z$cJHoqvppfp
zY;YXoX>@?l;DyP_@QNfED~<nkKd}qY*6x2rvNvDG_%nI1zx}=7jwNJp2mckV`I*X}
z@@$_)<&~RIx*A=B=Abn4`#012_Xrv(8PY;X+nb;_9~7Wj)71;56YqDQz?3UPnFaJy
zfQhc})p^YRC7<i5TOja{tBlFw-gi`_n@i9YNf*P-UYOaso+He<tsOdzJnTjtWPNRG
zRJgu6yF6Y=BJViQj3LM*eFdtE?u5@Hx>p)dB}8yr?B!FM<E3OsePLd`s(sB)Pc|RV
zO(l!X7_GwjgINt(ZjEp|=%p-kO3KxYrq>q~TXzE8OK(z#g+v=I<Bbz?(?dzIo(&yG
z3l;k%(40`@4@RoZAQoMZqdo<PJxaJSw;Aj-7Z#Xl0dzJ~R0t>WJSc;z$kC2}N$U0y
z-M1&85Fd*}d#-(w8vZ%on)0gyPUpbU2V(*Zy-%W@9aYJe`rfuFD7*?V6;+4mKg4`c
zs`fH!KGr@^sFQ=C%gc`mvASj1Y2cn&P`)ar9tP>jP;<`{1ubBEN>83se&9q)R3P_u
z?PCnZ%?uAvS^eT|xJ?i^WkZTwJ}>A;aEH!k@Nl|ExG-`vkL(hsXd6vXrN=PPf!T-r
zks1Bl+kOoZ!zOT19a&}3ONzXotNaa+%v=fnHx#cr<*NBVX(-{ygRyf(Aa!JV%iu7A
zN#9-QPX8AzuWP#CdVH#V_<`VoGdw_-@B&h(_|e4q5(V>3huiBi7I1?5@)PLjfZH)|
zdI}HK>9}7Nhrhis<X<RHECL-%-4eu^fw!ZDrH!jFy7O4mZ@Ex0!$!s8rIVpbQ@2iU
zp(ARop-=e-o{6;p3$ddKi2_8INbLM<r~5xzH)B2>FF5wKs-lT*@}7#FK)2j@(k(5)
zy$kO2+QZznG7c5G74am2aaogKe#MSv@^zM>M?wFG-g-xK!BwEIVMXJH3vo_#uVNzN
zq6_J42_S$MzMD}HH)`BTTs)CbS##X19qLSpx8B3f%rb%`X!(UVzJ!yRSh|-VoCo?q
zg#^Y!>k`gaQh}+JDwTpiwxt!(74*f8Su@8XFn@^?2r=b=L7P86iJYx~(cr*rdKL1Y
z{So4j>sVE(<}N|Z%{3m~W60nRo|{&pKf7wlNrEde#VgK1e)S?;<vWQlFFbXYE<ey-
zzg_r5?V}iom5OQ{V)nOYP&6NX_LoylLjDLw+y5swjw(FWdTMy6yG>Y3+_W5S-}Y3L
z5JJ#j^?tUe+YMQ1GIh|(w8xdI_&tM)EMHDPh8{AGB@-l>?GM_x$nI8u85ESxC~X(F
zIkV{=-h<AyM;IU8%R~Az*RlGnI$XbKchbKlk`im|E*eC<ywsqMH_L%F#)R710Y<wy
z&_}5`&KUIsn!Tp!jTC2%WHibnj3>ra$fJxYa<_8+Z~4l)Bq>;6Zf}%TlUw(=3RiLy
z;(^dTZ--*ScrdJB@N<OAl1(6|=};iX<!Wj@X8`gH;agsA@dF*=%3aV{IHr~3#rOUF
zm0%4IoE)}{&)WP`6NU4CD@S_vXP)DBn)9LrE$Ptm`6~vdaH+rc5^_~VWa`Ad;Eo9^
zeK5N-sg56ESi&fcW+t$41@><KZZxK0HP%bid>O&jFYuA~9tV@5YW(+1d{iAAv^OTk
z@*5__3q9Th$0dyct9)$|OR0HG<X*k^Ss(l4@`)f3_pK+}1x8Kz4JP7gPIO+D@K5|k
zsSFRHW);TR_&{%@!QphUf%{72dCadI#b`96f1_Sx=Xzt`aHFufZ=?0(lPihK#V}t?
zM%;}EB+9{}$HH@I3FJ8Zu)+J{y;2x-GW)gRDH^A>K3O=CthlV9VOu%qUoWYl$&x_)
z1}j+t7wiuyY;Y3D;f!!+{{A+KgpalQ%TL0=(UONT>cD2{*1vxdt`ts9J>REElSU9g
zxO%XI00j8SaYac0`41FKZA47+@i{Fb%rC5V`Dcp}<JL2AWf(AFMhwXi1=_w3Fp6Ve
zb!I=AaY26W2ay@(MtDAh8{V?Jc^}f{7h?Tit=GP4MEu<li?kZ$C}y0l4)I!y2A!d=
zo&{qgGJK2usnvTT!!?7LB!ed}P6(IAMVui^M-&_X#&UKnSgEP(>(h6e6Fkl=owjf}
zX6iZ~I!5rugRxy6axWWQiI_@X1wWq-#M9oOuN6l^oNSP>#gz&3Exn)^Wv~_5_iq1q
zzD7EKjB<t?{yActpV1mI5A;fIcdZdIf(9W%19Ki@NiGQKzXTKayCni!Yok3^k6<td
zHD{oLy--G6RwP<tVD!^}`Pbx4fT4;GKHJTkgC9+MZTrefH=@4A%TC^5?`IxxVXDKw
zdlOajk1w?6gWk4(QC&LQ8-WY_0OU#jfibjQat<F(<jOXEQZHulx5~^cwW=D)s#{|)
zZ%)CSGZ5}}ub=c?2+^Rg?XN8TH@P&re>hG^%80-LQWJPOFm?7XGjV_BI6b8-tS|cC
z=oJz3$;*+kw4tfz$x$XMI#diC)9i6ldhF_*{4(MJ4}o)BJ%aeRuZoilb%msmROB$p
zG_C515Y~Gt1;@cz0+JNKlR_W(pttN?C(T^m-!r$Lm$<aoUrw`zn{~m!u4&W9455aG
z&)F#0#Bgd?1gFWRzGe(&Laz+{?${{vg91~Au};12mkKXj?+5I7L-@d9t|8H!C@Ger
zOOIPy^Lv)jXfdVYy!a;FMfTEZndKQVwM&uqjnsdFmu{l%Ti4zJm~Rn}X)%p*(R}xd
z)^Jjm<)wXI9f~}134gj$6+HOfxPQENB6Rh8%qQ{}_2G&)J}m8vh5A^okI$Q{<7L;9
zCPNV>k~o+NLc;ZXD0ERZO5Ty$dh)DI)sHuy{KM!;IlvzlUmo&GOU3C7Gi03*q4dXG
z9e(Y-hC*rkrx2ZVI)X-(us|Q{pu^~oks+SqZm>kG8b5iVo=GB8i7G!qU&mcUS=x<&
zX+8tjuo^l-ZGm&T^o-nr?ueUN`<rLUyVl!Ex}&*hxDZ3T$}gsg+9Q!xJ>llN5eEv0
z;HT_(8>hHgkJjS^I67R9_}iNv#Ez}uH4ru4a5=Mo6t+vN*GrycCU{k`O?1?!^Q+`s
zxKYTsI_5(SRehjRHV2z!2)>s|Yr;wY<sG8Wzl3&@MFm@T<uRH<u}2(XXOrMJW@SQ=
zU<WCeW@-N7y<MmqctK3CqtiJF_rqD$=W>IyII%8$$igwFf&b$%>#7%x#S41YlyIS@
z3;jCTHL9X_SA<EEV1&tlqa}YKDOvG$9{y=nV;IU`Y8tl3UxQj`tWtWewX6)_7yMnT
z2TQ1=x)&thqS~Bu4oxk-d1IZ8kb~Petsz<==Ayy+-N0+!rcSK4a`b_A?t*vX6Oz|s
z^Vn`=W_e&0M@L*zjtnZvB5vyqVa8Rp*kDfkIqZ6KhEMms|B7t<!+>&hns=}D1rc=H
zR?o$s>r-}p4COEiGisC}nu<`}ACrrkC+M8OTp4dLb=hR6$F`;~G0*60aGrdF<GH?^
zZ<jYN;8tk6JD)++mLeYmKWr3dpF|i~)Z^AyN+rvaav!^HTKq0drzF<F@OX=snYguf
zBnRws=zJ?BE3J`KzZC)a<+lquVWB}I+S$JerD~g}yQYG=k(c9a2MSNl)CY5qNi`zQ
zFF1q5ZZjrC?<qO|&Z5xW!8SBpzfAX>S<cfEdHXHYiImJlxF(v*^7&SDuR!&={H(c{
z!sEi>_YoJ44eU9;L$lrDHl!4j%L7;}u&rGpub!1-uGVezQ=zh`YcXQaV-mhY|A-q5
zaV4Kis@>B>uK~TWXY%q#zOCBL^jVpiL#U$8SJ!#%eV2O=YnrVGG`eeZ2yK3AjK-6$
z!~|N0xi;q$l1|ZGznl7wtki20{^f@5NBmg}MB%VkaWhl=+5L?wT3cL<hX(TAqZDd8
zU8S+Xc`~>GW9yKZ#O}`ZzSR}SY#u4=I`=?B!=!0n=PlIc$0Oa~iA9ohr`>TouB(kh
zaC-a7oK(O&S=}!Tr!#eu`6amYVc0N@!{A=3L(?0!NEOUaU@(Sm>8hgg)63j74}4==
zq2rL7;b^F#j*E}LPfpnQB{EcsviFuK(hhfBDu$t(8ICoQOi(JixZC4V`?1-r!adqW
z$;WQ&^E&(#7V)1I1L5OKho)R(9&p8isS#0ZV8r`R3F&na;oco)Y;KaR0>yYtX|bbw
zL)WVt0!dS!uq6%AD5cE8&CTbtRXjH!{W_Y&B}mYxn#L~GW>NS2G9X>^Rp|!>tDLwb
zZt)`Ouu(zOSj3sZ105fli-8J6^J;$iwtTMnFvIE2s=;GFS*lt`E|CdX$~VnutgqnB
z<w;e~DGJoa;urx>jUtq^L~DdxVJ33><}8K<#XG&!XKl1gbh7zgsh^9BRD;!1N~ap-
z9d)9PaUXei65Os-I;g2<_i69F=bf$(y4T%4DTxUwSw4NM%-73&7Ulhs>#J^#ptw%@
zZn>z1qc%KW*V;ck5@8TRG7jXh=>|-#JBcjYSWXnE@wP}rSwQ3lrPz$W>ws(K#*(`%
znu^SFVydPF`^w{e=UeV>3sEv!N}bp~`h>Y|V$7v;yIjA2^0@G~kRxQugFZ99JhV{v
z+R?*s`b#lc&h`$;<viJ!>yv7cw}C{Fr3ttYr;7?&o)BX)SNjXx>TA3ntDj28gd--O
zVqyzs9=uj7>z$Ujan4HU%0&D{&ZpRg`zsNLmr^H~UYt&!0w5I135gpcm@TPCl1yOq
zN^Kl;wW?qBjZ4+=TfU|`7~q04<>#hj*u#!2|4hgiVTqGi`>Z>=ZSj8a1w@?2(R7o4
zJmy^s6ptDj^E$x7W>XZwfPdS3OFfz{nJ8lSU`m+eZsdhHly)cSt<m&b%oyoqCxQb#
z=ql|bu)!j2<@c#ir(rA7r3Ww``7gcJ3sle7=dD}%=iKhtuwGuVsT8er6dJk+Z>}gm
z*xe-+_`pw$l(p8?zG!d1V2O3;`BVRh#@Jvj=5);&G<-SHwkDQ|Mx2B9jzop><Y?kC
z)$VFt*~nZ|zeV8VtjE!)!$X>wjQ1u3JoydQT^@muf*s68rI_UT;RBhJJO<ZC*l+{x
zZ-b@z2}DNzr#Z>8Ylllg^iTui^WW&Gm{Q~yle=Cx2-~_KleFp!@0g4@#R~guVaz`!
ze-uP;uzlBm#u*pG5e#)@+eo;Ux8HP=dDBmG0z2nb8^2kZr{1+cGR_4=X6s+*tTFM5
z9s8`ZKCo`JpJs5r6k0le^ky5p&2Z~>!{OT!A@Df}oT-_5$<%({xhHaYt9SJ?Qk^DQ
zdu{e^s=XU1@#jqX<bLMY{RSWGH)q&~6G;?<+_$_F4#ch@L-qH#8(s#Z;cX9kVGR_o
zh`|yN9jtX2AcOu&*%kCstQ7`Ao*MjH&$`vfmC&73LU{L=$F48m23_aoT|^DFcy_&d
z`iMx;FDCznqWi|R(`KjirQ!y5tHSx8^L7uIU=>Rz{LS|Q@je>}Zb$WxBWRtAe`@dc
z%A-8DrH?tP9*#5G3>#4NP?8n7+BU9Z?X+$*ePFrVHyDWtsH7OeqCqDJbel@gqnvVx
z16G5>qc79`oxYRyX!+dwH<<eUfo?xTiJ6wxOMMm@WuJYrw8y=pk2cE0qU~HcjweUj
zP%EmZWpWL`<kPtAtG>lE#~_JE8Kh9I^cbx;yY{J+(J(MA8}qoB+3`X;SQ{&N;2`%t
zYu38&%xEaxZ6lr<+P?dx%;~<o+<b5-9hVbhB^WGxiT#Aa`$Ia;<`JA@B`ySt`r_Xt
zKA6utIxl2@geP|1V(AxO{`UAc@<aZga`yvEbClcJ+nr6MEoro*Dd841iqT9ck#gXQ
zHUX>KKFKvSp9Xho_$|PE7rTFC@xCHvrKxN)+lYCzP;I}I|FpPTWYBsrTFnbW0%Rbz
zwv1Z~=jj>9ARyFgvl{5J<<!1q>dr+|uo{Jk!G6&3u!?iKQuN6vv=zAzkvmOtxo2lu
zWTguv!kSGJK>u-J+eQXOAkU<jFPexB;z70MY98?+1az#Me;Yj<**cGt*mkbZOj7cb
zIjo=mnTQ{ph^M}fUAx!#x9|kV@1{gCSF>9tYS2K8?Hu56O7D4roujG~WhF^MH6s0Q
z!YBFO?$+|H_nLc)BJVoR;Vr`n!t$WIc_pLkccPwJ<8hfJc<2)>q{dnRB>ajJHUFYy
zY~~|E7;ih>Wlr1V%)5|%xvcOU1yt+2jVlWCyx4LH=m%N^(ZZ}o)tjBnKK=~&0wK<!
z_Ha^e5&VMuMeV4gWAVoKvhv=mUGXX>Bj=$bXs1bB!7jYN(uLD|{SVPakjZ8@4}a0S
zns>QqE44l`U=#hXswgc3F<0?wbQJX7ws_Zm`{Oz&(wMP7?BaOWQSQ?}qFF97U`!$s
z7Rbqg>f|?7<){fYt8&(!zBgpEX+m{$zRY8QgNh81j22$dm-08Q1N&FA?H`hl7`@SR
z+cnsrmR}w@_8$4eCq#oBZ)eX!{!<P|>uE8_3}Y1jt}jY)O|w;*lS^AO-H%|qQf!z!
z18BtaHZmG)4r*M!qF{K_ou9z^ba1t9AB~AACs`<=azf;pR?42d<4;_Y@wBTa;*g%n
z5}efM4;ljVFmV`q>D+gl1TL~AX5U|JC%mYcn66Hxv*>0?DG0Y#P9hd8YP=4G`tNZ*
z9hA`oLIwbQRf$-B)6_f(tm$2udAzI_J^II-cdO`=a!`MV86xIwSMzNZqiwZ4Pko)V
z<uJ%@F>#%jYkq7%_o`mHEnGe(kYZyG`kf94RQ7K3oGzhpjo}!n0)<DEYPq4<XpJ0U
zR=uYPPU(wj9ImoR^yJh8bZ9a2sx-+>;hwf0BX2egDxIvxUClmzD5^kpLIQ3JzRO81
z`wcKrraaS}6t$q_>YTtSsw^3v>UvK)Z|9#UcK7TdB$IxjQ6mW6Uru%(1Y_op1I$@Z
z{`Ni+d~=D<cTYR{RenV|uYPFCQyUg^`1*eCF|O9&T9;mP@o&>_(W_?*0m`ZhY)eO}
z2*t`rBnG!)6E*RnG4T`f5#F>0eozx>5UJ*6FB4{!a4eRsF)B`}6!}QR7cq^18<Hn%
zx&U9MK=N>OjkP;xM9OU|{0@g*YD|%bpqItG7tb7hD|OqHl*G(LilYp6g4h02g~fBG
z=J2OoVv4o_?k~gCy_1JpiuWfb{(iv00sEUom2_Ls@v%$wd6AFI@trhc44Hy?!1{xW
zs`sKn=Sz?0Ut~$H_oyE!GyeIo9D1Jg^1JU)FIb7#X#_U8gx#+h^pdAN3<}S&awT#D
zE4YG@ykL#+zGXYTQhy}4-xBkG`JN(QWPHLv@p5r_us5`Py*{I@3up9ExPT4xqit(v
zS1BZrZ;r$IL=0khlB8wM3KceT+ggx*Z{^J>fSO#}9S`g8K$(52wI0Bf9Nfk;@y@u!
z&9JeH`)XEx{#uYU3(r{|RdILF%YnVuz>3Syu%taVD*617X?UNx88sDsZrf25vC&0g
zT=bnG1^SJ)ic%<Vzop29Z9e3M#dTB}2rv@BNSfhx8Y@@aU68k%n!;HVeD=&RPqSOP
zMJ<PvyphB(xKH3PIQ7yn?-_C_NNpF+i6P<+&8Jl^LeJp8;nQw=6kF30+ZsFn+B+iz
zqmVvlF7d08?!ow??WP>-y`<fttQudzfh|@XZ9Q^nrWLHy*7x!^PmokbV6{1-tGw~N
zG;-+APbt1E1q0?+dD&73;cA&PW;Emf-wO~+RUfiIj9yxxz$Nv0@K;OLMN}RYWp!Yh
z<>1N!#}6_Cr<%^nH?&MHPj;+JGY8-<S}i26<Dn7~iOMMs_fq7P3wQX}M>2KK_1^2V
zd!OLWACgR+mA+Z$Sb6w=@YocI)924^KH-#%lkS?Q&Xo&@%2ev$_#Qu?V|ZQk^4~8!
z`XLS1I*q#1xg9V)(ki1pkk!B!=b<(}?WsYM#ef7LEODn_o_Bv<Vg0%AqN+<JA3o>r
zU17cgs;E@{=fMq!K``(op3PZXs4@0r(zd(sLN8q~0z<!Xn7&J9jo?x7Y}K#@`gQ(L
zuF~9NB}u0){OPE<-b<lGO8E2`<J4wrzs0olr%w0Mkdx&&r#0EV4Tttu864O@)q%3R
zg{jg^edrk*DGx`tJ2QXE`CSCOBcy$H5mwp`H#<BYggBXEsGhdHt;+&{khfU_yKP@s
zwoDnh@)agXcd1*yYFG;*DJ&S(mQ9DWHX45?X_`ClWxxeF1%Wi$7CMs>8W(R95=<$E
zUUJReFxe?DCN&K{e3EH#!nk4{64btq&1iQ3#+Mc8{gt6J&TI3O?rm(X*2<era?Ahx
z^PfyM9v!>@M%jAH=bX%n<ax>I0sm>~bdf?gXO|8a&pQNL!h<%7=aHhP)xzEf1o|A5
z(D3W;$|LQLL<i3x^()UI^wbv-WaqW%tw}i<bg@O<R+4P;6t6d^-t8V!jt|?C_7USQ
zPCBXxpfe&c_oy4p7GHH7q_Q1M?G5XqTlF?}1{OFoiTaq1jMR9ci9LG?4oiK=AgzAC
z=1EgaJdQpoRPVK5I7F|exF)jI`;tJo6WqgPkc#?Qt+ND8L1z@gwt+FCJbTx{R>djc
zH0ac>(r>%(;uCvE(ncWjNO<_4!T~87^Bk^x33$-Orm-`v?~jkFR>#$O<Ku&<OVZQL
z4EGwU??T*g66>QT7h1|}NvPbiRV)|@5V6wdQcmyQ_)Xip3rM!-!gdrorixg6du@hy
zuaNHocB3mLirE0??E+8H6b<|ZWXWB+S0qWK<IU7~PhZP<*pDyn+CqFUX6_Zk?DrPR
z7cY3uJR>iL4wuKgpoHQr%o*3<Eji&y7bK{Q`Rk&s6&NeLK(F_$<g+?&6N9J(pgnPA
zrsM8LEe=0Pd}Y@daKxjma6=Xw*9L}gf^D43F<hSAG@6!L1Egx;kAKQJx>DkvY?hWQ
zMBmEortJ_8A16WcbaGtCvRV4*Sx;PVK=*VPEpvpkD7!W6trLQrs+RZ~HRW3$_lt;w
zQ3JP4rR8X->BGVF-Vhy%2>kt+LN?uml+qe&7*(M`18x*2J>1$G)n!#Ap$*UZ93s0f
zP%YLQv@0#iA$4IEF~t0Q)i#>}yZhIIgO(6va6SJL+_lWnQRqDR;4G#4WI{PnBErC}
ztXfk(D7VHazkGCFa=swIAGEdOytJ@Mz!1IgJFCrdjG|KY$Fz4CD{{9(KkcbDaFePv
z9)6`qIP*!(55u?caF0n*n_-j!Y~(kwwjfR98a03x{3WDm66!jNlgY{k#9S^f#cAZJ
z<RoiA1?%d*8t0`xQ{SFZ?O_QLE~I)~kzEeR>hb{_<GFd~$2!Kdgw$)a?A8Przrw$A
z@|Zy07ayk(C9Oi=l^yTN9|J@O%~IU>yoHn9$a_r%V%M1sgNBve{BGLzvv-eQJny|D
zCVjUs7^Q0IxIzeVg<|5B{t>-PZw$3@W*N-e=qSueAYvWWxqy#kC-pk6I~;!5T0IUG
zSo|^B;)uKb28kL~foIUx7NAEBt_4C0N@C-sAwpfw=s(@*R{Tc88|-K8&3s0R4BwdX
z?2?H`5p#zubV%_T<|#A?TLw!Qg=w(bJ1lVi5UHMCqqcYvz*F<f$uTi!a(MR^8T`m<
zxI9$lc+XQNlBO}i;wXC8<<seuW4AI^-QIVzWG*A#9rr1wfbW4jL-?A^2EC~SFR<3J
z((&@TGx<KrcwQnK82~sNSL)0BRQ-{;e|ZfsAou5>A6a@VU2wTDSf2KpEA2(xjjZHz
zX~3-18B-i)A<2L)?`?~o7563E7OAB=ZTQwgVLnobcYP<gp11B}l0HkMP#?>on8<og
zLJ#SXNbTjBV>kyg9<_X)X={wVRbtkc1uM>+xpHBl>iB)xPzacj3h~7N(&UyDnxv7X
z7tBzp?Grn-dksyU4}OHM1u{AH3e^ejjgXlO*=;2V*%>atsr(hLp@%5Q8F2%sW^fyH
z4_5nYoU8dIB3ani(2!rGj|d-#wL~+%Q=C~R5zgm}d+87GkdeE8EQhj|o$IhZk>+*g
z3MpSuo)z<?X;kablJ4ixV4pp0EIMv*zRa1?)W<MVax*^9_eoYu4!OMr_2wn(@y33*
zP^PQBlUsrd@AdiBftn%FX1uaN`YP2rgad$C$1C_`Ol?NAcO(t`mPDKQIi{r4&@D@n
z`3Z7NYhrRjo|oiQi^5<HoYwT}oedY0qG~jO9XhFzp(ia4s+si{zKAWKOLJ3*BG5TE
zHS0yDS=oPxKp5f&ySt>LjN*(ANgki|m)D$S+=r7aYeYL;4T}vVjr>U@oj&6f&&K<R
zm?~m3mr*~I{1cXms3=M_7HX3XUVsUB{j>;};v<({hg2_VbDMd1meR;`^M#lsILh<!
z)L;RBN6<X67DM3xMt4eU-;xleJNy6~-y<F+Y4gc%Xirjr-^_4(WA#k49_<-+9}GRK
zn36}NtPpYfY4rRPyNjopM^6B3AGY7~h)Z?I09-L1UAgT$(wX!D$+uxzIon#|NvyQo
zlu}BitT=63<PKN;K9U!eK5w0m*XRx2*oE$~9GYM5FZXgegkw?_1kHV}{Rb)M<r3ig
zA%iAdp|%h+stC0~gLEn^qZUK0TSsIHsTv5MSF2iujt+d1B-3nKV{xquH6^q|u{goL
zCIcrp%zCs2jd>7L3VhO4JqnsH&d>wClt9NK`cthE(b%wZ<EZ)%*GPnpvstLUb54v(
zP+!Xl>2BUayN10Z!&W~Nm!}ptCzz7>3$1L!xa(i{bl(QmtF}wFzf1Q7DiS3jnP@RR
z)zu=#qPF~vvF%2jd9g;-2`3BQ-3JSmQ}BYpC98PgHFWLE)&MDMUb2o3?Q~k52krAa
zb*)YcgUQ=SEaLXUWe*|KEM9k{;-yt%lO@;Z-?hd0-ZA6FOIS13TeNI?Hr(22Ko?)m
z^yC`J3s$tHKxedHc64A_ctQB?Wiij7+uNw03(|Er43}up#f#anodw=Vw~ZNF#dTug
z7d)R2>0@q}m2Y{LgZ`Di7t#!4w|p+`9(ZC!Q5;Lf@l)Pt!C@bW--d>EZEEm@`0);F
zMDHS$qP=&-2%j9ElA#vy<E0skCTmy)^1tb^X&%Y!9n-*YB{_fVl|2;W=pF>A&zX)d
zsc?(3GKjHrWMiomaR&olNqzeQ2{i@=W=X=ID5OKKJ%BbC&{_s5hsLI*+4cH1Ky8jW
zjAE%OlY6Aq$O)L_JF?C6)?f=Dj?RFRqo)#@EVQSp-5@O*+<C{3_^dab;suSWdG7i2
z`rMc|-m!=UH-Q%7y%Al`9H41CHY2(ggqxS2D3B<Bi;`gKsm(e5t!UTE$TyX&y=#Ab
z;K>U03RhzwT&Zppd1gAY>@LPmz3KUc4+xNaOnhr&P73i--JD@g3?{W+GPhppWY~|_
z`>=8hRmDOHfk{~@R;)_Nk`yKQ7`f7MYpV?_nT|vAeFqC;vOGW|kz%TIXXr(k`xqAr
zU13)OA2bzoQCy>5oK(~4p7TQ{Cm_pbBH&M~#Z5A_+uR{wFvZqp#koQH<h>%6_gH0r
znWw;Ic1rL_oqiko7~T2+dxhBRj?{a@z|nE-Fx>6XxwODx4NEDdmQ(cS=h8xy1-)G*
zrwf*XUr027O%NU~x(Ee4w46*2T<!B{X2X~!0#Ac@#%f`>SpkIunr)@Glo`(`*<>o&
z6p)*4G*f<bBcAfwo7EfT)Fb+$6gLW6&TZ2UrTlDk=p30;)#}6Pj#uO#m)B;&VO7*2
zCcqa#gP&`?{{X-ZxRdM?x*3J(CDk>+JS%+rf!kqkCfzT|ycS2mh`D$%wuj~IJJCzK
z?Tq{1#p$}YAxn05?~@O23QBO13gYn>b4J6h09&0)ssUlLR_QAMQ2~!xdo8#0rE+kO
zZc*d0+>*oEPJW=0Ra<UG(39CTn6XwFxEY0!?z}QFm1X1F8|6k=lYGB0rMBW}@?BdG
z5rkjXFp@F_T_x0$-G9dy;F@jrR56>`Z7>jEu+UMs*}y@|ja69(q!=#W3AGMsNOuZM
z%*JB=f)FYEJ}hWJqK6y$_DGVRNSK-3fq;!pkK*64^VA>V;B9YMcGhXzkSm1I!bw`;
z^LPv@Ki+-J@Hrnptc=~OtaQdM#RY<w!*~^!rGRK(SqhD*l*6N1Pob6DL&Zj@2Tx>K
zYgPAIq5fpK?qpd^|99xBY)4DbhID=QpE;QE^aj1y%`+5x57$xzso|&+`K$-(GovCI
z?VQH7F`Pfh?uXiJ?y|KHX~91)Sxigfkjycx!HVYz`vA8jgikWeNr~1MnYQY7jIw99
zn@f2W3rjQsMG!laCj<$Zmv%N|q5P?1$nWzzTBnshA(JO<a(ob%6@G+l$jzi$UTkJv
zQ`AFxrk~9R-)kKg2Gl_BwFl+s8V#kb@^CRqo%su;0Eats?6fkh)LDD`x8h{z{a*_o
z)oLZCA{{y?oBOjSP64wcABe#UClzO|XMpA!-eJ)VEV$)-PeB)KKQZ|nPl$eeg$^+U
zIkZU-eTOD=9Q7TEt4S#hPavQwtuYm9w6Vc$(|&*R1cNs#95wXil%B%Tr~g*r=LX&;
zD(B#v4FNt&L=H6_MHsV@O0mdTw+@|AHNtN8div#svCkI;PVr?Ak6?f2f6BR?-ZU+0
zI6XT4^4>rct>F6TKxtH)okGm(AwY*fbguwu0{hfZsZ`*!{}@A)wS2hT+3a?&A{VtK
zARE1~tJ_FNJi(sGKupdsD0#oQ?=W0wn=}*porfdz<lP#tk-B(6jm;|t>$KR$Q;^UJ
za;&>`NL_Z%H{SPrj}tI0=nb)*OZ9_~A58I?{)mV>q7qMbb3dYJaH*pv7iCH`J;1He
zYC`E-3u%)si}s;rbH8qn%68p25C@>;>uC*$QAkhcNC}W`C@RKekFYkwrJpUa4N&rP
z7r)oq$$f1yFQTMFzHsq5eIkYQG-{ImZCDo~pH_=Un!Xia9mn2sM)<WyCE9Q;g8|)E
z+Ko^3B_XapoEx`}q*eh|jlFliOfi8SZ=+Zsv-*N1*FzVv`ev?1IYE(8Kh8}cF>~GM
z1%p96kT+~TmChi^g7}SvR;eechB5jK-WL6lvBGn6TuX<2LfN(N(2%QnML3rtfNni{
z{3hZZyOr>1&+nc-DLTV)N4+FBv+%8JKY<C{g`Szyh(+baX&&cl7EyEM<G!KFDfzw<
zT8@wbw-8-!oT8S34DE<BHTNntKoh&jl9dWjgT!Ix>r91TF0To>dueHFXZ>K@JiTsh
zi}T1nex|9p7isQk({2U^tW;`be>E4)z%l+{$0_Y5osSzdK_n&qzLqrEzSP^cOe6uF
z*l`13TvFZ3Wp|C@#4tP#i7>lal2HZ?ac95$N-Ta4m1;GePeSJ0Rt9UzYdsJMq;NNR
z4kT}O8|hp<U?kaW&}OPJH}T@rqv-3$`%r6ImTK$k!{a|nZJr8f$F~lsf<g<^Q#_Dd
zgJu2l)fy+>(3WVEt!-0c$knm^*R21r?T90>>qRjT(FlA3^;=JW%v}Aw_wl=Dw0F~~
zo|4~&RyOBSgd&z`{(kh#y#V*IZs%5IOF~`Shm%!eB$C+K3@t{`-a*gFd%=Diw=|O3
zG_XlW_{>)PLu(31fVVQ8OV*L(B|8pE053P<jtw^RIY_GaotCaEU`a`ftsTiG^&1Nr
zw(ZEhNG0d7`5d$R(&m;Bq^5l(Mx`;FL#6{z=WZG6xAOo=NR)-7(HGirJ(jE^N5yZr
zQq@bN*CEs*km{f8;ddh*K4C6CN4NBg!<w9T|E3LAPNkQ6qk}A~E%d7Nv>+k=s;w3}
zRr9j0JMXN?8~AOtnACZji?~&vDg+rayVM{tY2>#g<ntY8l@mSZR5ZhDt`4+B8)PW?
zgep~FA_a2NoPit-lScNLw{zfb+ERV)_bIjXd0M+Ppg2DBqINgUe_lHF>)-@_4r%a|
zD|wOL2L?Wy3pFKlyQPHJPl!nRgeAz<+?_zCF*Bd%WM<BL2QCHV;B<=eYc#G+sa#N$
zEv4kCSdvwhUg~)~I(qz7pK*)SW_CvT33X<~gF$SBjqF}jwEE)w?9b03*Ao4bt|>VO
zcY)6CzC4HFr27zCtKZ001cji%BF5h4dlh70CIbrj4<;iKR-lbT&tKgSm4HgqECR^G
znw!EtDh?_Fu|aEoVM9`14Z^HTRQ(ZmTVp2~L6ZfbE&GYJ>EUG?>+HtW#&iE#ISWOW
zlmT=ePd7MoiyEm(CRsp`2hG^x+MYvWSc*RlDag{Prb5n@-e!xWNjRwpT$p2(Qb_cb
z8s)_NX21p_Mbgxlf{GAf4!wT9YmZ%@6zPLoHUU+U&||!WVf@io(r*}eOU<SwZY*g$
z`uVgCc_K^j&(t~(S`3%qo~vvkr72WT4QjCreSU8V)e9*PCQHyj$Qsi7yG|j_@S713
z57uxJv$5~5lC@ES;(FC4q=pyV0t0$)zd$}rtv<V~kV(`?Pt)#XuNSU^9L1lX1U6^t
zSLkWQqksx(I0|8DUupDvxu@lXJ%wbW?<7DnnS@h1a|(<zo`%yy)yiW~R}`x!<qdgb
zON5VaP`@`yro$FtM?BbF*BP+xkR1#|BbTMRH>BepoHc}6)YcYnYlT~oz!`CYW2Ona
zG)XU5^d?^j{D%LP8sSs~;KL^P1o(NG7iKI7amMrfyz26FQ|me$#a1_x9ftcy;ag74
zXdr)2O0+h#`GKdMEmN}zs>{+dA-aLhY;i=AX1NX}U^kA<p877z`by1OV*u^;n20ac
zmm61?zz?6%A#Vg$zn@@p!5PxUUbK>V0aBJ=Z>b6R3u_ERtlv609E{_EfCjJ20bL#d
ztC?ldSZ)Bf+LMPQ=<iZb%o~&*V>oRIh*<&RCR=LgGX0&7Xqs_h%2B-lfgz(z#N)HP
z?Q_x1Fh*$!j*lX>D5NC0sf4-fkREpES{p>3%s<%}9A1`Y{UI6NG-*)+Pmp;G66lpH
z_N{d_xBbKc$$!%^Kt1Kil^P5DvKb&7K^5w|VC7hhL^<$7?<qmp+$`Pd$TgjJEm7@4
zoS4q3?Ff3k=$_unSrEhE`HfqY-Aao)kp-<FUJW8Azab3MI%IG3-6_{xn=#ts?f}Xq
zk2fTqbd4N5K=>!*(SSzyQG``uZ|cYyxQ%1p-@JP<qcAuMaK}4;^cD!^3oH>la#4T$
z38_DZ6&4o0Z<u3&I`mu5{#wy0uG26p<uKnhzpoDa0u{+h*)n09d$TyWfHB3twiUKC
zaIi^SF<vK=b_W`@6@dyevh}v9&cCfE3LWvpYKw6z#uNcL)DZoB^V*2Jag4s339@Vk
ziP05@!7km&Pu4fHNbT<Z0!f~vzbKBMl^WMD4!1GHg_^0gIVg*|M8qv{I{9kvmZMqC
zV<y!`bP=;E6C0u0;D$!4K`bFcNURcSIa;jv6O{yz$Ab}4?d%=e|7j5y&_{4pWQOEh
zglHDg@eQ%GDu{<MZw2}dOvs{1{`fuc*8hHh;|DCAJCU{N*C?bX5~(j96lbg^{ZemO
zK+JYFNzf?l3ML^RE%P-KZd9rMK&L#Bj&%M<Kyw3Y$JJ<(30IQ*LfG_*W|L|sS-&$5
zt64Vs2&}Ed>Af_8c6qG6dwPiX+N}6_B<@M3(^ByvlHtuv6Nt=KXHA&=9DNOT0_F^S
zG8ouTHV|(`j>y5;$0B8oJz=jUkUn49AZ!b$d+A21(hn_07Sn7%PqP82EA14JH5F}K
zj)Y&Gjf6QNutdq&@|0qD8nQV$$2vOboi4?Aaatdo>*7QAayJp`@-uj)xRG|B1bk+X
zCiMfRqb4mV)PzPUW=cLgzSJ9V;22j#YgM2Xf!eg|eoKW1z0`JoGUz1vqdu(Eo`nQb
z<i#N)jocdISfeF)bn9Msx=yR^Ue+*I$NJUl`;gms<eTo$wInwZSF`f$<0#?EmMw`x
z{nmF+7DWUlGRI9^ecX0p<F8;4YIFWfbQQ{QgWH-M4U=kpw+-_uP&uy|pWM>AKmj#r
zYUt1FB&#q8kX)u^8A`JxRi(QFWQWxZRrI5DyRDLb%kzjbidJ(08Es?i`~>e+Nr=R0
zpVdf#Gh_OE+Uak(5qG*q=|c$AqRU;PLl7zT2^a}x)49XUd8C*ia-F(-f_jOQKxtdl
zh|-0{a?WQ-ngTH|Rc9ay`>AWim4$^Meq*6Z<RgESnFTJTlqc$$qj)-@Gx4yjZiNn^
zJFN<BvSHMB=q@O^K4ik<&7}#(eud&-mIA(EIY;s^NMMQtgnqmN4S(Q0C_Z^bf#<i>
zxfBOWy=pIZQ^E4zI;uM2Y!jqA`KVl@HQ_a)AL5LPGph7xNXbo1q{iPr8Ae?^GH4ab
zLXX<Qu!0QY7UMxIH$_!$W!jkq#&_*YGJZo&t$PpNv20<GfV7~gXw#@fl^#21n+C8Q
zEfy9z2KN9{yJfD@P7%-grx)D%RQn&j!gRB8>50>+<g-Q^S@ZITsm?ca$)Ej4mrA;6
z@uD<Rt8_aznKP=X8Tz3H7bSE~T>Cl33v&SObEP`yOrxOJr}yRsT|DxqPc@p64wNVB
z7h#sWZx{rNOk1skU#~sSnIF4(4L2=>L@9-vY0~^8soTDR;*^M!!??XS>1+{I(G;nI
z?xtesbH`#;<aCj5n0UPip>Z5iep5H^Mp>ZH=D<f(%HJAWt*3RtG}?kz;;haZfxnTj
z#6mzLHKwoVVjL#is9S09g4csVt{zAkVfw;d5ggwwy5kzDCBQC>ElkMb`RTp1dGgB5
zrv^+MR*5<~mW+0_2_{EYWhRZILP8`g^a7snAg1d0$aM7*7j^i?ybSAzGtRmuBv5cF
zimG_hS1(5>>QPU+YyM@$r50RbvnrB}9t>}9;VXXNwiYNMML3P0w1_~%83|`(jUaJ0
zXJLFx6Rjngjk`9le)~oGxGp?8rgQIQyZqCeO@Fcpx_rgI3x$SDu3w#P&Ye8fC+hyf
z0)Xc7;9^zM@%$!?OBM84URcHRt|PY8d%%zBg|f{jtwY+r9!cnzELcl1z0?}nF^d?$
z4YTfTxd8l)RsA34Tnb2fNn1f$s#Dj?(m3^rQ+$`2aF(ooPk|(TcF*2Dk)zM##gmd`
ztMv3r#{B!@EC+MXQ?>S^ECLXW`~jtKy5-@j-{$`qo2`I~4C-@p<;0-V=V~&KD5F()
zV^;~y)uvV4wmj0!IIPA{b(fC=x%d_Q6YeaU)~t517fc*-{UELWVaOXm_`z<P^CQ)w
zM&+F1T#B3CmubnKS_E8Befe2rIHG2_V;+b}t@H0c8HV=1w{$wKdi(8N^7<=Nd^6Sx
zvCA-$NFMef?2!oKk-cFKxv{I)BzP0^PAYDIW)RS#9SV*eg=3)hm-YnpT{{edjO;e9
zQ@U^8i0b^W;~Rs2v(WT+l~q36tTtwxhCKx7l$YJb*KeSDP8nqKzu8bz7SS~~BV^a?
zpXqTA?fx{jei2z3N2(FZG3rRE&iAz4iDB6On3ocu@lsl^_#2kKRxIa7VFah3v5+Cc
zc}&n9;8v`~ppIP}EUJRjUrGyOeqr4rCINfc?$by9KJ}=P36G$MhmzT99*eqFcxA}D
zTPWh3U*Nlr^MBauPsl;9257nZt&A<6=?{UZJpV*Y>adibze-H=V%V9n(V2R4p;|0>
z7(TijIN(ZAMBQpp?5F8KtC~_mW^R(15zP$l%531Yw)CQr$CHxGFD<ADeDe%h(hzUZ
zt)ttrQ<^O-Nnmtj(!q4T3ZzA*_@($;i|O2+|4vW|o_Bl6m$(30!2tb<)!~lbzKR;=
zg!sjz94nE<ggr(zdeDG#`#OnOuDH^DEO9!(731)!Yp>kpi3|lRnF&mf&K6r6`G>BB
zFuUuJA7e2RAi+@~`%%>c8F7aO-}+}gX3le5$sgLG@lHW3LHmiQoqy`&)J}~`1%dJ%
zNR-eMQBVdNu5p%p9<kL%l%8-rVA3KU8IW8Wsnf<JV=!US1PQa~Jq1Td9w#L{J+>bT
z7J$ZG&HpR!JA<Oyx;EPcS0$)`pb|uaNE9SzBsM`LiGYA!BnwE+prWE=nj}ch8Oa$$
zg$Bv7fhH(92uf%|lV&ydeZTLUs+p>pA5$|`bBZ5rbI#det@W&DJ$uWRXavk)oTz)6
ze*JaTP0%$txU$}gjCXv}PS=*a?f8=Blmdrl#&v}t4P6ycwqJv9xVi5#hZ(3@@w@+>
zT{k}^Bg=!qys726uUtG*TE_d_XTV{22eV^%j2@lrm2pN^D(ZH3{=F9fjwhKuh@|)$
zDaUb%G)E&plCVl>n|kGP+`U{HokUpA%;d8<&TO4+gjHP$*f{Ysa^B&L+Pg^hjI>i3
zz2Z{2mime}2erE>PBOdjR$IcKX?@2FkGD~DW^cXD&aTTcv6DQjsobkhkjteUN$yxy
zu|nv-!?Y*}fr9!dQ83<!U8(#ALZzBTAmF<^U~@R1jCE(7GcAwf+iF$pF^oF$X#!l_
zsUC!5aQgFLX)dl^P0d4~%#edi%lv~j7lrkQijanij*qSBr0GPk8}M-OY<K$FjDiMn
zuuqgtkUqzaKz%oYn_ei2n*CcNmA$3)&3mT7IE`iC6_10ySkTNx|J7mD-7C}b-X|a$
zrFon<&Jb$$ZS9)^;*xCDeQJ7YOw+MHQ*?$C2WW>9)or?krEH-}trE8ZuH$2?cZ))>
z=4>@VT;)r<m4EV#+FzA*bNQtCG5Qr^$MjOb)hi5TATMM+%tZp4aoXrlrWv9Kfv=-N
zOhqc=Y0R9oMr@b~vt#g2n!8dkykP32Ei<UcsbOz@CaFf;bPm*$plj$Sv-dA*7To_l
zQODSEd0v>egI7O1m@e%*g^ApFvfpUcIc@He2{k>A2+SWtLeqCS^PMljP6hiFH?R7q
zbFfQ<*`qk1GQxtWT3$uycwg5T5n;4ND#TG~Tj5{W4m%EH>5aK|XEJWh-{j&|sK^Q#
z_-rOLyQ$fv{A75^lruvBV5|Gr(b3Add!#ll_GqoUz`#?tU(K~N`!Q4UUD{Ww7pUck
zF(x+xu7xior?mu`q0V-!u`#0UeeO)3d+j!0Kb39Y%oJYrvu#FBK`!)!9uX-MY@!^^
zR}-oqa{0Zzt(X-!L0yxZm$!5vn$>j+KVg6yDDcmnmvsKUeGM~w>F8(MP>_brKyQry
zR4ym!LVCD-N%h1<1mfv?G)kGqfEs~FKS6co1vD9)yF@YCAJ)VBb^n})#8nO2`|{!H
zVr2Z*cCvRn&nog_*e?CaPHoNmeyK5poy=mi_euTUyr{P(-QZ2)xY<&!U2%H(2!)a$
zc7I~$^@UHYk!MN{_)Z`Q7f}!q&kM#Z?zx5g{5dbUSPU1XweBkP{Rz{(!=a*mO)Z;o
zY2p5WrR8Y!j)1#d`1h|ydZ;duLhZC0YIsK*?egAqS`*9k?lg@jjrV9T4%la$;C4TK
z{ATx;gg4)ebJ7`))Rq;`@eW2dqY8s~*GK5zc+p4s*NJA{$ZmO^H=3PR%p#-nn%p;D
z{$zz2Ek&<!FuSHKXU=PD$}QkM*!2;9m*dV#A0{`Qf83h$UKC#=nfW$5?TL(mOpzje
zyXKg{s8-T_{wI{K83CVk>DyP<c{$rP!@J=}s<rpjaeqkYuht>nMpl`W6)_gpd`SB=
z`9MCLr<{{BXV|W+@z$+84P|CJ=)r#CujcazL=QtDZ?4U~T;8ATi&xzxqHp*ZWTZr=
zWWu}ffkZrfE%O^~%QlW9Ud*X*aVqeakz5nx#4?hceekjPd0;?*Hr;kBwX)T->Wr4x
zY0_-hld=we-m7ZO&K1mb`(n}efiGCf?86sH&rh=5!Vc;3SlYAxU(da{s~2Quaf_R~
z%jGZGyJ%V@YlL195uyfj!FAKr=Fwp1S%DTh5ruV+^c#%@CCLT$UuDnj?aW?9AU4TK
zjh+TA*C?|_tncb~eO6_@MC+uo?k0YY$S(jr?J!>GPtC?`RHW<U+kX4pd%3&BOILL;
zl>Ng`j8+OAan5R`VB0hG{^TxGA$dzJX2AicDc7skh0n?-$W<8K9Svt%aqhS%sn{AG
zGCHTk7yC2`L4&$VQ7D;X+lO&JI})vZX|rvaSGh{lkUs}(`mk@|+r=kwbqGYVvBjSo
zCbVcv%#YyazXO{8)(c_Z^O8(@)s&pt%NERzOu5fC_~AWzI@iwXftI7H;+vBJ_~1)t
zbv?-i+OlUaDWGYp3pI+nHH*6~KCEQt<lJg6x=~+e#3RzDaGT~C=7)_f2#|6km^h~5
zhpO-84)?3<)1i;z^meWaJP49A3Vlh|srVZ1l)oDbex`i&e<Fiy_Awy-cq!LPiESa!
zN>l!!SL6xAi-y#u^Ve6VLNuJU?gpRL4Y8gn<)YfNiF#94A|b(0FdRLz_RxVPIAn)v
z%fqSXw}#diojxTam2<n|0W5}ecRD65cZd*(7b$ZPN`AfR&s0|8YT;>btK7i6OpzRX
zRoau6``3KVL5LHP8X+v5lW*l1`lpHx$7^eSiJGnm;?kwHu>2FkL7H3Te=7Rj<OA?v
z?+zWyo}o|4x?f)zu^6`T7K)e5tSO0E{~h2a@#xE62*lgEw;$3n3GYrMZwxq~2L}=8
z`Edq;IFL3!Ej_9&bQ3qyWVVw3z73V3aWN@`3ui4CuHKuTmr%}okW+(Hd%ucC=d-Qn
zaQ^CVntT0s*H2s>{jJBIGdLyBC5o%`o4GfpmT^IkVPwx7gJ!Ch%5B%&#T}SM3a`X}
zMj)h(evrUe!k-7Z47QZjGN$x4AG(+ijT)(>G5HMg-z=|X%=yHVOTls1FeKv2x(C16
z2jjEog=JZLDq<U|-De~8IU=i8uHKI-&}mAanUYpC#(3AU{O%_ZpL=F;N)?LATYXh(
zG)!1j-Tn(w2tER=lx4qX)uB*hpIY`Ae*KfN=;{>jpd6h5>m&JnLt_tN(Ww}831u1A
zU<)nuU@-o<6I)y)hl)(m0F5iocP~zx_LD$KYnDmCL3}5K`rGuW7rM<pMY$g-#>wg*
z2lb{{Xfg$ONsyVtEcRR^n5vh?bs+M&4*k;Z3+AcH!FUV3%37i#=3vhC9Da*D!KDRh
zle<Y6g;}-Xb9>@fHD)4gki6(YU1$42`3S3fRdEk5r?(MWivPP&(ez1?RT^@|PT4|k
z7-L(_tFZSGweHRtb*5foND1JDuJ^~8u*-$vk1(xUo8o1aE5(sYagxDY^OX|p<a(pn
z3o*;eQ5h$qh{zfKP_{qN14xlf$;DTFkqm8#K25Z*5AWz}E8*0MCZbfU$;}O2ZkOdH
zJ?}P)d#Eg2xb%0(enm!F;Hg~BjA($>16e<6Y5kpg`rT)F7+!mm5RS%`@X1vyZ76YC
z2$`VRxgBwE;{x&MjI;pnQ9X+OhL$P?#r<_&RvrX`=I>WVp{&7wam?Tr9y@9VLoQM*
z;j+~JmQ@m_i;g}``G=2O5P?`=0O;+zF}0B`{1_r3<z3_Q9d%k0Z3ajFif^kM!w$IO
z^d%Fn>0W1a1t%5pGXa*Mt5YFw{dX(?x@FZrS2*33S0c4ngdZfj)!dWv>@fS185#WY
z#YE4U6JeJ#WQxAa8V$<Rt=K48JpGncG2bhiY0CDdFIhM3#!LF{bsLVKG%SWVj|YXn
zd18dO1jPivb1UTc1*d<VM7$_VfxzeDR1q!YOn+@%Rx4SUAfN%MpzcBYCsIIyud+sT
zhQIoI!&kdoR$43cZIPJP)s(i6{0fqx&CJ#?-+*$-5OiUJC$N3*sh&{n+UN+~9Hexe
zsZwA}QlyeM3jS-)$*H)}SZ!G8bKZ?cMI5i<%jcLF?yE>rUPTk-lkuegWT|uK2<|pp
z$ra2%Kfv2u2nb7yX<aU+T=Se+p8jS9c=C*t+yggXQ}xA2_lfTw`gV(jv5DrBBt=t3
zyCi>~<IW&jiTT}5JeB*rjBMxZ@xi8<_!CFnR?xxOQ#ISbz@CLtx&~Q&j3|+aZXwyQ
z?)X%A)AgVBTXo?<68BlgqWqGueR`F04Ml-Yq)1DvodNs#U2Xa;tFb@V&3~UHH1pyW
z!FTSlop;yyMGirbeQSho)34hE5^5f^MSF?G>OZ%AA>7vWXM{eO?dr}a@_%}``|94=
ze?B2**uXvhxx&&c{v1C9<_Q5f{@2$c0Yu<!$JhV2H?NzCArP>cHpmL%vKbY?63hvy
z2b^IvU}xo-KQRJvPRYjc?%q=BsBu4MP6jI+P^Nup<~9lNNb;yNFX}G3K7uR&{7%z9
z71>P7)W$ZmVd2)Ar*W~dh3tH*iK&<-0uk1YD(t<vcdfiqGV=h}Df$q@j6COI5Le3-
z=mjjbRYe2L)As@}xb8=lLK5a1&=T$9bPe}#b(ta-6HWmZTAjh>!T8$_kJ8q@`i+nQ
zjh(rLp;^ff&J|tEW{lRh{QN)zDUBM4cPa-Wmid|qL!-ua{<}%<44~~Pf~yz}V?XpH
zzI}}<<gegQ9!Np4ke(lAU-vA_6J%{fYr+fkhGROPpFl__=GwZH)pBR1WBRH}KE*%1
z8{8KCJ$=}YW2zLVW2ox6G|f)B%YUPT@d{uS4XfU%u#KSfrMYygxJy^5*7EW;9zSdE
zYkg--97iEa<4hzbhd|uFPR8Hjd^>JH<LfI!vh${T@O$;aP9&X05V_9_s6Kxa;(X(8
zq^_F1qEC_H1ij=SAn&CR4cT~tlPWc)TMvAA&cbxYgH8Elg<^C%aNKljA|7d&+!*lg
zR(Bv4HK=iab$%?9NFfL2`E3|@U%LQm`Ra)G2l37d2?s+ad@Ta;IQjT?Ni3?{`d0LF
z&qv(?T*MkDux58D;9V|8p)-{d0b+$7c&-tewAUsxm7WcFAOy)gY&Q`z;*BdB#f=%3
z*bfUnJ8yuH-AGboVPw4nkiR%z3}*$8LMb6ispeE`WZ0q6W2+a#kN`PXx_1^4My(+4
zl^e{eqeG3+Xx3~suQV`XaCBmuTNu8Sq;Pq+gpvh8`F3<v-v(Qu_EE@gNRh!iJudEQ
zRAi*Z_sCl{f4-a-a9^L1lxSHR%)|Lh?aK%V2*e57r*=slc_nynjg;7knP5+z5@4Uw
zwHX*ZY<XQ*ks<AWXk+q4!?_fy0iWQ2ol**B1frD#&}fDL<0dL6`A}N}=-JAMbM6v+
zsJKgoL7c%?<}g)#$tMNPzaykG2L=Z%n_itsIBf7e@+mN^ePE3#91Oa0r_60_%JyS{
zn9HvoI8U4rBO$)owY7pR_ug}<Kg2KB%qMzM{&l*qfh&wmUNRJFBq}6+6PJ=5wA`x9
z%z{)y>LSsDAD*0`Gq`j=JoG#}%+hB<iWUmt-r`umu}xL#P^+=e(YQ0~y!K{4J9_mF
zXYv%n-1dYtt6p?Q>>YEftzRQ?x1YRHR#BN<bE6DJms|F;av9cWUlXz&IL}j;3};+r
z_HW^EIQ=@e;zyXlC|#o*_l>z*7yh#NJyBn?w_1x^%u2G_U7v*;?DZ?p^jSU)S7-9O
zD8FB^g%MhQY^aLTzPy>sY_YuX#3zbKgc87fQAX#)^`gi(b?l_0RzbA^B88lKKvmDV
zbLj=y6C<wZK6;dowQH>JpFH*Qg;`g8&g0wJzPmBLy@5Op*?C0`HT`;ZMGg2|T6Ax|
zUg>DUS#gg|i*Yy<yatY;-JJ0JZ3zdcUKMe8I$Ck2&&I7at1nrdIIv#yf^Dol*JA_r
z^^sI&g0p=SkU^}6pi6*s>0I2(5&xw#JfBj19wGa!?Pd}ulCLGZp`l@%&m_pmXV(<j
zSLePVYIBXu-?>+w8i^VzGES<nd8t~=nIrr|$31)Y@>>5Ggv+9$9{>x&FLPzMiVppy
zprAm9bI9S=-UnWjnz6$HE%hj7o8eE4HdsP3sSaX<0ts>3!sXX6V@}4|Yv<mKa^)2k
zMl<VZYdeXZz-lPcu^{pL8{PdCHj1?_D}9L)o;8myzn9~OCk7fN=tI3EjYC6qm7ZbE
ze6v}2^!1{fB`y7EEqetOcW6aAN;n;6ct(w04%ZkXlEuJfp)t0pwW?@tLv|#7E!20+
z(%sN7A^mIcaABF?o9VSGS~QaG57|>}$b%xI0`o-o%y!e8NnRy~%Z-N|IJ>0oEEEc5
zn^f9!+E3<;U(+*x-t;i0l`|(_-ZTGAEn0+82`Q5aahk(<P?7nU(h<F2{AYw=_Jzw!
zaH#tOHMQWmj@UG01go?D&2S~9C2=QN1DlT}HvXP8trajA(%jR)+My7M<W*3K;U0|U
zGT?WdQyq8eS2HY}$5ljT7MLu$G~34y98|;zlS4-{Rbx}(gshTwQ<Ih;4RQLI$=~V{
zX4a0?J!kx6ST4PH*{i;<7afH!en~6BNkhUOs|r76Vv+gQADIf5+37>gMN(53FZ|VN
zPwOCOIIo0G2}UGS0oj4lwJlXXaI=}(*a^Sq*e!us$w+IV;WBP`^t)~&#H`~(3Y<b;
zFkbb_rraoXY<#B@H83^?2`&Y@6*Ttp8MaqsK0eK{OC$?BTV7tCLuAJnt1T-lTe*B0
z4)jTO6HgrNe$!F0;zDW!Wwvz)zkq=0Ki%>Qp`F{_YYT7u<0if$srYv>4RmMxn0366
zv^T6YHaD*jbj1`HWG(h*S+7ksOC*XI7#VRRM;m>8D@|rr$7^*_{670TWi|w&GQ)RF
zV-_N)I38@ek5_b#TlQyKwxL++yAVeD`f;M3dAYgDCg_lrvaXikYb?mD1Q&e>9(dH@
zN<=<5&GmvHUj2$^u|v>gj#kzZ918WTyqcptNJ!Ycix%0JtCJ_nWoK(U6enQKj0Cqa
z`}gC=^O&w??YO;CIXM)0OUnsP<ih1tW35)ls!KRk?CNpR4+<Z}#<?wgetAm2<mr1<
zTY9E?QekhJVwST3bm6+LUJFgc%Js0MxVX5dK^htw;mokzpaC6QaQ;pAycQDq-pblK
zmhOQj8>RBBlB}#>YOh5!(|5@tOkQ7~CIkS~^EblZG3OK)hqrk;2`iG(iGJ(&wG_`3
zZR#xclbY{0fJ$FM<pg`qZ2o9s%b53{CgNIJ<)2^4r3O$eDm7!7B_*2V;(uPeYI+>g
zPF!MRJUBFzTUMrxq6}=-Ha3<@Mz>SX&dyp{Sxq_Wb5U53GDto`0i5dS>c-5zd;4~M
z-9bP|=+4XN*6`KHsHj_4gHcRYu}=%~^HouiO3VYp!yoXgCtifm9An3oCqaw$niZ>E
zcRui$1|)i|BMXe`^?%oICz8>NXTPVox%BEhcTe?vB6j*p)2WfmFh{NO$FUtI2qDl5
zqY?dqzQx0X*={`Y_CDI1WEijkBffa_mFE8*o&C<&^5Gfr%YEG5<dKo}P<}XbXOfih
z9kY&B2RRz636Ev2`t|P|sPw$T8fEi%dxdDGFZVY9GL|rnhvm$c;4Pwize-DKFTRZn
zz9wWf`#lPk4hi8QBuU$7rn?P$OL;|Q&qnx=ePx#YWsjR)aVow}F(YKD?Fmm7q*kVC
zFR9R2*zCL81KPT}%ad?G7c(;ArutB)&}2N}TgcLjNk_X$EJzAg)gGz+uToD%ZXjf1
zg|H{CH?EA9t1M%RVwlOQdNXo!bG1V{;siL_dmpon)_c0LE8Y{CjpS7HOChAi9NJh>
zXRGKm8Ab~W3#HjdY?R*}_-O*bdo@M=#=1nOHQfil(G-~yzIXgN6s;xF2QDycfgo`v
z=>>BhU;lfh#rKaZh_K0f!aq+NdMsw3(oI4o1PR@P^nRCJ5u!R~{PAsOpc(eiJ@}5$
zveU0(bf&}1&U9$E13%ko78aJAc(KLwh!R@Tx2aQ8Q`#Z%Zd32|LWpEF*eFvQWI|8A
zlxA0C=in%%C4G_7b!<@ZtK(}y>aE4UKE9xpKlWq7KOHEK0f6waRpZ!42_LU3x|h-+
z*0pSG3J7UN3XYr0&B^Kg_MXD|aCd(BS7~>&4R(Z#{K^9ZgFN7JmK$@O_=U7kZ7r=d
z6zY;%ET25Jky{F@dg53X`nvx)6-X<#b!!S&Oeo*WZFb!0-mH=o(G&(Ql_fOcCi}bb
zsDbeLB9npzWFOI^{S5(!nO56MCdII`Gde$08=AtC8)RIw?GWh^ra`<TWG8n`(IIp-
z(;;FrHMUu%6gTe;NNkU#Nna0p(|Vg`X=%x6r5v+W)%9-@cm9Gy_8%IP&HkiAZ8&T&
zhFSi`MG0tTyih%ak_Z@t$0Ccw!Djz>VW|hs%Dq0pw}|z>Yg-rEetMGBC|$?fyZ$og
z!|P~!r`af7tBE>y_&fHawXQ|04STjLW7S$&Nj^1~d5q#$?uRo1G4q%SSLFBHdZzZ_
zwoFDw2FR+Qsg9OeMi5dw+J1BVw~sF7k)v7+HbVBzJOP9IGE@XyS0B~yueVFkBf^>9
zOhbYjDD9Ty0KiPzpSdYI8>vM}LKLM6JX<FNW9(}vwbMi{N(cA0o?g%Wq%rwJ>HcIg
z$AA!m?<KEW_=Eo>V-R;IA?pIW1o&PCDi&pf^#J55vzu&q<{v-;`tcC-q43Hbv+TOM
zM-?{hC7fqYp6HAh(URI5>1Hmc6bzYm8cn2S7bhhpg+u1v5RM;A@IBnEC(PvAQ^aHo
zBj!m_%o6LZOrp#ArIhF1Jk^T^%@({8_m$!BQ+ugF>&2%h-F1eu?Rty{gci=>NMER?
zGzFINwk*uX7)LXAeiSO|Uw<x~AIM#!gevnnz{>=a%AyIJFNei=X?YpT#pHjkKY)s3
zW@FcBqf@B8B%dC`qIeDsS&kVQW9(C{5K2EjzjEBpn+cc-rlo66Z}3FYnLR(`M+b(6
z*e?enQ7u7N`e#M65<PO6Ex*1c^}9^{Ho9V3-kFvm0ETB!@vIl3(iv-0d5eZ*s|Ng)
zCWzMqnq$dPH<u`v54ld!%CY$tPh_L<uer|Ia6kHdG50^{VfDBFpohd<+$3C3M}!S%
zet3O1#ihM9<;$gD1?NfcZA);5I7x4hqzXpGb6-<r{UI;0Ak!2>hk)0m>f}AJ#8%j_
zv$OXN<Y);%G6zRW>93gii|=LKT|1b=06funO2dgTB)?m@ptG|xip;mzDWm*HFi7{b
zSy-N%Gg)rpg&8U_$|#@xX$DvQdt_A6P<LtfT4)De=33eioKbp`T~H7;;BaAM?xWpO
zia)(FSD-?geyvNvP@!R7J5S>ya5C>&n>K>s-Xy|=a>0c9!-TT-!h|wkI*Ux;QL)Zj
ziGgJVuSLoW7-#40ao4=eOa;fej^1=+glK^*8C3|Kn2QAfHU1~LZ#K|WNjpVwnIV-6
z<d8P~={=Mj=eq`VbPY3<m{T8eW;!nqE&lXp7=ltzz<Ehq?V})9IhV*Q)$2t)bW7LD
z<h}i%Cx}+&7U)7**X9Vmc`&M?^qi-7x^W6fTZ?@eqa~(+g!<&8ZB0%)4yHUziFxz)
z(c$mh{QQA4)Q@_KY}&%V^ms^5)fC>7z-f;^|4Pcv%^hv32TLD>VpPk#LVWET#YzPW
z0Su^W?t1%Gz?5HGx`v4TcI1(+wp=vex%F$&C1GyP49R7De?Rphc!_?4SM`MZJYDAp
zUfIfV7lX<>XCb(M?#?G+z`g_!D#Peyh}{5te|qv&bDO7n2E~numGG=!UhMWgbX>~6
z{GL|EEKGOd9iE5|+)IcM%I_Jkh6w_TL8Tw;ZM%xibx+ho-ke5<G7Jod(G?I&Ee{p+
zZ!Y$iL)A*}tH$bo?kmti7G&jwhjo@|L1~x}KmS5WJ9h;0!T!7yrpkVr8NelV#k9cG
z)YPz0OItg#h9ML^yRi%3x00-%`<G)ckP;$vio_^g&C)Vw-=~o|2NG<%v!Lx58&Y(o
zte=RB`>7(<FX`W;^$?{>Ms*o&FFGE?n-E<o4@L$$Z4x4)(D(B!QWRL2RizJy!|7OQ
zqZBF}sfVn$-|Dd?uhuDDYcJvdA2Zgu<d(Nhra`?WrtM|6<DzN_qB^b<?&e9{6gOa`
zfCigDFkIX{cYgY_l*iZi;46np4VWF$!yftU!kmEz*SF6SD1yAzwhRJ6NrjY?(PmbQ
z7v>e<PLK<^ey5c~3E=1ojuijuUr?dnnINVcZMq25v+sKp2NfYOiOFbtfOAHwvI=<P
zor38lMBBq}Xz7#tO!(~C5E7s3xU+6!%M0j~!SV-ZiC{y5lLLAige%GYZrCeK1Q}zX
z$^q8Vjy9x@($Mstswk(gq(V-k(X=wGVY+XE@kFO&y!{>!Y~IXgQ$&5D@5jHH?xP2K
zdO5`OZqiMUpNFM=dAti_)-(y#iyoZ*;6yIDliU(Qhh6+VbN{DRXA*?be3EYi7S|g7
z<vwHz`M}HNwU3Gm;Be;$kg%5>t{g|xIcgj!5eym3jc7f+S6I7R!-18#-WYUMK$-f{
z^OqniH2D!z0b1>^p{)=1cWZD%23jV>j8cPxgBjf9DE%(C#nml1K)QDP>8Sz;v%xqi
zpGITdNA~iS%iv-uRaFlsce`9E$^3PIFW~VWQBhHrL-~68WftLG!qY*vmB;aQoam+h
zK^H{q=PUooh5zB*1ZMI-<0~Mm5Pl6L`hS7k{O{h}_X996a`HD+t{7qludtQp4vD-9
zz>$05K;B%9G=+?<6#`#16DuoYa&_oe;i-@Gm>it|R*~_{U+^(8IsLEoaznSDZ$`*S
zKVP2~W3SJ)yFD=L=*^@tjcB!-ct(tP>NhP6i@2wJ;t8ZVPq!$VSK+<f4bpHK)LiGs
zEaM|QCgCtqm6Zj_;JaT8HT4Z7{?Au(=c)HM$*jB<<yM~>v932ek8@87VVsHghh=4S
zi-N_^yp>$~m;xTP7`+<bwvAUvZ8;IvXz?fYamI9DUHNrI{IlUtDeZnq6{Gb0G$uq$
zoHMuA&qy<-G+c?xwgb!~IWD!YlS2BwmD`uT!t(fEr8Q$xcTDp;zQD0IZ%8CBrLSY3
zq^s%a>D9*9?Mwxhn{~!<DxAJ<uoI%puyk!E&MVK3*@_mi8_esb5%vrl&6|nE`jJb1
z(02tMH(t06er&d;Gm|xBKT1Gs(7xtyP2}r-rg|e}jA`QZ%4e?Y-k&5f91v<1*fp9|
z8W2XTpHH%orXKYf*=Uxkx_b30Gcrvf_`3s$e>2oeAAdnkL?B}2Oo`KF?CzYg8?O;4
z--ol!qnW+&i&^+AU;NEr1}P&5O7jHgXd|%Ay$BjwQ70v-qup-j@3*pGi?vb&t6JaD
zp6`)55IOIVqMRHg{0Fr7!m6orbj|}BV6HfROO$Or?m_+0;od3FHJd4~>0qHUi{2Xv
z_AOU(v$OBBQTCm^8J`RX(rZ8^{h%^t>0t8cz}V`j4zxK``e2@}YuwKYsO}+PZYDTs
zT45vZiRScss6=zNwfZb9dGAjJ@{~PoyOiM6ha@y8?(WdIg}vXAP?+clQ%49Ppeb3Y
zsW%c_MlC&{G9DZr9-cNKW#*~r8ym|f5D19bt3tN%Cttn@k;GdVI`ztXETs5rt||9H
z3L%c<S&R(c<_%X`5Ng~Fp6gC3TAN05T=*+HJzcIpQ>`EBuY8P=X8~rLT>8oKikR|u
z+hAGmh(vE0@BSlvra*|(pzuMCU3!YvG@$tB3x6Y=)_Y{huOJmryE75$9)`a2T?sk3
zN*nBu<u3TLR=exUsOd2bHi}Jf*ZsL0p~@P}+wu%s(Rtdhi5tVg!y^RDw!UR+)Ve74
zx$xM^29ghm1h1N!+QFoT@5R7o53l_pY))396-WgGmU~i%YoXqwhhW~-evisYA5VoX
zS?u;2L@mEM<n=7H@M_gGzWn#wzgL~S5*qhsd1}48o_u{9OL)*wvP*g+HBHyZJp;3S
zA7yD_al3`4)Vw>9sVioJ;LJHICdChnmdUsF9_r|r61e5Eb&`mqSy6F!>>accAKbe-
zdGF-yne^HQURF;5GMpNI>kLlqXz7|spTYEIC4B2BaF20NsO3I{J5PMmd6E+2!}XXN
z7+|g8AvnY&KJ;jZd}X5EfKWbwJ`%GG+JtpiyoiJ(hC*FDMZ|Fq_Alh1aYmji0tue0
z)d@a&>XmTl{DR%#oE?#OU$E*ge*&um^~Nx((cFQ7%xtxcP<#*=vTJQ!Vc|W3F+q9S
z+jKKrR523Y|MRtGvrom-o7_<*b5hYHh(P66MoMqm@Sc)3CK=kM#C1MdoS+^M;_B0w
zRH~mR!l{`<QhNTnWW<o@ZLVr!6Gxl2w)O`|t>gUs{4^M6HAlx%y>hFlH8*hu6zobd
z!*4Dsp!7hP8fo-xw5LS_RV;O0e&9$bZ0VQNLM4~@GNC$QFTV%W!hI<xOI1fF3dwva
zXMU{Cy%ZRci&Z^Tl0V;V>2hAaL-2vd-A=oXLc>~BXv*FYQ&ngU6+!1)(rX={HEVWn
z)J`50V$q`0L3e_f9Y;)KoNEb^y;dlQ_FMEOD)T-IJCbj|KG}dG^aHFsOxM^>x;!A3
z<#iR6%(B7Gt)Ur&KnrYfHlzq_?#NX8*a4E*XhKx4M$P)vMz@q5b+>7Zp;?6uHrk`^
zDW4_m(&`*7x9%L&H6C9jp*F$C@OTS^-Fd`!Ao&1NZZt50=4Yd2Z37rg)|NN)d8FDg
z%Y81^yo0N9tk3#a{P@Y4P}QU+F`7xO0inf7nC*i+yC9^xElC-;&B1?h8-BN!+;F-r
zDBPVdPZ8I6#rre(R2hbE0%v!ro&a*KxlFLBVWHyz4s#7@w)5v_CS2HIIy5gKkB2|K
z2V~aBcsbmG79H<{2g_7Ljah4G2Q_zn;YP0_15W>2r7}#>!*4!N=SxVs8k01SE$)a)
z?fmfL#ss3L=RQvEn_1QKH13y#7U4fN%rt}QuGwAewY%;krRck!QAFf`h+N!WQ}@Md
zSwSV}J#<8kIYBHXec0;c=X$##E8C7t1=En+V97&=3+An#Z;@VxC^4R<)butX5g>c9
z|K$Rt$Xd{Zbc~E9W=kNK@;e^|-RbH9v9^EK<7WKxS72>c<Y<GpXVxsxD0QiWGUmmX
z#Ps(qEiEg-_2|dl=Mzxrp;Cv!CJ~q5-v<kdQYhwwt;)vER!67Ewu4WA&;+Sv@q7h2
z5tFMZzP52s3`~{LZYr6i4iCC5O*Zb0|Mu{%3-M?Vgkpr+*h>TDwBB}e`KHt^vN6dd
zyKQq%*yY#5kqR#4XqCN*=v|^0l5Lyw*C1$ON5PDX0Evl?et4)55q-f#JhxlBK%%gX
zDd!4kfLFWJCYtKu^BVCS$i4>eJyEVn(5bBQ%M5I*<>3OT&oQ`u;XeEw=uQvMAwBbq
z3W?`Gq#KR3@__v>gRKS$+=GDBl;_GnwpBVmD$LAeZJu6Z@=0v?KI2n06~MuF)l-)+
zJpE6PWkTDJ@I^<W&c5x@T!w{r!s@ai>5Lm!Jx}3mUlN#0;qL!!a{WTjoo;a^WSo?5
z0|;i62UFCIJuunYhK3ShJtvYCV-LO38bmm?_FmyL<&BrFEw;h{Gpz}rFjF@|hG64w
zZf+ju-f$y3n$y(K=$)QEeewj-P~D@~<?cJV8!2)(dgpgSHb2i8AK{F(%>6gzZ-j?U
z{+3FeU&>Qm^@kE+?%lMgreC-i9i8W|VhQeTH*v!9pX9-&7SLw%glKYpo~_mKk2O-3
zed&@?UH1iPeoyc3Za5PxbGY{#)2}X70F`1!bW}`?_0KPFY8Ep;@>>mL!=c`Y@$dh|
zJpM_QS$Ku34{h&HI)#FXLP*U2ob<~tYV0AOiSa_)hqz&x)?hgZm-m^DGj@!;Am$xu
zW>aedlY9Q^5K__TgJ9{s`L82feLw0$2`XiTx?~R2**pz9cL~HW)Mt}PK277`m$_8`
zLT?%l-bA_}f=Dh<^=;6Yq>auYWY0=xqPS}VLW&{KevN`3b*n6sVu<Rvaiwj7YWwQ|
z?;1Ma{j~!RZj7_mbRft!2TQt*eO(D+I>yF4XnRn;mWp+Yj1q=0BOjZ_$aj|@-~F<$
zUarAIg~^P(D&^x<+nB|~Gn)8v4RDQoCWDd_TOnXOChWb3og4!Cm0eRKD)Hy@DO5V7
zP9tNlHTh_!(=QrQ8bpGwnm-OHYp^sF>={dLxFQS{uSF;kPnbWbCJ{A_4zu5bBryO(
zpsw>EJ-{^1<Ib2Ikl}$)(T8S@7j4FVxR8?7{)fm>Y9<df;6n5)gV!IKr9)7EHMjet
zsBL(t4w`v<dRK#Dn0jN{1ve7V)4i~qqj?1NT1+VoDPofjVrfA5lw&$BI5Kfe$(2Bt
zfnu;Z0ThaZ0w!QLA&%^WERu^`#+d)A9u3*OOhN>JhAA|J)et@X#5=TPS*zRp;Vq9>
z{k?F*CZU$VnGCsASR)9Ip?TEs<JI}&-n^8S)-`mFVul)44c~pV-z|xr$SvAM@l-q;
zNrehJ5AD|sU^;Lh;kiYk@DOqP@qm!O#xzMs_>hpMhXWIUnf61-{LpLI7aNwf+IVz0
za8TMMs&by1djQSgoBxnA8ecmtoLf*Zxc9rU2<YGMcVzb!f@s2Pd|~b^5AZlETifUX
zzJ`a+&gJyJ`}S&a{5i%HFE-OxRO>)aNrKgd;guS{a)$L0(1OYxW|aj51(mHvj2V1(
zQtn9|IN40p>1noiv_W~5W78K)y*=irzH|tQCKc9+6&SoXHLS!;f<PVJB<S`n*X&fA
zM{=chQRMIyL8DA?_<qX+l85t;&o2j>?sH!5hiWHzqf3mZ{193!rD8{H67Y>+6K1eU
z!x)W)HxtzZpo6Cn5@}86XL7!OS0;F=@_}b_02P59!`cpKJIYjV-J|}hq*oPTQBG{N
zN*4Itfe(G-WKrYMUSkY1kR1>UyEKJqL3{y?6sU8Ln$URqQ@)%x=czMi%c-cSEMbQL
zDt+>BtCF3ECkAfYJAfZpkwRLHH;t0-T0yvjwu+zFZpcz%MvWRz?#~Hn-J!--6cAQO
z$6XdX0?$&gmWf^WhUA6|w)2wpX_UgbfvheHbO85_LfvV!0j5J;&~73h2$#jTcNcS1
z6#SZ!Gw&tj0@qTqie2-7g$}(CN(4RxKX=<Ip2?Nqod22#v*U>{Buqr{a$)Vg#-oGZ
z0$w|Rf+V}wXWO-!O?6>54!3=ebm!w8?~NvI6r<^vH!J@EMoY6r!YC|Y6riihSGETR
z29~>Tm<7{`RrHrQ%{4=8l!486y&4+MOvw1IvtT;o1XMzS(bu*M*77;bDA|nHlw)v+
z$JxgLJl{%^-VojwbL~`tUHX?Cdr+{G%?Xj?x>`MN)txBeL8w(g5C#oIqr!%RmMjnA
zmr$rX9UW|2PqzdUR%SvMMDC`9RoIT_Mses8_9LA0{|eJi(00qbb{$q=B@^K9-tw|2
zU>puIz9kF*-y+n|YyCu=W{;5pv}WIZwC@0DQO@mLl?&NvAt)OxeyFk`14PLs0%za4
zROK|^3>2zhX~{|g1h<+Ql!~^71|zmYM?)h898evyRzKw66xcC+54$B~><|m@Xd?GX
zh9~MlsjK-_($4R-^W0sWo>10J?oDiOe&;A!Agmm|K1)$T$Bm%9xb*RBwmshNI#_IC
zj7AAu_ufCQTdFFW`Rg=dCYZ~hV!>b%ADte3_U>rcK=9?f2L;M%w!j5vw)aYDHRc{U
zw6S==+5#b5V5f*6_9p`ITrS(rLstnGEEjz0dgF7#UY=7I{yME`LRY9?&W`;_G7S2O
zRuQ^{vmOZl2pQ~NBPschg7|;DKaG%0VL{Ll^w3+z{-1a_INGro*dtsM)BUe@?ElMy
y?SCi;tBizehPmtiq!jW`BmXBvmj4&teBdjqZNvU<UAlnK8>IX_xx%}TU;Q^uX5mr*

diff --git a/pkgdown.yml b/pkgdown.yml
index cee155fa..9d2f1a13 100644
--- a/pkgdown.yml
+++ b/pkgdown.yml
@@ -6,7 +6,7 @@ articles:
   data-slices: data-slices.html
   gratia: gratia.html
   articles/posterior-simulation: posterior-simulation.html
-last_built: 2024-11-25T14:17Z
+last_built: 2024-11-25T14:28Z
 urls:
   reference: https://gavinsimpson.github.io/gratia/reference
   article: https://gavinsimpson.github.io/gratia/articles
diff --git a/reference/draw.smooth_samples-2.png b/reference/draw.smooth_samples-2.png
index 79c76f9f449bd17fb5b1c28f6e36b618dc7e9e88..957c1e62ea861fc7de80943058ed728daf8e4442 100644
GIT binary patch
literal 177624
zcmd?R_dnKe|37{zg@zr{P|8RVk`a<JDiV@pW`xX;QC5+xBuPk;B!nbcA;~BiC1h_U
zo65@i-jBN8*ZX_>{sG?~J|8#N)$8?&^L(Di^Ee)l`+7VB)zy^Oui;!np-|QzJF1{b
zp{#PFP*$4IuEu|9yB9x!KUSNlC@bJke9avx@Wvl>7muEGqENQ@lmAhLew#c`p=_fZ
zQ;^qkdoa@B>ZWzRRc>PZ_nl)5`>#`WQ6D~h*#66tfR{Sg4Ajpknr(1cb;-cBamwKA
z+Moa(#;P~8AGkHO#J;Rh@xRV|`0$Fsh~Ez@8-C62D1IPuWCit6(dg<KCxd*+eNMH}
z^VRE|9y)#c)gG$ANKK{i->?3QPp0Xw2K@KSj)A4{zkjJfA4F%!{ogN|a_YTI|ML~*
zYvL36-(UM6FH)lN-{0$h6xGf1-(Pc7SDj$~pI>8=;^%b>{O|8=X8phU?Xm;!++ohk
z%R7{!)tmL{V3w<-t12V4f|AnV6fKjAn;WO;{mDw{QnjUFZ5uWorIu$L*+T7n(CK@1
zNset>cTdl?ckd4Kv#u#9)6mekb|OaHcxJ3a^77A}s=O?Rj~sF6@|-bVT9~7yrL~-V
zbBjyvR;KCu6_m#b2_hH1Jjq!TLtei1T2Wh?QUcSeRjZDkIHBm_A!A}<a`E>MjsE`r
zfMEq?Wz+UN2f<FhzZX&KPiJW5%r2z=w!6faiqhpZFSylxM5LfB)4Yark6z9??bNB)
z!wWqO-fwTL-=y0f#dXj_(!|uXuE5p)Nn)ZrZ+NzL9@!`jw?;b5)1N)_iHT`D(Io9L
zD?Ijk__<!I#J{)k<Kawh6x7dSq;&W8`gAQX&A0>_SKi@qn=EIyk7{3aRZuVITy>O)
zFvmK<*70YjQz?|QXU{fcQMSB4MqYoqj;S@x*3VWO`*7eL_1^4fSr(sqe-GAEt~MEY
zcL*hJ%(3eTI(cwTp-}e9Z%h5Wi&VVfbZo(^T3TA%rduq^-<z76PX4YxnQPhAR}pd~
zbgL|d;=S}M&Nx#?SHg|#ZY!rP@il8G*REZ2TbO*G=`yOjS;mbKm+^Uc%C*oa=PdU5
zgS9GR>Z$idOjmo&k4jM9Hm07bOV`VF@$^clqdj@@WW(35D=1G-#@#aVUfiI0p}tG>
z;KIa_Jw|Wz3q1$F@La!sz5Tb;!MT2xJ!fB1Q(jnn+Nv5Oo*wa-5AXkY`rNqqaLv)3
z%noxd$Hgc%SvY1=EYsbI+Nadjw~60XzP0&6<c*&h#Y+ObuIEKJzFA(HHxDnMytp)|
zxo(rVYGH@)-BEM%PY(r!is!z^<Z|sH$27a()@O5DhJUZd!0VRMXw&ifw7y}XTc&}D
zPQGJe-oi-Mr_PJLYXTK~ycdU!dPg(f{CTFB?N;DCeB;_mniUka2a?wo78axfqW*5M
zQ_7ZEIW+&)gNCJ$I<+xqhw`m+b3eavNICA5k(G_tOzH0H^ZoGQXmy+%myyThsypG~
z4hcb9t(%^2h~ztQ|5E9IShOnb`t^RADc3bq%Hoz62H47_oQJd=V$Nonx0IQRaXR#t
z(M?VlUjAkG`jfSPfx#P(xu<QEx8Wxb?h;!5S+!TN^#-{bJ0pt=Svh?Fu4bk6#DV;V
z1})WxG8=K@1a+6^$BT0}jx|0?OIv4XXh@;>9j5XvId@7rO$r<1M|~1Qn)dFie}1mn
zXI`B*_Gi3kX6w?Qrs5Tp!lI(EurR7vHNJxnqoM*ne^#G#xqMlwy`#haimJT4{Ly2_
zs3>WAx%9zog0Ba~MMOk6L`<l!{#on$<HwH}yCY=J<|ggSKCAb5ztzNwf2+TCy4%@=
zhlhujfx+J4$edDboZMcalMiL)o~}PnnfUc9Ya_So^5RV2P-AM%^Gz%)JH+o&Z98z_
z^|2%QlMeh5AK7?Zsm?9TPWC$DNC)Q(b`=-rc}#T6N2xw#vsM24mQS;?&$hArJ9O9R
z4Ed&OrN;wU?&QCIZSw1T&4<d$Yju{_8?F5CZJ42KXt<Srv*cBFBM&Cs^qdyY*n^(B
z`gT8fJ~M~P{0xx1RBa||{PFHC;g}lnVJ&U#*6Kg=b;?}ZWn>B~-#9utmYw(ZUS@m$
z*{!3Yp@Ge|@B8=FeuuYr)>c(nPyQl%Pov5?N8H-oBqkyt;DeNK$#9v=Nb`0fE@_u_
zxEkM#jJ>RDR_$b_!*_$#{wu$KT&oP{Go2c0%o~3*f4}35jZIpf%(=H$^@2=0UOUsS
zTSraVxpU{amY3&yO|?@u%yfCneXOdYqSQY<QMz_-;pY)AM<=K9Z-zxzvS;TCT*rlk
za>Yv4QhaF`*djL_mtiEAG(2Qu`*pfPd^i4i;5~H)OX{;{%#pr6K4X?fA$)wpt=YqZ
zwy9bIOk1|BPkFV}mxn^2pr}Yi>GWRq+<f2??IF+gSzKPTJ+F@pa$a$vOcia=edfRr
zS}~s&-C5)wTs5*@9=D=F{D59ec2Q={as2a~2tGRR$%lJ{yqD$%T*V6P`|eu`rhB&O
zIAzy-%`mb*Rcf=mw0P0kxoGU5uIZaY!R23iopJtng^V)jH<6|6h_3x=ZfpN{BRPAi
zP8`TY-80ZP%(2<HV#SIH+~#fK;%pHS5xK82DrAi|HO;-=(DYSY@0DfPY%>ZfKD$Hd
z=%$&n^ykm*CVvf0U2&Mx<}Fc2uW~dnFbK}wK}yiQ;s4L#Y!b6%oE~Xe$<EGBIdtX9
zfu-r(V&mC~o{O%o{z*xEY=XMikNDH?k5=f_4GB+m6_vK{Yc*g0R5jWsQ6;kTlAe)~
zkwgCEb!&-JIy(G9G}P3u+@yU=wh5&+4zbz)-LisBEv;#zA*8ZDt_{VOm9T9UyL<QU
zxwafzvJJ}an^4pZ4R1{iH<^BUa%^IHy5hM(!SjN+itHi*0RcPdt@-+P`);nIu<g<L
z{Fu``gil-L<VpKUCv!8ifU4X(-Gco5jFFp8HYkw;lAek)*|zEL0g2^8@rpR^r@Zwt
zH8#tQO0&aiYV=-n-`Hj5`onKmno+KP{CHeuw%b>yz{NIe-m={{-XXMEZwiau-JEGE
zJ|nv{+b8&p{cB!{ofxN{jM?~yupPaQxRZJuDYol#AIiGtxJi@V^4d8zv8J`N*&H}(
zXaR6)lV%F*n>TOrW;)#`Yb3f3Wi<8qG8|;&^6v3tzdt%E>@j=MVeU(!O72v9(#g0A
z>x?4a&aJXn#IFA7al&pd7`$Cz?O%yn((zkrcxdSMg9m#$w@TRQ2ls^p2N#Svv5>9a
z=FpGRXG%_A@-I`LNV}u)>VM1hflJA5Myjf+;ZvLv_hQ@^%(0NdsJ@fa199HBa6{vb
zioGaTXJ_5ex)yHCW0!D<+eEp!<J3xcc3Qg|zIsJtfBt;pi4O7U8{6;!Uvep1{cjm&
zDQw{G>FMcpbSiD(-m#<n+m9cXv_7AEJ@wMHtMBhu*^@F+(6vZvr;lce!lg^EbWBJ4
zLggy3V0L*K#S7o|_Wn$Y^9mXo8XA7So7Sf$tvXb0$t{@gp=JB9D=J{F?Am?gRAg^e
zM)ha^PxWUhRQq^NxI4Sa@W`=al_#1Sho_qJos!%N9MaoXqx?_I%+RPCEsO8nn^B&b
zrk~I3&%m{sa%r@6gI<me6JIj^K6fxs+WPCBm_Y95`|&4##Bs{ia+8vUxnIWFx-kYb
zz;&g8EPQ2cA>pMG^^AKd)iDwo<!$6zQ0%^9xsCs~*ey+Ec$d6?Pebq>N`nTv_}6tT
z)Koa7l5P{*ureBkhV&Hm)JF01jGTHe%>wg0eteE+D`#Y3@jYYYdAiWE`z;j^WpmoE
z`L5-yp|aB)SO46Y4Y7(z(Gu{OYCO|B8t3(g?$lFNc}K^WQ9rQN-~RkoHM#KXQ}A%j
zL)lI7@$rH|8!6&8&D(@NJ&?-!ccpddgEvtF-;o*^=5(7jS5zqQz<S@3&%KV+A)}o|
zn`Sy(jj}?&4GaW37b%5q4e5Iy+I7(gmyH!--_*<Hy^wjrcfQq)W*x`X-@h+W-eRXt
z_V{y&7d+oa>-Ig`!iKp>f_dxKTj6{s4=2>UvRr)pn47FKMflPW^|FeJ9Ls;lVt;Ht
zAea4{Io?jMdE1zIbmKj2!ESm25r=uh_s{FEmko$1{d<|e&M_GBC2FQzoa|MU`ThBb
zpl(-b=}PyRu_C)88!4o9T&RCaMW9NZO-I&hpKaaOXtpwL`Z&|>6nY}G^?R?aewz(C
zOLMNhQP%MBoy=)Q#anlud0sf^Ad8oi^!#%vm|!W-c^8K{;1P@eR7Ab{=(e;&p2tMO
zE6sj2N#?I@;Pc2ZNSZ9W(Dc6-cG$r|#A|Wt4EjV*Ss=@Y+S(gR3Gzpe(vrmk1mR-g
zQ@gG&R6O_w*txtV5IEkT@aM<fHP2_3=EsBg+ZX*0+;FX5v~PK-n8x)~J2fmK!cWNH
zwZ7}(moJ(&%`d2mm;UU=2c$}<BY>p(ffP&V!9V<eX2!ki2W&t!evfr*l9G}lB@)0?
z@Aqrc@akg8%|^4>vRgYXZ*b|pX|`@gn{VX*Lh08%rT#n6Jgl-gzkU0Lvp!9q(zqEt
z8bEm^#isQY-LHX9xlW{fDqNEOTf-jL`Vd6#IrlA+a`pA=1Avr`wRNwaf2)dIN%8RX
z6mcEXmsb9Vf`b3P)6kwM09ke>rtM#k;k2C%wz@9mb1F%#ry=DOWyoUZcVP)qDBqWp
zyUpaxraGY+qf_i9YbLt!`0?W}>^cs(`#Zg-mVfKBYAuE0+T|%R_GaEr`fm$Sa1HwR
z9_w^`|KB7wlji?qg3FWZN@-Yn_Uzf1rF)>liiqQ@C@*x=g*ijgneLTp@Uz}No(MWM
zO<yk5?<m#(^G^n2cqMgkxyeAl$dPpIQX6r6w@I@W;D+Bmy7TXClvXz7LSs6f$h_K@
z%4b`%^aGdoPKRY5va~Ke=pGts9h!e~YgDZ*P2#b3YWaB!8ShFnQP9;S8Q00eZ|wtV
zH$EMtB-xe%J?W%tN2&6<`+pe&!oxkJytNd2XO-P5^>{i-J2iItPf^ix18|sEy0$T+
z%U7bhoe#TS;f|~{`#gAwGE~eOn`um0=&CF^8(Qqpm;C&B?FA#6ST&mP%KlgHGwLkM
zo~WVw)En$vG1aw7NJt1j;qvhY^C!Bg2XiJHE#sV4uoTuN2`<a!gN~z^Fj6Nic;m9-
z8?7`>pZ<1<j{afpkoR4uufrVKx`Gcg9!xDZS&FI6cO<Fvm(3JEeOfWdyucht{rvfJ
ztfAW6l6AJONQse#MxFVm9g$U|ce>99t|t5S-`D@&QVO~CfRzq(K39W+*08d&5`KfT
zo1C0{$ktZS%G!E+*ii}v^w-(h89+vzpY^k)*k`M3AWT1hf11L=!q15+>~8MvM|i_q
z=8_4K0zyAa>vMHz$SSX(piTO&cqx&Gs3<p;c6_O?54(5oMq1h)Xb9URB-oV_CSSMV
z#*0+CCroi3P1L+J_+^!sm)DQRRQ5m9qrq3NQrz6!UbQ+fGHJ~3-o3jNhjtt7ic1H!
zh}*SWj^8HlMxj9C2)K2tVtL`y^4qSiuFjrUbVN*gaJS^#TcF3X;+r?u>7;)Z_f!MJ
z^zm7N&k2mHUD#DswRYv|b%#)2-@SVWrdj`#ywNZdH5Cf_^%k4e>T=I~@fOKDB)t|S
z-^`Cm<623*ot^Cq-5%lcCMcU|SW;oAt7T<n-R=2KM~@$8-dmBOoKzo=p7P^s`X(^I
zp20ynS65de>7+KkLnB9p6pda7)Fm9Pitps~bj1@V*5e2gfQlDmWn+8%=n(}!z(^f`
zO#l6@Eu~OsC=^-u=`si`B4~cpG&HUsS%jH1EFAhjFoDgGKJ^xec4P}ZK@pT~+qYAJ
zGW)r?1MU9!lDHP-Z-nD?+d1fS$>C%%fj=((>^}zm#lTgTfq{W~^=fWmVMfpSX~hh~
zqVDhCdGDRe>lAmqyqKbZE_$q8{QWs$a@(S#qmQbn965e`4al2KPEA$S5ir2Vk00ZA
zZWx!|Fopv2qpNs%ecyJd5vL6dLeO7mfi6l=q@#%j_?UkCzI`mIXENB7*+Mw?uUX43
zH1et`oUEqvVTeu^pB|RSAMw9^`}Sv>-qQ(n+xF~PXKZZDA$f7zYnRbhAwFr~ZLh^Y
z1YbspU0|j>%E@7?tgJk$s#@l^mi=~2%<kt#XU_%#FG8s*ZEM>E=<)tK9g9xIwSeT6
zvp*6-AF1<CRPX!r2&kyDhcvkOygZKa;^l(?ixf&{XD0+v?bPi0TlDA2im~hhCW4A_
zadyI3>&zL6%k6m4Qhx?^6|N6vqS6X2cePU$&CGTJjEQ(Jd9bpx_l>k<1q24lYiiz!
zdL*=KS4nNGbf$Ha&^k^rS~M(0bipLqV=PS0X5`Zj)+bF?Y_dO;uH7cRnN$P3Nng=S
z=Mks{SN{Z-sHmto3^nXZGc4lVTlD)*><*hvfA^lGJmYN$3(3j+W##4j2h&sQ2ZgN~
z_-}5JW`hDnFS4^7`;GA0R~PElAqE0T_0CTALDE~L?p-}OIT;Wds@(M4Km@==^?@Xl
zK*Yz|Fcu~$e{|zQu!?}2H_Kiam)co#B_<|*`0&AUS_JwImcB#&y2RI6{bY6i){)Bx
zp55Sb`{sB#_~UM8-Vkrrd&d(|>{Jppp+Y6eOjpEiG82tH^=375U|EXX{Q7>!{3H78
z*-vlujaJ8b`j%W}kXv9plVRwmz<sUc+|?Ev!l5o)*afBgQPXhS@7dADxd5U$)11%L
zA$xWK!p(<|A75Tp<qjizBb@idVKcKQI($t``2_{0e|`@}9rQc^5<dxzkR5`T{OQvh
zD34yeVB92OS9WWQG+`;KaR;}cFuUBkACT;%mut5Lcm9P<%geS-WmVO6OG`7w67x~$
zVSJr7DyZROV`H{UGFu{IVq%Jpyx5(uKh`i_ypKzolH6#Ow8e8<c;$no#iqea)AWSJ
ztrz1ADJr#eM(u=JrKqH|8Vw9b@KK|cv3N|_ojWVeoH=u8ymJd@$ZBXNHNAN)3g2zZ
z9WrJ~<!Z612-~4N@#~XZ!iyK5zl+4G`Qx;j{Oqs%(VCqa%WWeK(ibIRzjf>7ex?Tp
zoY$b(D1uf~9zA`!y1-@hu$C6<Q?=M2_wk~p!M*x{0j-BMHJKssX1Yz9B%Nf$eh*tj
z)ps8L+VfZZeDGj%UVeUwrC9IBh}{!Ee;)37;|byFT5GE=;SEx~U*DsQhxu9_9T{uQ
zUW>{k9_oy%hu-zpz_nvlqUMDp+O|+=aQHVrhrWvKQc@d>7k;kArl)!2rBgfBQIKih
zbr6+dC1C|g3x4+lK5T^^{g>Qyb#W0kn&1g!y_P}tw|Z7+<u^wMle;;OG*h;P9YxV!
zxB5to>`7&wh~x2E0>;7|51uRdDsVHY95|b<%@Lv`t=zB#d_uOF*YqZa6B02sF2ffr
zPU8OgY<p`a+9G||g*0W+EhsPAsg1Qv*>c&s>;FNSadq8Bh(Mt@MQB}MBIr5*8r*_{
z>!BC<#>H`ACr;pazAZ0LJkiDhNOPWAka!eZ9yD5&Pxigf1m0)eu%SifYtl(BP}8-g
zH`Y6S46J}6g<oZ4X7(8wu~|UdkaQlp>sImxbzr2cn2X`y>^6cw{e@1;D>N{9U}?~M
z-{KyO=eZbLTJl4!&zN^}aG2aOx~z2fcp`_lQ}`Vh7knyw@87>)EwUSZ*~F*GMqK*a
z{CIEgia>?lyYHU^S5ar+C>dq|aF2ch#bFCB-__*}A(Ge4+S)o+sAO;0(MQ^;Sjkqg
zl9o>Ttug^nN$9a(2L}g<EQ2Gct<832dL+=U<Mnn)Ne<Afr_D~ef^y4_`RzhmZn`B7
zUmp{rYjB&tv0}6~x##iQmQb!_@AVI+oGGni%h$Hi1_TF}-V@Zj9Ui{pAUlPcnz|e0
zPRY$J&w!7{a-kj7;Cbfk{CuAuyWt^pkR1Ci!t_=o@9lVTp`Lf7Etdg|;r;FHM|Ahs
z-k32Ih&W)dm;ceA>!U0$b}^5o*tywFd!>hmyVr3lsmOq#o!Dx9lvNvrvxJ6=nnG%&
z=bzsfa%TxBx%@!Zovq8`*ZOMneVV6E#n%sFQ@woo@-r+502oscW}QMehop@B8i;v6
zfBv+}_T0r9q}>s>!*RNgX2_lDQU#~DU?ZF2(5X|WCda*(H^J|qOnC3}fE;xWb_&OV
zOP2mIQ`rd%MmRz59?ARO2HLSNUSvaoU!^XmH{Y6VEn@b8aSsd(oZs|6g=t65cI`+@
z@Sq<l$J$*SJEroWvgU@1OHO<o+CxL8e!e3qAATk04x!t=(iOyh!lixAFp>khG978o
zBrN${wr<$%+gB$hF1q`7m=g@Ve&44qA#i)%U)APWr|7MzhWuyHrXa626QnK>(L0fY
z0-u!DL*2k9dQ-_Xwb6d}-o5YPWN<;7{P0x5^3y|E!oF^9mU?+)eTtKV1C#qu$Dpf<
z^)Rr+y}d@a^715Y+Vhy8e^aj3C!Or8j^39CCt<SNm!Z?4ruW8n$y&+k4c$`lRG;W<
zz1092ZKpiHq|c-GKA+cdaS><RYq$=2AS!O+mITuBEbc@``kzkKj(_=*dDpI8fSTCb
z{6a!X=gw^d<|AsL+|p!da!N{xGMDFdz`BKBamypGjv4a0#eR-us|W7yEiY`ea{k!6
zaL6<Cc#780tM^^%1+>X=7t7E7MgGC%j_=>wi$+hh*%uq5?}^W&OO%!}$lXMNg;?O6
z7kCJzJF0s>XULY93WR)jUKti7f^C=gvh1_eR0uee`1$UMo>I4=lq6xx&yU(Viy?$7
zB|sFv^l_`T3}3@3@TQ|X)!Y|G9n{YG<n~!}oc&U#6otC0F`%KXeFL490S!-4SC{kd
zaUK*d=+8{7tkfY&zFmuBu2K6e=!1fSCcx@<$jNc7@mm2zS{7;Kbz%I7B@1c%EiH4u
z1~MJ`lz?UfW_GczA(b`Dyk_Wkwoj|CmOy!bWw?k*87&y~1{#K~)D&RK-VWEUa)2OV
z02Iq?U1?=#%H*piB|W_yHHUz66u%GO9Ptec>aq9o686gM;<$Hc^gavj{rgXkd|TV#
zs;2wKBVY2*%PeCcqYr99!NKH09!NP=d|zVPp>(dZ@bbjuWacTe*hsY`g4u<=7Ti33
z)kv7M+<;|QoOiKj1yok}%e2zUe{DGhg^-OQO60%}e2P;>*#XtVqU>Q?>*MPi52RMc
za<Y4@U|h+@X3xRJ-={KdqM#aUr|Db+MUduFY$I2a92;F@4(Y@gcv;+8U~yq#0tMIf
z?Nut5o3&EZt5zA82W@CBa+h!!Yp2U<x%u*?sK-PJ4Z)9FJ%4LtKIDfGs0!#iQGZgd
zMS2}}<d7MZq_63E{*eB;_wHrTOwm%Em>zCoaLnKA{ii-o*u0tpFpnli+?KhDW*WK)
zz!@{(hiR9ur75{j*$Gdc>|B8AretUL+$~Nwy#hZka{eQ0%z=9iIrx}k#mipnwj5wJ
z@|xp8U817kZYjXF$LVNK;a$A^_J6Vf-aY@!^PwM5FLJxQg11P!5JdBK`NZ#$@o|Pk
z%|ITNdk|4q5R3x}w9K}&v^3MKlG##>)7`Jjj=1kuG-w*luP(5_ub~8&ufR7*(Tf+t
zs?nmW0EU1D{SU`q#a7l%^;1jo4Gvz*mwLgp=*rZ)P4<P?ItpA(vAQpG_VDl7Qzo}O
zcT(o(haG?A^}v3@#z2-9M*tAWr}_BtBTDQO&0T~hrk>fGtR+zR`-=)(=eO^<yl*x)
z>o^qP(DhdG#a+6uhK0pyapS%e7Rs2NKfj%ij|L~C9B`hIot>u0aPM<4{+@vW|LSU0
zkHx8`rj9sOUf+TOX%7z%!a|8I2!f=5!c~``%mANjS~5*7BVJ-0lxcEqAZ`b!3)&ly
z9&xiDx{pbr%{NP{lG4M?os<a|wQzo{AQ|p6PV+We!a`p=53dIJ!pYls@Zdq6Op{e#
zzI-Vg{azazj1zI-w8bt$HOCGa7;J$H7fQ@z2&YF+o@5^$y06M>W?_*~Sg5aW#2xnN
zA0N~UzT&65hwbjYdyikeVvXEulyg~CAmUw96Z_KA5|-u=Y8^b%ZIQoH7hR?X>uKzZ
zrr5-7Td#x4P!5@xa2vXh?%!ir7&R64FE7zr0@-bHrg(Wvc6K&|8{8Y(HEVn#BiUtS
zWF|iD&VX*w%e6d9gJViX!M=rB{_aOzg6%X9?F#gg?tuY|&?6Vvt5+~Gbv*9vHNiPv
zZBhF`9*}k2>Xm!gRKlG(uUuO(#dja<(gdCjewhZXNK0S?aQtlQ<G*m-2R0xi1A*PU
zp+(<9?c$xM8CqLoVPUaDT6&Xm=+;#fkY_?MckZNKNkgxI&&j$*F{Y3p4n;-9<g_#e
zAbu2M4e-VL_c=M%ttQ_gN|~D#SEgvK$=9dHH`FP71Gu4apoHbUejVg(XJ<Fssd5$$
zR7%pr!#X+~aKadh3`2kZbSzw)ZbL1-Rlmrr`nS|IJS1@kr~-#Gno3PF-`a@p5EEm?
zE1+PNbS+Iw5XT506qI}w1<3c@;Fm;NI=al3Gq}Te-B6rnQl`+przM|=6Hu(N_HJZk
zADAsff;`EM>{od!m7`T}Aj5%+pPieNhe!teMI=^USk`)h^74%8XaSJ2^^2}-n46n}
zohpx-cU(oq_Ury=)x1lC%+PDE$4a{%j#uA?^@0ZiuU8p6+g|Dhyrj&21;C}BzkX@x
z=mbD7h@sA|vn>2(YE6e_al3Nm9h42?Q%p`WEt4G#@gb_lyrHqN#6~>0cnqpFzyQ8c
z7;sGV4oqVUC7g^qNOX;`hY#uH8DBULYd?AV)THjQqMAl>+`p48i;A;O<=!571B2M8
z_Vo-5{wNPnX(zGE{>lxWQxVNf&j485An1}q7~=^w$++hJ{tq=Zeuv}VqUjPX;KNq`
z#_yyYy{1R?0r~I75tVcudvIcJ!)jC^4&if@6)RUdObwmhEP0W#b?a7>qOaSJ+~^ME
zdw4Z6vdQ*6bg8m_b2wzSTj9GN2K7Ph1gr9?irm}Q@CU0%crA{(xP9kMEao;*QD*QN
zyzrrLM|$4yW9?@EsPH>z#x8A{UAs_skeR`GEj1Gb4SI+F$69%7r=A|6YyZ3{uV=i~
z=g)fC>mBXw-H>za?d@l0XaAmQoQBtz22H}L#{i5v-TKxO6c2lVi$a{#{QP_!$kgGh
zsfj8B|Mem1fI6=)r8HlDkKKoIt((pU&856*Z}Cn1X*^IH+bfP}0DfjydwaXh^ai*%
zAZjJ&gel2fqkn*}a$0E5xWkAG_`!LlXsbW{mUpDw>mCfb3?SS8A&f)Rj8-Xx)9=+Q
zv9z;U6;CZPVq!Ke&Ayj2DGOk-J@BBbvGMds?AisRmK)g5_}V5S!UT!0_YY2DbAjt#
zbh5R8B52Uqz*vYKyLQo?d9Kf7P1_ZKqMMs5f!kh3q`xay4nr4(tW{pV3bK|!#2OG2
zh+B|<!I^!w9b%-;+zk8*CZa4aPxUe@t2}TJ<w0FnHwXwcv${0o$JCXWGMANzRN!2P
zDD7C%Lj~~h@iCuC&=Md8AD-8aZQCeRR8*uSlh478e?>HAKlmd~0TE8lP4<>^9Q2qy
zVYK&OJ%9ZMY_4#+)jByf)wj~j2ksk)1Q-t!$QZk!8x?x?J21C$Tcw*E*W?xlPm1rb
z7GztCtk4a#x!{?(+1c)?CL{YV#?ST<c=2O76*UUPdV0L8Yv<YII!j?Bgow~mQc?o2
zhr$apmj)Yd&yoNw#kAsPiD{_p>N97WrVmm>+XT@mB|JCjq<^`IJa(Uue%@8^n;lTX
ze6~G$_G}HJCh(8M@S&b)rZlHEc61m#ps4YN$AdqRj0yTe_waBq{M)w>pRmZ}VC?Ni
zrJ?n?SUO8Bf5^-0ASex?9cYqfW@gPh<Eh?ubZmik(QBL9C@|e^yI_&6TW;h6&3q?w
z;A|6PaHEx?v-5rwZs-z7F@bPDs<VVlQSSZCQ3|RpIav4a2LS<rm@(C<*U#n;JS_cQ
z6QiN8zYb*<fV~H4j&ek7bVp@({#|`XHlJ<8cy~h#CBA96e7F!H9_S>f!QCj*@9yun
zCRVfE46$!vYi16rpN3qQZQaxh3$WJ{1<!3Pf8f9rwKav%Y;=dRPjbsQ7w2ar28&wZ
zy%D^UpcLYR-8TNIwzlHW)UYkD=C9ws%aH5|M6Gyby&%CIUirD$+nYeY?zh*-nItq9
zg;(qJ>A>@Rz;=M$5b{@{KwkIvH-)|ug-n*dzP{c393)awPyoEQ?ccuvq_}4KmF+E5
zyw|Sdx6le7t3(PBT@?avIWxzBST$7zNIE;98Q^D0Nw>2TTC--2x|Y_?uYIeASi0A@
z(fR;$`C!|lV+02p=j#ip&GLp<s-HTwL+u$&17F+;UKWT@kLm{r|GgU#VLaBJ|G*>U
zU-|U|9DXzReG>scy=EOg+X&W!e1v6qw~|5T>W1Cg1P>ls+}}ag;8-|s%gAmXo>kUg
z(=vC7%)VX8&>io;l{te-dLKZ`hYN#KS4Lpc7B5c6oM6z0st3ijx6^%GQC<D^i7&M#
zncZ;lD<I%tSCqwXW8qbO3;!C{N86&L-A>RF*xnVW^5rbCsj0j54Gi#e>R8NZ)zye*
zOq)j>-AV2nmWpx120x+;|Nc=&yLPR$W_^wbWXqVCU{vd1@21^;pP{oI9`7t7dBDub
z7YOv=<zf~%p|u6yy~|4aq@SN3p@P5&(8=$Q%8m>Tt$}h#YRFAr!wmK=dw>ML!^WqI
z&j>}Lh{r#AbZ8_cT{{&_AEzEB0ad>KEntEtNlB&k^(<|<_R3&ih8~mU^fMbt_5g6-
z5{T#E(yS>_Y6v1h(+A^zS6Yfsjcnx3o6xEWAhWWvqBMl2ErNdDl=od7I1kT&hJN$g
zijYm^wP$coJ{_F5@gMSN{7Sl9k||WnAD_3|biLUE^SgHu#St*N9E`bVc(~210NM~P
zgB|iJ<M=NC5%7<Y?(Xhs+9E~V^V$bezS#IXL_{(x`*7-gSFe8u@=270f0`Y)kPtoJ
z$%ke6P6Ow}IH?-qE`@H9zJYg;PpUU2?ErxfxVsGKasB;fJ9&7vquhgNv7an#O3+Lp
zYQk@D8Y2A19rXO1t|t+r00W+1K;SR{r>Lmt?S~I<;TmIRo+YAaNcQNKmbtFa&^Yok
z7qUBGr}!}wIENuhB;NDm(gBFRz_&p~q0R@k0Q;k&kb8|XMdS=3b_Z?P4N1(gD)V0e
zSjrdf6QN>~j{N~{u{`0G!sp&n@Iu6zMhuhaVdYQKLqi!1uMG119%@u+4~2Ip(dFTr
zpd1=B0|?bq&TgcTYuC}bm^Ttg*V(285DK)?4HnG5d-oB5O(KWVvGBIP@Nb*iEagZ|
zqF3viE~8gVJIh->p1z1&rN{heE=>2Fd~fVjx7i;5x+3=(-ZuR<>)etj$HFI&hd2Tx
zpOceQV_F>j?3n;bcFs?aW^KIpWg8SS7~Q3~GcgS`+5!>iMv^xexahG-od#-qa3zx7
zOEOTOuKo$1AYK4CGwe;EVz<ke19>)c-gy@?_va5R81;o;Q&YC0G8JZ`+ht`9My6u!
z+;LQ?)Q>pMbI+DoZ=h%R#x5FOi6jhO#6dF`hMSg_)}zA0&465gy^GVTvtJt;h$4)x
zj+^NNbsgmk*+U!&bv?b{g#`~NLWdGH<6pn#G|$=?7#J8O?J6`@G-F7j5kL_iY-3se
z7N39v3HWJ}b%A97?swfV!(U%W8S9Fd6S1h>j3({@(AO)DE)BpFT_Yoi6RU#+T}}2P
z`r>56sgo^t&s>5b-$_8n0x~Tdtk0}C1ggu-%>2TA#;)*Jt#s~!u~qV_%UR9=Kc-P6
zO{?w+O1g|h)hr@`LiFy1#l@J8=lO;OfVVn%7q`Q>L#xu$Kb4bC1P<^tLWg1+7+s+1
z5UB&o1MX#l;`}LOkM>)!LQen+yGG6vFYjuk98SFP2ZX@+?in3jkLc~X&612jU+}Nw
z070=*_eZ(_fKW%VJOr9+YHHpu^Az+ud<~%M-KS3sgmfZbjR3C%gv8SF@^81Sx<LNm
z0quZh0D<g7b+rN-J_zwS@tATmdzgKAPY6lnrei=f5S0j?y|0eLE$uS$mQY?s#;XXT
z7{j}B)vEfNmksz(XMppQs+~^YE|?pv5lG>rWCDs30TSlpVG?z$L3}LJvW`1bhMV$0
z&U<T|=X7*)-ErRVzkT}du2T<O4DZs1cY8vTDW@LLvoC5iZ);D%elGkolnM^%)*q%U
zjGGF^4tc42G`C9$mro+d<Og!Imo4Hv*&zK}eoGE~5V?(35;<+)9-I2A9I)IdIq!og
zruSv3boQ_jr>6OhXF*NqUW8Z5NCT;}Y~@+nREegIFjfGpu>HphZ(P5AZm1yz5^8dD
z@RSm&9<dgO2JYRx8xZOWR~S%-fA3yplz-ITZIM+yuqVI%woN`W{OucY0s|`z&{AS!
zLxS1&_V$(sV=b(5a_v{5<{$L@vu)e9Lwou@qgsd$)1|8O<DVF?Emk6;wqbm)pkO%|
zAxQ+4y?;;m`pCDo&nqQy?|_?rG-YrBc#-lzBr%*62L}guPoy*wa<Orr`MK}a(NKJ?
z!uz&8p6=rxCE@bNw@9jd78QeVG!##;FL_8bO(Z`{++5UDAv-4$(+yMJ%9PQPuDu(f
z)@ud@E`h^9)xAU$S65dr;5wzu&2->su6m;8uIX;83rBOaYGPXpFYhBk^@6e2Yxfo}
zysR-r6(;LWAi|%a-Oxg^%jacrTMo{4uY_E{0oZ9j`ob1_va4<y`|RPvhm;|Ivo1A*
z>e@p|_44uRV$!Ug8^Ng^M9EHwIri~D5Yzy~n>wG;r-Qp8R~Uc~dk8;SU;u0bS<J%8
zWD-g@X;%=RW1JB`DmPvn<pH9Vb{%^)cbX-6xmXGsS-aCG*{0F3_D4{5Q&LipdM-i4
z2YRDR{mf^O)$QB2lLJT6(r7hMhWZBvR=yCdJC~``jboXv9Sq8_>Ff7Kt8B1w`HL5Y
z(a%D0WA$f>GTCwCfq0{$I3CDc)?C=Sfm+^U8Au+68zCBq#wFCzX0`0O7Ys3zDDyZ0
zFV2<tsGVX78n}K1$nAm0r;XX~rI4{pYMi-TusN|~=*QTrnkzd{We|U1q%QeRbHW0L
z4ZH3J;oBO4wO0oO;RdZbp7>=~hAAg+?W(H~Eeb-*vS?tXJG?w^Q7?Q!R9)^}Ry=EL
z7U%)eR7$|vfOoB<msiQZMf^Sp(+Xropmm!#Ho>PQGXv;4lc33gwimE<Q3*^S^7ne7
zL`)7Q#c>}zXjo_uMbi!WU)$pn`k_flNzE6M1=~2_$(S!iPJc1JGI3yqJY$&=9d>1X
zzXR5eY*8`*;@D5S6<u$dJ}f|e*?a19{1GUD;V14Oh`srw>{NquI5or=ASGM_myO!=
zmc%AIBw2|qgT;-RKP|9<m{sIv)=bkwAV;r6VF-$9u;u5wC=0%a?1ssg{pf`7&&7uc
z+6QAl1g)P~k4^(>4x+Cr=m<~)QYJbDs0r9F31PtNT1~&1R!**%*ZHzEC(tOAT{uSP
zfPjyxsS%v%urTWY9!0!4uycqvKyl^e<^25o#O9h?hq(3zLh8WagrR&Bmj&nBxz+fu
z@}tmj`RpDx(%~g9c2lsP#Dn!Mp+HVT)kmDdwDr{mQmJN-<N?m!6Ed*c7j_v4@oU;y
z(_aZA0Ba)d(>Ar2e=bX1^2zO~vCnvDh*sO{cct&_Jm)5iEVzt!hM=5b=E;l(IFH26
zNHo7QseW*x*)){IBAQM-`S$(0$>N_WIPe?B311<nr^J^g_j0?0roR5o(YD+^G%WF=
zH!zXO2#EVwzVb*Q@p#~lfYC#iA;u)gqD1;6@GTQs#nR&Z4t{=G>>r5}mh5PRKpVt-
z!~P*h2v8aOf7)=>QbtAIUxQr<G8s__K+F91c)<!J7Uk{RjAYQ_eJIy^Tu@ka^e*0K
zo*o|CB96l)#Q$q(YC<hu6I{-%HtmAc$!6Irta!`WQm!S^tTH`TYVdvS{9FFG-Lb-@
zd~qH=V7Nz+N`@+mqzf`)7txqujWKTA7#Qchbhf<$9TYTxUHbg~OP#GRjD6iaQ#4a1
zlH?ZY5U<Fsa&Xq7kLk9RdGMrY$VO~I6FD8H4kKDlUJ>YJ1?dCLSr!b4p1@;ViGz3j
z#tjop8gL+P?(W9ap`fq7`OR$a+L(BkBRAHSz>0V-ZB}ZP-HRM!_FOq51=B@BGSkhb
zy|+!>5&o(;DYtkXn6LFWKRjZ>1;IkM0wi^lmV(R+_5)H(dSYU&P;Bwpd*d=P#!tA{
zAqNi`C1C%BwQJTOSz_}hhI42VNpV8dk%7B>`Lg)r`-%!~C_*rEz@;k!n08<dF3fEG
zl6xeft{ZZ637mmU?`1DSKFIFym;2V=|LF1K6*yJM_WtNBk_KvUKqCU;gHH`+dJa%}
z79Ext3aYUOHbCN9;JKg7lBryPuy())L#$DzRil8Jx%uPg&*_gHJBAqyq_{3RIAFMd
z8X}G@uY71gz*~I0@?k`51C|%NmdgR5%AJX@*BiVy$8Cj!g-Il}J;Dp&t)?OFkvx2S
zMiO<{kQ!QA*CTeH37Wx#S-S0tBBdi=ekgH0so+w8%T~qy19h#}(b#zVnzd^kFkW)q
z&#wnUw0}+7wWcf>;1Js+=9{7XjnAIBRZ+P)TJ`&@eWg|39C65#4lb=0h<K*Rayfeg
zC^Oz5uy_F}K`54cbY4)vMf1i&cVJm(F^d&Oq=bRN!9W5_aGx*KIjQVqp4{1}3n_&R
zSS?JHGMJ3E-Wb0Frvr)>2_JdSRT-%p7<8D;TNk@RDlb#8LtDWNgf1%y2I6xua&d(s
zx(?D({#4DVZ$LvJf~c@qx77?=*I^=I(piEt1hVC3qA8<kiBQ?1eU7zDJ&<;#htdue
zgFsG9ysegy$kp>`98rc=CJ<qQ<T;0g-R9oj-U&p6X|PTRL@PtQf!2awqDs*Ug1cY>
z9U2n@f~cJudU|W+881Q)T^NvCCJ<Ky42ftUh)q~)$nJ9Qm5Sf`iEhilgU_v-vpa~f
zXU2e>7KRQOAU3bYodxv+9)q)oHC(q*l-ep=w~(G@s7er(P1)wu1~xVmBGgt_vo>wS
zRkI1{|G1rSsLj_vw+njxtHE8a>BwX0wRT~zjEg;CyAUv^#Pzg-jv=+NwU;^ri37w{
zLP7IOpmw7d0Ffa6R6=S1RxwRCdkyBnhEB|6yN(+^IOzGh<vnpxpupwwKfPyp^oi6y
z3pp=};GdrOzfgX0rb`)N^x0)W2;iUrN6^7yIx28>k?4Q-^h6A(Zuuw__3QX|PKd_2
ztiFfiEhDz)BD0P1Hw#N71XdDqR0*I!u)RpA0d^5H3yUppNWrwUQfebNbVPuCT6%gm
z>wz3I>O-ZK63ESg;{ZRtoO^!n=uYuZ1LNbNsFN7L<Q5bA{W@+Xlu?q#=`9OHs52lW
zq#SS#QTrXd?yKYvq2l1azVKSSLTXV%vId6Po|LL8eHAt?MO_y`I14a^d6T%!RVb<8
zP*DC!@((u&S~#Xt5|Wc?<2-)Sc`eOZz>Odl1soJKI@}r}rVvU)5+>*&S-+S%H-Dm&
z;iF8;^1WK6Y@166b#^YXq0m-{?1~hJqoXKTCww?KHbACm2xZ4=;jR)3767u!oNM@N
zI?I4_Sfv@ESXkPMuu+Nb&mFr->ZS3F(RN|s7X}aEq<svzAty0b=de3`_4-ZXX_cOc
z%tlE%aDymK0u_+Vdft_(A?I*yh1&0^W;)qc5t1v@6u3!goAj4k=y@&oy#Jpg04jCI
zEso6o>@~az*Z7r6dlNG?6`ZFM%&jru?19mORYEgEUB{u_A0wtI5JBKLwBfR1kguHC
zCuhL(v9a}EU!ns&1Fp>PQGE%nfa1rvY?ox1c0BPN*r$lghz{H&5&=PXL%i#z+X{AS
zD)zmGW#z%_llY7!kY+J`2z`79BP(k^BcC%M2tbD^46lHopm)$aw?`Z&k*M}|eGg26
zf>6Jc9xwET>h@RCL`MlB6E9HUPM<yv$Z`V#Su{b+e}xe10@542si-4`s};;ri`vo&
zC>d46TrUi3Y;Q1jFmO}?1h0FOn~*eY?XO2!B14&gL6qe2fGLtXz-E7Wb^zqyMfAmx
z`xcSZ<H!NTtaDz=z`*XmR8d}TAR)B_$)3M_*0ZW;WK646xCKc|L)R@uh76AmZfw_`
zVsW1uB7)p@v?KHci4zkCe^u{`l67~<3WWsG+Q3(|1y=Wi&z0~%j%~C^sIx@8?fw?-
z@pW1P5zR3M8BJX;Tu)hJ1_e0341lX_oLNFbf=r_#rfG84*E;utWRKV_XUqt@^#<}m
z(X<<0z5!f?tV!@tr}!;25gJ0M5e2|pnCZUX)XB>p5OM;$oG_7x7OZsW(AC){TsUvn
zzV(%3)=(=UXcH;+AAyiZh<bfrgH<*$i&n4aybc{M9v@RYupAB<th)e__S#`~PR=0A
zILJ(vu^c^m^es}BOq(`YXJy=oDn^_a8JZn1hR`zPb#!$6e+cf1tRm5F0|VPr?;e*x
zKR{sP8i_>_wb0z07m4Uh<I>goZ#*`Vaa1_Z#9=)L!+&<V&7R2QFp+R*i3~%MV`wDg
zEjMqz>CffGfzrN(yRHgJct~X_TCF29q5CXDoqURp=So`rZP>}ODA8m@qpq&5cN75^
zP|m~LVTXhw6YS|J9$)__3()n3ky^gcYY>M4Ky?iOEM=<c4rX|EK}#uHq~{Q(KrZ+2
z-+Q)6qd8rnG#k0NW_5bZi;mycEdwT@K$gDL@#i(ByD6WCj$Bx|Vg=)-O~Ej?W1eG9
z(4qAGrq}Pfo?BO`bA(h#x5g#l=z$3#&=NAd`+m)L7xyi{ugD%;UU0&h(k|Q!$PZ)i
zT2hilmIEY%b0|*Y`@NtIP4<VM3@REJ-{}QRP&$eNX>1(dyn8<kB^(ndx^|*e(TC#?
zVUU_}^X5HYYu3X9NBA&ioqLYX9`S_LoDWkG%!N=xI5%pg=)QgSv!z-q!61H*cX5F^
z5=IE^*cRnE^2G%iy4XrHt2@;B8_>1N5ZRpsoa%#uDCs`E2U4mji~HHL8_`qAU?zN8
zfZlamWjFxI-XwNbrKj{no!~tcK9eAS<irmV(~9o&+|I+O>C`>TnXccBHMGZ<y=&I(
zOM4k~HQ)oAfY0|gW>!zNg%b1i?&m*^R*9awpm9p2rdn)&<nekf$75Bm#8|3$YIru3
zua{?Jyhat~e9q0wd!o-H_E5b^+^^R{87j4#AD(wJul4L4@Ep_YfpqhGq$M2sF+mud
z;x-?UDsL(FmP2fTM^-i##SU>i4BaNF#U9WvbPI#%+dDq~=_=Jq1d-$~ciM^cN2v#>
zB^^oBj2>M)(-5V;RjV=YG<O(O*}x}hmDSZ|7??jCg&D7I$kkB8Aprwk;w+e2TkrmE
z1p;D$qG#ge{_W$2dwb5BXx`0%e-DNDB-=wTT7eTmH;N0I1a{oIbqf+L1aovtD3}|D
zPu-o*flrxuWem=n!+Pexe8=Ihf<#V7LsY=^cW}tKT?{l@2zvYi7=Q>y+-f-&>Pcjr
zuF3c;MB*wGGbyL<>mU;Ecc?Nye;%iG_X97uih6o_y4hBW#a@e+4as$KPvy{hck=Sy
zxqbU>O^iffXeghM&@Jd-G>$Bo&w!L<dMBf)qoe4H<l&VoSJJIpX9fod5d9)T)M{}D
zYal=PJ+czxl!7J;Dcl0mEouf>&|xIL5J`W=I@VcS|44r2*XIURu(mK{uBgF^ZL?9_
zh6A800YVpm6w>I&(LLvI`lT?6mu=O^&BNmdEC`h_`BV};3etHnSTU=H0MsW)j0SbN
zd|^@RPj9z_J;EHQz`aYwuCuTbff!XNq3V*$YO1PG=vVQxx?`Ny!P$9($vAvV1)+||
zbANEZ{4^?BTk+t4in~!2T8`cVA#aC!dNk@HCqg0CA*@-6V%+@l{92^#MJ~%qOUGcy
zj)}j0+CaCx@bW3z_Y8U;j&P?4q4~T(6nJ|OP@vMS&Bu@#K0e}&+oheZTaK2f7*kXv
z3;va)7|eFwG33I93ji&5KyDp+-uao_I+IX$6y`eGRdaK5H=Z{U8@mO!Vc=V&QAK_I
zso(oK)}eJ{4bTU9ckMd1fEi>qya<D{EhIIlq<xs{N30eIH}&p{A6*(LS`E()Vvvul
zhcXApK@yw5dvS`LY=jp2xI1^=p(7IF4v*Q}+Z(_Hn62Xd`+@kMk0W^w#NI-nE)<Hv
z!9PECKsf{uAP=~};~TC4o0-_#hvMU)y+6SxATl+qNSJ(hAOk=hHm4676N)ntpa~ho
zRv$O|rUdZEpnN(&kB8i*u%jP?KxKbHgbDg>!^Xyjy+oVYgfUSFkvGE>cWnrE;WD?d
z$htCR74J-i13{jpA#lp9;HAa|YQx#$h3tt>r<1v!+F$qys6_I@(5@#YCc2SK!ng=H
zOf?Oc2XSz*kb6Z=jzJ*$2s4z9j?Neo#N);3(Keg5oDjZ;ZYuzY!b*ks`23;jW#Kkb
z?u*+VhvZi^pu}_Mco~)oydhnu(aFWd9G)hQ1?zAs{~D6VhCm|P!rlJt^ls1+?IEen
z;o;#@Sl8m^C24HdB8#9EP>c*S-j|e6p`*k96&DwOW#6?G^%kiB6)P(N-1Q$*f84)+
zS>&^^wN1e;hnPt2>9=prpSW#LGf7KJYdop36muMH-M@XBZS>OBca6Q^_)c>_&lA-$
zBxF6{<4#G*sP67_nL+wtWF!KwZ(eg>5jrzopHN`^Vy{@h?4ZMh|5iVgjX|c&kECg{
zbwjeU_F>wCj7eheWoPdvSC3p;ADrnv+#w))?H`NBb~eU7W6VX`h1|cp;^gBaG?0Dt
z%(w3DxA2+)-B+#Ydw9TkC}7JC#JljghqD|S7_B^u5n4vy_!oAS7<~vz<hd&&De$d$
zr|KR>Y!-5?Dew>YKnC6{!UL_vwxz-Zl2B863c?$GkjsT8_ah-fHe*^^T2Rq6vK=oF
z#Dn<|E$#Z;@}!Ci$s5neTA)ux1IRLZFRx7u^UdXF$|*TF`$@6wnShIn3&hxP8IkYT
zLqbA66kowh`wNIpKjIHM46zIcmf04OF<f(gnZhtu0}P8qX##RM_gj&<Z(YPrtePUC
zqbqR)h(ZrxkDLAZV}1Z|A|(MCgQ&sf7BYIHYGU%(?SuxaRpT>1fB*NzjuEH}uwWfX
zHnwynHbReMc5d$8Gd*AG>NNU0RgWE;J1OP1yc<?G$nGr=4o8%^c`u2^w6oz5mH>HF
z5rJPmHMs-Z$=7!!K-)}piMt%a6Q3R{1`}pzDW;jKEjaWwy$Tw@;Meq+%C1uxO<BNK
znO|(hIP+05aN4miCS@?xvCpPi80)t@<tfYWsLEefTT8!oZLq&f(;4=UrmD$!+(!!6
zuP)7q-mlN9JKninFg<^sC&&MLfB)T<%L2P$$;PTFtEs8^-RjtI4WR+2Z0~0Y>IQXI
z*{I@KUeoa}yxw(W3s<kgpbe{{NR@?AQhi{97veiSY)7Au>FDV#4_kCuiv2{1`w5~O
zoCBgmCUU^Yln`uz9EW2Tpp%4D0Pmbj>AQD?{J1RzxqL_5G@o^BY>72wQ9VawwN|x{
zSE9oN{^O|yZ%Bg6&E8-g-0GcgJf)F-7fMyxtpp{)$a87~Q0D+RBT6ZJuff*r-H3#F
z{n+^#t(V+w*cpz{M4^5mZcx(E5sQVdOF6{{)S%O2w=eLCj0nd%4=BaFdg2f_OOrJc
zg3^a}1(A##_@^+_2HjVdB`_>i%Cq&<)epJ3NusNL17Ct(bO)Ago#jqGK4t0;^z`(S
z2M*w1U=o)asGXN337f3@+qcub-0DpC?%iuhJIjKoA8IOY1TQAuS?7WuJb2JQBTv`d
z*0$C)<1s%5<Ss%Qf${ksz4wHsIbNEDjg1JA>Uomv*e~kpj8N3Ry><lRv4Fod3-t-q
zs~1KC?h$6s4`IZoa^LvKNHv~gPyt{)H<X%zA`V6PoG=F@v+^%rm{EL4Lxs_Fq3QV+
zK>J<J1FscK#<UXivIS0YR@`vc5WDqE>zSzR$@7UnmTEe7C2FVEOpn`-eLBc$w@c#v
z3N`FHMUCpEar@eWj-{HcxQ7pGp&XKYnsCG`%Q_zzen>PZB7z5Rj|5;F()HNWwX1e#
z6mnnJ^DTLc=VXwu$yNS0NN$QjNK=_vJzmn`z5LQ+X*pytNi126z7rEMS#UiceO@(v
zQi{61zTVGw98p+ShQR)~F#66f@(0|iYSSaa!b)&n@;v5VS`H%v1ZE_}YXi9vJ)yKE
z%R(^*t(Sz3({;1|G#$2t*4~t1gy#VH0MR;P`T>$saQY=(UESXP{_8l1Fs~2w%$$Nx
zSn2+_(TbFBgc&8Jq<Th1ZsPB2Yin_@dq+n0w5F%0r{27(7rnV~ru3XJallY9q@Z`<
zF*P$Isf|{2w6w6jqp<COY&?yUro5de*hn~@#8<)(TLBI|hAI!8DzP~)Dm=UtvE`~a
zp7Uk%5AjeOR<DD>&s%D1Zy|T`1*(JS<>v&tUYfWc>3gOk{adMb$k}oKtgI|Xsi2^e
zD3$}U^zQENzpoUvNHzE8osRmVl2jl4O{M~Lk5#v}x_XZjyV3cwx;mzWy4j3V?@pf|
z85y}=$15mUGqSlI5E9`8Ad9UZOSm>|ssiBJQ*l1xxMHM`!R+@Kd-MVxpe2~xIBJuS
z<r^C#(0w#B3}q?cR66JgM|&ur3okoi7GJRU{m-|RmB(n?&TljYbL{@{L&yAm@xbM(
z#hWwT@cnw1d2EZ8q+RWk7Kby6Klb0>$$V@s94#LE2*E^}HNF#=@A~q_(>>fP=rNfo
zByJjW)PCF6yYR5*;@_zM`uP(*<vWOreu2x01(oB+q57vsS*GB@JN(1AJ-kY|l|Uh|
zfgKzj5y3OLaG?@Uzo<nub#Qgf5c7Ley&m&EKga1g#R8D)N8ao$*a9?6iRku$Z{Gmr
zR^h)9>jlk*ch0@)xY<xk79UC(5lov~TbY5rA4XzGK62Norw9u3{n0ynmafMIunkVB
zDNg*$4*b}C5LHZs)I!VKhzQ=wkWA*l{-qE9Lpr1;kRy6&Qobfkvjjg3;pQiLG&4Ip
zk-mZ+!|FZ+>~f*?)h=Y_n7$_=xRMDJB_da6DYlM7_$nli*B_(qKU))_)LrC0H6c%Z
zyMT+@N&oWQqtA9^W@bVVKIHKfn>89qeyDl8i;(E&po&X5^j^m!rgF@iUl_l#ZW`mF
z34$Pq2~7g}A(R391-0h^mV!tGSv_(!wYNWsM8k>uVkNMa(DcZ3omE3}3GNla)L3s5
zVcZvzqC}c+xhYpk^CRk8{cC4y-CVnn$NcTDorB?ol4Ih%R9M+1w_}GdCQpMeZHFd>
zr*;v3f%wjGXZJJ6uA*rhgX-#>J<GIV1JC9!2*Me|?R#a___5y=N;H`QK_1{7##9KW
zwzWM0+Z7rKkUqR2KD7tkz=wz=1>x}AYu&H1^LxO}|DaVZ3syyA1n}^3M_yfabDIDS
z7q#z{v1`whM~Se?zE7T31KnWtI*zydKFRnVPsDsfXvijs`y;vl&?fJ$(>;!Q_2g2D
z*a~};ma=;-$fC}FPoh?J+@*be>MQ`25Yp@BpaX|>oGxAJ$J1w?fSK|xsy&pc#FL<m
zU&adU+<6^K*XbH;1K5b?&wK}_?L}qgZS#jug33)89A<Tc?9pGz>I0pUh$k*Xh<9t!
z)2D1Id`MvYXv^ITHDzB_q2cNE82m+;6^ZRjW9`wH<_0BGJm3#>8S}&t5)X37*WfE{
zOxIRHKf%5tnM-6ampufb&hhNt9Sl~(!ohJAK~Bz)8)#!hVZ((hLV-|ys0fsRJ9DAs
zB@gVCofs4Ej}YZp2Rqsnrwi>@=anVXNtsFNy@eA#Edz#kk1h#{{)ftG8X(hPj>nhr
zO5X!Pv24#1LGXtM8XF=&N(oR=Z$qb=K+js!GmY~OwTz_t!@KYttF*NBnm4pKujPV3
zbF1SSWug!^W=6R-Zrruh1xhDcBp#@snX`aEs>h$fB$E{*Jv}8D=~c~nhf|mZwVC)F
zb>_H8u*DxpqX$0kKyvQZR@sjzV^C$SGMljGgpYw;;!{HFSKN9G!AC-2ft^*-u13ek
z){qJfkZpdaAm#F}?}g~Wpv4PGGYXmxAWg&*nwm-kQRx5v-4THajTeGKLTbM;0#4??
z2;ag9!%lJgU8hV=G@k$Up_q9V&N09WFIFFng#`y5b3A}#n>KC2fa9^_$4ziT%WoN>
z*y6EsH5kSG1e_qk8S+nHfXxE{fG2yA2bYjdhs_E1cxhlpm0Lu@tX<%rws8O@<vq-a
zV@HoN+9JaH9>oUMH?eW*ELojZkc6**rKfP<8IC03X<<qliJIggH^1=Yx{r6)omWj{
za-07>NZQoew~(yyfUqjhd2b}sdY4I}FWN8$gk2g>Yhu9c(e}-E&^ckdHSRR(0cW1W
zE1h7Dr&+yP6p;r&>GO_-2!4=TfDWgspY58^XqEZ$WiPa3_yVQ<9N-*8?7_cB=8bA=
z{JmR9wUt7G@mu+&r6n55NF5T~-pl4k1$cn7#O&ImAOfuP^IL?JMfyJ-s`)PUi@&~~
zfQ(v-+WI0hGak`TG5gL3_(?gfMC}kIl7!NEagONS<cvOf@&t)UawQP0dG_y*z%zof
zn;iuA>>(**Yz)(HA0vR*pxP?f8R8yiX7=E<$Z*U$u7ia-x2yvd9KHX$?yt!s_Q-m;
z6c9Bbgp;Ic$lF0$<`42wOG-dKM>R@VzW)n!0Xzq$awm^(Lc$K*g(Qd~8=O_l{f?*X
z<oT5K@+FK(CYEpbeNot;%o~>Q{F!}7SwEV%rlO{HPkh|JyI6be-3W`dnORv!QXB8v
z3Ke}xj`G6ejovK$RM9Bg!_OapXTJrfKS!M;beZI74uY=#9&L-Xcz9Wf_67Page}5D
z(63J9yki6KQH7$Tk=pn`(qToBq5aP*pk;#{1rlJXg##Z=Apb$*>3iq5_HoU}k3o2u
z8vt)|cdwo?xOnLjtCYzpP0gSUPg2J)@PTra?>5N|TF7?mO)+jb9X<V1h$v##Uq`N`
zmO~-fhZKQBI|i=zV1&wj4B8zgtF}LrUycR%09UBQ)>jmHi>G8jTbrJq?%me_JdB*G
zqC>mC{4F}a=9U(S?DBwgiblu%)HzMx$!Tb>qF-JwYA49`@#VX0u;afFR&|Z-{k8u^
zO&na+tEB8Uu&w?&c*$HU6t|z*;gq@g(_@+DP9=X)(+AWyfj^obw*W(nG^whn>}v_U
zK#y(%T#sz3ut*Bp5P7muxJLmnz8P#xvcxD^f#WcmRS27a&K-dGuDYe;91fMLrYQBQ
zRU+s*$4_^+05OeR-BZN@a`gDHrsHYOYwX<x<8R31N$&s-!9FBxhng}%k0+88IH_}Z
zdXVbxmKVk=f#XvGs?uK`y_S3F_KI=um})4#G3XOPaWNA1wRoa|V#zbujDR6_qDP7P
zV%IK-?CI5a1LGh#BOL{iP^wG@E;>3g`@O;xLQ9^5sR-3`>`a;Eg<mrUr>kpfzTsj_
zF=WG=uLI36TIRAS0IWbJe~i?*u@!j4pg4FC(ieB4qCS9p;4zb|m;e^OY#4+|RW$d{
zfB^`$Plx79*MIpECFS&Vb_{Ec-Vscf52K4rxZBeYq2!0fu&?nm!cNm<72dY(YLlV+
z36EkqLTeS)6n^KK@VTj1wQ|?Qy)ByE`^Ozr))Z=XN36(Tbql8M{14P3$iMi}qiu#>
zbLUCP#KBiE!dy2<X*uYVB4;Vot_g%<1Da1JYStkzkH0xRs40pCd@u*k99Zeg$|(b!
zQb=NgeY>A|{by;5<EqD-Gkx(i(x5o-&rcAffy4NNhyaH~OVP#NXIcUez5ZN4;@)37
zl?Mn1!C}nyzcKj~r=6OJy?*J(=WFP67a=F&IV~SRF;xFxnu3ISVcm^MP2kmV_V8>`
zd)3Xw`>h(zLAi#;^bo6eCpK1amEn`OG57CVasR+^0VvvthgFqe#Z?w=LJ}cb7m&?`
z_Pj~iEw3Pt6=Ld`o6*(T8Tz=v=KAt-Km2Up*qGO+;p(a?Q{;1bZ_Pq}4vUKmg(n9l
zpscKnVPbCWJ}&D2hpjgO=Q?lS{<Se_4=qY75^W^y%Ty|&A{0uEXrUCTXjNKiLzJW@
zQAsi?5eh|0i)o=nNkoe&q*AH;UzcX?|8pGAa~!|lecUs5@%?^2@AtKw=XssiP;J#7
zO&qFKyneqH>%2?V>t@T|l?Q8*w3;~dt~A{^Lpc@ZL69iod(ggpn%__j`hvs_giz$3
ziu2_!a9VrHq3cQRB|c+S9`InAHfyGmR{rv3CjUxziK$}mN*e`%g6YB$fnUmeC|QQ9
zqNDrKqzk_>q@`uix-U0UBXpS84dxp9U4S<aB>dxn%GH}oP$oEyO?}VLq|F=)wsLV@
zt<jp#|40Ryb?4QK7rz4#EqZ==BvmTO`Ru&xmgtd|^hnM-W$oXfk>cOvvf48`g-Pzd
zRAEmvH7NOD<_`bKknfdGuZxR$415rs%EDb4bGwZgaftA|9PqGhFBQ52N{y|wOouo{
zYHLRI?Aa3>s~^G^9nZbJ(7ja5RZ>!tIq4V6(%Q=xC4odNyTTU{M;3}{jwU06)b~e;
z7Ogz%{yplw>yLWd|Ho0Uc~_84C)C@_nzV<=aw`E733m?Q=#AXmiBXmo=jNNEaI1IV
zrFkpGQEy=HQO~&6SR_x6zCX=mV`4)p?C$Uan>ROBH3lHjz_D+`(>uBsM_U&4)rd^5
zT^8hRYyq_IOo<)wE_A`+KRAg5ynYU;*HJ8)A3EFD?)+TJeFZ@zzmwp{KGlt|I5uGf
zd?}hSp-xn{AZ!)BKWLt<z5RXUxKvcDpN}|({DPMq&g}>_^Fl)E^(JaRBYFC7`pc6Z
zwLN{!U}MJ$3S^v_opua8b}2RW^`rQn)T{$<#MP3TMZ;UQ82Q~{2C=Fxxy_p=E-o0n
z)@Hl%p-XA2+Krx*va$u|nVKsz|CCbnO1<pOV<XdDFE?D;hJ3ItGTnaVq>T?#v-|#g
zp3C$qpA;0FTm0By{kJ#0?!Kpvq&h(OHe47@@I2zxF`F35JZI*(`F)DU7LVYcuuj-@
z;3{<($9p1L487T=Thn?raWF|3r<@w)tLoKttwEfQG8Mcd|7*m%#|w@`5fr%UQqEDr
zZXR;-{u>0oI!m6M|AXkg7VWEc#(vYN2quUS_4`ww`F*Bll~y=#H)7el@$)9~+g3f#
z8xP+AP=w)I_ntk+oiCv0Vy<TxK1jtr^;dP&X`Uuv9hYPib$**JT^#=Tao;83{Nl%f
z5M^yTblA>ur!!`F(~buHM5_(otCV1_#=OXMhAm<+^=HXfrtxqx*5cHV=vr%Nm~*`g
z3kzw`<f9<w@7M5X;NE#-yVu9#$d3UHmp<-J^yrcpnXV@<_27wgwE9=Te?)-r{0FD?
z1dy7{Hjv7861)M_VJjdV=ifT;I~if}(<S(B=ER_l*;5vnGj>ILudbJ8T>!8i(bzAd
zlCJ8&>DKV9)tK^yq5hKu7pQoYDsE@j(TB?)##IwV51J*T?clk-002LRL6Y_>!Jp8~
zYe#G%8`JXiFH8P09JE}<@U^!*p_rbX?D+!jp?$=b@s>|!+S%PZGkZ8l_BMyOSHyfK
z8>U50y-r<E`s_SvZHNB-`yXEW!9;AmVSWj~jg*^>+|_w#Ac&O5k8dkz8byFiY#Bxm
zUNdq7eXht!02;-ir)$@)#U&+D8Ueg(^Dizr<RTY|KYzYYbqzxmzsRtFeWPG}oavz)
zCoaCbdt|U_7O5S?vAdpLGc-)FI8vU%w^LTo!Dlc@Vz%-+q{2ZYB-<Tb1N}}|4)+2x
zGMO@^U4F&3-rss$9DFase%K_XiqR<vrfpZ%)>KS|=MdomvWpU8d2s?5E&x;~7|C0d
zo@ToNu?ONmX5S}mA*37{gffCb#BKTslHffM)(0-1Mpj5<L6}y>O$|Li7>19j!i|s#
z-%Dz{X{6rEsGZX}RqLpZGJ}&Z;Gr=C@0w*fgiiN9hbb^HFjG0%CZ?2*ZUG-eu!sww
zERf*e$GjQ|{Uy?ezI}HwD?$0RF*dd@C(&^6qrHO*y?nLK99#dN7JwF3wRWWBG5Mu>
zCHeOzf*iJdQ(rm^^NRAFG0f|Bb_M`%I!F$k-tcLvG=SUqiD(2y@H~3;>9dnx1GC-$
zb>=a;2i(M4KtGG+85QZGvjr<@&_>Yr1dr0X`aJr5%R<v$mAZZ+wWZF{hCuEA*{%)x
zgXm>LZifzhb$RF@t>ga|6hC@ZYehMWbr(I^E0W$^zHafrB}<mvp4|UZCdYw^)3_T8
zo;`fHAKB@2C|<|xIuIg}RG2sZ3vXa?`>fg-G9L9M2d`(JKDS=Zf=7h6KY$U)$BjyQ
z65JBcHz`bk-+qa(S;h_=Z>5BhHJHJpfypQ=Je*j#+Mn}zkP!ht*D$~$>c{neRy#X6
zRq_^=qi99{0O&R$$`Z>q6F_*n)od`{m#YMrRQ0+&y(rk-^>L3Tg=YzRWh328q(Y!b
zYqfDO+`RM6kkCvW8h6x6a4s!{Sm7gsrwK;*?wz?@34Ez+Y;5qvH#;y<?7=&ZUY~ly
z?a8U}+<ZN2&(NTtaU`R!$lvPXmjaMMIx781Af=&8J>7pQaiN=Wlf3usR$-bt1`&YF
zY@2sJM<?p}&!YgeE<t^}!VfUUvX?-hreY7lC!Oh*?M)adOXrb~#{uBI_iVlwP7VHm
z*#}fz&m|;8QHF!5*iK2-s;Aot#}oJ@7r7|H3IHi+>%9gog|=)6HiqsbhslvFlF$@b
z0ebYcA8zzAJ7~BAP2TsCuYj+jaU_2h!G2|4xv~ij9e(4eXel3j(Ki;CmD%Jc_6x9D
z{l8Z7tOp?TsM>!925mV+8dzTYdCBD;pY6}gx;Pd)0!lG%Pf2<CWTDNBmZ(dmLAum#
z_8Hx-A)MSqh`-Etle+krMmg~WrZv71K5T^}Wr8O$KGV9ClFh$v+m<ai#@P0V9{*o|
zctm0X0st8XgLMj7$m1SKG7+XY!aUZdn%O1BY$_bTu2nUD!Rf-VeK#lTZsCO`_FG$O
z=-QOj?ar86M7{C?m)60(|C~>M#<P2q7xQrBh8aDgn^RLz`u?V@ENq6<5r|;}U|7=9
zM)OFEp}_%EgTP-v$6z0ua-({U9({zWbM#-=n{Ao$B|UwNbhzMtd}$Gakmag?l6dAv
zMod#y*%3LcXNgTcQLPz_RN)KJMpOyLW{%Fz#yqdDkXGPRWpq~XQ^>l}QM94Pli<^N
z^!izyb2BnFqALKRw6sXn^d2KZ5BMm*xi$Q#R|GKdAJ1XokJhiNi8>w?wIg;A_jz!e
zAkCUrX>Q|S@Ez^ZI~lr<J*a}FKf!LccEPlmhP2*}pYF^Zus!w-CAc%gFA>kts-VSO
zfOfgqmNqm3z!xd`15Kqk9tY)H?|2<MUN->TOW3c7+gx3gc%ZE1ZR7NLDv-Fr^d|?C
zd>cqQLXB$%MESFHxKGl4>MWPEORpSAXnrHI%aXEMD7V+@)3+}<z4t>^wT>M-Vk@<W
zPwszZq4AM%XbB22(G~CvK&6rD`uWs3=Cl~!(qB>8VfiDsT0;Deis77kSeJ3jmi^UP
z<>zP`YX<VN8=GZsz87hrR<+mN^=r53hWj)OjD5}S^53qjwlFXr_i6IoX@(0JRXM*9
zm~?!s&9R5&Cg$eAf+0X<zxZ|H#a#+l>8dXO{N@@|v4uM%)8+rXd#6{t(lq?-;fSO6
zeoK|tMQ&7(DhIN>!P9}$?&RjSR}nh`fQOqkn_>n;waHX#I?ONPfb$)59o1xO-QdSf
zvW!)qjsE_8c_juwvXJ8L349Z&^Q!sh9Y%aDI_2^1Iy&DEr2+WKyj*6)Lck&?&|SnR
zJ1$6b{>PQgZPSJ@pU<FcH$%e*nflHDKEG%>z(kg%-umf=Lk%D1!g2LqUf0L!!H-VO
z8Rr-xOc?QLbmqC&1XM=X#=q#L)7%J?=6BoLdLJE-_5nf!Xx)EVF3f^H05x+x{j(eE
z{v3L8p{crxX;`h=?LJGLp4yEfoBXp^n&7BkrIt<qjWt}GKC_dpb}mB32BPLmeMKEa
zO9(A8chMpfEO%P30w0*d@t`0&G9-kAi*{0`QcikodGLN44c`u0)$MB!5)O;UEnvRv
z>6i^^4<eLqU%NKJ>U+@lfoHGVz3JwEI<(z`#|f_OMJgM|K-!kf%<*>)kJNj@XUL|{
zG1ImL>--Ir^BSt7M?(n|lIr;5jvw69GPF6ia+vs&?bh4@T|l}wC*E;Motmwac{^LZ
zn>}1xaeJJ2yW+<^cJn93fRSJ@Umq2vMT?Ac=m%^L(58NzmOVpt3%HhyqXfK_$O?*4
zg;i7ua*EIMHuX6>f0S|l|MAScbDjUKPi?wxoME8~QaEnm!jlL5#h&K=8_)~N-DT8p
zQ~(g-{9AsK4<-*-b?9($aF*OrV*P2XOU<$5W~ss2N)^whC{CVybKAGs%b&IE9bnat
zS3S4Kdnv<UKP>b+EBi+qPn_s)vU$QWvFMa;Nh|p*w2UndI`pY3OPH*pfW18P{Z#l;
z(mv02M!d;EG(i+4sgF!a3o{6IGIS0ON}?8}7wTvzNBI}z5?J0Hpejfn>bVDy#8zw9
zLg)eDG*wo1<P=LtJ#r)$2Ra%=sq(L6=>%A<`d_w*_`1e)E0+*f@UIIOCLFuFcmMv0
z@B-v#a0VWR@PKrU)UGf(TXbfrme;(wKe1hxxCpC*X$OHY2QMvbtUP;?ajOTHE|feA
zFUc(^NcpG#G{}+bK)$jS<;z804ibqeNkM!a$W>wUK1R<@F(Tm{gB;cBaxT*E2<Kw{
z^}XG}yYGgO1(@dn;n`!-Z;p;<zt3;(ZzCQ|`eTNp4N95qlxjb}S5Jl5-NA?tV$RW{
zN1t)HltZ#`NAQEN09ApnkB`r`VV5cXshCQFXE?k7&jz2AzbUN;SqCZgkv?`h(Vzcg
z>P(D1MDo{fwyw)<*!dttB2X=GMPQMr=A*4++A;fxj+(M18lE3zkaqQQ1RB?%59ZFD
zYy50?r0PisBbvbRhwtz?5iq|msB?QY9=S@+hkKfBZ;Tw`UfdW>TKuD}Z_CQAi41RY
z;zIlD&n_+>?)f^M<T}}NL~p+<B=k&ZEi>KqyCJwxrR`V`QTH9>3g5)EM;CO7+^cKe
zn@^gwU3I0_{2g|&Wqe6-orE9?3R3Zln09PJ3%J~i9;M2wFCC_J(l$FFb1q$nt=Iw2
z4P*2wd{yBK`U~8q+~G2~vdj%&co)di-gj5SR8wuw|M;}V#*6M>Sp0Y_QV+@yZp4e1
zFExt;h&Wgl@l@LAPZ~2ufjGo74jSr7VaE9M_DQq5M<-hybWy{Huua|bv<-EmKc^Mw
zBV<cAxTp<^zEntA<b|0bEyJf|(=M8Cw#7U_u@!KtgLkPKXu5SoQW(@Zd0POpOnn|c
z;0u9m>C+^U%7-Gr=kXeJJb*^xI_@%~<^6KHqLTFVT}gf3_(cdq%$&DPOtgJszub{E
zoq@LT&)*7xMolJuF<^ZgUVpzq=_q)b$ly>lkWSZ16>DkPmV4#|?>j#%%+B&hfP0Kc
z;xq$4Xj6(1Cd5dd;Uv*v4jZNpuXrdt+(NZUvu4dO?t|lH+KT}ZDtvwoA!qa6a%xei
z@jK}4Ba`gSPZ}EMbK&Gv@pw@qz=&@7rB-m<G}AMB;_l9E-YfO$)ytykv=4Ky_i&B!
z9opWOsZd8`ew;{({P9;lM>vWU5P^QYbppk&_}NaKn%j2Ieqsl2-aIg66*V$96XAo@
zwWEK2IcAb!q;pFJ0n0$Kx;%1Z&f{m5>-G{q0}^s9onkf5%t^Zg1(W$%X7+_BI?LQX
zqENlVjNT*mncZRZ`VkO0tyZ+8rR7xu`6)6dKzfDf^4YSb371|D6~i-@X^A;o3C|9{
z!ZTmPc^m&&xV&S176>xVKjjdR0N6vwnn34}xb!)ftR1Qcf|oB<RSomdnFK?oQ?5|1
z(fTcWbw`5}vE@ragmE(muNO_=t1a#(TA$&Emu{xVBmI?u3iL!c&y``SXOe@%;n&ja
zpiRkSVoLi7eQzIMryY@<d3_5QlIzjoUiqgXLxw2q1u+HU?caau_7rASortM5v(!f5
zQN!(NpAEjkzNnovfgbDy6F3QpkokIw@DGeT%9gAdH{Tv^YqAJ+$ZZ~r%xYPvZlO*u
zdGqE5kcaS&a(v0yh|;R)6Bwhp3LQO;Ch)I>gmr%|$KmS(scj(k!~lo@HX{1N6UZS>
zJ}G7W4wA>bB}-;sTh%9EU#tIdv0a3!pa+u{79M4=+tY5GRDqbDtp-BqIc(TLo=jUc
zwJbs}o(fun2k$4$q3GZI+X(mK@!|*wq@nvQUo*-KxmLS)r~r_~>M`pmz-7Yz<`=(H
zT?JY=aLjiwqF)qfO%o>~q+3>s>I@H^Og~&r#6z*Zpz>@AoFmd0=bP<bczCQ_S%#9V
zjQjijTX|g8;l6M#g0EmD)nk0A-vnvg6AgxY`9GrQ;FBsUj}8N=*J2ooV+LY}jH4G5
z>M*%T7MR0fq-!&8;^(~Npa}gJ06D3j^6gP~)BcTPa)CowTr+#tEb#&XpK)HFrna$O
zX_NR{$Hb7ne2<lrl;WgP5WIlj?J~sc)9R7DXVDT7%Hx||XF{2<#A#b}H=+q42SdnU
z%)UfSNG{~CQu=66O^O^fhl>qTNvn6YPm{v8?~C#<^H@Ir;*n3Ipu%>7*6iK4&ua1T
zm2C)(V#G=7dGS!(nbnnZw&>ucKWpZMWASF%?$#S48akGJ$U1-sOZ2fsA_4?sCBD2%
zt}!jm1PfJ;hC+URJ{LZqrN*#*IxRakSA>z2Z7IfK%zUkQ`#^lPP(2?QF#{<E3g10+
z=xsZ9o)zuU`0=Ahob;u|PYik?*W7Gh(7BP5&LJH;Tibx3py`X6yhm6J=0(M(R74xh
z=Woup1bq|{DKpT_u?>HEc^zzaC$PpWMhC?};eB~oUv(8!aJc#vfjmaszI!)-YD3ky
zJ3=GcH%8s3fCyw>8M6YT7Gm@5`DR~++r`@KY@q0S)U4fbzs4_fHbVu4oSiiy0jCvx
zzoEP5K%4Ut+FT7)@#r^ief5Eq-cKT0Qn<``)8z8m^qd@JV~g?1{g;nLzSzOASFaP6
z{g<(h>tK^+?S%G3{BpPa6PF4w&X?%G=yf_D3#$}f$9Vk9r+MuGuXqqMqZM+g#fulq
zIx;RST)H}6ueqYH?WkIwwBvib5dE!ov1eM}j<`3}tK!<+0r2coVM?)A=HZ~~)JydG
zIWf{V(Q`mArKOCKixxsx*N;m^#$n!(aadUeuyew(M;=VY=fHxx=dMj5S3@B&oMBF?
zCFP4U#M#-}tU2Uc8QxL&2S(m&XCBnj;Z0Gja!Ustv!fzx0kQw>`y7`o5)vW9;l#`_
zXh5h&J`R{TBPQ=29N6L2Z+piVuN%CftBR~y3258!&m{#H<S;ScUNs!HPs+vzadXl-
zzfWDex1~e1Cv?b8`^jD36!je*>H9GCl>rc&;B5BZ8mU4t7p8_%iZiY@73DJThEA+=
z=gtecAmF0&0eG2@34xfTexf-hDAAYfpem!rH2fv@vQ08U9>ER1mofR1XJtrAdM9;t
z-Aytt5d11ytFEiiLuEZG%Sub*m%SLvMEO*jV+Mri8vj_ZfU%_GGn+8Y(G7ixj3q8z
zI;S|E_ss68spN?sLx+X}s#A#xx4w>ajJn_i*S1D=hn|W6*-Ne4cr~B~_F`>a5~5oL
z%IpO&Rn)-5b83x#bj@f6mVw3)uX@>B1U-;ALJhB--Di=R=<*>Cw!9XNE=B%hmL+X(
zKqrQ&fGW^~%9MkrcB*>CoJa?q3^bWMS+pWDPe7)ligjPMOjLwG+IJZYw;a0XKNfD*
zgkQ|((o$6~6%`dBlx2w})fU-8zB;y~l^;JkflW&HN6EO4t_(S>_-|GvQ$A3zTqRGE
z@de*iIkr#QNBdKh+3~?&-%npvo^O7>j`&;ZZ8D!;aI(_!)?TP&FtF)JYU&A}xIKj`
zm%+2S%n*<YY2&Cx^=ZxHjcY0d(p}4cNTY!R1-YI=phx;Csu&JM6nDJUtf$ci-Gmyj
z(c0KvxtZ3T1N!647IxY2-JQ~I3#u!w1fo0^8;(Qzk8ze}i)}Pbt11dnhD>ifu|{2m
zwkikI*5}Gx4@TP&XP)&aH3bvyekS|sRbRqb@vB!e+$w&z?M9&OU73@VFa#PWwem`{
zns@ilI#0$#z#*c%?vK6yvK5DKPWlB~#;$Ex0k9&^6@ifMs`pPJU0%F@zniuSSTK@A
zhwt%rjziGFgQlLIhC~K{EP|~Jad>$_DNdFt0>|6xe7~g|6YpuJ4_{W~^SJD<eUo&1
zGUfv-#wV9jM$YTEZ;oi#kb<GZ9((H@0t+h9rz5N1^|L#_9Y9C9%>%*=-3*769OYS_
zzl%!zw>Bp9QyPTYGB$FR9NROB-55Ts7!|i3@nXXlYH4bfq5!4JzA7l1sK~eYT*ARp
zcw-bleiywf%Ah)Wk)>(EH9*Js%5vNpc{Gy#2x@b@g(^BvL9^wngN5ex8ypw8+VPHi
z3|XAlNv+Y8&(>K(1EmNhr~GgvAwyXWLVlL+O>N!9`cVTb7b*_XZ$rR{wD5@$=kORf
z66%@GgjJF5^4u?aS8F3*<VR7Y-+kjr(A`5)GvE@Tq@ak?jP926EQ3n9n~lk?3CCJ{
zzAxRoci$+fIjmMGRj@Px?dHju8XsBI1V+ZfN*QW#NdRq@vBbZjH5}byskqPKyl-=6
zmdUJHhZe5lwh<{vlUAXI&1=ThC;L-FDKub=fGZj)Hwi;Qb026&0f6Hs;^WOR^~OQ`
z3N*!j&lCUpO8{v(pOLm-YY&h+bAjPAncdoW#k!h02t~~;vY>EjbjYed_^~%(tv%+v
zNIDBkk#}Ecifymxwq%eQk_Q&1F36w4;}>snai?dcsZyM0rA<qvrAWmFF1YiOKOFHo
z90CYfL~Mj3Lb7Oygc{-oQJ<(9w`p>p@=bQGkX65BhyTTcx}=d>>bBqY!O4Ix8^~tf
zM;rjq6|lUMm1gMbLCvHWhvDWQczWWq;5m#Nb7%V0k<K4dNFbPA1)^4{#^W$_dkP~L
zqJv_e-65v%XL!4V?&h$zW($0MTOHMtRPNf>culRZ?djS-^E3oPqe|K~^wiD2&uOro
z;kNCEE^o*Nw2|4C+|}F$x?z3LW~O|h3A>8@+TGFG+FM0-%dra$FX|SJw&2qX=82k4
z`*O)~ck#yh=BsXVgpQD?LW6)iIS%N6y$1EkHqMqz1wFa2WHO|wEF%?+nV3tQg4&J0
zaN!#2W}bPqRMs9Zuj>YvO|vxGS3BxDNk{g{{5hcFIfiG5`H$Z)5>4XBk2dcRb;eu8
zP4A&U<L9^X{v2X#_g=xzN;Lsq#LUp`+ak|oyr&|ydiSzg+IUcmuLMaE+$#ER1ht9p
z;kI;XE}SAB1N+Rbb_=lT0fK1E)f0C#9s{;5dtr>{SiROD#USMzDvsEh=&m^LdudxM
zf!G^~S(7*+3-Z7P_7A^GH6hFxf<K&<?hvyCYJ!;`f8xH_SlmOoE#s=pq-|cdmbnf2
zo4|rOpGF2udOzV>Y^&U7{sYk(sCoUG_p%yAKk8i`n}}nfN=hA9bWy>DzOuw7H+K)R
zJ`p^_-GeM?VI&4O$=3!F1}oayyrXy-#u&#tn!K6zzPCDZBk1%&j{|MhufKC|wcG6O
zv|(1+sx7E`=Ycvim=^?*Tk`t#c1l$SWeBZ!<{>R}A`8PzN6;`A)bOMgKq}hMO4D0_
zfk8D^=G!mD&(tzw)eNIwaI>}f-D%^g_qj5fI2nKf$sXZ;o3dnjiWutmQ@VdN38k>F
z^tECX8n7pYe<8+auCC>duLLpay0&s->uY|yr;kT1EVCjejs8Ge3OwUYzMFijYIf+}
z-s-0o_$hWw95sT{TBe@RXAF}?kC)f6A%UCOyJ`@}Au(~4?W`WtlD?c4u|!s%dA6Ve
z433#hof?;)JI@4Kfu6O2>iY?Bcc+W=bT?3h6UFAgIm<Sc9#aIl52`wK7P&bNvKwvh
z%sI`zq@7^_0T%D$Yw7+%F_YM#UGl4h{2gu=KodIkECW|mnsTh9HG#FFfaUjkqk0j~
zcTV+>84uya@F%)3sp%h6s3FGn$w5w&a_i3WnjzB<?@LK<wq;`-eG9WZ*3sshb83xX
z2t_L)>H*#qD)9Tb;L-|I9`V=}oabIa$r3egg^JWGp@Ur8$t(td2+bV@7}S?IPl8c$
z@Vq!$%a$!G0&K^HGmN=j@Dr8{Nt??2DEJA(pIO{f<WoP(GfwFKsv=gPnFEan;Tvb>
z(G~WCp0wxqJOV|$clG?c8<=vXg!KjPk}QM&kI&dwhrz?p7w9ramvv`3!v-SOT;B!U
z0?B~>o8{aF{_s453;%Xz)H-uk51yR^<i1)tb=t_{{0~>=4v@7%RBN_K#<*Q+OY69s
zp(tfjG9Pj|hB9$xJL`CTU@A4`(8w}I7#}r#AG_G&{1>rp0V+h6jv$K>_l*n*gIf9_
z&5bw@)=bD~2`#ff#Zhl4g|(XHkmRwA3sZWx-Adff%pB6I*Q=x9!%7!TPD4Y02dB-B
zuMK|u`0<sVx-Ku|{MljepCiY3GcCMvA+rv!BcehT=LRAN-LO!J!g7rI6MEbIpGW!v
zg7e~MdS1`X&0c62c<`X^sYPoL4z74>6%cnWxn2%sqr~GT$zzK&=jl~=tJjauo$y)*
zW%iGZZ;jwf4_ozqC*UK1fB^R>uL1#0sNm@P1@_LmRt+?U>z2}=m6uq@Q&X=13L-=y
zc9Uf=!U_qn61;M-4>fWeIu>hvx4!*Bw8TAERE88JP8`m<FT*L<5lWb>c*S!Uw_5Fl
zMTffko=yHw3vlR=NG^~n_n9_i&o>$X%o96sJH?e2*AhY$eTPV4NiK;`a${qf&DX1J
zXJLJ+&GU=Pw^1fL8mhlRZ7&_)y?cRF)8ZdTlEmJr1-UFvHM?(9=|}R6i_r4k3Lm~7
zP{N%#Pl`gR(CCC^7>aAZe}8*2H8f|_?$eLm1r9kmCv6n*^*e=F@d3&^sN(3%B~S8N
z)N}XAzojZb!@dF)YCC!=eLS?k!j^f})R$lQ8(HO$gi=roG^n8=H0cgMzqWDUItQqL
z4=%Q_RNVpqj$+N^&%XkH0Cv+>47k|N*g{sM5?%+lX~q!pHQu3a_!w<OJFUI^PwKk6
zdZW-lL(m2nZK?870a;sVuHLq7+c{!l+d1d{`fDy#chtJ*hP2%Q5gDBm7rGj%^L5Ep
z^J|(jZ2reVuI+%EOLFRI&n5ZHk@*dfBfQqQbPt307OFvOKa_34)<Rp$H)Gmy<8@R|
zw?PWoOCNTl=9bLtP|6tpvj!ARauj644i?%iH1s1mIyi9nj(}zc=L}L?SBK-6^;hqk
zX-60Sv+YWgZ)Zu2r_A=uX<OQ_5`Oq*@#K%7s_oucYfbw}*?~$dimxFW^_K}W+oYyV
zuuK48t*kp7KfT+&tq9>x`KF6H1cX#(*ugFd$yhfKjw2x`q-Xu(5B5RyZTH;>=mrUl
z2X0P?3;QZ&^SN-eIjL+Lf9f~km<r!&RyM>l&|!NLNG!>C`-lVd0rPLG2lSc|WvP$S
z4RxHds`47{5JC>K&beeVd|8<S>h?r5MEdi#E;AtM8GsA5O&d(#Y<qd_P(D;d>@U$%
z7a-kxm%i4?Z8?!{pI4IqVT^?;$0d_oAcGrqY}+thQvUki@v&c8XePpYpR2aAp|wC$
z&HS=~+7R8I9{pq<?zLyK?#Mk~6VZ9mRVu0Gch}KXUNS{S9u3uR8z`L%4-ao$NaW8M
zU)0~Aw^S$uuQt7;>6DrShNr=nMV*AYm8gOmf`*xANS+dUn@GS5NU&-af^znklv-!`
z#&ZS?<Y_kkd3L8hLU59OpK~S8{XM2W6@w8r4djyVZ`MAj>xldgTXC5<?*BQN8!S=>
zpOgI@0<wcAz26q0!_z)2)FLf?gDp)9{ST$!v`~y*ALK*DJ&|%;hNsCQAJCMX!+4>}
zj!~Bm%$z*lo<b~fsw~Rr(PC4^+_|v5W6xVfTkY(izll1w;g;t;G4*H@txA0Fh`$~5
zbNh^wAg>fH<ShHfNkkKd`;33vM1h&SU>39R*NAc17zFmRc{<P0(F=+Nbin7aziZ&x
zDBbMX(u`-tZioORXWwe3BeJh^rAQ*^1N>3tEhmzw8nZFsK6E%}rgr8tFaptiQha96
zmfX00J)KNWK4m^i_Lz`1rbPRt1isy~Grh|u(%y?neAOvC*5MY>B%RQp`#q`cjUVC?
z$3q&)IpVR9T<N)6au6Jw+`V~ove|=!h{(}Ri{_irB5}kTu}3%9kKY==dAIpyS|cNE
zs2?Gbx;>32_4lzmKdH}0-ZD>mBKzkQFzykZA3oy#^F%I7sn3M0);E?U1(1rA6MZ1l
zNW-VT+(A#7woGxqr@7Q&6i)3fX8$|yD%QES$xAf@{mR_jYiK#-f9{vCDg9StsYO<J
z#0DGK9pa&efQ|~V1B@izrgN0ZbgFD><N@3>If~$Mei3t>owqO&z+WhZlJ2IXb8=I~
zBulm4t2WlW@4Uvk@2kmoKgW`d3J3aps;pFKz5BY~h%@6o`r~RQqwN=g28d?x0o(iP
z(yxl+LwUKLO(&Fkx5X9t^JmeiQR%ZRyn(86|F}z^)fv}{DM|^Pq#3n^@q9#A$&aph
z@p9=xdp|vkC`6q3QM0wy@%(lfzG>FwwYHrZHTUJXiq`V(N(P%nnGtZXJ5yxnm$yHp
zTy_n-T2lc&cSL#t${_y>P&mFM!?4${U)QaD-uA}x=p1)f+X!MhT@{T(B)jNjkD=w$
zdcO@@nGvGw{X%X;qM1H-0I}_Nm^*RYk`GSQqDv7qH_8OT*zqYMbB2d&h>}SL1hD<`
z=(Y3ZoaygtPu6u^Q=>84I(XFD?`tNz>AvZ{d5p4&&ZgZu=F?iYQ7ib(x6$XVrZY@+
znnt$yP0Om|sZBo0ma3gYEfAtCx@T=V)<<*m-o1rx9tGZY*Ta7F9NT*Aw`WHZX6ju$
zwB~H9i-#`NEI<6?9~l(JJ5vE9MLWdXNKx%msP3lN!ItsSvf*^G(erzA6OZRIDEP><
zlf&$CvSr-e7J8n;X8S(*7&*%DLiuRage+|?B5AS)1Wm~etaPvjkDm#8rpd$%U_j*)
z*Es9ds|9AK-bHdg=Gfi2RNTyjZ>TMJ{<g~R?em0$Ua_IuwoP0&_^5;RgKAgZ5hF*E
zsOHn9@N@Zi*FYRZ<(#R{CMvAPHqnK!=)v??3(s$w)OjFIg>W2K5N_h&g4Do2PodR_
zN(FiLL`~7{DM)h-1`LWY%Q*!yGec&gIJ(=JM`uWpBccyR!sOz#IK8Io_D805;`G|E
zrRdse{XO=slZ}_vph_!aUzE14uBqZGEt5dan0ZEF;F7(7YQ&Jl>0R>0i+9g0n96+~
zgC360BKMSWqva^jQ01WNC&)-xU~1kAb9U#sGm_`6=u)(fx;=BuMI>?5QSFy5tuVE1
z!*t5UcdcrO1g_`92V35~7MAz>#g#s;@iW~PrQ6?s#X1t}`JskeTKGi1oji3a9l=4)
zZ6m}TH=F>o6gXs{rGRQ8@|BE2rn6QKlyN!^qlLfBH1J=rKPWHv&@?&a3|=oaqZs@8
zpqJ>3Rx=>jmDh!O0nE9CP=*T?=2gEF0Vw;6M~IYzYF{*~aMn!gpdBdTsJx2W?;sYl
z;JbSP`F;rwdl?v<&51`u=ZMv;Omh&+#A26CgNU}+H|Yn&@bKoH1k3yP4`jzCXmS5#
z=+Z~>5LtMJ_dkJJlgCH5<H4J_=t|d+;G&j>Ne2Tna_8pF6ZaQjnqq!EM={{Q0U5|-
zYU(&<PsSE>aMFvQniA%i4szX`-4AeZ^<hVO-EjFX63i8Nh$Sns!Qj7M$hB|ZsMK%Q
zF}Di?KPEJIV6XAd!;jw2o;{1XrMS}FJvt!VL+=UJ8F_BYM6(}lwQK&X7kG8aF!1V+
zI@EHqkx54xWRj2`e`qyde4*h%)3gGT397i}xL^8wxc!KRa{TRuu0K0l{gb}r_QN?x
z<G;^2*}^3u`G&&=I!A<E&cG34ywi^}Jt?*+I;}oTm^e|44zS@v9zc2+=$hZhBD+1F
z)IJacWw}RkNmD)_e!^r|GbXnY74U#Rt~7>m!gfjK8cLYsNnKQcw2j&E1q>zjC_2xw
zDwu#=tc*{ccAe8nZoz%sGhJQcrk>n$E9BY;x7cQW5l;=h)9q>}@D`#Mox1+y*@Y#Y
zK!zi>p4qldi3(hRtPz#zjaW1x%mQf`hm4lk&{4j_4N3XFb?XQA_2v9<ja0XO)xaUa
z?lxEa3$W4S<P~~M^d3KbdM8!+j(Obad)@Be*L(ZdY*a5f^K^|?BPB{qofUgLm%e+=
z{j}VioaCsL*2X|3QP(FNv)FgwFC!i|=>+2^`{(9gCr>yQKo+TnB$b5&I2VFO1B(3W
zA016Rv+e~dv*$omX+7+3ASRGWzg@c(aCvt7<`dDoy0UUi+(ZAO#02q8Y84F*9P44Q
zl*5MYJt@q5WpF02UKY832!1I1I?6bj8$7JruDimJ$pO+oaKo%itApthWhE{Fz~Rj-
z(6#y0tFp#Tac<quA2N?3kRXDr?TcSb9rblH<>&U>Pu<GU5od!Bk%H$MgbMV4obeCb
zw8W6RtnPO=Tmzco^ldYAW|-^byazT%rgbgtuU?&$aDl0BKdBBbhjo`*vqG0U1%}G7
z*(6R5@>+07PJm2^kjdHVD?91hcI;>}cI>7Otri)o6O20|vgQ|N<mDwCj+@}(_jW{j
zvr^kGl#VATLW>9C2GDSlb>5!XrD~tF07~qfHJ%a;6bRtcD_#X_>f|1_1+$gqZ49qq
zMMqSEb`uUBz}&z&XJ4oF?^&U8omrysf9sMt=xyVx__+Bn2eNKZ?nj6l)=-$u6tBm4
z*c46?6{!&2^u`7peN;U*y$7(%ts|-v;8gI{wv2bIUOkNBOj7j-m@Bb4ZIDx5MKY@H
z-)Ie`7)3Ty9M@@5pnT=n(d~&G!@#R@Zj5re$)K3{-S!gcO_G<l^WOseuocMy4(jRP
z(!E)C4e~;ml#zY`4!K>TrldC}P+A>aMLhYzY4>=(t|qUhbdvdjL+_rlyJrUdf&wGQ
zmtW|sbp(+7c(SJ2wCz5nNuto0Fk$P=VgI~*Ii5~4F7<<HaK<wv@H;5>{EFVsUiB;s
zFxGk5jKI_TxJ@Y8TI&s=iNH$2B<i#?<OO}6@zxE5T*;Q;8_Ct5fNnUGReR4xMsAh@
z^2LkpIioRcuLc?BM7c%pB8dO-OVDhR`WfTO__-Mk<J(T)d&~d1Ws$-Jc*66)qYNI<
z2xAz;2jsp(KOADDEFNY=<PFr<6D%y+VEK)Z1rdQH4tmj$jM>rogAOQEfN$fV&={YA
zlNHCUOc<4BU$z4Jg^==}4&1X{L%LA#t{xrd^bEcOq?~kg06~$33N^Jqv1JclRhwY?
zpAJZ}j~~7+%cB<P_6WFlj=8_)jT|xcA<k7z`yglr6=SgJa;wJPdJQu;d^7~*TcesG
z;0ncs8J4z5&Ol8(V0;K&w6Bf2_3pg`oa3L6g4c}1QHG9BQ(i{RNI(ZzO#9&<64_~C
z^Ael-r*djyz3A~O(Gti-6@w|{Z2*R*H2t^-b|)Xpl)<v1F~XHat44ss)^P-d8S?_w
ziBCv~t(Y8m+J|~V_tfQ14G{D9{b}fgU5!k5F;CDCvSl~|KNvC=jt2z?i?4{R2R*Y!
zc$#!nW<!z9Q18daR~Mcc9;<c=AU-isDaU`Oz3V$`<}tvG&YwSjIm{y*CQGGDmpfo7
zSt{d+#n8p8o~>xl>6=T-`LM|<wfY^GOj4Vq=CfE41GgQ!c9}=bTLMz%igw^eewQD5
zT-*a@$4!TyvR9#!iU0V_ZlLBM^%hVajwM%)-^t5U<2rDef0KO^3-eERgcokVZWCA`
zikLExUeW>%&zS_eH%}7ESG4j2Jbi;d2(gIA*Qj*3${H^(E*Nz}58ej6sP&Wc@8%OT
z5Z%jAm-vJGXN|=HTCBG?xEIg#TW0?E8_$z@ZxtW+daxz=wTiCGccaF;oS!~<lsA7J
zt<0MEi@OGW7L@~ze|WuEf1Ld=aCzv~k$a+-AdMJt(nx(*VyfxM-ZB7gp~_b1b%Ues
zyn)c&Flo`EH!TKdLxqbiwrok_YKyGF-CdupB;wpdT_xNt5=V(kgnGw*3#poj0>ITT
zq!X*{J4jVct0zJS>-C@Cy?dASG{Gpl0}0+CuteQVap}AlOiCK*7bLm&B>t8$<yOj!
z)=LU5uwVw5BRgGw=ortMkELI_zL|1I&qNLHT`0H(D7a9bPi?#^&t(w98Zs|TpwRVA
z5sE{A(l}7nJ>0qHAGU4`)U@43M-7GBe|yEl)3!YuURHzKf{M8M_;E*_{=X9vRy~`r
zds>&yo$(_i1FJRI5A4ad-vetIaXWbH)~y1U1pomcpl!8!QoYD#7pa$)girhxD_p+e
z)L{ZaBvluTDjr@C4cL-3q7W712l(0#v`J#tVmWPEx3U5#+W9Ob5%893eCEzYTG3|=
z4>T|Agro$Q>@Bb&q}(iz+;3$FzF2;?HOrATp4-@%fFFj}i!JTN3)!p3mj#SByh0M=
zwPwwI#1`tqE8YcmOdWLW=*X{ka&jCIc8Smv5jVP)Uc)*uddA-uVvpq3W@=wOh6+JP
z>gABGqA_3vhDWeQ(5<4ry#{(YE5;=_SK8nYD57ibW`s@5P|w$`$t?X*Hx_rmFhgA3
zWffS)TsCIyg(dnMwMKc!DOM&6YX)CwoF8i!i?~sP;)N59LQPiUB!2q)29P|c7z>xj
zP%L340G3QS{f1b-tHou0`3mRC!b#&a5{>2?_$Ge7-rS-5@J++>-e$A;iUz(l6FR2S
zqZ7Jz^1i3j9Fh2tx$v-y+3iMjt7E_27f@}{Z?HB|ZJq{cNTgvz{i{6G{6BB%UvFM`
zJUk=DQhny0GdjvtoIN2y<?SY?#fYz?#pM1*MxT>?f~`RPP7hz5hAJlr8<H_(0dvEc
zyU0PWK$ty!`t+Ndvbbs>&2Kw-5_2}2B6P_6pj2`Yq%ot|xb(ixFUwKZILd1&PJkrT
zN5LAVI0~strc(+l>UEW}WlbZvn%%o@>IQt`s?Z!ZDJ7fq*r#-XAE(>m<LNDr7mHtv
z?5o$Av#V{FE(i@4M3q3v9Ux<|q)36M(6(JWo6@rh$W291geHPA4ICOcM?j1{BR_Yb
zN_x8_F>mgGO7Ia84Q_u3_;C$R(e-76Q^YU3*w0%Nw#NQG`(x7$KW7bU?D5nt<-O;v
z)dM-$RQL4)td#uqQ^rNzf3pndBU10y$@{_#hVE~`B&{^g2eKawlAyWB)7?GZdQ_26
zF^i3+L##hHk6upI+M_ZA)!Pjxk-(7<#rc<iKz$r^V?1aAprSxP>`!5Q&V-+lwBX5;
zC-(q0*snHHYLGTV7GLLF@yt}!)ND%|bwM;0pko;}uQ`5sF)^;O;c2=p-l{OhV#-b~
zFHGFxI%(F~X1NtG5HF(T6%WRmxXdK+g8MnM{1E<{h%ufsN$kWjA4MJoyk|w>kFB98
z^UsA6kFAy8i#m?icSB1nQ1gct&LCimRW&tA-H<=xUstX@nQjr#en(SlL>ts4$$wl_
z)QOwi6+ISbL<Z?7Z=u`8LIGtWvE%Z)3vWG#F&ukm^v|kRRZn0MSUsV#ae+y)hR0J~
zNx-}I!&Ea|h5n8v(+}F1vIIcsZ|0{+D#+|#hAqAP2iH!uT{}Ou`Osrnt?v^-55j`;
z7L8X$9gq8coigb<2T5N8F$7}tJ^I7woCK|wXIe}#2~5Bp9j%3|Zfqnz`H*QyO<D2G
z9NKTqn@0jxv8wTNviUvYF|~ZjhPv7#UhhOvLVP=OzkIJ|wEMZlMAVLBoyIX6y8{Gb
zcH%!b`MLm#4((sY>Xzi)zaIo13Ek(zF+=MkDtQnSirbiA_Y;=UOgYYtdQwa>!9HcC
zv*HoY1vr&Syj4FzWhA{GM-`;-<;#~sVT;#TRn<|dS%j9=RCDfFWU{mX^nokHV2TKI
zuOX$(bWjO=(Atoy49>Y84<=xhQBIveDP*1h*tOt7>=(P1`?Oq_Qv^|BU!y#3kl7YT
zXG=<5&Ldar9<%9%)Vbv$UBk?@h1rSk9Y~XpehMq)v@`vU10%K5sonUF@h{)8ov8@5
zP4XG1ZE0zBD}FCzX3ys0_y-88=#B(D5h*Bz13e{%{s&u)`ts)!Eq#i@-?gV6?&C7R
z;quy0s#)6qm=&XLe7o)*BSq?<DI&iRG!o8?NKO`nf@^P^!obl>NNL`prL4!*!CA}$
z2J9UEOd?yR(pWcVtLE$z8I8RpC0=%5iO-zlNi)n^HdxP}vvb|AIY<+cE!{Gqqrb~#
z8&yA=&ItVU3F44{vSNrbJgHw2432nw&MbzL_TpfAv+<_<n+IUUSjO~Fbf@Z6oCPWu
zN~L^vbxlp2)frZ?2{8*t;pXMQ4`gj<hDepU?jjL$c8+JK1ybr+?>-l<csSea$f(&5
zvU6_VmPs$r6_+E^&vyCmKKMUgT*f?}sO&fiY%&lvIt3dV5O9`jykp!9qaq{6@0)j0
zXV7O(sK^&>cmMA4m-?+<)nwn+bC9xItetJG@wcn%4iZwrXLW6p7VKqPK!}xXg2f5=
zg8=r4VIR#e*foZp9WhHM-KFvf26UL2>ZLE6nR8aH>dX)^?{99ngTJPU`E)5>sh_fH
z#1nv<cJciXq)+rKS!XZMFw*6n`X?9FB85lkuEB+^Kdr|WZRYuf0ZB)!w}I@)4xuk!
z6iUNlrH=1k9&H-fInk7RU|E8sR$ysH`G2=iWhXUqK#@kFE0e~OQ`@_5U$xQ>@K0aq
z0dL6p$AndP<muqZttWZ)dLakLeRbQMc+6;}OGxBa&4B}hPM$nzx^br%<1$0RvlFLq
z+yjvGi_Q+G507VcJsC4X%9L5w57#fytNZly!!7|(9lStiX??2n5*!r;k@quq${g(x
zBW5}MjsLVjKV(JA(u+sDikmX}D#}g;HrojVfcoD$<r3z%@~#{2zL}E)`&y{cWOvl@
zslR&BXm^%E55cVmOjS{5!`}xPFamAFr>uw&PRJ%00nk%|;+bH}!e62R+*f|J+N682
ziv~0(^lRE>TJ&)T>o*mOzQ3=Xt=~H+^7w1~3nY?;gedOoKcfq8PCIA^CiPPq8zD}U
zqYNl}#vsYQjZ^e*?^hC4-TKoX#tdL#WL4$$#YQ3yLU0&2{0<lj16w@njUWE;1BnS!
zTnq^*dvc^9nC4*jg16gEPcP=lyA>o#TtOiG8kDZc;jZ2S)wg@M+|)em-&cEeJ`gLE
z4BvsA2CWL7%c-OG6oDvWLbWv$_-B?yU|O|+eouTDa)O5v*reU0+UR!qa&VOE!}(LN
zgNQLYsA=-ajcG2Suh`kDz{lLNlR@>!%`y(QZy)j}$sy>a5Hm0Bom5M1y$cCFEUk-4
z;OTM3ODf2A3U&L{)*So!47@~KVOZU~Z@&Xto!AGtb&oIA6y5}=%>2<0u!Yi_Snlx7
zX!wvJ!k@^ht3TITiYDtX2wU#2OY(og{0!5-^Y0wG!CYvIeWO$3PNqm}i$Tn0P%B0}
z@y{Tq$)q8hSYOZ}>?`dMIU$<ktJFeW^D;TCW=@whWSmv9h08U{JkV+m$_P4e@uzT7
zY@(!ufCPQj0x~<XRh@(m1wn_l0B03PrCMcn)>gnq3`^7IyZS>5$~c*IwVT<N|Fi%x
zHe<Yt>b+;%wR2c?8a^Jwgoux0Y59MNgO27@lRUx+_9!eA#Bm#SQ+F+`s3%LXE!x>>
z@RX-LS))c~dc|aO^yi^b$4&k{)_)8`hbx;NI!8Y&hfNTXG4cTam%ZUis60fm!0K>$
zb10BjFTSYxEL2gYNMp#dKYyFDOJjHlb45khCO<=imd#;P5ZeLf1Xnzq2DnYZx8X<m
zoZ>}2_w_fy<QMNz>{hb?`}D(ZgMz{Q$!aawyWl%LeNzH8^9Cy^DWMeI48f*AGD7Ay
z!|sfOIO3vPxQi{*pKAwnp<}xG?UEW}sn@9)cLK=(6A54~k})u#cShx&!%&3*&XJYi
zC!Dn7taki&wjJYVJ_7QDx>Wi7DXAR)23aV;EW!aZZ5n?vEPgNpnH&+<XUrq?Zt73k
z<>1@2au|YL>`?zqJ?sDs!)?Sw#c!T`bGUBZzDo+I6qhV@>u%D*nno3sd?~Zm#R<@k
z1}JV==*M#w0}gQ0U-n9C@zx7=X)xA%uNO}K21$A*mY0?@Pfcr9+}(-+2}s*$tHksF
zx|;Lj6?M|~-$q41xVHBGDbbJQO=_TTqseOBzI`vFdna4sV2c<@uZcH!m&2RFKDL90
zx&@Q5Be(XXkph`T5wQ{<9VY`ZE5{^o0y;B1fPm}YcoOc#r9s^C*^1Kh^NPKZ*QThb
zsi|2%-*xhDj83tmmg{0}V$zg$INLFPd)gc<e4<PWx$EB4x8=v8ciEYk+|al*hZ1;4
z^5ex9o;>@L6|w|uU^DfhbEs|6ejj3^rB16Hd;tn@2UYN4rIQT#5M~f>(ovqnHOHd=
z*-bZ`idk7$M6pynBB9OGg!uS2laIAfyxPKdP<}e2wLv$m7FjMer#>8i?Vv+k+^hTt
zt$s0mQBhHjL5i*(dh0d;9OLKEnI-i8SyzB#F<O4XZ%gTC4{O1LYz0X3%f?g|{L)Xk
zCHcp58@Z_>Z$&#HGgMPflo0M|`NcB%UHMy_oN4*sJUjYe%~JB})kG@NxZw`<Qy{LV
zr!HLlHexE`u#;8G3!4sgi(R*B*!)W!0Y}{H;;;RT>EUZ0_t1X!bFEhfgGTRJWQG_8
z<)O}oy^9S(-(ECc!9MJk#ujc#5vTNrajR+UwVD1hwvZB8UteYxX&Zq3FjD9HQR{Yx
zszZZO^zx-u^z#cW-csg5{a+Bsx8x6@D~|r&#H+#kw7T`;_Jn@RuJr6J8j31_@0(g~
zd76!A<F%Do)4YgzpCAF3De&>Q7(u*mG#5u9nV66b+`4K^J1$`MV#KEh(}V9o&q@Ac
z&aXj^QU%HRIN5d8(`%!H8WegMr0X?GHAr9Mx^5F;4D_T&QShNdJMS(OKmUgsuJh)J
z4V3Im>vRHYtT{tb^u%$)3z<ivQj-Zrq8=08cv)7Wk74e|jGG*@ZV8H?QnWYnDo_eR
z3V>2&FvNQ=-XPs|T?2=`CHYUZ8>J>pRP>5n)32v|iUQS!sYqI2WRcNrbckaLOBy2i
zB{X^+{E1MY6*5xEVWt0?z4Nh?&%<QHPYQ9Si_%9XV-$g#Ok`9P`_`*aZs+D!B-<Uk
zC!}ohu^K8<i>Y_#4xrkdrgGcGlldW<uI_^e*X+JSezA(4vX^&;-rx$praHL_etY(u
zpulCdGopR*b0Jy;yZi{fU1mCYJ5Uo(vfb~+2g?wVYZp+r*bE(_$kqXPl7}uzclTaQ
zPG*Rzfh;}#U#kx2kdULKkyy}6go&aVt)L7v27<WJsf#P`2{B}p`5hUi!Te(N?6Npu
z9$Y~G*-14GoZ2Sr(lZnwd6&k+qteH#HW<N6pbQ(!%vh#A-7}LV=p9b|Sgw8!2oU1J
z_4}bg<QCV$@ZGjcWr~cW8KDC^{P0QcqLX0ojIU8B&z$Wu?v8WL&08pTh9Ay0oB)`K
zAglqigQu$+>|u+;{BvKoc6x5qPD@Km9MZX6cr(_w7&mXKY@BUuAu)o<1&Rhb)+tB|
zXkbErOGQ>?JA3bR2q%oxEGz<~=U={@#e510!>=JmeHeWAbh8%yd~59Ju4&C+Vq}Dg
zT^CN<Cgp)TxlS-2@fE((koLSyBY!bW_+|aw;1Ro8`Iko{BO3?ZkQ9Zmse7M3r|+qB
zH@1)k$&VVd@J`5+)gNEGE6wm)4sLYlR#9+0T=e<dX1{OAFl~xOpyXE(tBiQ?t@>h$
z=bN!6COaMt>C1&i*o6vl)|>8Prp1lG?XW$Ra+_s|iHQQUQBmHb_H~(k42&7LK!CO2
z;4N8>F9^t`R81@wYqGavYW~5DHGK&;-v2*v1$-5nQE*_O7mrauntye#vUmpel$gst
zK}z@dx8+z#IHO3UgW3L2FL1h(V%(>7pTkxwF7<uvWFC0>Ho;QHmOEH|`}&naq<ctP
zimLFWq#=N3*whQ%ODwJC)4==pxpPcMS%3|6)MLJy*D|Pmln#Ey1`B$K#JM7Jt$03X
zZ5k%77ItO11QQ@Y0|41}ag3Gh2hesHtd9xy0$N)TA9^~SK7Gc=J)CC{DqEw4(pVWr
zY7tVVfeQnb)aT<DY1ESin$aueXHYrF!cj1_R*W?D?$_@)FOV{9!m$k4cs!{O(zR5f
zzaI2ODKpS&y1M5MMH81(?8vwYkDH}(mEu#jM^OzCEhqqBCT`rm9mQoOxH7jQqt9Mt
zp$js-sr<m!#Hu;fh?Gy$QdM8qMHWy*Q%x7Y*hfmCr+T=gvv;g8Ta3;7FFq&YI}Vd+
zt4j(jQM-jsBz|TeW|d$kcG<<U1jB%6HvEa_SujpWWuO;j(K>(xyItyYH8Aq{l#qod
zN1h$;+vje#$%fb4w%8Rt-WMGhBU@ra2%gNlp&iRwXz0k^>LGD_V<G4mIGE|=GUq0K
zo*@hOYN`7iw~hFn06ERwy-MR!UGB|;r>m|hly*8Ao%gn&i>iQz!3|YKMsi@lADd?C
zie0fxXOd5K*fD?NzB}!EHIU~(F$!JLYcv{25r17~1I)vi)6TZ{DNx6su(h$7zao4d
zwXqUIgLxqflU9b_3@@}-kMV2UV)doMx7s(BEm%k1cVv_lOPkI*X>o9LGUQphMw&~w
z!>o}f_!5xtEGyC4P5mQ=MPw<;takp!ePpB=t=^~O8(*f;Azh_I5&4q>M30R8W8E37
zVtg`e6n?4k*K&IgT5OPRzsAXDhN73V{70wBK7;W%79kte0PMxgaA)Fqv=wJi$dkLr
zl%(`#onFrJ8wzpFD{~a$9z<R@&v%)y!pOE3$B>!l=I{D)nbe1^w=z+a{!qlJR7|h<
z^s-sZCI+Zkg`$M<eVZpIOG#%kW>{Ryk~yoK_pwVnf)YGDf(Bi@aqy|ryF1?+)A)k6
zVT!Df(<_`mc?{vZ_2pM++(gxEmIEr#v_*?ptL(v$IDEr4MW0XwA$!uP3&#q52j}>y
zjeT4&)2g1H=O1;Gp%GBBh`VMXF8m7fQ!Rcygonwm4C)j0F0J$w78?K($W)njkjaA<
zgG0@FxQY}}C`fT?^ER=re{5kMJ~Q%*f+L50Abx;Peue<0=N5qo2a)DS{OV}=*zbTQ
zL`qq9-%zC6)9#@Yuu^^x2Y7CU`cn$kQ@s&y$y80zVhU-h9x>0izI+1Gjdpo^)77~7
z!BT$56{+|D*a+}i<9!P{?hg*$v6B_ivd{?NF3&BIfiejUVGF~DxJTT3+x*|Ku$n!l
z9kH|mx?;n1P+(wNiv@{}&LPYs2@wb@!YBVOgzv#&p&71QoD`SdP6}^cNTnQtAYgk1
znxVc%4R!nVn~&5(R5ZWjQGSmXi&}&z{ef^IUaeJyQ1)Wc*k-)fKk?aKrNJ|FgWXOx
z`#%yx;{0!$X+y;70Rmy#;o0pG%Am~(u7DmF4-g8B9T#K<98FOto&FtVlUBov!pxmN
zzmG1XL8;aR@C>))eHC#Kb^C~!5qB4&j1lKBP?rl&Ub&toid%JZIk%MEFIa9K)$y~Q
zUy0O!vutd1Xb=ZG*tD5#Ne3GW*SiZ(e>fMH=VENq-L>yIj9a&5IbL)-_hc&ywQ~&O
zvpEM`=E_1tn{S77{UnTnBW~-mdONY~K8ig1QE5|7s4U+PU>Redp-?){h!e2l@y*H&
z>qb9#-=WdT{kMNmM=9S>;Hl;*R2=b5fmj{DBtSdV0i-}uZZ=pVhNF@78lt}v6Hidy
zjbH^Tg4VH5H;bMTEx&Q=H|%08CZ$AI7xlZYu(V0XqTmJ&9gAWwjuN(tQQ*5%z1OV4
za^PcFKX-XuLRt)6J|Wd;;osfpAH_<NriDOZ-DlhyXf44vP&O&#RFqOb4DQ>sYKF8V
zqkepxPHLi`#WA!hy-_Qx+O@j^npc`895w1ysP?JEwf2=E3l)`vO`U@uwydkEouQhg
zo82$}du?3O%D)PqZ!q>gxYhF=Boi}PqHre5^>BHo-dvHA2%Qer+~>~02!@he2V$6A
za9C+hZEfPU+|h=&o^O~P9dr3P3mmQ2uP`Ax?`>Y0ztg^H<$Fxupq~6Ix<W$~dX==`
zBf^M_mY?Z#;SiYKpq*td5%uRp;LU;aZ?zt@2`R=^o<ipIGXBf}9L_Dq9Y4d{82aDi
zdpwH0x<NcXYCDC>PoK0{E52aCE8XYwfEv+%;TzRD?vp6op08D**%Yf1(m;cKb(clq
z%P&Jd)g)(hwz%$5gFks!$kK(nuelol%mJrbcj>bJ{_Wy}>ij|%-_sdH-WEcxT~|*%
ztGIOBsf;k7QxbCy!1W~N>YSaOMbu>(P5l<J_9)jBL5Oe!_z3)EH+nL0XONe33kOsa
z1u}8wysU<nP3U`2I(kh+VYtjkJS-Q_p&K`LaJ#qPkr`41CH_Tj=x>=}v%1to^o1^y
zMbtcO#i`$3<Bg`j;7;fG(8pq5FLvmxV(sfti`Y(Psi{oG?6k|s)e5CI^&g_SV*dtI
z!kDiLqbZ^nU$N2nS3s*WdL9SQrlX<o2gq8~DAOd$0EASFUm^Sf$|oH?J&e%(iq3!!
zu>}{QQ7|RD&2V4=^Kj0ZW#4)$I9#~0YKVktpdu-&RG`+_g2-n<>m<G)baJNjp!f%g
zjSd`7E1lf=VFMTJ_Db%os+vjnTCP<(2OU{Icm#w%ssU{d9y&DZ%_uLFl73J=ADE){
zk6MNAO%K!*Fu`)NK<u)<Ql*Fx#(39rE-6vqRQL@x?hr^1C9QgeH};xK^<OX0>W*W)
z7-Wd{4TevT^EQsq1}#W|sh^uLvyaq>+;UO<e2CldWAfr0J(GK%$RmL+kAT!jFXT=p
zcV&hH2uF?rhl-jAhz8jHx8}`X((4TwKD-1kMgZ5W<1PT>0#X4-{W;=w8<vH}#xA)u
zD-57|c2#w?jMzA8$rKf<IPT2p*J};XODHu=(p28E_>WS;=;j$RJktxXQ+P8iD*&il
zVT0qIOwI>dIDoqeqx}llhRX&N?8<u7XOoW$5YhjEkEx8{p;~7pp*cbz?lejeyVz@s
zjj)avmiGDcBVt>LAU7<eoaf)grILkZr);!`(deG8T5e>`*}t7+(V~#*_bgONRY%!Z
z;EnK}lTjiF2#0!%cbl(%o2t0xln}zYAtsPf=lW5I)lB(gm#UcKn#IH3(fs^3ZP&_>
zp!&OE>zlZ;7T}^W<a5vW`l#6E0W)p*`llX<AL>Q-Js0BR|9E*lK(>IP8);ZtlLt~H
zPKF@lnRO*n1!HHh^6_Q0|22N%O2B4~eR&;6cny#d@FjEB%)4jhT8VV~N8lhY*FC2*
zT%Tcz<4bK2Tdpo1+@j7UDbf~zPKk*j?-Vz}ljnH9IK^a<Oz^X*MDu%36a@uVv+A#R
zm`4kTC9~_&nM#;p=H(JNNKq|vt8BOu1J1SWFm=3Bur$ql!URPMV;1a~cP=k4ze!3$
z8`7P|p2N!mk`4psH|J4Bte<DR>>-hWV%qRP(HS`<G877-;zY(}us{Z;8SMG#F~~P9
zZ8S6Q;tUWNC^I+IGStl-pi$tv1x7+@HJ=A3o+Wfaw8}JjvOR?7cKn`ON}|Z?h?GGb
zv4(#k*`no40_8vTqoP}0tACNXw*%F8eP_|NGc$F>Ykmkj7tq-uc|fV3f$iei<`hL-
zn)|$TnN)y6efw2eGrEepT{Buy91Pe)GTy0z?NmR+rij7%T|mNZp60#r6c;F<kx*z<
zNIZDr>|fiO+G3~7;0c4#GHOwSZtfn6z||?E01Ll8I66)V{k)E19Y?(Ts8L;fGnU=2
zLF~nNKn94ro<jjZDpM9cs@Ec!6;X_trtaal;=_Qk=9woOlcX$w&KbKlZS7YNy04fO
z{pxQ`<h3#)5i9RP!H&o(bPvUJ#lVwYQ43KkbALd|fMWp$a0r&1J79U4S;wcrkp)(-
zI;PS?2HQ5X4Dw(T8SAgR8LSnQUt)Ax-H)ZxbBwBhl23*9J3TYLA($U`zw*O}D~LYi
z2sLhKgrPmqtj`DYHI$+3n3?ut)*%i;BbizdhcTKz<B%oB+7-7=d$XxQ_7G5_4ip<x
z1uO%^$QOJO1O%v5MwjtsS|ejJGnsNy;5@Rzc?*D=b<Ay6%rs=sL|ze05$S(Cd#g_$
zE8PAsk^1Z~v%5&*WXdJe6~wLh_e2_e+Eg9}C7OPy8L>{btswOjma}xGN#LoYcN(Qq
zuCDmeQ!6_E-vzQZr&@?tsPIC3#O}Mx(SA_1UU+0N5(NxsEEO>?O!i|Ei!XmI>-4kw
zNZs}ts@l_*Wue6HLFF_hGL=HcN>qE9x6m;4Lyn$Et|&FQwB%wZXXhQ?W@75;gcM3r
zAmyu6ND7y~6gLAo%`DrW2;d@PrtsGjNWgq0(K1my<<0CnG<xNYUj*_qby^x$mSlk!
z7=H+=*Ik5Cy5MQ89n_RvJ7HmA*NenJ5z>dQsFMEkrhcHNbr5~7NuDP*ICz9)9v^`y
zO#Ji?95`T_@l4$9$*kw-)CJ1Jd<LBxXa)_J%sv4?;e|1&v8ux14L`kco}2yZA$kA7
zUV*xVE7|9I^0yC#ImaD{QmlrwT6~65FhjJbefqqlk1TFb!0e$HOD(c$dG+?ao0YEQ
z^+-46AJ|%IqtMJ>iGEjjgihz79-ATAWlbnI@z<gxfMoEU^XFC1JvKv&K~coZU?xCj
zGYLU2v9Axvm~E2Wh%U?|4^j1GEIarT3SxrJvl<9S6yNC>2M|zD(!AI%@uhyF_rbh|
zL&2&1`LibNJ~#N++}*{Qir=pH+*3)$%`f=G9USDes_n~59NFzVcZ#n+Evf+fD5@#$
z9jPEVQEhy9=zpV2RG_Pdi|F3FceLHRC45Q>n}c80b4;khY|3Z-`noNQp%IWg3UBS$
z<+P46@+cAxss;ENnk<I3z(A<Xl@RAVNnARWSVx~05M!eit$1A_PV?DubT&+$-bDJ^
zKzK{QloBW&)v(hnGv|v4I6MSs4qL*8Ig^J$Rmjvgdw$sv{gNIr;<)C%ibp?~fVpvE
zKObsv)vIn5h#f>S$6ej<wGe4MeB3zx3MCf!;oiW&v8b}?)3fFdU{>F&_Fdu|l-biR
zH~UBckU@5vHGIy=Xm=EV$Yd}KY0Lk<^QRlAm(VE)&q!O+&tcW_<vl5$p`$c_a+wN2
z5`yI~Wj<9X9~fKkTM(pUc%#&cpS-jJA9yb06i2*GX)}19p*jc}V$ZOj(Fcl5HpE6m
zj#&|%=K{mC@q@|RSSPbVYCKe4k_Ri_xF&>2|18TzSDv4k+7&^_I&L4JuQL;46wHCF
z=JVmglkn(p)O>gl?C5H?@8P#*Em|y5J+gpzes>Bcg_{cUuhsJ?nhihZUxRZ<Nd9hn
z^cRz|$8Yuq=4E3&1|I+Rj|w^w1avQeBGHzi91<xkRcPJK+LN3{$U;p}6K*i-L79Lb
zI8cQ_1Ev-CApPeVi`^|1T24-m*#GdyQ_qyeF<2XG^V~39bVbV<oMkeOg<{lM_k{1L
z8WY#=8-(08_0bE=gG%xbt~fC-KMcWb`?{Lb@pbnW8lo<tw2wVgaklKDnfRekICe2=
zTZH*8=5*I3;jrLjkhk3b?O|p=ub++GjaYo(hM#L&b{(}d%dtUDMYOONkkgKC_J~#Z
z79cWW6J!v&kb}<DuK2OZUZp#=<B;9w%nuF;@%wC#*$K9Wz8zv_b|O*fuZq1bE?Vuh
zUc9voUwZa4<MA&DL}IkafO&PuIs3P(=LYWIpUog*25E=4q=i+|&mXn)NL?9DU2<e8
z5Gk|uMaQ$HkfmO+PP}nX!yxRotQSK#M^>aXO-f1e&*i&l;FPm<<~{p>qrpdxOa&&P
z0Om0@swDiE6c=9sDKaj4h|CE(=(@%@BAnYe8XW*4*#<#{!*ya|t&tEAckfmmub@g0
zOb0pq2l7-9sz@Yz44W}#xLsDv@hvF^WVnYS4m-MY$kj!WaevpdQ{TLSQk*G*IdvWk
zZbMevMi<6CNHPrhH_}JLv&$^(i>ur8F0H3jT*jM*I%Vv&hu2f=hdAQ0Qvh8j@Ck<E
z*O=Byf}bmYqpKPH#$}q8I-rf<ZeW`80KZzSd{fMU7IDjCY#dO?i}(uVq_3EcsjJ(?
zo=EqL*gjh`=3CERjXt3px54cX33P8t4IU0Hh?$eh7B=}onr=!_Af)ubW`l@qXLs<C
zRV<jcbYL=`$T4iVU14mN+j~OFz1WoUGbz|k1bU6Z!*M(c7;5beUmY;Nd0zK(l_*Q{
zPN_{he7n86dJY;E1st*^jX15v^6xL!uDE=<$<0rN^(?OC?A=_Qy31-_$9hrCz3>5{
zW}<hdmW;PtmWb33H@z)gHiZ1A1qcN=2n@o^`@q&;6an*kbn%;E(>S;1mlTAcY?3O;
z?%rHsw<>8D1;SL2Q9Y!##@`In8L69s-O%;D>p*K|y`ubo`RQ;-C5aQ%O5E?qkR7tJ
zkfk%^P8PmiOuLnNDHO=B5_Q|NN(tVsm8Ove@#I;VP$lsY%*jw~k>ygj(Z$F-Asd`_
zl(f<qmbsy>c8-F864d8HGowky+DM!o*MBW-|L58rPZ3`9Z`e$p@jIzo_?GQ%9+d?y
zAe^j<NsFpk#Tx<j#nZp(J|*w$ogJ@2lr`V?@SS{<tPW00^k4-aYq#WG;ici?;yms0
z#sXT5rM)}!9n=|ddcBBoRWVr{<X9>Mi+dSV!d%1UNDWC2y{_32Q{jaKG2tumOho^Q
zBwpY=F<`l7i%uaG4a5&{fXD1pQNX;c7^Yw9q~IKl?-W$hL}0hg40<;50}<eI;%9{?
zRAY*8WwBB9-{V?|<+~%;8gG#U%Z=}ZhLg_`bE)ag1eHe7e>|cUWeBA=&4jGm<Fhg6
z;D;m&d_vY^$9x=WK6R?7NXC3MXh+yXRBIO-LXG9l)xv5_cekMKS=ys`qL!Z$I2y<w
z{mcp;Ja}Nb;!d3@TNPdNPiNH2(6eE?%xJqdXE`3_DyQNLRQv_%n<?``mxNCkTpkB(
zR)EK=2HVaF9=2=Jzem<`1=RRo2w+grm~eHxE)nNBw4q=ZXE)TXM+4~>@h98Ul$EdC
zzP*b$M%tqLgaGqjRp`BqF_vd57wHk6ruy#C#-r6?EjG;??hm`bM3F>3V8r*AY8=Q4
zZzFa91`sqhe)zri$WX*R?4j~;nfirB5Y=gS<atyd{=>!=78Xi5$k@%_6riTr0K@TA
z#H1ZM>0>D5DZv#)l8;``xUzi_W`q&k_Z6$nc~s`MJQylIp(B?e;uK)k#wNe-rP0J|
zkck&%Wr|~v)S~ovMj69w1t;UV>7eCteG_p$9;P8gZBRIbWi@cov(Mg<-|lv}7Ryf#
z^i26>CXVU)c}ufnymx_UMY+Nfw(XmZsLx>)8Ig(NGh+1@kX2q$G1bhhIVz(waW?}r
zW8_jlEF5tvXMU>DIcM0jF<<X&gI<e$cJ0RoD#VD(Zq_&7(0sH^XL_?d(%kYL1F`kf
z>|3|`Y4swVNx@Ex05kV|w)+T-7B=z{K_YZS`OiUTQB@M$(3^Wv&ER}97aBHP#w?Y-
z)AhZ>(v0W}kN0EV5wTgojxbh|3pj8*WP@x7YOXO%71tzcMIh87RNR8Vfgi{oON0{B
zE_Zh4fhDee?`pm8T{M<*-%M={EU)JM_jHCs)W{SvRUcCkji3(1s70bFMY7y5nkOJm
z3V7<Mzj4VG2SpaCB6KE1{f&kL{3FXnK)fRtE_a&Neb6B5wX*|OG%5`{Vl_?Gcx?qk
zSPpcm!*OqMh)vWVKD?6x9u&F*2Z5p}Fyyy3Vf{7Q(3=vg<e9)j1J1leogPJh9Q?Dd
zTN%Ogj7LOAbO&hv5}w$+cw*3&Wr+`lFIMK|q5+ecXW%(zvVq`<e)u0}A@=8JW1eCe
zo%xCOyoO%6Rf3&4s2CYw;)(kls%wbAgm9#88WU?%!ZtxV<W8T`!fk3dB>sPVy?I=Y
zYuo;PDf6r-GM1#3gqC?s%8;qjXs8epQl=7-Sw$2=W(sA9NGYjQLS_;w8pKj0Lkg+i
zcjvj^Ki|)1J<omLYb|wM=Xo6au<hHn?Je53KQOIkb$`>T=dM`P8hlENsodb-?EKk@
zrfZ!REC@)S6#?lbar4XWAUwTcI?AE_3w-{cAH!?QdtXULOKPfDLiLxOIb_U{XR=hF
zg9BTNwHP1=G0S55^t+_T{qP?bIo7I99-2aym9<eY&)f8{p$We`I(FNmmp?b&?ERNi
zG|gh_)H{SHX*_!jJ1AOIRwceA)JdM5RnH&$M)>+L{8(_z^a=y~$6y#p2}EHo`$Z2_
z+ZhFGgc4JiFIyIN`SQ2;Y>X+{a9aN;%-N>LqY~{Q?i`);(#~VlnTCb0#+v4hpgVlA
z@cR4U#P-l)4^|DP=B9~WQ~i0200E3>7tg;DOMt)lKI2F;DmdDv*ER>-n5Ei;KJPo1
zDpmkHVx6phZ<U)!Qh$yVh|n=$^$|GXD2ZlMYS2g#c?6eYvhK@F3V1{f+y^8Y3JRO0
z^1>S59kAPXk`KOr+YQ=IX&-tg@n}5HxFn%b#rb0ujfiz!hYvqYZx^(G|2QO6;5l&0
zr`yhY_W(|aC{-oe3>=89v@dcsVmQ}Rig}s>#;i(GZiR=7I*?$tBB$b8#`g9Zpc(XO
z;9T2@(y}?~)$Q=?9t-|c)kytIQ+LBmSFBNi70l-+$(|P$GCbf2*d?s!3493ZMj+<v
zQV2Dr5s=JG!N=v|E@2Ci38xLe+tWz_!P-5ZIhpmPxD1Huai@zSmp-|$lzQ%Sj6;;V
zV|La}=QiU@7l(weSY%YW!tg5Jy)A}sV%s2cDT+>k{HVOJ3$H?%#d4`N743Uj|HFp|
zVHBpF=$F{fgtmpw5XxO976g$^%ajNMz_lI7ArNEX8X@9szPs1$p%9_P8^-s`gE4k?
z`>Iya&ql0>0iyta`|N&&=t+FSiLT~LB6G6B&b3tyC@f{I4rhd)#vq3|ozT2_q&5+O
zWic`Ki}&!>X%5s!Ci6&TVP4Mg9af(Id!CVWQz;;VvOm0G8U8O6QzADLo-Fb9EjO1(
z%As}vM}$d?R+Dc;Q7%1S>((X|_xo8S*|f~ea+quHcGqZrS}$L~ek%;+>fPCEP{a4B
z??fMR+$N79Q2Hu7j%>OwI%~Ja>!VT41N2~3KUdT*hvRj%tRGLQ)0(Q8Kn8+nh}#m)
z@2n^1BV`4qwVL5#ilEb{rpzEPIqMy`_qP4q3w`UO#gsic{9-4NRy%!KcU*D60=rY;
zli2VrUR<^K<dgR>YrwsCKo<q?;ZJ#ad&3$Ka6(ctJG%9k3o4B%bVAYVk)w#yUDs=X
z3(>cFyIK5(?v@NOvVqYvyeu7ZOG9tGiyy34=Esa;Pe_b?z|Du<IqhO;Ko1NuME%g4
zIRw3g<(%J78;h%*2$@ipWq_HJ{z2afrLvSrd=y0Y3OY_AfO&q5*GMt<?#j!yNQy-g
z&dKrpvONJJvB$DyXTLAc4u_be@i@tJcH6%t_K|kmN6UUNA{N)6Ra67e#n`0Z+dHqR
zT1U?1Zj&*Y_lQo3s+xOGH@I{Eer)f%@t|aE0lxV)e?#q##9x7dktLmm*BlCcr95t|
zvF=FY<c7x@8XKATH{R;MW2kQ4dS8FvhK<a2o&2Vyb=0?Bn>yjATWCFQG;-(hmS!%R
ziP`%n7%qK1;>+9Wt&JW(cDg^W=QFMSS3S#aYcG9}=XoS?$gSX}!B+QNC-}xvRf+MM
z?tR~X_Up1!aW7X{sW-M|968$M@X%nXKie#6in6YNl6TJaTI%$fix>9_$?dMA^UA$*
zw{|o|->+o;o4b`N5@k*y<Fzh>2md@-=M~YhwzMer<+$q6eKjUV1pFwyWLZ{^=d<ud
z^d_p0kjA%UJt2e*jlM3v4B!Zy>9mcTi=Mdjg<@i@A9MHCt(NLXh+YhE1b_4h%1K=1
zi3o4igWuQOOu{Yw0#@$2{Vi(1@UnRz5)jMt27D_07!8}Vj>CyZE_>=sqKKuj#akhK
z_>{BhiAyFym}6N@BkKPxz>Q*0%FA8}L^FQQjF5Loe%SK|B}j>?i$+^D<~y^X@mxF3
zxHJ6;y%FY)fk^SdH?N(W4V`L$I5BWpphbM8&HWe4t0A41T1WaXHJ;eG{Mp8pf0QkG
zfl0MB2Pn#5+UULgfSyB)L7BCSr4y4c2#TgK%!-Rm2`}O~4|j{}0+21UYCvCE@5~E&
z)pxk&ILoZleHNsFclD}S>*Fbg?!t|#lOA(d)&w0!jnFC#1@J$h#KxOCY)GG#vatW&
zAGXKCMvgW!yVI;!;rP%4d1tS@{-}3#?6oylk1SioC$Pg4lXp<exKsVpu}P;=q1xeC
zO&vpTwXM^zP}tgQ^Ij8Q(=DgtRi+p45n#L*3zuSUIX7XG@!Hy&Jp%_5IN#3&m-A!E
z;l0t=Ig<A~LEZe3{Sl}~TnNlN+}g%cP9Vuv1C~6%sDch2ajf6-Ptd*1vMwdG;g!=B
zvSQk|aO$~ehXwbnRNwwBcY9#GX^#W5sNyv>N3Kp{$sQwyXd)ZAO?c{{;sXZ{Rv|uu
z$nltE_yltWRx2S?9^!SMUoxbWj#eY=rgM3GeEchS`*(~XBiMgIscjHb*NX;=P9f}u
z;T#H2uP3vgmR1mp2>}GRUSAIeznx1>eQxJ$J&R!b4<#j2kNrw~z2|9zm)Ct(MHB~_
zuhwK|$vmFIi#U%Hx+$OOhK0wlHc!mX=_oQ5_Md&x4&GVocCYwB#f;$!-`JnA>G=vy
zOXFX|gj!v8kV^sA#EH(2*?Rp;FYcm85Lud8y^l2!yeh%I&`^P548o^`zcB~@5A3gl
zKp^=&TnQFU?4qJ(Qo&B&;D*m%HAdG>ni0&XPBP9o4I4KOepjn;Qs?d;e}wn1IpDJJ
zN^-pGrah?3p<>8}WahKDd*Id_Y8P(B2jz~)wiY-u{6gY9bMX9i4LUjDXTfmAy6Ben
zIK9c+Z|83+EZ)_v_B{GGFE*yZKhe?DT*u9$PqI(+UcP5VPumD98i*#oT&UeEbDet2
z5@`i8I`f%P2lnmLRm%9X;oCDl76sUT!#Rxsw|cda0TlGvLqO<+%@)F;|7=GY&Tidb
z**-n|Z{WCVOh;uNKmH2bSKfSK1vi?x#ft4EdcDR4JAZ^WCZiRXSJwuaUt9kUErvke
z19DmLCk{f4r|SqsNjydH_&eTe9;iPGey!8$tfAn>ElyZA4h^+Qe|P!?x{2sJU%L03
zG4aTsl<=&SM}7_ODuz~*>UzQ^<Sm<8w8n=?Tz8}aXLy2knS8tQzhhf?4YT;|)r{`N
zzI?%g@c`i~zrUD=T|>yWabqo;S)Wk4HbAr^-@l(ksXj;NiM1@{1LenT?;fv|iH?!S
zj2y;u4F2hGtgnu5Bja^sXsH~p@jfGH`9)exRM{J851t~zXQ9X_usi1TpHdm)`6-i)
z1p+dPfzXWj-C}*ek7G!@7NQHBdSA+&5Pf?sgHRSHq|uK0JjuqZusyU%h5v?xKQ@OD
z3<a8{G?+LvHY@_IPO#M_IxVcodTKb}{42sO1f0uK7fEH;XVyj+EpABCPNW0aFqjy@
z8KT)q(+d<JctL*LP<R9ciGb)489O@9BO=ERF^LGx`WsrEXPCJ)oxI<06Q^qHrv~-w
zt1#6<!B01$3890pK(=?m9~rnIoEF=8?p!%9L_8{%IGmnvf5(qXj%RwVNpM)`shhi9
z;%tQuj=YDWB88`GpR~p&B12h83H{&dt2rT0tA5X$Hy^WbunYM=x2kxttvopJb6ErT
z&iGA&@I_M$5|r*#uVzyA9{7kE_|0d})O63Epxm6!wf!@ZKzNZj|J?&(l<<O%#`$3x
zsy3eE^<Hc6JY_LFxlB8L<K{m^R%?B@F-RjR_+_`VLV!JrovrQTYxxg(9X;>A0AY1v
zB>(65*<D=EZ^cb;Bh+$oIF<MXx4gk<OILDT-F`gBdbTaPKY@kJbIbc_Ynyyi-pLE&
z`a&?3GL#uO3s7uVzh=12xfkuxab2)<me%-AI@Psz@0Or*BHU`;d@jchE+l^t;u6zr
z$wqLZbZx`T9uG6B@AWAe&cQAE-TM^D137;1A%uVo;~Fq=Nlr`_ghSatE=Lni>-6|#
zjgYqB080aZf@Dd@@Val6%OV&X(m?X@bPNp6-z`G)brJ?0P+|1bRf37*xj?L`pZu31
zQF`*zG>A)#ps<Aqs$k*tc)BK0u+yOuBb6V9jZ=MLVoK`jgd6D}TMwQvwYWkW4axzw
z=WJT?Ao0P_uT5#T2T)0NXHkcAuQ0hq?jW(Ax*dp0_Q6V32^=DFd!)1*aw5aqtEq`}
zvkGNrAC2uY9>-?Y_I`sfjT72WdKel8Lp0?f;-=Aq0w6eMGnhl9NFFJpo1v)Fukaxm
zQ%Y|2^Gk8*y98{Ji9R^R(@5(kc(gD~lAjI!e&uw}sO(Z%!y?Z2-0iGBl1+TgA_|YY
z5qZ>7F7ot4e*GTe@%wt%THD9Be@ue6$DGHvQ2$x5(R^#}g+wA+?JcCkklW9Qw)j;Y
zCIN7_%;)~1ZxX`++IUfC&1PnM{AA&FkTkJ+XHdGXP$LP|35ppc5LCGQ6n~#JT&Lsj
zU;e&B3I9~(F3!!?4r!qb%;0aDxW9ghlaPP?O;qnT_T8<(FA57~ujkxLSL0ipx?}xr
z&@n9{2T;ni$h-I9N8D#0#GPiRPo6w!NV{i|?MBKT3g1(%Hz1){+VhTig{cesKe;e-
z6ulFced=rfjU|z6V1%(X{<=t1FJ^Q~$b4jTEplLYx)PhJ;fqJ#pSD;Zq7wfOZOxsb
zYV-jl@3Q#5>3R3<Yq!%;yT)j+O2x6>0>(t}q6p@Rxs~N>RwNyaE!zw@!J-ltI+qOJ
ztv7GMf*#QpC}7R{bxz1er|@UbA{1DUNzr!tkN{>H?q}yNRw%3_p*rR>6hhHvXatON
z3bY=n%#^;%wiI6d4RY0Xs8_j<eIOKLi?b#7?%vf`ogDWXV+w9D<DbNtm!ALH&xAX!
z&W!y_F$<X1vsv(&p&MdpX=d41SEv@1Cavv$w$Ev&iBvK#@bs5;jePdIMd)UP&;d&+
z@4`Ai_#0DW%<6jL8N;5|+=&|!%%^+CYR65pvI5jQot570?5I(r(jPkjFawjWeBD5J
zZe;@q)9g8}1H}N}##IXOHy~>=Us$q#p$Vo&81~*I2-T?~+5?xQFiV01gppASz3%*n
zN9%AMwI2TB?EQ;}#I)fCU<{(k<3_D3v~1R5*tebfS>~p1uEned!XVnV;<*tAp#j#2
zfiVOZQ5FFf_6B*gcbGsEufkYS@*uxEt(g=5KA#A)m|W+3ckUDt&nOmQEvYh6hKqIW
zXUQ26ws2GZ4;%A!FzXk+`hnBq9uG5v(C&H``Wgoalz#30Vjc5?D}^zI!eWAxQ~t!g
zXIVVBpCP|;`u53E$}PFHy7nKA619yYeSd<+_R-gDs1+EHw|?*u#rbxPe(}Y9MEu-O
zV0#ieX;Qpm6*9<zt<HM+IqR=UiAaX#LN;KOLTtl}G}^^Z%2-hp5(ucn(7qjp0~E+;
z7!!%0X~WMmW7|l57}#l;8DCmfRYPnI3fBm8y#(aM@DDvuxSJiixPMs}>41)hw_(EU
znkaVuW><>P_@slZs3CFs)pup$PK-iN23HHtI+zC)3-Ofpt&5?@c(W0cAcT|l{XY0$
zk0Cx<I0|e1%p4oM>{G-S#U~nv%zIPg9#wWL#WE%tL_!>esju`sZ`++1z5PE}B7D&{
znLXUzg3OA8(duCscy;pr@7Mviy)f<p%F;X_dv2Z60n~Z=_0{&y<wL8>Q&|S`F(go|
zYa*|T84COs*=mf(gJ^v0>~1mmlZN&%3rLLUOJxefQCW?!@ET(YYN*vdK2lpU2N5|V
z+%>SPG;nC?iHNbW%cKH<x#0fxZadr~1~F<T-TXmt`>cLHc*(1&Hsh8$`HC;poKB*w
zVTq^{1RmKZ3nAisWDCS8@2`s=g!okgq^~PXcB4Rrlmcm8IMahRGseb<0IWmd-#%po
z{@~S2+N2AyvR_ZcmQ6;iPMX%Uh*Mo?wbC~hxI4DO=Y5SWPE15yd7MCFs+lI|ZKL1B
z5ZrjPub<T*VCfpaQ(3fhu~tdu7OjW>c%y1@!0b%cEMm`}r52=Ju>w%Y6p^0a@Bl~h
z{qWChYCD!)D$5f7df7mWN~Ro`p?3>eT`Es?9~-MdaT&}##wM=S>gCIq&*3)7kVSOZ
zd-jOYx>wQm0aeVV-0W=rJ7VTqcn<cIu0f;E6z;Dq_iMhte|zNudqAOCKR#`Dvy>Mq
zpc1bx69UJ$$1@mU%2s4~8uVZsTUtB_L((e)IrtVncnfmKH%PCDrq`!b#k6G$!5gE8
z7OkV^!cU(^S0cic-JN!8zkj7OzVFy9W@f5*g~Lw81xYV<<Q}SA6fyko<}`^_f$L8Z
zpm1|G?>pQzFTNkRZ9@1WrZV>dY<&xz7LRjRKMDy)^w*5-5pg2qzE-?&mbAaG!0a0k
zP^dRZK4PISw^Wn~CI60K(jl5i%%SSYoP?Y&aGn2#WSW!y^sB330a44jfB#@F;Sx&_
zx0QpOq2>`Z!~z2<p9pDD;}+jDjWFBOSNrR?Z~9ZNgl&YEVRW_ImGicioulQvvI8PF
zJ*Xz3wARh-)aQcV$KNC(VX?_}{`k`=#4J4eP9<X6X%)|-YS&Ku-Hz-n^$C%gt8olf
z5L*YJ*U4f}q7=e)pa^J1E-^1H?c;zm#mewgQte@;7&HH;V=*2z%0ABhSi|Pic9p)a
z4gLAb_GIwVBP?qx`8=xldhLMF&er`)SB(jG1qFm4QOrQ6!|>H{xq&ed5t*QgQ5t~k
z8ry!pRCramOPUn?;3@dexinxY^b`wE@y`MrH0vWqzaT2pKFnje4)$8FKYRQMqqa=l
zLDI!;jw<E<c<EBKN0dFrBrP9f0-5`8kYq`@aF4G<jXohZjMxvLl!C!gz*82R>u%#i
zavx7LW(tqV1V3UZ7%!C!J5gGRDdPO1y<uzh_9yMB>ybI-ORPokq7^5a9Kwo~X5x;|
zT`ZKK5akv>dUen&JThcbc0W9dmM-sX9HVEyXWqH9D?gZ5+b{op_1eieND4BJ;-j%j
zVJD`J5Y!oO2#lWSFd4;R!{-i;rPnw%Vgv~iRZjimKlK{#7)Y_tk~d>n2fD|#iBXm|
zm~@s;!>p4hYszfH?^IVOTyJV-9oeVOD`=@(^)pc)*}m*=nMH#_lEgoCoRyVNV#%rh
zaRGMgdb$tWK6(`Jih#~sugIM(G<Mxi1DTQGiT|Yu?39>ta>cBb5c%!mmqk?N@upNp
z&SG^Drv(u{;|)oN1fowry{HPpLVIA>0}auTlq^sM#+M-QZ_xC#QE$Lkknu8Ml|PM9
zT`@DBS-jC%Blss!WZIl~q-!>!!%(lCcr}aR2H2#Vw?0ot*7O0VDa1dXF8bV0t?<Ei
zU$4cy*&Mrl`@LpImipu;&$@WaXgwjMClu-YfL_-{CMi@ZM#j<6(L`F9C=F3VvL7fr
zjO2WfCCrEBbKOa&bnbx|gh6lW(P;r70o;L;0M=Q9OvVN41(`)?cGN<|wXgT963x8Z
z{Z(j)US)kp07}Vnbl}dP4<<}EwTN5xUIpVy&bXa^al&D|@F<2UM#Of2K}}i7zM3CD
z>BrUU{r3?*xZxasJH1tx7OA7=VAfHmwc|Bz`^9bl5Ht9<l+B|Yh5;6DNpU)TVRq8q
z=Q%k6FM}wAj(0i|^SQhn^W5Fj=B8;#Azj>43h5vAXK}Ie@X+kqbyVHOXK&xT->7mI
z55G6<=jWB`En6ltAq-&3&#~dX1$G#812i-}_?`F+V<Z#zbPW?o$X4^njMP!YG)e}M
z`GrTjyRj8Ze1hOLH!<RMudK+j?csGf_LScJMgT5tu0I$u=ib+6+!#aFW^}^&hY_Mh
z*;3AncqWVX^=mWbYZL%)gr-s12ZI*olRe%Z%_QlhyXr4En7aGh?h7|BvPU84n*ai@
z!~0<zi$ma&RY!6^op%g%u(Go1`qmA6S5ReA4i?^HmMn?ZO~`-uZVM<aNFrNqm;#&T
ztNCZBGGm2FFUY{4LH5ocU*_Z#m6l5H)OPT)(*q}$R(R{HjgOhr&Zav?toO%>CffFR
z{hBC`)f21oqFk+)m)W)spK9jyY+qA{_cbh=@JVg7mfA%w3%nlp@bI+7)Gc*AF;nD=
zUk}eWNT%R0byxSF{IlC~Z#EEWheqN^$Sq>U;<>E)az)Nggfb;_IR*xrknsRAdSiwv
zL}^18{zw{)-2;wplVo7#2)r;NlFuPCf-lYHR_oo%NQ@))4PLItsHuCGF8Pi-PJ~&F
z!kilxkwAAsvq$H+a>L%R+9>TOL;Ljky!pXD``gc0t2yH^1umVDTUAw_PPgjzEJ$Mt
z;(vUn?6!=WGUa`RH?SEj02}puG<I6abO6qKJ|+EWabD?VxY6C*f5GF=^6okBqMc}+
zAb<$$f^LBU6Yrt8coQ>k#Db8CMote1!4m%D>C;=$ajKqpbPHN0!q;?A5yt%JrZPyK
zvZrD0CHrfy;ZufRzoNt8EWEaR@7{UKN4k+~sq>lYop@xIw`EGPaD#t2?oX;^iak~7
zTII5YwBUquo>NX)M|#oU8%20KP<w{~O-n|gNP0qjVAiCpjKYitD<Eo1({@WTGtKew
zxk0$8*QLxWBsVLPq!uw+llZM~o#sGY8Qq*5I16`ottKqp4G~w6{rmqpELItEuoYC`
z#0rC5nkTbMS2B!WNw155lq$soHo2?}qgQtG?!d0mV-)#?r8IW-b9A01zFD(B*4xWV
z0cmte>W3qE1ssi23lqTX86mSR{*3^EMLZOy3)mo_JMRi-gsUkq_G8*w045x3>iI`a
z8NYrEY$^1Rl0le*ad?HpKyM=h0s8e()FWqBhJhmOS=B#zl~2!X$nDleL1Syf9=ZI7
zc_&4VaI|4;?U9XTo}QIvK6&6$FgUthFXLapT>HPCi~i3bD{jL(#MmLZIXO9yOw(sm
zyC)X^mLyzx1ndLgOw-U8lub6(V7uDD|58_nwZE5p{QH%5j>Qz$@2+9C&=!j#?kyD2
zK+0~sJgU(73<4*exZ?yR0(&g$01Hkpsej7qQtsrhHt?u|ez)!1*|BUXU){HG&@LX<
zx3#onD;fRMANR&~PR)KJhek;L)Qu?A+LmNTJrJu^M#goZ+-enFMrsYU8Pz8(C$sPZ
zSTQ)Ta8?BRL||+7?A>1N9#eXD(_jm?`N{9kE7~deJZ&-yYj4lG$m_bnOVGWJMl1Ht
zAoKTu!@%jRe|qx<q-d<}6a&SJUbrA6mMPacJ6D(4vq#z3gj97r%OVER10Ln6vLSiG
z(?vmigR@i6(lN9bZUxmW?KyzG3h!Q~FyL$0xZY-FMN0#FM25eVIc6V4!}zxXO$RBu
zO11%n<oEBiby&IS!i5X%J8iHzGY6gfgaiGX0aoG_CfjFmCy4q`Fs;p?#XTqDg#{a1
zlnc~ZXjvviWHhU<%#^*sQA*(~kFC<1`U4(f^D`6T!JIq#bYKrz!NE{{JfA{vI1tB{
zXNJRSz(oWlJ!H}KK)~E`UMY2gh$Q~3e={Xbyk|vw(+_6Bdm=*=i{0JJlhiJ?Xxupd
z-0$rj27PgBF)n^fF<#IlD2s*H8O%Z;tq$36fzjJ+8hxYB!Ri6&S_V%w2Ds^mKBr{k
z2yZ%cXhyef#(dEDY!(dJ)?NAa>sMGIP6S!Rz4>KjlZdZg_<EYCMxdCnMDk#`{V6*F
zj37d>!k()u%k5=FI@~%8=T6bw4H&wapZ%HG`V|>9(`v>Q{U`!L259xzc`!C43zDsD
z$wpZZ1-$_J+^8tog?We91R+#BZfU=4E01v_teeob3+aVrEg;a{AiY1mXZ+WRM4YZc
zH<{ja<Sk8nYI=_u3Bv&<m&}TRf=w-$4T8X`Q(0ITK7{1PW4S%GI@r0m?1f4*^P$IH
z@d{w*XZh$__TyL2f)b_m>lxsRs1n0B41jqnv!1kUI`=}Y`}_p30RId#*DiwdS~gm{
zcCFY=fBZrCAxCT8>4z?O+SBy|+BRjA$>|B_rsIOZ3f^_JSKJVjb2hsU9&DN{yklfT
zna(>MK8>jTAu13fk1Ov%3H4SVe4LkX<n1{i6=8}gWDku5EKCNm1-V1<2@%Apy6SV*
z>fL+<>t3uL5Ggbr$QO*$;k3F!CQe_O`|Mc={&gaCVCY~>UCfmbuCDCTVp>2w%hs2p
z?3RQ4DI1?3Jm}t|M-%V|a>*9VP{{unZ%8&}-lrYmZRoY5^9YBx#t>&smbnk}C`@ZM
zM5q^Vg{6YJeLLdu4(rIPxGpgLJ>E`mINnT~yhDpm(h||k0$nItd+zFaTM2Pn%FUaj
zmsvMs_DN{i)4Q=We-7vtGaz_MLI%&%x-~8M_7F;Ro-ow^*3E{IaV%Xs(`iF)+l`uR
z3RN9p@>XW1tfM6<?P4fBzE`4^+g#CYl7z%AOCe%+!M7i4gidYG&3*m<3#F$X@uTkj
ztpn7_HTGa96O#M1=|{s3!kfPdYZ8rVIiv#?3*O>|iG3BrEqSUedO&jlX{WEtLj_3Q
zQLi=;41=8w{pv?4n6}PgB@hG@tIU1f#zn5Qx|0;X(MmJXp_gIz1-CTwV$Ig?^E88X
zxNH2lb(JL-WbfSN(dTT$%v(CU%$z7VQcN2f86E69-ryiGqUnbAl;C8R6NVEDnaWXT
z9B+3ZhV-d5<>ybGWg;Y4hoQ)sj}9Q<z05*$KhDccH2b@PWX+oE4Xf;Sg)W(N+6|{g
zBy~dmVMx5L@aRWoNx;)Zl(eKY>$@I*5V_J_FuF1T`P%E&96)cNx-Oau?JEM0J<~o}
zQ_@zKj{-4koZNNC&>P};HT!h&pILt7Kf3SESyOSvJpTTXMYy<VShOZ#>3v-8s>iwx
z%4o(JHlAe&2xet%KzcM}NJ8k2(<RVKp{b4U=X+TV>2%?0<`dmjyxsPuICa697wlxo
z&u{Z7mKuHtIjG#RV@D3+Q8)#n-IE=E^n3WKt6?g}>f-K11|}k$Y!1+@N!DL#|BUl~
zmA!v|n+9hrePmf&hh2Z%ZtVMdU6SHS%j@v|{qvV+UtlN$*z{-f!TqwgZ*|^g?HQCo
zvz!&|V$rkRwPKb#nl10VKI`<YCfWiXt51gqG%s~?Gnbh3q#LxSKNt1;0$ehSSVg-4
zD|7InNz;$^Tm0&vX#ISU(oB6iKyUKI`24d5-MWuku-wNfHPrss>l+pJZJtHO-*42<
zlr#t5PZPI5Cd!O6L}^M#62{3ARRmCDopP`dF=#~8X!cQXl{Yd`fh$7pmwK;W6`{nu
z^(G3WkOLim5HEUjzsNp(v_im9_y9IF$+amO_ZuyF1wGr@2dt98q1fGE{c#XnfR7G{
zB-}(hJIbeKF|JuRaD)SN{R$TiWGVgduvIeGlMarz3T}T?R~CTC0z0Tu6vgz$H;7K)
ztFXe-7SK9C#7~DW)GcZV!!uiE-_3xyIn26^85c#BrCyE7`*_H|<a%jGCdlZ_yBN;H
z)hC6kzAQ86)TvYWx*J3{!&lr??K~qdJV%}mN9pk|FQA9#h;H3yF1Yr!S^9|BU|*X_
zlLB>@&s+rT53|VT!0nGEC3<d~sbeTA4{%%22{M1MY^V0@c>IG#{Wf7VLnLH$OQ4Z0
zYE|~cOBEbwAgBrXo4(7)WBwL~!0=tCxH>Xll<{HV&}?)RLHnDepP3-FVp+XcJfifH
z5O8xK;}3dQTb}h_x*n(IO`pC_0>;b2DQM4_x#d$USB{@s{tAa5ytUwrtAk$&H>l;i
z*iz*L|KltJ+sqgRxJ#U~Wmyiq1Q}#YP7#GIk#eai51I<m#CL$cep3=C#C&}lrI8*t
z=U`0E|F~9_m^jUFgYDe%JK{t%EduiH;yp=Dk3Y()6(lKz$73Nx;|5e45*ur$4~=#@
zGmS-yZf9nW!m;AXdyEL$s6Tr3>Q!u;fgx03+PcmEL*y7ls-|TWD^lT^KnP;^j6!uJ
zfGwDJ5o=dw0~m#e*#Bh5CzAjGxhBaxJh5W_foP91jjYiJz*f)_oq@R!iNAk(J(*oc
zXosa$<pkTjuR6ExpwI>Kk~oiohsgoO1>LHhA)*nIop4#;wY6+1mcjrcpuk@^pRH)z
znMwaZ&@67s+2_mGuA4#sB;L`ynG7D_3wCU|Je+!*$&W1i1|YM2@x0r(c+aeTipA{K
z5y_kR^vn;FN^g&??1g5MW;|`icY0-c&u%?$w1-L(;Bu!v45?mMJ|Q;%Y#U-eeaaVx
zGrg7;=u71givr!T?#iW0mxA1gLJS3j5JwG}i2~CyToOviQDy?+SMV*$_(7llgI*Ot
z*TmFRT&)C&=HHZ&s`S26@7}*ZY2m3yE;Qu21eJP%qBDDR=sz>U>CbImww9+W6&n&1
zs{ZkF9vOQ$4Pb{Q23oYGG@7llgNP==j-Wo2orw@yVht1j=|E_$s0ioHFy?wh1uK3`
zNG0xEvw@!v!5T8&=_L(%j(hA={H*KjJ!hp;t9=$XVf0T=KyKi`mhIa&Y2B+wJ5(%-
z>N!0)G$eyJl2-H7HD=&Hu?t^V_O(_P*CrP#?s|bnJfZ0R1hhHu&9oG&ApX;YZGcjv
zT)%0{2ftW64DvTCVi8`sjL4$x0#{Om%HRQrefy+S%K6NDM7PCIz%mE(r>#s^TkASZ
z{fSCs+MFM|?`XdhQdH#@D5^4V17OaC=>d<y(X8Hd+&P(2!f6HtVT1wE0hZ2uNMY0i
z>0q}IMT8d1rL@zFD`im$C!6VUf9%}q$?_!#EDS(4BhlUKeU3xXXHB&iSHT$zo3I?4
z*@y<R2jxL`>etE7O4Wy7dT{7U72Wfm^lo;LXBMygg=i1ymyYYU7mRFb5qr_VVq~Wz
zE5>2;*_JnyE4RRk`bW08d5kgyH$aGp%ll1|uTtv<GeN?vPO2Rye03eYt>uejBYuBV
zrC<jO^ox%Nq|+K-<xSU`di!=GUx-2MuPrd^cq%67s|~>8DqHUBi0X1>aFKVwjk4jc
z<uE%liQ$yI2#WAIB*s`IT!K+^*)1=W2Pue_lmM*$8K1S3?oJqCLeUrg3)cu^t*Nip
z{91$_+$Qq-Vh)bTYk|nRQ&FC89Hjj|@af+xhJ9N<1*=~g<8FKIZP#E|e_A(j7*tAc
zH<7P%RK_>g5}CC+B*t4QbAIc#ZV-?Yv!J`iE2e;0&^fqyE|Gf{uS#_<0&A_W{TU__
z)B0dG!^~=3VZgo8#U>%JTgF8k2&4@Y!UIiifQ}|0*FRZLoB3}@<pHTWuH_94Sft%G
zZ}6bSGT4hnuSYD;pg3yBRg;x4m?s!i93%v=h0_cMKNf}fNRht4gAN=BusLHIl`SZ|
zP+jSI#!Z;edixU!XNHYqs!#z}aqxsoA{-q)kiZC>wLPx4a!4XfQSY*e#pUJY2<iTo
z+4jXycE4p_gF2AZjo2_lY4id28CnCDm4jKP&}Ab-baN|fR=dYJV(fsuM~<wu%{jvc
zuH2^!``enG`k*379pY9i$4bCps<_wBhD#FR-6-JJrOmo{G)ryoku~tBesE{f^PVA-
zBwH^T|0}Dsg5YpwQ8a5UOfKATHrx|>)=rf(3VF(`-{cEAe4OJ?{bK6x;F$2=d0J<t
z*LU3WZe!?#sBCyxi@hevf;^5gh(F<*p$Y6~LuhLf=*fBn`ajEc+#?okSiioC_fGkC
z3LZI!j7q_j%p#tWV01{J#F%|*tXof|By{K1oF-B^Qx)Xl*NT_sex&|fpP1<Lza$%%
zGpIC)9Ve<OtrrYkE1pXW7fK!ASr+`RXX({ITund)LV;DHspx=7Sa8~EsjaESX@)sm
z_<^z{jj9F(gfD7pK0P3^08D?#rWwwL9<WQq^#ynY2kOO(C)4=Lb{C)&CzhW<QImzz
zV$%jpt!d+L{P{r*Jc=T6!zo6`a5}7hJP=S1lEmX({iKvvgYcJ`YnLQ!h0h7AeQsWj
z!JkNVZ7n4zGss!P9rjq6H5+uIwBlz%<DE};&dgQCR^pTEs_mn@x+;aTtme!x_4e&E
zsp-?gN|Db9tP#6>-oFSTn8W!6Of<DfYv7HGBq9ctiGP-gK<{%~SxW^B_uA0V&~2%6
zWdmNZ$HhatcFo?iYqo-Fm?bgchr77`*-9}09q!ycN7A9LFwWploMk_o^&wT)!9<D6
zQy1+Y{ZB27|2{a*>l*0(pW!#DNbLcDU{jbj;aA2_nk4?<)3>!iEk<Lt1kID>axb(e
zLO%I4jtT&47@%}uR-Wcj0#BR%Wh&Mla+=PT`@H&R(Bg9%o<rvS9`L_~6#vBB$(%|+
zfMEoT78|eFGIa5}KLBja^~At&W@fVx0?45G+qZ9)JCJr0ec?fhl^X@_XE<RVNOX9~
zc$`9xp|IINd%yr4(K$}pCII3TWcm5N{9S&h&_KR@c%f=)!N5`EkHu?N000XR@&p?T
z2z!wwV=N4)#qZ8ty8p5Hvt-C9bzjkPoin%<VHtseC<`MhuP3EUk-ZfcrzRM_YYU8F
zYnyGBQ`^ij_Ku@Pu-iy_Ow8+o;*1u;%%caoelGt!d<Py^;r+1mS*>1}(%?KjW4nU{
zq@2F%Ac&T!L4ZJIde+1y;-?c83PVdODGpK*f;*L%>4dQRL^Mc}YB^cwXB=SA9$?)Q
zigG`IF&P<>^$>%Z!j5IU%cArRe=W{o^$O-AV`hFNScw<%r8hq&N$`zgR>&QC&T3V#
z!a1E-JTKlABx<C9?WuZFX-}@M{$WT1EzkGRA^49sL0_<ugLN#!2EZFm5dDnAHM&_8
z1u7aEW}r{hMs4mTN1e)YosTFDs-Dikfp&*i(OO&e(@9>19Is?}-H6RpOk&8Fy9nos
znE;ep<I8P)sA`Vu9A-y)4(8M%_V{D*lM78@>yRIO@?fj|aODG1oLcsq&O(5Z9202P
zjc{JU{N(~p$OK@m&`WiM+C_>~Wf}-$N<=o|1jxyxoRfWoG|9wz=l>5)yHEX_$)Fst
zPj(x9c%5_vtg%F+#%JVxVG+A-244C?t#=97eM#oIzLYL}<HtNLYCYNpv9}Qz{pnLH
zem2pH(+!7g1C<JAbUH2(Fvn;Qc=cll4Y2iqX1T^<i-!fxx!l>VaD|veL5yrZW=zJN
znpR_jwd?fLO~yT@R32D;+QZ4L-s5sArH|L8GFp|BX(uLcVk~wv?ux%zxJsQ@dmX&}
zNb&x6UxDO7<jHB=VUyG(zY|C%z%adRf~8o8!4U{F_IE-#2d+O}IF`=U@J0m8f`iPS
zg06fYu(5}91CT2rb<CsKBeLVmn?|pfOzuOq{SIFTXU^>(K#bz(+3C<Ll^TCCa^!FX
zN#qh?UxU@$8Cv#t2bs`fxC;2P3SD5$&-dNI93XqOx%ULx2<r)N(Ch&j6D`q;!KTEi
zPghqr+1U^VofusMIu>`@1v!8qegcqQsJ_u_>vBNmP^m;|*Ad?^+SsFU!=Hh7jPGfY
z*akL+@DLP8DrD|<=m|3<XnF&g-70{F=+wiM#{5AKW8ZS~v~WYhKTUDLqi8g_ty}0?
zyfq&m9jPpAoQvC`R>0<v2g2oo_nS7V#V`}j7%p0(wvcX%rD%-p8h+s&gaB&GK$xTP
z;=)X2YSGB^$Hwz<A1@kDu!GS@U3_XnZ4=av2*2g|Kn!b1StvahqGQIGLVrgi{AZgd
zL|*fR6^j@g6DNRg+5o3Bhi2?lOtZu{l;@O?kZ^kW+F_K1q~VDevGdt>b!!Ww$Ogxh
zthV6Ks_Iu2)aiyqlGvCQ*z;bTwaFz!DkIPs1q3=Fq%$G*d)ES3e?_!3l#%xs-HD)-
zyrksB4-vr_q-N#x7gxyieHXp7TKI$sjraEtx1sjbk=4UurORQbLHD2VYHj_MYy|+t
zu1g0(c#w^o!oe(_Ek8bNTDWdwnAc7KGL+xZcZ(Rjic*^EhT*|zelfGeH`Kjd`uD$L
zH6=5+{uD|{DRr68aUCY%Ye{ei_p<~DNler>{b2GxF2FGQVxHaZ#xGgn6WVvaF%O#!
z7pB`?YlMode9dw@V#Rxr)@RbP8LC}B@za-~-oLoKDRt@ERT*IqISwKu^ZkpiS$dBu
z;;9A&!JJk9&&lyLf@u%GivfWYLeR^AI?EviErBot0E>S;yE6{KaPa$hcvaaTK;CqL
z1{me@Q0KAj?k{o6RA({CgmUY+&SI$HKNxaw`=PMi*}Q;LW4YbAuf~f6&t5cbUHbI7
z_;fanw3tYVsSndn3O->Xzh3>dnJkjG@N7izLCwTXNr&+a=oP@vf(C#jw$+2B1aN$o
z&(^RCU8ZZmRFDr)q8mqL+t%b0{mP8Cw<u?cwWXESvM29|XVxi?#_qxA#!pJy{s-0-
zhQ<M2R4@AUj=Zs&Aa3OG)@_`Vy8oqhlQCFw?ThcWA!Sx$jdc=>@&@5|LiIR_VoZ<}
z#(FUQIst`(kaLLVu#p4TEBn4PAomDug%l?=3mz#I39-GCFGL{x5%1H^Q5(rRKFXEe
zv}yR>>S~;4@q#bc+v;8UryhShY)=`(y{@gvfZyYAa(*c7msE@h1yk6<O`j!hCt|Sx
zCYe^DjcRuEsV5e+aRl*E&{%2NRe#$nETU!00(T6Mc`Ude^{5W~2N@aY%;$^!>pI=n
z7D;LCZ%^2|x9LWBiC2q1hOL^+cZKaOLl0lZQiiLm%5(1P=&`(y*ohQGt}jy>m=p?8
zQPHwk>{a(4|8K5a5pr@atv+a%_+CO7Ay02)H#8POOa&w#M>+I-vg{nE31H|5uU1JI
zBr;D5*%nfF<@h`OsG39?OCTd!udY+=nA)$M`?Y`nJR7S%(xCHE{a}slv2Fr3V}uIC
zCHs6wMt{{H>awL;=-+T9MPX_Cc%b|)RNI`c6y_TyHj8f$vO4i7pNS(;K~_xCL@+Sj
z+CGAhUKy&<CjTB5a7fA2CteBrHH}^Zu8uZNA_~Uy^qwDwO#S|ak0$C~nD}8~%-HK0
zVHqJijKES`x9calk-J)0|Bs_!n`D4WaYkTVA}%aEU641eR9$H2d`sEZbnzZJ`9A9g
zZ!;suFbIdU^^`^)b*5SF@+9+1CS!^O25d$<DU(L39kJ`jU&1!~!YBLxJ2_@0rEXYv
z^C@<FPhhr4=Lg2Q4meYwWNfSjfF<ukq9pJCt%E5J`C?>EP)AmbYyxST=3If$)M8pJ
zZ;RK|nL={g=-JDg&Vc%O0|+bXxURV8PY9kC<C+cw`8LQ|ddF)rLK43gQZiUWJ0!z!
z=_9`svGMQ?3}Z3QqSOqgtt6JoT>6uDEvXKwsbpRXerzDaXkeC=JS7TDsnhjzyTAV1
zg(hwg9$8*N>0G+#Qx8^$_dP*Tl*Xn^!UYLVgfv4u%k-^73s_qndBKB<WSphx5==A2
z{TvCA)Q0SZW2A$G7Xq!gzQN}lk4(*+QwBlOkRb`^JSG!OC6ZFWggt*(u|4~50fw-m
z9{x|GPZThU6-y0f<JLU-L6vu|u|_OH2VWl?ySBp3!+L94NgpKG$f%5fD5N8YvL;eE
zP<NhW#0LjvG<*~)3yvlzxGWN8Uk3vp;3jd7WgyQ@{rYsu<W_2GI-&0GSWodjjsT~(
z3-Z%~G%0-90M^)utOm8W1=J^|K%uktbSPoQil9kJ))>L1jJp%~70C**Jn?QN^B#z{
zcV>JX@|R%0+df#S<zIfip7U*pZzVGpMLlw1Y@<vC6Ljd1PtGrq7fE<FU+poNs3l{h
zoL{s;^WbC9iu&zZGwypCsCGF{VDw(q&9{SFJ$(2uXU(B6u)FE-p?{s;`8#g`V1?{e
z^}n>Q=EdFQseiQx%zLBCMn$N>Het?D+4#xh$7h$kxcCP}H`Q--xqlmvNpYR!2Qgp7
zyldz08Ms<-l>+W+v5J7;7j7xGEX@4)1g##7Sn{SJ$4kk-mltY_z@((L5k5C>-ZZ2#
zu^8Q@S1*hg+QXuz4-mOfVEgmd>4T2xEbozI`}SOAe8QniHA2HdNe}E25@PdzG@QTP
zU$4rSMQd~(Op|sEGp#SBZQ<rIs^AeFs7jd9x*oG3vv67ez~85?VsadQ;X+Fm4~`tU
zu4;lcWZw#JS~!63aG+s^IN<$!sWebm?a{R6FgRBh`gM27jqAg-h|^$UJD^vus0oK0
z7qGAyH|p!f(Z-dkbTbs6>k8k*Pm4IPb_8b6JWEQU1$$nUG6pXEdhAcr_dPIJ{ZmN%
z9mmi0scyR0jPNYWU3<@_8M9w3FTqrJmk#0G=fF7R6iE|Jx#G-gZ4>>Pk`@BgR@`Io
z0TClF;r~ieDox>EhJSVAmFG&nlgGEtby4(6{PwtGJ<q>OiDd5vsvT~$iMzxQY1lZG
z{!Pvix^3;iFiEX?!nu&Gp<j;$zg(|?cvY{7+U|JWMS1ZtE+ICMs!7n9(=UVi$@tP@
z!z;oltqqwyjn_Yjp_6sfkJsI!{%9v%l|a$?mJ)S@YL6v5)5xW?-t-b<sKmfnq*W4c
z7HAQ$;4u1ZHBy(Xgf4bI2?Y9wd{42=%FUflo-rZxQwai80r1F*Q*vT4b_Zn9{{j~~
z$>g91LT*op5rwrt-aRV^WzX9@4Cc6kXAAQE_0F)kQpRClm=s9mSctKGv~}4tIuBSd
zpx~_PG~z|d`}->b0PSK_!^)<s#tu>OQ0(^$Kzwa4HF-D<vghTf;HeWS)d-RWrKL@b
zAXfax!o@%8`um5Jlj%k3DA%{T+eJ#3h{ou(WU^YPHoNsn=NKcZO*lwB_N*KLe#!7k
zAqq=4Bzl(~ki&7v>|%J3I$gX(D0eBgAG_Ml&ON_sj*RiF4iOvev6|?0y-G`-1ngmo
zUY4M@<zc$xhs*2X$;yd_LIIB-0GgFVj8K8EA3-gat$f7J+tkCHtC^64X?>1asYTOx
zBHEZo#ff{4xppnF+kS(FYJTGN?~`r^H#++C(3R5ece<(fRsbiD>ehKiv1M!1Z3U1X
zChyN+U?*3`_py5G%24(q2t)I>vr-pr?Kgp?1vk^u%E9xL-9NtU$2Uew4LwtqH_=U+
zHfrb#Lb3N}q8!JtRVkp}#z|{_`xX{EP^5AQj(T0KMJLw%@B#u1-#0xQ`baUUY|eL4
zB({1}!Mc`kOAk<4W~of^7j)b4UyakfKlt}ramwNFwd*shF0BJ@N`u5c2g}(MCqSE4
z2y#G$Ag_>_#NNGiJyQz+iUOJ0aGIn`;9asvQHeB?OfnEk6774dGkxnI|Me<87mM-6
zaMG0XrfxHN-+AL*C+zBnvEGsezEA|*8D|_FxFFnF#picj0Jvz^=F!3YF8^)&+xHtN
zyXamJ*`aH_74NA#sb^Y3B*Z3Sa_-EEiVE2s7<b~P2c&EKx!8we`cVhaiMS}THkwdI
zBh%^zkJo7+4{up7Zx(73?&i=RqN)mFeD}A^K}*n#whpm8Vt`p<_a+OcF)3R8<I9o-
zRRb12-w9wY0vPZaS=+@}K{i~_LRnIum9#AxmKyqJ(31V{{s|(oVJzhV2V7(|uCA{8
zU*4yfI)3=tab2`qv{7SGts(0mhO@jzY?=XlKmrB+;Ah@M%6;c)lV`3mfER(l=+HkB
z^*UUnSh#1!RE73T%g&aS%8ebjY}~ljZ;7{e{61YrN6$&;&iC=V<r(IjOOH38saHr}
zOWf^d?dfQ1N+q?L@mitJvK1@lvNehCpZBPSSOgW;kYa-pu^hKs#ltAa5fr*SxK&la
zGwTY^hKif)U86sjRfBd)I8h8u*4F&!&afMo3T(DSYl^Ou03kBEqh9?UQX=~8*wG!A
zXX?RF@TT1_7DIv+Ikm9&;4gH+hy~&xA0Z>oqbSmQUuEkBhh@|me)8lOhAB%5(;A7b
zHDM9p9I$&&qYv>32@uDWT00&9!vxKsyfQiQ$Q-RXfF`tGp6+AjsPq`AOpjIq(B0{K
zVL8j;z|aJ=99TQzd)bhok%k{TS5@BK|EhapP)nsSbEUDWCaqiQpE(`Sqf39A8F|*N
zygXOVsAuAtx5VDAG)C*Wnuq#w(=>yIr!RNeX|B}h{m#Mt_B-7Fw4?g|%6Y-#>I6>+
zzB(eipRFM^ZSK3fcb6W%`fgC=iLCE0`l}j`U7NjF<r+{%Dw74pWu2ivyd$$6R-JQ?
z&*#-{8XQR-cd7h253JJT^00s-3%h#t|8S+s`E*9ny4RmREqhZ+dqnls_WSkrZgH9)
z_FfC<7BqIFy~~g<SRb7rb*uQ*<uAf*o8V~m5+YYJfM;^=rje16?CGhuf5K)UhnZwk
z0(uoPIAuQ*C50D~a_^7oGlIs}-8?fsw<bl;t1ru3{a;QwR>dm!z`y{t6hHs5cU!;g
zTz&+9OZ6pYEBcn)?=$m4;fN-QSATP*wFSD_9TFlAhTKNuS)axG9iJYVHY$?o`pdVB
zsIIflNI^tMK^}Qyl8ZMpCfS&K69Coe(hH~9oX>QT>i8Wb9j&px+)?%T)z@YX{f4he
z>Wp*CKs*b8YSt6V>$v=On`&#@0J8tLb<T$EcAIshxz%K%`{~^9BQ2UWYv$tWx}LQV
z5Mfy+p^!y>V!(RMp%)s%c9!+GAP?I>-R#8NNqxDRuI9$3D}L*w4B=MGwsM6p#lW!>
z*9c|RWu8gO3R(O>DBfunuBkd#+^u~&ySFGTZLmC~OXJ~N>nW|9)WUW9_U$uU_$ye~
zzUrIrt=<z(mHwi=ySON)e?h=&?6;m3IBU<Czi?q1&3W_9GtC_w0vV=oiG)tdG(<Tw
znh}Na!AafQZcdrAcgALvycIvUMLbP8`e^J7pXSF+FMrzLQrkiC&0gVMxfHNU8qw+p
zwbd;&+P7zixNm&?V0zIM*O<ayYkxGU`I(($Y_PX07Gsn@Y2<JYoUbE&_=Sr9FmT#(
z>acg?q+fZT_$1A7{g<eT`45gk#wMxrG(&%#+qm<;{1pj8d7;s3hbKp?lBBPdSy@@E
z?Hj4Bf_InkiATP3>?t_hJ9J+iTj4d-#l^*F>C%4zBw4rK;9^|AW<ek!XrPXtJXv2n
z&l3EGhg<#w!#Qxl`K$GLFFvuKH?}lXR2-<=%W+6^g=nLJt~)EFTZQdigzd+=)Pk07
z?_g9nhSa|5$jz5es+6bC``5O?k<iCk>fJZb96bGG?5;y8wE3TT<V~^}18?lXDE0WW
zHVx~F%U*JUdp^)}{=nvhkDZytUWC76=DMPbX}tX7n<axpG&nrZ!MLGjSZuqd^*NvQ
zp+>DwEii)AQBqO@fp<O5<?s?a@%NQddsk&et=R-Xbqo(>=;e}Qm~XXOk%$9_kS#8P
zmk;dF+{)Uz0D?ybBl?aN<9@kwqF?UyUaIZo-r3{Aj)5(<)K2fw+W+*BU)ohM^X&WU
zIj-eC2915{mNRYYR1Cp=!MIo}Z@%wIU6(vtgS|~+Vq)9}AA*F%nzhaFl<K#dA2XGm
zUy5^IPd?4gPW@B~vaE(n>y7xEJ38+m*otSr^sb{0D=SF)y<wRkdW9YmIVSc{ZCSI+
zIy+B%{-t%K$JU#bYsyP|?aj>D)7nqXGNP~Enw^V#cpYd^^egGLwpWF>&-2@Q3MZED
zF`3NZMAuKqGswM;6OwqG5Br`zJKevHZm(X8J%3Tttg1+Vw?Ku%vOKP-W5PszHP5Sw
zY+H9dRXLzvzZp&&kJ*K^nKY6kW9;YWBu-00g-SzH*B3(UyW>YzfeH<zH9Bk4!*#&|
zPEnH`85=;xTQ=UL#41P@+t+9=Ge!ia{nWFhC)5QMzpxo8wJyi3^fo;(g=2DRgn84s
zBqJV|`@jxC&AW;I=AlL7;lp}`j@4FS#<%a@+ZD?TkB{^WZ1f1jyHWaTsk1JA`?*vE
z(p>l|MkLLT_8U26%9IV<)YUvgRH1ekoa!j@DUfe6Np2LC9Y9Xgz*xECq@kfKLZfRa
zB&5~P@Tejpe@;`K?8H-<v8K^~0kf4H*Bl@5UyI+*T-+~y?YFGKk*5XYdzzjJXcs<W
z-v#~ld!k0v*4XN{Mo#}6tu>v;7UoDhH>$sOxSHoywR-sF%a@(dI_<pe>@t5oA4}Cl
zs}+#i#<cUP1-_J!sl6#^+Km3<Z=yw?J2JwGb=6~a|G71yL1r{!p~mME$_4}Kot&%^
zHS6L)?E>|lRdr@Q@*3KBjic!TzhMgFA9r0;tAAF`&FLKRY2)w6ex?5@-|XF#?>ZcF
zkJg0Ma$eWOBZpfztgH3Km|?$?OV5!HEqwBR{`6KPqk8+<_e#K{+}v$UKqig#+TOO0
z5)I00)6QW8mH;rphMjga*t2I3{k5-&R*F!Ao?c|yAF$J^EiPgJ{!jjGKdPbT#&4rw
zQQSnZ5&#Iu{^gX9D^NzVyj6DKjzo?Nk7hHlQ0|JHMK>Bou+>7Ql^rKdSPDK`gkDed
zqxqG;B>J-7Z~WlE7B)d;v7?pOXeGRAsj8|>rVA|cu2}iAM}$@5b2d#bz38yK>+}D%
z+<x}afMEz#um@hk7e+kHbT-6kxQX(@s?L`gEl<$dWj=53WEFdTHC_mR(X%zlXv#6b
zxXUZ?)lTm>*GPENeJZkGDxU0p1YT!J%%Z)S#l}~)wxqT)Y&7>Whp7D5Zx=VWz-78V
zllbK$7TTY)sak!#;h3kfZ8wi(=KEpAsx2d|TFhlp^R3v5w-+7t)y6ol30)JEKMdI|
ztMW*flv+U?IBxjmX@RxNFT{#QBKuihyF`Z;d?{`ViF8ZecMvQk1xs;WW*5w88CX;f
z-gx5Ju{uVjb)Xm(-{=*uey4az0Zv_;9=sdHA{Q`E=!&#R8>e&tU|uKV2bbSnev?j4
z_Byt)rE&d4M0XlUcZHr#q&IhTY>|-H9H|Whtu8ESVAz_BGgg3P4b00fyjb`gs|y)l
z6<qwXV{LU;zJTaeX*xwy-7+9Ojsx>wm4D5X#oadUY{CdhHlRW`P+*zv)OPf~CYw*$
zobh-uXD_0lI{ZkMH7GqCJaX!yMLlUr_{5BpH>I7=D5{eg?bKH1O~$C$oMwk6HQ1Zk
zZNi<Kv+dQGS7s2@e=fXiAi2KP#BvDDttBtgndjtD)%Rx7&`ocy#@x2`QL}Xw1ukWM
z%nYBQr=Ix1#8D{D#})dI-S~V_)6{%OYokp~>kSM{{e)3x^D&K<XNWRx^XB?iwzeCJ
z2BPi*;4gv@T0b+IOM|cDi3`<GQ}rh-BggOw$YevvEACb0GvR><4fV#}GT`G@%)9!+
zx|9MuA53>k&d1S+@tNbD+BrBng@59XJ)ILA&(o2S#uFzt6B8K1S*}M$M~h#k2Y-^a
z*WGlDX*@0uC;ORU21j<*o0?e<mSQV<@AVJfrP5Sfz)!-xLxThNHbY2%{&59|xNG<B
zV|3>y?^yrlfswUtYxjgaV$3>XsqrzF0mi-i^l4=~V|g3c{;h%@Eh1s>Gu2P|#TOE{
zsgAE4@hwL&nlA=lZtQdzzU^!!R8&=&e=$rAfLpLB*Aqs15gGR;j$u1c3s-;Ho3znN
z&y+xR*yp=FmkEH&!)RMw_?D>*W--$+_~(UpesUc#GMG4VB2#TJkaa{3l>Hl<r-0<>
zZ&T}pAY#qu0LO2)`nTlSDz6)>ePi+3U(<&E{`Tbb*|XwL0MS!^;q!(4#hsA7c+Ft-
zh}nwlG*kEjbvIzDbz}aWMR|zNWpb~yPC>5*2jw#)C1b6ieA}d-)=#^wJaLhVwB5di
z{o6AF*$4=*MSJ-Mt_*0UxbEI$52L(N&Qd+%T=|dH6r#p_L4F>;Qbb_dn*a>+-3J;^
zTI|L@qgZO)xE}hQQD)7}v{jhQw37$T^*dfxHL{_mPdBYb?DF-S=62OI?mS!=8YbRV
z1DVQm4fURGxQdN8+@DQ@2L-)da#iN^1%n+gmC!Xce{mAW8IiSU?kHYRz4^>nu+8#Q
zBYroUJoIzjD+^z&X9&A7_efCCx)Id{ErwnFCsA+Epmt1E)PH+XOcQ%=xW%4`kJo@$
zYOdQF<H<%~B6Soz)-?I~`L!i^rn7@q^VlCv)Y&zYQMWu?t|aBSg`=Z9bS^{@&+sNv
z#6FM%WJd(>@|M&BK?p9Qqq*RCg^I8&uBf?`5tL{Y16v%TGRS{a^vCu52JQS)iasQ-
zP?m#<bw&f)^uSYGX2iM_R>FEZ@(nZFE_{8xWBdLKjYf@9o;r2v!{^T%2Ba%pwA+S!
z7cg8ZRzv_${$$%lWTeXI%FqcqGnXy)_Vyk-i%zs44l4r>dXh24CM!0!e4MQn-0T<T
zLYwU6mAON?qGl+l8Bz0jN8E?#Omj<1C4ww`fI`glwws)uxx8<yY2_c)|J_wB=ra5F
zgt%~%{tG96@kwMYnmRsDfuZ&W%Ig#&ZnHaQgHVx8C0&SX49e7XCV&0)+lz~`sqQag
z0s+3w?F)@GH}1s7$1eqZ@w2=>qXj;-#2Ny}aaz7wBC|Vrwg@0b8m;_ggT;SaE)6|s
z(}er9lU*5&8I@kVdj5QK-lcTinn7dTQ=1L=_TszgW1~;8D#rCsg@mZ#Bmjp*8LKvv
z=eNHAJIg9`2mojk+G_<h?0=vcfKocEUbhoDAiDJJ*DvM$GGpF7g~<)=jZAr%85+-@
z{|~L25?KM3;-wc&H_t?52tTOlRue5n?Eu3neAi9L0*;?lj2rs7wi;Ky1{2;JwM<29
z#&&O~09J%yp}o!bH=p#^#4+{;dy60@C7#Ye7&&IF`3h}{T*AnRGKN0^-hnAH(Uye{
zdU|>-8`j;)iH6;@^Y#;2{>2j}fmyD~Srlt(=*(HwE=0{;e&BU=eW6>@JPz&D0;UOR
zMAp#=v5lZ6haFctg+i!3y}iG=Q0`or);YCz9K=eg>$-4Nq#D0EordB$+vUP6|2cp4
zZRqt^cZU!7b7WJ(uU>5jUUd`D4&{*F1?3%3e|J7*T-U08BWo@OCT3{_{EctlzTIiq
z6pjhgCCURTynQ%jP_19Unn(~S9O1!;sGex<xu0kdHjUL@Z>5(mJdyM^woZ(<jidLr
zG#>lff9(27puZcnK$k0@&*24XS38=UKP>GNRKtRTrb3*izS;8j?OgPPtaNI=W7zk$
ze0bky8k6~l$IADiTWV}tBgf9AKU!?;+NJJdKlK<w*@qQF7}Rs*hruz)mlsKX;%)2v
ztnctYE`ZYB%uQrb<=S6Kwe7E1$?i31RGYAyY0jtDoTr<}LzKLaf6lKZs=qo^OlzbW
z<Qy|~Q&<_NH8t?5b%Yfg4piiF1Iubglve)ope7Ji6Ukp?bZPu_gWgS*8K`i(X+X)i
zvluba7>PEC*rvp>JN2Z|(U@fqaQ_x2cF3GQ;HVO*$%My0&Ekr;Fwst#Hf2iJ>L@F{
zRD-U&Zg)}f<NRzKcllQS+Z73)!J>~vM6j`<8N_*ZNG3~S;cF{gdiu!L0k5H*6v<Ew
zdo#8gyLqBiu04{fi+fqnG7zi8+l*3+cV-#UT5{pDv-TStxv#RqkmpENUI)@DAADxE
zC+Bk}{J9G`AD!CPr-m|Oy}iN-uxP^??OyVLeC|Qg0n&i3^Pl_iz7$+E*0Zn)#_V|6
z*g+#`;wjdOs6<B1nbQr8F^8JOLzh6=fGW1>)#-t7A7}FGPM<%2Q)uJFoYVsMhNYlA
z?Cmk8^kjZ65K#eTqPk9^4i9Mr52^iX-;#xD6GP^t))<ege+or^3)Wnd2R2SO7y|;h
z=0~REW^MNt9R~U$y;7WWj4p$KJa*G~c_ibhm{^l}gPd+|ZM_N1ov?vdzM5NXd-2)@
z_S?dTZZEqE75Sjw8B~_7I(L5U=0jiCL`^@LNviVEejdF!j1vDlWE32!t?o{g;8bOu
z%6VaXnza1B37JEEbjg?S|6W>xO;|FC>gBwCeqfP13qc?YHxW8GSP9v(V98a5LdDNy
z-@u;#{;Qfj!?PooO@u@Rd{AQ_AD>%GZ#x?jU}wRele<U(?F}7S)wtpN8H6j9$IoE2
zi8q&AsSX)4aTAmMd~Y8d>7_p#Ad9A{vEUFUNgenznlo=X8_E_i)b8u<%^d}kk9Wi5
zLH4#oBAktSQ2scgBoLR+NR;xalV?BIeoN6R>*2|wQnX8Hz#D~#vDK)HFHvmdfqW|F
zSDaZXs4<RcOS~D4X61RY1aAgnlw}*AJ}urqK}S}3gHTIw=4>qNf1tc*w^_I`Ll(-9
zO=-+f|Anf(4eC{sZf(5`_cx`b62M--8oG>SKa13e2sP_eu4xy7JfM!7h@C?6#}z`7
zB5--3+#EGn6;EWoqOt-FRDR@{^Zw>(^bBKbVaRR)Vk+Pn*LyhlOXl25;77-!qg4O}
z^zc^kr(C7M&4+F#opWl)G$(sC&%EX8HJBtRD<FDprn>x8?wu%75BmP6r*cM9?gddG
zLqh~z_wPC;u;J@PFa5V~Psfs_&-};Q5bvV%(&li^7!`;<3!AVy$Lel)@J?AlDk7w7
zNxw2CFzr@Q*OcB(sf{;N>mR*-y^~w)kT++Q(djYw6mF4d>$@#)r0MMI=e6FG&Z*=5
z%h(+x#B@+dQ8}zr`}aQ85r=44E=)a#CQ&dF^7W{a+&1d!MZo6`6bzBXBy(^_c_8Xc
z={=;BN5n$Bry~TVFxT57n8%gHuLFs}g)jq@9{!wN!gUDRx9_>z3~&Z1d3gT~V{^8n
zP;V)*+b(2#=xjrY5bBW}5`elC$SmS2uh}T)8PQMHV6vRL5wH+DeU3cXLOUe#XxZjz
z0gk%sw)So(#nkc7KUXZ@(yV>`2rK_Pcg%!qM9F>n%$W^2AM2;~CMqqt+Cdfo5xhN~
zOdPOj_v$XR0vAlK)x%`MXNUfnj>YTVBP&dnoc8kNRK74Xf&2kUKbk=i+BgUN73H|@
z^rZJE4}hqt__<4tsq>CzeYCZ88_B-9d&~wNniMzP;8yVxajP7QYCt8H#3u+F81wB5
z<8sih7xBJ03eVwbiAd(`*|QB7T05I3?P>YA4MiDYt%!#<kn#VPaCLfW_<aJ0Wf~;t
ze`v7)j2ISNGHLnZ;@88WIW#A~_Uzg7hT*J>ZJFl)^@5~qnr^U#-Ay!zts2&)6`-K)
zHNo<nm0l|>Es;t)!K4CKF*`M4Y%Al2?b>P`&hPf<3@7^2wXe-7ME=0eada4cR(^T2
zvzV7^uxuN2Y7hLazArz`A-Lsx#rd|1q*qQo1vZ@`=YVtnr=$aKI>Gi~YA$#IH*uTT
zv>+<+N}jiLhXIb&O5BFEbrkZ;2%y>72LMifgF6Z75+u|D?>#0@o3{Pe;-tAQ(uf@t
z$+F$hH6}jKkoCwY2}mSH4ElRp#$LT@fBH-h6+b`)mTw?GrS#6P@SX`Of9VB4$|EE>
zRDASkFi}W=Mp4<bSvYAI=dMXx+}KV(#3640-{|Kg6P0Tq;Xz(gAMYSLUi<cVyl&5)
zn_hn!Lv9h&CdVzIr}I?~hs@@0RB+5`uf8ph`|~(%M(cO-`DTH_MT341xV>RrCFAJ)
zbFr^(&}(j^)>cx~@U>?%wIW}+#-2xJ-U?gK&9|M6!7fl}weR0<sNJ7~5RyJCg#;;u
zJ1n9OI76n+p^%qHfgng0jegV_S2VCWMH~mEtwa#CInYROzdj)kWLIYIrpyl+Fz}bQ
zN+T*o{)@s$D=4|N;y5>rk_^@Z1vO;{)eO$GbtM64hnxt%++1^`Nx)WZujPYhXde#w
z))j8ta9WxM)FcdH@NOIW<yKeCj-3LFE>K7uo0J9Y3Ch_%V&|VbF_BgU;80d2aBJFb
zrt4kz;N1p{Zlw6;^c}^K!Xe65H|yc|r!%;)aw!<O)Oi@%^g4aaDA#u_be;mIl&e}C
zMrlRMl!jJn@N^r4y|c+_ZHKJThg8KT@AfNv_=mzBfZ8Zn&Y3KY{t;))b-SnqXsY;i
z2^tHN_hmPi83yW<EYl6tNxPilO;=V)Y9v3<tLjFO=r^?8<I=^ajGOtK%}q3c%rWSP
zWiyXU-_V`KYpMzwK*?^b;E>JV>_%36qY{4VRFQ-S45`{s&c){R+`W+ac{(a;i)F+{
znnXw`<zQ0<@ulY%#2yxmfwIMvWn<vN!)vN$4+SV`^<0azypjv7YhW;%f5tKgzmgKy
zf^A!_aQ9(*3I;()oQ*x=$I<Vv%y1;ytR60VV({-t_|k-A*EF?D%Z5hB%yC6xJ;-Kl
z7gMcPlAvRA3YHBXrE<GzsuLxAoJ6ku`(HF!wUrpYc2pyW$EWxyvgxphF}q^wf|B99
z*N=lc-Me>>$8gF0ci@n+hSN^?3pSxXmuaLl&m4r5XnkHRWjD=-hK2@L?$P7NK>r)X
z6c8894m{&g&&Ww-@udbsX*lSNWd!5;Q^oj}(QL#0;w{K_3X!W*&6O?tL9;Dp(7YAA
z0XEP~lySI$>2&}3M>eR)zdtiADzkWfmg^?|%5!Uw;lU3j<mqSL_*4mOb?nM{n={P@
z1%O@28QD2jn?E}ekjeMi%j-s^8y6r!Nd~ojeLj!E&@j$o{oZfwfE1A?@@(W2Qm?V4
zF$I+QWzPgu=rdP5htE3`GMP_DkhhkGE_pz<0?W*td2Hu&q8)z$cpDq9>y;&|$yS7G
zz;y7SGvzH5g0r1`W1#75#E%9~8MwTpH7w*n$ZX5)7Ys2lXfCi#HyxePaSxb!!A6&X
zJWH0O7%EuRG{b{p+SoF?cnK53plx&P0lku5x7a)r(_S7pqrr{eL+031C8#j=rg0V1
zjdB<>SI+#vM5+(tK==hMyc4f?_f(fmwO-{~*aN{VMZcE+^+$5F{@e$Lw^N>8%3A`H
z`39qM#pOwA?(~e7iMqV<um*l6S~5gT#bM&n!-tA3sq5%i`u6SXTQrbxO36bkJ;9<?
zCUf1+LyTEbiYK0??=y?p)ZU%1PB%C>Y?T`GPk9xTZX6l~C$-^MpSn;1jhKj2NCZI?
z1a@}su;t`Q9ri(qHMu-w9#4Yv8REvhxvB=kk4zr>AG3j~-)j`-#yM<AQ6xHawwxkb
zSf=qez$}EE4rp8iLox;NsCL>twzXyml*@BrmRNbStSj%($b9j^Z*sw^KQ3Oo;mP~D
z3ORSo{6CKV<(23lPMmSW*Va^D_}5En-#3pLaw4YgFnL%->s+@1C8ec(KtmW78|$k9
z+W-Soh|PjTFQ$2izKn19KD+8mLn<j_<5<2=*cyWoDbD2;P|%^33iw6AwsC^aC=Mnd
zo%p|#LXSp9npGJ*31L>K5>F^FZhvPlHH~T#^CEg&oCkDELH;*z92*sPqSxo%A-&sy
zx40Tw>P3}e5xbAw0|p5nqF2nb<-Yn&+MVAm@qW=~SwRPor=;L>7BVzn`uxcjEQi*S
z#S}(NBi!N&HV@yvNB)S;sLf4Nu-e$Vwz{aZ!k674m;(vt9`9@E%19$9!?uKL07F<3
zPjN#F;9n|;%5(eltq+|dD5CK2X-AShI?yCYt4{9dbYQ?@w*h*wJ=g4ZqfjRduO5EH
z{k!to5f4L$zEzEf6iX2EPVB@hN9~c0*(QFLy|%qNN2kYk+zjRZMuGd~fK`Mb`I<;k
zyn|M4qX=;_ZS+dlxp!Odg)^9;WmX6wj2?UaUw{GeP5@0~%(OD%TNS;^Cda49rkGuC
z>=Y1l>a(N!swHciYp2Zl7gS)k*_^s3{5ou&*=27g6@*qtGn31Qs=8`?K2q<TP5Xke
zT2VtXHm10C>8#3*)c~eOtATwKzLAl76ZZA4Lq#IBJ@n^#n~c}_`!~kgvjrJuO~Uer
zeb%*noKwAK1Uv@0+?x9{TQq$?ftmv52lj88dov3ve@Qn*cw$tvrRSrEHeFi!W|x}s
z|I=r!BjCOMv?ZgcQM-;!jjAjroO%Kn2*2N=ZO4P0de$NB?EmEbvRO3XRSaxX1D|aV
z2uKA(6NO000$%O=mfgHg?93>-8kX_3UzqQ+`;+FCvR`C?^g(r%>hx~9K~gY?!J^8x
z#6<oJQ!8bK;2zikI4(JxI8H0bNL!_%wC$>0Fqu^}xR9s+A9-&cm-D{9@jm9vvyvf0
z=Alr?P{v4xL`52wDU>D+LX#nuWrz$>Qf8G3Nt#Q`3dvki8c4GgNpqd|Cwu+&I(z@l
z>zqH%`Rmxd_Nzrb&-eKZ_i)|Uechw_nb#<E%BtO=C_}>eQFF;X=YAyY#2L{>mQf13
zUggd^-f{o=)nnTwDz>q&w~Q*A%EkGFlO<hX;^{)#@c`1867mONADwj4n4ZCSc7{6i
z`d$+|ePl)GpS|a`S390Z2s>ZB0G{H#z_|m)_mIMh2xoUqdoBS@VI@U<WIVOj1e+f|
zcz?~=h1%M^2s?<Y3=Ms>0%sYxxUZMtA`kT!>NH$L5&r?lcRYXMSRvpH3>}eRl4q6O
zKeAp1mGq4e$V8aoqXLe7Rr7T2v3Z6*_uuwCyV-46lXF~LVij|6#xpy;<}SS!Yxr*b
zm+#QAhYRRXLX)S6ofTA`XJ6M11g)%dx)8IF!j=!|^#u9CE5``W(@V{F${OCwbt#hQ
zm%doIao;MIc;qP{N%j2hgR*9i>|);~anAUD-HXcN0G|Y50?X@n|9(HBd{T=8oC)}^
zg!n=gQ&2y^ua_-;cuC=SNACWniXAh27}(dlS0p9^r`a{?7REHQ>>ys06!`h&@3oGg
zf7J5SsW^MS7;|RH$<wFXVp`8HZSS2j@@>Y1@#DYT8^mDgheaPm0FLz{ttOdVbI81J
zk~J@$D3T>nNyv(2>Xw0Bst%PylXrhA;x1qJTvFj-MEX(jc(}V%RUl@|Xm|0o+i_<z
z%j}`kuYC|p6@%zX^@y-OLxa=|7&tu;`j1;uUQ6(sqMROY5k4Fnk9!g!GKWhH$qqQR
zLMmc3bR76^^S5pDKk<^vqCu40VMy5q6Fsduwr$<x@B)Hb<{ps_^*GX8Ub=nJl^v<3
z%9tj|8IlNtZf7?W#Rvlh<^#$~&>ZpRRl@k>PFHM3Ej{NdqHJLGvZoX7J+dD7G&x!5
z3W47kpF1Ugjw+5IxrkCCCd((-U0gA#5e^z5LhVri*pi?d=1q8*`G|VF+pI>TJp~+z
z@OYG=UTv`kIv)DDnclLtLYZGgkdk=RlK0a$^9kTx9neu&SBwjVyuPHomU;nud(6wY
z`+obp$(}2U+PfdGs&xdv6$}Bs_$DfgoP#8s56x`7BHP$D2iG&5^$^{B<UN<RA7}R=
za8q_9!xHft2sb2%3!7!6+GYenj}%ad@FSuwL9&C(x+K4H3wG;|BS)+$41uqcV7MKd
z#yKEKO#(YC>e7x6;(3(NBB_iUL{q|jt48ouU|oZr_V2mJ0=t@1s*qRx!KQnosel<5
zyd+~>N!5+kpn;Oxl(~^{>mjZ_1z9fo&@)Q}iga&W>Kv)V77(bEm6gfE76K_^Eku;>
zr+iZO!k_ugyJL*NriPv;5$NnTujhU9o`)B(;bb%<r0hSD7w`!b7p${Hf6;SrS7H_^
zY3BnI6BUQ%Osgqe4$?ugEh;B@poVdN9c@v=WPz;deI7W|zRU6(Q@-o~AQBt|Rz-b;
zj6elPbF>HNQ=_L%@lv;Zcz?Gkn|1Tdy0`6=RikuVO*>3`z5H?(=YZ`OAfZK6T#4*E
zQVkP8GQFCA5qRwf$9|iZ_KB*+>VTn3D9nqu6sS983&>c=wkV?LC*nAc=gDp-fAnO>
zm~M4JWk1E3PQ$G-N7~e`j2fSO@aVJrMIm*Ls_T_Rl#EARGM1bqB)u~+W9a+wr}yTy
zYuyS`4;9m)mj)O}`>vM(rT7RJ_5@A5U<(2<gO?NQxAGoqCQ5YVpIRj1grwa(5s-ff
zTrcKylR+Ci{yKu3H}V+!zUa>QBcm=l2z0PjqM}3o8NC?;-&fMI<45hSg$YH6O7&&|
z8AugNlJ7bIW!*vZjL1oV7X*hQ5Bp*tr4_IqI7x(2$UzJ)KA|CZ=l%8v3>M7D$|i#s
zSTpxsoX)`rtXTLwU*5jk@jNX(y+fdHXaLAkTPJSp(<>q&UjU8MkGDy(CJlIa;Oh9~
zdGfsya1$sLF))i@12CwQ+q}iGOPE8Y;-u=^qmA14w8Bca)9Z3aMV(o4jjKe}s^`ex
z@)i}Dn3xR1H*wR2&RI*z#a(M`@D%%~K|K9Kd*-9sO}`VcX1uzQ(O`((uIO>UTx`hR
zJBgS(DTV2@f_iU~+wBH`JWa=UtQs%w`Py^F4zVbQXVIRLG<Z)JQ9UO@767SepF`9K
zB2RbfZ&d5{p|3di=6lWdY6sUVI@Q?<k_IvVKBJIBggQxiG-|fuB~?%l?jh$`(TbAf
z@x_NSuZdBb@K*ro1piMX;ePm&{O|V3ZC89x?Ok)=Iwk>jW_qa`UoQ%KH+`3OUSL*H
zhJ$+(CqnL6%DU@!+{vwK&x8~Gw5>|0M+iPkd9@&G^^m$3XbjG<=Byhr!N@5aW!htu
zyCmMzEO@%<?!@F!>wy9?IG%UlaQd%&k#aHoj7^i*K9)BlUIeaP1iD2ME=r7-KDSoe
z_h2}9nU0PQT&zV{{3V*2E7z^d+;$X!4w_dMF=G*eyWQOULENkp7w3?}j-xm|V|Ke@
z@a(=Lr;05iW6spO1g<sosyF0{7Nio=fy5QQLynNSCubFAhj?G$gkqJ&jMjzI^<N8s
z0I|x{ul4{PxYqd~4v@wK+f>-*pY<!F1)xRw&SkDUerDdj-mWJee_nKh+(fa<@P_>9
z%WFE%k$L~r`of9r`KPWo+-=yn=LlQ0X8ghQ(!QVYxj+OYjkw)|i64s;PiETBPN)Z#
zk?LsRHg9M{&d*z__}?>2+J)(NJ*M7<FTR2DgSDn3ij)GT-26*gwvlQ)DpHw;nI1L-
ziwvKyY;x__>Uu!YLx5)=33){tWoM@h2W{pDql<`|u-Gi08*svKMpG5^#7*~FNwDA)
z?B^#DeOMAugx^4ndmaZPkQzGZJd$M?08uR1n>SqyXZ{oSbe#N2FkING0n;r03~^~R
zTVmDY=zwY6=9u*grEs`AM)hG&yI{%-@M(u?WaeHqd5~FiV5?>C!wZILY;0=o95bOt
zRr0XN_)uais`XTFK~5cJU9`{q%jJ6Qfl!tU{sy=aNJ_$QLc0<?gR+V{-i%82=sum+
z-wUQN)houa%<!6a$FrM>pP8bw7osEHJdcc?749}*{%x36Lodxle+hbN*t+AsTZ8=l
zH6Dma3J?rYoR+$a!cJz;v5-gSFgiypDvT)8a$)JmiPd>pW}hdHS8#KQUR+7_AQzTK
zJ|l}Uj)KU7U!r6WgRFf+@d}rRv#UP{=p6!r_TBxGIgcMF?X~iXRKKzpRKp`Y!87-&
z#ZCojuZGPs>N0mS?LR9RXAf~WpBtCp8<Bf5-m$-1?&~fI*SgkB?rOdfA2#&GsD4M_
zd-~}Ho+MD_Aq9^Ph<72}d%~FmUf~U+vyYQL3kHv_7!4;lhHU+TRMWB@au5PVsT8*A
zVolBY5K3eY_B1WOINiE6U5blH#{i?CgpBWZL`!?v2i5Kq3t#Gzns9F>A@D*A!?1DV
zP*;jfo-VMx#h2SE6t6<fq9-Aw<Tpsa=VtG*KQMe4mTa-!pBpCq4S}+<g6Rm<PyrAk
zDiz#d;lX-irX{Wf7o+?9vD`E4{dO>4$9MH$U3h=@B<W~p3BWyHnC_v&4Igc=Ui@0w
zp{;3!8?zYBkjQs-H}gK%^`h<36$2}kZEFud5h0)T^npXARo4zM(46*QqJnXtr*F-X
zko*A`ipn}A-+lP|I+$uO;I<D)R1|lDX`46dPp`NTnPmaaJG6=8SJT*Z;;OCH`%bDR
zKrFg}2uQg#mf1&H|3Y9JE*lqGg+HR^GG+{F8Xd)2QaS|g0dk&Lv#Y&VZqN6xMm|_A
z)x6eyYJ%FQ1>U*oYO>*KHMfp#&35+NY218&#NoOnKVmM+Zy*n>fL79aYvXSd3?r2`
zSPBx#0RtHXtQcd`mceD?iMqgph59La=EGlKI_dl{?^`atBk;LVk2IVZ2ly|Lwe{u1
z=b{#}M=91SATW^IFRktD>bv-u@9JYA%_St3;hB87B~@xHqj{pC?50xd=0}qsw70Z+
zcy`5WmSE_&F@Wc~fr)32OI`djetQ2eJ0cG3=<N9V^yal6T~v#2arjy=r}c_5kxL(k
z=YMm^-4Ov(!z~Ud-Y6t|o%<FU_0jL+jb{Xe?U|I*9}_QbE3qf#Qo@BHtQp*8MD-6E
zOpMF5>o)b3^T-j+ZI&r*SUa_3l!|-7$=U|@)30t#%Z&Sax~aIUxO4mRrVB7;`XEw;
zU8L4qH|-aP>~T99p8jGuV8!#9<q}C#RfvnIU*a3wtv1QO#_tP07JbtnU_+7O9kyMZ
z_0Za2vB&-uciWG$-rsK2-dg>84aQ9_v2W+D=(M~~eWNom;Q@mN=eQHMSB^b1bn?$O
zU;pL;Od>uYy&gj80o-mO;X5Vf_Q&()d4whm{~c7B%Y!pHP!~p!38K!a=L=|C=KQl!
z7i$mftoVKN@x1uE+aEtJ4EBg{e0U+}2GPgl^S2Bq7-QEl^FGT*$p_YIDji=AJk=|v
z=;orjKWU-S(%0|gXV>;6!w7~A6n1*vMA1$o@B@J?fIJEo@m|7$p(Ts<Eto>qF4P=h
zHvI*tYK=RyE460PX&tb5S1;d7<6WKqY&V6OiVOsXa!vGw%f94nZnZ2k_iiW0$*1IE
zo*H*wZt}fjO76*+hUNx0HXqY8ckG#cn}IBzxDT_B%9`i2maa>p*4yI=xLXk_-<s)g
z+RW5+tl(T~VqGNh#kPWYMJb8HK_NuAJN+O!HpG|gF}_It%YJebk#XW{eRUv_6F?Mq
zY0{?0Jt(>fNd?g#eK_}i>E}iLE3dVg*jFu51!(5LRaN7$!!s(%rQMTy7i-lGY<}P7
zW!d);`;P%KBw4Sl&n_%=t5nloIN<!!eyQb7O)_u1ow^*)8}@zpaFL#c>IVsLf1niF
z+$f)33%Pb(v>eZ0!=GU~A2M)*j_;bxoF?_&L0zrT7@I@X7nR90?+@!+`{qq=(HEo?
zE8^e=aP#t?^_-2kH?MN|q*<LChH+S2*UNm^QywGGcF8@9BQ5miUh0AHwK$c6bF|?O
z1AZUo@35=iRioY6e3Wrs^-8Pa4#etsvB~ETM21cEUMACjfMZ8WN*G1K`6VkS5&&Tb
zC=tEJ^NQQ54(-~IGkR@E`ksW|iLG391N%N5ME8@R_MpW}3Me)6z5atu)>qs5dVQ^|
zg@QKl<L`A0W`KHg?5UvjKJ>38)F;Dv*t2Rj%-gsexicYg<PI65qp}9ySyepTjVu9V
zXX&x6=w%&9kDHd$4EuVRih@Y(tHGK&`$SR>qCPw3VNVfoeb(l1i4G;92nSv$sBufa
zc5Cfup8oaIsVJ&lf8m;iz*6_3+P=aSC3HhzB{BcP@*9d7sqmyX+>vU=Twgk{H?ypp
z#^?KVK^*Xk?RKb*eN@c4;bS*XUi&UGjK`)pKy^2&#;2m09xB#dN_jgyRb=-=3@(RR
zj1dTgu>ko45VG#ndz0XqI#p-=XvYu$ss3%D4M5;qf_0L*;c{xPIs2YvG<AB!Or-x>
z?CK(mZt9MM-km;udby|BI(<Qvz-{2~ACk|&!k9l&Gf#77#RxfxRG8WR-jUL~Gj0&C
z%)8_Yw!<P5xPY26O&1^vU$XoL^5hGNV}9&*db=WafjD_er*mWG3{2ucFsQxlp>BhO
zU`$&;cEFa6=wwUB(EL^X$xYg*$o703rcUH%MTJFX3=W5FcGjx#>^h6Nj5KKb%m1#B
zowQ-W)%inug1%mKY1$=Nbkqg}Ec8}>XH+7EKRQ|9zWgh=27sQYzQ`<0pFuY#k|uT8
zFHq{un}t1<ftV_o75aw-(6@eo!TnZSm$jYJrMGq^ogDLrojfo|XTq5!mr_kRv6t_*
zAHi}@ey6psx^-Vzc=Ww&+gX3j{%FmU0g?J9L*7HSCJ1eHZ}|@^>f(#0Q}f~%c8u9~
zDvHb~f%b{MsZR_ke$yb;@yQ3^Q<_ig0^csAg9WAt-Cu6P6+(BxbauZ?0SD-Je#h+2
zf8HH0C?3?-1boN_)g&7?z1~>rLRXkomR)+)J7%T8h!8Rz!8u&Zp63^e13+YVf%g)*
zk;GLn3aEz<nppo@Q^8I@+1}4ODoaFm3m3XsZJRkcd^p%N2}JLS1li`(_CZzSq;FXH
zS6ORDS+t^Srblq~pYrV!+JFDE&nzsl;DD4?wU&VO?7eW|g47~mXnM@st*y7kuj{?g
z7yDG`t|3&5jvh#5=;qCv@tmosuw`KVKn+8s_b0sUuu_yC<n_e~ejplYD4EVF?GjD5
z{U25fcrlWjQKI__YT+?knb}V{M3*nmTl^V})DPpi&Rtoi3#wIr!ntxo7t2q>jrgs{
z>l>pJG<Tn-6ib_nTPY;+t8ZSkJlx{AABAeo{YBGn_*wAd-X0Gewe@bisoHU%<=qUg
zw6#>L6y;1v5_2eDFd#+b5~350dRx}uj(`Pik38!2GnU{PR<^PB(S*(NKz3KP{jCVy
zHo&kIO=2X_f@(N-<~~@hd)J>kQRY4}p?ARVk3O|>k}bGKNMM@NA;H)3@Pb|mi-#R*
z9=}-L<K~5}H$ykhQ@XUV^`PF5<4>PB@$&0bh?y5I9EF%ETn0x2b_c*~&)-}2_o~_H
z+_EjCk8aHaWMxrEaHXw#Jt}BarOm@*r^hl3Z90lP@9HfWl-pv<?>G=_M)i!Okp3qo
z7XHBH4XfK6(`rxF+NaaPtBJqMoL{OQYP4Uq^N&Yr4U>juEU5q4#&+qDHkxD&^xw~0
zGM;{mH`YF4ciMCs;ZWa^x!v?J%J7m0j6ymjD&qTu0*?Uvgc=h{jY;HFgNwDI@V8z|
z_x(DWW+=doU~&5vE4J*-#E!hyVepvyQ4BHo2s7y+EXi;9seWh)%r`U~L@Uh<!*l`!
z<e?wmt|s~Dw){q=$A--<ycY>9h>zbKCT??T)7NFj{p9>VY)AJT+V6<ls_(N0uecGq
zwYv$GCaj98ufpp~3#d=^Uj8cX?fuPu(9FONfEQQp(w{M?x4wpaFSth!vc4`p)T^hH
z^-{l;2`#7Nz=1^_vxofCVRpHu-GA@z`c%T!Q(qNq4H5v)j@j!hW5V9ldwY3J95eXr
z=J~PDK!ZQ3{$kroei2fkk`Vku*}=>cEr9?6EAS)_y1G`VHNOkCsN5X`!28=;-3Jq>
z0~A&gq}cDtN|^M$dS~C^$#*#ggdFSBQ<`&D9tF!-mnj!5SKDh)!lWGoI}iiUI-570
zT(euaa!u{N^a-rP=LMNE`+dFRjNw^_q_<(l$;QW}1EMWdzm4HvtJ)?x49$Fz^f=^3
zpiE-Ws1KFxl4LB@?E76w_{}xdlqyZp4mJCo>t<%ScS$@D?LPfDT!*r>Jux>_-OV>b
zn}tgc(*E#hp7i%!Uy2$XE=@YPZiSw2Qf~Fiota0iCbe(f3c4jljD*rEBipDn9dGL~
zlkCF{qt3|yX~RbcMJL|=Xdl&KdsIJ**;(f4+@c0&m+<>)HSgXa6fntLeZ@VqL66?8
zjS`HRix<1WOaHR}qB;P{o3Z1@C0Gx{04vn3nWg1!qwT-?UAIGtuksq*nj9~z@?TAt
z3m`Jd=GN`ok9sxDY!1J&QP;RSXSnqR>Pi60fXhUj{MgELmv92Q%rgE{x@y;#kTU(9
zKd<22=3VNMU2hS-XU;t{^UR}1j-A}RRzR&P-U)Am?FuWm64yAJRm>W;zugFj{S}rM
zny4Ohrr`3h!=1PQSOt{%+Dm@1$aovxQta+AHNUZEmVBHl5wW987rHunmaLsRZ5pI=
z=3Pm3T-Y=T5bhU6m1t8SdNXimV4M_tB(5nY_}so}($RwhQ#-YUPknlAq@*_QlGK1J
z6hZ*c-V{gY$~dXY(YdN5yX<jf;-klw^B>66_S|p2@pGP3xLlBKpw<AD>F9-Y!XWj!
z(gLn_TRW}U0h%y!3=*2x_Tlj<)A0mLj@tGaPUT)dNV?BPP0h|AU5k%$uPnl&439Sa
zJe2AFaf9s)pNFIW%$2E~@!IDR&q>G7kTXjJAWv1|&YFR}t^mm`xmGe+y=hlX^_q&G
zhmK6@A>{|EGQ(%$=a}sG&fCF`wEf#A-<@|91cn4GQD%R5ocU2eO`ZW(kNTCBGAXhH
zc>>KtzKfbH7ft#ezH8QC+O+J=)pu^_VicbE5h#|My*~43u3csQB!i;~WnP(O{dzU>
za_Je$hZ3!yn5*kCOZ3mjmT@#!@nNoNZ+j=Rzj~vKd3t1~$$=v)DZRmjyrgBI71b{|
z<#%y#kABSPa(IClCnx;zfi8VOyC;&&k|a&6k(tclb3EFxb}Pt@ZeZA(JOd8dp8DxX
zLQ({EGrtuQQhu_je$I^N-;zx4tocydSlwiQ;Ru9r>a#z6-*5k#W*vM*W=~|Z^HB4@
zuKmcN(G;eS!c9Yzm>$Gc<<y#UcsMB!?}Nup_jzt)o<5^AZfAH((UYbq03{A8#nc7E
zif<d{sl909Z&(_TB5cl+O`NzE3X~W*cCc6U+_Vs5x#Au({&XK-pmz0_iEFXVJto)u
zCEtA>(I?(xhy>(q)WcW5{sp7Y57u>FQt|T=X!R$8vjke+=UZ@NL=zdbx_PE)kRlK;
z?T$7`8;K|!6RSJPP7%M^^6R+u=F7dhN(0WXM}Y2O)tn&`NezC&I(b64RCowHT!<`E
zyT%*mc`8V~fF<MW_`M_A;-|X#ioy&MCJrmMxBkcLZ93uCwtQjH)?xQ^GgXe4{;b1s
zXFKF|6;IKj`(St$DpcfgAfi#A1y@c^Zj-3!IdJAbU!;N;Y3GrgUqe%gLdhNG8YJLt
z9%g%<K7D}xo)Me+h??KupW%QlYq)Z;eB`~?$=Gg4brQ+J_4Cq3om>pi;8qINhQH>d
zj!)j#Qkf+>huQ&wlFJS^5(}O?efgRdTvq*{gCNrud2}yyA)ye#gg3A?zi`$(YH*XK
zm9X%JM_~f(o#U4%^Jv4JigQ+zzL$O-t$ud%%%iW$s}CJpNrS9>FSlbWG4KAX_vJmf
zU4OD!&ETkH2&{RMD!V-~MnWZJKuuX;9<)7Jj1oVs(Fe*td3Ha3GH*I?<HxDM*j2Vw
z`goY;i=B_3baAXO%d)TEzwxv2PWJS}U!JRrOO#n5;k_{S_J2J3bV6qKVD}ya;x0bl
z@Yz^R^04&j-SI62a;!o^eMo5a%8FRa0ph-=m&O$F?2~9L{<&FvB(?2F6UwhxnRC&*
z9DW7lURY@-vr(_yj&j_B%eN;@?qW07T-_?_ukdy-LREc34840!{OicJDRq~H+o*MV
zYCVt(>weFyq5fW=ec6VLqt7$Ktb|wTtQu@rEt9tLXhPEKua1NFL>hKJekaNtmMoVC
z4{(5e`yAVTX#BP6wKExa<>q`;+y6`2od;JoKAtn{`M1VS5e9?rmqnf6rp|o+t@_UR
zaRb962W(tXl{4Jw!l}jb5{dt6*;h?_9#rqPlq;;sUwc;2Q88)XV!v@|_orvAYupLG
zK4(ae*Y2h6WwLP>FMf&BB(kylXnbXHnAxDP{xR=eJgIKiFPiAN^IfoA+O*@FqSCTt
zW}2o;=-U4nRy^TkNrx#P+}`#Y(smAvI19ROB=J-o_X;J|>BD+;L~G|Eq|^42HQ^S6
z1j|Kg(Sl*(PC-U$C28OsQxXB^{t-sHF@DW$iNuQHmB_Y!w^|_Ii~bo`m!?WziR+e&
zyV}FDHw5LpYD3cChQ&vp0gz_cpl-$U6LL0-IG_H<Ne_FTjJ-!;VV003aupKxkihl$
zJtqb&+P#Z)n(3NW9tx6l%xb|^g)ACEZu<L+o<G>U!v`&Qzw)u~Kkg7@C8Ga--zmCR
zsXL0w@!vo6ShBFA===Zszoq*75#ImfzlXNzGr~c{{eS<^Cf5zqfkOth?Y{;N5#>~g
z1TizfVHRS+yp0=&W2o6aSfQm=Eq?`T3)+)J$Q(D{*;)8%WomVElJTkY<2Mc6+qRd~
zkuQtq*!Orpci2|zgs{G~V}t6<;?@MF>6;;^RjM7<uOaefKhIBp=x27i+hNn8+gHZl
z?LDuTRLA#IPMxo>A3jQNn1<Ao@#jAj<_&xvk@z8^^88Gj@<N|<#{!$ay&hH{)Lms{
z6igHFlFKV@-$f<MN7YHcms9I~i8Q8twvX!DtJh_UwzFQQ?3H_ExptkA(M53U5gPQH
zEdl?qI+D8rdgy8f=g9k`obm^`g7`azHE=^c<gbx+KTytI6yI6$>(AUjF6sX3+a6b?
z99sSQe#qjU6|H}LJLkl(+;+ddbsdr~W!olq4+(Z1XDJLhSTbvWgrm?8^`Hbb4h-dl
zdHvmGH_wMx@8TJd@9>8OY>GB9e1cWmx8Fi*;RaF?*PoBmW>k<mvR>LgnJ!4UBOVc=
zX(4S|cqH%nfbtQAZJNLTY_g9qh%CFGSrYaY`88Wk<oR$F=^6^TPhY%9hdgkFj>xUx
z7>r_iMMb%a^x*!hOv`75M{HV|zUgDO);PR#&i6;o7F!2t#^t+YJ2(3q?msfE+HRw%
z?%})~V@JclB`M+MqjJp);CS9JZu;)|-C<TiGo?m7dYwXk;Dt&_?N>TDJTML~OP+ek
zBs;s17|shyZ@Xp}Ci?^#L}(*x_LgGE4=|o|2ahRT@PARy3l+Ou$u3KHR9o`E?8}}f
zwefL|VQ~sS``>&`9otlV4r9cBn$TG>sAw*987Uhu$yecdyPL~a-P-JwHPJCyx;|o(
z-T20{F14Q;Z`gMo9Q!W4YtY&e$6Z{$Z4Mf%Tp61)CsapE>o)=W{sq6_7){EVum+T8
z4R#mZ1fs!EkwzF(QKR`@df(~&$hXg=>!bbr;vHD<V~9ZYr|3NiC*8$8X(K$u{xDIL
zeJz_BYL!);>2Ya!Rpt@w${X=%nB%Sw$^|i>9dnKMX5>&gZ}~^WQL=$nLhR)VTQ?)m
z9Qcj$hb+TM)tQEO$GLI$M)+$&dg1-evZ)1?k0K;t+x`f<6_V*@S!=2`?fUU+Ip5`s
zW18OvCk*XT9~tbGsyMx=@sVCwz<Q<TSYH(f?>mBEyCW(~cJ}Uj7@bS#pT6Hdl`3>G
zttn@+EYg9ePoFxK7yj$qDg4~Ebm`H-d%jGfyL~C@S9deZV305wDIF44RjF~az6T^W
zxzySTTNX>*7~)cqWSxGZ+M{x>OFL7S5dE+v9!emc$grk9=*DO2ue9}qcs0_?<=ipu
z(~qb211f7j8MtOBJ<?B0FMAm780H(ZGU>Ht?ZpuHDW>VyC-6yBl>`6$R&Ev>pmVJt
zr@HdV_Ul~|UlQ?;)KGpd@^*jlo5cYaH2;#z{JGg%D|Aj-C2CyfUKp=gQxsPA>9uld
zUa<fDz=%$7o2GV|&qOVM`SY2JD&L%YAzj$|?e#HfcVjK2WBWdCTsmlg!?JJF&HgYq
z4Q_oW#DA@~+PEK$ZKG3!4dBv^&=|z<Q)?ecZJHi*X@2%>S*W#sKWl4kUMqIc`RyNX
z#i+SBEMy-!xe$0ld(5M@>dXG0KWpm<U;mA5LdFc<vh2>X6rCYIm6gpZx46an9=WBw
z3dV-VRhWkQA)jHi&eV9#n>OAxH@@Fdx$|4qoN!Cut4~iEgXfqUT2gVXgNS_k<jG4=
zVX;d^+&5jiTsDWkk4_srt}=Z~u6B;*^A1gYZ#L3H(7evWmU!^(uBA2;r7VusI$BSh
zWA3$KLTRb}$8~!@yr}ZuSRC${ELW};X7;A<?@op}UOJy@y6fr|)pvC8_NTp1Bggr%
zskY8-%B~SvU-i{)W@p@2$UNZHZiDYkpXD(K&v;z*ZTfM;E_e^bCmR)7?BNE1r>sJ>
z?FZ7$X)4t{OOlLz_i8#RsL0jm4f(lKL*p>r9HTU38LjA3y5oO&vWyp1LH~HN{BkF!
zG2(2<-MaO%x_8CYRTDimwX@X5&1h~@_|2iDvDNmXUpF<U{zKlkq7^Q6r;t!qqLr=+
z&8fet7>dJO<ndkmPhJ4Jz|j`A53z<Hw03SqAbB@*tSSuAv<DMzelC46Xa27>{Jk*-
zbFJ|jE7`Yq?^Ss-E_C)f^p|7obZNh-%kG@5vWuz+n)9Ww?(Ll)>zY0$SNk#&qOv+R
zcty%s>xK?9e;ujQi7RW<bW?h3pQ;S$v}cUozQ3Q<xBi1-7qNmU(tA@(_LplcS`-#M
zbJOD!otqSpQ-$WTiifK@<=e1;o^LIz1L3#qhmkc+Q(nIOb^SKDxImQgxJo-?YGk*<
zeFc~P!yx9=N5r-m#G3CIMDx0an2^#{p-TofCp_=uq_)M!WbD+bx*pguozfYAFdk#o
zZh3w$n>YWL;cm((Jh;Ay9<A48D9R{E%%-Z?^`gs75KfcAwaOjH*G?oE8vgP<)lI6q
zzF&pgiom8~6{4$iGjk1k%=(Yp&=BDpqr=VLwZiYXxcLdb^%(<NcDKvFb{7leaTV&A
zUj4|Aj#y2Ps{~PXn=@m#_ckqXcQVrHgFu`<EaX>!V%bHOx9GYYsWAg8`<46q<K=dS
z#dz+u@Jy}Q<D#(k?g9_a!CUpd(=T-1H-Dq5@rBM>-q+V8*F_rEoQ?Y#8eY1lZbG@#
zXKVE`S%;<%>Dd8zi{|TMaavhUp|5Sr@~=MU&#%p!0XzAuydhJ?o!_@yf!f`^zQ%oM
z2h~Dq;DBCi(2O0Qf4}<l=N0SLP9IW#8nur!)XjS#io;cSL`@H<UIW4=Ja3oqpIanX
z|6W>b(ayHtr(dz}I?JWZ%rfm(uOydM!bZ4;pXb1OA=)dy4fI+1gR#DT034Gg9w2}D
zzYNoPh3bg)GU616L7A#5&oPc`EVGR1_0?qa=JTVHGA6!0Sl9WpwO)FQOKpgVLv+-i
zaCXF~QR%y`y+88tzjw(o#@9*M=l|Iy<NAzs|J<dGG(TUV!x9?IHFz$@%Wtgmu%-WE
zHr3{Un>Sz5lV^HC;s}}>Nr_9Cg57qR$F}vB{Y%<+f57D$0<zV@MnvT_24^q)M&phW
zvQ2~VSQ+y9AoLlx*Y6+9AYlze^Y_0ckUJQ={X&vn;l96OJuh@`NcMDkSSw~_p@AQs
zrCYe0?OkE$yZ_(U)BMZEl`gTnn+NC%H{+?0?P8jq`_VlzEVkuijK7s#o!KQ(vGva{
z6%q6lmmVH<@#eQ!On50Qo>1mGO?+TU52V-ZJ@-WbcEi^>BQVpVDDS0JcGEzx<f2h(
zwTG`hwG1us3fxY=F)fT&cFj~%GbZuA85%=LCQYuXqJ!I^L+KHNON0xujQ-b??_0gL
zA=@t9&|_@K;t!15(ofGf=pn<0uA^9Jnf=pYR>0VHFS<`3>|C>+v6PV+8x?-4Y<>3a
z*{@Jw`$oAVb(N3mb6(Ck>zv#i-Pp7$s=c)F#b$+TCT9o*o_K}@yp67RUKD?^zw;{;
z4r6Kyyj!AA$R|Ft_vrF*^-N6z=$(Cc(|4G%?XR1a>|`OM9@P7oUY1KN?j^-%rXFVJ
zB^3ZWv>mBHfI(=$6iWs5X4I>1d&3x+u*yeU;{O$TRy8%jjU1uVq*T1YZF1qhc>l5s
zPEF0zKG!$@dsl?foM2pDs3}}e!;4JedaS<X$Htx(uBUpMacY2e<=6WD3I7hZRsQcH
z?1le`%w47b6`Akr(@NoA-?kmonzsFxKT_IlqyP2o|AWx|KVCR@%3$|Jc;eUHoh#bV
zl0y)<1PK1x#ot}eO#GV*u&bayOFZ>uN*js9^$~EI&#YbAWEH7YudWdRk4BFfEpoV{
zt0)@#c}q#8H&5x3C}lfbB1u8z{UX=;8iHGqX?9}pRu<ys&fEL)%oRb%2AB~7gb6L0
z>qwES-ItO~^cnuLt3;xthVf>kvoGw?wS~KPez1nH{Lq#T)>{9~Egx7(Vd*PgB$6@K
zue<DIrq&IzhrsA)13nYgq`Ft`&YObwBn{emlfBnkM*8CF7G{y9j~z}&rsK@Ti@_wd
z+fZ8BYtc_)(Ca#ygAO|Qa!Svizd7{j+4B+wvOof4>08x#a?NYRXOQ29%tR!u^o?1o
zd5YZDLWG6kn6or^m?6^@_eo2JXv}b|xxH)d;Ug(W>JvK1os+&OC<`mgVktASd>XGQ
zT^VkCFiry=u;a00US^JYv`%R2IX3gy_Y?AV^!w6Q?*+d{C>k|FKD%WdL-t$@_2YDM
zzBVTvq1Ux#O*AifCx>U*uM2u1-MjY{IHSuwYzbRnAnp37!wdVVRO+U5lB9Ix=)~L7
zxqMD~mN&B@Ve4y`IeR8wgUYSXdf}kz;0J69eIF1MipACN$g<F*#8T=QONe?d8TFUy
zzcj^@PANR*)mP%xn*-?j#3XbZX#%mLhy1b&(^oFP@g~{@xr7-G6@K#DFg1T1KOXzr
z0Qgp(@+1<PLx<Y!oWiJ6B~P0jJ7lw*c|9eU?3-&_`Kb&gy<9CKVx~W6dhmFc<(g}f
zJg(vwOj&V{t3=JZ%+!03d&eR?b9^w(2M(_aO%Z#FKs7t7=st%tURHLMSjG9Q;saWI
z{-Dikf(G)_7ccHE8B@8%ijwOq<jnQ+@&YCXkJRaxCEdGUKa+`32El(R_2*&wy?H4<
z+W<b>oMZHWT!TZW@tVA43@gR-lt#qmg6&G&DhsEM8#5-N$mW7v$-#pMr`~<UFW>e$
z`1Y3M<p(Ps<?UmWP{GR#`su>ZwevLpnmQd#{HVAn-(=G!gC8nT+|4G2Gks$31aT~N
zl&_k!b1gW?Ro4m>(xMq<BHxjAT1r?TufaZ?S}#LDN*}z93gazWO=!z+rT+S@1LC(t
z3S1N(*BB%}Mpf0wLkY4=<%WJ?{gq8mUwzs+c(aK~TEth#L(lT-3{G>qB-4%Oo9Rln
zE*HtZG({@K!pWs?Zn)E^7u=E~kTC`@hgDy0CV7yJ80Z*m7sF6h8{35h^#qT3$F7Sd
zANFg>#lgLa?3{ph$jUv*$dG&L!+x@XV}}i!hbxHunPCQn%G9ypK*!$C@{Joy^im7n
z&(7(+V%GYJYBXSGJN4wrwx-Fh`};23ENJk(H(|}~IfP!Gt7}h%GubE3<0|u$OPNR$
z@a-GJn(E%aa2}~a{JWAiAc02ctX{p?1L4`A=Ej`nRJZ^mG!XQl7d6VU&WksA8Lf0*
zEcc%#246Z&mq?P|g>G@jA&%0BM9~bN9o;!?7qSQ`R`%$4FfjaGyHa{K13@@oghm$G
zFo&@|GVl0I+4nnmxixO#rn(yXR9C}{{WQQqCwmeh&6|g;(N3&YabkD1?LW*U5=()g
zEDjiGo_S4EI{)+Mo0w1pi+!<b6W4@L!ls=hoj&0ptQB$wOWD!r-(t{*Mv-qG()i1Q
zUet2`6?Y4NP*qUoWd~D2WE?TRF0^Xa5s%1L$x7lg&s}aM8Iq`|1Y4T7S?qCE!Pscy
zrSTfw4$bE+NtX4<2@(7<N&R8C-qA4{zFi>HyQK*71#(jR$jSyFbX&q?;uy$HQaeF|
zv@By=Zs~7r|LuX)|5t42KO78?%AfxaJlFrvU)Yy+LC*PU4$luV=Jt^)$4q1nw(mCZ
zCW*?j^OHxsOkvwxrL{s^OI!=$`L~`$>VCBmz-adn$c2hJhQyX@;k+xGJ|M1_m{al^
z!HysF;*~OK1JK}<d{xn_`6NcE;BE|gn7(kqg137zOVdLHiZVh&0A55+PTxUJCAo5Y
z1J+)9syKDrI8E_VB*bBMrMlvkX=2j0w!Xt5y}~%mZB)f0At9s!fbBN-#NbLwhvvGa
zGr3BWE8r<FWQN3TZTKFIK-3!=j&#ebAcI!%K~3@2Bo9@)pF!^s9-Q2)7)k*6PS@=d
zc#PI#w3N-x5PO$E(h8}rBjtR9Oidubp0TZ=DJbVHw=uH0c6jg*`7wk$1Qv^2*R6`8
zV<!umRAl|vD4m>g6(h%Q3pHNDM+#%PjHh<zrB(Zk=cRQ5bp@n~i?3Lie$AwvBwk<2
z;iC>bn>WOjBQ)Y_Q!E?mri53D+#DCCcF(2Rxo^LIPxHP#F0(I1Fkm+l+I@QW-WXiZ
zbuI$a*b{SqCL%|Ssafvv11}8gy8db-?^xk;T`N?5!Ga?EPOK2w#bYc3jToz-;F*+n
zGAVB*%PP=S2>agd)-!x$87yotQO_syddq`h35=9CKkYNYD0+>^h_C(m=YqJ|)$l2P
zAh}ygP;v*=cdf!mu-uZ&@DM_+G5&YspTBsa=ipGY(6<n6f>&+2&l;kke|OVik?!R<
z@K`m=hM1)ttbsM-&p3y3;VAU1$|5qeQ$5(Sn?u-5dJS=r8XijYH<&YMB7Q<ZCPzXU
zqGRPxQty<`+Hdn4(pBK62N}m@Z!s4hZD+f6Df*jV`pwOVb&8TNMvf<quI9Di5x20j
z9absDo2f$6v8i=wu6Akj&ZAD?a`3*NWfBhDOwWTbKo~El@ftLc@h)S0?}>fOmfp_3
zAz_hXkj<G}n0`Yn&I~Dsq3%gnKv9ZLE*kubGep1eIC7hDLx;|Vcl6e0XSP2X;e4W(
zkjf34DEv+GBn&bxw4s{{q>ACAO9^{-<0(p0pVMu7-aUe2UFu|g&x@Wne&Sp>v&>2w
z-|hJtkz`=>z-s;eOdiLz?v81S{)qA8pBWr?cc*C|%PP6u{`%J254dXEm?Jy`Dqc)l
zGqW6Qk>59}XeO){k&gTrC8gD1?OXsmi&E%>kto^`smg1J5T33yQ^RJM(r*xw*3>w&
zu#9epzJmrChS;ZHN6S%M=+*URflq*mlmI?7&bY|jGKjdlvW$nCQ+CTJbjfQRGmWYv
zHR?pjiCR;kv{M)5aS};dyp&Pr<dl?D>P#!k=S55`Tw#em&4DBRo_f4xM7ycl$*7M9
z2MEw5FA^=GNu4)q=tu0>e5G?f44`u#gXOypp7kKxy(~4*GvdP8v-YpX^(T7pThF`d
zn%@}v;mp{=+@S6)*K@CH`ooLl95(v<pc%w)AD=#ceBw><>0c+Fwns`GVH+NGYW%WA
zZ53(jvCIrK8zKypm-wk#MCA9Yys;594e6CI(wTsuRVXr-1;nA`bytXi?p9l5E@aPb
zBAEPK^7@))jGP!<>vr&<SNN1EQwC%WW;qNV2Pg+7=L_UgplD{8tb%4S-J0sk&Gf1N
zvtze`aZ0d-9#<h*`J<=PH9HO6lK?SK!{Z0-fW_Qg=NLI=H=gnExGV}f9)lfnThSS{
zPCI{~35)X<@jrPD3Qzjc4%RX1>)vigK_rs0LrNy2hN3^Wc7&C<WW@?k`3c6c(fS`n
z@K%0%NlHf0x@kVst><}|4&0WBWWE#UP*!QHR!sPql^Ffs+8xaeyNzkLiCE-5_*jOR
ziu)`d`)>4lW8%Yy_n?5f9a!x=u<+UWE6m!Mr=8!hLC0+e;Q#_1dFqFv5CyEuu8!T3
z?F^$(Rw?o_tu+OcuMFkw9TUeYn?Vu8(j5Ev6%W8jRr@C&gBOr#xc2Q)q5rNWz!QP&
zO1<jYF3az|d-qP$*>RbTHz{~{#z2!~PX)cDOPB7?9Iex2YeBmyOr2BmUFB2_L`t;(
z{+V;<woXi++I;};CY$t!a_LTzL4U4CB)%b~cx=fL42^4nJ4eO2(mUr|GO9|q!OX5x
zvQ($2#}u&#SIvElw_a3xk$$b_>jx76ow!{&F!*i)uq$niDh#)VlV`Htmu(>}QyF-?
zy}81ySsx67I~~Py<{IkyuXP_Zb7Z38CYghjQpj=c?JPGUW{sP?sB0YVql9po2GP=(
z7O&!^*4n6ZDK_-%__ZiFkJfn27_c9K#*IsOY8^^mjJUgopp-F6<dM0<dga(Pxs8{#
zvCAgnGX%W|`RO7~tdKNsZ@J0?yjdC3SVrHicK(Zc1ek83o~7jA&Gj$X!KlD1867Ys
zUTghLf;NP{e|vdhXMo{#&}i?GJ8d}LN!9+#IWicXa&vJSUN-cs&^I7$WV(uQXoi$$
z3yLHfqJTy!tc<~Vl$2D8Y(Svh%)%FCcBEYG!;?N*t*-BXUcQm))b_X0QS5<mFid0{
zp;7W7A{4>C%SA;+;(3v?NXdtUt#00#Eei{sm!j{9Yd$Z1=+spaS7w(9qw7~YJlE}E
zVxk=oei8HipkiyQs~=?4?oDjuBKs&q?feg6)3R(&p%1kxZ|ZZogcGMuy~({V>dMS)
zGz@Ddc~#$E*m>Nzac|JRxQC2{hBiYb(P_%|Q%R34ZBY3@P@Pj$#wPmNr$m?p2?-^=
zsF<DQ*}G>S|JSjRPNhCET1928IHS1j8>l%5;JTnut+9B8*}w^kxJ8?Y*hd9B&1u*c
z6y%glx};95uC0tfx8g|CXUtM^vNA33sNrY^47?)*-@;5&1Ll_12SbAKO3rCKEKG^x
z4VBbeRefb;_06+$qobq$@ep>Mx^_)v9L!~^;NH#Q2(5Kan;EwzxVY(v%FeGB5P{s3
z?OYa?eviqD0u%7KhXn9!5~Zqaqg}fq2!*bY32A!J2Y7E|Qa5(cct@tGHulaud#d_;
z-n{Md+iABvFQc@j-?-LyZs!(1+NPC4N&X+>f)zc@>Qpi*1*GI7fc{kR=x)nQY7ImN
z(?a&mqSPQ}CV=X^K?OENC5>b{2zrx&?H--sI|CnP0h2P&$WZ#(sWcfKX|q#Tfr6B3
z#y;b1D=z*PuvQ_J>GFYDU*CroE|i-uOoK9}VKoo#+*oQnj=K8g(z9pJmLRR@xvRe6
z$g0(F1;S=9UQL4r{-+G#+PO}Uxy9IdkE>X&5<~{4DQLZlu-PKKLO~=~$0@AyLPyUx
z#Uc(3Ly}2D8-7t3%S130oL%Pc9`JKX3%C_xA#BMRFSC^0{v<r>f(iysv$(*Sa(b22
zVlTV4jqdl8dbV)?AS&Ld+Xa02vfbvnVxr<|EiFHcL|2e;A|;QjR3C}%uKrav8zG~D
z?^y7X)Po`eryXgUB~+z(bmXO`MUW$kay6q0|1|T?d{%60igy<<u>2V0Z<wjEk_0Oz
zW5*Z<{+Zj<?#ZdYm>y2Z;<_Xy$mGn}MQMmJtb*or7~SHlg!-nXvIj3=(>ZdbNJy{p
zRdxDoJ0qn&huZJ%FmDljnceR1`8`TrX#j@ekKccP@Hxo|<(o+Kh)ko)DGwJ!8*z?b
zlJ1s%&_za!_1@m^6Jb%Xu9Zu84DzE97?5Ea*r7wnfGOy#+X9Z0TCm`ht%lnCC}fMP
z0Fx^(mkLG#;piFR+22L~E><@>P$lh_#}BB|{*3cdwiqrizhJQYjkKR{t1_N15+}e;
zabNmn3_#eH7lXmof{aR&9zEK|JZQUqr-=VoVWH%=K{!`_zDM^FXab+h<kBn<2vXg^
z=RYs1g}<jdFU?Yaxm*S<psGnPXWU?y?@vY*d<Tux2}$BHa$m2#W-{B@{Nq<bBs-LG
zce_RP5>brqLn)&o0&x9%TTco(1@m{e=OXf=MfCJGYNw^hv!{Kq+K<HXC}@DM%vok{
zy-{MaeAT;+?x_#5iHNo|ev*zmAup^Y%0&X-D*DN^FDIL(+nkC)YR(4c8TU@Eu9V8w
zMBIC45ob2UW>BIcOFwG!rcI+kXYj0586uC~dZ9o&RF}4{G*0(DgqL_jGislp3?g6R
zF)txaG<L*@g=n9?=7gsbY8_Fv?E5Q?ag{cjhMD4*#sRQ!Fez_$w>=eY0_8>);749y
zbHZV^`7n%kQ)(;LVCzJY6jTEhE%Shhe;voE7$5X4nEty39;Yu{xP?;J?D!z8=bZWC
zUbq(Q6Yuj|yu{|>Js90<CVGyGpYZ)SyGMm?nZWy>z4*Y(@00%Z@+3pOfmH>}U||uA
zUUJo(X*XU`_U`UdGhgG3^aCU_H@L17;r?Hj@d=WYRyO}e5;pt;VEBIr;^xuB)BO`W
zU*tEywKLF&7<8M#AOkMpTLw2B0~2MTwMpgLWX<~Zdy-`bQfO~}d2G&zM8zLZM)?+r
zw7|{>N4vUPhlsy_O1)z07$i?$(t4b4HT_4ZZ{t$o_7^NTKnQyJ{CR)EnprQG@u9_^
zmj@L0g&&@7Q*aa%cZ`PD%$>oML0h-%lKA&!gQVA*4eK|4%@7T*sphgDLE^<9*WE*N
z&0DjQfAk$EPrk9Nla<-8f04B9T);n3=^1t7%@_@0mDhxOBGE7V5U~ndRFV@HClEd6
zeW#l35!4{a8ON$=>kMEcjm&2LdL7sJ_1&^Z=5ny>Ad3VrKO>7nyc>3ysevWwyGV`|
zB4d0TK93^gwmDWUu>9Q+DF(bT;1%Kvu~&Ja=G^HgA8!sUWuET1z(x%~WkDqXJfz~%
zE>gu=K-GZ#8rtActPdZ&*$mT<Kh>e_rRaK%;igSjsc?_bfPg`++uiC&g91g}gX?cH
zZvEx5Z&6&eg^LzxdJun^4H}q!f&UmsZbY86*W!;HV)>RDusGkmYls@c<yr;?7t+J`
zyeh1!pYo}-&Y-7^v#g?6UWDMrTM@D|$?A-!rx`GHU0id$niE2YV%30DWODnFtB5~a
zJ4rcsOQNzSYQPVYRdsy8{%y;arceNxNgYsSH~E?<`lXzoa^rbfR~g0l2KACC20s_S
zu^b&3+i_Z##0&NhmkjK72;GMfaqaGBmK7{MN8J)Cf+YNLGd7-lZoEn-c|Oh1p=aJi
z!bwOzcA!(ojw`ory-7h}*5DO>-g$a7%8`3|jIE#KncO$GU#%~naPi{bN7TnuzG#eX
zYS2rS5f2<NoMWMXfBUwH##}_@;3*=rM4*Qeo9B2j20@x-WrKR?6;{@k6k1I6PHD(X
zF0&83ef#C87jI6}X&nIZ=d5E{Ve{K-(a|3~)a)2xe}<m55iTYD1`IH3AC1Du+0H6~
z&Lyi147NwO=2(P#x3VNT81ceIN6jYmlbQR`qe^YEX-|)HJlut6@|)2*d!(OfQ+NLz
z6R&zLSh(<f^70#@&s)k|QuaN2_EgQZPxbWh2#cF_yCkQNPMqKAu`YXcsurF(Lt6d5
z%-$E)NosBRP8MYQ)OC_|y=53MF}X{kM|pE?Va&^yTFH&ywE|{!$k;wrGq$<Vx7o?A
zTbH*gk9N;q+q=);!NExGi3%15bIJ1MvA_9n_e@L04yUX&Uv<1)X{FLarZk>y&b(Z*
zVV!OtDg;fPk^*kYSoyqoTBvq-V|+=u3-7%mwy4Wng&ELFe#a7S<LK8Rv_+b?BcTyZ
zkdyLu<=I&tIf}e;yuPDy&^dVy$a!TutYW^p73?K^NFcLbK*sl49M5_5=bz^i?70#7
z*Y!Ncx~S+I8Cg8>f5bIN`7Mt9@JL^A_}H;C>t)Doq-<9;d~NV=$58SKDq+8+{@C!Z
zldrQZXq3)!1zMp#R~6D!!2PJKvsW1T`BPxL9Mrr(PB~1x<><vtTQQj|88c7^U2ddC
z>csf=;>7T<88V$R7BDxm7)Izz(Cb^wfLToE`>S@P=x`fSFYP*~^w-Zyloqy)UbdJ5
zMHS8NrXL@=>BsC2Lm~T;hz=`q6Cyzc1VAunnt=BCGpF~w_{={U)j2Qx!;T}Q>u2Xo
zYp2MloIxr(7EprKUAy*tmqaZ>dI1+Qn%q?HG7l!zkIR4W-rMie`Y;j`j=4(5x5E+=
z+E)ou5Ujy8zE^l4+0R7)7o~+y8U5U2a%x5teeN`zIoC=urvjDnz$PF$Q)U!<F-o1c
z=v_1tUuQ5GXG5-c8blH4;B9=M)Zl;;`zY!aTu4$tcmbGITe^_Ng9@ovMwBYhXlE!p
z(|)_e7)Lgz(UDMrO!f_hmU~QC>dL_FRo9oUbRRssb(WGa;KOs<Q5|${cp!2PDqW>i
zCp_dCc*0kBW)No=i}&QC)nmZIVp||Xn)a7^AlQ7Bk-mdwIcPBe`ZaCN%D{#1x8I)=
zJDyjS1qi-L@&GI}xy$^6>$g@GqzFVr<iqfQb)l7<%(CciyNVB5vLL?mnhWb?unFgl
zoT_wSC#;6&K(GkI;RHS#+8m|{+t4NPggloG6Y$g`JdpAP=M<|qcEW^IM6_*D4impC
z0uv&tRNhMx2dCDx1luMCdlVfUVDXais6YcI&el$-z+)2^At111*|L^$J0t-~I7Z6P
zq;0gT8>Y9^(xu)T-f?pgv#)kU7zjup#bgznOJ4GWMY#MJTmbX6Nty^QJTuF5u`=w*
z>NRV&K7VB<a}fUlVbgfb%$e)hzc)hikDPes2?fGpKXu+(#6rr<sv?iwM%DRgO3E2|
z9~B#Q6qqimf??9rI2R!17lW3?!g{tFAMoJ88oYP3ZirD3MBFo{PoKMDzmCw1vGQTt
z-)j>C>#%4XxHEL%;u|<cK0ZO{-`nhA$r#Ii8oX+A3y2_!QVjPP4fHMm3j$5)2h!L?
ztwSXDMTpO(DyJJ<VZ7)B{d<dPi}hH&ws1V(qMDQj-fc%N7{DlpRu+>uZ6jJ6#qUFV
z{0Hst#c)sF0s?aq<P#M!S=V_%<N%kDiKZdvszir_zy~OonMelq>NU(e<vJ;ES$51i
z(7IQE<Eda^cEzt38aq{;As)hKlUan3RW|-~Z7%5bHy0p0E){We0ZUF+AcWm|m&1Z;
zQ7}!a#S{%hWKBDJio)$_<{4KBE&KtWMBPuBjzc%%iOBaDn1>V{kAS_4I(QI2Bz6op
z;xTqn&;oUJrkSiIvJ~mPC}js3M(Gk-X)RxVT*VMc%vIXj9)Q+<0(racZSk6UQ-Q4M
zR%}n?r$5FClp70mUV0mfcfVgm8Hf`oss;z5IRzS_@3)t{x^hU+a6|yeDFBx_!^s(;
zL5b~{7{4bXB4Heax%~b2-zDKxH`bhN+$pB+wa${N(Ssqd;=UO{P5=I5L?BSGaO^GN
zuc@YRJG$DgGG*zi7+@zj$zYdSD_5SZjTL7_6eGG$```zYrUdvSo#<Y`)HVmoYmxC4
zaU#0vOUPi~plN;@WQx9BJ?NA(oa*k8<e&ESRUp12AB8Z*&OW95j*U&3_}To`WjC$}
zfSjNAr=9rF0-e9;fl{c6R@w=wBL1RXhah^s|BS9Uky$o%fA?`6J9j>nWf5*0H%jN{
zBpu_nb@OzNwJNR9Ng1p2a~%$q1pF=J0D`~+GZZ=v=-c-iQ{c@Qy~vYlo0wc*Th3M3
z3IHHX3D-~%nnt?Jy&q8bqS!Wou<K>Yx2OQKWsC1_%9*e>%A#sZPU>nMr~aECMW%eT
z65zSV=N%E_cXhkBqfEbQr?)(37n`@cQH8$l*5BO}T7|UhSy8uqRX2yOi7#Oo2Ln=+
z0MipOPTqeW4(n)TWgT1?(7)E)xe4vuEr}snrMH}9;JnU88&Rtj#SxLY%P7tfSSyyR
z&!9m;JYA@R{R|)K5T0bC$_y%YiE)g(z2nl-c+L26U}X_Gr$}U+nxq_%$q5nXi|RlI
zbz-UwbjGhP^?b>8m%I<XK~*F*)@hHhBLW*3+pIOqD%Vw@2LPsH026}r!@j>a+uEhI
zQEQ`d=|OFy#sAWR483jQllydaU5GsTD6sNt<oLF8AI#)g$ReXN>fjm6!r5)#eef-i
zv@x@0MQ1b$gJS*ZfcR7(KY(DG+WC*OnP@#4+Xxdr6!OJz?zT8!xE8D^`TYeW=i1HU
z_z55`=ch>R`od?RUBo*avY#Ktj?rjp3Mfq7eX#jKNzUf*OuJMay=^EjXTSySr(kr8
zsNtz%7yLBU8Y3!zdXV`=X$ef^#ID_~B(5(=?f1zfl+Ijc*S4I?`Gdf^h;7p)q``Tz
z%za4$IM=!Z!d`-1XEA29&bPT%j^|hb{iHzZP$k{rA}9iw7p(;CyUJcg;SBBq?<@`*
zRvjMssog({1PH?42=8bSzK$<?UbH<Y8orI$v(x2kH{Uzo7fY#EOx~xrp>4>S`q_tj
zMpvjByXP_zL|EA2QCY!gBX}h-2>A|@u$S0h@ls4V2|NBL;%j(-mz2Kz1A`-%K#rlm
zy5NpbVy2L_QhatIZBaMdXqaz#G<P#^QK~E6&9i6AzE(=H)18OqR{30)so;w1Lc6(O
zU!E=82Vo!CNdVruBqJ=a_99VV5lGG(d^)G7Uh|^dG^%F8s|tIM`O7aTSmJ>az-j}7
zTfl^YetvTZBT0(%@P0%<I9?@yVz^to8^h|o&vf1_&e>tt^rwv%@%81ENqxGKDF`Yb
zc%r~=Md24cxl^la&u2qFpc-@?Z4Ge(fK_hUdE+Arlyr0|e_WK4Ehl}X%DWmp6`xCY
z)Ou9(;%Zoo$QAz!vnX2<f5gPiHGQX-etapUP|vuAg`y4$@=e5q*JjgG0amc>8Xid3
z2GAI^8gRm_!tF)CjHi?X&6t1#hLVQdL%w~K4*A-%XG;a!ljaI+sSd&5`}4uj@j=mi
zE?>D4^E0DO!c(`&t*{OY;?44%R}9N_s|Hm*GqyemhYT-P162Yo4yVqVH8;WUG|t$G
zI#`%v?T$gJj%&D#+Sr#M<yC9eoI|-<Fns7DFzg{)4%Chc_DcK=WyHP%29$=TuZ0Qp
zUzc`3O-_ea73ax|M3^Y<2bT9&Mkij72%Ak7g4v>clAilp_(-g&*L3U4XjTbcj^K7c
zK^9j7tS3dH2*HvqB^nWG@&cWoVoStZF^zkSpn%}SgJV%0y2FBPX)nRvz2q9%GFY#e
zGCA{_6Q3zek4ye&wUyNr0KFh|h{vOP?lCs5$GT9HW7vf&yqLiK1*cB0lnCOtL6nGw
zs0$k_o3MvzE21jJ{;k5n;numdB(DC#(Bj;>tJe4tuN0n98);fMQfbBveKd~+fQSPB
zSdFpc$FJr$=|dSpO_j(3Lm_S~A5cJ#`+m9)80UTf_Q-ID5$Y}0RVIP6Styn5g*vk-
zPK=S&7g$8Z1fHp?vLYfZ0gS}iVYE9`uBn;Ga_}rgXaKy%3vgWQc;wKy>G(z#=7z9`
zJSTjE5M`yG{ltAte`+^wxI_PhWGSQ4wBVFoZU4uobabFU5mD9=0^R42$!}Dg_lWq1
z%ufIscIbFIl&P4Nj)u!4XcZt1W$!~5Y3CnS0=ngR5^7_O`G4r#{6C;&rG8Yylm9xL
z?Plc8cF7VOdHp&-v91URNy}1_z7l-l>VaoYq8guKvtb#Y)Sp@1q692DRUk*ERGQ$#
zKnJhm#AvpJ-aL*Xm1f9;8UJ}QJ373&)kExMM~67DO;Pd(U_QBChP2?1!-oYqsxLr=
zxPk)Y)$<JvK7QBqvlWL#@HF5h+AKgf=`gh}!XeRMh7$=Oe4|EsL@%?jyY9r(9UVG4
z5O$8M-qWj<LUf3K+h`s3Oyn7uER;K!i$e^s0%P`M4YMCad_^3w8tW{|ociH?NxunI
z4Ft?KZ{9uvO``YF{*$e0sXBy_GLeV~PA)>f#~dG}_WZxHH}m4Xy}bkZ9BX-L4{F0H
zCImC-Tm_Nf8#jECHtN3fTAkf#Ou8_%jK~?!!6Ya?*D8b`e^ZQ9?C%tff|0Cj`}y+l
zW9#$hop<$g)f<)@wj*v-&x(ij7`1heOhU&HjuN-6g4r!>FdtIGg9Meb71L2fNe~lG
zW2Jvu@eM9bq;2tKL8-*bt-|pU@K5l+dr_586c$!66i$DRz#s1J)lXOJia0<wc%4|(
zjt)WVqIy+aUOQBHjUE+ZT!x;$zSplv9Ak>@=rdqIU~_eIlP`Yp^Qpoh%SGqUox90I
z%&kProItF%NCBucnI4SBCBj=sCzCSsc@aG3?teu#%an6x8&mBKj{d%cA<cigIqMcT
z2QDtwl~kGs2(l>!Nx?{NL+wO%mcwn1$C>lzxBa{Uuje&;;7=$jU?S7>V4wF*2ZspX
zup?^;k(6e<SysjwSLu{atwPG%&d6<q!Se%+7&Dqc3gp`2^+hl)YW7@?x-St5*RC)(
zEl4<cI<1xjG+dcIb7p3jU0hbgv4wlUBahuWG1BC}c5C0z+@M|rsZ@8jQmv;$>#evy
zHuJ>&VV}7_<e0{`7Fp2iL}JcfkVdA%g?Sc<J6%CDL>lp6OQV$H{->IA#nb0Up5ah<
zx~4~ljccXg)wi`#PRC6xzK+reNSIE+!el@+JAsf=#ik$d;EMgPHOoe03<JTLgn7ZD
zh`f39=o9ibs@Zg63Ma++s9YNr^S_vjRq^ZKhAb6{KDx7~*<yrrOsi6Gku4XbpkVCZ
zu6iA5q829hjF(pcD%1o@UiH*q7YO<l?90{K+Ly`f(sx;hmk;~D+LXz;hLN`ZqU``O
z!Ncg=`p{;D*%#ywRo@c7p@TMq#gN^G(cX)jD#k}~*f?Ln2DmXyF)nA5*0UtD|K5{H
zxj`GkGgB$p33f2V-_<w_j+DGk+7m2GDx;R{&{yVV36w9TDSR4)HQ8BiVf3Mlx8a--
zq$Cv!C>zs&wH*WZ)kwLw@cx8c!UDDymc4ECR$VCvr%_t1EMbSVm~{d*0yj`nZbeH|
z2I@%KoHtd=`SWK!MfDdKEW8s}QID&P6(g-!y>jKH@85Tb$`&_1&@Maj`osUW=Qh90
z0-L}91`(#^?}(s(ZT-(vI1+t)tuWWPi!f?nX%TiPVLdw%MemRvS15JjS9_3?AS#4h
zPd#V>v`Vcr+bYtI|Mjq~j;}{#R1@^=q(Jv!BZOxP4i(&DCjn0zXS_Uj{5wqwdiM7B
z!5M^sCCnt%KG$vwN1q6+@u$A*)BpMy%7<L59g#~C7n&hU3WpLeaWq3srjsv_!HcdN
zd^rBb+b(q{0nG?``6A%*w}1(J9GxmPu(~uXzd+2O9P(kC_9;ARv1A6}y|CLvisAO=
z&s&A0+|DrmIA@9t(3ubp)sJ<IAYdZ8AT_zgh>8;0RHVE*pH}f%-dHd#?v*&`;0_|8
z;*_zAB4(q~`6E59DeTE1()jq9|5Il;-^$p+!eUPBx$z1L>0IA1CaG@TvPCaSfjh(%
zPsIX4CUJOW4h7GYN~XwGRSGT7U<C|qQk1<E*Y6{xlh<h<&FSC_-2E(9gZC+L`-;Vx
z(7)my4Ffo(v5YMNbe%YP^6Q4zw*?$KbPK!JCGkP7<N9{w*F>>r<!P?!@TGO}UXf#h
z<=I5ajbjX<@{P%0S;Eh!1JheE?&di&2xQ0+zHc67q@%QZ)?m$xqao#dwwmrZ<J0FT
z#w-L19s>Peyn_}9n5f-gY2Sxbrp*}GEmVu#wX1AbtFIqaW+OTGlnNrTS&HRq?hgkl
zkbt*Tkibq2depan|Kd=~R^E}7iz?2ZKE1}it+*rk>aBd5OS^_=nm_RyLi+|0>e(;V
zShAFrihNqtt{MTRUDhnJ>FyASPN;cKtf8^vH#dxlUYRO!7>S4(6AT%SlD;tR!Oj^z
z)X{x#uA7$UXpP|I1kS<2r*gJ`T~gu5IS|Wix;YGTc2RG8A!Io!-zncY9VVv|KUQR#
zz2Gx2DBcg;tn#5LHt7{|vR61+aj_j(UXx2`bIShsKq}*<Y<5<FxB>|i{Am@P`Sh5G
z{7Q_P6RjCBS&ktfH(F#!Bh>13)GANjs$G@DhQME~nU!n$2a~$PT+{#Xb)i%JcIWNw
zy;4l3v|(Ux@f$Udjt(C?L$TDqv%!s&MfQZ+Uss>{?m(M*4eYHnK?p<QTew_|I))Ik
z5#%gAa!)0Z4a=!Ur<L{sJm1z(^G(FGBE5tH?{&!<;)CD@F*~cpRj4{eQg~>n)h2g#
zSL_hdZc>Mk@|r!}vQb16M4Ty$j;gDxmk16dk<=H@LN(v;g2{pgD40YLWko88Ng=!=
z%@=qnF*vGj-k;^<buT6#RdG1KUvERVmSD}MV(MnViQNF5KR_9asbLDtPXu{D3#q&3
zhD6F7yo8>i00c$Dg22H{sUFV&$^Zu|4YRM`*ex3$U(B<-Me*U12o3>-K~)Pp)0)zt
zPKr)zu=bOQA72Wm>NWOXQ9P#v6n)~16_)rVB~;x77{h}T7%G*&c>MS!UO+H5L=BId
zxAl6}_x1(3VfI;12Sz_jNFMkrryfYK)Net2>CC1s4xyo;Rq*Pv$ERUj-^RJPG}09?
zEwP`=caufCNQ=hc&KdoXig9^nE?_0aF4zP!cEn_T+UhkCS|jm>XQUj0K9p-vz7UDG
z63w*e_>)vGq|033ZGv(I=wZsxGLx4&#_V$oB?9@{5^O#U6*ix=SnP3n##eXO>SNQg
zwW#e}eL7}n!1X0H)I)?n!5Or_n#~m5)&F+XJ~Ql|evreT2|!68MXWr0mFEo+*2~1^
za>&exwLHcq8ZtElu!1~RRBiqdlU>>GO$UzIKj5T)#sA0E+GL{uuujaGOX((PCJgX7
z4c()#vM~5?_{^t#SW}LPz=iNATbh=>qR{Zs>L#1HzkS#5>ThEH`VVJe{s$x=(QQnQ
zkhLRuLS0t5^9>?RGeKsF*%RJPPW2YPV@ge62oFk@l0_NROU34T-OsW>K370i^Oo{w
zS1!+Z)6ndH#_;|Iy~CaV54OGp9_zN-x+GKxO)``@AxUNxMMO%YnS`XwWJ-}#W@TtJ
zRFnoGWXf&ISf)HO&vRuaW8_<xp7)&decw61U(frV_bJ`?|9@S3uf5jVds7MV$n!kQ
zyb-g<Z@TEk!xx_W)vl_!n1t?WtKDC#rm5+2F|;D`_zYidSkZ`;W2lClK=bPx>t4Tp
z$0Ib-eN*I1;=99V+d8&Keq8ljLi9{>>x^UY0N0FG+bjRKzYmPgEap~k*eH=mZDeVx
zrxuEV*bhd})fe|->^g>!Fk<}UJ-#&<Un;-3aCB3GRyn}N=ZXq5kV_xZ(mvn(&}0=C
z7k3A+*5SuL{P1jHu#&V>_s6U(W=_sZAS-O$?7^{iXj{M(@DsYaYS?8d5atKK+3}5Y
zA2SnGQp%(j_t;u5Z92-AO2gav!jDeSc{O;*sCNR_2`&B7TvEnTUNO5cq@%5%hgx4-
zC$azVIu%<(#?&HPT)KQYCMjvXT4-;CowoKC$R4wEb3w|pyf;HPTp&Mr%dPzQUi<a8
z^5r+rVVuDiXlxkG+@q>6&x|>y*fN`)Ee~G{dh2sLalx>{@>lR>4%*eLF-!Xk6ki|w
z`-Fr9thNZ0_tK87=^P!^baj=%AUiH`ah3$Fqo+@Y?R%gvvKezxFtH4&l+)njl{Hx1
z`}FBkrKm)q`~Kit5c8F<T#?Ah$-!{;@*Zm}@dmCL04S7^m1XAOxK=yCel;f=Xz0(#
z=pV-I+qYwnRG-RG>}ti<_(RUL_&?3E`G!NCXPw)HsEwANSyom&lPhOF>ck54y6+Se
zBy-uF-2)ZrDAPqJsYXGHm*QIWnibX7hAqt}EqyW;Yn=L+@oC$tm-tp*US1sguZE|m
zOUbU3e5~<(>MH)u&t+w-d-d1cj}%--UJ4hr5&vEHXChuM`+5^X`Nr`C*;1P-kVJhX
znK1dyAkQO;44fSvt^|!bI<KdlWW@Q9D(UQs;U7N^VqOoP7G{Nm8Z(1a5nDz;cnS<i
zU}WC>{8f<Q=*~c!urO8}`vl}1IB-BZp|+)_<{&0cLErl|GqZt(#UwTQ{n_v985zyt
zLcw@O@9AmhrhUdvv;zYJj(uM?VQUQjAftQGv<ePUl|Bv;c|JnYE>UGvN$PC8sdgBV
zw_Y@Z{hzjdDq<(jRcymdwQJX}dwK0vcX!V=Ox*jx-pL8r`f$74#r1mn`np*T{9AS#
zDTnf5fPTk(eTFMWk`6mpwiGlq=_W4K*VReJ@XTO4M<F~w$M7mK*BS2{j}(;PJN=Ac
zc!bP~5}kIty1SQ;n7r)z@nh$XFE~X05rRLi$nD#||B|h(wn6CY8jG>%X&M><HLWd=
zLhW2!RA1HB)YnTjKbPUq*4EZ_#SDTwhCBRYvwY{ymMFD1*VmVzzl<>-dFA}+8}Ht}
z?FKLW)X&$q6LMb8pr!899a2&cLVg7|3SV43HZc*pIv2*lh(E`$5f^2|%goG2xY;d{
zbRRP_u|G=_)0p8va+>|*c<$VGaCFOu7yVgfd!ml3Vq`Np<gljw$&s`j9U9^gFUWHp
z=V{5XeU+K%+NiUQnfV~bS|Szo<Z?&&`1pui6Sg0^<8}LHZtj7@{RzPw*(E~J(a}dj
zUzAZQqbKXfZ-zQ*4^kLtHqXk*%d0N1=_NQfe)`C3XT`kJOVO@+0i(LfX_!zhYH5ie
zvn66V?q7EsN-hQIYz74dDP~nb@gWx%X}|~*d|(V^x`D}^HH@dR?uq*59&e}QCS$SI
zGj{ZA*HS%;iz_*A&kz?I>y7i*xG(P18#+eD^p>;)stcHEj|_$xq%Brw;_5D|9X~E0
zEBhFq8!=h~I~HNUNiptdX*r6;k@$_7+1X)_9+hF#Hip?xA1mhz<kL%d0G%cD{{3QP
zp(YG&f?0@~Mq>-kEo4O&#6NtK$7reKc&HM)^^~?N&OdSIUV{zW3_$|dMRoHHF!+_o
z5Np=1H9?V#X{|vmayi~hqNW#iG+JN~8mE!=RS~h-(WGxP`B&GU>9)kY_H%DWG5zmU
zujTe@)fo*lo*zo?+b=eJ5I7rK)1<}4%7AA(2X$B^n;uJ&?_v?>qqj<y>rouf!mIQJ
z>=WkcQz-XGPz6xa(4@c&N^}n&jTmYwFGa3Qc8;hFJwshB=7!Dn6GHX7z63|#Ls#D<
zma=~8--Bs(j2tq?Gpv9`r6j<>s;!Bk#qdX9eL~v@zA}7ma)H{~z~zOL4eyB3eY@Hb
z$*>&hDezcNUtcjCuWHART|n<VqkWF3etIB?e)~TQZhxjT@{++7$)-bvvO@)k73rT~
z4b;Rc15r`Yi2-9VBO@c&sua3l0wwF<u>I(MeZ7{k;zpQoXXa4h>0+=nxJFl3*T%)P
zrhsvd-#Oi&7*cQfgg)|>j2pdGW4AS!AI-3;7lape4*u};%rUZkc>7CD^Coed)$IjZ
z6R2BG1}byeaRAJ(#sbw8_lD9^A52m|A0F#K4?SxEu(}^CKL6t4qC@5TCr{Y+?c3Lw
z<+uZtZ7gRT#+K1h7qSLX>VyP4)G1p<zUzr>J}kF&1{+f8zeVoaz55jg`{Itl#YG+*
zA75kSLiWdxp&=m$@kOO=TJIINW3yA%hlB3!@}Nje@oRQ|KE<x<5XiJg0QPOg?Pw_i
zk4J-DH})WIsz7y{)#Rn8KY)4_0M+~+oDeXV@ITc3ErinVoniI^V3YUQ3P3EKqJ|%-
z>1EzylF}a^^fU&-V=<Y(YCRJxt3^WtKR!N{SH4s%&sa33!|O!F7u>GIQN4b8M|80X
zcE(>i67?Pe#BXXLra*24m!+w5OZ-VRx|X9tm+I!QWZVQN94Yf_Q<EPs0&XCTCli=J
z6msvL*mJfU<i{KT+M47R?^NEEaQP^s2WM>ZIsrjJUjU<s(hqtGhJz-0obqJ<Gbm&n
zq_Sge^Yim$q9U0{nAEUg%N7!$P?$?a9YE#;ZG|TTll?OUITRqL3=MaAIzM^xWELla
zoQm<uNmEGQpnb@8oNWJ`#KgpFJ?tI#RyTtO!01yiujW`l`mIkr|Hy(`#g^VDz{Tyq
zvfl04F6(d$xY7Y@yNsVH7kB)bmIgV46}{b<8D$K~z}fv6FQ$8jDNJIpX2_z!udA{x
zV+a1@1=zni;Bd&4Pnuo=<yl=vM^a`x^O$G0T6m+Gt(`7S*yk5z0`9;fX?oTVX-rKy
z)AV+E{;4!iH5~B@<qJ}u3pjZxMJn@i(GiUHZ_*dwz4_c%z;e1U`dEGL!o<;NkNHb;
z**&A9)u4ZRsk&l=V*Tr+rKMBhg(tfx*@BQ<4Gaza?%v(O&wm?zA0?s<Ih@L2k~L<>
z{DSoGUoJ?$J#i>3ZTv|F-gYn+1Z=}DA4mvhlaVloW2lEFcG@;k+$IgrS@KBe?qv(Y
z5~@J-0fKn4_IMB#8R;7kz>D~aIn_zKGLXVlFcc066}vs_`(;<IT#3+-482A`Guq1K
zViKPk7QTl?M3kfa!5R;xk|C6@T8Rc~1_rGSavu|Nn85q{tl<M8i0-%z?CkOa>?wW%
zmNB27SfmI%dFN%35?9oERzjCO77ZR_WEs7^wpdpPpI-;so|Rg&ar(y|F}ra36&RC^
z7(JL`6M9tg*Yvc*+{6V0MdZQ@*ctJko2q1YrTwzu72C^~KVG`thLcJ~j)+f@!~;Nn
z$Sy4G>;W|vFW$cOhgX&1Qk+y{(_nWuSuKycrXotdA?|c35*F?o^hb=Keg5LbE%XO-
z9~(|I$|34q(?=)af`S6R9XrCp!<9j(8ydY*e`$i=5xDwlYHJH&?bpW3<;~7S!+SY|
z0dOpsWgdpv)e$3`*D)~k`6w)KLs8v<Zh$z3CUlWtjcQk>b(sLUss*mbXQ<n;H<Z4G
zW-nZoZc~|W-~NmfF2Vdv01057O@>V8o{7AxT0b~bVk|a>n#VIdmTf;~m{tQ!kj<-z
z*VsPr^XE^xwQIZK6(Q@EnZxLi5tst$dL(V9CML+@TnV-}Nl6#Kgzqe@+PQ1jYU~nd
zdY&BQgo&Z3a1qG9HZ-VW5hz|h=Cgk`vDMVP+p2+W!9nn}K)vw`4E!+XHa0cIFE1Z~
zyH4O>Ku-PD^c7eGggqXQp-$}uKY-s#axxcJzG#vo7!)#af8VdKN{A`gaJ^gll%=(G
zALb35hlL9hwZpJrtM<-y;6RdIdr1HSbB?{OpA+9>=?%^@86T{IUCHQ3-Q*}kz7^e2
z{_Ajy_r;yVwx)b9)&vOF7%L2&jX4hoJ>m_Q8Vg&{ozd%~?hD}Bw$@m-3J16UgomL}
zY5jhQqQ}?(se0zjMyP#U9kV6ULszfs`?4^^D0{PHxcRd%d};5GyT!tg&br%e5E-%>
zm#PXTFE_Lp_;3vV_|c``F~1p=5Gv5$(N9c|;rC`tx(xmJ@jNExCT<qQ!po^PJs`Ht
zlBM_UTd=fbmzI!#H<y4DjFy$E&cfVWvt*YSxwQ3^oYd6Q&TV?^l@P(3MRt<n3)W{W
z_88e!F8WAA0FDW=l)3L}o)$K*pxWV)muL6%Tw-?H<>D<^!=)~<{MlFGpXwEchy7@1
z#ayTKxR@5U!_@IACFNAKE1>Ko8vO1$oQbW52JZ&xuL-8WhNN?)Cdps_`R9<Or7+l_
z7jbd^$EygAc)zZP-<I^+A#m&kD574yI@zX84hIFNVF78r*i0PjniQ-kt*IB+lYO!P
zOWX41>ylJ<&L{f>1gwL@3_tFly1K(C2grfa)-GQcfxq!0-tAC(4*TxiyJ2HN80LZ0
z7bnQJabpo)4*8{w9fn{|@E(B<xt=<XcOMotqLg6($|o#uCVxU!)~$%W{IUcaxFr(E
zVMy=o??0udC$z2!%f)bbnORs4Vf`pRg_MiNr8GUH7wlB(Lw7-MZ|@01Lk4<@D}Yd2
zu(3kmz=3GURelcc<eM`Et}rbr5Vqs^RA8wv)YW+%?U0`<D-VbAVcj|cY9FW}4o3!I
z6;u8XuU`2eDcix@ge;1gw`fejSMm)GHU~Axc>U=uZ*3XZ``@Wp7>zDubA|ux<kTdw
zmk4)KcVUp2087<1G_<dG2AokhItNz4#4hDRKQ%o)fwtH$OBp&M6Ni3vPEZqLJ+Oug
z=WiXyvGzd35)iEg@cGDKa98EP#Pqblo;@MJNw8|Wl`o+9!LCgad3VfVw~F4fd2<QC
z?-30Ris`F+yJ(Ykp-gJRmxpq?wci2Fa%emEuBfO;+7z+-Ty6!b&JqnJsjkjVj7H1<
zU}7}l?@!#&R1ojx#y4}ajd%b4{WzE!0<{g!FLS~iS1bJ(8A&gn)ElfQP>k`6H?xn4
zsr`XXjl5DvNR3H~feEP-wzjq&a6WEeW0T{lRtU97(`zcO>F5}}r>y&`7R!g=<crD9
zW(OPKe#q8#H#EpJ{V*1NxMRVhs(xshcO@lU^tYK`zXE^kDxX$FJwG(0VQZVzU1W_C
z4yR7+eU7$Y7+;RjU8M3~nY62@@)C9p^4^v#Wfcy5NRi~?x&h>at>~!dkWpChWH2e^
z%UzFOyb+8QK!T`h>FDT;(W_G!9HAlbXU21B3tSy-qX~Iod{`S6?*2u0=zi*_{LBlx
zp4uh3MTCWI085G`ra_=QyHPOP+K%2^z0}@rRP|J$8Z3WPw##n(ZJ8fGb`KBl*Z6kk
z%^6oWw>{WsihP45gdq&n-lo>p)-REvV4ARl+&};O_a0XI6gxNJcW&7=7L*HB{oUQ4
zFprrmd@vUK`t2Lvx<df3_+RdTNe28u#wjQ*EsZ;Z=%fZ?8KT6Y1g+Eh`oiZjUcNlM
zITf%3?6-7+<^EIPEIw1^v6R}JaBlxumTdXX#IW(SNxKx=WPS3Te`~ynt(IXmA<79_
z>-G1!P~$@AYipN+>t0FLJI<kh%+~GC(r?eDyWpYMvo@wnY+jv|>~A;Ld^f~-><@Z}
zIw9TQ&S0)L#Qz>-B=)cXw%9b5%Tf_;Ki|A+#ByRBB2WqlDeyh!cXT|3Gix|qK#teT
zZfq@W5{seYr-FE^Ew@f_Qh02g=zEF_-1jy&H-n{tU-}Vfi^hGW4Gr3&0Z8}1lANXq
z`hkxMFVL#L8N3bha9Cud2K4vHh$gvFXAVlbiHv<WHH?Pnm|&DY0F;0?fbbF8aZdl&
zuV3%ct%bYb(<Z;;*pni#NxBeG+SpRo&?YT^BD=Ahc4I<&TyQIB$1qhh8C{tx$VgK$
z^g`)J+lsN0<s?-3=z-FGZ|v1(8YxIeo3zNPMw?V@(gg_ceBr=VSJ$zMz13ov<L2&R
z<1VHr{QPp-6ZD(hg+1+Sef^zc6ReBDPDpTXN}+r{ITzKP58<c${kz>-S6z^GJ>342
zwIslM#vB(z%(R3a>{Jo5)A-aBc%cYXQ*)8g(c1t;j)JcPPqn0XF)1l&H(@snvO=-X
zqfGKXZ3hVrq%a56Cwn>rif`uSEirlL56ijc`C~FOUI^SA5ad^6I`sK3wAor(#!@Je
zaHJvD0S-xRX?Bw1Of|3M0n*9OR|HrRT6c^vUi;ia!Spt_w4g5j_~yj*419hNh5v?=
zvo<3_927>fw%j46^`Czx(wrg_&*h26kKmw~!m;Ii*FJ`egEDQ($4McpF2Oy9-%i0^
zK+HraJgK|pBd3ZZ|CvxekKZjF6PP>owY@!9{UuI2&{+{uE?H3u6qMqQ2rkS)QNOYh
zFEGWBN!@A8pHjCc(1mui5>#m%U>+PixTLm|9Nq<rCMG<vB7&jnc~E5pk_>gdEH-(7
zzLsh`0=fWx`4Ai@&|#=#UHUrH4r^+zM`i>}VfY;ZVhE248#_(Fu1a5LR11~ySdc;8
z*Mp3s6pST(fRs9Emp4J#nb{b>>3@l>^5)d-ot?&6`tLrQ%{PBMT7T#C4TJUScJ72y
z4SbQ$zF~vX_3QgX`MMAbK;g}e-}T%_b+Xx~W}Bd;rKKA>dEQ{t3^rwJ#gG5`#fPkh
z25mCnBU^d|TrHF9eHpcl;B3RpUanofQX9^8a%zA}L3YMC1fV0+Ne*zAB&TX<EHu+t
zo2Nd;<}BPOT)NaBH|v3?v5WlveF#iNzB$eT#h<{Qd@e8VL1ZWlR#i~}p>h~{;llOz
z#R+UFz+wSFiTLItJCin$Pq57H;j^$fYIBdtt=O_<i`!JANpfrb4eR}?u{DAYCvx>~
z<y$JKsCVw#wd(?YK+2KLn>H0=smNiB>Ns-i(w34>igTzLCNNaQ4!~gU4-+RR$lC2F
zgGB&hopW-@UlU)X7K1Mw#V35WbCX-V>4}^9@{h1wv#H5lI9VFIF)<<pedA8$)zP2L
zy0IY`90@Y&o{Bf|4TCk8)09<JQ&I9jGgWz4U5Lnz1#|=GAuc>H>fzxbaxNE(Y3Q?x
z?vhLjbZ`MuL!rrf@IYX|0DGFPHqJWH5eoKD=O(J%_-}s8Y_Fn)g&-Ex)6k^PT{Fz_
z|MSPgO>AarD$SBgrUs8?k#(n+OH;}PfCPHAP~^{R1M(eRuw+NIalW4Uq{AxM($dm|
z9XmL8N^FU6lbDFHy236$a+$?0d-?BEkK_X%-=vmuYPC^Fh8)eyH5BixC~-#cE-8)d
z+qQWe)qeN48UjuQT^3klfqL9j{BiG|mf~Uv>ICPba;|EStF`q3&Op!XHrOd8t!z4b
zdzG=b8Z+lIL|)d_)+%8qErO}i&n<<DO6f?e(0Hz@GBn-#OX2<{?~7WUc-B!&<38&F
z`qtIe<ts{e-?(uDEzc0j+cECRSdrcxaCCTEkMcnY-PG;8HD(DpIZ?#)bUgjlu?R9x
z`7<!9s*pyu*k=Da{T+>Xf^k^$%i8$XB3TFK))4ulE$jHD<C%`tqaMaT!}pDNr964(
zN3Q#ffaT21_;_Iwsep3Ob%2%K-+%nr%40V@Jxxji7`~g3j<Cc4n@8zs3dNEk6fgll
zA|wzFRYGE8j*+I4lJ`Omh<3;Weu&lYA?1Oeuyb}sMb?`3IUlPC!T;+PSzo#oavK3e
z11ppO;c*!5z%|x2Fd#joCl?irS{W88n^f)=l+QR+ufVS0)?)^ljj$+c`YgBj<tjO7
zn&R}~q{E0@NW{CKtzHdSdjMm=CRUwf4`g6uUl{AFA%L$W8_CfzmF6C}VKq0&zf7)A
z(GGy-q?PootO!c8UF$sscRV+}pDca?My)8csG)9Cf!u=3gBJ{v9(>*wo%hBCzAh%|
zF}=#V<k;ee36XrWW;s--Z+4~a`T0{_M~4+%8s_8O#jNx+SKZxxd~*=;4JfW5<e6YL
z07#3I@pWLk7#+JWr|E$J<dHwOii!$M0e0hk?#ZaJ=*3@0$V8+P<Z~pt!j_ibiA6?f
zdX-3>kTAG}g#3}gAj$L&N=iwcI&(%DB`#txmcO01*MDjXbXI}b5**BkZ$d+VZ3i|K
zk_!s*l9m3BzLWw9bAP2N7m*5m)eZaui5qk~Irs9jeay~AqP(wQOllSqGpz-fMk$kw
zkPkPC**Su0GlwCGG05e`o`_14%8(;a&Z#5v;sb!R!|t6bH1()hS!R)%b0kgKn@>$6
zLh4xb*V*5{Q!$v0tgjBCuN4h^nIKAJA`}c}uq8SaN}}XJFeLl1cT>i08h$bSe~Zp<
zDB-I0Id6v^m1t!ICd8`S(|US@^|i{Sxp?tnfAv#MoJyo}2VE-5_V%Wxou0w+^71?X
zf+Dc4jn`l+RUiepn%a33`ZF6nQQJWkN9|2t9@W~`h6ij6he*c8%exIflEM;c?(6$j
zULKyC2-X-1dFQ$`ipY~O2g5Cx_}Pqq4J5?g1}r=oQ?L>j1VnftMyjsO2V>coZq1G-
z2BoLAvGEAdhUS4eKF|5ur)bAV?dWV9*-J-Ok&p@=-@(GtGEAEVKtUa^6S)y(#A13n
z$8}CpYK_zt{Dx;P6=!H}34wf}lUCuLt)l7U<PR(`lb%7l{wKKQK+r&#0_)=a>eUWX
zO`^Fi<K2{VhG5egZVt+)f^|TRPz=NSI?rBh|0#IOd!`;Zsm%NyIBI5AR^QKNY&<$Z
z_K5~rWV0`9#+^8isAix<LGHlfOxcUIgHw2;L0}~j()^KJp!_KXt^>(J5I4q{;H9Wv
zyto65g^Q#rCR3jW8Dw_h0&18{=V55&tt)6oc3|8*_wL<yp<Z<M^myOBTP^myUe8Bn
z|9)_Um0)c9exE?{MgLRGix*~(gmq9VMr8o`+KmO@cN?^L-rIUs#+9>--Ih{-K~FY=
zt(cs0ACI6}mZy~9&#s;xC8(0fM|dq2=6kgh4Z0MTCKZ72=xMyY(BIRRI(b6Ib+|M~
z&UK(Nq7-vOL-^LXxVR9~8U8VB$N1+!VA}n!=n|NS=vcS|<`o6|ZI#%A@pYlm(ZL|#
zAWd<O-?9Avpp@Rp%YUEl$oHY47bz*B2rD=d=BM*(EhGr*wqZjdI7gcw5|_dCz)HX&
zcf}W>qj}%O`CmW4<6^hlwT47~wb0jzi2~Bn55a#mQZPt=GY+owF2nk^Hu$vN%gW2y
zoi(QcXM}|VkcL@USp}EhGB72}WO}lcEhj^GWHz4KDlks{=<$>Szk+-FProcF)}gMF
zp$d$_RUQ=)(svPk(H?LMU-0Fr&n;iJ06Bx(IgVHe<HHOkL38<O$+*V9eLFk4^vBli
zCShY-awj=q8d8-*2CCOOw79U42_+5X%eT6^m8(V$VZk*@PY_OJWo67caiLL6%a02n
zR(;!yqAtuS>Xo?Ec4O_F;JRVZAwBQXXRw${LRz}K<K(h*9|v%aCOi9~#yyaL0N^<4
z<MWY`byLlg%|5f>kK_Z~f+D}ymM(Tev>V;)Xz%n!k&MmQpvSXgi?2Wr&&yK){Y}=L
z-1zuWPS)H~&Huk3d&&{Z!>?-hz~TT4-T2JR_WFAXA3r`J27sOt27E(%N5|8>jSNJF
zCCo+XKNKg46AlEg$10;B$CGtf`3G1ZQC(cf_M62d#>GCukfTZP7|g&Hn$`RO?mb2v
z6dd|q$m2&su_TihWFeY%9E%L&rcFc=Z%Ii=jA3*@;FmAzq-fCWMUr|XafxMV+IHy<
ziV$Bzyas##KT)o(&a>(34*g(kFYu)}^)fTIhZI{#IACQnW=EiS2`}G!&8Z?<{K8t?
zLCAsS-D5c6e)vS?RYv+*Nm}t2;O-%!f#;uv>@SzHv$J=Ki0~;Wcs33kd{r9~8cH}K
z!ZXad0<mBHg;_bWN}JDvWJa(@&jh+gN`P%gWzxrk|E&Roquk<SmlHE5#Z<3z6aCkY
zj%u_I2R2#t1FB?Qqzz!gm1|o|LAZcJ>a*~7OyV6t`PY?ck!XJ59b@=*k6CVnKT{+H
zkhUn1w{6?TCm=ws%(~qaF_Q+C|H8^Ixv7o3_A>(-Vy8Ton!ks-Mmnf!e?LS?hpHdn
zdN1*;WK<dRsdbZPDD*n^+(=E`ix^72Ifko8R8KmMqgfi<t8MMB?OV61;afv{nYj5!
zJIk(%3ZEqwt5UE)`)LEd41Nd^;h-(29p!CL_WcXn+mj`2o%qzNh|Pm_6aFqZLnxS&
zf)tgCc?&Hp3-FuR%yWYY=1d>H%$MwOvH~hduAnI=R7dESBd*V&0Q17#bLcNG@NoA4
zhA`$0j1&a}f|j>@aCdhv)rk4l)bttgj+E+fULT6BF;3P&xzC82)2cNy1i_p+>w%9C
zE%*=r)}P(|{Q{cdN$2tsGyPIjVhh^ZqN=Kn7h6xBj<17;xMEd+yLzM`6my6mK33XV
zYmXm4o}!eNo!yeCf4x+x@MJ+554F+n^OELQoSoqaRRU~-H4xvb_Ul*U$vq<@BdB?p
zwrnXyG_#U(fQ1eymIN~ZgNV6j$U-pRlLrkJLVoiZ<WyxNqivA6IMQEV7GnitV+OP&
zEm;isbMnLqMSx}?CZx=A&n4u<6bL=Y)}uPNw%Pty!LVE@4T@d8dKHE0w(iT3-wDFP
z)K=7$X0)w1k#6Xwd#vF3$C_VJ0TZCH2qn$U3VoO{1*NRu4Uo%*hleMsi2{dl&hr*N
zQs>bXE=-UPAjzxL)D7UvRASHLqoMP0tX|(XDB=_=;Xhd;1ozmcNgwJ=+wKiHbzN;J
z-^p|5s!`|>kO5J?zXx?ccuH(EWX5>MR8?O<It+*7kg>Z#g1Es726}o(Mz=v9P_0@u
z`dyMxfDqo_BOsrbr-k5jL{oEzuyA?*B<K)h2{xj?!<P=B(X`C|+_`f^AECI8zn9A+
zkapQ#M9L^Um0^EAwQFGEDGL+jAe7h3oA<$wc2D~ukuay?qRy1E!}`&SB{o|p<a3Ow
z;oQY`s5WcA6NfrE&yhd*4t;%_f9(|)pT#`g5|BKqvMGMvr~xr0lGG=t17lCUxyR>s
z<HiasId|&~QJsZ^AlrO?Em1uqpZ}uNwUdgxRikrJ2)tfOqosAU`Ct|-$bjMeqN2el
zZ_W=~adJ8d<OU0=EIMH-XWY@7D`Jp&5A3u2Wg%<v$7rVDM%u%MR{DTla&k}5oJUMi
zO(H#AUEJ$3;Fo|c^d9<r0Ig(t#{siMAhd&|U6T<2i_pIhWVxWuu7TA7>ImL!hTAo}
z*QAAacCOl!QBhZSD`EtrUrh*6hXx1dKu{e5ACGo@R0Xqz4K`cIJZz-1m`YOw42&6U
z<ZfR1u0?`h?NqPgCF@4J9s0jaO-wMwg&xeq&`<?1w%bgbOP>R6&uT=B24mDjriqM~
z@Kcj06Khq2Ii9073sM0pu6U~O6q8JV0E6V6gRgBu91z(Yz?029ICVAaf#JDxPrC)a
zfHntPTVrv?VDbEYb$-oI?CUg?s|j|3)`0?|re_=dx8x5O{X&!L-)ZzFOsm2k)#Hql
zi6?B2@%At?W0mj6ANzOn@!dju06hHG4~#7UI?$-ivCs7$vcY<Hie7>_G(++R5I|bm
zFYv3efNSE#8vqA*zGobF>ND6IXw0ZxzL-8t^T<uBvORvB5gN+&2_2~YdU|Z|CCo(b
z4&L-%74C9@pT<sfBy@7Eq%jD@{6%pzQciU4llg`S0WM_)DbQzqJZEW3X@|J@)J&ia
z5EuzVf)}i&ZMBmSN2hy^N4!I=L&VA^E0s__)T(n;F1dwdYV^Pn)#&5{C)7iWgp(*(
zTD+S0VOI=8#01le%!jZz9?fP6UHLy6>rMuKY26KO@rxvLWU_m?h^fSepgKU0b^&g(
z&wZKq-F<!h^6v1D?iyID1qS&-a)Z{NA(Z&2FQA&BL%scjSF`pdYioW5g{L=f-aPO2
z7#=B-*dd$WD}TYTEt_Ah8<jbFS%Q#{x$mBRa|Y?i5k`)?zr{uVJ)x!jlsaZXSr^>j
zRZy7kwb27=p%^I(7K~@Nv3pJ8)GY_Ffw&nm2P|k$>FAVsE|hw{hsE8<wd=O~RZ&vb
z%5QS@9P{2j23In~Jyfok(Kd#;gZ_YY@~UzZfj(X-a8Hna4h*idgBb;KO(-9cfs~c0
z|0Rc978m~O-o#6U3oR3Z%a=JmU!FDIv0iApzM-LIUp0z(G&th^BgPuizA5{U7vRsm
zl4`A!Qy_wfIMvp+e`vhDy`7vYRM73tyxqiu#37@I3;?h+F`AIQp3s72Wu54LP0cDh
z2NoR_7ls0`r7LlZi-!V|L-M7#i_-;e{?7v&L4?5rga2*x(asgqm_MugXHMZyk3z1q
zle66+JHVgpVo9wNPjq6ap=6-8n^;0hc?=i?-iD+2wSW}v_j$MpQPq%Y3-A_vcu9%M
z!otEVuDh?lKbvFAjnxSEV(Cn7@rq-f@YG=_58b+TJ(v;;i;NeB_26caHVKhvFjR<T
zq-eXGcJ)Y`k*B98wrs$9-kDswEi57x9ZO$6UDnrU2YSL=6hpNJUIjQltB<c4j}ouy
z9=tJ5PLiODIAk5DS<8w8{QWVG-5Avy>ZXwJaETWx&RFS23dq6w1YXolD6`6M&cbXe
z>o&nhz7S>>L4)_S-nsv;09^dYehl7m2gq;6&(<ZMHMVzf2pyceE_r59{Y`xQ;4LG!
za6Ug|&0y#{#Atz=KJsJkmN$a8pZ5$fn;EX6j>u*pFh~%c7_|;%wPj^U@V=f4{J%i4
zQZOr#@IBa%z^lfB_}A5S5N@gU`^NGOr42UxtBwJrvi(ag|I7D}CJ-wNiX!;+9D&j!
zrguIPzu+%24R7<YKe<2}kSf(-PD6>9JSPX15!xjGfdhZie@+1%6PFK~TaN;A1LRC}
zbd;9|pn4@`KtxbG7z{`o>+vWH`{i<Tb78Vbbsaa!${p2FS;ox(?$eaI=;b5C*uIB%
z4Ctrq^X;=^H(QQ|OR`mFMxV$?OG`$h6Y;ewD&B~E1UPaUM>FnN1Y+r@rl!Z3Bf1nF
zVY`MTI=}pg$jAyD@ZSE*2yF}Cohh-21*U%E5-0ix1|~6q+7aHGClL{S?w=7y*Q{SZ
zfWd3S*0zi6f&X&8Ad?utI;&UdbNV5T`@~YW7yO!<BGnG@?7?$@qWQgBbVi0W9EpWL
zr;W0mgHVp9z=4J%EH`8i+(sN2PoAiQeBM<q0B{GHq!F$yK!<)vxk|wpKa1O@gO-NM
z$oF}9Pci=h)~e%5N`?1Kf#b}~%~Q}H4X;=3^!2T~{te>6i*E5bq{yK*qGj5+Q5icD
z#l(#DHY66T`PF#>7>ObdpZu@$QO8R+K>NNn>fVvvCyR7U!SGraez~TP(fgUGMu*o?
zBqb%m_I;_W0~FB3PD<=xu{-gCx7w9W5U4S4ws^OU)6b2>o9I}GTTxba1lhs$$IFK!
znbA-F_vch<K}Ye%Mr&Ac9-v&pjlARcLhGHAdrx;aWag5#Hcr$no`3ei<p|``5*hh>
zem)e>2<EDT2M-noySuuUb$o^<1QMSxWOzM=RY8G)W|+PUTuUzu{a*)R8GCLoHqPiJ
zHISnO{8jpGvE2VZz+b^OZ3}?^j+A_P^UGtRaBX6gF4Yat=@F~OjUmkoE>YlZbH8Km
z0^mvbA2k2Oy?Yl3aPt*R4G0v-9{-{r?x|<gJs=zUrBgEFYb>DDppw4+q3{$v3h2^p
z?t$=UWl+=7I>PLOi%HnUOuLm(*5S@4_U7Re%W;e5q(eLGYt;2@6x>SKBza-Yc0_Yx
z%i9)qxXa+p>sTCdnS)iK5{>$hmf21vK?=Q%PPzA(K?t2(4H+HSJu@q-DUctM0Jv}X
z>i$LV;ajW{(wuSqh+`hs<%wYRLDQa|9-z662y<{4tX#ED9<g_*IV}K_Dc++go_Wt8
zvBoJH2kePu1v9C2JRIX#E6H8Zk6!A)o(3TDeG|ThrjFF8W{9T`eFMSHf>iRRQW0=~
z-&-2(>${D|1%fY@_5K9xs!~8@gkNiI)qrspVI0*BF+-GY_k<Spj`DHGvfFRd%Yg){
zf+HI^I5h0+#9)HTz0_tDP|OpMfSx*LM9+}oc6ewWBqoZ&oz;}*xx^$rjPdaZ@EoKZ
zMKwr_IH92?dxJqxkVPWIT?At~p~g8mIrA0$d@&%)m$(gqT0$N_Rx>u{66^F04h-Z&
zB5Q8egQ-piirMCX)eweYXOpq(T!9H(1BeehVDYFLeNB8j;U$lD)_euaRABrpBV!9Z
z0ip((H&905N?mbrS+$f=Hw@&K^1A@0ENmUIKUwDS4jz}ryDd9w#^Bn#P0&HO<up{Q
zam=d>`n{&Zwe=_b9iJ2y3KdT1>z6~m=;dog6azMZy{{1ScdoZo@A%gz1Ix@aiEzD`
zgS`pCpg7-ek2Q?xa?Wb|NF40YNSK)m@K%Ty;q`0Z`uh6w1GrNu=*&c0)6Itu8;+HZ
zK(`_H3Bnhs!)0|3LmP-~mWU-<&GA`Tj{)2<kz4FtRx1Vx>7<%7;FY=<uvonSrK(KW
z)$<D=rr8KFou%yEkA54ErY`L?FQcR#V84+DZ^1!;nRSL#R8+(SdEq%1bXLLuZU-|u
z(SxDHJX(sQOoK(HFHXKWGog4XS*PSmip6fgX1S=52i`@R9upgMS2#O3z+V%&BihxZ
zRZ2TG234>G9mjf}hmb^oqQJ;GEY3U4jhaINx+M(`7}ZP5)dNuGD$FO3$t!D`!D?G$
zA(5}B>h|&&(g#CTRTZ2QLdRRR;_D`~&ZaiKD^m+KN+R;z)mHD7UL7CzH*eE%3{1XJ
zrqAh}<DkXL3WD~`eVMF`{$Z6^dyN4rkSowUK!ul8Rhff2MJ%^N@yh+l44Nv!=*CT(
zLLNR;g?kncw4a!*P;ah~kdpc`^BT_{Dn=!^Mijj-5#>p_6|ft@1JyLK;KJ8Ms5vxh
zl(Mt418QL|Fny4U5{!G1>$5(7B<5~RE`z%Et-k(?tO+)sqgsCfOE1iJH@&@m>SofU
zNurW<4myi3zjEW_jG$7t!6UJ~t#xv?-d=8bX|f|nVzlD>EM-FX^mL|8fVZB6DQ_vg
z>N=JlxOV&P7~-fYQX*txgRCpW3G}QKnu&9(v49Ff<r_ej!Y~G3>D`(n{2Vc6p-15l
zFQEASjFsro1N-#n&mR&l@w%tdTeX(i#J>!z(i^aeJho}O31z-)0yq-N*D$<m_wKw)
zJ<!0*LEgYC3S1#eY7C5svWL#5dm>G$r4GC-)Y}ISO5m2d@O(P{!-o$<vB7MuJAKzk
zel#<CkNgOE><-7^;Gik-X+arnbcJ|T3bse>=+Us-JKT8*{|aHf$Ff~6lmL6!Ay~kV
zkoZ>n?OPE<DeLG|=V1~LbkKC|9UH+)aM7$B?I_xG%J7<c`NVP3FcD`j#+1<#^xkkE
zsa6RQ1>6haM)a3J9#LWjhK5A46>M(*FO=&4^$-AG!Rso!y2=2ci0G!=0!ayI3=Mfg
z$0}JT6WvJ2yKw#0qCEu?J}jnUAid+>yiwbn`7Q=E)E;=Do+l(QV7=@4MOHX&AIrOl
zLh6LfMO=;p@DqgqUEh27uoR60W6O3MTJWt}@n~a5S*`(WZDC<?@AmDH5gTX~XvpVA
zpA5e2ml-Yh=e-ng`M14Y_EqVtvf$ExDv(!))PRy4NVU*PY6BA!@s0HPz33c)+v>Ny
zFl|zkl>nX+6gpr}C=<XI6X>=J{;W0_CEl)mXbH{}JFH52`}Xbkk&zNJ@y}sCz!ZP5
z4QSEww(Ea83|>CJLOW`#+%7k<3{OWH+x3ZS6|GcQW($AdKjz2Z)?G#iFTbqpR)|c*
zIfCZd${8LE8nLP9@9Nq!l*y;Yq2Lh-vk|_V_)w-(sSyU0Ej;#KIi3a2Bka|T@uRTS
zenQ(A0bp%xf0mVuwX~zNkr~0%gg5kdp(f5DS?3{EESOl*Q;h^)2WX3D#KOkrUts+H
zw|X4Qzu$h8Bsi4cp?uHL__1!CDasj#p8MlfuIJV;;QvLNAznK4L#&bqw}F=pq=2?I
zfaFm9dezg|Lros|7Fh5;_iWS9piL^%SKZPDR;#D2z+D(o5_Tv!UhhRk9YcW#bon78
zqkINV!AeLhgnYzuF#z?@M~{w}neoX**_3r3QdOme`IsN10gR~68wSClKU7daQU+N>
zkmq~jl{Y9~SHZ`l0y+dDJzg;)3hI(WFHezD6zu`%^8JvW{?E5>)bx)o;F=bqJ@Lo)
z@5Zpn@_MX415`uqmwfa<0<RZTx2{gkgTJok?DM71y@EZW3%^06$%yk6L+hIE(PC^z
zZ3Q=D5T6<v!0-`fUwO{&z|=+QwWwk?oadK7Aq95g1~~4T*498iHI#%n+;HY}qmu4L
zYR>!7vrP6c|Kom%vI~l0rz^<=MYTfjn|{lgk`s|NbAttx7#Lwv2jZ)txLr}`uo+Sk
zcnYK+K(vFK4QYE|K7&I5cTfY(pN-={8}LOJqT==0Zv;sRJ$aaqa%Y%P-hEnxbZ1~h
zp^^t^dbC0gz{+EWhu|_bpXVPI7G@Btafvttk0rbq%{P8W^=f}nLUJ%};lM6|Ftw15
zl~Koi^b)jQ<l~R`Rp;ew&f-&JyU%H0s?wR$G`gDiE@9y-soP;7h9|REfbhrY&>6?P
zE*XkxV$LHvDq_omxjY6M2r9hVj*i_s_JfMQ<>NE?N@39t)K$dYE&ol$KR@H84a`sw
z9TR{Aw%Ul_UtY{^DP2dze}O)}?lR2RfW7^Yx)6Cw`rVKx@f0?NeTJPi;TX3ugi{ik
z5OvS<bKtz<yQ5e}H9vm-F1qj-QF1B3fuz(=jh+=^rbh2E<7YUBHldOxy$<9jxcr{*
z^2*EOQ}ZuxGX6eu?sLaV*4+H()Mte1?lmNR%gZCJMYGbgqqa%z&zh{9bqXsAWDGi|
ze*FsH;c1?#>tT90W=MuuY-|L^Y#QHt0clXf`=ys4ARusO)MF+a1gS?j52u^J&HQzC
zU7uEafHOP=P=y_Qc+sfYK2x>hCC*mbdYGPm^gkZYo*Lpf2^Ni^Nm_hVCh2NR3(yYU
ze}#pu?EvaCjQ7MQ+!EY&tNuOp;v(d}W@M~?^J?9y#IKlB0P@`r#KZ$(BUIclkf9e{
zMnh1Fm7ch5@VEDX>0inPw1P@@Qb$K-Xb;@urV#h=zRFGdqPQ-P`Lj`CXhvtFGuqdY
zdc}dk$VZ<RG@tO20oK)jFJDK^s2a+Li3fdt-can|i9t{eb}*yTCvq>;Z(P>LP@)iY
zYJYqm9eo8q9qt=~zVu)IH-Dus<t*0;8X!gTsR-Ld!JOfJQ{PTQ^JIJzYb_N4n{G()
z3J{<*7U8$G@%o`MW6dEO;h-?&9&wgoqDM(diBR}%Jwej&pXcr|o9Jk5tw7AHAg~Ra
z+4IW!cu`DY!-`BEnw+z+i_OPsZa*X{(hB+o(=sd(iDNPkuO(y%{o&tJuL4B?&;@~I
zsJZs<k3yfiu^8)}cSpz?Hy4WaqI}0(m2b&Tra)4^D799gJ|rdzcq&c?s@n;p(Loh-
zG0N~xBqE3FDR5xO$;r(E*N`4iw1LEOz9j&D$hS;?HC12N$jEWrb&z>D6DQA}Jq&t?
z)V8${UcTtt!IZG)u?B!51e!ZWNv{$Tg3zb~NPt)l-9HLeWdj2PBKsjcaiL%#^&+;w
z!H>6|e&`mdI*2tJSQw2NK-UmN%>WTwJ(m_q#YK#=&@5Mw0E!V~$Kk62E3*P*_l7v?
zNb)glw);G^fY0YV<Ypkp;O~?Eo61{M3E)pxm%yG940v3NYY0yc4qx>3X<caja{eT2
zTE~}T$}ScLYZILpvO47Of7;rDj=w5|yoAp~a_sf82QvThNC42k+*5N@0wU!VW`4FI
zAHf>256XqZfo+IaaL}RC2G%^*^n?a0san_-ml<9(T=!bSH+c^k1`m>y4xviK#EPc?
zI`vLn5!xV$xsuGp5{3=<#j|HOaQ17Oo8^WU;5)w;8hTSDHtWYA{>r~4g-M0}D9!&>
zQiN>z1{c}V?~YO{sndpr`%Tu|_C8mV83+AKv^xIOn_jPfOp*n~b@97kB5=D$?&`Te
zedl^#)F3mXkec{*S~Ez!#@Zc~Ltn$Ujt+&9!=fMn_4M>euS{&LH|i81R@kj=D&Vg`
zvV5_?it-`bsheX^Mqa)=%oiuA`2Q~*9-RDtONWAj=t{YFvUhly-3Gs#Bt>qbDcKcY
ze$p?@-Ll<Ag01v=m9zA+d3k9#xD9|G0h+ifQFozn9RGr}E`c+Hq5vchM-?Ex!?sJ6
z+;x{F5mHf9B!gJ{`0*q9+)?xs`vkad;*@?>ds16_hxGg~tPF5r|0{??yduE>sLu?Z
zJODvNc(do}si|Q&IXFG|@fy+ln&Q63_fqOYOdE8HKk7UyqFBdN{(TGeZtV1|8xm%k
ziI2Lx!H$p7Wy}Nl;1)c-b#-TRli@#Ww|_qCC-9(Y13@O|oW{sRihj&#4-G2;ma9)R
zEZ<wbZ|{@y&A0tBS(@RA(Da<g`iRRiNTBAaTVrc{@xuxWSHbf`h8tzI-uYLi2KxHv
z(Rar7B%TSaQXYT>R3oFUuu$X3qtbYZ+lG>vI~a2VGVOb4fRuq%h&Li5W9FWpCm;`Q
zHbm(!d$tCtKyg;Ku^D+k-vO+HZB<(DbPnQopg##dH{5%SDB*O@efV$-1QN91U-6BF
znPnMJ?5Ltl9sgG?j};x$V{Y^1^dXtq*h@Q<4_Zx@L*LqEVMzlGSxXDmOW2oGVc{d!
zt?|uk0;(bfyT(B1e{F2^)rx-^AHNQP6*2@d2tna`a2Vpi-}wU(f`YPYT2JyV50knc
zB9Rgug?GmoBeWDeFSFzsq@amC_`Kk$@oho-;sTN`YnK8X9b)YT9Rw-~s1o8fLLiL2
zs5<uyIm)4<35E%@Y;j-B*#^Gt+N>DVp679K=4Dx^o8G}y4Lc9)asRgVKlGEd&Djth
zo~<6L4DS8dv19OgY=C{2Fec;FMn2NIzcDm%$NFdRH+#kI-pwQvNL+a^Lo9WmRTaex
zfpGpO*NlW?OiI{;2X`P)f{z9s#epXs)HyjhM0iK{FQ}R~HH>Ew@u+zNNn|dFdUom!
zrV7(wowYGSra|tt$55iMYUz@NMZU2ZvcmV_VW5xOz;n;h#6B_c2~-QvMheUY3yX_=
zDz>;!5P$$ui-1lEF}3;)2p(eNg7xbQ%;sRl_#kycZ+jabPZ~4{FD7>4V;4rc@b1{*
zgMw%qMyYcL8zDWh-o?{~AN&$r*T~-w9&FaFfKBpwR+d{CPZddJx}8xLGURlYdfOe6
zVI-Hqo0CFFLj2|VR#LHbml2B?ml?hKj9kfPav>-z&2@Ff)gIG(!64;>6Gy_CP(bq}
zOm1^yh3nCxjN0^X56%Gf$<TG(@fuhDyAD7Fy&YC?;96)}CxB~_{SfOzFlS}n3ASEz
z`Ku;1^b8LNfs}^l8wUdbObMYHSq5ku4Xq%FcmS~R-?$s#dW86uiiK?uQck=%6M5kJ
z6L5`om~g^90M0(us!1Gn5-FkMsWr;WwO~w@zU{;D@NBjq6%tL+0Yb`AYmEFOV=)pE
zhJ$}v*aD(K8D4>k8!3&%TBHWEkn*Octzgb@gH>e33)(vS`%B1bYrwY-c+f1Ytct*X
zp%8_G;H!lM?$zW_TlUO@Hh;pMlB4oP>E)?CIM<{>8zTAZ`1n{v1XzMeFUhZ8MW`nE
zIoR0QQM!=BL2BCV3c<ikAWSi;+ZScWd&m)BVN{@Xq2((7Mp)%f)@bbUKod38|8b1u
zZ~%M+PelZ0v9AVX@HsoNy})yT_FO%wWwMm_{R7SS{^{Pu`h(S=8E38p5x4B%U_gO!
zes%R-kUPx&9z$JrUJbRir;UO}M@JuWo}(&I?)F}P;GSV5LXdYsJ_egShDT6PRUlKJ
zyd!AgWX=I1HL7P&<vdCAAo3<Kt9pzqs{5d{iL_nPaZhN0tX&Tkm2S<NF4P|{AbgRl
zqOP8KZB87<I&_7Ub(a_?Lod01^ClHkP-0z>lq5F1f4+WI^zb-<IR;^EueXRAbRslw
z*zUQo2esh^RI65abvRk?F?CY+`|#-KMF3H-Gd@V?BY8`I*6bpE)TQSQGgH>4d*3$D
z@7pl45rhH`8~$Qs8}xnpELZ`;Nv21YWknXQZbuIUnX{t^wFHGiFyle4_wt0^1>*|0
z_}~<j>$Ni@)SZy<x<K6E!`sdOx`cL(zm%N`oEGqaK-@*da`?|f;97nNw)kqUonEx$
z7if%UI1K2eK@XV*L^-?{R4sN>KhNq8B_t)if~A@49Oef&0~I`uPPGYKH;kGsL$v6l
zvBmEv^$SivOnw#UWt^?K>p=`;Pd(;$f@dkPk|KP!h=>^m*upyV-n^0xFF|FljhT+x
zbQAi&exJ6qj2CHU{L@fVW3S1759K-@3Q^{74tIJ&ay6kH=Hd{=0LhM}oJc{mMk8V!
zgp_4aN>5_m>$;!vSK9%eU6p$_(v3U-4ef{-H*+feBO};GxgHj|SdPb?Mm2*|`xbxI
zQm_IF!xT<wYoo!x@yGBmY;xnUN`xW@qt_s}hCz#5nwjPJ>cvauJ)rIYv3m!iDgrCC
zb6gl6byL$=1Mzo>iNUx?czx&sT8S(Q-RTtYv5k!=wlJ+gG;;)NLsVjzsXb?0JWH{H
z_&R$1mpju@E30~XDunXcli##$-w~umf{>xf#x7Lh#Hd2B>76I-5oCyu!umgLQQvaY
zZl_)QWISF!p9)tPO4!$D-ZEc3yJ0&rO{$Iv6xL*TZ<wzey_8hR+yamM@2;OezW`Yj
zqW-^IY!V{>10>f2G9uwaVXCmYkX9ago5)gjP_4(1FkxFnEss(Z>5i02c(rAf`r$YJ
z)6gxojqPlp#D<*DPYBCWV^-+_GeWR+igVJnv_K^nmkg2jI;I)Z3&=GlFv37;NQGbo
zpW|mZ;9tMCDC-UfxJK$jofn2Ti6;PKu5kk<Dp3k!g=+sS$=2?8tP!ead*#hPygBO3
zZZBW!)VH@??^sOMsZSeCG0}B$GSt@HGIf2|vv>Z!&$x~}WTIx~KH_sv{L<c?VH!_m
z^As}stxTv`G{n{V=L{OM#4EN&XK9){R4rbsUc5HSz&>_0hoRjQTPV>J{t(vF^dI&a
zo0kn0n|Dsu$KhYkUYj{8E&CrYzySiokoBCiIkVf2U%2qHdzI1Ivr_>Bj$j9%B>`xm
zaMuxeKyx=DMbl#4p-HLl*N)XU8+6{@t2PXW+blYA=dem`+;tyHnxjomvEY43-50?1
z0b^<{4|RKeA$t9?U2GBZ>1e#sY`zCm)8XA6Kc4q)Z>VT^ipV5Zv7}lyhytM9D(^cK
zUxy>?qZLm~l9=o}hJWiL|4j-=XxY!d)+=Lj4TWo>-j4im{I8rxI}`5KT?V10&Frpp
z^aYse@X_QN3#4{5gnhyg#7AQ1selZ?JLYy_;Oi%w45Z>UG~U!yD*^g3ai%Y&T$C4M
zYap3Sz$+{{Z)0$6yqG*fOyIiH^Hrao0JMy~Rr-q_3h-A!I6I%w^5*T^7&Io&ZEXgo
zMus{P`mHlg(-CPCuCkqV8R%5xYfT^csn0!@zEUpg_?$yp*gd&jhdQnIV_;d?u8h_g
z5nFXQ+>u==*t-zbT){3kdV9hOKc2)R2h1M9$hX_o12DrGh;E_rUWqh8I!Um}WXs5j
z7HaaTFkDKQy1i=WnAGhu{T3ShUE2<I4rfx?UWZ0@|4TazGlKVBrzY)UYsfW3VQ&D_
zbRefZ(PBwr1A!3BUyq?E7jdW!uGAbRF&|Gts4`aR*Ko^22yEiU`8S+*b}n!bcZaR&
zlZ>l5Mu#&Qthu|=|CnO^TDxiV(N2cGY3_o0lVnWw=};^}AFJhb<?!@PCfAk~A%Yp4
zQV|JfOYZZn2`Hvo7#KPE!XA}r3YdE;uN9DR&|>P?Vx3|CdpN^K{)1Mju^2j{7;e=5
zu$tNO<pCG)2@(E4aGf?XLZjTN2NX0nZGZ>@s`fh^^qET5PH?*1_j5q6ThST{3ah`c
zPJW1Q{17-F|9p16kxiHIr||B1CraXKEY1jgMn#n%r^eW(Y_r~%l4i|BWjn}?hsixw
zzoWMp7^WOL0K2YCzGqRp$Kw1R%r$@~gfyLleF(bMj`}}N^Jx_j{u~b1EYJd*VGnY{
zXYA?RDsFbfhvh2qwzb=>qojzyn?S}AKpFBs^7}M*1}N{tJ#dilpvMg^L&)eT>HKf7
z&tH;?kpI@<=*UcE%Z`+!Ab6LSfo)^S5$c=k3T-lKF*YCP4R;A-jW~7#W8%LCQcxjh
z^t1cK7~_h`fl&2gKYnC`;R{{pDIx^6!!3&@#51Bi1U!A8{hwcV$5{O<*z{1cV*+9U
ztX1SIKd##Wh#1Hw=?@ygG@}JQ^p3m*nn?|N?Yor)jz^Zk6$o4Ru~_ye*r$7e^bTCS
z=t6h?2)U=$ySy@Tj~%jzPfU!1VWnUamnJXspzigb?sh&yy1@V81E`xI@@Vg~<#n5S
z%;dT{%D;Y=%7-pW(7S*OOQ!1DU+te5O0Ay2Qx#fc0@9f{kRdUSF!@#OJ{(^s<d;&~
zlVz;%G+0H9wfD!Xr^J8{uS(AqIY@nQwWEd54_)l!cA+;F+0@Rb<uVGyzF9m!9-EkV
zTJac7A`?~R8*ilb=FAtqolJDf%2R(dkx=oweX~aj9;?@XH7gi)&Ukx-<;x!G5N{?|
z{q@{q&Ng1sw?xYQ8?%9nJQl`Pi$B4O&s{E#N4;s&VRiMjpsM*n5aCuiqL-4!R4ugV
z9=8(wGi$-f6ANdM)*W+mb8%RmRDwC05+B+w&pNBzgkk5Jyq34-H?QFyPk)*A;A;?T
zYE!FXenK9$EeH>!_6IJH7F#Q~@{@Ol6RYIqjgapV$k2fdg9mA|BkinfT`a5*x$FF7
z5k-7=Mf1Nb8aXi?s;FmN(1QZD2u+nXAAbmJgEiF7r};A7dZR8hop?*Lg!LYo>fH2!
zfkdfW`RrQ}N6d8NK-z$;8iUi>5ymZS@`{ZKgtrXyf63t=pSy-SQ*!<{@d@MWu0ZhI
zg5@rfjZY!e0WPBBNF!UsZn=L`X|f(;h2CjddJC9w*GPSP{qChT%b|K3Zm;QA-PRW8
zuCJp!5D0Z_s1g}VugZ9N+b+uhq^sbTsBQbN-je0<dUlDR3oL^OT|hFyFrR#&#6ph8
z&vuDnBie{dUJu5tV00^rp|^C4+wRgr48bnMvKuQ`!HM9USx8{>LE&Y@_%#|vd)Otu
zzC5}mF?&P~R}zbcSCJ7s5O=giIB-yi0f3f~gA49fgn%0SPGF_`5i22#EI_2%jtLrQ
zf@9gfy&ARcPEk>s{L5(*;OF0?h{77yO0>@EAA8M0_vFxn3s}|EeVFsIqK>YvUtw21
z=F9%p`y#~MYtPAIF{qHZyH!P1RkdjCxq~gWBULHLj96kgMp%4Q`6|B;Q6|0!>m+Ec
zfa8fzl=bHO2@GQInfV=ZVQZS+G0>Z2jp383HF64|dwa3a8WGm$Ie;9<I|;U+6f?t@
zr;=evByWv$fpR92DoSrfNXA;~o58+(YQKh>!cg;~Y%hgljSS7eZnQ422<mAjhK8J=
zGR+~ULa0Kqb*I4igithEb`b*@5WvAZs|2wXW0=TMn8Q~<0VSOBHT0$teilziDE$Cc
z!R=G$C=7ldq#_tFDRbdN%kag73S7O)kBp^7MQEr%`=a$fq6e>KtnP|4yzmhF>CnNP
z3cm2^^@Q3Vy_l1`ZUqQ)G-G6^6>gw=QZz7`j4ODvfGb5rI7$#NZTrYqYU_#*muh7k
zdg<`z@l3j|$YBtHOw{qn<2wLJJFd@+<Kcv$f*-LXrR3N;VjtxCPJbj+8SB5$lR}HH
zA&(z60IAm)yG}Q5-YjHdbL?2qKgR!9Nrm0JcLVSQvqR{pF+6bfGYlLgcoIt%gbw67
zXJcZk+wwpwRQuu=zJ!~*-&xD!qsVom&E^#*cQV@#Dx^1A?K0VgNjB)u^FNF@*%AIO
zmmx~=Yx^Z_s!?3CBtE<W&3;mi;XiHuDYr)(+?O5@Dgny6Oj3j<Ly?O&uNDdml>kD`
zQPC{E*5_+J<#W2UKA#=ihVI*pmLL?MeGYg4a1eAuCnIfqd?3VIyND6IlZtuyi(BD~
zCmH3CmcFIYr9}nOrAAnMa-etaJgFrB?&`$p(@J8jD5=j&uz`~}oYVkM;2xCDX!a)n
z2F9(hr%%^Mj>{WjOePs%MJPuE1J<AtyKO04^sf81!TA-QK%p2g+Y2xIo#%^lzk~k^
zDE`Md;O<?RxbDI9g|T{KY(dNo@+x1mE=g030>sWD4Xhv_4rVmPm!}ZdBQ(O@XGRkC
z!416^g@tiMu@KX@0ZeN+J7Q%i!3NZd>9)*AC3rUC_QNSaWD5K2HaK=W$6Yrnw4<vC
zQKS^U<N~yy?tbtCkK-V%a&H2a*Rxb8)g+X{7TSXbwLVlAWUVd?!pAOEyg2nH{QhM!
z5m&tgtu;f=@<0Mm_~vIu?_vc1J9Z96T6ncD!scn<+;oBbiWz&pp=;a~-RG8Eoev&9
z6kc+c;mEVN22lm|qW3PoZUH_%^k%qJz12tRvk{?_;~!Shbp;f<U%ETz7mLvSpOH!H
zo%T~B6e-3R^D{-RgF1n+DfcdH)qQ{z28L7o{HI%XJ_2;SFdo9#xk|(!lj&hOnOb9%
zv?W0+pkeSO0s-WdG1O^MNQat_hSrhqtVF&uS11T1{6KLm$}!V5hj(=W6a(_<mRGfF
zB7bOXr=xzaN9K<Tp3Ong4nsN<nLu*wz;-Zj`CIRHfZZXcX3UU~5Y1a~s4#bI3_fk=
z;hNK~^2r*DVJ2oD(q%rs`G$06hlXxk34bIrrYM@2#0b7%p{qKS&Fl)z=Jmd9V#YTn
zX01M{F22o}n3$+vr@<vdi%l;DgHuA&^k6dK)b<Gw{U}IIx&{~$VtFO;esBua>FthD
z7<f8oe%s4%fWZw%WEX=k+0F;F6ZK8=<&e+x=16D}cC<W}vFGu~dy0`Xk+4FffK#U!
z!GELj*+nx_rT~x!-;-NXa?{i5SF_UypLo-tX);<W$rvTXzejRTLO9_lGDKT_*piWf
zF=G(JjJ0pI%-==lov5o7`mIl$V!_WjW&>wh%<N&VOvUD)e0FhT+9%eT9_b^#R0l12
z%Ehr|o`;E!t6UrfDjpE#hYN|SQ*~CnSVGy}k6<u>91EKd6%#$p_|+WD{UDeK&ynZJ
z#2WGr50Q7MXxHCghk*YXzi>er2#G_&H?m{?V*#xc5#S;+3*s*_Mz@Y%Kmba!XAOgR
zTA0s<uTR8ma7`D{JB*;cE})pDKX&L^S&9Zs&v?;jNIb=Ct3;JpY0k?;XYjQVpBD(2
zY~bO}_)`r$ut$3{ot~dF?^sE3@yS!Cs3NmWH@>|h(-JD3)DXXqOnV%p6=YhtMSNvO
zMv1KXK0{kEkd33ijC$A)+dYg^n{*I8eAaI->2rX3Ip(?#VFSXG8{9}_0_u;BJkP%w
zyufYwb#htwqGc^uexwt>o&lzKLMDRxyDh#BT{a*%@wHIfMpWKG3m>>%JYVQ=n2msR
z08Z(^)4MS6kbY;()h9h;=)Yu~;dU0J7}>^tX}RaCk3AMRxklmx0b22Ai<-Yaz}jWs
zOo;b%nN$ZKSs*|mUl2!#*WRNIpfaoG=I8HkKuu{~`IY*nfE#Wx;89shpk4ypEqfe#
ziC^}8lT{a^=GS11#}WlHo?Y}nceX#aBr3O`;7#wq1i1xF&qTMV@V|lj;&O&!iX^(-
zHrLLTq>&P(DN&8U5tiCy^h{kH|Gns%V+7g{PFTfPwao17N6D_k0!#77Ks<!Yg26@H
zc}NE%BMyU8eRMbRIu0G=&SU-=1E#Td=_=KseCLyOjvPO(g;T7LExb&2t8Y51<8~%}
z{3*nL5;8_o){nRdd`ancS0%S=cqJK^t*8n#?d76ivr@X07}qcet1l@ryBLpsWdy#`
zx5>H<?|SjwieYEeUU(XrZ8q>rSAhI1BeEx=V5ZV$e8vLVzh84<9{>fm6z)NFgDlsS
zY)Xqe(hbal%DBrp8(c%Cfha0s?2H)yF3A=|fM9qybv}s2Mzmp}fy)e<D)Y&WT^K}9
zmoc?yS6*HoVR)VWlbLu3rh8!MENleK4sd9w9XX<ypp`Kef(ViRtc{MMZREy}N*gyF
zdE>F9CFwCdChF9J4?%)4QD@OO`ZA$#ld=qky%heq!a`4jj$}+<0jrLE?Q|{{H=Qwj
zrdv91jsXZnkt-i|0UTu<g6e_7K8JDSsd2PYdcjg_%OIDCIPCT8`ZABpWw8x-5}0Q%
z=K8?XMtalFdCc)+0A=H_Gtejq9K7V>e>7ri$Rsw{1mRPG9R^$GK7+T!esXw%Wv`63
z?{D>SE<jYm?uJud`g(fJibV`lqZXoi3F%%=hb)e&sd4Yzxdxm7a3gBd)2C}csbc2l
zr^9&Jv9Ym4dLZU8oQ_{V>jE&Kj~S7WVva#O*U%J?u>(sdqJQHzMYOdZ;+lPV4f7n-
z9VIb~8b1$+MdelPZFz}<@+$FlDHuU=;gTM#WOi{sFwrgl$CVUdwrJPxHDu+Zz0C$Z
zdS`#4`gnDILjO`X){%B%TUHNXyD@SR#;b0Ms3h|o;n>01=?uJoqqOJbTX*vMK%e{_
z#TJ2jbT8&niGKVZBewmlshMs`mz~7eCx$gy&&|!Bdx%z|HzD_6pY?TS?M6X5>JT-2
z)0H6F3rTM}W;-24Pf|vkv%dGD!jy=9irE^}o3@2?tEmW*1nJ4eN>2sZ2`#J|kHoku
zLe|Jpp^(BJ1g59eq#q_(FG1a83mr9N2x92KOZkHlt<iPcj}qFo>m_er0wjiaI>|R7
z%?jQ81&sO#7OmZS{w;bvK;UIKhkKELvB}^IAS8)I5a>wr{a@h=ryhZVPmPC_9_>X(
zfU<E^KpeE%NQ1tQ^<4U60JR^CO~$d}V@VAGmFAYl&rjsBHS~-de9ffsJ~&>Gu)9+L
z?FWw@39b!5;^1wLa<IchbGx~|hm(W=2SSvddxE_^{-m!91=k1rD^vxK`!54T=swuw
zerM$>8sE{_P73h!0($RO7~=+@&|7(NWP;ZFAkL>w?loC_ET3&xkF<K>VIX>_mg%p@
z@ISJgxb+bYm<iGsgO;I5pgN-yXYDP6ke+!qgncd|PXyWv4m+8s``3J|h;xIjn7aIM
z(7k^126`n8SPk^AXvN=0nIi@=k*r0++OI|)*(<Nt&b%_pe4db!GI8^*-C3bckXZd)
z*;6#>)_jC_*@O|V;>qFc4X#}kp+T7Xb(ajPg+>4&9PtM%KB)u)RZKILHYy7O9ykrw
z@0_Mbo|~O@70vl|%6{c$i2&5j+dO8^ufU=z4FIeg-Q{Gi7@6*aK?WFIu*92t4*B@&
z@k`Zi@r7MP$>3(C7YBKim@}=1LLIr3jA15I^d1^r^~0#{7lwfZquBLRuEI>?fUr7=
ziNh#-q;eOh_0cpy4wvS-6+rYY!*O7C&=yKcw3G|$FtVigfxgXo6s`&psy-AurL!3d
zi*C=)B1ygR3Ock{X_D-FpAjME@TG@Twsd%UXeRv*r%a|2Gu;Ye%T8_?R-Z})W4d#v
zpM6gWxwy1N77|1c0TxED^C@McjFyBm6S*irkyEiHPLz9sKxX;4{rw4l@r#(pf-Bh$
z)i(!ROfI$!Wn2YT8Q@;bWtAPm5vMwM9sO*eXfC3~0X4;Y%=gDbtbmbK5$07!pH)w*
zRk@f>+`@VfH58nH*;Y*{w^0E&6f#JQIreLyf*^)zJb>)B;JV8%k?@goY2ulmY&ds)
z@YYkC7_nnqOdt4%(bG=X%nlK&cwiE|44_+VP1&tA8^95eymSz#nY3Hr%fOxx?@JMV
z7+@^sixLxHQt4H#{r{osz2kcR+y4JPC>bHMWt4;@D-C;Ogpid;R<hGDDiI+WS*es1
zvXed{6=_&W_R45c2}vm-tKa>|=W|}y^}T-AALs2nZ%!Jo_v?8)kK?fpmgf4HO%s}N
zd2>2q0+AGKSC_>4eHk%qm@~)a96NU;VNjAKo;pkAvSluIxV<lxa7R2U^O;h`8i54t
zdQxgp#VhV-w|=m%#JN*bMg3E^joUiNFt+IV`uR;hHUHPzYP0Lv<6AiXA~$jqjTHbH
zz~$1)n(oyVgZKKpegvM*1MIsQbNX2?_-JXK7<gN2@jldU+;833<8zrj$idRFXiBlH
zvtcmR0RxL}6J@8g)~h!5wXOM;6aYH*&by8>*fPlh+p`1>gV<NJoHKo~{0-B#{n|9%
zqx&Fw=0GAj7?@j2v?E9cUl9Hzb=Iyy?G75vZ_wYGI|yYMCmuo&g(8z1!-)Qsr6Utn
zel?#Q&s!Hkem_6BJBG@DQ=@7qKE}ljSSx&Z&pOhnNG$Idy=HVh$-L&wqF|Qm2Py1K
zyNWS^rzYvA^;Yq<2DMvI#9Da;I?j9VEA|BjDy0A{T<PgZjR19B=n_8a=%Oe4-&RI`
z)Jaeeof4@69(m~ZD(mBq4m~i>IX|@~{!(d!cGe-IE2Up0|AZeN3kIOdymK#0{OrO}
zBatpmiUqbST91Ikp}U@GD%SzsC*|+w+JWF2^vZz1a?`TH@pExzw>h59RGJurxQu(r
zu_wK99xaM(^;kDLB7zd-^J(|+U&)`I_NnWWF|9}W8T(qPinQ@!vsjzl)+Ufn6g|5W
zi%^D*Y1(dqoR6<ru8q-hCju#>_Q9PlYtl6=`u+HD4!3w0l^5d9X}4f2JyqVoIg2N5
z)PF>iyM9n!qDx4imI=l6c$mkj^fgp1u$f;=`gL8OZokO?>%yIHE5G};jDNK~YUGhW
z$(FO-m)Qhz`WfCKK1Z_k=;4WS+pzfg&GI7-AAXyC$ppIKX3T9);4=as4TBoM?WX0Y
z0x8giReAC{P&pA`7x8WgZ9th66o{Ea-PCE8`Rxjws-JaV{nA{}*!+7@^pi!?3x^-i
zEnW7h$Gy667j=qKMVX<mKcT4#?Grv=b`92FHp_eLW?c(g)?f05OD;TGM{p7fMkk%p
z#nG325OJY+-RgU-CUXByT2aMcMQ6rFd8xqZ6s@pa7{!4cTm;r>)Wmt{Kp=LfOWAic
zMiZ_V7&x5HH-Z$&Lsm~hUzRu!!X)6Z)&Ng64BWJ5B`dqkiAK_0=wijYy=aj>cn-G!
zA47!iWxRYXi+WJnKVnH9<kTh+RU+-!(S#7Y==r=2hkKnr_(ZNSt{o8=(1<pn4QG7f
z!}c3~6&9Itm>-*IcXiI$e-7jMUPC53YA0tMQ_eQ6bN7<f_|ZF7KN!}k2Voz2Ur*L3
za_@FI)*y9(k^b4(H$To>(*fw~*Jq&e{jz(bU9@#lhIWGsBir+5$+<f-i+^38n=TVu
ztWFl8bF*Wo1=N532DdSDD9Yi^(`i!qP>as{(6U<&@Zzd*<v-^bfFTkjHTC|YSo(~k
zA#oQjv<%AFboJFixB-qgBcsiwm0jUrJAj3%wjbV|#rAUVb~flgB;(=!E-|Z=fEiu<
zd+xcno`M^c(c<Iy7op!TG+dip9a?pJla9rNsdt;TuD-3?<a|$Bp(VLV2SCLPG7#;F
zi0VU3#+B9~2mO1|_KSo}qULdsZX>Ikw!fl^Lntn7MuxED-<8o<8eH5!X{VZwY9YNF
zgOsUOqoe!JAp`EyM+0~TVVp46_V&bG8TK)5P{%vGgRybTbxdIAlJtzuHmI6#p)wvh
z>OlnY^Rs7KSvlMFn4x?qy=$ndsm`4ZXOmJiPu3duYk!MZ)5dDbytp<w*{0UGc1;da
zA!r-sHL&SoWZj8<VuHYN?-kDHjR7;8Kci>$=xc9bwJNw@Cr+MxT_9*u2LOm~<bTRr
zc`4v$7u}y>$ZL4AloRH7IX?h2O&}0hX~7_o$CUr^?>W(9j}Hwj^lko)JS3g6@19A%
zKU-6b&QDuv-saMj$ffPdF5D<3MRw!$-^b^M`PK0@7GET%rCmnM9P=<dZYx)=j33&}
zTmdt(5GZfpN+jEN>&@b?KT!Ngz5r_!f-nF&RX(xaRyemr?Nat#et02v%@B(5j9VFv
zW!Mm0jGyYwp3#P?EwZTY78HF23c$d~#*Q@TDQ3f-UeK`S3T~Td7yU5LVA5{JNUjKr
zbJv^#MlpEBOxUE^)b$ZE=`C9pUT<&Omr^HsSV`9Q<YJY<S3-9vb@`v%%BV3lBrRw=
zeWog&NO&xQ>W@E4|0mQOP^e+s0ioC<S1vfqcV6D;n)-<YXnA=EUBp?-4=*Fk!bit?
zz+7o(ZT!n7unN|?Mf<hHhdfvyTVYg}-Y{G+otM!4P4N7wiEcgUYip1HprJtNaeuX`
zm5>0^*$rj9Hoe;7F+7|sCz7d)*>0Ns7=E6AQgUJ6+VGgb<LYg1(|=lM5aoIZM$SG5
z8bxQ0rp-ZG<Yy;*w*sZI#TThnF9n-{q2agWDx%%vH6cnV(PnZ9eY?%VNx4(B&CC)e
zYpBK6VdfwktGj!-y?pl(ZPy@bG=%g7AbMTt!5~T%$SpmN9f0>yh_>N-*a`rIE?#L~
zVbqO7KEQ#@9vZshg{n=oVf}x#0EBqGezm!+W8PnE-yDG=3PJN57AE+8GmKk$hbAor
zx6R&r#N2>@r|j)K@rqV(yN3u!B_e|$+OjOyd$_|OIH`$1@?5N<DnxQ5R1@v<!{L5<
zb(Rio*CY=)%3QYa^5*@WUT@E+qlc*~%S&DakVlFCgD8L8L|_H9Ka?WI3MW%7vI=I!
zP7@nFGw9;H7AsjsDJTL1m}E-rGjr8*IG511u?;WZPy#COOv<))PhuJ6{ZjdI$<~lR
ztDS;U^vVFJj|KhNB+1)&>Sol56XW$i-3t=^{=<g^9gJVXUgU?qJ9n1;?S5R>qM@eo
zcHjtSJo;0j?e4Ikiz9vrLmngXvxox&)@VdN83jSf`AREv>FgesQ8?jR&Mv>J;CAXD
znoU;o8Sig**Qlh<=BuggclI07UCkcoS~^NPQ1~Qy?M$&03BrK6c4YiezfS1vd|sC5
zsWqgR<1UAm^$!kKr`uD7s*xo`1QvbbJ+5FMkG@VcpE({TM$6aB`$#M{OD9b@W3L+Z
zo#WK!fJnvR^x5#KS;DRqdm8YD>OD6CrH=fxwofZS4Fy`6aqf4ED#Xq%QwMpO!zNE|
zp#VpI>xeztfZL&F65Z9ZPakeTRY3;`mYgN&_eYO9Q#c|Xy#sp6Bey2C^s7IXB)G>7
z)p3Apc55mpvkCVa{oM2=G?@229xq!yg$=!v^0}j|03Y5~&W}7Dpu2^Jc-S?SdKanN
zvp?-!G8&}y29mq!j`6RQIPhth35Ts+I{>FR1TVY>Z;0d@(ydX{)q2`1Qx526y3BE3
zqq!MG5*N5S94S0(t4J3<g-Gr6rCC;aOP-8L-peq1{@cE#=?{;Nwqa~VTu>w;rVVR2
zFJJ<@0d7!?Sh%>9ck_sTx@fAl#h%w0Z@&I$@!(JdCHpGG81k=crC)hd#ruN>v<nN(
z7nE0nM?sUYoPMd+(r?VlU;%k!Ox}+)TCx+Kqqy>dg~>-qImtYhUug2=fQRk$9`9S6
zm$IIg=S;#I4P*s5?=2H*Bn{r#jyq2{KIU8jY=)7DXEkF+cAmz3<NfOWJ&Y*i5u6?X
zF(7I>G)BZJk;9^<@M$pY>>G2;;kf}VoE%@i*L6@sjSt&+FM#sqQ=_7$O{Pz6q5Yz`
z(%5XvhY$Ry=ckn@{?K1#Pol%#ddXK`EeIHUP5X-T)l<C+q9h6uVq~$)OtO8XGHPTy
zy?`<m*p-&A-ME3jmT&WWX*=y)`ZVyv36tlad#)3cK8@av_ek}*eOI>J*{G|(UwYr_
zMDB^$B;V<iCT)E5u3xD$MiGt`GkB*j^#QdowXF2WI9_piC0pubO5{}#M;mnasIV>;
zO{u4JY>-bW>;9s?H(*?#JcXRzZcMzXOL_-UE{J+(@&`!8VO*fMUtUY>9ph>pH6_w$
z`-)516wC|XR@2Q>B&T_1g|ErIu=c9<P#>eI(_TziUT|bg;5-(B)JvgTd%3`i%MW+6
z870vH`KRrg$|p-(&oPg|Aqv8#GfBOn5*1@(c<8|ybbP4p%3CQlZgU&8ktlL`p_neO
zzM948CJY8Hsc<oW01hMP`0BJd`{aqJs6t^ISvE}d9G2HeX~WiR?~Fpp;>B@A;aHzp
zywvIP9$kxf1yRKN7FQxreSokE&;BsJE}-E~mx>R^p*L`ztM_y3PW|1wxo>Fj-{RPU
zIdsjOu?mIZgYU`B*B|tMpkDrzA*?*P`bV;WQW#r!fmnGr2d@bTXQ8*E&r`X|Lk9=W
z1A=ey>B8@)q5Bse(g`4b7Abi^8mG$o!@)Y1@R$i?Sp^SR(+7_DjjXYQJ|v+o96>m<
z3FuMWOG{cb-ky&A`_EYDy*TcrnOoVxp5dpXqUMo2zvRXFdLnc9P;#PWE82=*_usTY
zrf}wJ$6H-35zPS8bWuQfxzNAfCq7g$C%=6e*A4dOXE%?5R;fQKc*T2X+79I|gz{%f
z!4)Tbr?C3P)y-NY?q1XJeAg~+#ja&}gC*|ViT~ze*;{dFc*G&3|J*(;;0l>5Zt-;@
zJjZsm(47q0YM1o=eU|}0KJ)@^UO#9ZLl;kJK=`<`x4#|T_|Y)!#*K&P=8qe`T`iTS
z<sJHDvyOiOU?l=B$?cXpl|cBeh}ai=s<-h(p5F(3-@A6HUCN)9>F6URu}XcpR^P4m
zdIc`<QN114tlsn{)C;HNgiX`EK<}5hbO^#4k^jm3NXqI%nT9rjW?+#52os!gtB6&d
zpjNk3Rzszs`$rSYP1TUtZ=fMf-^-Qap=P?L4G9q}Z_BeT`flx4=V+uSM4rf7I+)v2
z%emg;d#X)qBM`bbtpr<j%|yL+Wg?SK*koHo|4aSrO<0gVYD6LEJ5$nBn`so5kt0XW
z!V_84@67$8%-uVG4LO<Huig2Aeh+u)b~Z>Gz|LdL6{#4@<`4M&Z7yH~Y9%gm;(P}K
z>+r@q@R<t+LLPY!a_3tv{Wjj~HQ{!B3*GGk=OtCW(H!tCulDIFU5xkpfqrt4umm#V
z!3o_0%MNe246$!)A09F1PF_V7vmG1}qxpdF#yitC7-aB`mKbaD57FSr9xu>>cTb}<
zLE1XYD*KI3zKhOb;O|d6KV&8Nb4>;zghPcUG3~N|j<Ch9Ax@90KTOu2Z2r3>a?+_r
zUfGw%zc1>l;dkuOuaa$;=bD^v4t0b;_WA^E-PhBio;vjI!<@ND_cfxPm2Zd7Z;hCl
z$i&8Uf``tZ-z&mGh`fo^x+jaS*D&${TXS}`bpgmkJypoIW^<5ohA>2sM{H-pCH^~W
zu@Qsi)jd2uFJNE2&VY#!eL3gd6{8RK5L`xbO76aMedA;rk12HhIqNE)n#@lw?<WOG
z#b+*^*`3B;ZQTEJ6H*eUc%GIxC~od$$-h?aYu`-y?hb<tkpN40KUmeg7$!w1WQnT&
zR5q4N?bv?DIWRjq&ixvTEP%(>SU^4c(~B3UeR^Z8!5BgD;~xHFV5XfHjoqVneB>&L
zd)U|jID`}rfQPFY0E}})n@jefG9``6CGX_?p1~qWC%lCB)|Eyg{K;C8t$6(^o=IV_
za5qFD2cmr9>gr5pxVt-fdHb6yE-RMvnWK~xe<%$^_0F_c_Qv}I-Z+42-O7LV>{-k7
zZ2nTl=>heg8;T%F7ESUl0f_O@F?T`;@^z_YKwR)mLVq)Oq7jjNCyq<{*+&l_s&T`j
z>zGnMX<B`&RL;*Z#po00qiP*GYH~<7qVxL(7ZJjI3yIZX{FM1+tMxI^1N)xN6?aTF
z_cdCd7fL~HVQrgAFPOknmP{SWheZzbHsPN~m^|rML<0LLVnkzUlzkevHQbn^r@Uh0
zOi&`^V%!ILw}<s*rZshKl=*HsS2RB|PZL53UZ2p*<XmP__zB}CO@QYJoKsy*ydgK_
zhSYToIQI_@24u=(#=v=5!-ftAVk6{E8_<ABFLS2X9k9`O5GT#vueR^l@rIzUxl|5G
z2cplivav~@yR-idlqIRWEZ2$*!KwkW@?h7Vb>3Jf8xY7LySwrhuYpmw{A8|stJrl@
zy1n$^(x}pumRA*ppCd;`sm*0d9UvZ;t}Gh(!`*07vuMlX-=B&kPm36z%}_}CbEij5
zkvhU0_x43>q|TSHD3c9?+W9WwW9>VOuzExee_z*k8~8`!on4S5pP|R+w=>&(wgHJu
zQCGd#d(F3L{z&NeG2DuTg~tSVA@?4G@b3@he2Z<vBevYUv<~k7)2zxW&(^0jok!z8
z;WvzWo%zhx-sTN#%qx^2M=xBMa#9U#Ar=`?VG;va{vBvckbU~hZX?TxTprb8@rhBf
z?I&UqcD>g>jdeeVfBv?&nX<%G*Mi7U$Y=V&tp~egxcH}?nG(slxiw*}O(4kC*0Pt6
zi&d0r!&C4HGmF9i-EcVf%E^?jdVI(rLN$<5vTNm4^6%p0eXMnKG5r0m$7N)vDyP?q
z4wav?YMZ0Rs687~XDQtMrtD*0HWFJP%!y7+{WHH4a_R9F2Rpkie80Nu+brMu$fqZj
zg?TA1Bcyx7sI2=AaWD5*hA}kwopL5za_^{z7luzNyRgG6(I7)W0`?1?b+2#`{af+e
zP9wH%*>VVkT}A(=bm_c6Z=av*h*WMC9uc%6MpZR)FktU?wH~$3J@*{!ytTev%^~VL
zQ;{6Zmwz$I<#S7Sgm>VEk4(>G#&@Pm!O%ti!ga3p2lgc&xZ05|-KPCV!b1h2gt@~(
z%4zcYQoT)Y@$NN;8e03ux&n43yz<f1a2vIGWBT}3g}Y2^6WNxEu|@=l8;ZPk+`Vsa
z88{EyLJIaT*+DS&_(at?LxV=K7|-bbiPG5y7ESTvpG<mET#0)H#|AY&{f(aXkb22b
zj(XD&rtTh(90WX)dIy)2I;#o7d3wAU1BvxQ`pNUla!zZ_cy;8#YWK@E`8V{hr~8@A
zN>PBgcS?O7ydb9Ka|@UG%VPADyZupGz8s&&%V(;!L#dYIiJoy6FMdx<V;+JApw-fx
zRCw(8@o?8z2UC$K&%6XgAi6u47xH~Zrk9|^;RgXjr7PELK4YVSmfo0s0;q!9>u{WS
zafatdPSQ6_`X<(ep(|HjaJl-+lP_j}Z+d&PRVBA;ExQ!Az%+MM$`0Ens!fN%mjg*z
zPMHu87+C8*6%f0`!)M7}0|5ypB)Mw%DIu*VA^-?rkkf35B;b0hDDWRA8AFu#8o9Hm
z2t_21K8}e9Xn?UzdQp6ay%s*y#kTeEUC?@I?~%g;31rY0i;BuTd;F6#cyU7_HtB~B
zO`m-;L^?E19n9rOekrC^3FmIz+H%sQy*ghoD56Y*V_Dt@OcSiI^PH#}Zb*So2GL#&
z-+>>AI|87QkBkcmtcFDygF*j6hn5#V^@HHvjn><hj@;x#!-wydXSo+0kSrMtZ}v8N
zfa+orLKHQ?V2kgd`@y~&=>p!~L>t41jIfrmWg|d`gzRHUKc`p4jRFLx)FMGfnIlKc
z?XOx6euRm}1$;>Oy>n}Sc`)HWP4R{sbYbLX`~jDRJ!z_3M`Nm~2=u90h?y{}&E`aH
z?bNEFQWf|NOnX}}S^*_`%)Qawtt_wQ(qjfql=~=^e2k`h5`!brKoZx~XmfsI^~@3E
zxi&g5_PAR4{CV?wdsv=W_~L|B#EU+OzOwKYT>;GxVF8)N>SiwX?Prc&f3Q6zW9s<@
zs?sw_`tI>@;WtXZ=2ui9T;aFc1*Rw)((C)s4vXhUG*Jv4qNze}Su!bo4zYlM19_~a
zI(+b;LQuw&gXO07{fWtK5uHWz29_+-A|@x9U3o0b;H{Y8r=7OiW{|=4z=OO6;vq{?
zgns819JL}kv-tWeCqN)h*dg92*b!$yAPBUZ>FKYyUxg?-v@ADCWk)fnt2o?bKLXt$
zD4wjNy$uI#aCcR$!M%L1$;YS6$A{%*!j|dDJ-Jtnr_N~h-uixXoG*Z~M>R|)1@8|D
z(f2s4e0M(`jL2e92t7xVAqm73x)cO0SWkcaB9mwCx#xR_{6!G_$&;_hYcJqGNBGr{
z3ap2YsgGa}t>m-V#X%F+msC_C8<hYS01BBL@VPS0M8Az|*Th`8<ksQTS56W?Iqde5
zMkQ~%8~Tc#l!%Lsm<pYt(M+Bdxg<!thvMfa+pu+d_pWSpIq2^N3t&a>GQC{o$srU2
zBnF*zD>|@BZGh)DcRy_d4V~2|N3TA=;9F<(3^h@aNc><@k+PedhC8HD00r8K_j7nP
zW-#aU4nf^%JlPHZGG*n#h^a=XFC>eMPs<EMl!+d2Qz>)TuD>-~v>-+3D56YU0KH~k
zslyN{U}=Q8+4;xEkM?!WJWd-7VgEM^_!u3A^4y!Qy7>cUSp84B1YIWwQN?d~#2Edx
z8eb*50C-{VDK<u%;=;8eS>s+`!`Tm-JkQ9|d45=XPsP&<L#7&?Z*$=KVQt-lmPm91
z>uV4GYMbzOoT6RSl$R$oOjwn|*>iniu)==hu%Rz)FNp|SgrA%bpyBxurHg35f$gln
z4|qHR`2~{;qSrKL9mdmk;@+aou>A5Dk|uav9{>j9PP)xns_gE+%&#CTw%2r{^P20e
z1Lxh?(Rws`4F$ke0YhIpUb0lJ&x(}0WY4)1jz}{hd!k`0YsuEEzpv*s1>kng=o*It
zOBN2s(`m8kb5z~0NXK>`IB*A>Pt~^E(aMRI6@c(kS@mD<O`SFC<}PhO)$+t<eYMMm
zf7uk%UOCa}Sf8(e6m!9>WkRbuJtUwLs}rO+%(<rgr8IXZHX0+`*K_cNN!bASV$^$T
zsA$Lg!h?Pux2!GF+B}r5$02=^b8|ZlXM`rjjfHRsyHO<GLYslm`1kX#3mdGD8$c};
zL{YNzfZm4fQ(hBDode(JanauhAkRM0Ous-MLKG^<yBlV<EclLt*bVfo{f#y?acke#
zV#lQQCJ*IPR8EY-%;PrChM6LSuZQrdxoR96kZsJ4^Fxpx!DE#6m<ebI`XXiwRqMh|
zb#lt;C03fP`EIl<@skoMW0NE9U0Gt>3rPsjRY<?KA!GEo?5@I1lKM55AB{m~1E7UT
z@HQ%852SE%7t(@mN*P^nSMEXFO_~y{NQOWi@fC=`f+PltXV#Ybn?l0FqYX@c(Xjxw
zjGH>gaUl=)9-Z;if+VbmV$C(mF#WYed3_ut;W`u{B4PpC2bHJ86}2&@FPEHr<@F{p
z-S)N~hKVk0hSW(6=MiWbfFQdB6d$U$gR@BEY`fygWC^T7;$akQz%`e=_LmyK4rQ}_
z@7}R@`VQHw?}kH65K&U*Yc@bNZ!Ws98r`q;mVwLr*Q+Y#F^EHYX<YUVmNkbg9roZU
zI*{btdm+6S3)ew)SGqe%hg*{FR8*nwmOqS|ZbF;IW6Q@;4A68m)s%@E-~6<Nd9!;4
zyRi`c?&ty@mwm_d-4a3k$2|2QOwn%h)kBjCd^Tob#&syP9v*Ra{m~&Odb0bcoyCL*
zy~Ru2OB}oLgw`l0YBT1#Gbf3BQu7X~vk3LRD-Gb|X0cSj^6oi0AEI2eI1%8PvaFW{
z3<C0LT%NK=*QcR6@MW`j+TZqfTiM;7-o9_m$LB=7lPtU1yZrql-0m0$bgC^q$Lfy>
zb?J}bs9*_7HaIGR5%A=hHR{qOi25D$iipXDJ;FO`ml?l#xhz@ME>PKVSkUt?fA??`
zQc-{s*Z+Ke7I`Wp{H4)K!l8%uS+SM30x_G$q$o7#6wM4t1z;Bq)xHpPNab2ixZw)C
z$Oq|WZe7!VY>FIlhSR3)1{q+H6X-{P$T(Rr9ejqT-QL#X>Y~{e&JdL$UwAt`6t5$y
zJ^D;+KCemTTzc;JQ?F6+b#j`-7>IG-&b_b&G8x8iCO7^*@<nKR8o<5xB)@sFZqMf>
zD<vrgbj9h|q5{xd7H8&%mUgX^b93yT_Ofbh@ym92D<k?VC!$~EKw_ZH9&|tI%$WhD
z^eG#33VharbKrULy6aiv%UrsQHwpw+nm8_p0L@Zu8ZE4c_9R2eD=eq3&njF{l-6LB
z{E$Bpgp`P>h~B=FS(}5Q)sQw6sO_32I|gr1T5)Qv=KP`m`BcS0N>5Nf)MkKI$b!sq
zdpD#WbUAc}%vV&5>qJN?)}IDLAGdBMpWZ#M-@RzG!gR*acz%+EZI(_Sv-$VnS#j`+
zvUr>P*<ib4R`;7UszCXZLR5UfndodO_VHvcL)OcF3zh;Xh|BTLD5f%+4=X)u6rwfh
z@azW3Els00T6u}y6Hzfq65w!V#dU5MRan{CjYU&2C$-5|_3?>5l^*@+T)B|C!$W>d
zthn29p?*tolBMOpl2`$J1419wiX(@ToG~24fO^QaYbvS6swuca*}BN9(J)kXP@B~s
zdVo0J{T?<1R8iNgW>eaQ8r#;333)1YYW*P+u_jyVyU@+wTix)oUGAWi{xh||QlQ9u
zFPWUDqDo82;gdti-(2?hIKD5JPxRW#E_BQiKfvtyK#c$(_|=|{S*9LeKUz|4_<zX3
zA8~e8WOcSl@EF0`5sp1LH~-$)#nW}(MD0Fs3BRlgL-;5ctAW~2iz@*!)+tS%cPPD&
zuyxPQFPx}$_~Sn6fHp&cG}}1Ha7kdEg#m$Ov&4xL(mRlg;>VEZR*H6}u4945yObtZ
z4}j#b36w~UMX~zTXLdsOb3@>@aY3{tJ}iFJnkDUxv=-zwcsY%pg;!k|(kD+kXYROP
ziJP`?nrci(l$;1<4+z!63l}@ppk!E{iZdb5xRag5wGK!Ebj|0?x8Nq^5<(6qP&CNm
z(lRgP*)}xEoY<7$dU<WDzl|vT`FkR!IYzr9vYc?K=Ejtl9_M{hG-F2Vcj`3$0`*~T
z?6iA7UYvNa;LrMq$o0GS@7wq6@Em{1DSnxwSaXb_%tVX<oS||0AaWwH;mU0Sio0w7
zemae@{@#-jF~j}L<$&?T=y1ia#phyuW_l(1|H%ezhnRHg+h9(P#&g-)_i`P}0*5t^
z4DYn6_^D;W+(IYQy#vj^jhh#kI?MOwz*`N5DJQD%F|3j-XEb&l)Ec4@QX9CoAGqk?
zb7@T}^)oy5jY-kG;&ASh?6?A~EHtBeyyHOX#_9Ar<Bv`tfaW$=j4V144KPk_c7-qs
zAcO)d{e#MegoHSZJ3^WfFfI{9^oYD-Xo8L7?H=zOadg5Tf03j({?U0_&sfodIb`8B
zo1BdcmeuMOsyaQ_qqu91SM?hHE$Hv9QazwUym=Hm`O65n=WWQ1$aX9u3w^?=NKT3H
z1tGRm4FBH>mmgF=vCYoS>=vr2aJT#CYLP?=1C5dK2+#(Di7bNz@Z2^tVn2Nf&gxE7
zI-5X=_KHsWt`A^5amu!o;C@B~q5`+>$S}OL9UxV-fkL%S4cH_^g{0V^58@5{R1%k4
zPRS$ma=qTYO`;TFq0DwIvzR+SmYi4mua67RclZAN4MN}5YO_esm<2cks(-rXjq%@#
ztEE$Py#u!NI6CE5mLF1H|8*c2g(HYAKWOvhQ7CEQoo)ep?C&4;=K!hKu`~4_gaKn4
zcz4|T#$m%FV0GjjXXf0n@obXmi3umu+Es4(R|^1jC{tOabvK+%_k`NO6-V^~egPpx
zsfoZCzk=Qr$2d*2*^~sa%Mr{iQZn&OT(o3{#L!GSr6^i3!l2qPbyW=~(}yzQ7^~N`
z#EScqTV0j_g2=&u{n@n(0M<~$NSU}P!A<eHXXd#Ybn7;swHCJ$5*e!c{r;Bkw|4XY
zDtvdX^8)6nbDc{+G%kEfooq(Rir)ty!chG`nKgRcxDm%6#Xf#zq*>$VkDz@dMS7qF
z+pz?F=?1f);SrdA0>rq-J|g}#ljUsqJ}&oDscN7e1gPWVRii5bF_2X&q#yoG%*0Xm
zWoyV1;98Rd*-*>$Z%fA;|9qT_=eKlw>9^qvW8Pl#=MLvOTh2fzdPw&c^6ppo#+5`4
zG4XdPNd+%VrwyL3U0~h(;t9eS8F_+x%pEQ`{S_2(yd-;gPBAWpqrk#MD62P6#Iz`X
zWt9&E9*%!_ALN-5v)-kP{BDv~QWkzo_T)TMObJe(dvldZ6J;Zth4-5SQ9!q&m)BuT
z)aX4ADG*U$!Kh0*)w0*^iB*MN#=Tizp_ugh`(h|S5Fy^+vt+ypjzp36%PlR!>N?3>
zQe8w7Em{0jO*mRfBN0u{RyV^Y=K4CzS}Y$9NJXtzxt8^2QsMjM7nO7Xy1oV(^+1f`
zCujgUgagH)gx^aW%nE`WRtK&?7Bq=Ml@sAiAjIFu!uUnSagW*Kl1oKSab=8%aXvg7
z4COp0U4+7n5E_?riYj=&M#BlR16IZ%B+85s4;TLOQ1sqme(7Z13_ucVTW3WDGH%2&
zq5uHNLM@C-kG|0O9(?ppK=FV}-zf%#*b<b-<+W~NmAYC93DMQIzf;JN{QG4IO#qV_
z-J&?}S{WS^p0Dqr14q5T9_JSRXAIQr7yalRf`1)mL{E0l2$xS|kL77vVpH<Hczb9q
zK9}Z(Gm$mp8soa*l3P&5M<*jpFy`g`?mn?a^>dogm60(bgCU;FXg6;Ve_?^S5#ZJG
zQG+ksb)+p_7fsLb1XW|EK-n{_zZQXrNlauromg)&K$mV19t94R-F0td%?xsM@_&b$
zEV(CNi3BvOjam4OD2IKv-)t|`$FdP;@_u!OvM9taA@e5SQDN!ShzCUEj6nhvFD!6B
z_nu6jiHZG0M8YcS0EfkuLclTH%m)kx+lKlOn~oeg|CQJSZK}gB_u9GbTvD^@)s8+t
z^qMF)>ApT|OFiwBU9)N(F%FqBBX)A}nG2Ogj<FZBSI$^9{aodVmyV51vR^EUoxU?F
z<jsAJE*4{IHC$Co_rai>gV*hH+BnzLVc+`TqOWzM-Hnxdw7Od9nH$|IIw;uPtx>Rh
zQh4c)D$jv!Z5oXm{`7S${mQK!Gv>P}mF`tdWZzkN9cM%>l4i2Dbi}~#s$^b%*q0o~
zYbUY6OM~i;Q5sU#2+PdXfxjLNl!hT8VHycT$kpdQ4Nlb9m^b;$AhU^z(fhxD_mGPK
zU~$A{`S-?4o()1ENNA`4-DKnN^kwnEfddB`Hf)&ZWQy5+?${%!i2<R18?eY$c6Onf
z!4)Gh5#X@w+_)uQ?{S+AukOu43^()p!yx5GjlMjq-h6gnBkxu|ZI>YS@{<JwRIU2Q
zA8lc;baZAj7N#*eBj$g-c;=56z82s4DpjP<8IDskY>_&oOuz5QDW~#UI2T`QSEtmE
zMyGY_*5r@NPRyZ0YfIW5`?ipo-@ku<UPjy0va7anDJUxfs$Wj*uv4$Tw%ew>0~X$y
zchJL00);5%*oqUg{mG(O7Rbp4LRbR9z0Ayli9ZL0+5CLTu?rVgKK_j+UX=RwJ<o!+
za=jq323&{q9iDB%vw*mi4iAs%m__P=-G<zr9?<$y{nWmLZ9V3$XdK^NYoDa2^!s$J
zS$m~}@Pxkj-0u1g$!)f!N#fO?@2!(xbKe#a$&pePREAAZJVlw8$7#C-^dEnBd1~>o
zq@*xGG7u}cPXIKd4sy*wPzeNz!E_ec7VQHo`W_fd2Sd;~XWHgHYh#JK?mol=&FP;2
z?0k%Xlt;f{8zb%x2qEI-LeJMLTCSSt_>YVbYmom2FBYhh%82nSkM++5Ae?Qq-DS51
zLrfvLZN)Q-*ksniyVbn9>7!aCwtM|_%OK^P#Z2WcECgboto`nn_wRRi@M+V1RGE^Z
zVtP}8-NxR%wIcf2J`3g4E_F)BG*5JtczK>N735%4lx54Qjt|(Ie3yxz*1OrCi=5nP
zQ3%~};C~D+d6bs68O0zxE#hTa?!v-aboFF{sl{H#8^X1S<rgyj_@tG|`{MNbu597G
zYAY0Af#FBupxb*dOnuV6PAO$vqHEsX*G_o1iySA}%K(C%d3k2d^VT0$c9Edyb^`~7
z(<Q;6_4`XpOKWe(TOThzb$xSNo0R`uyPA;S_o=*`w>IA;K}|#BFai0*Z(h7~$;|6_
z;!oFvqlc+5!bSlmvQ2j2yTSR(n<a`sq_dbOMI>0ec5Nh4=p9)uBSAjNKQ1Nb)3&z#
z)K`8a5ZoO<C7wH!3{vIP$h>wo>E2&Q`Nmh?Q1$*=&WDOJJ&icN!ti6qj#))czcD0g
z3fUgly%ihyZr{E)ZDEHew_12xE9V3oxMg!rWaTTY40(IU$#=x&u)v7$!KSLJ66Fu0
z43~h_7CR)3vDmj3E_}Do^&-(+3A%*(E0cA9=<mPBJKFws7kegRLM*0V6#cjS^Q5i2
zM>O$nC94z?x?g_u&OGA5-|XB&B(XvBMRb$ITWvZ8=(%O{8}nmJ0xm-d$y)CAhd&e-
zX8=%$PY$j7#6{0Dmp4yT;M=`oHE8L{=;*XXu_y*Vefe^OtW;p`Fwb%e>QuXB-7K04
zd@kT%ch77I>>WkzYF8(^=>~gkH^ne><(%0{N=gMwY~yY(b#(lF<zanChJNsd@^R>0
zY4TVqEgPt3&TPx4(afpE#0F9&CVS_ta3#+l`KkqQ3`N2=(M=VlIlY&W(cZE$C$>ef
z@UAu3u*=C_o39Q=03?!uzAKm9UOkqp38;~N0K#@C%6j>K;4LN{_8E0lhuXqR3MJ8d
zVdDYU6DL1E8`&h$TYb|*#b}IJKpldzv2hueoHvLEQ2pitVC|tb4N@aNte<ewyH&e(
z?Gy;*uF;%-LMTmY^yoEfeyyn6X#M*2urm^+PBTmD<bjwB^W%|uic-KuUqOkyz9_b*
zk<mu7*-+{7@?i412Lq|Stdv+<<|+vPIzAoKMdb;7PtEN<vZc3mp2glyZa;qfAldRD
zrHD&|LryMj!l{ai%1RXLf!Hc+-pb*<efCl7xCWf@=%)Eoj@Gj?A%|I1Eal}cWFD}{
zkJ>~vB>|M03!@_+J3}Hr_pRiY+1ZxJk9l4Z8d_JK3_|g`fC;C1D@yh;pdpSz3VOr^
z69}pF&0FmJI@Vl_oAO%R?zGp?tpXC|UiHs$$a#?CQYS&S07x#)cGs;F9aDFdw%g{#
zB}mP8xLwI*l^<Jyxj~jMUw+fy0;LB7N8}S0y(=p#^N0u88NSKS?`0F@dS}f1)COeD
z-^fh*(wp4mop&4vZ@Ib0gi3&lE{KF`+ngN7I8jJYAk_Kf_^(&<2mVkexQa*EP~JuR
zV-0es5N7S;c!0cl4|8(XPEOWjkqmDlaB<9ImHPEBF0#<6!(rhcjVrsz)6NscCMSg@
ze5CFYWTCF@fnbIDSMcTy&A(1=x4(#w1Q|LKAFsCUde24GkbC65v1y(xnG1<)tgIgF
zbDrboW~%33&p+Krd+z0bE29>A?s%<1T?v)Hof*A(^w$pG=Ik|aW7!-pmF!qObm-7u
zYG+ixZ)p2-9g9s~`+V*8&tpt$nK)|J#Rhik*m(3Yp{NbzSxrrC)~s0ytd*oqB=cW{
z8KAQ1FA1+d2VPC22Rdw4OI6U7j^BNFQ2OBVF|87H91qw!?8z(n__2+)3)?W>efW@g
zDp#d(<BoD;+$N*zYHn_|&)pHqb~PvSOx~tf&o1)_ad&Tk$?MU*`=HgUFL}<qIO#Jt
z&Dq&km4r#oTFoU5cQ3mgiFNC-cwHLYETkayfmlP1qi~_TCpSLCzA}78#PjR3O&iV0
zIrZV$WaS)Z*^@f+=jg_3-W6wV7-^p5!E+(B?ODvcZMNO**fGEYt6{^g#~TGU@wR?m
z`>1?kW4~IwCWtH^SslN`HE6(d&&w(+hq2?ndGl*PID>}_x#C+Q1uG+SNzdSKFOz@0
zv{>e0@}QI1$gp?b*4Lu^GpKx++MYb=J){8H#+OsC=U_=7>e;}Bok#+FFEHAXk+A0X
z)FEZ{<&4EBMiYwj`Sbn9k9C6_QBm_Y^Q!thdZ){}3<?z=1_vFo5CVH7k-xsbY@(*{
zL^|=toH^t8_!DOlR1eML0pYxPnlR1`M{L4c`S=SJ2pcTV&Fo_bEeOfyXN$+QOnlL0
znNE}2T4I2C_~60%Y)3E=5>ITB*Iqt+Q3<bjY9U0-8<fiZB~p<L-4v(1;{5!fg!IZE
zpzJc18vF$UNmurD@pFS>%We+zJ3WGfv=Q$#^VYSHqgm;aUD)!>u}OYxmV}jqV&8cp
zxUY@nW_Udp#*tTy7;UvZu3otUSVqu&1$Kb+!zO=jZ%Ewd?84<2ggPT)){<_NW}d|m
zTyVrP1h#!_c3>=mwrc=cd|31TxT3jSQurtfnl}2lr}4^52Y6MO*7OoLZWz>kMsQif
z9137ny(;>(G|yii@ylm=cN>0EJrRwHtS!Tr=FRpf$^ge7V0Kp4fPp$2te=k?GE6D@
zX5N9$HiPKH1tDMqOD|JXU+PI(+K7)AescL|1Z#q)QY?S}NUqil%D8_2eiMnCM{Vx!
z*`nFzJ?Gjvcqm2tPVX3Wa&IakV=gC`DQBj?wX=M~?iB4leWnE_P)<-&ahazTPCHX~
zcWRfOZZ+R}#r(3hVWWq2TX)C7zxXptpKhsN?rwWao5j%>$o!@)MOp`AO25S$aw9Pl
zSDhx*Cm1iW^nZpgPlonx)33}XatM0u{l&JvTihdK=4J+mrgjNA-MPup`Z)qx)8BJ_
ztfvi#n&Jv{dk)&wxvx)68$h-8UZrQ*l+|}@{z3VP>5nOx9Qj06LNGMb!nF@t_AR`+
zs_kzk?p|D2ET0fh5S9f0HR{d0_iQK#Hgr?;&#<p&$6D{ji!~=ZyjpZ)74Cjsa^+yn
z)~&nl+&xiIYDP^Dl)+6D-s<v(4IAWyJb%tstvuxii@U2$#W2P-PMXN#$4NxvpdL9)
z(GZpR<DtC+MqNKVY+Lu>F)*4<H(&j<se=}#7q$p83XrvzxLE&HyZ?3tb#hy3tSNE?
zykkbadxI|_gGhu2mu(MOJKFzKK4lK^H@Dd1h8oYEt)*xdz~yP(DNW@MUjq+r0t$tl
zBJ&2}LYs5X8#WvlCRwlt%F4<tlA7L&C}?M+Y`4B+Fx|98Q_{eKLj#Hf6Rex1At1pm
z)P7qQH?ri2EHG-*xAfDHKjpIp4I~3l(Xi-@!ow_GpwD0p(Qscoeosh|s)s>I*JvMa
zG>>N9`^R$W>RNH)Yj(LYOf@0d86C7(ShcT+R(-64gPyd2sxvqKR|`1)aNaMv>u;nl
zaK0RzoSIO;xZwb}@-;U8>0#t&ng2r9dBFl-AnCLZIhUvZ9{cs{S47$jHY8I@U>4Yo
z${^NM)nt{gS<_#E_em`2-@bht^P=~dI-?rx;iGli!vyv5%_VC@WLU2{{`s{JA3nC}
zcI%o}uI9s!Y?0_Vpqs~E1W3Pq|DO2HrY&3QZL?GR_(ZU350CHrIxPIN)f-XR_z4qw
z9$V0$K?62PN!l%~uPoc8E91Jb>Y&BBX5x}9VYK%%E;QWn?|oWJu7{tBYrZ_6-~UVB
zT7~uSh^KRnTF?S=s$5)SiE}Nx*?#Bt;<GgacxSoldpYlU_VQ)M#Bu$H(fS~LBq)t`
zba;d+p9#CL<sD&5GuHDBtXK(if1^&E>tC`0XW+Ng#>4CU?4{D7W(wmbRnr?%mc*cm
z2XX1O)%dak6C4kT?qOQgmuFRDjQ+Vanj~iGyJc_h;L*BKT{4aNhJyyB_{QJM%nYIo
z?DqE`>f}_m?}rv+&8broJ-=c_<-W>QY^{@nK%}(ctMQ=b?0Y=1G}|TxHEY)LN0#k-
zZuz&JZMF`xDIa~_sr^%I@0BX=?dUdo^ysR}XfMC#OKvQRJ<O^hHsgTV=7&wz`>#WB
z(*GLE4#N=9OKR8ES2(X(xlBYg-a(1Mw)6ZnOz=Bi3%bgmWLE@s90S@~xSt5~-cLsC
zAInwx4@8jnySlI*w;xIR*dPUq2SSuI8}1Cop1YlbSDPW8LtW=Q*SMicl54f!W*?vC
zRR@*_by6RF<j!)#Uq8B3_U+!rtD5j(8EZiF3NzuHxX-XeE$@@9{2-E*I+m7&)a~jk
z7ZZ4vaOi(lQ9nNM6*5iIjU@Si#fI$0kR_Xp*=fz6@-An&WONn#JzD@DQzO0)IqBsE
zk2C6M$nJ{3F0TwT>s<}E9+y?SgGI>R;iqT_n{<#;R}=9JcMh(bQszDYH1Fq!`2#h-
z=GT*UK>Cooj|ZLP1`7ZE`8czF5AGMc=%5bKd?Eyb|D@^AB&E6TL39kVR;B!#r^IUz
zY-;(YgBzXdro76>rJY)jTN{SR-=NjreV6NA4ZcNz-rC%=&w;TmJsWKvx;xhbV@pP1
z$XVx%R)?M}nwM%~YI-6*zQwNmFQ43YJ$?4<CT`6U78dEgnRS(v$|@?ZI-K*HmL7m9
zkKLziY5Q?yvK`WWk!+&l-|?;$2|H4WxtzHJ#Ets<*AIN1W&1j4u@CSHhBo|5)}Q;c
zhKh}F<U#qb(Qmc`9n`?;vBw{^|2oKG5242zUw_$cS=iMrdvGuxY(d=$534<S%p8Se
zCPZL}C10K63s=ZDWtW=C$Fyb)$Tgkif%IlakwR36*#y5<Q&6d~$0L%H+agc~ItAJH
zEb5!9-JyfxI@|%}tTDIx{i~Ee!v!?<>ATf6kM6RDd6q}X7`BAS2zR~SfdATD;#1u9
zfs@${{p>ul)HaguY}K=O_wKtSdI-`!BjZilm3Pa_Wt=158+BA#^G9^~LhpoxY<={O
z_I@922)+J@CC8^kvJoqd@`PD@`SQ=(XUtn^wpZ)Y-GZhnk1n*!IokHxe#eUoo8B|;
z9#BOwrZJL$b(0bbmb>{!QZ7=*6iFBr{XJ_IfFj^4iMe;5vz$C%JCAoAe!pt_dp%7f
z5E1MP0WXc$$xeVeKsh!^wboB);!GN06J>`!lI2Dt!(E5yp%J8h!B=VYiz}gaHWOuS
zrhb}hMmTh^YRMm$Hgw674=2{3KVT8eEpDB6ECAxO;kO$b*01w2SF+Rv8iGx{&fa(+
zssdWNLYgfp&J2MBoGU=r0pF6G9bx5{Rmc0?ts}}gP@aAsqt?Fj4~5mYz3?%BJMQx3
z19A|`7jwmjnOUWtm&FKB58z?W3ASKpzyY*ZZ|%cF_=U!d=IxMfi_~v`3Un(S<_}V7
z)aYSm<_NB!<f=Q%z<Oj?jF=I=d|62|q^_k^fO@FIy?zx9xi0)Yd)G>RqWzl0dmhjk
zv4JG9CVofL%M@_hM!E;d4+CFy@ykARNL!q=1_lyqLbfeZa_yjp@y*+|ZOagY`26tE
zqm+~}vu7Vl+`7`qX%~H=kQ@yg>O%HXt3Sm?uknM0fMzuH_<W*YU>v&!G0{rbbdJml
z<SFDh9MvM(<k2RJ&JUd4#a#H52M5Nuz2oI?hdv+gzRCBy<gVUuoFH8zcDTk$f+sif
zexODg!~(}shjY*8j_wXjMK`<E_C5JR1;<v9G`2hA7Z@QYacb1l^Z^0&(sFZW+P!CY
zN>qGP6__-Z*T@$7_ii)doz}O-&?MRRjK};Vv}!zFQUaKuc|bzTNG=PuC0X87Sy`z-
z5nyNx>}I>w<N!+m8O+Cik2g^#8kZkLfv4A|Yk#MNyD!6rGc~aWD9@~BGY@-4+536p
zlKIs_wJs5@eCGO+KBQE)u51PB{gmG)=Ffb1=9$)vntdf``7z3S-geUk!vL3Y_hR;e
z=XXXNX!TN~p%O029>HT!%z51iaN=t_IBfBuz;GAVY;2-;FbxfurT5_Z^K_(}w>h|M
z-%#>NzRBudopB((y5S@UG#TT?*>Gc2#+{mnb!FqmjZU9l4<k#1SQjx%gU2RSMlExu
z5HtQ=7ndhPwE`ns!KPMr`}QhvXk!6d(}ybC-P8V4=>eFnyfZD?q>KzO@BX2o?d?o7
z{7$nYkX*mPgHt=bJCRw;PcOT|2_NawX8x>M>-b!;X5iF}$U#twSHee15F&2L@doZD
zCH{7GTi_X3kTH6X_P(#{)V>C?jdzANv>N;vVPe7dXmTUA2XAeBHPx3z3XJ>Ts;b<7
z+RoScc(GSl$}bZRm%oOA&|(r2LLMNtUW^1%+h@$^+W1Wv3sRu2`nmY;dr!w|^B9_D
znDH)9-`NWmjNCo=a9muRXYR-1;)z-2HUuUzx_NmeAC3u%Tb6ql!s0C(gjga05X(~%
zsj-Sm50bTLnd8yR#Pfam7~YUzcBo&?1~Ha~{s?Maaa?wh;inunc26a@n+4$Djoyzc
zE5=O8>iEB&?E{`Iyh2}A_jd5$(ocsOpmf8CZ4b7&L^}d=WD{ebK6&1g3+y!Ppp|8z
z1kIIHy?XU(3Eu&yh9?$J&p%LuZ9q_H==OAboemx5P3zcMo@(@ngfPcz-o;lZrpELg
zzwwnupO07GHS@dUkT2Pue5A6JSuL=eEZwU9)VBK7gSmWl8s)<814muS$Y=oGqVWC4
z4~)G2p2M4s8$Z5@+|TP>1`SfcT%&gPp!J@DjnxgB?=`kyNjtQ|gc&m|S;b0)b6dUo
z(evk9Xbfn?nK#&^kbpkm86pjWc@j~rmG27=4o;8b#awv#vn7Cy`Pe<jCp_KG^r@un
z)_=3K8v;X=GKHB0HIv1@p6^D_)WHzO1&ds)$EXWzj^DF?|F(4di>Ia~kZsEi!vNvy
z(u4~-XBFC3QJ)__&Y&W(Ne2%OGVHOkGr|iFE)S4Sd*4du(|hJxYrAz+i1V#&p5LNz
zuSbfdGk?F|bfC|_T7dFa62c;@Q=TkBh?iefB;f!cpp|dW{uuzd;{%K>0P-b1DdOZw
zS+YowAngRw6lw~C;qCn<jUhf9FO?X7O#0y{sDRh6Ot17IV*kvdq4D?Zkhycu2K%G+
z?NsB&HXKss(da+(WqQj*M=((#4bkhjxBFUISqMim4{t2vG$PD_qi(qKUO)#RfjHA=
zEHWaQs8O9ww6QUJ|05bbJBe;6|6!5tzj(2gPI-y(&gaXtO-=U^I>1>xzoj0xJg5FI
zkqfS0{}(e0wcLMApI#wg{B-lO#tv)&d;>>5I06+d$Gxca&bBlEo&T&C`kTYnrlcrT
zf7@9-lDm{n`p@tQZ^UqQbB{IY31}Wj_qs<DCM0sZ(jr9Meb5yZ>)NGP%J6wo@Tz7T
znk0r%kp4V{Ri~8gOeo%<y{=;({3z_^6^yBHf{l#@p9Rfc`wLWnEYFHNOS0lWZMA%0
zcp%Q$*q8u>@bR=f%$6gT7MGNCHJMJ*ZV|hrk1~cM9@y!gOBE#6cXvpL$MqXd{#*We
z8(4%~x7Eagy$%MwY|5abgXA|7cseO*z@IL!x+=U{kcu_~lm2~GqaW&R;lZ%dOq88u
z8mMV$9XWee72zdYM;X7^@Wq885t5V@Y0!y9*@ppA`WyF|c+|Voqx!q`&fNTeya^{^
zIfQND?=vtCPo#)t(DWjKsLjLq$q;n()Tu>}fA!_YVZ9CQo&LRXhVxFlN9tN&SMi}(
zpr}myX~&cz3Ok4viWY%c(ilQr4qqDi=-pJ(0GPWB>=NB>Rqh|}@}ZBvVzdKv_jJ9V
zC+5}!a?_qDFfIl!w7zQJ2j5Zun3=2EtExH17(YlwH&xIo`I{n&nB&KfkGPv-3DuYT
zyCj$M&}Lv=*ac4Kkt_xiCY@Ps=@**Tp2rTfNAEr_@IQ2N>!tMO?dn>L8bV<%EWDnQ
zvhH=k1;RrY7&TQ_--5;-*(94<nNvsHtW7V&9`%uIO9#1UKbI?>&<DA>_YklsK<f*p
zhZ}QT`Q#aFKe4-OKSkrCb7LRt2iu?QvQI3t_kX;2n7nAm=S)#+6PI=NP}6+lRaqaM
zOq;8p!_;Y0*2VdQ(!8{&W@albCmSx^(f<jPr4HQ@-2)ftFiKtbpK;$&5Q-fDL|YCr
zA8SjdYk~My2V=u#5M>e(mz2GIfv?Ykxp$^R>axgGmQV5-HPzJIM#vB>WmqOm1Q($*
z0s&KG-Bgi6lDa{I5QV|41*J73(->j_yhAjZ*q+&zW6-jrW{0#hA6Zn_QP1PBS>m*)
zu)TVo6Ni_D4fF~_+ob@dmA){0x6`R!6BU<%1BhvswjG-ZZ!!R`>|&S9oMtYKq~u0E
z9t8ZO?Ka!ZW*XN&hwIyqACLHTS#`rUWJ&|Y928o&Qzu`hy@@kt-lIDiGlKBF#~=A^
z@g|Mkui<d;!~Z(S0a|esm{`SV&A7MdD>{P%QB7)9cl2o7rk^itDfyF{Y-!MCb|IeJ
z?8j5bICKU>IpWM&8-WxfO_CPmoO-YT`oWmtfu%UIu5i_=uf9u)cqg%RkKr%V!U3u6
zLC+sf@1ekAu7Z032;!(`l8GYbTRsjqw{J&oG#a&I28fvg%GRpr>hIkYM0f~c*s*H6
zgJCbLvQc}k;pZ1ABwv)ilA(Y0?Afx6E5@^|2J`sO<NB5pzIX36)R#P4wj4do%^e54
zgP?*3jj#l8NT*%9br&ue!|Z=!0h+u%(947N=+btJR~>7Kzvl|;|DUrz*(FQAPU+_+
ziH<BcWmc7G8;<(n)2B)e!+%gcMDPioF5Q;AQf=Fg8#iue;zM*b40Sa+C^G%ILHTT$
zaHCC^RhE~(7H>0(EvcHEM0%Zq>SaVVfJo59Vd8glcNcxyj8)m!Y32T&7S;Ziij6@_
z;FB8?kB$id0FbV0oLjxRyn(P33bK!s;S3;tow!cP_?va>Nc24;hr@YYiAlpG2j4m;
zcA97mo-9iG5*ux~;87bX@sk%FG?31-yH-IX6%|oJ0EwV0rz_XB*wf1+YDyOtsxxZl
zm6nd1Kfj}>a9-yAHm28b*>D>rB?yeKukXHsE{Q#X?*h>4WcBHxoLX$QQ9#;sXPD4O
zqsuNb_iBSSVrO%vN4Cj(QL<88Ye?oe3G@Y$vVyD$)_{7Fk^6~Jn~EuR7#<DTIFW|u
znyabZ6bB99KVOaejYV^XYhErfM$RHRak4*w<}3^89-&r$(M2(~7JQfkmdwIDrcub3
z`)m*m4{u8uXg1-CMY`~`%~CcsXeGB1L^L;d1+wk>XbWk!n4n46nK?6eo)<5Y-jAlo
zRC-y;tGRRS=l009hI)Aoq^3H47cT6|)vFx~`VEp39H7}r$kJ{N8`#A?J}}Tra$|Uh
zNh!;^edyAq>4AwvpYY?631^D%Ojh+RWP{!1%dM$1u;@Vj69@QPrw@*qpeU@sfBG_g
zGe=-{kNS4fozu%>>lC6aRJ(8Y?kSuACLaklWZ8n)<G@$wy<d?hjLPTWgp;s&f)L?5
zdc-DobMtGHciusahZsPF&xC(*!eB4xA3HCTx&y>1vJfiZe6TJ_GcuTdp&|4)`^A1{
zYiek`#+L@iGtt&|`yGe&Mn;b#oOkYlHZX9T{IGfOrsB%)ni{spZq6Iipx=|-&$9hm
z`6PGZT(&X0TSTB_VfC-%>NJ`b{>bHvBi}zd`S<sY#sfqYu)qiws>kfKQ2p&s{3a~~
zq`;8e{JEbgByOX<EXZwPV&$@G?&LWg9lzG@trn&`V_Nv>6DNELOHo%_Th@wKg>%9#
zKv^En1;$I1)(AL6^y`xGKrJZvHSo)`xfCoH@8}Qjtz0_3-txh`jfTeivukE{5#Y67
zi*vPHS`Syd`*y;M+-t4lE@{rzSD}@_GPzDC%~8$ni(}l&7V2?h3!e3!Tu)hf(4Hfz
z&M%iAWyPC>$WEW$_jP4n-Yf}`A%grtP53tx<FmMB;aPdiX}7#{4;;AU_UU!8i!uf2
zAU{49xE^cPQIP=|<+$L~L;N)aK@4!Uy6eZi2;S%&2&~DsbD6;FyE;edYGXw(@EgC>
z0M1fUQk-*310XQh-_{it0phZoP^MsBu;JY2;K3s7$1b<}r<UQcfBQUI$ZOr2CMrmb
zm?ToHGgv{I`~pYE;XFUUW=6Nr-bACjy*;$K6H6Uv<rICwy$2FzN~1q6FJMOh8u0$|
z0Kl>p|2h8h<>zmO-gSP`)q{ll9<MKhjq*eY(LWY?*&St(ptt}gBj!2%XR6p!+APz<
z@DFKK(>8gcOp&&U5J9R1nr@N2ELflx8--&DGpUPv{jcx+H3$DW&x2rbRet@Zn%sCd
z#$iA&i?WX|<95n^ZfXX41*|MI3Q|d^;2V!$uMsv#FR8zDk6Q<MY8YjXCMIgBC0WzP
zke*@A<*xw0uh*&5NW+R+l4&Za2d0a{Ovy!OXzahnB6Iuy(LDz1x@8l`fWWfG4vOnz
zL}II!EpJ#Sf1ms`dS<r<Ng(91xB73d>fcs}@3Axz0VFtGbYCS(G448+@!aNzy`i;O
z#KAXj&XPqBiEg`TGbq<-pF5)xiBa3JWlO>t*IM)BPN?tDyceP;Tf?N!N4q_E{Fnfc
zfL;q3iXg#^(2eXGv2$R{Ka(ESf7E_mJBzZ5Grc5qPPUQS_FYO7$NAn&t5N%Eu)tvL
z*LLgL)w0p{?5wpvN`BtS{ds3&U~WRwKUH7x|Ct-~C{j#@IE*ha-E$WCmC-SoBz;l=
zz+$xJc7~awGRikW?4<e<qqt*Im4mplsagE)CGX!KMgei|dB68+-TQ^(&EI4@GycMb
zK-y_qj(%@W9m1u_M9FDMMNtrPcUXE2nLY7XddRHT`#M`bmCT(o*klx45OQFHv|iBE
zKQoelW*jl<u4?&|jebXsBb(Ixk8jC(i^JLFAWL)v_pCeU4_WT|G0P0)-Mivq$z4}f
zP0h<|2M58YmU#KH#i&uQnT`}fO`W}wP9#`dpEM}lPk-+Vd&}-USmhTs)X1ip<<q@d
zaa;8kKf09u;t92wD{~gc5@hPKC*JI@D70pOP#h7Cu>}FrleKBAMd<iPxG2rkFYb~s
zes`uEa+nY+%y88r6d>4SSsU?DLb8gOicY6$83H*D?=b*opbbEosmW{mOTsG{(6c2M
zR2vVtgsP2L4q4a01Ot3R={fU0u?BJ0Tg+M572M8^zf{GtT`AJ+&LOriqY!*zTeN3k
z7ZOs^=cfJEsk!A(u8)!PbEQzRTf=e3f(&eiAUHs+U(+coKyw=z8Qp*MNbXApQfTa(
zsZ-Dm0Ql0;r|*elbO+wW^d3`u%Fj3l!ly=`v@x$ztEMw9h)2K?+|5m57(}NA<Il}T
zqd=e|E*U}!Hd7Sfx68hM?MS!?^yJj3`?4CU985^CNg6lJpru6iA!6qpkjEBuvq8g#
z_B3-$t-A>W!oop6Q-1lTW1Jw>64ZcpfX_(QJBU}0fv2XISZ%Q<-sx2>Ts`cB4Wjd<
zC9l`0QPPqNXKMyn#a>%^Y~jRE%b2Lx0Z!=Wqk+GgVVWy{b~ai0zjs_^*HO&|UDNB(
z!HRqyaBbGSQye+=A|&kJzdt90aSMS6`(_Cl;LL!w%B#t{0Evt~b;6V>WL=jZF<5`G
zIfNn4svp$ppXt4(>5tS(ExwWc_A7jqHUJKN@2s;^YrJMd=rdS&p-0P41Hh@EOO@p%
z_-+An>G5|TJ;mZ64jY&}-%wdaC17;m42oxnudlMS7dV!;Y~OB=K*;=<BkZO)F~|fw
z2872W7038F&ng$BD9jM$Ydk@^fy$xExV%yq#WvN@=%ep5!Q}%Dwr)>T(@u5U%>106
zG`PW;+sj)eV!W65eXcBoD90ah_v7h+lqtWnY_03VMZIctG81agljpgqO9MN#GgT(+
zG~*56!Mu6nprfNBkQaz3NDvS-KZ3|2)zn0nQRw2#y+hwx;L?aj3@qN<ybg_HGv|c4
zhlPXB+3g#@!c)~d)3pZ9+mk<qo~|Q0$I*Qp*=GT|P4m}2Q~SbH0~z4~-<X6Eqw;pb
zKEg!`I16$~f5%Svx<2RT?+*zPWimH0Vjl+T&O5thdQXXzSXgW`D3sBsrOY|I{e-_K
z-7)Wu?6;)c%;j?P%-76q$6rK|IDPv5evL0pYM4pNs>L4L9Yg*pcW2T=bS)T7*AcZB
zApPCTm#d#7cZ~0I%+)^@)pHUWU*X@-ofrm<4l@{LZeHde{&Qn!XsGkRk56<$PH)j$
zT(hq8*!eBxPmF9d8*t%Z4xC<<E5<YSyuTKRjTC$+G=6Ss5{5!mm7ZuO5s$Z@vk|^a
zP=NHtA$m<an4B8#;on;|&Z8ykOkJsgHD6mV<n)~ay~Q)H9j$oEGG;L`Gt;oUx%Jr@
zWabjQ+=>@eD_{^3O*3Y3RKgK$_SC6cS3(`9ObPT1Eyk%M$~b*}?gh6oX|GF4?!12z
z$z;}1N5`8IzyMXFC4BO3IX@p(T**FWE;L!b_R;R(Ec>OS>LJIl6G6PYdsXl1kIgS`
z?|g50%fw{(95*tcNCJ0xdWP_`S>6A3`!olj?SO^*gxACWd!JO?B2mHJ-CaQN^5kV8
zi&(vo2{e%XF|VAYxNq4a2o>(yTQ<gY8_SlJst&e6247v)xO&i}NsmouWIqA#?<w7V
z|DFGy*nd<l)%)Y%WXU!nD&TnNfnq&)q`swz{=8IHKj66X@v=(wVKK34>AU59u<Hnl
zt*7_K=iW>5qbE+>0OT(O$i1Zu!ZNPg+mpgXb_~%KeoYhT6R9(V3~oV($t`c<x%SZ}
zn)c&kwkTQb5r&G4(RHjRl6^n2oFU_{OsqoOM~R82E4FkD%R|Hyo<M8K)faO6QTUWG
znN3w7uf+=+5f!CUL$mq^O{)6wqhb|XdZov<-RU0X+f?PS`GyHW7q`g&sBJrAi_d6_
zF=J2_JAu9^Fvd~8i`ZLm2#Sx(H9>p&#EH$29`(`>81;L*nN}m1y&5%z9{>;D5r$Nb
zvEFIhoqK{3I>(xCkYUN{aH{K{@5I4{t4-*jG3$eSBxXv$Im#QXG+UZ?rTIt!1%Bh|
z+1Gn<VCZ9|t3C2)JS6yp^Nl()J>3o_hrX`{&*ZtHu%(h;{P=AWT`{_Ll9fe@CeE=&
zjdpa(KpiM+I0z{bTQ+>{7QsT8c`4ru7cZvrAgQB{^Jc{0V@yc40u0%^VAcl<^DRF0
zM4b-*_ZP9zVkN`&b=j_4<31lJF6GCqoew1_qVvv0{bmI3P)`IG=~wx9BChS5%~oU0
zX45c`B0xhq^8O|*TN*p2CAKjaaas%YM5Jm9n-yxN=dC!0YmD>Nx%vkR<a}mbeWtEH
zsP`^-jsu*K+($tY*jVqrdezCW(X~$k<v`}Ws<-fB=bml>L8)DE1`M_bC;u-E8s8x(
zJg!lrMlQq~($N31vYU(STPAVQ_9awH<u5Pq)VZ^IWm!8!jC>3X{3ec#&#d<F;E(C!
z+{`U{nb1^W>EewRhpWsjq<37rc+sk{+dq;<W+G`B@(aey(VJg3Ry|dg!Bh|`S{sSa
z5QLy9`_}hN5g(#>)FEamFV}N(y<U&=2<h2cv6SyBA<PKS5f2ie3QsL-jEaiVYB%ne
zf?%h&W7;LVnx>{^l80VsEE9puMP5}IGYc|7`s2x87i3~!b=5ty)n5!qv8mbl$&!=J
zMAKB+<=<_S^m$9d*yH(KW>(dgjHwa(r`1c1?Dyy4{dcwxq9vdx3LaUY2l({u-@R6#
zt&`Bq85e#IS+&YUvQJSQsZZ9SC?V?}Ja+8e_`nf@Y0%zb$F%CbX|}1c9l)OCfbijg
zbAY~^zHuw&_d)}Q!UqAC<+S!cJ`GpSsW%J3T_|b@wY9Zg6&2Op-f-^>!=RssF1md>
z92!+6u63D}{GFaY@_(l%A|Kg&&Q%JXN1s4{O5+7vwT1SGJroC;X0-kvbWG#a23LD7
zL8va9nbzL1@A-*K+g~(5UccQTV06b_!DxZRhR-FOU1m3v=boJk?o&re=~hN`dmSBN
z`@q5zMtD!Gxp)5up!BB@U;geMne8~A{88Be6tQs@!34zB6wMn}LTiqZ$K}e4B2idj
zfuO5dp7nYXq6pVF4VgE%vUhy%yEt|lDmj7NDUb>jMyn}PjNj~I{jVhhHvE;pj$v9e
zA*i4vymlF?_^K#vux&>Ff6Kr3s+NsYxU{*_Q*=WsKil`ImjhlE=JCF>_#ogHX{G>t
z@PE@^2=RaJT-B54wY@upb^4E_hhfvy4&9z4v{T;OSFdL8z348O6y^Pe5G`1Jx<>se
zq4I0iT=vXWN4<p|3JAu@>2F9-vRw$6*bMQk*(dEDO2qOy{{t_oX_HWf4OkZYd<hKr
zSrnV9$;T(0M6HuAt*wsEoZGqNeIc_TQi3HT*H4tI(8+MukhawMYn)nRG}mHq0sD4|
zL9sdP=!EpfM>lwTOVE=9fyKtsLEUCevhkuNlTJ>8z1RmsC8~$ao?&B^A+`XJcXS$F
z(r9n~>(^9X1)pU}AG_(aTb5r~cywg&^2=jdr53v+xG{Q4iV>M4{`-Rz>(+gA|Ngq5
zpjJ$_KVP>>Zr8v6ntq0_d{qC{_Dn`>n5fu|%doDjXe&I}lk#4=rgSw8!^8;_u6X?}
z@^T?M+pH7wO$Y8!7<o)&?W*nQRe_u!x6{{arO+RI!>i@-{Kdk|tw>p$iF#e_-^t@2
z`ITLDVk`rixV3;+-L4f;JP=T)TPKwVblX0iF1@f^1va^TPA|2FBJV+N2kX!~{ADXy
z*iGId;w@ZsJKU>xvUwRML%8{{eii^bckbMYxX??(kM4F4TJ||}Uhf?le+f6=0R?ml
zg=4R=ZR||=DiXpndRM_l;&R`S==bB;hmw-0o9o|s1RHvPh1Q#oJP@U{E7K(Jb%Eh#
z1_-W4k>ijiZ8mSK*MEmO-quTPFrIit%Hg~%nrO5oB*xMt+nlK^I<dSydAfmVMm1xU
z8m7dd<VY1UB;l<1{N+niEv<e-J{#J`?;@?_0BK=a<Hzg&U+<pDfMuwe=aF{9`{%b{
zDhsgbFy`<@T|)b{r0$W8$<<I$_gWCNdsC|Gjabjx4JD*d7A=jRG)Xt;!1sOtz9MU2
zRG~SRd!!<)2l5I+xF~Kh%|cGywoi>DHYX#n-oW>268*u~RgeIugev)^+OGMm+C9|y
z95HbU!2X4AS~8XXEP)=o7XH~3Cz9mgi$hDK{O-I@&x!D-r2#R*Z-T2T-^C#ENgu->
zRsHBi+H{&>n5m+;P10|;XS#wea>IcDC%ApmVZ>%mLsElw?RDSWsz&hLYSE+PXD{z_
zAavnrqw~|E63yp2IgQ`FKIXxKYwwotct8A99j>Z^)qC$a{G*NdZxmc$(C!4y1AcWu
zU}U)mLMH|q1nl_m>AnlbrttNpZxK0IW9yLp`+<z4mVVWy^6)oke`^E>zi&e<sxs}M
zG-Qe+{-D@n7Ge^}gsppZg`n%$Ieez0Y1nq_xObmsUA!K}Dfx+At6R6^pFJh(2mv(g
zfhduKcEq>TwGg#2DLH3ZDj)(gnB=bXa7g}2;t#|LxtCNoK6=I$=ly*hSD9*U-Ts^q
z7tdtt*UwaL?U>%t<wMA+__ky8tsHk2B|f3w2DavX+>9;`4O>u#+nGL1|J4F?$V|z`
z5XL)!RjlgJ!=nHe1O+7&22P(E@1%{-i(heY*yvbH0I10z+XSv_6Lcfh<8m8dUv5dU
zc0t6aqzp1>nW6kH7eNkIR2fEr+eD{@ubd$~Hg?6eQwf-c{i%o0Pu&gH93Q2Vb2Pqa
z>AKKCra87RKD+A#_f${hh)A5Ug5(L4a$OsThF)fOu<M%bxriMgVFs-|RgI3c0)BXn
z%UI9QZ}8_;n<}4iLM87)xpCv<S92{Z8|E;hg`toIW8hk!jmnli<8@!P>SlLsR(F{s
zMXJg_x>2Ix*!;iRuE{UVpHEazbmX65xLDgmCsI-bxKPyKb?-RLB_-njYwu0NdS2W1
z?_bd}QzS`}B4ZJWA{lE5nMFz|GNe*bnMy1q$`C?>siZ|E(L9l4s?4ELA)yJGQl8J5
zYhCw!-S?Ykd$#Ar|Ha?7wsl<=)$e<L&+|BsW8e2<Kg5(7L8u&BfR<WCMJuM6Li*$K
zcoJ-seB8tBsOwuf(qKy+g)DD5?c`>f>Ew+cM|eccD}rJXGUiclw(5Z&oc#G}fUT*1
zh7Y_CD*xwmVq&oHuD{n?t|zF&@;cv+OaJxhYQN<#0i_oc0T5rQt0-e@2}=4MkaNpa
zFQKf$3la7<^ZbXo?%#Kj5Auql8^>h~baPm<=1};@MN657CFm&VZzB{9+&`#o{a@s<
zQGV-iXfm63?ysyFd1WCZwkJTVgsO__99sW&ZPQz0L#2^Ef4zKtL;&XrAD-SWk4h*}
zLv|59JrJchb2&w4D}|;f6vsR>GvigO24cwYqLY{Qxyxl-iDE>W=ta`e;qEnnVB^ON
zf)6KYgmAb-kdxh>+{HkrB+z+s3r_Ga2>I=55}@b(vFZ3bSi7ApWdfrYu5&=8K;+&q
zgmDG+0TJTXbs<{s+7%oCf5BT)f!C~0#l-~xDv|5L?}#&8KLzm*=I?++bPwDZb~vr2
zi_A1I@oGXspcBvl0ZPn86XF(l32=_oy6YWax{lIdS~u+1$6*CmN(mVZw7DxPWm8j*
z??_&wq1<+NrGIPAxaTV0MA7Sl&Z8T6tIZZBm2Dy4Wf<>QU%$s7wMbAvYFFO-*hw@-
z(_KQWwfwXu_`)8`{?pFjn~Wk`bePDWMv99lRgSfSX%*e+p64b_p3LUzr><^maY(_b
zHEEU4bf=4fc3nw!7_D@Z>qpy0vI%nG8F|%V^UIa(2yz0jqyQ8A7&v_C{bt@waoox-
zA{I#i?BK@*T*tv6K+|`Me=C)o0!_AgVWrCBDXjkmaH9_)m0LDOL0E}8r8OK)4W@z<
zBrC5&bJ*1j6m8s}bjQ<>Cw3bd{oKgqu_TF9Ymmz2m6h~nA`^H3;7AonQldWo5yYqv
z6R;4mpty6K(N2*+3^yo$${_C`EKhm>?x6Z>lB=abi;nB@Bvbsh)@k55ee5L-8D<Cx
zZuIZn)Bf7~?;GEvnql^6j!!=sMgQKT9bQZ^HT74{TH2WmmO%uF{$8RXJNsra7mXO+
zgC1vT!@D>3ewHD>+vBJTgv9212%MBE8wXvCAS)LvcyjFZEck=pg8xl(?<kU;O$EMT
z=|I?kI}_i$UjFE~5J0opKgOMGnvu6KdFi>o%M6f32<viTHl^H|ycQUjuX27gb2O^T
z)jSrJ8v$=yMChhF-Tudc7$XFYB*j8%Yqb6o+N9UQ=$T6M<KRI0k}q?AfVK<yUtHY0
zql$!VST;oYl(^K@5a*!Xq|<Cd6kRCXDxjjFXFWYVxdxD1k{ocw9e~C?bl|`QQe)ZF
zfwStC?z!#1gvLkK{U&pBXg}e`Fn0Ka>3JPqrXS08baBCWXz5oPH$puB+dCNR{OKv>
zir_6kfsImC$d7R0#0(+%?-$94Cq6(Y0ju9HEF4>IfGiVQ3Mn$@bIVIl$9{U(P){I`
z0s<GYXI}X(bgv+xNEj%hPy^XL*8Q!8g9zb3hF<E5{$tZIU`vq_(g(r5f*7!<Zxz{3
z$d(A<tUH@}?C9X{+TCTGNMJ=@^x(a7faI~9g>XbP3G~SEXPp2zz9I0Mz?f^}=n5mq
zz~4oa$e7Q_#59f(29qRY>S6cj3lfH780#Of8cRbln~Q#MS#Wy_Pclg8U@RGUAxsD#
z?c>@e@=!mT4{nu}o%(qqWl&DIAy^WQz%YbUfk44V4h3~i-gznI%Kh+++URy`b(92|
zo1Z`H=&eWX`nNBs>YCuTVM9GjFMSD_jke(uN9`cg8w{L(^nUzNqyk-j>&{v}@&L~g
zb5t%=LI|PX`V{c?zmuz**x9{VkiDJU5yk}kYBG(|=f5(QRUkL$X>4npQWu^i>j!<G
zNGbr4XzL%2s$F<wUb{9?*lC;_(7s(e{L(QMe6{qCP#cX7CDR;Mn^{`sVli?TNXRH{
zH1{qtb3=?-MzmY0I|^hOdPn%&vE4<nx!`s&=M1`B@6BfM-D&PjowWgfn;(CVT!4&^
z;y9e9d|G>_AL8Y*QAy!~$GV#?wFx?U^d|lq23uqbb~nC>4Ex|RI?(MF(k$PfUI2)M
zGvCxN)8CJ;YS@%L^KfzL))3Rt)vb&DT&hZ&&1;%QZ@*?;CL9v!@MO@#%)Itujz=lY
zWQ${MWcX|Y0HHx%KGB#|=jXIVB-y~N;_f0<MM@{srRGt{0&ZQ$xnus{^9OTz@NRW<
zZcsnN^^oh9lS&5Rql^OqLQXu@2)7mQ-y`J9AEjwXTedp4Df~8RFHnH$;K9CJ58VKs
zRw&I+O#1w3yozN=#E3_ljvEZuA}vY#uyxRg5nZkC-BW6j(YLM;1efS{5>d^o`O`@Z
z?4)*eTQ1yx1Q{w=R^&)ucG9WT;+~9P5<PP^%gp4jz<{oB_!NvnN-5x^Xv@7B8NU*(
z_QRf|=0l=HmNAV^e{LZ4Fjs&?6jc(=m1sHwng7Nd*jJbBDQQFoN7lgTyDdq7WgU4Z
z-@$0a>Yl08%KZlnsAN7io$iA2>bA~k-VT8V;)<#^?NG2(58Oo%(rS{TxGm_GBDzdy
zdf0OBE1K}oIisFw1c8x?S+5-wJx?frEMT-aU{)xzH|Ivb95H<Owu}sM8$CUIVgepV
z-qs(9#Ui$Vcy^=hyF3jZC3f{A4e({7$D{255ZCI$M&E!yeqBIg6KockjNS#Pa39(Z
zKDea}-CBITpvg69wn@~&{Rh&4Ay9voy?;Ms<*#I~=yQ6z@HPECj+}N(&|5Wq{Y`hK
zl^(yfe%}&&<%I7jxBDAh^ZU340hEh7tL)?AbhE4mySk^3AJ^eW@Ao~F#!WOpUp9o~
zlea6rstxIC);n_kU}_kYm-Cl@`J3(}!&>I&IuF@z?NRc2z5a@8XZnfyC%Amwn41<h
z;~zO2X1cXB{9wY<=zrl|`mZ+4$&D`0r%TD~+x$}HhPT?OoAMQZ>cscV-K4Xp%4b^8
z8+-d*)7~ZQ(Y;zd+F{`%jhS(eS3UL_BimC=RVz5rS#mk&b>64O9i@6w6PFiB1s8m(
zzoqVxlT^Do&){T!V`^=s=XfokkFeDHAr7Zh+m4~&8;_n$<NLt-vs;;ip3zxZ7+kqv
z{(Rvi&opvJ-4SL{3vFLqSuI-37&(YHScmYrRadOmS){mi0yhFi$>RDwQ5-XS;Bclm
zu-K;(6TijJ8$-UL4rjwSy;WZ)xg<RDnBY6c;QYwMvtz1{O2<4{HHBO1vn$r=+X%m%
z$uqOjPS#ON&O9-TksN2PT$!&lsq~R06`N?LK2db;$(mRQ3E-7){@vEKD!z~z6d1T0
zzGp%5n5BTFVt5I%ba~+MvqpZC7{vIvZ05=Fz452*KFn^9&w1vU2!|RQYL;BjH7kZI
zuEPyVJRHgVnumc$0|V7@Wu{a1I=RP*?t6NT49Z2zdDM5JUxX7PG}sst2a3sD-}Lf~
z>|cA&re#a-$hF&1?C2D0U6hzVZ}5@X1LfzwTohBF-mhP4Eat8uVcQ8?2-MOSe+zPG
zbexj_?Kc11ekJd*<)^2}90ZMW3~Np(#V6&qf&&n@g9M7^{>wSfdTZ8^`z|9_&@72~
zR0T9%a@vDTCAq^yS?HY|^p;$FzP<zCX%4ygUq~zxlKw#GMx^!TY0DXwm0Q!J&4d{t
z`cJAd(z~{ns?0x`gGHW@S|~(Ze!NOYQD>Db88`lRkIUsd^C3kJ3Cq0m(xu@MH$Svp
zc;V{pVZoN!nldF6!gEbDD695TAWnC+mUKGX?&khJ!@4(!fivTf6GeA^R87xsv}MXJ
z5z_E+n}|&VA`h>e2i-AVTYdT43|sCctlWY}UYC(nNk4Q<D6KC%1$iZW#1CrSVDkyi
z!G<b<;S_%mUbS)X7vf3kZ(4;j#ef^)*ci>`yoalFxGQFQrVr^+UUWzjwsFYon^{>|
zk)_gZD>kL2-~Mx;p6LVV|Jhbn!&2r<XuosPk6q@$IZs1aXUwy3dw07s#*5ywYy_m*
zOaW(nKx-B=@Q2n!SukWDtRpKZ7{-8dM=wo#@vd_5xXPIt-T3mb2N}3eaLYctB|A~}
zME8xo=0(nbxQec0nUvy^SB9~YEh&rS+sp6|Hy_sv(SeqaP6!HfM=7YN_@~q)ci7ta
zWh(2#Gd`foG2Oa)@R!XKCQB6^KWkmIAgm^OqdhKaTnJaS47LZ#ELBrG^T&2oRdw}N
zC%u}Jbi%z?;Y%$fN!W$Ib#sWr^BpTn@p?~Ko^1T_!-l+<vl~7Ash(_WDGE`Vyr)8D
z?{zJG*RW~Ilqq<5mP6ubK9y(uxvvnDoE)Eax3l$u1>Ww!vELvTq;!71(qx4Db6djj
z@U1`AOET%;6%AQ|H{bq=m+va)^7q0+&f&*|_S?5vA+8cYv+%FOOZn`)`t>`fK82SA
zU(l=$#PPE)w)i02S|oezXbxqfY38Fx-2@7)q>%p|zmY;6-{+Qzx0HG;u28eLSU%Iy
zaKnZiHG79;anILbaKzLDA#RRp{`oLJzYFST4%K!vjYtu~W$v^2gg<4)%})3`wgT=3
zESm`8!+$4jr>GjLFn#Y?_v}G|H*#LA)g7b|dued9)t1mj=euW&;SJ7}GE;s%B(Gi3
z3wgP6w;LzV_!i3Uy$lzH^(^{8Oo(paL^vFmJ}kX2=DH5LyX@fA=wk`-{S$`bjS8G9
ztE{{ar1X5YnD1LbC%|9G&JYWXRi0(sA(h)W*YoJ$xr2`=j^E<3+cvoylG(^p#S30&
zntiIT_kOVIaCP8En-UCghTJ{*mc|%n*4}bCXh3n(;Ob(~=0^G?|5gELDRst7{5dy$
z2xu3AC^2MiC*D~aUpp~gN*Gcw2dyaD0hx&{0S78|0N}u&CbfHZaT6ow+mzqfBecv6
zy{S25HJx!RB6b1^a$r~2D*{$feP3Ut6Tc{nBX&$@byvH*;OEhUM5FjP0|TiE=2lq`
z4x5B8@yvaFbm~)(4f0)z8ihV1oy(_G)nbYmk@IRs`A$pYLGtZ?f<|1Yt|eeSs*;XL
zX}5jbtnuHcBoD7|XvhZd-YYJAvuF!%)wXSEQOD3r%sY{EQjb=lB-XMl>aTVt8WG4c
z@u{)$Bk>WjkL0rghrV1byCH)In>Wq<xDa3hWh-I&;rRvl3vd*fZnnJNg~V@X#*_xl
zjRe1Un0s+lP%iF)b0;ojFldjHd=0@;5hFxI6Q>@mn!RMnUf4h8Eu6Z1xhu9kiJKY*
zqr#hGW;Vd@NB6yV%U9WG@4D$)!$nz!u5qS|X87E8GzJ3mtC8UC_@mQxKzT7Djx|&p
zKKxS7u=7tBncZ7bcxZ{QRdV3r!<lw*SI6iloIl@oLe{Ja#bj`beC&j@(#1QvZab%b
zeaVT1;St@2mA(FP_v}q8o7)>yY$nPs-~6-J{)^ojmnGHL)g7o-PW9{HZQ1Z`)(M?<
z1Y={UKPFK8MMWUV<QAw;DQ)Tu;(ilaZ<}Y+kE^RU(>g`Tkqw9e)v+7NF(|Rcot`5{
zjy!n$_#R>gR(C4gj`js=*LrKS!7;w<YS>gYOZ;_mPyA)3+;L*<!{(%O7(zos@l`tP
zA9#EDby{j-y?%MNZr$2iWDDcBCSTe1p;{+zS;&(=l5Ib0m7SAzqTloL*J`6nz8(5J
ztNvnFYQBCnJH|00;X!~=R&_~QFVbI$l2)d;Hq^(C?S)s?L=?CLPBnFPNPTt7b-Q=&
z+)x$S)5quhdm!`c#AR(fCy@yWCo6oQredW;l=ek6me5$fG;+Sl*)7=*Vo3=g?b>$i
zC>)}w5-Bqdj|-ZiK?h$sb!~3@J4A7=dLw3vAZZMeKaDTBFR~wwK}W^m5~Nm(lGjJh
z&+LYmzKAMqx~PchuV{I~4^Ar|sOcO##`)c@qY8nXMH;@Ql62Fy)%n!v421xDj<?tu
zLx*-Ge2Cd(Vo$qOXVioZ0dcym+4S=#6WEsN202&17=@~R^yYnuc9YkyzmjF!k9hA-
zO!D#b%jO~%Q`L}&O*J1<&eRPtpMz;L+hExHaE&-T(mi6*pUZeZLwCvuH=;KO0wrAT
z3VZk)DSKUZ4=Ys-jX!uC)03YTMm)k-MG5Mj8e$^iV~TO-kFTax-7QE?)~CsSqHOEc
zs))sQ&c8Tp$dD{@eBsb99Ac^Uu)=Y>J85OjtE19R6;8L7^lH}3n>fIIyQ($qIufPy
z*KhM^clJYTIseDkS)L!eVWr$dPz&K5&a>0XLpRr4<LvZpD@{P57hbK?Srjbq1~UFY
zs?{IX%$9uUKoy4046Ccj0<d4RhAAfDQ|8{AdSi7rWDw)VjVr6FY6rFr`=>f^ph#14
z4ed$V7?Ng$`Ej&WhCdsjf**q~(Ap9DX3CMO8u6o@N~R#6E5QU4yI5ID_Uja)a!N|=
z8A(O1#-&(><O7q$n+VDQXr<t7M)f?cYFaNVFaH-Gep2Lo%v5AqRfNEkzzhSn<l4oZ
zW-gog4ACY=i&fp1FGkkZsx=wEJdKq@wIWmAM77^3O-(9Zig0yh^5dD?O=8xLi47;F
zrtLv;x|#`Cb%`6BnYu7&!Z2j@a(h$L&V*O(6?f#pcx1sv9ZDQY83eThvnwyMa8uKx
zPhI}TQx5hu7DGHS_z6r%5$CqumG7K}&;W=oLXCVJT^a1tBqO6g_^WMB^-YkActU2t
z^|oC#bCr&Xx+~)vMD|r)`m1@g*j~qu9dn!c@_tCt%A(<;a8~!n1!o!a8{Q(b3fTF#
zIBr~aZOJNk<DWib#=fAl)rmx*pLkpzHEJ(Xa6Za6u61Y3=kQ4bO|<#=>UQFm@0WG<
zJ8;s5Q+K-WGMzHT;Fw8xa>cCdB@m-BWComZT|+F{&oEmi*T|yn_HDF^TcmI4DR=>l
zScL_sIQh`s58EQx+J}a+?8csM!UywyhL;xr0Hfwq=-NDb&Q>-el6&D+UfxZDm1x+E
zTx3b4C%*${Uv}s4pl{VqUSb0jrEcs7ZZ#1Fis&Au96ZASvJ=}U{ft2tq!AE#xBKWw
z+}k=_!Y&+<m8y=8Jb!5yIV3blJ^hiW`)rtzkgL`9>xbt%GJd`urpPQ@QPz*1I*a|K
zMk`t$5PC6h83CB^F<$Nbo{4xcr)nHf_^ee0$ugi$998j`sCUI{#U0JWYF%YzrMjVE
z9}rjuCFkInUxtg;#e_li7*YVP4Mzq@-l0BnB-W!n=gOIf&z<&Z<zBt>#{&L-vIaHa
z>UZL@nh~ViNt|OJpLLr))Rc3FL(8pKHQ)R_etdLF%EnI`DqcV97xt;Bs@fxfPtE4k
zX!V%Nj!~MI?R0}M|8WwOn?zhK%h2cnNaTo%95i#%>DkuIuoIkgEU2HT6xsW;3>xG;
z8!lMs^|v!_$?E*}UUWV<Dk-<3U_p<?uHpXj9)%Z>oYVe;L%T8VOnIROh2>|4qv?Sr
zGDbY6F)MT*jMeC%jC5WEw*|?pgR{Oj1h`t?zETXCJd|gTF*=E~7=A<<kwr6{kQcM-
z2&@%o_N!y}Ufl3CwE^Y(q9m6Ud%q&AFUK0(Q8&MYn&lGHp$K^|F)m&EB^64OhL0IP
z62?@H+@^7Mv*7eK#}~-u1@9;T(mSCZ1B)@e_ee9@PhXI|EY`aGnpV%;tRXq+y(AEi
z)+3HDKc0~6;(b5*PC-GB?W(s%IK3ZK5GAA}Ofv$BKJn9JI%y3JSjP;#(hr|4eP^0v
z&+Xw>TjGbX$?l{a04Hx>o#$)yN5^%y&#YDZU6*hlvs-lUc56Lx0;m?)D$;f;j1zuE
z0C=108M~xBVe)u&|6OkVu9}OkjonUP{bSyLG)v9K_}i=S^h3X47Ur9kCG-tGeHoi%
z0@2dk{53<U>%=JmOeWyUk;{^@FMIbmQiS}%DY+s2wf@#naNoLROC4t#V!RwcKjz9a
zbG{z3-Ivp5vR7mRW!`O6)-CzKez)|_$jH!9`ik7YJ8}FHacymBdfM&l&a8xR>L3>F
z=Ip3*M>5EH0(am151kN%a>&-9seH+qeunY8$TF>-VVH4>*NgPXOBf6YtvE(r?2$m<
zoC%ZXZs8V<_+=3@q{@GzJszOV{5>1I7;oLXHwDQrD25L(ow}A*bXHES+uqbs>6VUD
z=aKrEVSRcX=MK?w&n+th6zrXC<DF|a8S6{Tm~k6eEBY*}Rh)aFqwPQasPIn_4RFNj
zIkRU=PAwZwN0X2vK8Ue0q(ChLJ#y4NmD!ULkP0sr?wuZIT6tth*e7raLFJ#CJuocN
zfSa4d1Ih&f9ml!V&Pifd>l>K5J2$pnYP?MPxm&c#HGtxNxhWUsM%FQA%a)lPqRtZL
zNfsHNmOqZJrYAz!a3D;O@D1Sk#GEnHJq10-Y)@$P>h_|i^0m8isV>SgH{?>IdtNI%
zL6{aQ<Dk(_H*v(9gNecGsV3oFfdJk<xzGbzNsPM(Ip|Kep8be&7)ohs|18>)!;hSp
zI`~L?-c}|zo;KH{M!xIof#;)#Z!efbG1SSpPQQzBolEOs|N7ULi|?O`AIHF9PVUU~
zwJ1~tszm-Zpu6GNLn(rr2ft=zRvqoJanmL-FOL9c%mtzRY1fNuOk|{uQY;)^mvOe>
z;`_<cS6ww1ehkM(tngjGd6v?(yV5ZolP+g&CcYz=FdF=29BwM0``c98dN7Oh;Nioq
zi4fzr{;+14k61oX7n!};(wvViKdK^TDqOF6x2UXoWA7E4FJ<!0l!tX)T(-f>+idyo
z+EXkSR0rCD3Am$DP4a3v>0(R@t&Af76(?oSo{8vu0vyO3`Z7>~x;uAi1tXEg<i0Q~
zh1kqOZK{D?GD|C$c~jLZcq}ek;hnH@X@WwJ9=2rQz-smn6E=Eyl!-Q67FaIfp6arH
zv;gw*k6#vDMT*QqGTF>5E;5}@VdSTTZ*^-%g51&U5VU7+$CX?r!-mBy?kpUzg`ZNZ
zOzD`(<h=S@f=Guum)gyG{kL>2he6EOTdKbdb9`^+W_Bf7(}gXl`aMK7R#-TcdHk%;
zX|ahO<18&1&g4x&$C&PUvY6FM>b+vgf2~%vzD&tY;V4QEn7D;VImqDx{QW4=VP?u+
zzLb(EGEGVRING}+%cLX>mCF{X3y_qCYs#KN0YHJzhD$ro(ySZdw4slJN67Af)+tb?
zBy?->CXPAH3v!HuTji31OiHvNVE_XW8=Y0yqqP)Q_%$N>I<6wSrd5zL6vx7$ezjN1
z3E)SvDv|Iay9GfO&!VX3(odq`5!q(lRPO{97hoe3^GGS3q0fZ#KSUdr2f|rj;HDDm
z8$NO(r#^7x2u>y*QXoMkB_-d`VnPMIKD;PoC!ex(PCV~e!Q-{Z!CTT!1w-q6K_<j)
zPu3ukt?Uxk7gB-F!TILa(|@NS;@JhEy}B2E6@N~fHj(JVmz`K^MZ!)7CrW!N@XgD;
z1B+Ob|GtCFR0^V~qOVa+??BK|rAy%Fs>g=82Q5D#p@s=G6Cv9W6EaS8pD{|Q6NvA@
z0|&N|5xKwHpwDEW_CldgT@^m;#b$GFvxavl@!owj>!0xMUV={4oPqMFC&(9Z^3LKa
zmeT2WA4u@BPU&3%Kg<l$rD&jN5-!1<udE?4soQb*mq_?I^{7r$)=xDu2Cw5EgY<w^
zdmzvB8EC0?cgpjutX8@~xi1Z0SEF%GyE1z}I7so%%O>I5$tuzQeef+b37@OfZcW_t
z<r@z&$An;+y!_w|kR#ih`zDee0J5=8We7TXX7U`1v`yVvLTsBrWfRdQApzJ6&@3f^
zjE6e*7WW$L+1rAcJ~R{s$(MkD{zb-s|8=R`OHE_hVK=9Z4?Yqn-Kpv5w?}jI-Z<El
zIO+ziI39=SJTh|_xBLP3lEF*kq&iudj*Y8*_xFuq?yrv}+^$Z3!gUY*J(?IMs3+8w
zU@lCZ5~fqgSurA*%)PifVt(D{&r`Uh?D9kXK!q9TGQ&LJHhNq<qC`h88Strv7bs5N
zeVn=}`|H4^2~wRh#$_~^t?@rKe{gugy8a3tD{eiPDe2TpSJ<lqYhZTp4ZZ0OVw_~v
zw_QKko(Iwgh0-D=Dh70oW#~4%pRC!rdv_aDuBBpdqv@VU#}}0ynPvREtKNV+L+dVQ
zcDR&T*u6ul8jfgGL24j&+zV7ZynamYv3#BcugG04PBS$dau^Zom<ffgFRdJc*rA_F
zwp+IyP>D>gc}Xv>;z+xWf=`LC>KB(e#BeA#M2FO&b$9ffR){F5w=@A(&X)M8sBP04
zjoO^KbEQ*?I|JKOiqmA7g}V+Pwzt-2=sLPt#F%68J%RP8^|Y?WisB|WH=FB;t4yrt
z^j1=qJ7t!>n-jdatf8)M0tV99Hqe9bkMplINkM?hC-QZsUbLBz6+Dz`aW|_Xgirt_
zQWB&6Dt?3j_ml@Ik`AdZ9|raca6NNRdRT{4eruVOIppD*tKzc~S;FK}%JV;SE}7ZP
zAS-E39-vTA?k>@BL`g-l?xQ36@Fo1zEb!Ce40VDnxs*A*bB=V(r<U*;{Ar=M3XXi*
zU{5lIPG!gW^8-b_Lm4bl`ccH0hSj@<K@QwP)%=+^OH2ERe96K>nKnF{sQNX$w~Om3
z46%u(qQ{4``1#$FyZuM064{KMzqBnO;u6X}*AxrGyucbmdL2tVDmtwmRg2!I1xTXq
z5qTh}BjsqhP?4Uv*HjR4DRW4vf7b)<|6YLN)4vyhxQVU+nWuBr!~T*^Q|G5G?ZOJs
zHLg*;c;m(-op-`;5qzs{yLJcPl&<s+RQCHPZs}D<iQP8j>-&XN-|!7HK(kO5+@PB#
ze2_>M4}^q>QHU%RviXbicA8N;VMb`n!Bre?Em>o4kA?QoVlVJsx;#`hH9Mg=1{0pC
z(6m9P<)PiY&8SGg$+>hctv{*SR%o6bixRnosLDKlHI`03J-f{7*H<r56-qMEmeey*
zaQ(LdQZ+ZFV^YVO{I@d^!h9({0V7fRQm6qNQO1sFJ#m|P6gJl0gsNZ7O?}&Lf1EH%
zWrGlgcx>7vkRR+TWR|)N{O1~b9h`7kWEVv-OL|pCQ_=(QV2Iq0VtTM<s@Z}CJK=?>
z_OA1S`O}cWp9jmg2j&pT#W7t6CD<CSHjVC!-oJY%I!A<P38b6mk=<n4S;h^QGp8H)
z#&t-YkTvCn->~&)X#N_O{!Mzq<lR5OHq_aMf(4?dr9;5&)>PUI)sdULtE02YmS~cM
z|9s@$gn=$yx-cKEvjB_!{`)$)B9T9xqmIJ!Iuk0(UcY{_{P9!-A0)rpYaS1T@QP(;
zXaCQg-MHa>RMLe>rzc?P!Kt~Q;AADQpl~KRSxJH#cMrO0K*rXTww;fgTqCq&$aPp@
z1^9(A5YV<h3M-~hn|2L5GJ-!H9&av9eZHtHTqXO7ObM8kNK+nW(0sX_+C^F-qzfY6
zN(2<}^Gia;hVN^9{crrw#|z={0Fju$(^CvQB&>0ZNOV5GrheRU%{r5cfcWc+I|%)0
zs^i8fw+!Z8rMlbJyI|uY@;0-)jYC@ESW|L?Tg$^iTw&X~8%YX%KQgE<BueMrnarA%
z11>8Hs<N^&(;l0$r#|Q0XxrUW6aZ$5TNK5eU>|w2vk6~I(>8I^FFcC5Ej_G3F<PrG
z9q;kEa-%`Iu$MLDT5xW7w}2#<i3ROQ#l<vJCO*AB{LF&Aw$c5ti|JvJR`#&=P8cBX
zF%?Y>xU4>`z&@^I0c}s1d?q9ypcNrrWPRMC*Qm45zF(6brl;H?H7`p<I+wg9;=7pD
z6hyh-vFzSnhDu4cPoMqh9*9hf-v!*I<?!gyT@?OIE;VKg<^ai~ee1%af<sKtUcm)F
zv&%vXBz6&jCuOpXUCY5-;r*2|Tqt!^H7{cD8nX;^OtOvYwF~+M$u^9>e-elS^<UKZ
z{;wzM?lCoOQ^R$hUzWE+l>Jk03ZNf#8ds7xL>4EHQ}DloYvTs{C9Y-+9|pPT3p&KQ
zB(ASI^f{O+$cEF40nYk>sqVeymQ+lAPEAc8T3F*=Utih*_P_~r*1U6*7cww<a2?c2
z1tyM1JWB{CW5nPkk!XONaDj9>`=Bj9gTb9jMCf%U7SaD}<8|YBPX46=5hy+_qAu{1
zxg)lG|G0v_kycO=q6X`b`fB$7yeN~b65>a@FBp8}ZE1!f{xS5h6olPhGI#EFU|=kL
zC4>y_;B|lQD?hWA(VEw=<_priW2d8Z@Z`z;v=>n>_C`C#-$|@EUlM9jwjkwa_i4I}
z+ZjoL^2db9Z9eigrVS|c`qgNKCQwSk_(9Fw+tWoYA7FI$ny5&)+JWyoBc<9_?6}(`
z{5o;7xVH-XM^5h`Z4%527a2Y%VFhsp@m+UG8N?1krh!cD-3xqo7W?SVn~T{=<Cv!*
zhQ2Tu&4<PSBto4hOg^Y<V9*;z3dU^;LK!i&JbF%=fecF`AE;${-MG)zetzXCv?2+5
zSTcd10B0_lC`(B#BrYsz%18^g>l$_E(^n~tL{ZPOV%!?AaES(ga3*v*H=KZ?*+zm5
zhA=-v>MuC};X%%^uaBV$ou8Ais-JA-E{F|cFV*!n9!lN7!171`?TdE*00j|?fipEQ
zzy61Y6nzDc+b<~%1>m>(FGqB>{0P0OX8O#`PsI1-05fUJd3rFb`Pdog!UJ#R@Im@O
zaXx>Jho12Ke729+nRxWwK>8K7^8vsITFt$rQZ^Wo0Cl*(wJ)W&%!9r>QIyd-VZ#La
zhZbo#ETg+)OzCx|nu&ADc*tq<+JBipwC+%r0}IeXvwG$wfy{BIN>+_|kp8W_!|Km9
zU@bV@)aI>2%XGe*{^jl3;#D?u1cfA1Fy?iFR|(d4wPzz2GKr-1!?0=Mt&xjV+%Nd@
z`ST4r?01krmlFZNq>Yv?4gGBYk$E$Cq%iQH54J9%6px9CnQZ%Z@S?K*yDuXyMePta
zRa@j**!5L4g9BBNEJ`F~@#B8hWjXE%5Gx~oDZ}iy0vyeR!m6j`qXOzfSeOEV0j7w}
zLv|xTa)_4BTOGha$e)1oPG7kqPuXK+V&bb~ypEehXhq1$g{q?S*6Qf(;%>h1q&<!_
zfHT)@V#|;(*<YSi1+wL|`p64FpB`W$H5~(ge-1tgvUF?O=f%WDWRm_ID%@xQsKpJ2
zLzXyoENJiuV@GHJ_ziI-sX=Oia0n@ZM0Nc5l9w?B3q<}Gj<SWO_AR){Qmnvlo>^ZC
zEWGaZo$iutFjk(}W+e|YVZ*lka!~-z7dVno2eUtRg^YD){PJYk*5t^Hlx+ooki`!d
zn8fakbdKxcLGy+~ABap6>_TzGa-S@#hzr&vVkTrpAk`>fuG8)sw(&c;^!T53lev*3
zVwG21J<(Te1AZMW5-|$R-3_wWut_nbJF0DkoR}*=e3BC|O>kF=n$~>AO`l)LnYg!^
zU=Tq0cJk!O&$TqNl)<t<3}tV87a391UFr8Xm6J<4z!SnP?GkEXnqmUZeY(PAH=5(H
zzw*JEP8Pyuy=z!8;qujbm^>ztrt54PO(jeX=jiYI&nvOGP0b~c>iP)7rgqFKSv~jd
zOtOUz!&0}2zNMWdZZ_V5!CC|rm_@1km;Nr#_xv()v@yr<_pcfuau&@46jGDf?~mW@
zTD%<Vd+Np^-8az&SggGIYo?yZia*2W>a9res~kZ+jN>8ou(*+d?ub0Y9bV|bk=+^;
zQ`&vtkxr$@$JqCAUs1lf$FHLm<GCLmuC#JgI{C{a{IYSr!{e^?pN8e+dsetw)tvOw
z%W$6%deTe0<af$q_Y3F>TP9X!H}1^W*IYH(S;DBOd%=3$kS>YVmZIK{f#I^PC$4js
zeeAlz&dvtS1_~v%5i;^2M*hlGHJ8K!o!{&FX4XiU#;}8vNZSq+_CH7x#?~amMjBIF
zv(6%~u&@^lv1Im~IZ_v%ynEwgF^n{k+=Be2rgQddcv$|CV4$R=5pX+KE-4h9At!Ia
zf+H!Do%Kli*_M(Cbd2mXIvb?!V!&^yG!6#_ri?wTfPLgL-Y$5?!E--aOF*@ed)d%y
zK{bbi2b_?!gw}3o-iiaqdd<7=WaiT_m+t;$M^2;Nk8oHrZVvhlAktXemcd~pnIctD
zQK@-c;69okji%<V|J(~tKv`R1;XWv5%55G;VeizVJxt)KAMSMFV}8&&i~$}P`RsY~
z9`$Q?uoV>=Ns43}&M>cco}u!aKlsSJt~a9b-kum!&`R`U;}#M8#oWdpw<!22Gep<I
zKna}5Moi(OBj>lQ#I5?pP6%~Dve6}?zNyzTZ?j#&8|t5YTX|x;Y8h%33e%mcZJ`D|
zfEhzJy)5YN6R_j!=daHX-88&QbK>QqHZX9G33tDw?QPMir1U!L+=nh0R7-yOTa!GS
zH|Y1~^9?Occ~&d+(V>wr?Q^RhZ`eA36<bzP_&o@|cajr5WC4?6s0327HjLG}Y-IYW
zAftJKQrcK@O@R;Q{&V=`=D0KT<u6eTV>q8Ja33iR4By4ggOH^T<8CnD>$wM$MIs{4
zY)>Ns?Y3RtjTx?}l`{3;@$md7@j<}D1eXHnF)}PWR8ye?H-;byDnv`8!)a=K7hn3j
zr@3D8hoBUBk?Zf1$G;zkfL?bHED^%u5WAfvD6`6tVdgL@LMvqwndqx^g7zbSl+wAE
zd+R~UKVSN4pkIeHsH~ao)6hNzwohZPIr={yg@8z}2=aH1l~orBL9gvO(-5|G%W+6)
z*RkXJ7xBHij6d+HeGxkZ9A>?{qE^ATZ>Xk{wYz@hXd^Yl^55Nkp!m<5H*O>yzw7rP
zad%>u;OBX6ZID3xvEy3tvP6qnW1A^i6m{1s3eFiotSenr57o-(0}~zDK-{}&fxpH$
zlue;aV*r0#J;n*Jt8i@xQiaoETjQ%K*)})w1{789??-#=Xms|$>Ow|1Km4=L%5+mR
zyWxuLM(eBQ#l3y+f<Uc4s6z3UQe4AP&1c_J_bTkoE<T+YY;=~fPf`-jNkVtqqG?0d
z?0Gc!4qzJ;rHtfxXe$&i0!uqqFpoN|Wuy~*_4}RDzyBeyDb=<}odr)F?^x_BTK7-y
zR*}BH?k61gJexTeEUg@YYn1s0%TUV-+Xt`*(TeE4yXB1(Q@n&ciq~9PB;847sdLzV
zi-Gd(!6A<2zEN@aE>)m=aB$tzf`Yr3Z$%Zn$HFsYQ|N{Tf9yCt&D+d>s;Qq*E2J(P
zf8u03zYL+y<)4RUMb5{R^YgQcTEy34uM$YIXfW9e2Uhi|<eSIX3!Ua)c;eWzXC9Ty
zVQRMbF&X}(hw|h^0bN(N+?wKFW_Y+{kIWypgNA}GOk}(cgy{@doIG}D1VDT+>Ycgq
zpBDdfIu14hcrE+>@imqP?;m{1bM@%|&mYM(EHc<S(LljNoWkF~k_VYSn9z9x4Y#5i
z=c1xvOUM97ONpC+lRy<O5iIb|NHq|JP)uY3<w#~PUD}ywKi|wOi<l(|%RtN|P)x}~
zH@<|%dR)iCs=1LdO}L(u6LC8pq4G;4%}p8g@pzO86~qyfC;g0qa@mjf-zcJ!7F8w{
zag`pyTi~s0u0D(>Mhc2BEv-=57Vd&Lh|oT)5?9wO$Ar;)=E%05KYzZ?-dMmIWYd~8
zIe^o+Sg+DKp<pZTWUqWS(LmdX4@y%lnIpo%GpVT-zI>U7hsU3Va;>E}cjVUgU2Uh=
zT5Yk?%wENA9dSIYX-kA+cK$)GdGunRnF=O{t3$CJ*4Z@VoKwthj9Pf%inX`m*J>O2
z{aMvEG#TxjboxL%9rmYZ4?K~*$kMVCV3LCE+ASBetHm;9Wu3kKeU)wyR>vno_LqNJ
z^5f#(O7FyK`h3X3(2xM>i`4}{Ktlq@nBW&1Zd*LYs6Kvdwns{n$NQCO(b1hLn(-M)
zu6EP*EnK_0Fd^O^2!HnABj{m3a%yL%JWb496=_Ltl)YKd%kys4Pi?o_r%*-_GkJ%;
zf*jE~VVYoX6ZEpiZRWl?<I?>K_uZ^M0r+{?q(2bTgZ_o1>Kk4>ioWVLUf2Ht_%Bzg
zWVr3kj+A6=rW9wRAs@|~MN@Qkri0b=t<LT>l|+B!Ep(FAJj>Tp^U9cKSb68CWyGV?
zd}Ps9@7a|6x!LkahWk#>yfs&^1B!-)z4C9en02=lL?59gZ+6|MH#7Ncg<o9#@09lX
znjQ7J)_={&n!}H1!@^w?FUX?4)#eqtS#zpSd{Z{AtR>6mwtSIbbT-zq`bD)%Ma|)i
zB|{Z>(SO$dda9URu`a?bDqJ_~c(wWT#@Mv)DG?C^c6I8ti&m|(;_N8jS*t71>vtAx
zGiXg~-5@ZU*+}N1<^$RDTjX<XfCW#TJ}9lE)Yo}bA$A_{^6a@khbZqYOuBlD=oZ#`
z&d-d*>c9UU969-FVf8~d76WXa7Y*DUbYoRQA^j>Zlx=6u?8J!@BHYR%rL&B@QX1s^
z<mOYGi&0TWY!EWc-N)0c33K#|<^IZ6B#Gm@EGc4Uy@Hb8xoP{#|Gr*cv2E9``THK&
z#ZA-p{gv>1?aVKcincHfS9)ttl<#ob&z3y&p}mi``Ou>perbwlk8N4Z|I_9ZUr{{!
zP46;LJL#OO(f}(z+l!bV&)ZiX->Zj5cK+FHn}arKql^#*-x!!~ic?O2?ef?~=KEsn
zS5Dpc@~NUzu_?kfY}>fcuk^|O=4j9-Td`NhubbMwH^MHYr>lUqZ&y8q>clq|W4Oip
z%2!(c{KfHb>Sh=amF*GHG;qAx4B70j!{WQsW3Z;1(#erXqaVg+H##{KD0$`Guzh?`
zyspV>3vLuR&I)?z<2G~cEXN*a7d`7yI3a3bq0K(O{Tr6N7!{klZe@x_x>r!{)3Dsc
zQAVehjk2j9SrF<rGp1m2)WWf8mT`-QR<9{g%2|YfI{R7T%x=FF_T>1UPl3`y`#u@N
zgx-yM^}+Iu2X3-^-&NjF1S4V-BF`eQdFFWL&H29BzA5;aaq^jy-8xJy_gwq4dF+V{
z_k@<+yL`m7%8B+(7Arrxn>Dyr-!26*e7@EQ*m!=KL*z-<<;{+<i)xN#Bt+!Cozbin
zdKayagJXgD-ukCv^S+33a5DbrUT@%|{cgSfAT90u)lMlI?`AZ6rs!sM|DJ+6pQ&K&
zQxk3u41Mpgy1;zRNK~w&{A1O>r+An)ObI^Y{ZIV%IUhuQ@c+Z_=BA95xTFz_BGwkz
z#9BkaqdO8<j<CldR$}I@(sBcLrC104p|t?oMucbw2?wSPG5nQoeeF?4q-<0d{PWP2
z4UBK&=S^FK5=2;hRP{o>A|&4R{bmLsK=_Ei)8xO8SWL!U1qnw^?^Rgs<2)*z=*CRr
zx9gJ?PP)F2&kxqy0WEtrK3)Oi5y3q1o;Y`sQ<mEC>z7n0_UdD=ee=r)V{+=9llm4v
zeXdx${^AnfIa~6ozc@_S28N~qzDgHbCgAqt=q7eW7eAv%k+#JT9T+REiNv@aWw&VQ
zFbRLyS6g@dQ{AAbtV-Jo-N}*jbqn{~#U)u*+HQVv)&It`kQY~J;s1#V`*zi~EPN*J
z1r|?<l`48?EL|4{JqUPlRj5?QdCcDAALr8X(14Ymq0LlU<zTXsZ4BA|gMcm3f4X~p
zglBY=gVn6sN}M(K;R!i^!2*$;fRwbk-|*4>i)-E<?NOt;gC+@~lAvE`R$2U})*#%{
z-riI_-hO~-a3*zD8@Mdo$b~xE;EZL;qsEmfPw(3S6}-GVEaL{egEoQ)F`~(6mOx4O
zskTK;CoSoyGENb#t()tVS0(~jC!Lc>Sm0OnfyagXOMqSyRkXpNJ&yi97TMhUm}UJT
z5du?{anEw)4i4e9%+ha0Nap6MGsVIETir5&#Yup4Wug|of8%rOh|q7pjvAKM?`jyg
zi2RmmQzp*;-1U!6O0UG9QiAmF{Y#7<BN06oq^@SBzxm6>^*^y_@P)!l)>Cuo4<7}G
zB%$y-6?E2lC<)6yN%WM*GXd=XKm9JUxVNN%e6TA}nY2XEO}P&aT!6sgaV+*M@)Y(F
z8D3tJZREYKI<HKa;HCdHJ701;2gESEq7PNvn^it99AWbw-l;O_pObnk7o`UoP6x@u
zU(?1TQP*m)1k!Qz-({gmil*Nzs0zN-8kX~6eV<Wl4vVLuyxN|!Lx{M4|2p%fqTtED
zpJ+<4e23Kkm%lswqFCxcBR7*ZWSy|BgN#JtG66}CHN_35&9N|kWW||2V>TI@Ap%E?
zD|knATIeG?gCc=B2?s4+#50GJDF~w(UlRSKItBxe3&Hva&hF1p_wSp(wP4+6Pb~D9
zkT5MsB?b2ffp<O3*jSU|ui6%gNhaEigYaMEx#)410Shm>x@JK{LcZt8;5~Xnw4IKO
zeiQZa&8@A)$5XqPeXljSXIZ#kX6bQ#NIfxl{<>qr^n};H<$^f6W;3s*4H;daQ{i#v
z2VIoYPLe%}(QP^@X<=caulgFTg~L}r?3c<sc}-mp3Q~mXKa{qsUeQxo6EZR{E>C50
zXk+xti{x~+!<5E(tPAk^ROC)qn=+Jt^7Cpx!w$W(|MS!B+gkkKpHDY-g6eY`D!XFa
ztV#{to>q~J*4PV?uL*+qdcIe)D>jXDJl;noSfo3wygAiy-A&JEzJkr64)@s(Fi8I5
z!VA_3g*#E_^~H{LQ<`Xj%C4;C{D79o1IN|)B>Wk*?&-#-$8B%9mX0dAdJ_EpFVMwh
zl*!EAUE|V+#m6(QyZE+P$(hk-v`Y*O79R@x7n6wiyZ`wV{#%R9O9r<c_seoml6|9d
z=cPSgQ29{&pPO~k-&kd1YZ54QGa&E676?r<;u46-4-Xy2j@0_$J@QK&6J>-)h#)vA
z7DZqmD9`NJWBMBg3mRTZBKVz!8TIoAKkug(6qg_8uiUcM(M=2tmX-*u)D?4wz58D#
z)ou|S^7LV8e}SQ#ajz|bl<)jpt8uG~#+!@GGc7rma3^T&mkmokF5Fg`F=V*AgQ4f(
z6-KhHx8R{8xnEv1lG0UxhZOQFEUx}u<^oTZk#_q&|G@I4bD`GNS!Uz%lWk@g98CEV
z2h4}Zo8-^Jh|nEP<Mq0^F0?5b>||MQ@!nUvVSdfCf_qCgr-zpOg47=g1<!3K*@nU-
z{@XE5L@NnHpDg14`5}(u3_DexbvYUB5xOD!*d{gi>Pi&2XGv{^o^{{8_0QZQ?X5Fr
z<=qHA!ecBk8&@Dc#*wjKMtyo38WZvL7yUqXbYn<KKx7jvi-)1Um8^s64s>YQ)x7Lk
zA^I|>PWSy>SC?2_As`5#Vn*i5LV2{X^%WJOxq+&5cec7&^qA&qgI(jkT4q@`(rub6
z5qkQJC$7gTJ-^=gUwh!G*vX}q4WFA&a?aSw(sfyFV*1;t*;0+fJ$Gz;x#3lONY5Tk
znY|mIuhp#xaW9dQ2=}H{4YA1n9+pxRB2T}GHbIz5GZQoX$;Jb3^DLZy7H;my!QVCZ
ziz{+YFjXlDm`2q2qyPJJyxUqF`X8Tz$x{;ZUrBQ+x}2sL@zd_hBMaT^EIzj~t6hoI
zLA2kJH3=k<fV*T%Fa@$}zQOKx-83@kpJLadk`aneI=9T2-qKM(L0or5FBjcP%E>sJ
zbD%QaKGD{&54g5uxZ8(C;J^vr6hj+}d;RB@yYTh*mh*Bic+}{HFvTd!am=goqQ07|
zQ_a_IUby?iz29qdq;auqYoWOk=I@{~zVMG9<M)RQUp6VO5DkTB>JZv|F*i$ywBYt3
zzt@k^NOFBsU;9@3RZ^`I#gUcg_x=)V2$6)X9rK%N{Qhfi-dplI_eo>2=EfnJpE6S3
zS2Y<fQ9e$uk&J{&t0I0s-X0RA@tM`%Z=U{Vsj62_l)c0p8=A|O!E1cOv8MQClFyGX
z)k8Lfl?ShStX-p@Z$7qVr#xx=j+F?7m_D*wK-TA{9qNi|e>|%V7^%IoKOy-hhtMJ^
zruV@F>2beOcKqvXu)`(W*pC{eS~^q_ttIex8F%BW?<Kx%6a6gT=1?k@h%=52mjXd@
ziO5J(Lf;@cb@AfGd4&hZx`(bz_~)nVp|g=3ICkt=<f~nko&`7mbDQ4TT7AA{dp!Bq
zHbod7pz$-^t=fF=e&O0DQPtOfvEug@Sy@|oa~jUPwy|j3c5iXrfLdD8seWEg(x-7L
z#VsoxU6fB1&HW)Gp~F9AUSWAs>DAcM0V{3tOju~YjTcgr)>w@l;UweH?7qPu?bk|g
zGdl!{2>7HVR$fgT;u`&o)FCSVkiegv_}2sH73FVNwZ<e=fT)7}!7)Xz-?VfB@#EPg
zj<Frq^l}eb8P&Tb2Ht7f^6Hix)gonc#FGVfwSV#0Isf_CfB*B?y+5~{&YtE>6Xj#T
z2-lVVYVLtBMk2~$yg0i8m4QC4a-LfUF=360*W>jOGv)N^ym5gq@_Oc2%{avLkCp>!
zwe@Q5G#B6!(cen1LG10nUZeUzOnd_;qv?`N_7h5krw<Zl{<1RO?`2=<xpIHY)oxc<
z`iu%$B7y8%-QYa#Ys&to{)Yc`VuGE29#3%V6Mf6F@I=ct>6Y{+x;^^2gwzp9PskuI
zxxBukPz%U)1DCN*X-i8#q|>Co3|*a|Yw$Fyy6@orCF!x7|M+HrD#0R3>1>I6@}g)3
zLMCGmkHpWl=(5J)P#<-3EUm3Vm8TyP{Um2j;rdtK65MLWmMeZNa__V<zG=eba^WM{
zYSkjYXIWXCkk?7I<~%2I;^Uv4$o92m)>=Hne_V$9uH;pcyMFVBueXmMi|hJn;YQcb
z^(#wWq>YK|vT9aI`?{a*quxYUN4VX+aHtSb%1gR+2ViOL12qPhysRwKTQnY`{?Kpe
z7M{1ZC29vJos^lFSRGbT-sFc`_9YS9*v?KvP<l}NqGIA^6$(wEm|%cr=NP*gxq3go
z{qqF1mGB~xPKMKy=N-D|TH4@PpgTG}O}(@+?mzaao~u>D0I^R~TDcZ{dD)hJ+JBNo
zM_<=EjLrY`Ys0*u|HuAr4a^!`#9pfJi4&!;>fQ!+ExMzD$EQHQh<<6RlyOW^71kzn
zG#2&N63=ye5&IY?Mn+Eg@-l=2rjYQ|*IzrPZ7Z5^L_3k-IXfTmc^fZJQC8D=yN}PN
z7O`9(x<PvPq4#7r^M0+Ne6@U9sJHWx;5fiUSE^zGkTt}_GH%uZ##5&}_H3!Hi-c;N
ztBf1Q&3#}?1|DUe<kDZL&fIrTobp&rps2S_3jzVIKzMdNeB6<U1B6O4A{r%74Tt3n
zgV%LSDfW{S5|x7%T%<3V{&Rzaf^w)GkSbRlDIrvI1?DzKRZ`YG8B7aBqPDr(nWN6W
zf2gUsByk+j4GKA$WF5{|pKD7?bLWVqksFl&A^sBskv@_#h)1FNDOQ$~n=?=O8dZ^G
z8x%NgLQHHd(O%tzD=-~<jEJqr4dQC!Rc52qVXE`zuXGgtDG4oH(Z`BA!3#d~K-0ZW
z(tw>+*jLR(K|E>O7CllbDi?^lZ|G|U%M$-BaeAQE%K?Vp+fnBWH7J8_QYXiZrcABE
z?xwV}!wvyWyvlx}H$);z@LEJfot<|3+*%BeuBLX|cE7={7TJ@^##eH9!JWR{1d&7k
z5ZD5pk2g?wJ$QNKFPc5WGf)}5s6Qp3>pvG-9FLRE6g+oTZDL~LfBTMmCY{d2)KWup
zJEHV9anFYq|0;6L7xEIxCH#s&5jVWu06-nfr76jTWt9M2+`iTFU9~I@#?f0X)K6%Q
z4xdxljCwrDv9;WWY2sYX=|%zyIU+6LgGz}RqY`V}r#x^>my(E42JA=uD_eP?&=04<
zDuk7S2zbXbTfV)na!B0Mr%z2jH6R3KG1ydPn43vUQk#Fg7kYx=I#VAX{aL@&%QdR-
zpY~M%F!;I)*?>BtFv``tfx~uVuFkMNVc6Tb>E2A)*0@<Ii?jj&7Ngd`0C5JSwh|%Y
zU6u0jIBfLA*mG7C$~D;%Q+g&aj`ghHYc&Bz_<fYD+y>uCtBPFMPfcFjZZ$<o&#<sw
z9|Ty+zz(6)si}Tm?IvuCg}?dfg7>u6<|Mg`r#e4R`G|O`c>N<g0FcvvHE%|fmNc)v
ztDIh|h*hd;XB0#yb*162)h>mwABCsOqo1tF7VW)Qaw@W-DBhPjw;)*hxpVX`@Iw(M
zs<!3hVqJm8scSRk7UE`R1}mfF!*%8_^{fGEE^WoOd>c=zL7K*|lFKWmrzF6c?MIa)
z1ol8IW86L{q8s~gUbbcV-zXo9(>*tPAdy8@PR<*JYz57!54Vcx@)7Yfr@ZRea`4R$
zQ@TzI(2$a_?sLq|+e?Hdgr<_9c#!{{gXW#X@3{|2xj;t<S|JJ}#Yw!qSKZ}lywH`<
z&fNOcvh<}_&8sgtbd%elU@Z2IKsv<*BC-Bf_p+sluyhedPlv}OFy*r?ej^hQuo6tx
zA?8$QS=n35_`vZ;ZZAL79RGYR)s5Uib#xa~rp!dJA#_LtD7XNm85B|?-thL}qP1yq
zpivo<XoOjc`s)Ys3Liww>cf}|X#}QPTY89K_BcRKZKcBn;a5TAl5E3R9I(yQDZuB}
zL2)wAw9GZJhKZ7FLja@%jv-O<{P+dmH7SX>-wLd}(y}Z2(0xyn4S1>y-Ogfa2rjNE
z8XjWr9om`q=AuOO{8WgDN6|-%dSRHN1uu!XWJ?l7UGvxaMgrH{eF>ANJRTIhgMHAC
zfds}te3G)*9F;5rO3?>xfkwsB<R(Eeb5y7D_$ok!F<ZV}`t0Oq<eF2^#Nnx9xU!qZ
z`j(&E&iK_uiMSdf6^G(jkkGqfEb2~X=g{}%_C-@xPETIEln1};D?87TZ+~`pN3T-}
zi82Ax*7e%%ruj#L?UfE%nPr`C3{32{V_HXr9p_e8+#9}cc$?5Rjwu$2)pZ|#4VdPV
zQanPrb3i{>A#?|iz8<~10*L6)@^LDao6<bL7dJLFSKIFIDUle~M;~v<x{x?FDA#42
zO15R0RaXzGTNU-Uwv5mJzQy#=hp&x+Su_CV0ej%r`r768dVRmUKxL_gR{0kTt95s0
z^+k&a-cal~dgi0B)h9kTHqFooxLEK$*fe6TZV>)F3BS}cofbWih+nf0cS~*gKU3TN
i-~Sa0@$YpokTRK6vemO>VWarZ)25nEiJoM={r>^*V3)E0

literal 170409
zcmd?R`9GFz+dg`Uq|8H-OqDdCD5(e;QYk7)<~hkc&$A*VNs@#lBvX=^P$Wr`ka;c{
zGF4P!ZAZP&v)=Xn1J(~~_3?b}`{~AYp2vCY`@T)b8G1_X2;&;=H53Yk@z_x%bqZyr
zJB31RLPv|g>AICXjepRZ96zFjf8x(i#WFwqW3~NJZ6^waH-!9;Dsp7zJcS}aIi@79
z;eL0b$3ySt$dcSr`@{>eiPxL9h0$*c4h-CIWl&xFR-E?MuE`I1YFiuM9{B#V?SN*T
zx>LLLPm9lIEM;DQzj@J|Do5bn75c!yhZA%RzpEJb%ug}<cCT8omS%mhhYUY6U!>R9
zlBw}$U*n9wj{7#bRO|{Bpj&a}zkgnh9TnB5+VtN)rF$5r*;xPcM|(T1d+2|^H;aS1
zT<O1mG{dM2{`+H_glj3g{_|6;Q86a|vH$(CS1RIF$N&3dW~p5LTmJiVPt^ba{Bh3b
zZr;3E)!WOblFAk@Y4gO{Md}0#4b7@miWwS%OTVXzoL$H&K4gE~sDcHQ2~c98seT`G
zFY&_H?vjsrrmOBIB%IdJpt-DI9OZd4HrBMO(B@;29sl6akg)_eD;L+5RebK0xlfHR
zssgCH^x8t%w~84*r=c8jc9sx2ndn{gNR>Pn{i^pinnBh6&z~!pnVS!GmpC0gaiS*A
zwC;Lz^k!RjUfzi5>1oN>dP!kc)~HEi|5&M>j{FPaCN*@Fs&+%ai1Z}i&Q0I7ifnm%
z-G>-&H-3t@oePm$q%!pRel~u?Tyk>qQ56-1O5Z>BBd^!KGJEp&(N3~eTs9sr(iRpL
z9vB^U@0z^Eb084+>KfDd_hW2~!)V8x{T7G#sVF~w{BR!(-72CU(r)0zNWX4p`FZ<w
zlxKf_=w9WQ<8yIwvDx9)`sgghf4n`{wC)-&Z_ZZo`?ZzbEb@Crt@E*;J2+|GGtMN)
z`f#say_(|h*5}i{DP#S!-(RJzyNgeIAGGb>e^<tfQ#plID0+2AM~Bv9qksi(b#*oa
z@0nW37+yX;rd_*sQL4ssE19C}Ka|95#TLF_mLvUCo66|jJ$d&}aTgvNzo4zDsp<dq
zef-_{cix?`Hk$aRuIC(#A2PmHF8eMn&NWf|0|QsMySqDonPH@**3;8Fb><8;rFg#A
zE6;D)_h_o9UxQYK=U&cT1|qW0+-EigGx2(Mnc^B*b1QyuQ~LX>gKo?Ecy4V?BpahN
zFtsbIXmixjni1(^&6K~p#O(OTpjatYW8>tHr}*}mUp>tI`uF_Qx}dV%aSVNa%RZUa
z`V_6F7OcvVe4?_pJg;4)BV~TvE-p8#zagC~^7pxve0^s=x#r)xF=5?Wyytt|24^#K
zpIJspWohLp=ovgVsutavbN#5GR;hEwf;+aD&Fk9nfq?+y7r|7N!NI}nNlASH-0b*$
zr_YTs*|g<U{o6f8{;@`q@ebccPm36obD6%p8W0{%ud!pT=l8M8{ryI7;`cEp`7IrI
z-_qicd6jR2q%}+IR_xRUQ|<_58d_R~3=Mn@es4FlJ0p{>8ZY#9?wum@-pt5qqeFUn
zcNZ4ww<eX_#8vWtb^Y=_wd*T2rAcnNym+zYq+G#8e$sYIon%rQJ7YeYqX|7>d{b>T
zw0qAU*0pOdWAhYU{5@5<{H()1S6M27DJLg~Lb<FEP+D4QUh+aGIw2uq>Cf*M){@i~
zM|16$|IGI}{~l?G9m+21T|p@>DH#}R&3tEBFgTZykwM46z`-9I6H{GZ|E2J=g!{CS
znVA`dBKvz{Px<0(Ls8FGadB2-V`B=pEyMkk6uUBJvhqD~i66{u!r#8CrvLXWg+B(h
z2glsJDX?wZ!$!st-2+aat`_sGUF*^nyqofEva6_rr6h#kmuuz9mD#rOw{LHgyh$Y>
zB~^M@vE+-#wwTwFE8VD!25UkG9W5;_<A2P|&##L$H8C-g;O=_LPd>5W<VmZ1*DZzG
z=<?Ro)a8;BDUq^lTLqo$?W=xFc11-;QwLR3@;eytRbdcWUVd=#N0*&evHeB~Zq>~5
zTLcBKG(=R;beg=qdtmULcuMn?j5F!?pFMllA-Q(#+D7v~`em-7fq{WKH)GICcK-hL
ztFVMgwW#sU8|xO?YuB!w+Qu<BR9;>#$sb8JbZ1U_fp!^NR~EWbVE~oEhC;7-TZf+u
zGehpP^;-ufCmAm*P=zY3z*m6|BPvqq@L}V|JA2*x16JvX3UA+jOpl3<&f=U5S!f1s
z2|9}X!1I;mGcPv`I+^FMXO>$Od;R(~h0>CKvd(a6HdeEtzOiv_tdF;Mr?>aXlP4WY
zvZ)LxT;e8|io0(Q29QgVI!9J$@81>jl%;jI(K>%VwfLa>bU#x}tFz;7H@A7$$xMwM
zth~I88BYg>iqQpl_S;h^(=&Dan<XVV4=FO`CrG>XI&afAZDP8YT%@48cXFWl>D8_;
z?YT+CZ+cls4|Y59e}0UH#?9izFX>6H7Os712bw$Y{4UIBbBy<ysa@CTc>B+<@mza*
z`{Il16T}Vg)tgz)74q@%4bC>sJ|@rgbltb7l2+$T<6|&BKAZZt2Uq<5S~fVl!LB?Y
zfO^#+p0WbE+m0POI?rmhB<8Blc^A}8ImK6Ca}GLXYGyxw6}8!}#IGf>`0UxUHnKSe
zx$9!P_N)%Zq>8n&vg#`Tw|(^gkE{MTJ@CR|x?lcdR~I$q$&)A3a?5jabc~Fbqodi;
z?WSiNc418lYX#XV+a-!SqdGj|KS^3ixQuCt=sX=Q{O#%G72cq_Rj(}J{{2l39{p7T
zHodtyIcj>EnXNXx|7(wM|J@@p{MfR#zrKChwtai`HD2lIDr)8{2Wlzzva%Qqy=PWp
z$T95Ow{N1SY(8y7qH*%9fq`$`_o#-E#^5Nbs>PpkT{mB3eW|h3*4DO>{nVOyu6E=H
zjXOO(eeFmZd%@p@{5*CH-RmXkUZ<PtD7)lS)s3;)$(X&WyxGywaT>j4;OFP0kP|}G
z6b%iH{-`5-)xnG#bVO4wG6$#0m~@R}wV%%n)x~FfVN7)1+?91PsK-d`smH6{=9D9~
z=l%SaCEK2QFcr!A{CeskN4k_{W#W7DNz1=I_`5g5wE7<P436I3-rM4)w9gi2kB@%$
z%+f1S_g>oN_w#sW1+RjlV(m!yx?R1^dF*U#({H7xu0+;FY*=;Ac%l2-mliwe;ji;A
zhTp$`Z|n7@vs33mdU|-|ix)4t^!k86;`!pqPXEehW&6K(`j=ZAYgS?sT|J$m96HX%
z&aP|S{xq+<bFS!&fx)$lEoq^Va(}*^sd@SGc~cWJ#r|ve0rd2RTjw8Y?6}G&qvKuf
z^ZQ2^-~7DHuFlS&qWkynYxhe22o}KYJ>jvk{NHWogOI$u{Gg-Ni;lpcpxTkqQ46}j
zcb}d#n(yKB9Nj3M`ut7Yo*C>N;G#`ZQaW$)3Vr{0@}LNBo&Pgw=f@r^_j8zC<LucU
zXI*)gJB_aw-|T66>M<CaR^hu<?XG)WzS!SM^f~Vqio%!_#Vbj%Dyx*Su`#k%_Z~bL
zF6<in?m5wsf5f|6qTOcmxpU_he*Icu@yahgHn#m!DaPK4BiDGSD8*%ELtS>2Yqvg}
z{{G#jjeR~jgwIm)bj_5^w|9!CcEnI9(e;-+<~~(hOBR(*rm4oWtXsEL)|Q*+S;EU-
zE2^ST%65GELzesDq^D(kLFT_kzk#d2!1#FV&cnF|6=H5*^-zYpwuy*vI0df`^bZSL
z16(1ne?Qmx^WY5yIz1wc3TFTa?iBNE*wB?GB`72`I9b%k$-j)|{J5yduCS5)<iP;o
zGVVi)ii?^hjhL{KcD*99O9LVNVyHH<0Rnmf_{bFYTbfI%o#gYH5XXcq%y>(gZarrh
z+Q2NkU=%;U8ISA!k1wGpg6IE`*>$?f^|4Aa%KG~H@hz+p1$KS12T^$0vav}?yFPvA
z-e=3Bc6Yx+!xMCu{_$~P$(r_D!@jKQKfiy`($EM93)52`TQqOVGN>qRwOJ`&b$nOF
z2a$JP!q@^=A|pc*67=H>D;<VtyL)=5C@IPjXjJO)KmO-B>TNVBT2IVY%Pvm8Fdu(o
zzUPEc^n%uo7zLr|0?(|Md^Oqs9<S@;Rk;!B85;ZjmJXUXCHACECL|=xH0-MEewxfc
zF)%OyjFO@V?A)2tT7DsdM!qUbJHIA))@Qz_J?n^#)pwgaUDg;Z`<+KmWp=3uvH=FJ
z)5<fcwP~{YG0}1K=urzNU{Ke|&MiADmzRG3r#Gc<Ez%AMzG<SIlGkC_(!cW^%k-Z=
zlB!$P%GUpzv8%kM*}}rZ3ECr5!ok7e)!Vnh2?@MFdSMPjF9+~|UOqS$EfCvar>m*1
zeq+D%%#-sC)AN<f6|Ix9OADsrpR?oRH{=*ptjX5&>YA+t3>Npht&{!o`*?eEncGyn
zt)Y%iSnji5Yk>QX=;`q=&{D6Urk?2!Vjj#$PYkTAT(*hqxG6w4+Mdgg{?-5K6WC^;
zWW`Q0hdzkQ<Td{KS~fLkxcsZ#rok$r!}{4pW`loMPfuO`;R||A)XkeT1v(K=p6o`g
zA3vAT!XxL)GcYhf@we$JH0~|;;LIV-leK5{-=BM!ZbZ=h^6$!JddkD>>{{UO3j>|b
zIkn;Is43+Y6*CJ%k-Ei_<N}=2|3b6-iGfrGW6qn8T;rH~P=Nt%^P1J1Q7!RcX!O>T
zD=cnD-G0$cswivS-$C%<j9Arix7u}v-c)fXf^Xbl^ZwCs;Z}}X2+HT@+-GV^+KITp
z?Pt=oik1J~<>|kj_4LUqp8p4<mHGcP+bu~IKBS+EsjT?>xF8#z9L2xGI=I?s^s1#c
zFMw5sMp$X7)bVq*n}!_^49^DVvZOTAMc11RaBBzaWGnml$c&ATi|VfpKb4^Y=uh#-
zB%yDAx%+|o_M0kO&`SmA0#h?vM%+~`s;r*(R0{3izu(z2?n8yk(?(BW%K{y}!rJlJ
z6GA}>$&c%;*5&FJ`ZQWDPEbGYkSte84aBuHTRe3>TY6U^*`EwkE(Y4$jpp?2563fF
zE#oWvrXG2ni#tIZq3qQDx^g|B*r_AChCyp|vN;dEh`lc=$>6_iz+?S=F<$bTrInS<
z4;R33G42RSuO(MkSCdoe_wPUdWXl#1Jhj5h%M0V8!%Q)>{Ss}VcRqu2<6hv+E~kCV
zNfVQk8)H;9(gmt6_3emJv5?@_(A4bI^ABpz4(1=R{w}MWtr^}ftn0HpIWb}VqwHkz
zsVh@{E-nR)D;Kt~t>)q3Id$&bh3KhDuu1C0k&zL%$^DKc-J7Tk(k#gE`}gPnb>+kH
z_&I+}9T_~QQ(9VA1_w<>ouzKZ#3Vm@v}W(#y$>@p74d{m2(jaFn(bbXEr5kEwzFeC
za^y&HadA~^>)OP`M0;oFOG=yazHR?o->l9058B6*TP;QPBg%q-*B>-nNImsyF}iX>
zXx!eTO=V}S;-;vN6F-GH);MUVs>=&SYd@B}m3nHM^$#b3m?bxXD`@xWw?yeysHmve
ze|f(}uflUJ&={Jc^08yb(s>nAnyc#SR{8q+PS4G~xV};GUHZu+*OfjS<bGQga0d!F
z>-1zDZfWw*HpF#jG$zXBuKru#`n*!kbRkT<&*he8n7>W+^}TN`5yp^}Pgs8`s@^Pt
z??WDE1eJiU%yJsXn#1|=Ub>s2j-u&qjP3f=ezE0{#f8OP9BbEZ+_Gim_U+q`s;b)T
zz1<ZTkb(k%MzU$oo-=x%hlekL?rszmq}jM}<0#N${A=~h)&$L6R<%|ua##D?7_$Ez
z^j-(Neuxy>_N+l`ssvtvQlcMeo!M!w^!%mE)A@K*1V}KY%1G=UK+>K^=Utxahyr^#
zOC_Xcny`ZTPGo4FJ$o6SP2c`%cXxO9z4r0bXU>>@d8BnIKR=(Vd#9+VjT^J3Jj?aC
zxX?4{Y8ww6SbyNafue2$q3G(K9$w75(Vnt>K;Zp@gDY08S_OO#ETxcq>X1;hJPXb3
z6GAAq4<A0<!)Cya-6kq}*ug=3&6+jf@yrzleG^>!Y`d!lUW7O-OdC_G{Hdr}I5~qK
zKNf?Q@$_Ld3Kjq#)XHdgT&K&D3{~!GM%)AT5dKaC9b+`iN<r7pdL!BUEBw2=^{hWW
zrUD~F^<r#HfBQD^;dFA(N}?}*7|GEu3ufP{CLk-jl(z01xSpY*Ap_q*E^LC)zDhaJ
z#m;WsPHV{*781|#)!4{*NKs4^-K#=$rTigWH-=-_n>Qx_a)w^V2sf@MAUo>y>*p9d
zA=N>v03}uoy^JtDsrt7R@7q4Qe}4tP-(ud!<gH?2EV<>gDv=36mx;VN#@5=}y5Dtz
zi-}+MDp3eZBVuFMVO)xNFW58iN_|U{*Z8NL4S)U|&YlCVeki>q`N<Qun3x!}o_*Qj
z=xkUU9$wyov9UF7Zf;mHqYn>H<DP3`gwNUkn9#l}<+7%>wsz1rc9-ExC<=K#zg<t<
z7NcrOQw>T=GORP;`qwhXBxr-EtlF|zr$&H{X4-V^zK?zglbL^mveqT{m#i29wApU?
z&IQDYj-H-9D6{f{7}xcfn9EoWy%NWjkhWemG>pHyWDePE+m0Oyr%#^_|18M18jC7s
z_KM}roxSvGYHCMMo~(mdcuGfS$5(#ne*olHc@D6hl>I5_?d|=wSeDcn9+g9&fKkIO
z(PHnPyZwVTO#J?g8xBCl8MxD<PjubW>3iZE_q&X7_%8jtfTvO9-rLebkT{o|??KP`
zuRRk7uz26TIX_5C+axNQefagIE|4?uT7RAFV7;9o%@#`f`n>4n$?fB{m3QuNiysw%
z+O>KP4a!-HclYky_cV5FjBS825CQO1-O;fgf1oz9jepp`Q4f-%nAZ>MZl@+08P|`0
zB^K7xZiQ}B1_ZAhdf~k>w{l4ew=SZWSr+;Cc}qb5tBd5Xi<DcsvbeZt-n5-#&85qi
zH}2T6MlqP-A}5Wgh=`J-<8DyEpl{#4>872da0}_q&dkUs$@=tv`LcTJ)~%$D%FB5m
z%3!S(jg2?HZfMY*u)1(zm9(^UacL>^5kA@=c?L>5zA-H=t%K$?>fp<lFIU3DgOpPO
zii-9LsokZ+hQ4%(24jkf0*MOPa}6m1r_XN>A3AjCh=Iz#O<@_bu;4z?S#aEY5%;-K
zOzd6Y-Z81as~-|~LI~{4q$yFR(XnlXKblP1$)sz?qJ=_Vn9#@{($wSxs{@w1ge8c-
zH90zZWusy+ku$mW+Y3O(3_-_vkea$tMur!yR6$McEpMvz^@b6Ay|a*LY-?|CPeH?g
zLh$h6!$U4Edmm+HQc-~Rk0>atK>aWN`9oCV?qUbUgCPyEyG)_U(XC!xO~?u1x@87R
zN^Q#NEyhcWKUrv~FvoR^Yzv%0&#t@*hG5n|*Ce-&zv;F(H9%jXQA8;xX%KEv%@!O^
z6=RrQw{G2k^ho6dJK7Gu1TeK@8<J0LyVq=Cgu#VAht+;wTf6e>*RS>t4qrGGmsN$>
zv7JtzI~Vfk(T=+ZJ=nZ{Ok8ArFxisM46a5x+QU}MS7y7-&Ywq9r<$3WDRvlA&K&9L
z=)l^t%y&C66Qqu-ud<LBe14f8RG=!>(6{5ho4pl!_Q8V(-<7%9k%5WL$i&2ig*uhl
z3Z3hEy&3ND`}gl?3lDR0l+T`Bi+yogT|IEaK3f4fIeyZ2g@uFE(iK%z88L_PU!$&H
z_phl@!d;l4qJg<griP$hZN^lBWW0;dV>l{ajq8Y#*(4ZD*jnLJhK9_PNtH}RQk<vJ
zQy?Z6xbJHF*zxif0J74DW6&(HnWCbi0<yA1^vYZrFJHdQdxs%wJMbU$G-!(Jz}gQl
zFZ~(#{JH)9{KX^KfE#!1Vm^HM@V1>hmC;?%(obn>3jHv%IbnF}^yvU7+Su_Fi&tgz
zM5;$)Q_|L6?>g}@00L!rwM~TqbB4wmd|^8-mEeIiHJzNu7@wG6z&sqPiyWFAX+Q^g
z!nNNb0?Y<g*Z=wRDt>-`axu2Hwh3%3>u9JT$eznJIE+1eBRbk)cK8^kb0{NSf?5a`
z2oLPr*Pac4m=oFMpjM!tD^ALM=XiN_eT!(|HzlQnFQd&V#r6X$%p31quJD@IDV4$!
znJIqrl+n!AxeSI)IfQO8J(iiPup>sN*jzsx?clz&Al-^5`S}iu3l4&Uf&!a24;7Y}
zN@M5@qB@G9LF9kOXq}#)52>t_Td`uryUcUBUAMN-tyn`(PuAF~J%^~muj=dN&z~Q$
zna|G7KCGt31Y&$gUt%veBa|y@z$@`>I@ktSS95vWe_Cs*`s#J-*6lxVKt5YD1-%3+
zMis>8)0&#Wj~+cTkG{9}hrQ-D4tqDZx_k0LEG#Vk_wETS{QOys&AvM>D|^?4_+x@>
zWcLHMVI#YLYd&(?&@d7bgE88{0@Q+2FplnREKv0ImBqe=KmZK?mvUn})<CVAfnD*U
zwY8vS{Fsv2YSg$Eq}(*MgfJ*S{|r6-mu9<-F-QT{Xes-prH2~#^<7d)ePZ8O%0~)y
z@@9N|cyaOmn%Y|V^p?<zoH{Ki4R;TZ>Xw!>CyM$<M?Z?F`fKdC{1^Ev0ewQHuIbKA
zO-+TKCy&O7Wvn&a692lX=?E^<#MG3E0+o|X)_cG5Rjz=lszV?#bb<bN@A8ldvUl%V
zG||CfBO5#cmE*^WHHK}6Ht=f15!Yw>ZL}Fq)}iy~w_tJ*sD%FJAT7@VM<Q7>bKuh_
zA=-eI^8WQ^3dfFx-hOt!fUgO%8`=_-Y65%}l35I+U(Vf9Nws%zIXu-{G4S;(6Eib2
zMtT(n8{lz*XH;y%2$!Ta7Zku?BCPE9`6WTN82WZ{VIgst@C6kW3kn)VE(g<Or$PCZ
z^hD1!*fSH2$CWhgCOZqzGTgtU#_9|k5wEE<zR<4k+Rd9B6voX4>vkDf=O<hwMye$>
zWO=g25Qv0@kL%@X?5I5-p&SG#@Vu{&{~u)XOaH8&{pgW`tE(izR;YYn*!%bH-RqEy
z2xw9aVUo{nSHwocT*CdXxjdM5UYKJIbi-9XzrL|Lwf{hE0TmD}1^-G9L<Uin;;)%m
zyQk@35QJss(9qD21y<|S&zzyU*HL4FB|~8!QBtD$1wR_!4^0dUun}@NzLT`X9Wf7c
zbJs0%dyPq9lM6*BmzOWzl6lZ*o_||&aIiL9G||n$;qOcx8lon>V7#4KR#ujmX%S)!
zRR4`o!#+19)qPI#y9U6E{Z(xwMzNriH3Z55+avJmQ)4_CZV&}n-@bjzD<#Vsm3sG|
z1-MWrw?tK-WAkN<1)e73&yiTeX^2BtA`Lx8)-rI3RR!*SB0{`x+vJB2=~glGDOp=*
z-S#TXKoM@nN52SWd;t*r)iaSFWYFKALPJAyWgK;VJu&eb0L@)#HzwFX@4h&1+qKKC
zEnjO>6!Cd??%Zkb<P?NvLCeg?ASWlc@ar^7_Kp8!0e(Gy@gn8$pIpPrYmM>yNM9V3
zLXZ5?lD--E<MZ2u7tqQsy??JRe!Igx3&#GcXdzVvd;7;+T5p@0>Y_Fuv4^n+P8ZrG
zN&B3amzR~BI}AQ9c0#4L4<<fFG)TqmrR14h@yEJ4I-V`PA0mG17uzRm%izzz_&7Z-
zih*Z83pFi+1!u8cByJ7L7%jNi0vv~^TepHC%ztQWLzfKmxKp4bijg0MO2P<S53!~^
zWeCQrij9rvtzCxK00kig|GM`2Rbyi<?n2i`O(@zJeiU!Upy*y+dE@ZvFE4K@N2GXL
zA8RSRtg)k@W8M&aakymP<itIY#IJMAi}QX}oPEn_U%Jtvh8?>YLVF&yL>7>;LPhNK
zO-%9}+m2q|?fm;C$K?p#sIQ+s_o=5y*vz$@P;%Y?!22=(0y9ysB^CiWVP)H2<(1}y
z*a(w$P|5Tb+7KNkAK+;UU^{mNEf8R!0GQYc{0mzs4gLmW{BeFhyHls9hllaP^Z;aq
z#@7;D3?D1J_WhXZ+ohIBO9==FxGm+fLoG=*qNvYz16$~Wj+Jlf>sR9mJs+-*uGOmo
zc!xQyXl*S59bV^yVe%;f2?-8%p%YawP3-O*Ikze5P-d$_X6t>szGvCCrn}?rwR|B|
z9V^bk$%%TU{QUWI+qP{Mtcz=D<WD%T^7Av_OKu*A#`@zWd@uis3ZD(Vq04Q1@Ow8a
zUnLVG4`7#rg9Eir!RN2JyaId%=>7<rD<o9KH+K&JjLj%TT3EA2*(^-tH=WYe4c`YJ
z|5^hn_g=553eBW585%GiUh>BFXhyVi>b541>z)*hZLs<Er)b1!6}R$W&zHzwZ8_q9
z;zY+MqHeV*rK)SYT<gRzG{JZn^cnh)ab^aIfPu7tv*-j6DmM!W(ZbyI&&{3E|FhU<
z={Q{$N(QbFrfr_nXEjiwM;)F(z%WQmU*Fu0j;dAkDnSP$3{Y?oEBqq$cx=sg+z@P9
zG6u1Rq#L6f&V6oD#JK6wFL!4}r>t52)3;381y=fyy}j5@{nAUQu5k7cD#rRTVEsr;
zY2%(!m&R)=C1`x%iilf&*ftY84RMFq`N&yh(^dihhF!TrcXj=qv=(3an0hnt{dL5&
zuBxgsZvwKx6WX+I-#YBK!AmZxP3#s&zZ2etUQ2Xd>?2eZdC;spJXhhs6=tLts%O-o
zw#0lF-K{&GtVB0Ea_rb@aBcx85z+O)Si!(0@Mi=>MH$06MJ-zdJ(QGg0rUogkCp!#
z%d~fNL}rAbWo2kN<<OxauCAmIzT-X;bI#V&FQ<Jx&-<sN_xILwV0+JD;fKND=Hre?
z6VB|}A{Z!o?bu@n-UNywbOJz!LyAr1Z9+azR-%EgxqM%{>dh_4Uo*$^AR4xQTiN#D
z!2=p;x344{a_9TUv)C59&#wWe=(X%`Y;4SK@S28BF@Q`+Er`{Qv;!7`QAY_lLD5|b
z4wn3VUWbK7uhe-pU^Q^_ArB9ad><fk2;#`AfFJL8R*8v)!hF@#WLH?BkX&_h`<dzK
z7tE%R90(==8zdU==;)~VMPBX$($a*>V`>rwAuINQA^?Q_WanNY(QMvK4K65Z(4Gb+
z0$Yg{f&~N&A}H0@AG>RxyEjmvA1>#|LYoa}Vzs@syW>urut`?W=%1R3BsGi@Di;3i
zxo`T;J5rcCWBX?FTcTZR_Xf!MIx}O8?<${Ofe22X@1I}h`!pc9kX66eYB>WS)Lob*
zd@iRSPdr<8YjpkeXiB83PKDpel;%UIa~7KLYtQpIO7xgwR8~S?v3_t(^twu_$!{Rp
zQ27<Q_d#o;>ml=q>c{MCU+7rH-6>jhr!P4j^%2xzUHI#Fk#w_!XppGy^4@D>PurS&
zu`;(rv;)EODU5T^MD$C)rKOw(q?~R(!ao2fh;8aT3>&nJU2lb*_DMW@s~-Wi`&HU{
z2HHia@9;AX^TTm7ywLt-yHWN%<_qc<FK$PFizLha(>5S@bqx&FtFK<YiZ{rDW&nCO
zfTbe9<>JMQ`+XPplV=0?NFp#MC14Ebnn58Us~{W4kEFwEc>eTa3k$LqSTOmWv75JU
zT}5^iYW%dme$QYr21pHhB{<`i_%x%(91+SbBDr}2bf93gNQ6Y)xzl<7vlEEZH9ncJ
zlp~SP-@au+;HMT*3}!pKsdzq!jRL!Ov%LsmuDg)rLksW+pfWr%0(7BgZI*<D4}oMe
zomRO6d3Ai_oHDhcSML1!6mNG5j^<inP3Hw^U`=N1!F@i@QWDr~Q^oaJDe(-B+hbJX
z8K+;HJu#;;(DhzURo@P^Q8lp)bwXXPoYE|0NFAecK0nC^ouIe<Lu;$H^|p#s3C3xi
zu1H;oB!7N=AAEHdP~kHDy54VPBA|QV!*wKM1=H;9kv&eIP7u1YWy=<~>>S6EyQCTB
zeAv&bdGTUKv0Cz}estnmxS(aK2Bok!;C?Y*iHNXiR+JDBAaRBxsHd_O>Ch1{h`1yy
zSnzxZRzpkz0gj+*f!=QhkX8Zj0OaWC=&H&+=13N7TZ{@<gffq`D?{btEdARgxxP)K
z6Jujx5DARmKR5=QJGnsLd>FqC+yD*s8k!&wH-aGO1p<&6V|VJa^{qnN{rFLfs49TU
zSFT=FGBX=iZEA-!XIJ@0lI$T~Y1gXu_H`IP#PDjixCCtn?e5l2{V-hqP{n+2{@9|H
z(uX<0ty`%rZfE?O=cc|4knBDjb%a8J>Lm|OT9)NYnC%=1Bq5_l^Ui)(nf-tQ5g|)$
zl$B{UbM!(VJP?EyrwB_hrP-)0OPdTePwTDRKy$A?p?MA>!=(&MG5<S+DEEoHx>8lK
zC&&vd|7`MO;pOcSc`O`tRLpfk>)=9v5Io+Bq1j!A&#s=7`@?}9O|&V4+z?O%!2i?Q
z+S`{NWj0&%W5r;GkU|323&5=*T=C7*C!hX-(x&PwtedgvTSdLbdshPs+u7N<dwC)H
zwEiJ5InX~F8yk@U#qSX@0D&1}C}UhQo)RQ92Fa)RFg4KDVD&2gO8}0N)*q2AnV%;U
zoNydnU3#Q$_<a|sP=n-NOI4-lH{Dm>cl=mO+h-Eq?A(N5`}@ZvJAdW3_s6-UoGmiD
z5&?rTf2&HIKG*c;R8+{I>YwrPcCRFyV#9k0X3G|19x}1=n`Fb~gJ`uMyDq0<mcZ-9
zIJ>8vs|8#Nh0x^Fn>bA0%Ar}baTudP%Zs1oAaKCDR)SQ3JtGcPaprz|`bmDcLB#we
z!lSV83+S9PkatkARn5(;#94544bIFIetC^IygqvC{KCOF6c-*5?%)tg{>F_PEd2bD
zNdHkR7Vd5l6550W74<5n8bxNAOV_Vo=Zpx2v^(^IS*`{`4q4Bc0+DUo{MFLgorIbo
zYYJ;-Rn>&Di#n7lDK%#4m;Fd<I)CB95UlRYxPPPxyLM)+#piujBD|t<`H$Vq4DtZt
z#?LQl7Da@HR^JlUy^h%nxdTla0=g;KyH>tALTIlG%^9Y9q5AHg=&jg*;Rf7KYLF;O
z=$6uBd9J|aFuZ@=>Jrf{WLR2Sa??4#DL$>`%j5G6oOcg6DeI-HXH3_ypELs4=^1S3
zHQP<dQkm-{vIQ|GgcQxp1o?ay2AH-6?Gh0ocSwi^fFZOsYOu+mRm_}oG!c9<-1+8>
z3aFH*q@+j?J7erpAV!w0_gQJEATXKSa~u1R0cp4_vwo~>Ga3@uJjt^hJ$9@bGdm0|
z_)4@U=BEkV@6oOzqA{_O*-ifq>Pq>~Pw^yo>bp2|1UeZK5=32tE^`?Larp2`10+5d
zznt_d9bw1=RB2oIQ!}+4>~-uRz*LrQks6Qq0W=-)l^SQz6x(z32{uP(KYi+yz^Cen
zu(0QM_F7rrw|P#P;yrr&_zPXTidJidKs@w*NNpr@Bl~MK#r5+E4w`{N#SR+z;*?)(
z9i=FF>)!J9=oW*eAVGtd*BOjTe&O1wSL}~_;w~vE8I%fD3WP#TZk2Z|k6yytkMum<
zLTh=D;;^tV$cO#tZqW7Mh#0@QE$+TB5JLJLBv|XNLPiu9h5Nk(B!3Ri3O;l(XnAOL
zU;<F($dWw!`Dp_Kzbp^nsWMO|Or?-9Ya1e>ViIwJ{MG_|kR2b+PJ?}CiitqXj8AV9
z!Xe!_$N}rt@%#O)IZW#{a9*J#Gi;D#AgUJ>3kZUeHl3G{@iFOma=yoW@NZ6N)BU%a
zz=<#t!ZzCxj0&=wclchL+3o?*S)~&v7|7~sY0*L6gZO+R@!)!(-AT@blF}8_<&zrP
z+MxvLz=?7C^rjlgU%($iIgpfN0eT`?)L!4;lhQl<ssmRR17J#*v?&$V8;Kc>w`E@m
z4P|Wy;2ns&kSLEHT(qz@x44)KA!Ms4X7`Vejc@O>GYD~<kbFP>eF16`=^lW|&&_tr
zC!aC`q62<?4q_o@)h2>K6Y@MPsGaP$kGPRshFLEFXAM_|{3q&mD>gjSy^UAA5i=s}
z8qERD5g9do&wQl4nc8Xq1~3-*NzuVQ>c?s-85r>S2Ok^P%+d<LB9m|_GE{@U2woC9
zOf%PjF+pt=BA)0RDS)=vkU;`;SFT-qdFG_dWf*#Wj-NdK$?R4dkc`h}Y9`~`DA4&s
zF~|!?50SqmbmqVTZHBLpG<KjUPTbwkig6X7XDe-dvg<&L+AMlf73vvCwdvs8)nej-
zz%+S*2@JO;1eTxln8O9(N)eVG8xTCLN?#eEyjmA3Dk{W~<%>oOti-q;L)8>#w%rV-
zN^S$tbo3_W8#oZjFA2!V=pVcWm&x15M;vql`p5u`FOtZfotx90n3|ZlN}4qWy6CeO
z+>jXHhIqsa@WWbz_b8hM<k!gFG=<8HC9MJAGP-b~eZc7nClCdi;%zbG6-f0y1~nrc
zheG<RUa>vZUlrQt+i8anb&x}*0aA>S#4dK7<fc?%_=B76hxduSOIQ~XVy4RHm~h7;
zt{CK%pNEEmFuAJVw-R4m_`GEN4D!2VNL2Zq>HOnMW9s2?fXwu(R~d+z_cCGw(b6+C
z=$M)7&?#X4h&rs{d_l331}1^5yMlsx>KgQb0EQ$lKZf25fXf+zy<OIZ(j9^=A-lZj
zB+iUSRg1PejZlnhIR9zrRx^-*K(P<$rE6x2!!$aPBwKOD_q;^h38H%vso}@V7?srN
zA3u~eG&mqq6SrE#@LBocU+XXn(RRr4(ydujYGW|-%;z_0LRFxP_G2Hm!-N?`*wCaT
zE-voXn>RIW*}6oTdG+cj3Oq~q+r6)A0L+LIfNB5+i2w*lE-mGT06{W5-rffjy}ll}
zbm<a>qNlGvgIR5Y_CZ$GKiHWUn_UUt0x28ZaT@<|6Be*Z_IH-*pWT1+{Yo+|$nrY;
z`gW1xueY=A{^4+J;126}68|!M#sd~c6akFuiHQj;DcM1oRos>ZoouZ|+j?0G6}7eF
zxWr&`U37_u4q$|e+wU<e2#r_&jkV+umbwmiT;qbr?LJx1E{3=>Jpda90_loqv{@a$
zKmiZV)Uad9f1MVG6oo6sZjpgUr>~cbC3hS?y0EZ7s4jFaE3}GdKH?68O2h|6pPiH~
z1Q{f~!u$;v$GUYxAa45|hp*fYX@eM#_ORb!kOphR%Ew0#Dm3PvpR09QOZ2Wmb~0UD
zm?2UP6R#AFJPS<CFqF0__OUxcLI^KF`^|HoF(D+JMEu@O3%g+4g3GK%cYwhFwSZ6+
zQ1H@}`K!Pq@LQ$#yH5)+@g3YIyVR1jM1u&Yc^8?QBd=pH{4u-W;}9<oSB>eA=QyGQ
z#Ng)ATDFqby(?LK`3Qo3ww)>2;-?(&BhpeY(SjWMtEh&Dhlz3YKWb7iGr#Sgik>T9
zm4JX`e-1~R2Q;9OiD*5hK$GIk(pG%;9pPwDSl#^4UJ|52n>8JMmz>~u7s>-UH-LQD
z1|+~J_aN><1tlY$fmShmN(jEg)eZZY$nJsjAvNihhDmc<#!He+e8}m8psv8!94$%&
zG$UnRg`q%!Ja{Q2#B}0go+Rso50(Yw;DZ>R6Bh3<k1w7~$cGOaZ<6HrVLHI^S{UMA
zro$)A90><lA;&e4rhpoF7==w(0a7T72yo_?c-H~g!;$#|r7q8=OLAsrC{pek_>p86
zAm$#@6p7LZ&;yk9v7-YMgs5i#b&o@T*T)<C7x6-UrS#vbXyl$d&W+^IiL@>zo23f0
z4u=wGxWMk1SJg7}-9dx{+PvO48|k-5jmycFokwig2p;6*qn7M#9BW|SIqP_|DS<-#
z4X^BDr0GOnBBfa?YbUV80@}&-*x0Q?21(LxJC~+I<hsA}D}y1Sd}??eRi4w(xJ1}O
zeSJNJd+7pf6*A<t^UYmezx#z%u<j{cEA2XQ2*3kw9KLAep@Z$BW#D>e?G>YPnN%2B
z2~y}vWtHi7Ebq(?n#){=E=4x?*q9{&F5=?i*kkNTO>f`6!~hSUdiI;Cy1JT-LoKb_
ztqhqL7q3E(C^+ioAPmu&EFW4x2-10w$>BK@@r`t+X=p|A85$;-?nE&^XXvx?p0UX@
zTtYuYc;f4mRlj}v3bFRx$FqxSG?;H0jknWK5UT3=a_10vXl<MK>|w|GCAO(lq7SU%
z_eP=%SGa)CV#L3+amxMs3JhPbaqnFN+B?PRG^xb~`0?)J+=O`m52bO%7W4d)#sGHo
zZ8`8TLVh4cY>YaJ<1s<-_sB*j^7o^&nbU~4_7!Gf;|zcLRQO8|-76&UK;_n`qf!M2
zt<1(hGi$jzR>3GCKpQv>@C-A9oOr?XXg}{_F}YlYo&}8c9Bml%NtjlSm`S+xg2Q~t
z(#X{W5s8cl6JSS>U;$w-T(|&ir?=RA4B!yq4Rip~jJ~Q&e{)@?M}NZqp*4SVI{?WR
zG<!n0u-_=$o>E8!a3itA8T5`68rWm_`ZHxI(m&-;9E94E7>&z#TWG)(xC~kIh2MS4
zOrVI^^sDfT7lxxx;=rDxlM^co3kB<g@*<8b9N?*nc|J}NEq|Qxx|e@4SJhp$e=k-4
zB>-<wWJ!F!R)J-VJO2g_j!T~tz0}J6IBBThc2|GxDI>S-YJU+=1gQ=ZI|fY<_2BLy
znDE`lmK<?rJ1Ychi@Osy+{FMb(z;-07fB=ma?}YL7r`wE!F16EKT!XoYA5E=a#*IG
z&<7+$lol(UpvQtw(j&JM-X^?nUy=4>CJU}xa(+wRfF>@*=kn^!o+rw9P2+e28CLM1
z#o|toCk00gQ3~*vTKe8gtJy)rKXUGg)VYC31l*jEtx4W&uvMUZ)S@+<(${CaH{ew$
z9+In^QVq<)f&yXPto-fO>RR-|JnN3#=u^~v-gAtw$I<ii5IYvpD`tT=RR?z~Sz||@
z>!c1chS*ABQ*%h8A`TPq^r?h(=hM}=DEFnGpW`}1{&xJ*REV({mwG&+Q{&^|m5U?0
zOb#{xXZI~H?SnIkgo?w@+4GQoP)|6cv}xNmI)3kIDv(6e`j%12(nkH+8(WUo2sENT
z?Aw2QA^4thNM62LwWxXgJB2&2I-xbZ4p}z{ThJ%iNLYVEe-Ca;G|BGn^r%|B*=`5A
zN7Y6w%j<0-G&3(NB!k@MJD(MS&~Zi49w+JSe!e(zft`j5Dd3k_-WnG~R;odqkmBnb
zXma<$Atg9{DJXxy>QziURc{ldYYhdc1Vls_02oB>-o0C+*E`wz(aH%Mf{e^H^z<Yv
zfg+|bDh|Q12JjKP_+Ai<aFd{*62n+qTU%I{s}KrAydNg^`y7sy5y(JHLkTg39<bl(
z69aJL^}BaPo(02zARh|#8|jQl?5oZATui#2O&>E)ZR;N%rb$q1pE&YNTKxQL4uFAv
zgyrQSL1KLM;^Z0$nY%B})QKZ{MGjGYEOyuoWSOH^vI3}>Vf}6zz$+pWTUzeGk3h1<
zDYZb!R4latJGVuF4qRY(W7?%VIGfkDUxFOw;u+`mn!7WbR7_7ywNVwkdP0`&4Lth=
z376x@9SjnZR6{E#kqu?#;M=rkQ`H9m$Exu#-@&Pq<eyXfdNw<BWa{d&AHZy~`<}iw
z<*9^;7FJ{zDE^Q`yCo`ZheBGKnsU!X%GmcG65O&S`RUWAXBxIVkp#pu^!>R7@iLK|
z4brY62EISfhcgW|H8i{cTy7SLECj6oi~fRxdr!w<35nE-RlUY9cP;-y(|v>k0tF7`
zwj&G|7SIDkKNNG$pc=Ay%~OW9H1Pwz78ct84b>8)E+Z%e=A_0D2G%l!>Av54K}>?1
zQ9(gL%=4QC&B~QM-w&Ce!y&P{_o>J8F183cyst{~{YB105?IFq$O>tn7tAjRbRUL1
zVna22-RQsowKVToEOeh*L^Zw@JnmeirC3H%JP)OJ$2Wd_dNJC4%KE~Ekjka+xhCmR
zuV1U+@CqBA*2AnUC3Ew%{kxk%-iW8V{M&Cix6xrAT%tbT-;A&wwVrvGoPDrd#@-`T
zWyOf8fT}e{rVXbu#;ixkY0h^hg_YMV@1Fl-9?iwt5tvpvxcANhCm{@7tGwh<>~x~+
zP2+h=j2B2leK`2%Tg#+cI6Axir#I`t0%(c#g3JxL5+njfx<GP-17S1BJP_($fUXf<
z`Kc+%uVJyj?d99IYIpC#PFG?0VE_t*55N(xLFmu}Kzf5L3Gd#WLRjwo81)@sw1siQ
z<*=SIH&YZ8#UBsMr2fVht#Ws9toma}mk>pmkO_j{5rQ5^IfTQFods4Vilfl9>(E-<
z;**>a+r=M|<194Zf4;XP-L~SL*snhk`+EH&Bboh!nPM<gkI2hY5&JnNX64@_i&F4y
zh*)7)@tqSoeV21N`D_p(fG>28Dw1!*A>O`%xqQJ)rfEsqGo?Fz?S$|NW{5NyJpf7-
z4vRHUctF}Eo8ZukteJ#V5F`d3W?reMEw8sc5khQOtHf~&xE;;Cf#sQw3rV$b`b0I;
zArBK~fmUt8d<MOteEK;(f?pTUtjB7d%h9Wj+j9Xu;VQHUDvRQ2a>5Z2Bdi03+tC84
zGQvZ_w1VM~!OIE(cl|M{mkG)bs|Of|Bd7uo9{yzcLMgy4>S^Ql?W;kdpm^0jJe|VA
z#%5BmdLAl0uwxL+h5?+>x(rp;)^;Z`N%r-5vjZ(@;TR(XVro2X%xyesMHU*sKa!Kg
zHB!A3lPGt1F1hK+)g0R*awgtoA0d>)+4>9mFToq9kg+m@lYS(mgNPDJ1vBhnMn(X_
z5lAaf!&}%SCN^DISC5J&Cn|5;cyH}9bhIw<l_hn#>Rci?L~+A4X-FUwy;6p@5B`${
zsMwH~mp45-`y7YI)}*S30`Pc4l0s^t4pAoE#^?>;iHsX0aWs?;NFcHJ8Lj|&p#s<`
zsY?U*f`WphYHBaBov0{4<0tr3B{Se^6X^kx1(n3&>-pEhx#<0GHhsT%$ExU}(cXgx
zd4OEM*?2|bFceW)(D<MXUX03riku&A4k_`Vl$6bI)PM_-v?GZ}m{`cblrQ}npL~7b
z1G@y{kG2mVh&6zy2Y#3}Xe%B%wgN)bwSZjBaGG$$-IQct=jQTDjoh`gxF8hgS+?$g
zi%Cw<;poXa1Fy-w__lq=+povR3t!v0ezzGNN$p^RL(0S1r2b{oskcL>rSo->mFB9u
z*Vmh|hKeY>`xqqf<5s6it7e!+QKruEi#CU+L<@?#k6*NN2=6?*J@r<uj*jKJ<0Xt6
zpQs2%o!uRG%>3<C<j)a3Q{n81q|ZYGGe?CkyuHgFcOa?hce&iuKHxS%9KXHUbd+UB
z(tGZrHGw$hAaHXnsG>2>J<rZ#$VQ#3BuB4sh7so_lyr2sv2F-&;rOg#%#O1U&!iKZ
z3WsHfqjIC-;~NTXx)uO^CVr)8<QgO(l{|kk_M`~2%9biG1b=>qZRq<GUg<cpMkRGN
zps7G78*FSA-~T$&1^?Nd=^GW8ZrolgP6UKaA&LMUH-G)S*YX3Q4b0v-t(LVWd9OO$
z4+|b66+gSXd(~<tCUYpuet&3alJ6cnd6K+90A)kf-``&aoHbnLy+w5U_E3oO>}%Jq
z{CMjH$_h`Q;E4HTT52lxel?l}WEgLwJv;L|PMDhuk-=qXh{<A}9|x+5SnzP*oTzT0
z8S)V$g>7G3?zeog=qeNi<*2ombeb7F42cJAtRI+T))hKmO<mm_lraR-6>uM@=VUbB
zk+iJCMn(dA%-{6GdDtu~YhbHFqSjK*qXcl1lV6<SWRM|j2gmDrP&s%8POa^RVuLL~
z&5fKuXSyKWh_VOYf1cDzgaFKH!`9Bu!{i6J;-m1$c;1BDXOyRf{W*ZD#J)o+Qk;w7
zLzZ?ljyuZi-fxwsMJPTAq{h=rNlUj>`uR!WWL_Hg56%$v)zfV_)0U$-Qjz$Sp&Bh`
zX>6lRlyU?&*~NIgFA<=_GN<E!nDb<34DMO(@z2f_gzS-vd5iye0I6so1Csj$*vDnR
zLZdqZRi{C}ElyzZEb8S=qD(xnXw|I>ZJqU3G&5r$aN#HzXIWD&j=Ihws@S`)BvvpK
z>94nmG7V^{!xKYZa4Id^b8!BDBd{!Vkd8^k8-n0XwEnP{UtQmb%Q@@K*R7b@%EryD
z0&sx!eE4wNfqMoY7TK&j`VXsiez4q&29XYDK*FZ;=^Ml8o}Qkc2uxK&IU<zsh9Jol
z%+Ah^_Et!PdG5qf+slhZ>sVNbsslDN0@)jRN%-hZfh*~k?EMAlo-f~=cgQ4_L#J>t
z?AkT+nZa7eT`@?;fz>j7RFNc*`|*}DJta=70F0qp&rDAX2nhHC&%Y~m84vlo3{Zk;
zNd)<y;;HJ#zz4tdRW7s7*KW_{cwe!7^%j`)vX1u_i(&R-;x8#eh)#Nb`Yhk!>L+r?
zQ`PHx7#q#!d>R^r=`quALQmIKI}T4CL2{#U$BVn;633RJqoc8oOo$QsPb$BH7tQ~%
z`*Cqkd;xu-^%BNLH1cyM65MF!Sk)ZE%C|V~PNGP7T+H1``T2WP??{jXA;)ekK!}?t
zY_&j4h}O2Jbjzx1SjF|yuGVq726kdxr!sBk*AS7*@RshyXhGsyqCWP!edU1)<1;G$
z>*r5aHa1cj8L+BM`?6XrBoODk5fwF4^=2YH`|;!JIN}$E=JYBO4=WP-n<G3XcvM1i
zd|X@*ZW!=jkrM%8$}p&qT60<9?b4sjn!VM^U;vt#^>5yYOe$_)TkE4rQJECF%2jkL
zu~N$1hXW@RZmOi}%{_b4Y*_#$1hV_ilf}aH4n0pI4kIy_dbe)Bl+^3$M2yU<%BI`-
zPuK6^V$c%dguSow`;^LEgWTA2ht18+dy9~CKZ6jXySx48w?v3OKD;;j#tp(~hbxPa
zoe>9-K&T6Na%g#JYMIjYr~#5B{ss_NBU@*!pPyf^_nfMf%h<OnA8d+z9IeM9P4L-u
z7mH)mSL7~ViI0yzSSS2e90$~18i{dqE%E;<-Le!`g+qT+@kXdHEj~R6)dR`wR=`DS
z>5x%^cAS00@jGu#;Hg=dicq`?dA2#Yxm|!U#++rNe>Y207zD~3m*pD$8=i}p^+&O9
zo)tVFWSmD?$jc~31)Cch!U6N{NZP2NrvRlIqcQ81yEogGp$A~bo2=_*T)lcA!L2t#
z!i_IU=(2+R_pw&;ASyh+52795n&fl*yyx%9Zrm7vEsaKT*&sI^92+%v6H-M(K<1Ns
z4`~aqr$Cx0M%c_)WQatdj8sUL&|E*h2k|5p4vw0L4f}|FhigLrMr>>pm2xrhn!rsD
zlWxRgQFPP87@fD}{GRc2Gt$vP^p-tzLl8*lIn>c8!BANDsCn*)V8|vQ&g3$2y10`%
zs7yH}3SvkiBH<ldK7x{o;XLdeiHqvn&z0ZO2@Uqj(AeQ|4@&CL4Ou~Xme*Ijg34g!
zo`{!>K+Cx`N>bC0ad4q*XFRlF@*V~dN5}CtQS@Ar4n@~ze$$<GI)zqL=PA+iFj5dk
z7K#qVBzb(HiLdN4U_y|P>Yc~%jt<PJW@zjNf%1g=uzlC5T9DZW7cWsxdeqDa1a{E4
zS1yUwZKGeoQ%3N=!>Q`XTb+O>Y5|}Zx-2StMLlB}Zx5-fuMdUDPW9Ndj!A@*K8ThP
zq3+jv2k~+YG;>-&uMdyThTp!;g+_$E@j2e^*+x0JB;u|~oIJwvT0-0w5j++edH3B=
z1igQKNq0=kN>ATP(tef&PXO$ZKg!d|CT2;6!AW%EBHJEvOdd=%m|?x6;ij-w%MpZE
z%>l<;Gk1Y`ljFGhrOr|3Vl|b|e|lqTYI^)PK49DVd0MxM^77k~Hr&WQZ*e7cjrU2+
zf*^FQH0mm_Is&<*#Q6&whUR0dC$2>hh;<6BkKq?jt_Zjx=rCBb8pF!2h(|5Ay#?T_
zWVIG6Vlj9DOe5mEXY|r_zTh!}`X0gF<T~gs2Jqtecpe7Q>~dwT$pB)gde6M$3ar`(
zr>0H<@?7jKKLFaE>bkh*av;J1(9cJN5)N1Q;@{yrH_i9?F@Ma-Ly88Ep9RebwhIpH
zA<&L9=vV8_?)xn-4Jro;92ENRDvITIyn#>=8X38&s<!qru^%&AUqb^Mp~)bDqOV`o
zz&9b*VmnX+Z5kk^Yv4Yxvdjmh{0T&yGV$9A$@h>1##ul9b?ai;Wa^$gd_&`oS9_@Z
zUB<)sVf?;yji54g1*4y{BeR$3toJ^`Y*9uA1Geo;YstR(a)ioto?6IjX?>)&QhHfe
zhrPq?yZp!PXWsqr)4<mtJgGfalD6Hqf!ofBd=@o^rRY4>n*dqORD!z%FYN(bLU~u`
zuv=jq0_j-yJ>x}Neg6FUDCkj!ULCeLvK5C?)$?4&HLEy>TP+KaZzgY7L0zo+_~d=&
z+02*Ox<!b2Yz2<SNo%!qQy@7gRmcsC1eOO*bJY}|#qMCC8Suq^z)NoKB5k>Xz&V&9
z+qZ2qg7|_HG^b9TI)qVxON6m-{l*OixcY!OL_Ya8)_T0RkGw<Xu`xB;+EcFt(vu-~
zsDEibqZt!N*=2N&z8NpD$*>WoQ|P#L>z0t;M`|?Y0;QlblqfFn2}-o&?R@}FZ0*Ym
z)zIjsysqI82T<P#PGdi?Y<-0IRZo#BoLE4q^Vq7cj3)q48m)3GlyPOSoukHohuvT}
z9|EBRzxzH+Rx~<zgk%_AXiD4yZ?$))j7)-z*N?yN+87>IBAc-F(OPi0$est27m29>
zPy*YaN;CQ#kPUReW{Y8DlBoo6C;<D#w(nU4_%$qToEX3fU7~H_d=~7t6N`UVKR0C=
z!r@6Q=cQ$|Lv&H6rA3dbHfb|BM>rcvSY)$BjoEHi=$=VQ{P;D1<ZK)qN<jCdZg*9D
zV>4=4u>>Oc69@Pwehs*LdhSNA#98^;yCZz(g{ts&o-6oK3S?eyqeCOM2wIq7BvdVr
z(PI)V!QR|+;Vn<efw+*6BcZfzy9~--p#n!b58*PA&;m%=XV?3*OEX)urQE|w%5~x>
z^pev%VlZC2+kMX65QLQb9NzbYC3Q)0F|puakiNvPnEa~X>F!Px9t^lMS1$-Ju0Ooj
zJt!#1143@G$D9zhD@2f&Im%9OBoK8U#!IF!_+M%FV12jjULdQBi<syxVfR;!mmC-#
zu0by2FrFxQQnh6PFo`_Y0l}wgSiULEN73yp3-0{+J%8rZsVaQy7TObm({%rxMRIsl
zq^SOUKUY=w-X0Z=+(F2$%EcR){;)$p8%fg#^o0aUxRX@EHNczqxrTCr#|EL>W*JuY
z<^+@k1yKVLrvcE<#~kCZetp8<4f*QXc6j3pFDCmu$3;Evm19ppJNz>!B_Yoeb__R&
ztwWAaJw9Kn7B4v&aGoi#5Sd9d$R?n2s^L;ObSfZnV4(7B9Q8z}O#{^M(~=y(Q1+f{
z+=oL`s=g;OG~O3jox|ydlX$}d+D<{4R)IxxAUWQ<Us_sjB%dQu5`Y7~4qk>-zAzY$
z2TV5APs7u+?qr>rc%X4mLy>JRb+PI!v^kDLeR(U@)}l2a)MPE$aDF}V2@x3ocm)RW
zXV8P!XSS+f*&!=oQ(|P`6a@)}YwS1xB@b-?36ri0W*Yeg{{J-Ag{1p2nTMBt6jeew
zsfH8~C3tLSE$PWBDr<mpKV<8!gFOe4S`j87pH?S}^&fj$CY~#0WnRaUG)vUYSCq>w
zEu<{E96kDq-`3BqZV+Ntc)$~Zn|q-9!RRx5d&9-S!5DiBWMl34xXO(hoIQc&$BzGw
zjSGk+7T<{D5zyO)ZQ>;^Dqs>?e0bQ*Pdft0xB%LWtAETh9JJg@`v}MbFDa9O`gGkd
zr`ZBNe>xU-lkStdj&5%0=pZ<oa2;bA=4?GwA!tE2my`DF-c9P6sNQ{jNkCG~@Ix7B
zU!0NS!<+HW-IBi{xb$brZx4KJq9a2XYyf1=(-J}W5onV$d}U^9ShW$!9iWH7$$zXR
za+OQ`zS413It7o7FSGwLemY5X{~PhL)KHo)0j`%&IfDkH6%fuiJZuKNnV6wC+=3uy
zEFeGe7JPqw*;@8Kc&}BP8S*eDG2^B?_E7aA4ODMfR$96=@Wbsw^V-&V7Mf&lI3VPu
z6^^S;KrqAWD%2-btxQZzVB)OBiON{&GRQALV5g8c!~6hP!;Q{7r{8t3ZAL3w6N94;
zqYUa^48NO`6Y+qbTDFGax?oGmJ3BvJbOM~l%#X;;wadGkd1@9K!k6xnJ=owvd-g;l
zgXuK!Q5<b|*aoi;?0~i&zt1iWc+`3HT>wta$+QYQ!h0Kp8^D^0e#!6ijTP*}480Mr
zpHV+`%3nDp8K(@u%&@&7M&R8lO#YBPjPta(4mdubn_cr<=08J=3aZyDG^^n$ukT}r
zakNJ7(MB|C^bU}J^f|$O`(gq6?`dX!LG!`sO@wNqaLUVrA7&=r>vaWI{uQN%TTfwo
z9&U{BuhU`kocEjUX<AE@Y%MGzVn*2IyN2rO0NlV@tN8|y74ot_@;G1_kWWB79Dro3
zJvS0cU2}63M9g<+Cv|tSsFeUIz_ZCwsag{%>+d>V-$vy@)j)_AzqOpigsjdH^Tswt
z#_#$=UX1``s7sx{C?E<8jcWudFcBEAMXFp)`_bEokQP*Z2m}1fhYWViH=iB3x#}5M
z6~QE*AL}cZm&&})oJS2Jo)rfN?Gjoaq8!MUA;=1@78)MDSy1p2Qg8yt*hm`6vS2sb
zIqUG+OJ!I7hg8S(K$m<C&^<hb->GfeU0jVZMn(ENKpK=vgiNu+KIG_cz!L(QD6t<1
z)MvY?p78)xg&o^&dp>*Go#TCvRp{~N9&JU$+m9TnJ7Q}nv3;ek)BXFj$DZ^5&a4!M
zL5=xtG;Zzg#-8z731qU_V#lPCYI;jOe<>_y`0jGv*8pyjKN|)X@y}MO#`|m@0;1>f
zqzE44k@dcfOz<r<T^!^*rmA{4TXX31=McCrz~p=~p6}3Czj^*Ko!p0+4{GCx!<Wz-
zQKjLv3Q)T|aFzj}3Wt&t$e>0F=cfm4Pyo>h0NqUv^u0Iwzv;E=`jQ7Xx4||=vJ|E2
zcyC!=@EBepbqA))W!Qkxv9TBCoxslUi*Y!a9`xN~hg;^{?Cfn=2>2KZWZ^wNznsyk
z@Hc6TKYiR74bUZ^PFbC^6_k*;2_LIrx!sMuon`}M1)XeEQ~p!tRE@<rtLM=>LVry}
z9>1!4T-jkM5h5oB_R$+==hjs|!PEW6u5F8W3U{-#i*}72$kx=|{#hKZ@9b^4ef#F0
znb*ezU|me|KAu0J$>~{P@XY*zxw*QQme%Vd8c)y#KnTd}geh0~G#4eDZ&uItlbZ-T
z@W4P~j^0qhctY$0aOAbA5IO+J4#41iBku+Yiz>XaGR)<2COGunr|Nf2847gtMMf5u
zrd^INuI1k(B64EsWTof#NBY@MR+W@FdBEZi`~DH*Wjw6Zl5?e7jLH_sX*&!1ZJ{|M
z^0pD{+ErxdH1|mrAA^Ij#KRjMJ&H;h?a2RZy2=$V^zxejdLCoiwK^Qwa0UYOyQde=
zP<-)%0BFM(_8B2avc(RuA=AoRr)HqW!)8XP$^`ukASD>1xL%b8FaTr$7&`#JqdE2X
zI>4bT(a|kQzN!gob=Y)qc)1g%8=~ZMxqg4bK_{-G5s+9H@P%gP+G96{<=gP!JaE8r
z_)ar$j4_@*Gvn+Uh4Y*Lr{8O6!WO!Kzjm}Qf>#J87lfNL&C??8^NSaxd%*TZPrwd|
zMlKkF5`7Mk7EEY03yX)_>ZaFa;Ao6&#Z%@L)^R=%d3lY(A29RmpN{JCo3363xq#Il
zSJjL&Kqhi8!v=e^^0uaG>=^M6sx{euOC(_yde(jQjHTIx+*y^BW*lm#==RT@jte&$
z-*2E{#@mlWyh0)(6B+#HJUyiV|IHynpck|8@~T3D_N+(*6NDG*{<Y*YUf;ok_pnBu
z6~ZuthO*FJxl{}7YS;<fgA6o?o~N^Q_Ch8AR@sGdTI9V?E4_tytpGWZK|z?NeeXYw
zj&9p^GUw`g7NZ6lM`ve2Il1M$J-5&9Sc^zA85DSi=m+RJX?m|WqlqJ~`UJ^EsLer<
zkxfYkX`mrW0Imn{N(~t7-OqgD5ok#CTk-&ac!7!jScfEV(4%+6tO42qGXqoVN5qBN
z`gj%w)5YfdmwRqFJE%U!%42>zsP@m~toBpZ)_Jt`W%ib#+-utWwK(6CcvSP{c&tk5
z^Xh6U1b4n@_FWm*97(K0K+|K<5VYva?)}x0zOjO<^E>o9L(~VK>SW)|!eI=<FPdVw
z0^n`){GTV6Ck;V;N%&=Uj(Ww4;g=U7cSA0LN&?f~ku%pIcdL{XH$q}K=3?Zv3r8+U
zLJnu1PHn2z|G$_z54fEBwvCs)BYUq(Dl40eBD6>v3YBpYnHd#PvZEWR$cQ4Nq(qVu
z*<D3sWR+05txDY_qW62|zMs$gem?i}K7+ch|Nr+p$8j9zae`TScACP)im<BK5|!jP
zCZy*Bh)e$L7O&m*VA+&*tMj}|ncVmI`8h!~|L-f?x0`M2w8T@(bbvv}@Uhq5_PH&n
zlau$a_{v|@<?p#iQ$}>!lXe3F?+!RpT4rV;T^BA52-&`T{d&{6FB*-S43Zu`nY(1x
ztZq<j@w)LU7d}Pf?O=IgVn=XW8P%n$KHpw@%f^jDQN1ZD`u*tMxzo)w+jY7!_Il@j
z2ktUL?+fgc*0|nw9<xKjVha>&zgJxTfL!E@kP$VXNl|{q+neJg@*1uc0ou_q^kq{e
zWf<*n89+;VrYmRx!w$QvUfl$I40Jg;U9;W4b-+gDJKn5JLX_TJPwtXVg64t3k&$t+
zFSd<B=wAmfJ_NC(3QZgs08p;l-jgZJo4Io?pQru}Rz}{|Z=P5CJC4a>sad%eRGWA%
zbGk%ow9!(pkT(Mp>`}xLLVEdM@0Bptr~G|W)!f?$GFmHATCIKb&!=a55aErAf6E!{
zFw|`$5EL`wCbqV%aA~6@<pYk3U+@@CuA70uK_2?3Z--JTEClX{cmG`UsZOn00v3Wl
ztG>|imKvn1XKYSenZBD({jkYtZnqlAllokpHa0+mN=wA=gr3A=_vgewslhER?>8_0
znR}oYah0pRgC;Q+70cCciJ=J2vi%#tk^ozIF|S;@CJaegG6GzV^fHU@Rbzb;+-iFm
zl530*9-RF=4Idgy?OhtDrk?XqMutD`!D9_7dcJl^oh6ElS-q6APVex>L&?F>X5OF!
zpWAh{YsF;H)aWOsA(xBt^15Ws@KXlb3?g@s(R<v>^WU*!4j7u-(rm^x5S$o=Q7><&
zc9={-N)=76q_w7Fqj-jpdGq?UAxdD+xBi1eT^xT8$T^fc?cnSk;pXoHY>(GSRv|FV
zC8X|*|BMn#Nyy$_u~ny2v%Iy#crQD9ltEsA){O5o1lUP5JqGb=Wt_aTK)~op7%$NP
zH)YUTv?gD^V6(fewWG&+PUJ}2o^@gy|D)D<WCmdP#pQ*6RbF(ADMa1-F0n8a)1z5~
z4oHvAFu~y*w|rh>8H3H~#k>6`*77*K&Zm4aKoN5vTNomHHCT22)7Y+bvM10_Ogf)Q
z`<qr^g=RvkUq*;A7TpWqp?aZx>v2aXI~qN{yvF^*lW12v*C_2fA9;#r>Tgw9qG(j&
z+a{sa;0JS}zi-uY4f=UWbN}gWTCU+~`(5I`X$(%luSi*0QC9F}#^kEM!n-Wy$k8!S
z<^|yY@xUc;!xj0D4<GI!4~fO!?a;lgxA^7GiOzVG*Gs)st8pG4L+F~_-iE9ZE&xmi
zqlIGik(c}oAMr$WmVegy{&w|taO#Xh!|i_m8nEEeR-)9?$3>Orh^h7$7PjR@YVtlA
z)1d6tavB|J8!gaw$dF~-;DC0Ix>0q9YRjH`KB}tnoYy?86+xL>Gp(aLPjP6JcWCW~
ztr(^6I;qKw>C=%e*5aX6nOU4wwx7JAJo9nh+UcAm6wLQeP3b{3$BlA3GFVe{6K2mL
zgs9hV-<CZc+>&y967$g_`-rw>7;O^Nuz$OO*Jhm+>u{bs(U%)Xa7L<s6!Wd~oDS-q
z?k4K|9fHP#NE=-k57lKBO$U1DPHLRnZPrVN<gtx&-*jJ4Vq27Gt_Z!owOw*}otTrU
zKj#eeK$n=GxA;M1YcMvDVNL^`HExL@XnAzYWGk75usQ_;yp?`Sq|pxX3%))rD}B>#
zK}m3UWu+H_RTJ#Z12o79XlMEaIawQb)Au;sWOwS%l7-PcQL2E6wA(5)MR~>ERX?59
z6(2uw;vCrDx{89<8YUpd0sHn%ShecH^m%qiABair!^>sk4!8eL3qV`aMC;Y1mD}O_
zMR}ieyZyigV|nRd3qHM$h#6|qXP)YzVJpL(uB=n&f46)&w$73-b4zm5T?V)zD_&Kc
zyY|(*B?Ly&>DA24Vzu%;Ahjq!vD~bShF~jfU)Tee?<51Erd3|7xlZQXkSeTUX?_}2
z=I!2l67V%;G^QqX;`^BT-8LkYDoS!Kc1$}wa(C)9FMeY|X!_Phj=NKJ^*xO16u*6$
z*Rw%UtJ>FT(EGM&L{_COSr5#IW}*l09vZje)+y77aDF#F0sC9^atI8G+thNzh6MAv
z9-H2DC0K5$S2EH=f61T&At;mSgX=DAUJ7YZ6`^(!xO7#~%e36fYYINylo!Dm-O80W
zk?T;;TUs_{)ZpIi)~-mY1UvooA|bPP-@?zI?;^`^d49epqvNDem{h{>&K&&;dL%&y
z7JxnVbltbUUWrj!9gpoC2XxArT#8So+xP|$(W3z|t6~#z*=)zld{vOTmcg4gBfW;G
zRQ)c$B9)EUdP4hQw}#w(F6oVvnPy1aU<ib;&%jgPHY(oAskuaKzUPTawLG@(tS85F
zN4*k1l^FS#;?A`_Vx}ivBRcwlm>)UfL!Cqhq9diuZiiPjKb=Vxt>MF`g*IO>pq(&f
z=vXJeTkN)7CC2h~<zuyen~b_%>8W9&bJk~^{ZVV3$=WS~epGJPpF=ZsaPn}}m_)VO
zPkSE$k63$p$F;hG8>3UA1i!ghB>-0kK+p8_hO~U#4T+%qv+lH_Gf<OwdCrENT(w~}
zB{xVGl}QhvGf;byD3J1o92Kb_!68@w^Y?nNI=*w3#b16TiHg4#D%y8x+4G0F-s;YQ
z8b=>UA@l6)>@7JKb)s#}_yj`icCTwoL;VIBcyQG^LGxtV{^l%k7ZuG`)MK5!>rYxm
zG~h55BJU}nBS70}*`##yrcF!^+|XaXCFR&Q*K5DBH=c+daH(0t2(v%W`QIn}=P_TW
z*)K)7G*dqW4OSubD=s=hV$gYt<KP^9Q9s|DwYxBnW{Y32i>wm-sON}g&GLE~Lyys`
z51oAm9y`KoY|HCT(ecMDwe#=MwZ~gMDN=oLs3%?sJwL8m|9MQ5OWX`#621b7QAS05
z7L_1p>^ueJtw(NxhDjxf2y8QyH=T4FYYmtsx6gXK>O|Lyx8lY?jjz?UtKmBHFw6S{
z-#U%kZKSV-;sIFXlaB{T$yr4Z5S|~hxW)y;S&TYGwC~Vj+0v!EYOW4C`&av?HKsX4
z+nzDAX-7-@l+h`H_i|o^r#d_Hw+zJKs&VpfKHX`D_y)0u(!pC#IbBIP=<kiX6;$Hl
z>aRMGKY*MMkC`;b4@k+hCV63Zat!8K<b!D+zQ&twxTy8g+rs0y(xKIz%Qr55ecxP_
z5L_d<+g08F{maiT_}$HYeQgu81`Wa(LZrY(WFZva=t;m*G*N|_1c+J8;|1*oF1P_<
z)_d^$-rKo38%+Xajz&5?eoRX7JY%e?&C$)Oz}TqUcbqgj(8Xie*ll%w7F^`ky@A2o
z?L87_2TIvrH-7$tJ`>^tE+}?QGWP(o<ms<pn>TOPr{zgZeRvdo$~mmRu;y_o_n>8I
zJ8XqEU5*M9D~Ff}DSK7*J9}{w7AN?DnL?|4?o?h<f+?7vdgHC?jWzAVU(bYfxXW=F
zvihHsId3T)-^6D-cGuSaJEd?!zzt}%p>Fy0R3Jvhbq_`kid@784<0-KH)SgTUJ%zn
zrl_Qwcj_cc1JObrOzWhsE>m+lGj`cgO&^qMI^JWO%91b5RCARh_BK;>Y?jwj{$;JD
zPr2L*hx1z*T1DVuVrJI8Z{J<ydeC6V@a`a<qaAM+!mbe^Lk2C{Uzw{DrL)>6!Q8)h
zSy|cp=kZYnX<mUuTGWqfOzhRf%e`08_2#u+cjeB?3|4+IQ1<s5sX4qf$a2xlf9j;x
zsF4P`vNNZ^m*&uqi2l#}Vh01>Ev%8--}k#FrCmV0NG%>g3rt+4s1wBm0-@M7DKIrK
zvfTy@IG@#fL~AAdjasf}VMp-ITl{koWt0jJ$&1Ug;?vXJu#T_ayy-S-e3)g)t@cZp
z5}pLHWN&Zpg?t+-x-kGZ4Dt<%I{3xg+1Y#1yHKU2lR3VWl^rGeS)a1nL;<5Bf>3C2
ztDYN@jS9N`2L!Z$g;F5fq}Kq=HljM?0>Hf^Ol$bm_YLE>;K!)q7}%2@n9Ayo?!rx3
zWgflWncdNiN$#c=aG3cm!8j5_UtC5-r02bs1}0#Lfnn25U9_||8*guwmikasJv84V
zzm;L?X%Fr@<+%CThDEEPl9-MH552)*%FjE4k01X_7@ulqql^3R&Xp|46wpwQk|SPT
zwEb*J@h4M@egp!Nv87ow*Q-)zfLJ(s|60kbbUHrFyTP@$C+*K^NhK@#!5Qb`r52+F
z&vn0!I!J%<%OMqK)`09_ds%m3aR}G^E%SP*&PM=UWbeYD1M(R;N#~cm3a#H<>rU|=
zD6H5O@3X8jU0Q8?qw(MKtA)wRJh~x#Xc*KM+d=9U;;u5P1>8MJ!z2K)=W&W#AxaoZ
znpn#{eXIVSRHhZ|nE+6II?S^Q;I$gAe_V5K<Hym{?0)}1!OV16*p0s+LJu(J!+aE+
zgOJ!BbRC^^+c-c!(Gh^hZpq2gJ~VMW!x(=9d4qoLasA=j>p(d|h}ecxbIf4RI*+<X
z_4he=I-`Md(8Os8@`aLP9&lB;;uC&@WFTX}tbB)uzLuRZh7_88!i0K^ip-%Nm9K}m
z581(8$gFEZd<{0lBXSNUlgw<I1U$XsJ&JZAf`2Fv!1%8*&`{&)0{>Qu(%#%sD$a)~
zYu6BEWsrZD#x_dZGHBGkgst6ay4FMf<<N|bmvV(<Fp_G32ASiB_JC|jp?qhAowZKT
z!T67V|9yx1Di&g6yIHptA_vt-;FaHnOxWCSw6&Xi_P2{G-dmmRxVK9^wYd~n16_ls
zC&74cL5KHi<_trV0AXNx`;rw)-(vVA7&(LdRGOk$r*+-!oeO)#l-ln*NgF_T6$sdf
zk2ytOP{*KiFe1}c08@P@%@s!l{SElT^YeJUsdfIa0AI^yi1LL(BR)&vnLq3iZJ&ro
z&{^!>>$tkU|F|&(f;N4fRhZOPbw<Pok%XuG_N1iY@IuHWLTD6dpG-NY;M|B5`!C5g
z;5pug*QlO9(l8Nw)zP&)H#Hc)V_I!td9{lvhwoz0CT%3e6t%x;gm?-u;st}=Wfa+b
zpuva{lTTU8aN1E|kchS4uZ@6VEC+)r&2HRoz^*Inp62)@ZJ2=a>!!Q2Z<m+j12j^;
zziCKcukzx?%4-|zdq4f8ho+JSHaGikJ&<BMN6*?&i(nZPvlnLUy5L>)oAg>3cDOD*
zCjvN=CR!~|aVI4#(rsX;agfLW={zq11;A0_HbR46XZOd~H}bUMGvX3@hhi8UpfO^4
zAPUA*@9|%F0w$3>!rhB=xu{YdV3aQ%H_r<fhP;9T(P@i<RFV=LF@K8UUtkh&lcRLG
z<R=EAG<}neHfjEQjRWDBzvQ^Ts&=Q<YR-YbPI5x}3Gu}sH$y0QEb1e-fB-}K+K=#&
zbatoZs=yGHCFbTkD!=aOy}kEe6YnDXc!@wHQ|5P2N?Yq_neQ9T$LMtFa@rh`t74(>
z;*8b_tS(5pM+pi(y)1vyk^~$)*yyR&4cJ`|m+OQ0ZBn`?xyQ_?YmK;qTBsUV1S|t4
zxEyotGEs8G>VNW>Y)0(!eo|HMT44q~bza<$ojC>>N(NIXD%18{=V?7WH*XSxMVVMi
z$Df9|y0=6Ek=iF?ZAWgR77?E#1oX!_89b|IZQIs=n)?tWCIX>@O1C?@)>Hqxnx1pw
z9qbXs6T7+sfO35RL}6RQV`G~_?17X}M?(9i-3>w2(*`44f?<SLoom;5QZcE{yEY6n
zzIkB%mfoIt=ayzu5o!Of*`k)J4+@ny*zBPb`tx6*)+xVV-xUd!)Py5O%vfh0lU+vR
zSVrJb{v^-MW1i8ii|vVrUoZ}7+M$EjtC0lTEoz-b*ydI?FxDatDBgk|U&v#bqik4N
z11MXb26VKpUuwgk;E5-Qq`60kSM0y=<@|LuQI}$hc97<3I7Vv0ZTruqMZUx9gyQLx
zIikFS3c?@rL&Ji&jEK?jwC%319{_-XR@{WzZ0s((llT7v085V}WMq_HN&*U$M2vob
zh2hGNT54kh#SjBvT0vGL=Z~6wIZ;EO>IPaMrg%ZrNvZn1=IVtNcW>V=-2LL!p`nup
zY->>~%!7fc4Jac`f2W|XG^na9qlbT)un2@4hT?40Q6`O0;WQ^g$ITdVjY>dP1%ZjG
zOsi7_PnG{}l6w+jD&AX<kt1^#UlILsX68<~qV%t0ilO8=RDT-^8~((!#YSbfAw%|2
z@|4QOKnw%9)UV%eu-Sbi9(}A&WWL}kWztFHa2*bomY2^%hAr6u6*b^C*;C3h_B!Et
zT7I=lj8T|#{+->Ns6f~Mou~lX5heo@se6c?uvhK)%iyFy<eVrFFc&~7_oA1(4`+vs
zsVh6!SkBb5)8&#OM63h=fzr_xbZE@&j&l2{NDoePx>Bmty}5cJgF{f*V%b9Hr*cH_
z#Ba}C%NO-9QAY<dhgbjsnZ7d{f?Kq&e2wReqz!1*d%E_#!8T~uMX1H?OZp6#C*?F+
zl2toN1c#^*WZ)d5rUm8w3FG8ksFdw@_Ni~&0`(=71DEU^;3if)0Im$h20>Sd5E@eh
zEK9`ucb))__Is%OWoJo69SfLc=^1DN<q;yR<HsV{;X83KL3nF5HFOM{-uz!Nk5b?)
zdA-!q<Y@}6agOgJAGpwzYoRsduTzRRn}xT-ss(Y8m&;Dou!WN+1ZV_(GgM(D@C^qR
z)fs52$oN2^L6!<mr-Y9gvFg*lgTGju!JJ-J^XySk8Q`A3v&!H>_`^Pn(q8A7<?HSV
zFK7to7_nCS=Dmf3UJM*r_VufRTL~9egUx!(P;~<YL5E(#0DXGq$`vc-APJEpTduel
z3Z%w7V!Gj5p1M8*)=3eajaRNJ|MaYTcCmO+BkHz3HLLa5-c}?6N=kD;h)?>Pwu|2u
zSmf`6kqaBu*^oS-0%btpZZp$uS^o)$XWGraWPsNhJhtJPyA3atNuasf8@L$~DD4%y
ztb_%^U}`8c{?#<_c#IS8((kwr-3k>xdFY*n_kkiBw9J3}@$1RQAAWvVYuAe57pCPa
zi}#nDZ#iSRvPQqfz1n)O%=n=Vdr#p8Wyc4(`E8gY2)-SOAK2|_Ti5xWCvH<;-j+8q
zE{c&2&y34{`^T;SeD0;ofZ)<*K?VR&H)na~>16mJUVMvF<5Ts0MZF5`nnfHt)|X6A
zJ^md?%?vCiKm)itUrQ+hTMB^E;_HCI)Z+J1r&^8taU%WhU5n5)EehM<bUM3l8^1+W
zYHM@Uy>D(2j0PlwOuU6EQnYD^*7L8bJ*{&Gl4K6Ywsf{kH*c6@s67tt5(@vJ7o)O2
zwjH%D*<g2{=kK^%q6t`R_3E^Je=6*+KjW6EVl;D|5y5<)skp%qJ@)pmscG_t(H)#>
zLLk416g_%u;eg@{kJLT3q1cD)AG~BDP_xVZBkM1wYuEE8_|;sJ{G|@<<!p){3<eNd
zjOI*Ibzr{ZhO|>4Q~pYRA*I%KhIayq#BO*_&`biU)D?7u0<6vvl9fugJ$5IX{ug@V
zdJh}6pQdLibS)MygNcDO+Po@+MW|)lLVx5F=S!DkD$<rVxpVMgb!Dg!EmBiX1v4fU
z<JaI%!JAW5SpGzB+SysP`8B;#yAv4a0pq};)QTOr)06%J)+E3?6b%w?ICJfWg_#EF
zNUTAl@?rh@Cqi9Qiw?c`0`5xYI6Sl3w{M?LmfPm%*A3w<r$v;w%a(;vgS*|*mdQ69
z#g0*$0d%xVFCN`g=KxKf>`ub%#gN1Zn}N6=8Lm`RK+F*|`QLy3OwX*}s8Rmms>&W>
zMbf(WbgWs#`{T{$&lY^-^>yodYMQ9~XK!>f`LA4z^0Q5Mr0YjbpZMeqiV1?7oE9+0
zq`bxGY^B~R(3=<LHQ9m|WA*&bH*$EB_e3QJ56}3CSKc5h<Jr_qD?mCn?mRQ=dz4w)
z_M&e-(aTpX)4Q&xMop~~^kKbb4IMw~muCZQ7|?G6c;mC^*($zFXYWb|UHX+;J)*GY
zdya)ELY_?v7@z)fQP#2gsI*6-3nOa90b0-}-lFbgLZ%BYS>TSyYS3l#a``=?&BXOe
zQIeNxP?pg@VpX%O_bcdapDPyGXdymud+)Ejc%$*cXK(orBmldU>Mi-vxZvZ^sDuSr
z2Jt(1u-&>zw{G3aBI68b=kH}fwiazSVhvst+A3ktD6!R<T1VL?8<!4Zzb9_yi*u7r
zTwGid7yE$fAc$!BV}FS@OD-0t&{A%Xn4XrSJ6_p<_~Cfd{x_;ZlTIhB|38--c4o%T
zF&PS-+XnAuB3oxFkfoZwrf0`iIr4!C-{qrEnG2n(7YLm@)9ewBWj@pxrL67x1Lm>A
z4WA<yrU+0P9G`v$jn67E{)IKwUD1D_(6%X)t7%(%oiE$;gHos6fYIcEmcnR?yrDuZ
z4jO=7<WB23=n5wRU6}>z?XLSvrl=9)8aN+FW=LT-BuM&;;X{Xh0ofzO9M}>(UV17T
z?L>nwYW3^SWImmTK$cWeZX<-!BVo|dz;pmHxNX})Cn`@E>F|vi#)jG$xcY*Dh0Y80
z9GB7$@rroEkojj494N+~6b;2ASxPnBR{jWnq(mdd+$S<!4Njx@`Oe%Boh989DHEEp
zRY<Z?;(wZlz}mz&WNVbU#t26MAbma`52X#}T5Hn5U^AvnT|_2CNM=ExH7&cp>BNy6
zHIvbb1y}7toLqTl5y&MU9u6@0>hAi!mdIwmAi^+9>4MmBFTsLULxz+-JTRsb4V&C}
z6lJ#)<>II~)q6rQC3~Fh@NyCFM*Izcd4#fZ%xZk$h?S79aD7a0h{lh{GP&-Arl#BI
z6u=!sObb}gG`P{YadjZf2U1sK+VFDm8YpA&p^+AWJOO5l#IKaL`aVM1Jn!`E?Y_yz
z5!0wkwaxwuj-_}+i0p?((Sp*EVK+uF88Lx_6Lba8SE!qcp(tcwE-qQO^Q}MWu#wK|
zfkOdT9hF|?X6_%KSgfn0UDWiP5OmM!b1Us(na^>#qBPs|aip$^I{g^!Sbv}Bhxz<J
zMA@Qmp_0J~mBy1(A^kH>fSv8b^JvjTYUx<#5#5UbK_~|B+q{^QwU1#Ik#ka&|NZoA
zcL*ZAhv^qa7#Lu}ck|MXqk0VIl<J>Axng%o5$Zf`f6<2zrfBBR>Y4R%6+{X&NZ>N(
z;Z>9y{c+{`d8R&~M-M1&db_Kwz8^ve1*5n(L6{9;I6f<r#Rm6fZO^Z(M<Ddg$jHbV
z8E6%TzEiyc3b31T`4y7~H`~6QnW8gqJ}v!>i<;Fb3Krnd84F{k*v@AY)b#0i4#mWp
z1$G{np;+kOjme@-2W}fQ0jp!CP=%8h6gPHzwfKozX^JQ`n88U_AAq{*THxO_RI{j)
zE1~@7V5W9<YQEb5_*f|`In%TBL#Q$7haJ7X82|Y8d0Nq}m)^+M4tS;99C`~okO=hw
zefn#*Lq%vk-*svOBl|dvtw`z*Ts*3Nm6K*<vh&l+#3{M=?%wssiVO^M!}umm5d7H3
zIy3Sef>aq9xj>O16KC67&L7Z@LhTms8#2q^phO)Z;(J~QH$gCj?sNtG73@*Wi4~;u
z%{dqNyn;~jEH_gIPzH<n4vLU)>Swx6Gyyfc)b-5#Q%7Kla(>1bLcsryXq+<>aQHXx
zAvcP?M#F?|iD$>VD#VcOHC0LJH-vGFDTtOb@Xh3gH^Xi`d{cC0jP0|cRt=5p-Pw(L
zVd1lG-c=$2ff=Lj`Q)1fx`A&(?Ex<L^LGC<<3UriickmH3x%|_Xw(}48<GHo0DNWZ
zv4jh9oAFY+rpk*{>#-1$i3?4cwq`QD5C$iuPxDBS4D%3u<9oq=G^cnTVKL><sV)t-
zcxNfUCUFZ}YDO7pU-|D2#%Xu&z5zV46t-SwhTto9X<*5;g$ftzH$VnKHY$b^MfB|B
zHZph>^k#R%n-PQ~u0i0Hzjp5I^L}q@plZapTz9XoeGE}Z=@Q4^@tH1sY?RA70&%R_
zzCFEr`QG5*Ta3brw*p-m`#;D=0Orj3QT8*aBo@(hN>vrtow1}<Y)2eU2usb*wmXx#
z7U?X`gxwRp#_R@F?oZ>#3^rkwk^xCjI)jt>5tTsSb72?VOoD+DN&zOecZso#{NLB*
zh9HZb0IiGsxeJ<H>I=++;0GvM`n%lkLM_8gY2AfUiY<-X%V&k|>{Zaesh`=ln6(`k
z+ySVpc?9rCe4U_;k`#;aKDpaG`=uyQ4u4qodpCur$W?f%_);DrjDpr~0!aBC(N61T
zypy3v3pi2vw8P>EZ3A4N?2r}QZpWi3M5SJ0Ov@ub1vCj?vU#*&ZWc)#YYh`4OfgV=
z5TOOEqt*dIZdIH{_+LnYnkmqVAouvfQN;j3^__jIkwE%n(b`Yj^}zn(ASuL+j*+wl
ztvN2jsvjfqw{Yi1n2k*cLLT1W`o{}c+icYP3t~9aiiMk$C&T|M(yC|$;75jnAaw_I
z<Jq0Ije6Pq7yuG7#1gs}yLA)M5{MT0U|?74>6*E7?z#-Hin8rR?|S&dI<wLNJm3oY
zp3WDoV|X+qp4N5iZqYyYY7JtIN#l*@$7z|U%UW5sai}beUr8?{6E6juCsSLAw-x%z
z3eWOEj~K4Z=lbEJupM%NDR2~hsdMxvN0_!0kpP3OurP>P&~o1m81-ycmufjkP;}b!
z4Z<*TI^4c~umLLKsE&i{i*lOr<UcI{_$!cUA2qFqSy|#0G?6@w(i239?yWHr(-5rs
zU?Nn2J^g5<mD$&K?P@B#-0=;`;b{&x7slB(yA@p!;u2@;HFi(DFh&+n#_k!2I>l?C
zJ>uznoD=h@l?uQdV8!m-E@IIcca8%=akVpv5rsY}-@EaW_Y}IqqQlMzeFy4u>d>V5
zr~bBhVog1N*fF5b$p<bX1%WOm4vdN)6&Wzh9~HN#&>t^QKqq0ccZgC&G8&rTajd21
zqOIz9&20G<wNks9lZ|uQs1Eh9YfHwk*0D*fa(<cQ-Unn&bz#k1z7^TI*QAjjQ4RuU
zekm=DTN66&sNqr3*xe@V$#RH^6wky{v@H#>d3d()I4>5b6L}Q~uC8Mv$LrAkQl{6N
zDHF;p(-#YGKX@?t-t?#)HRr@$UP5<<#4aY#uzirJF=`alFa4odXcV~w{2M&{5P-}_
zPXHPN#enXBqWtEwXCX-MxxJ<l0VJe1g@xm(Peh*w{q#q$!jv6WQCVCz{iOZINIk7}
z(tiznovCw4%cqyKRfoaD{c7CNJe9n2+IBNg8qke7pdA>e2$m+}kRqml{{5pTpK@^U
zq(cV}=9?^4B(GzfkK{xJfZj{<oqo%Coyz7>P<Ym!NzO^EUoyg?x=`kE^kxQl86``T
zRjm|Pf>#p6XG6$}{0SHR^ljFBd<u*_cpqk$5R0XpsbB~c6P4!)OO_-kCxltbSVx_|
zgZ!DCuKcOZAg<M^z$F@i$J8_PF53Z8AqzY-vB#u3^rMG|tV})@?1^bkC<Ll79RS!a
z<>#OmWv3FOhD6uQwW?>co)&M|+9>zL(Qb1}ev{(mCc<%Om=Fl}?Aq0($^ftey#D9<
z;yyv9H^Z1WxwKOB-IUJ6kBcimwgN?@+Y#{a6VTk+y+@852{5><-c^JtOB9)Ty)J1T
zH1G_7Z<UUka?~c$dUyXhbaprKXX^4QEf=mvoTC_YBZ#B07Oes0H@I<DlwBXh*PQXW
z0uB@pM4XQfFUCB0RkOCqE&FdMcPv=+(xC%De|<GI@BBfu(sRUpp<?rnJ9)iEBAju{
zRE(_?b0@E5+hoP44cRUON@>zXDNBYGl5*Brr2a5tU~800h-chp%%E&$bx&u929xKM
z*nVo1tG*;|RZRGB7xz~-_=?H!`-3Y1{55L5y3Qa;<-ew%c!<bZO@+}gs}Db0ZX7Sp
z58^))ZOf2zgH6|ZUwpxLA#<*L^)nrj6(}rfmFx7~1?$B|2@=((SDsB^{7-xZDY6eX
zX#28dt=Ha!ZR4GtMzErW*m~z#SD|Cc6IHtoSdRd|f|YGnw{BfHR^55O7ziRr+Fx4H
zlShffM<&rcyR0I0L)cI6*?;23xG)3%Y1_9mze#KA#ik;p2ePIhln!6~ktIb#&~V4$
zNKTzss~&t*Mi3_TnIOj%c}vyp+_hWA+kD#D+52$L-J;{Wn$*`&HS)RN!m`$G-8^IC
zh10`pz^YBK&^B$DH7!tOeN(>);|)g}r!<T4u^QXcDzH_zT2@`cj+N)kDX;wf)ajmu
zxwfg!@R-8ao^ffPhR%EHHg{zIvbL->Hk>~FpnjVZ58n&g4dzAg)4cfc?c3|r@XV?#
z@n7OvugW~4uE#6PEhWS8&mrZ=uQkmkO|qMN41j&F{c)EuAq>XYR4l9gdSZiy4LKw|
zaL@zPWqrZ6kiBA=4zST{VE6a@R7_Av6Z_S>|0WwST+kd$Ov6N)CfaMMjbtKv*@kAd
zbo)79RnfFdU=9amoKi0Y9SMK&CZq6Rq?}u#&=z_a7(V0x5wBNMv@jf$O2KGW+q3HX
zTQw$1bSx|uC)ft^g2jE4TfUQNk7|k*TT9ct)-s1~n;2O=EcyEszk!%3J~{-w0#uw7
z<GR1xc=s%+q*%i(l8SWDph0ZsIsnKm6B;^a)>{S*2wHoM7@(V|BP06HPIz}QPER0t
z6<^FhO||BuxL{?|u^d*$efoko`T74`J@BY<;1;t7r3+oVn^^frSJ6z16$O<zDl%#<
z;gTLNKme7Bb+1i$CGwr8%rJ3B>-@GW=KwxrCQM2+*s4Iy&eEd;U!4q4)VBc*czmba
z&{KP4J<A!(kJ_FOU){R)(9;%QvS%P*!f%u)(yyIAdXVN#n_i_@4Lr22wcgaL=*XC~
zZ;0ntS5hOK#oN5R(G-|L7TuAH-Qu`{G#%qbKYjXCZ2r>byy10{N&CZTFX-=My{EW-
zL72yj(b2Y}22`mHOw4hxk)nU9?Me20+gnUyON3$5-#r<P7O^+dD-+@r^^@uInDF1m
z56oxKEOAx0XqcQ|_$+`Itn;N?>ZK7KqKi$f0uR4m+&MIS<1gd(J6=w@dNg?_EcJ9)
zZpU}8NPJUtZWnKerEd7Kd5uL>6@OQ_=>UzMiG8-0E@cX48+o#)qH&#?oasUw^WkYM
z+jaoOdqX&7OTi|dLjdc&G)#o-!!$~i1yK!Fy#M+`$0Xb_=h690BhTe4R-FD|cG8Ul
zAfJ%G5+`crxZQ8bFfw8W{HEdzxCt9Bw1u2TJU4nm%QD>qQWy+O(Q|%#qDJ+|Dfrwe
zo2XV4{L0OCu-S9)yLqg2r{SCBDw<_q8az4SG1BhPM7vFP-Ng0au@m43q|f9*J&<2s
z!?Sw!>@kLh>*FS$<^IGtbrwYb&)hwM=6vM3gr)95q2XaJk5o@COS;#3MsZHNsGBuo
z%9Caq8h5m7`dg=W@4sl!afq8hrGlPy$gpAI*&mT0t=^Vbgg&Ruzy&S1A5#J{hg0Ix
z1hCye2@TeE;DW%o#C<T5G>flc)S3=%YZ&V$WEqI<0O!Qp#$(6!#q<t;H1D1Yp(knZ
zU4ZpCqn7KckhjU*W-eWI>ReU2_CoQ7kyuLnD+u)f2F6Zo!SppoCx<R@ONx(_*^<~}
z9d-8~`Qx}^LLW#Hl3yWM6A>YiQXEtx+BV8OHZDHi=36w2sgA-30^*fEnlfqIx&%w3
zh|Ijt_rq$+=!yzA={P|tXl?zfsSn()Cw*=-!ytR=mjb^a&zH1;+C$Z<B58fqlF!9!
zm?#%=Y8S?L?8M}iX+&f{&o0ZCFBfeSGmMoqKJfnV>kr<qkZ7-5gMa8AIL2iR`KV`l
zGX#l@zjJ^_KL8j3ZfN{e%$nrVK4pRBH+VMm^00^RS0GDYaJi<dan^!()9crsJO1oO
z1qN!Y*wfc_`H3~xXvG1K-=JMUeJKOSfP@0|Q{$+>F(1I|f?PKSgr$NKLv<>gElt1p
zw3!}wjhu!mQRKc%{f<XAOYQ>1@6QAge09myo?APOwMgm_nDAzPTff37y(ii9YkyuV
zVOo^$0~dzETKG_JtO>3nI>Q{BhIsRN8M*IMuzIaxw?=(;_woj#Cao;^@WJ8nUdsru
zN}DUcpG<dGVJ$UFm!p2@M{Z=C=!Z=!o5T+v_$-nTX}^IVXe~4l#UYb4G*o3On<u$&
z(IUAOL_C7xAWYu^Nem-PDmr%PP}HV>T&tE9zwB(zt~XUHY0_@Mg!$U;Xp{gD>!_*i
zR5!ea<NZtbrRVPe87sdIOSc-Z9onzGyWW(*pSDLIymTL_EFD<{LFJu&1zhx1_R1V^
zFxz2e6j=1gBkq4$phLJ{xNzZuvA-1dQV>zqx`|e>VMEka*Kgh0tTyty53v32prD(i
zHj!`zp6{xHQoWF^A4gG}h)W2T87RueQ*;Dqpa#D#OYcicPByD*U0@OQcvjywfgu@B
zZasTuU;hJ1W}3q1;jLSB_(U9z({5wJl5(0;Ln>vTQiypa#<LOpTmV{Z%x@F@RlfuA
zW0ErSpT8eWXQBt*7rHqF|6X$IwHGfAz?0zKfRK9@$_+7NAccy7sOBAsn;e};JrVRX
z<a$<FY#4h?q5Y#bg3*{=s)r;0-ufhem8$|84OVg|*r2#mfCkw<-m6g^h-8cjp8g&E
zh!$fDbO!iSa{FK)8Ry0wZQ8VH3L{%WM_Z^!1f1&qS8Ym){6}R81)nGP;Tf9`RA{;T
z*v#b*dH=i^^uTh}^5r%2;MCkRHmm`X_PG+jW~PKGKl%1DgenR(hK+zn*~)3Ye}H6T
zd}N_%$7~DHCngI4!}CDSeOEM`G|w^^3t_;mrQF9)<OraR?eH~Inr<A998k0X;`^^6
zbILM9z{rIvuRw%IBsbbc1peaIh3kRI#EDw8;dpV&1)^7!aVKUJ;2GjTjKt52{5h5V
z0^BSbA{34j;WWmaOWkYvRbrTaTAVre4yY-dB)}S!?<t|M-Q&>%DJ!QvXb|FOm3(=T
z|Ayu8epaVW#}$=a$#x1>M45ZDNn@8rnvPRdc=^bwb&j82=4SNe$Fu|dC+nh*OY}oT
zXDI6kpoC>n5wb1O_){1y9JJ@qA*4W7S^)zw;i0X_LJYBD*o?%>j>d)OZk4Z7<ot~K
zP<0_|&1k?cC}U~ENalv9kLqj)yK&r^^#Aik|E@Y4z|0{g7-(h$3rs#j+2ZkAyzNhy
zzHLzBD)@gIXoexUD*+LpO@*f@SO-u2U$Zi{>u6^(PB0HqE`)Knbdd@ONlvpQV}~EV
z4w#>{5o^N-(bEP;9XhL3@7m?dtap*2ND6fZ*(L!b@nBC-fB?8i0Roaj2Kxi^s=#$n
z)ChnI2eCmwD`MM|Gc%$ZjU^lI;JB8uCJ8dJ35_mqLAFkS>F1W~v@^VTVE@a-(Ry)u
z=P=X-yD<lJfeo!(o7s0YRyaXkISSAFHuK!yvKGfQ!YEh)J+AY~;IUA@XohH>qZ*b{
zuF=}4n3;yW(ak<>{Gd+ylkd6NAHOQI{^2_+B2EunwrUj{M(q`y@1>?1KI<?X=+`PZ
zO9t|&gw<QNG&VBwjVqlJte9i3)u+!SlQhfop5>5kHwpb<xX5o3UfOOq4cJHt3!n^@
zwKu(=t2#2sh@F<ZUwN|y&D6?gg4@9QFIpG0TOYZ7)vKK#hb53>4;Vt`;M%=e!qk+6
z15mD!zy6s*){3gUsOY(~MPp<CZYX<#{JVX3^4t93WnU$c88HZtT&6)0r~!!l3kpdR
zW5LX{g8uU;LC7%gF;3$I@k2>Zw-L+-`9EW*1;#?OlvE5ytxmDfJMMYgWCR5`sBCM|
zc(kdhY5ecsXS9ubw&Qu8cGJrE^|HBN%69lRRgW(>TYF;9t%Aq-Ou$=0WX<5s*$w+7
zTh7}2ma5yMG_A*OMF`#!vG9WtcLC5`vd?rxnQ7<Sf(~FFXnIh){9dLHEUm}8&c>{j
zRbZnx<us3+a^}wdti{nT#;G~bKUrsH3>R~PGl!Fu4N6*d&cD#gST&bE025IMSEq5c
zJbJ0-g7GxVi`TzdGnpaWmhJNH-n%DpOJD_bHM-^PNMC@%T6;{MC1729l*z^N@<QBx
z!(?3F&4Ql@A|9nyVmqh((a0LfPlo5@b^d{+8z8rkd6Y8R2Tn2F+n<7lJTsr19%A?u
zo6>E-eUuIVxQ#f|B@o|1mW_+oD4MP*!G?8e*M9x}eHxQi&k~oP$m;j?%a@th<VXkv
z+~ETq_A0F}yFP;zvOyk2FRxJMzv%&@LeFAi_lt^F3=Wm2Q?S8U4qSOP;Ffq_n9dsx
z6Al=}B=SiEZ@74h4hS|Del#M(giBAOSzJsvj++!uTokcj;I^&LH{GbTHn_;sdEus3
zB;Jxn78>dLh&kRMPL*hT5G4q;3dAiOX~@v&=MB$YzB~dSgdOb$`Z4r>>6&lSwd*m@
zHBJquW3zF(wsx7D+brY<pzWJ@Qh~~_C$Q?2Etc<=Gk~qN=Tyez@VY~JKnfTH=3;h|
zkRVcpFJs?Kv~4=>!KOQ&e*61@#0A|0%_EK<k8@6~0konr!QA`?;1&XlqbMNy&)xC3
zXozmTYoi{<Da^3tN+XA>n%!RoGz>5Q2B9HDdB`a5k&;<Ze{(w!5s$$TirPh#Pn>mR
z9Sk9<;RMukcid|(1oHr(MQ6{QlPwQ4Pf|&r?YhabXn|=`eE6L>af%v>&`<Dm8m4X+
zb;+_*FXN&nHZEY>!h2(^Xc7P}f&q|C$aKYkc0u-`lt&8w0-}RRzZl1p?;r7I_f3A>
zS~c(W>#)kGBBWGe`OWkz_CBbtCSm90f#plvezP{l9r*4vBhhUW;Z%@rgjmYY7rnDL
z^wzXHW&QN|^BTc79zD4<zzHDFHY%zhYct3%3hZlyDiyYfDL6GK(P9yg6bxLxbg45$
z0AkJ=@AqK~Dyjv=>?@pk-ZFT<3bm(f*7NuGS8qC~oxg43fR1|U0Sy~AH2ZXNu?0k!
z{+<`@_Fi1CICX*n<WZwu>Fixlb!d$*xt{!?foBx@NCsbJ%zG0<(D&~B1=l5`V~V+B
zIOq=1!NeXEAB2lA(kn7d&4C+;S%O_I9HiTGHq{X`Ba$zmm(T-Fdu`~*wkV4E;PQ=1
z%eLhk!&@|cY-q6{AvobpYR>pho*TJmwo{iTbbHwO+4JW*uNpq_sIrQBj0S_>bGj&3
z^N0T9q|Q?wpccrA>HR!kDQxH*dt2Ma)AoehpO@XH1s9&{BK80U!W8lzHA#^F`bCQt
zNk52P2em!Z`_!GJGekYJ#T9#QX6V}6jmM`S&@f>sM9=J*i>QW!Qe2HM^g{(5(BPQ9
z&4CVR`I`;(NFTrQZt4L@9(3bFS)CH6#~1GiWnGQBSj+_CtH;0vDxr!Ta-=;=akCyg
za2a6gQ(Jr}84o++R@&=Ht!{N=?xgCvmHfWE#)xW_*5CAA4(xr9+!b#YG6=c!j=o$)
zscelTYP2YxF_1RvP)4=ERswVTm3SYULo1VP#xn0mEmDRJj-AUcyZpTPGaZtjz#O>Z
zKvP3IvHe<mnn;It?^4Un^7EpH@Y`3GE5<f)+`WN6-0N_M<T$<Mh%Q(QH=3rDYi6z0
zkvoG=ZbT#BckkZeAi(j9ATCy$exY}@vb(yqL$aECliUS+!o&1Pa70qw{%;}ftZ&+B
z)O&mYAufnMwj~qO3d^$0`ZOcgf_qt=K>>=tnp4Q4+OpP$5+W$`08T;Mc3-~TY0VCP
zht1+mc;fxd-PdOjnmSNbX3gpF-?E0XwqE06=4C{ukCOaYMqm9ljcB1ud0tr*rJ<hh
z+7|3XC?cdLSM@A#D3B0Yi)tMa5}`tkgG<A*bwy*NO=oYdiqcDW@Ax$Cmf<MysMKK?
z+OmpsH3dlq%|+0`OPJ(<bOHRq1_Lp0ddQ|=$ZxT*g140Af036OTpM%zd2J7c{X4d!
z&Y~qG;HEr!<tLmX9xf`7^hYzgG_U1)fw~nkZ~L=_EPT9wcJ_GOc>+V+_$2}`^Nj*M
zaz*%$G%m1h2dE~a6AyWEwS7*Un76HQ09a0^vFOM=7~aN7x0a~cMcZIfd$y}<I?_`{
zYcyH&Ob-ZAZh?pa1(S?}i3SAuM!`P5JhLtjh_oVVM5yNfY~eema79-^>S42Uo0ICw
zpn(^rUA}zz85N9;(`M=at?O?{6|4UWq#q*3g$uC{4*HowpMHgDTEr&MdO`*=`o`3~
zwykFUmf;TZG6CdCIIN&yFSZ|qm>C$VjSmWrL;P8re+udw#dB=0p!p2T=&%6A1_-<|
zqn*^UJK9cJV0RC^Kwe0}m+m_ad}Cr_vOFIyh==bUmu90?j}Z#Ne1uX)Zwu0ldm}pK
z(|QMxBT+bo(t9jp*_XK40_4Q$z5VnlgKvv$-S$+SQmTb*xN8xMeH0rWA}a!Iub{=I
zGA5xP&HAWo;59a>8Y(`$MK4|K3<P2Tc82aF!c+Diu{uS9K<Po}$A5!*1k^GxZHlm4
zSOJ}Rz!cEeRfWgwZH}NNHtk*^v^uq=d&Wl_uRU?pQr5TRAp{*fm_d^s)WM23#@*9S
z%|UC@7oN&0H3yzmJvE0utxT@DcU8H!vRLL!Ap&(%)X7^d;~gxVosR_6?K5oHei2$|
zQx74yI~b6XN@MzGSj!Lw2*F72{@GD5ab(d#hS`r&Rbc-nCIvJgQMS>um`rWyWcWI)
z$Mb5>D(8-(k7SlFgw51Qzu%EuO&j!M5gaAU(PIjv`b2)@?V|VNYQP)+m5{sItd{$x
z1~!XoB*Qrg5m;LWM5RvP487|2dId_Rpu@&v>|^KByCK@;W7M7+ypfNhU9s}T`T4@e
z1w~{U%Ulw_G3Yk7mSH?WPz+Oh$5K42O7MWSwX&#X7(0k`C#zLFJG5<^283oB*wexI
z{&9=;irIc7bd;!^Slus&Y!3SSH)eKO?9S<9<xl{CaO*^!fY$8TwU0dvlK*|-oz{VO
ziJ=^3j?$iac_qgUJ<Jct&CL~GPg-<E^HnARG6qb%8L%YxAoEh&ypoI}P7YGdC1vZ3
z>f><!?=!<oYbbYg_TlT>JY8AeXyssLX4hc`dDRVV#W&lLozVoyE+pl|FlT#*xV1l5
z@$14W&w!(`L|)b&YdS6Aad}ci=$_T$Z6D=<h{V3&3!0up{=znWuC-WpCM+i1H>L;!
zKVkMZ4LiEx{c7jALqv(ah!<bKd$(CH9s4<d{wK`3&NhhJTe7tI=W6Fd50A9C3_(}K
zB^zxf$SPam4pCAljBF5<&`mI+PWzIkR9<`XBp6JDx@!W;JQ7%hmP36b<5PTfN@$ac
zL_khuru*|pr@lY>c4bwq+Nlm+$YPXwT#xCToH2B{s<J?j8=%+kz=U80)$8$$VKpSQ
zF}gd7{&py&Yi%}r?q)ys@Oxv`nYx(-E6t<s5bL$Kmwj9<Tv@_**hVS1tp+T0WIfhf
zBq$Uud{mhOfu1Y<>ozl$&O{TkPG`D*6&AGtraoF*FsbFnE9(D#^&#NVlv?4(W&;n_
zA(C*DX)WbYBS?VM2Qz5rJm^3oW>~y#qTFOMKV6K`=+Re4)b*(yW?*x`eHzoVrROG|
z_&F`Jmu^*P7lTjKyW+YE$3R2F|1l4Y8ia}=gG}KNe$>9j%U+>Q<&c(Yv>dM~R;#Ca
zBbLklLX2D@vQ3)S`gqJXsx9YJ-HX?&;4DNaTSmCL;Bv%;yO@;Jym|96$j~9kdvxvk
z`maI}#{=P^#qU`NMcfso0907YEFCR_^K!g5@6?F>PYWO^rT0xtdXRTar#7Xr23V4<
z6F+_cjW?fTe+#n?I7AQw1ZHHuDM_mqLF!~CdzOhKu8rOJSfj^USEoICRwK|q%xCyt
z9xV<(dx+S{FR8EL$<3rrHm|n<q@bas+m($)1PvM1hvEp<b_xs*^H@5V@<_4*L71O;
zv@(UcBc=H;+%%PMh}m?Xfj@uakivkX^C>ruy8wc601FCa^(wm-Fn<f8=F*)%w<w=s
zYy}4lL4#`Ww;SZer^xg|61{4N>4oQ1+Zn6XwGST3wrNUn)F8ehm-zW@5zv;HF?o85
z2Lv9)z92d%z`5(B=8d=xFuW^Vg-1!Adi#0}jTBLwccd>mqh4iZ#lPv_0~LaGGZ4GS
zPv9VA*kb6=p?BiUiL0^k@da4bA43Hhp)7uxcy8=pO6oy8=wW7AwQU_@@rV+c-7N*C
zQCn!?mQnE>CE_XHQi$xm_^LIZOtv@R?JO3@WL(A=v?#7M&4L5#UCeoJKm%Ak>!i59
z5CR?*NUZgOdWPEDU8-H|jqdKnAc9dtiTwA&#=u4WfvPB1{e|438WA(<i4&VpACDbB
z9-nPrz$B0-PJ2Jvi%Y9MufA<<;cMA8(!z%qHo79w<<$2a3CR_S9`h`lc3LG&4?-PT
zbq5s*3QRa|!qm}m6b2E>@yy8sjP%?oj(vm;z@NoA!>~Ya;zQqn=&0pjO~a;<*3iNp
zZ?^SNG0PjMr*}HTJ`r67!=X~JQ(>zxa;PrEIZ%7K!6d>buu?;Uo{z#l7YKSEGig*2
z1yj4fo~iLCiCHc$C@ee&4NMrg&M~u~%mr4UNu-V2L<u0S7d&e&J;yu|?jJImEG+6y
zlv`pK7s7Tfl7I9Eo%%*!>z91{@E8Ypwb7QAyRs5>Tu;3a>pXNADofFqlW&9E!+b3_
z{jMyV#G@V0GbPPYkjuQgNUv!MO@m%x8WC5&{nSC$x*q~N8F=i%E~aL1<=$CCx4}HE
z*D&D{TZhazKxx3BH`PUi^7-@UGRVPGka-<`CU?F}=7typToW2aO;wGnx;|-S5^x98
z3UDGBqe8kKv?d4I@V}$bJzZaRbu%3|8wlax$hG?!Q*a*8j><<WLDV}3$$PwTCH53f
zQ~*Mh0=>%B;4R4HGYvlRJK^cR=_dzq9<p<D9qW&ch=`E?T{;O^GOKqyYYdlSZ)9Dd
z(f<AUZm(KlT5Lz3cVlZa&`B2NaxEgb61W!Hhg`OODKd<U%Z`>_B>U0z*EI>?#5(ga
zm9J=g?B22Uv1z2QW6V4vzVg>ExKMJ~(Tq4^v{;M?o(W5U{8j2ffy)7u%MI)}Bci#4
zEgp}u2M&S0Fa%91qoM^4)PG(7JtpTU|Ha`=#%fu9B?k@A4^5=3%mhfqf0H@;?J}b5
zn%$cs5O#>6Jj-VjJ~Lf7moj2j_km&0CufpD$JscI-3@m>nQ<@jN7~)H0+nda{dAhB
z2M?qxGfsEek2AUDUWcOv@7|5PFT+0<E?n?Y%x>nI)7YUJ1`y>?$#N1iKay>g^?>{-
z_E72Z`H#mXU%ZHf#UFLH_D<xxN)%)8^a$7B@&ccY6CEc@iz;87Y74spSF5Ub;Py-1
z!h-8c+IeVq^+T&&p&x_Vy3)egcn!2+umw<tUz(t_xTNzTRJ`|idex33+zD~xD=4su
zJdGU^^SFf!cOdWDb?A`v8c@LoUms1OiO%^t-Pr#J#imZ?U_{BIygMN!q*^WZybPip
zWNOp1eB+0c`bG82S9jDs;F4gQTzApAT>t3k=&WyNZ~ffs5Iye1!`=8Qt}O5UVDG7y
z>&#b@IY?_hnk}uFXlUs)x!~c)uXo}W4*b~!qXR0YRLXksLZF<pD*eQ+MKZ%p2IsGF
zPEqS}_Lu+qu@Z{_v9qK)yTu}T$Xn))?~#ck&wDOlggo}#Id)(<6s&pm<caP3B%s4F
zECABItTQD(Bz0Ooeh&=vg_4F|oi!(Z!K2G=b@dqYrjauFO-qJS9mr0^I-q9*80oQo
zM&&0gP{i5aRIH^eEaeXn`Ct>dS1Mvo%=VXNp6;~`PX<d~<wJ)XIFdX%PU%5%@9$YU
z2iA5ySSs&G%syD%nU^xGSfDEqATsjoB}+2-PY9avQS+tk2%dIm<Cg#8{#UR-)P~OZ
zP)jR_;i}jhQ2DM}wThqr5H}@MG#Rix6|`b9k1s%*j*cG{OS{2~ci^P3bHukc!kCY2
z=4;t;e&mlDN-UqQLlv^SyfI;cdF_r!<7g|$Rd(a1*G_H6h^tWgij8;CyQ3k@B86Ki
z@Wy9C_^kf&NB|nC+1o}8TbidLD^vB3W}_^7Q>IammQXwxTDQIp93<J%F?k=g3>)a2
zB2OckC7%dpCt7!EQ>g5DvJi#)Q8aYlS7!Y8?BA~xr30f7fV#pS0+Mj+jd7p5;j_=u
z@82l+RTROvu-~B;P9(ljDv<!d(BZmxb-NfbJ}i!4#MtH?I%LQb1{7qwEZr&dem>M4
zqGx4Fyw#%7Vb7y%d$#*+iWBLlPv($S^Y0%~Wo~WYDjlzt-K+)#?o>{F9>{DzrU1u|
zO^HoNm`5Rpq#@4w8Gv?wx&lSWOil$hiw8?X@N1oo%1T-KK(%I_lBctG{(9@$yk|ST
zv3_-ShZ`PqIr(&rFPe?TIr|AE=;p?dV&E%KO`XkX98Kkl?|UKPkthp+CnFy8mFHa^
zya_EV!x5f(7EDg$OiUFcz5e4wk<Okdhq*PKj=999v}o+D`8>K{62EFlH{%Tg7|E}G
zJ=5X=&WBg4e!Z%aQ%z5Cleri=RLiV>IKv)r&-L6wMIw7!XjttT<wBh!{uB^0M1U`t
zG|(DAC5n#AU_k1O>aT-&aRcfk4DKZPI?vzg4cCCK`?&Qpc!80>zt5&MS7oM6m3@fo
zvpfh;OczYAZzKzCE0M6CIkq2^!k`U2GCI|$g%ei-RHnd}4-Z~Koy<cz0nw!bN_Ebw
zqEx-YvKXl}Ce;%tn((F>x$?s!S$7S>gBGkV4c5GssZFaDN*g~Effb8_%Cl@Pp3x@M
z4s|xDLbpm~z<qp{?Af3Ck<}r6W!i&=6(X2Rk$UUaxU?AA>BQ;T0mNlvXV$c6TD<v#
zF4m{x<{JEZSZEgT&(+&er~Jl!(TV&`O(6O-rKb#;Fz<BAjpC8J?nC*hVG_=h=Bb(3
zr|6~Ylg|r?HxmNn$`L&bk^kKP+OqRa>r>XYKS!#38LIue?7IG?U+yg~f71z4>eTf1
zUR!MUXkHlL51^ngHajXZ{LC@FiZW!z+zP5#vFi|HYYw_JL3}5zZr#E&DhqBR29;G0
zoF|kA{^PFGVM*T$n|ynYe!n?)c3G8=`7)#A%Him*Rev6uGKe`n_{r4Q*SDCMn`;aj
z1d)jsqcJ0xMMd|hKi|{*!9Y|ZmX?}DIc@=PKK;KX`I0(RiC#czyQoVJz9k7JaRa}N
zs~xs~8s`ZuZ0~+=SAkPR_|9Vz2RKq((N)mlu<KkE9kpo)f~xiK##iGsdg1_+QF$3s
zWMC@A;ZsUQ=Zc94#XC`7C(W)v@}SQUC<GXZvqz|hFQ_3X`-;rvFcu<>!!sa*!T}lo
z^;fUp1O9#z-JU>YJM?7NdO<7}NWAvf=MNtu8)Yy~YW>tRQWuCD+K1^%(;vNE*=xbI
zRMlJbo)H(FlyIeNLB*1?SFh;EJ`q{(Sy|tqgG&0?wyx$6Xmr?Y!7|gR$UX(@;V2OH
zh(8g(`Ym+~>M-5(w_w>7V!TRP6Zs&WDuTvsJQMZKoioYzpv}atrldAeQ!!`$#Zt}6
zteU^11+xtaXK>ay<Zs*EiyA@EpF6jFQ^%^m3Zwt&-@vBGuzf$BAk)O2ty;CRbtww?
zfFA&9H}HJD0+bKyX0jvObZG+TP$ZO(8cz*Y$odz?3RPJm`j+PA25SvH{UK-20VD=E
z%|o3{g4t^}1`C0n=@B?0pB_|q5epiS72AefX{^_AWB7q%16S<=3rRS4uKV!eFFp)Q
zv#cnWHSfETN$;UOLn?I#yl)>m3(|J)b6)5f#G3#uG;C}c*it{#paheR<Cd4t-8Js$
zOeU=<dRyX#G{^oOml~bcVL1@sxXeQlHccnSa9I3Ku8Oy7TV4x`YPV6xt3^%&C&4Nq
z_j*k+p`oQE3VNu)pl@5TMiGr3bNej4kxf|uw2};{2KcE20<0^my)>(S|3p?96B(sh
zPi<^1kLddww7>$-zgjqGRAAb(;w`pOy~^|d5rUpbZF(dF7=?i~(A#Mr3}h_e%I_b#
z&Yl(`Z{NFjuT^iT48p6<`r8E4hnNJoCFQ<ZFJnMaYdZs4HgS#i{$jC{Kz>@^*l_Zp
zhWs11bE%ORdr+9{qnW1QVMwY6>SC_Rt7ZK+;NXh+y||@8iOiutK&+)t{j}!Z)vx%>
zK;hCj%JG4+rw%yA>jt({i)|E&fL_BS?s;*IWs^iySSu_X&vK!qWxx$Wi1JV@WSNae
zNsN;OE)wblKIG+GQsHCTwl(4L(wafrgT+g02%Mw#>-3D&)mz=qcCl`b0i(1lY$LlG
z*S$jEb<%)Y&@ZmxQ|yFG>AO>5-b_8G2f7btU1MxUjo$kef&(Lt8YV-(ns=#uYXB#^
zA0#M7v9s<t6b&BM)H_^yQErv5OhIhM-qlem;--DdnhI9sUlqpWCh@>b^7V>jqRt%4
z_&@4=nPaD8gru7VuT2qx6Ed+`g$MO7#3OG{5Hson;CHe0q1<N_;dR7bNA=Vhufw|L
zrukYnX8u~@4Ucux9})mET%hjAPxK_Dl5?fsk;rJrr`<<T8Tjln)Bv3rjFj#LkluPh
zcOE|uLY6dh_UxKhd8N-l_B*D3{Q5(^Vl`NM2yu<pI3=cRCF^A9-Cp@<-qB5u3rPCJ
zMY&uoV@LcH-=$TMb0}+MJ#JU#%7kcS7bHr|Awxv&1nGa3j3CSHRns?~gi>Etl4mFo
zP>4BFl2X=}tG8)$2O9k5xCPh7jIZTUJIt@G|2VGU4Ro>QjF+RNL)a|i5CvP3F?{2l
zC<97>d}Qlfur?a}tEeM_Zc{2<&(2PGql@6uGL6m_v}j+c2bwxT!rb+WS?<R^@~k77
zV0v6_?qzP-bd-Il8JUo7Qn_S9KzN5j8M7xPs9ANaetorGSM5=RAU4oPe}{J62>BFH
zv<4djfUXe~lki!xwvo%l6)}QHYuHN98G|XQD#5_00#UF^C(q2XSd4GAPr5e7Bsr>f
zSOc0dF`T7WWj7{MMw6NULElSh(|huup+iF$uhxut4PS8$C}di<&%XHLp=ndT8w4gb
zWn4cDggTAuIBZ28lfZVJmSCaq%qvb0oQ0dW3)Qr$jMHNvTI|}IL0AhQhPbc;6r056
zwkx*t(hx!X?|ViuB-)Ds%uLwAv^#h5c>2`t4OA$E_zu*jbHPDLt|Gr7<Os|M@`I@g
zgJ0r2BRCv{8nkuf#kszg%f|ZHHI2NFRmXMCAJnd@UVB^7<6~`nZfqh&9L60d^S%DT
zJ)e!1H{Ol=1L9~FFXqd60`VH>&fxp4M%iR-^%#T&>e1B8tFmKy0%Lkqba+gwZuRq<
zMK*EIcs*@8=|G&t(ZQ<(nMNGD$_e9<_e^|L?X%Q$<Ii~*z+fEApBG^g(`OS8kGxBb
z0(4+};^Ax1K}6%6<s;lFdO_gB>--mD86|)<yKdkG0R@avf+3~Y@ptQnY9nyB+o&VU
z?oxhy&=r$drcCI(X>WycQ6Joy5jNC=gm@LbbYGMfPC6%wc;Q3vxhc6~KbO`i$gV5s
zq)s;zb)aFZtR9q(Y*HV+V8MruhlTJI+e*c@T*={hI23*K;stA3jbXZ8r@t?vXab1)
zh=f1a(b9?n9U*~eHTXH4NEz`))W=x`=-WB+hc>hPN3dZbbaD$ZW|{@{CvDfcXFHgy
zn`H2%7VX4r?gzZ7?aIfs!mM11W*qaH!Hzh4<<wB(k$uta790H}r^&1(%~42*tS&?>
zN|c)HTgw>D@M*ol7tWr2jci1F<sMU@7jCC=_HK8(rnOut1={qV+X(OjeV=$%3JFUO
zxaxN583xgo@`I(MByZ5q$mYjnv}EN5ubNLXFr@PmcP<q6fWT0vW4VrO=o9q@X+Wp{
z+D-dS&R08Yh0SW&qD3sDv|uJ21$0ZxSkEB9?E4keyIcd-%C)23x%K2pD`f2i6gIpR
z5=!pG?Ld<adL=HbxKUXBT43n$)CMO)N;=bZ_e-st`7bd%S&uL-zZX3WO)QuN8{lLy
zC+M7A_$Tpt!w{;}1B5D~MgklV3Q~1Xcg*(^{bII)%MO%oo59I_xJHu8gi8@Q9P>;-
zET%yyIYhz9BS4JOmc+adoJmAk)xB54di6vPK}zbHt{<m|Mru5nlMf*S_b7X|Ypx%-
z4?Lt7LxT;5ulkayFb+L&NyoHIf$7)E;@fZ7c!hd2+X*lKYQ^HsDKFozxTL-_8i|Hc
zz_A5iemGrPA(Ba+OP#6T4#^<U!a+bfs_Dv8<co(${jd!|e9v<i)*MwVLl&1EGiUD{
zA@iW@Tm9TI(=K_0%e;B+PVW;tjT?68%%jakPmX+lxBBtxUrxEFi(7Je<_u`Zn0-G?
z$GDZM)zHaNnsk}!9$O80AQ3BAMnmTbl4}|W122@=@T=LUrflIN`C)ff&o(_LN9=!g
zG1kGt4T%b<o>+Vxmr*<pOIU88P(;!|4j+DiGzdqu`M`k}GcvK)04$vc3h<JtVqyZU
z6zl};JM%TEppLQ|%y#nQmK(P?7_S<+Jw=VuMLBWt<X_SKF4|N&I!DKNPu_KX|3}Av
z&JFzCc|w47LGfA`Zs`yZ8=&hzp#Mi&sf?$z2wnZ5Ck@vYG)5{wlLa)&ZvZ5fXZ^l~
zwjHw6`P6{T-ZqtGF}6`=$(>K5ZocqipP_L|Rn&sKBUz)Uj2_qyHr1==T>lnmcPPt;
z1R4IrvjCShquzegP_(Xe&e~3mh#4zF%r>ia7;+T`vc}YuBa#g~r!+?w@wDQ{54MKg
zpeF?pW=Tv7&<LhEo#xJd!qbE<l&%&ufLcn_T|gh&?a$_Hi;kAYX)andVk9Mt3%k!K
zP-^NOm_LZV`cZ?%VRM^)ei9iJ8tORn_NtXD*?ZVE-E9-c4|O~e3;4wOEBil1<VDx!
zqnPi(eon#R&Lt#_k9rJd<`k2y+rJ#44#0EN8{J5+YTtJ)hGio2E2ZM4Vt<+J+1!Wp
z%2Qywnsv%ffK?$jMY9aZPr-^B=PqugK|!)hQgCRJlQ1!i@jLf@`Sm5tlrkd^TLB7l
z-Pq%gcWRhG^RN-w456Kk)5M77%a%>9Gf21e=Om=(=hP;5EsXxu+%NfvZ~SuxKM}>r
z;TN=<Mhs(Waw7eKQ_sD7_FU(g2Ho0>01Dd($5Thxdoq1n+mvp_XS%R^R<IVj=d-K7
zE>IDSfd5gr%5=R~h$+y38RF4H2P*)b*bg!v@oM?z4S;9NU&!>a57Uqj@f_3i{=r~|
zkQDHU1yjd_>2LiRDQO^3jRQ!Zl?L*xlCF1~he6omR1?@7qK`A0E8IFnl*&@3u7Ie+
zPdn7cHBU?oWciNHyqOH-pI$!lQWxtu<SvABarolN%Z44+AYmoID)(Tkdvo;9;*`Ve
zC!-)i{^C7Rbcf7NrjFd<i2q+GV>1^@br7Cg)|BCH<TBbjMXiR)w0P=o3fft;cBZ4t
zsc&Wj&X=z|L++I-NhP>tduXX!tn!Op;X?6u+AVljb&ivr;sAHHzDbC*2%3%AW`q?N
zk_;xEq*7*07p^rNPVoaqArX{!23v1PAyywA0&?A+ccx1V?0<bNQ(<uV<HR*tG(9!=
zCd=w#u^UkDe|=<C<#N)w+SF2`b?iss*ai;?)l1h5J`b`W8o;Q`u<m@VV4iab$4qzU
zC}c3ijAge=*&z`|9elN_P7ZtJvVRWC_ED`V12oR|{xqSiGaK$~a=8p{aa+M}DWbbf
z<x5BrD>DE*o*G>^S-@4iH7YS0MQ9Q;%zO~#L;yaCd_?+MoaGwqNvk!Q!e<|v3b34;
zFJ2_+&@ny^s8`a!sJ4DPe=4Be@b`cdbZ@uGbd0DHS_3PcM8a=3Zh)qVymnLliZ|To
z>#}w(z4s9=cq_q?n-=lQ#S{~oKsCJ!s+f)XdEe|%3saf`={A}=cOSN=>yg^82{nth
zZI1Z|pGL(@=MZz8J)`<k`rjfGDvvr;f6sc4pOfOuGM-}^218OCY7YkrVMgK|1gZgg
zl@gAhFG&SyeDH)c5T3fwQO11#8#lBJFK$@SIj@(twq?NKSDz1eu^x|#Nn9hTc91}U
zyVpFz5<anxHcez!77=EQ0q;kU=c%XO-n?<61os8)ov)aLRIgaWSKOGzW^#sog!lJ}
zSk$atJARs-Qja`2NxpBrz5Dl@Vp#;}QEsr8CX_uG*26kii<Q$=Xa!T=dT!c3=}*z&
zI&VO`3bZwR#k=t3g9o>!{v%)EAt$Ru`5ygPtSBQHj6OLmcyfl43mSDE_!Q^LS`l9h
z_;;16qkV7%Jdnse@i!XF$O$<#)7#fF=hLezhcWW0xfHt$Ms?yUg&r_Xqy_;(34eJ5
z$?w#6TBAX8gbMCi4vPJd-(OvETpKg`=mVK})!alBl#cB+wTdFR9O*HP6D6FAWD_P_
zPkYsUWdYaBrlb?4-(jFZQ0v4VTZ@<@q+aO=jen&rzYC!gMcIoepi{wS7rXzo0QSMV
z6>Y0WLuJ$8=jQ`~9XYtNeVj>oaS0>FaU(UQ^=@=ehJcVT?|nLn*D*ZGRyCKf9O_kS
zLQi@N+zOdU<OehLjo$h6CJSF)Z05t9B}tD_P{u`w?Wb!sLQjXea5qL$V?RBaVLNgL
zB^{rR<fW>&ehZ>FJH1KGSQpivW(E+~&UPdPWOIZX8NDl_05QrKJ|MQK4uM5rnlCwK
zIay{qsk&ycy)FeV^<Zls;bW)9cYS33FZth!I4yG!&<i?-&%+z1s*(`NnUs{lFm?~n
z)=<~pqFfS3dWLt7VZ0fe|NIb24z8=sH+^=d=0bYf1uHJO7oW**wFn(NIc03bk5`&7
z_<PZwIlg!C$Cd?mDEa5O=^6W`%78<M?$QhdD)#dQxjZTu%b3C2sid#@4GgQvy1VKz
z4iN0XUW<u}2l2a#t)<QEkW#M)5$ENhLx(!T;IPP}xeo|FU-U$yBflOCo(eAM9I}?5
z=CF<=v?g?Xv~B-cH|i;lX7FGzyp8(wt&)XmOHmP*o(!<8{qJeelg%OE&X8BugZ8j&
zoEC!!R)KdSgh&{HU9?TWM)!FvD8R2GQxgLh=Vx`P<C@^X>^D8hw?vH@h-ddI2;54y
zGX9d_1=Y@4wrwkVUU~$i!!%Q!s~ow%u3Yhsa9TpkEe=cEdcaVBqG;L6<(aaGE23+a
zdnK2VBK_|(3g8@N#2G$`+CN_L4$=9VXn&QTrnr(ee9$`4L@0~r*oy_c%%P~zzK?6W
z?Vqn-3&1;UhL!)Bcxf41d3c?S31+>1!^OLfNljq6H$DphrZM*~<?MWuZhkH<4z9_@
z%h={Rf(X(f4<YJ@RU{71=So8-nwW&`e`%!?#P5Vj0qDxfF8#H|!s2odD3-qXjA+Jw
zWH4ZS=e7KW<)`9^JITMky+B7%GdS~?%&Yl7z3Mgz<qLOfuSXGQPqvt=q`O@wr7vZ_
z##Mc1(S{+8n$pbXv~|$3Nhc?s^H|`J6n}j6{XrjdXzUQ;?ow3zRntJN|Nl_teRtnQ
zrW`c2nNmEhAT2ky6AR$NJJ0R~#6hX#quAs>@VL2K^%TsA5@mDZ8m0&7+?=4|X7<6z
z{JSCXqM73c(4EN~5mr~sK(p_00FJJfHIj_82M~O)Z=4afeg?1tKiPTP`xajxM{oL5
zd%^zxzuN1~%9hV=@ii6Sdg?d`ZF>G-N7i!xbH*?WVgqY$D}?!UhZYg90%b!jlr^Ii
zJt*=i@RkDPQ#D{|TRG~(BMP6iIhQ_;a&2_*(Etv%m2VQQDkZ|ztgVjFi0buzLdl9^
zGpO(~KD&xK?ua-n2>+$euV4fIxLVe4`H)P-3h&bcQYd*+!K~63P66kVaY58*DrQSP
zphn9d_mdOi$q<Zc`sLqne<u)+`e>?^0QH{keX+FfNs~WXG3_OM2r2Q9d^pU%jp%F&
z01zre8U_t+FzBCnFcOD73J<yCM*n|Yop)T%`}_aj8OaVMvNOuwdnJ`)M5RH?I!I|*
z5fK?Fn>fcz84Z<{mP%F>GD=94B#KhWNaJ_EoX_{4-{*GD?RGxrkoWubdOol5xE_zk
zH6JCzG*`B=2;r9&rq_}$O>Wi?6{_5(+PQzIa*Gf5?wtEMGfYD$>=&$~BNh&f_=jjJ
zwgnJ$0g9F~sM>%>pzH~k;Uq)Jw<~*<<S;}%4uQh|SET*-9meAg3=BkXb)$R&gS5k!
z%~s+%C5$02PHF2&*@Y+9M?84=l&H%T9W8-^Io-h451LYab_cEFkDDQdK+MLPF_f#?
zW5>puy`&@<?DT3F5TsJtIY-i-rE+=6Irv#H#OfV>sdc$u-@;vc_ACLhD=_bQ`as5^
zHDQ11S#9cszyIz*vukeJTzr997c+}%&)8hLHZU3CiXc61jQC#Mk)fAb^tzqN6cNBt
z7KC9aWovGZ-K$|#<FCDX^w628(f?n?v|YJsE;<?y1C^$4e42Whp#ZzyL<B9<SdWCx
zSiyh$h90y5Sq8m$Nz8(@1hw>>UKt&U7EH|gmOi1jE=b6R-D9|?V`E(elst}9(*VLL
zP(Yn&3fLi&?7kz7APmxb)tOYa1Xj#@Bp-CxvepcUz0Qv3$^<rxP^V@|e*%C2Py>(t
zKU4K}>FQ`Mwo|kB?y%g@z~BjN)PhpjQSb*6J396>yNVu4w!d#JUy*;eximc_O#cjJ
zOjQ7(T*<wNQe`&_0{{p^_#WREK#^ffGO<KB43`<07ylUr@h{e*+tTlXb=K6VoxR$w
zzEB%#pRz1Ke^1)G<{zmFEnaU^@1Az~sZJ^RIwUFk<xBDO2zdXXMPA+#;58o~<HI&@
z5<RUOHv{2j8`Qb?YpoG8%d(t)ClbasblSw$Q{*2F=9F)3sH|);wS@-H!%^16@#)Sy
zcmy%A05oC1(CF`Q7}phroCj4vKq4bR`k9h-;&P$ftX;sk2_D~cK<||M{n$^zX0^CQ
z;nC1u|HFR9G4w(L@h|IbgvYV*Ej!J6Os(4VIcMGc+2&PW%UquJaC)}tZPT+?581Zu
zq^hoJ*yLTa)IKrEUF$`7Z_(1KcX^Jks=h&|#ro}9sT!!cDm7HwZ-B4o;@)Sku6mn&
zb?VXgiJQKCsd>_SRG+J#$~V+bJNV|{%4zq;|A;Ld*OHl?n!36hvIH&&sUd~MHBjxu
z5uJ0SGjJe{Zfn$F%-&=@0O+V&Uhr1?4Mj#?Kc4FKJo4e<7l?gTqb5$A=ec3okZ*BT
zl~!}+1b_cB&02*?oD!j~QPG)MH!GPx=p*V>sH>}!NkSrPDo)k)7&4@_7pM4;u)zEc
z{*`*p!ZIm?L=jr>JsxwIvt9jE^)fC~9#fs1i?}=<jF93f)MUdQX3rF;m4RnsV*UX`
za0M)_`_&mHh(%#z{t#|D>jV`yK)Xe?8d<PUzrTY@g+;fms_z^M3JNX}3bCaQZ`-Pp
z`8;iypx}XlTJ-L+Glub}HkdN2`P|$F(|_PJGTX=~Gv3_=pVZ{~EysB!&vz}mCr{#G
zvx>vk&L&;vS>0cQe@sz+Y2xpho*UlZYgOrHt9~N1$P$IFdUPwZ@Yg_`uTOdGC5j$B
zf6~Jj2yqZtJqXNQ+L^OAtj1STf`n<jpLzZu`Kj`45eAq`Y1b7US^o;87m{K`55up2
zZqM47eT>n<!eX3b3)(s-)qH=|X{)N9O280%_htK-6n(lp=?%Tll1Ilwoa5mlEk%)0
zH+1~L|E7fw#<C}tmbtQ$uWFfHHLxgh`ydx}J(X^?{R~V*>?<T*!hTT&!q(e(d5eV0
zb(B0DYRRik*=^BbLxCcCBYD|VUcQ1X2@B{wXu#56<pq~rr}|i(S-HAVs0cwJKq!e%
z&Yz0Nr&afi>a%;tm8*af(hf2D3-3I3!Og8>J}t^@TRlDcZB2COmlGq7HQ#ZoY)8d?
zN<s@yCXQlcO->YLGfob_-E%%wvG{{_G3m1-!-e3(KT6YMRvkNyaMduVEcBIQu&=ML
zbOLd$Q=Fs0w|oqPFcd3;pCGDrX3X#i%_lZ>&snX($a?szu;5_PmmE5DNL;G1sQra=
zU5K(~5j$!lC#-RIcX=9n@SBI{!J193pQc?6EloOiuE+%$+>;(IcWtZM+czuR3JN)d
z+DL%bIsIgnDi{HKTn3tTuoh?#I_tVGQCTeamgbo8CKACdZ>!%BJ7teCuwZ~r6IwPw
z28x>GatHPj9hHBac$M*}aq|Ul`tXvg+e23sODp1BXgU-rw_D9ex>9{&Pc-7#ZaVrr
zY<UqfgZqv;F|%J=zql(sSym>KD0nqflh<}D(oKNT`T6B+&+BOGAZItzg(nNp{WqEW
zkiUOOx}|_8P~bxkPOxyA`ZK&~^z?!D$7dFnCf+~+J+AC_n^ncnn<FY931GY;E1s{H
z-R&UT7ueXajZUI2{o1wv5R`$xU%`b64IdM0wL5QNZMy5Ijv&Ufn%_M-qZgypaOEV3
zJCiBDAdN$uT6pE9*lN&s&{QPeEz~nngF`kR?im027U~XASA>40zIinZ5RgxGxdMqZ
zb4v#!^V!t{V$5bYcbyk{>3Eyk@*_#25#-?GB$o$qPo_;ga*Gac3EYpQ9wjt{@Y3l!
z<ay%naPHN^$92zHNC<=K+=ogga9VK5nLgt*316%c*iTrc+w;TcTxZ(<FP!W(R@)|;
z9kRQ_oh4&eeS{k@ySUlE50@Y~2B}YRh+0wpekzmK%uci5Y6wpZnHPEZ<;TJXdKns(
zy$r-T;Lon+oBX+<2_pz?Z#;kGXg2Imn;=r>;^qE6GcDV<Od6K;YE%4G-PA7kFOfe-
zl3>Ir7>*&_28OhJ5V0w4IG@AD;I;Pt8$ZWoeCQ$zb%Fqmn#S>&+6v?<eOr$Esvywt
zvJ7mEIsqn7g%A9_u7=t&Q)@h_T%2ScKYlz$bGNFVPu+7WIg<A79XlF{tBEYdICS`M
zXc8$|R$@RWW|xJrjm-nTK*o(PF72`2JbJPl$sFcQWfYjV@uMG=Hk1E}ZbA%BvLvoj
z&<R)5%sZat4x;=3)s{~fFr>c;SDgZd#v!bCleDxnH_iiMlh<pqooJH&TJfccNlMYf
zCU!C9sdHfG(4e9H=|~6iV)+R4bCJKESK>d4V9+N27pTI*VVbzeHT?O|FCsMFV$6w>
zqei&^A~S)oMj?rXx|BJf-arYptV#cPIWh|09sU<q1aZZDpw;I!>vzB2x;uv`E%YwU
z-O4=dI}5wP`Brr4(nVHiO`Y19d6jYB#qsg+6*$2tW&HeINBF*8`P{nbp=>u&ZPR9D
zX5GxFhw}#o*EM39T-?os5avB*W@gnNcAvWYAT4cr!oTKkCtis<3rG9qatoJl!+P%i
zb^LLlK~qurF@|KNS{@+BpJgw&NPwwT&r97Vj6HA_PTN{$oy6=A3)kWmo!7l@#SX<i
zInIR)2Zh~!{P?aM=cQ1`Vb<o4w5*q<z1nRiJA@KHIN3F~Ot;>~dv+A57X0?mS2l?h
z*^t1({Pw9CbOWxS|FQ@xU?`etYFhQ{S4h|zRNwtJ+jg@Fuxq%sP9xmn_eAewmbX5w
zvasI#`)#Oi2uavxaFQ(M^s=)3ZM6&l9yB`CV*jscwU%XjXhp(`{4kjH7+kt)%rUe1
z#9tqbf(iSq*$RR{4|}t-feEb@pSV6<nxN1^jjN2|RkYF6w90q!2Vj6Il-V?oGU^#F
zdCFxY6TVfvJyicV#NqgyH-D#;e)<%myw+)i^}z>|to=_0_4(m3``L^1##H5E<t++N
zTyJ^bVkIl~Ld-377n?kUk9A5!igYFXM_8PLnV5GSHR`9Y(+(&k++jewmbxTCN^n;A
z(T4maw?Ol9Ke4OAA@k1Y(}_HoBd<>WIrEl5lqwMAR?_)ajz(W1Lr>g+IeoeX&7lEo
z7=F*%&Th}MPq$f=&3~7kN;LXdQzA8V&2-&rU#@B96@AV8tX@oOLNCY?@g7iw$)RKQ
zO!JcLW6p7(j(!h!SbXW!skg(OEUP#A27Z-g-bAo~6^pZgIWyGPqq|u^A&XqQoX;DU
zf_JSCJLEI$Bd=dSP5=(g_Oy<2xWi2J#e56L#@+O1OEoXbVb~7(>(-4H)(1o&daIe^
z2kylzo0Kee@TOK)sUiztj3wR*n^vl)y{-K|QnaN+Tb;gryF);L!5T2c{1zQvWJKGx
zR2*Al+RC;_W&{~}<;JlN(NqgEwJ!?XMp=P@1LKvTt5?<h?rfQP9=5V<YNrYm-zc<g
zKPUI@vrRSYWOd8x`s5<mnA!?}sS9Nud{A1*`T?-cvp#zc9Xd4g#wD{;t4_cH2#MVx
zs;-9*Wmld82U1~Ey<EP8%Tbhi7OY_Ciud<4okpLS`Gy^e+^mdzxD@@Y=WbOEJn4%h
z2{$l@{ZUkDb{?O4u3owum=c}iX^J%T*$irgm<<S3{m}ml(w!Y$HqNnaeDcJJ@U|~5
zO#9JYDmAzMY8o0>)|M?Kd|Xg=XFkqU{ibuYWz!Oi{>!Hu8J!IXP%!HTgolVM4i{rf
zV!UPWigR7#k|riCGk!IB1E6G{@R(ekzxY($&;d?iBbwvp%*VL2{)aPC<MwRWWTcA(
zO)8S?FC^_Rm#xc%vxz_1OcEb2W>3`y6O&fr;}D&*rUT;8%oQ(cNR36Q?>acgKr8ea
z)Bb$$fZn}Btc;U*)w0D#cX4hV05zK~{V2bK`$hs=tp$Y}xc;(r(#G+^Z|)QAJ`3++
zi=|Dw>5tcF)8-a<zVH>1Oh7;y1o9&*%;{u-tj2wNJXxG|M3})-<?{*a)cv(vih6eQ
zrcLt`f0rx*=hR(1KkL(~z)Xji3F}94jJu5(aV=td?*UI97lw0^8km(Bjt3Wok9K$A
zrQq~vr|iRY9^y?BvaCLiniLqQlc|$Pw#0A-UzDt8fIY<V9YyB-lG)vg_K#`#6#(44
z=3UIHg&RIqY%;Vh@C0C!H6#QgZ$vSd__Yp^Qw$8s;C$JAe-YX-X-#~>s|vIcq^Y+T
z;&R0CW7esSmGxCO&EX)5d9{aj#kph7@1PBu%_~Ll2RyzMwn0&od0cjyADf^pBW_Tn
z$u9W};qNU#=IgjljV~6ljar@XU}H0b%!mxAJJ(Nu+h~`0*waOvTLG9Wo9x9g4&vOr
zOF7_=7$e_<W5sI+LF(7GDW8sN55nvEhrZFiZzE6t7xRe?@FVWPTiu-6&=3Z9bxrA}
zLdpTC+6FbY@i{Y5>{DhQI&x%DbT;ET`9m1EGj_d4Ef4gtirA!T*|qp%-By6mOZ*`v
z*DGd%;UZuVuWDhLBKB9uJZoacNopOzEm}_a*;RTA`(eCK2KczBD+&3Rw=uPHu?AI@
zRKHO26?6@E39cFAt|PmY4U=fxDJ?@g|G3wEug!=lD<Qy%+^?zvmx9Q3iA$(3qWZEw
z39F6Zu>%><%JMYqOjuPW9T(0F(}6;11L;M}Fzl&Wa@veC0y03xO3%4-|2>*Np1~GW
zqUW_Y<gPxk*YUkcDoEj>WabVqE7AG|1N#vpwUD-^SFFOh$*{X56L7X)d|89WO-Qp+
zW(`dECJO@&=SHYj9x@Ec@7Vxkw<vVmKmYubHlc2*&8x#doj~g>S6?R)9_{jY$rbbI
z58k{G*OyH-#q)FvPCO#9Nf$w5fcOyu@S_M{DaFDg-nYlTnvdna`x*;3@A(Z({EMTe
zfQW?Ru=BEu*;cy@gO2)L1_`t1b8ch<Y}I*(vn?z>6h*{r<u#14Em*P-gr`?*A0ueW
z#l`reZH{}I)O*>q!mV|O5fH%A;mtE~AwR7JhyybhE)0iU78SvKfyp7$5^9jVkrm;k
zFD^|s$~2D3pK|_L$TBbJ@5O&F-(+TcybXsTV%@PizX#T~#SVoRl6rRSs-U6>)NwG{
zcJB84LAd2&hMedzG$h16wJ^L7qheY^*d#VN{$mPp8fiOl__2pYQF9le;d_`=SK|U0
zl~D+btKBHqt~hY?#DXh6hwLe%3vf7I1}~WYhdBxeUm!6N*lIi)tMkcr(RQhS4+&{n
zPNa7H9O3+##KFUHNWOmV?%TKXa#~p)Te&N=cuSl`VQ<X{R`Xm7_N(gEy?f9rhwQPX
z*AMdI=OzTi69&bTNLE2*<~ozAiMDHZU*m_$K87)y?Q`OWw;cb?IB3+=CL~dHifr;A
z1QpCjP60MZ?f}>@q^WJko>a1yJX=aRz!*5W3jk>?msdPBYYtTmCts4^@(~qSLX<t8
z+%hvyW!nQO*@pFKu|sxu`QkGWW7S8-7f*osqIAZY4S2+kD7smcAlq)`F;kO=%!E|9
z2jY=f|DB#sm67Q%6}R?4CJ=J;Y&U;9l2pA7v?|p7=j6$gg?5Z6qv*sbNI_|B%{l8b
zuB%A^T3{Yueync{_hVQy=;5-h6zlV5UQeHzl50kLRt;lVB2qVMPvQpip9KJ>m~<5T
zelb70_Hl8gZo&7|0|#0wgiZ_i_od}DpSK|Z8*=Y4q*;Og$Def)C9hxha34yc5Y`j-
zB5~Xbw?41zZq%y{N`TY+22rj1`Zn?P^BavviA9({Gilj;gSSI7Wr`!z&uTQ(iz)D1
zzg7=a+OLRIJ#q=Yj!e$Y2IRMF(P9ZyN(BQd7;Q4(&J1**DWN5gHJqTe<5n`s@)510
zS8AKj#DL;2ZLNthn`~&kXt%|HBDmk9STH4?{+&g8#~U(-6*VgbEvXlv$%vg->e)@3
zH#eVM!vHncbrel7L50wE^zS}uCUuZJAtlcd3d52mz4!{~`0YF^&EW>T`g`_BiZELq
zz9{MFenYY@DyT9Gw|8*Py^=nmI~-5HJ2&}!_n2P=25zB|H@2|Y&Vd2$kQYitB`QBu
zRe5nrjl0zz8Z+yOYfmg20zINP(eg*F*`->SMp6BX+~bw%Z^kPaIBUYMGgVa#+N4FM
zl0p^8JrQLT%+}0TA)y;K!Y2$EGOLiDq;b=xVe!elqu@56Izk8|_`2dmv3x`aS=xb(
z+uukSX|r2%IoT+Jt3bbbiN|4$K%Zav^1Eac&_d)nrWJo<I^_nLtVPFBT^o+Ey%3Ua
z%}RxT;Gfa43T(z^SLvs(XK?Ws!!*SbJtxyvUC+l5>&?md8J@H1mPMFkR{01f1&i8l
zI7c7czkd>KCk9#%qgYyjEvtAKBD0kxc`q0D(XVe&Xz)COU=*CKveD&8?SsL3z87i^
zzk9b(meErpDFiCQjnpJ;kD<Q4?9IK1mz$aa?&*UdKf<*)4GNdyYuU420Yi+(VKmAo
zuygkgS#0`u_yM)K2elmehq9AxYE@8=u}c_PlAxCWumR<z70`bxfF@Re<#3R;S%jVC
zhuQFl((#2GI-HWNNbfNgNy&4?s}7TY)-bh#8pV*)CUU?{O~?IFeU}wgJg4Oqzd%_)
z#%T>pu&}VW%iL)1t-Se&Um-zMfBLas%Q0$-?}g!CDh{j7b^K}YoarA&t2vdM$*E`d
z!=S&)E(M4omOuuy))-e}1U8in!^lO)cMF9nl_j<EmA~(|rdKJs`6Q!L@5{Xmo@=h$
zKYZ$wuD9#X>og(L@6mkiMC(T@-Ml0&HdaJ(ps1(q0&>{1AaX)OYb2fFny(wewww{v
zqy&T@W*l(ENzd<d;R&cz&kH>X6c3;)Zao;|(~%MV(aq+oXFEK<@UO?7#wse9hbqoL
zn+m&d9{1+Nt|>ls=X9FT<}1r`z~OcyHFr}^c2ba9&(HkbYpd#RTZ5d-5i9?X3t-TY
zH!HSenOoxNw{l@`N>E^|!h+iAlo&h%Sr~xxIn;VWGOa(8J7J8^jdr#@NKO9a#qf^`
z7xR7Lp{hGoE%Vu)ZSt&l42nkV9Z~E>Bu?NTsBJ^ud~$vg*pLg>13Nvdu{6>mwY1#S
znaE687+IqvH%WstzK*0DTU%AUIo0dMrrL@wD3pD{V9@bWLvzV3oQ~WpcnwL!K`9y<
zC__1gtijbmMqZQb<9P1U18W?xGSy#MHfh5BtM7&VM8@1C`j|^&aRU%bX8|tKdxK29
zUiHj~bn%rVh|>3#baLS5u47J6(vZ{$19IMY3&_>yL_2o^5)OCk-*>BhjOKFY0bhEM
zH`XkerN3w&dPI7j;M#G7G?GE4lZ>UI_=Fg^QSUPfogM9NHrnl710f2;RbHNTY68@~
z&^?Ak)xboi(&JFMy=LT~yLg{FS%Vo?KHC5wG(*26r^9x2g3muNdk@NJ(xk)5ag?bX
zO#-KctpPiFMCdI#8f|fQDE3sQ(W7gl;R!0D?p*-mm^ctu9`T-{(hM2tp@glbE=vl4
z`3o79;-GPW9^p^9Jpanei*x$%0`p5tXEEN4LF_UY_f61d*E}DJ%;cU1&k@I1cmMI?
z?7F65<<D#>KtO$WNQL0(Ohp^@%;ZTAP^mYZb%$jdt=qJLEHTe3wYoPMI2SrueLU#J
zksa77EJ_%(LQ4Bk7Mr)dg9syEs}t3FO=BY|!GR5Y0LWduQm@C+`16e@z+mLc#u0`A
zME&DwA2OY&I)Jmub>936a7!QhRi5E%7{ZsME+4*sJt{7A!&&o5uhYuDeQVICa88RI
z`M#TTA)CxAGu(#GnNBt~^^M)!%<37`3``^)OJ~nB_gup2Rh!ob<J>>J#v_0D{Y{%#
zD`OGnbYM}^x8L`2a%{tUnsk{PZXr(0%oQ)aT<$Q>+uIxJwygK0-l0hSivYYPz7t9%
zNI^%fvsvswCxM9fXUfFdWX44}{iS?bHR3+&3F{L01CQFYDDOimZoYBKlT&V=HiVCu
zSQil*m3zSD3Lzml`Dz+IhB;jvN|`$(y&ABy*K%4akd1o&VI~m5fOB34B~Cy&Ax!Pc
zLZ*OOG5P-<IB=7gx$EJQ`h6zef=Yf{eW1M*GR>*5$05+8iKhc30N)|fQU)^%7`?hk
zjm*Z6zD$1u>vrwS4&G3#-ji^!MZ!uz-~_xZ8KY)7sLr@|y)wzXraxCW^5@tik2Zmw
zs~1iNl81*=c{wtcr~v;i?Bl1>AJ&VOeLh<L;&l1Kq~3MS<)4PUu<TO*I{g~kQN{Nt
zO7+kmB_Z@b5ueiRiOye4SWQee(>Q2eT=S>aq2mYtW4rZsu<Y%C8o{L);5>gHX?|>>
z*Oh6WHJ`K2zN)TWX<nA7d@N9@k9Nky>aX>w;Uh|X?E}Q+n&xqHYWy!D;LG5<?=;@#
zaWY_K3(=|~(hYll7Z5^>-mEqc$9xH4NSvMdNDV!eJfPz3D0SQz+Sm8*YuK?xBSI$C
z-%2=O{2z1%QUM-ux`jYmHsmOT^v7Jsz@Q<dm;$PZ)x*`?)HTztM-^?k*r#P_flb`a
zU(>>lQRZf>qTdjX5^SQ3+=xIOLPF59Pht=tz7EK}S^m?827urQ@CPZ5T184Q>OOwy
zP|!NqZ2k_(w%zpe-yhgl!%1-LHy6jYD{<p$f%Yx0A3O8J9-UqTW#tj~`!{^|s&}70
z?@qYCG1=6a;+*Y5#sd$ry9>a;tm5l6_#B2>RkXe$fdKtAo?BK8QPMoRm*CsRqc^R~
z-Q0*kMi0G>6VAOiQQHQySMC7dm3a8_lUSr3XV?K!ovf+uIA)a(zXGl&kYFA#!QXtk
z1FeG(-@6_)hN-Mq-w(T$GmgChzaDAnQ;E?e5UK6=;qYI=wl(nd^h61%<Euw%2v3;h
zc{Zz<<t~9SF){a@OrASLF-B7O_kH?s$rZo#W<Nl`7@9bw`b+P{*;UFo{=IS%Bx-2<
zg+f&OejNoMgN)`M(MB%3H0LocF>BWb|60vHWL!8H$pYW1sHo&#O6eR#kw1JwJ{8@l
z6OVfRdP$}&(z`$(a3uWdDs^y{$Viujs?S+H$(WUQS2{f!%<r`sQo3@k@Af*}_Rh^5
zz%EvsEZf#by8?A&e{|8AbD%5&deLcyB(>?#Axzo5e!Y64jtg^Igv{j~;FDQWJEG>g
zlGsAchYgPldOe*&?7!OZ7vXXlI%pmxFZ7-3C^CmQgGrnOd79B{_U;kljY7?fWUdz>
zn*c@D8x4FkGH2DnriPh6B~rq{4;a$U?QC@%R7C}<@#-bkaDOm}u`R0%&2)^KWTIAi
z?yuSLH+Ixn&vlB<K4wEIOurDQV^se3Es^ZyrFkCX3-Lp)9m&*Bc9+SL?98f#(b>EK
zCL9NGBBQos<qJ=W(@U_ULJRKQp5h`S{rkMSDoe((6X<jYjGbU$Qgr8qZQj$;u|V7J
zYn9Cl$M3lH<o(?inJe>0+%3yGcBTUvGNNRQy`BgaxzSC`dbR7q!fbAIu!lw(JZ1hD
zuwvUS=y%}}b2g{?9VHK%hxHiGcB$%`n&P!R0Gu>igNNWX&Hf6pP+*2bWx?z2s;xcc
zVdQUtlo_yO%<=EQb_e(mzwnuz%*YW!0koGZv{n?pZ{AmY%Je@s&ZQDdbB@`Sm!ixZ
zGUW2z(vIWSs_HtAQR#MDo#ZHP7L2uZ`u8_KHS*X);09W``;Q)-1G%u8&!QKBd}*oH
z#M}1&r;-22T*B{PG_tnC`*qveH1n9E6z$WVI}3Riklpf&N_-4I6CcGqs3@H(MSO_f
zQ}-3EM1xdhP>Gmfnwi<5Dk}~=te^A2!76N9rkRs+2!-bAk(6Jm&0k^ZM4L)GA=Cg!
zTxedIgd7BTD`QfBWp%LyIM*Z1l)HH46R8zQ_Js!>mdTcIXR&X|n!N15(emr&K~G#^
zMsavgiIFg!v+0*gux+SO%DO0pY@QfU?nM(LB^(P7%UE6E5ZcbbL_l*eYJlKO2lgGY
zU3n*CKE4L3&?bKv<=2in2wj#GxBAFF)xc$;BbDAz3IuNp-x}Qie|>$p!t<HYNL*rM
zer3GwX7lFF??CnN!mr&8UPovOtRInZ;Xty3u7qJRyk<&l(;wcgfWHEAs2zAK6fs6b
zvfcv*++A{I!H|FP$P{6S)`$^@$s3kqW$zyeKw3>wy0~*-awQB8OfK)=vF&6}ZvwCK
zBJGj%{G1o4A1U((Nr4U0v?l%pxGLHZ@F@V31Q4%@HQ(9@F4Dbw{oolI()46HoTbC1
zGf7QLqarn~AJ4c3N$*qCOlT6ZxQdB;tkM<qk5BRh`GIANSt@|##}qEeTSRSaJYM^-
zYSvSt+yio{SE^G4bMA#8yPQONxyY<A>*nCKhFIf=c0NYa$VY?opj*)PAG+1FZC9m+
z3`d&A6A>LE>5y4JP+}g|T>mY9&qXm1y8rwz5-~U|$9pp}cXYBeHHo!VjwNx$12lG@
z`^59GAh6315p0~28aZ6pcMp^U>_olb=b^n-wQ@(G4px0zDQoo;TC5(`k*Qj+L5r}b
zgP0t`Q`k<N#-vN009%QW%DrSKiSWv@Sr8&zIhyKe9=*CXFy345I(UhRb}EAS%<>aW
z-1^ho1Y0W)|4(m+%IxUTDd`ST@+3)~NO;DT-|Jq~&>Hx+sm1`auzrn{m4^_hB`|qK
zf{T_d70V(bg~Sr)`*7q1W`j|5Y2lqF%!^_*HC(|bN7CQ)z^8@L!G5-~gWga_%<j9@
z?^U-73)w4AX}yr9K->~}v3!U2|Lwy&=;f^%ICZw#Htuj4mxqCFGc!AvB}<ovg|z4W
zONp60jlm%3B9nHJ-`lWB`t$49gPgjnR@o!ptnCi5G0EzisY*A+)rb;Z{~et@{_?a@
z-`eJ#Rqy`&3kVz*f7&jn(^O{aUY9Qqp_xUiBcvD5^RUsonf}T4=8RnD|40(-s<>Lx
zNQ+j*B#!wmMp@gfbk1JAdI$rFcdN#TavHLKsN1e5_U{>7-1?7H^<VXS=|NDD*8}Um
zaLR_Dpc7())XP*g>;Uw_LsHeqBGWgAPW%#WH3^B!_JOA?i!d1IFC1|UFeBjl^^PP9
zq&B>y*DKQciwn_p*3&Dl(<;I&(Q}IDC&;2Wc!g74Emce=C(KRmzzjx^)6O0ryTNOg
z$YcK`a9db>rp`JK54+4VbB$A8aof&J;E97>`%(6hsNv*&U`)6a=N&ExvlF0MCIGEb
zf)E5}Vc_DGYMw+*bJHQ}9lRNK$MNsD`3l+t!=R8K`H}s{<kWe3wrjTC;gsJwjE;uC
zOiw;rS3;&6@OfvkS;HYDH`>erB}O@p_JvztnVR3Ja=5z*QPK+Nv9c*4PzPbH*iX^R
za7WnP`7e_ai?QdphM#ldS#&0Y$7w)k6l(La|EJus`Eqe%^tm-N_pJ-##!fdcc#(r~
z3A4)^6!lodK+FUo2FU<34h*cv0|hXnM4)Su-FgInSq>|dZQ3#u$GZx|b|s0wouRPz
z9r>7n#~Bkf1Gm-M0Yj2P0(C^s2-{yFpamXFGpUxMn(jzLwiD^+?5q3{cJ^X;QMv^N
zg<{f#m*9(QPREWGe*{I6DH;W((C`%m(iEw24>}UA_=A@n$Q~s1FH`YkHh-kSEi_hZ
zh-QJEDR_(!K0p@O0%07rh|2L_xSG&2ghWYu<jN8nbL+)@BKrRu{d_m?(#mn{|KaZF
zOfz5o&alMUg|dv~cI@@Tsq|%(07@5ME@z|>A2VfjI%n5A`Eifg57(_-i+h=Z8ipCd
zJoELS&TupLQ^f8&d2-Lo<-IT}XS6%~pW;<x+(vZ-N`(*xs7{Ulg!PM}(PM5UZn`fL
zJ;#nXX5?fYnI*GKNuiEMJutXwx#QNPsZ;L{cf4|%eI!MJhnP%#UK3Bf8vP$Y594IS
z{e_nZcSS}AUL9ZbaaEjV<<KHM$V{3XF87eyV2>fYl=B=W-fO<h^bmsrZq6mEaK*Tv
zx!!nZT)~f>mKnYhg`6thsf@4M4f$I|0RsPyhdX<sB@o6E!GLC7po>Vbd)q_X&^UH@
zOi`(cdAoq#P1sGvjGP!@>3*jtGm#8kL$1TY>H*J-D`5w-7SHQep6hD&BAdx$v@?4B
z0r`S1Ui=!GJ#FgLJnU9-;UPKf%c29KghthpLg06}kUN)|DUWk7Q)UKYPa)fgiX5j+
z^B;vjg8f$B+|=lr?bHw9t0&jH9yO)jb)(<cdn&!b?4PwVbgC4?8%}@7e8_szGiMs?
zxOG~2jk<w}`zVNFm{8pB<o=o!Y#rIWC`h?rF0V#YOss%I3ytBT)#$y`y)5fpZ=LIO
zT<ML+!N5sAjt}!(zdpBi3DNj><ZIXZpf}9jm>g|?xg4q6UMe8BwEocoda4=7+TRV-
zHV$L4tX+VGCV)x2gAP^X(S3|9rOHMQ!)_Bp6SZm6r;mn0C=>ekB#4kT4``o%(Yqs^
z(Hb~#FBBk_>)xDLTh>xzmo8P9{c%Wqa$>mGUAuRmJihdmWNDN7@fiEtHZb<zz5v#x
zfR(fq;_!^ww6&9)mtjzz0C@P<aR6go59#@A4}Q}?OcsiZ2KX2*p`(ZLsVw}R&TDg(
z1mI<2gaDnIiU6}L`&xeh5uga^_JyQ&O>QT3r%vo1kb;a5>KI`(DhNtqg~AkqN>Usb
zMONLvOC<29P^?HR9K3@kPd<Ws(QcUiKDC7*4Tr$A^f@Q1LWSobe-qk~%Iv<gFyvA_
zGs?Vz0@VA#N4IxFJIw?edTV2`{NnYQ=Ko{u9J?*PBo1Q)pSX^PRRddvQ1YQ2i;rs>
z*otjDp%!i%Sjd9_X+NV!QU;A7s+HD&2lUuXMP$;Bkk)Mc2bWF1I|$aVx%;mAWSMks
z#h{0!D(M_kB_VrK4a?`p63ChzA;?HVopM+NHC9fAe76rD4gmJS6}t->YcqV8;q(to
z;-CEN?JKq<K#REmfryzqP{HOnM;?B4rcZPPqeoZ<atAdurT{B2Qx>mWh64vM4RZKW
zT<PJQP-#)i-fWalrdhYz=nWfE+3QB%+ffUntVBMWdC-{y&9qh20erxG*HQR?ou?rd
zBb@l0M<`pF$JFEQE&=w0U3SC&mVCe4QK(v#>MMVRe*&6f06B?vmOvviY9!Qmnb2k0
z_5#U(vn8@qk@r4)SO8q~J|ey()bf|3@T969EV**YDSHI9436dqY<p6T1?c!5-FUiB
zL+?N9tPau3(*(^-+F0}Pb?i`?Q~*t}i6g-9(Higx`hzLp%V)i!Ro@+DB4xj-7E&~5
zQ$t-+s4y{vi^Q~;tV_F2<|{xTK;eCfC>Esy@>xjleb9e27p+J(%dPnNbwR+2p8M}C
zOkQ|NEX(k;m`r2~Ey&-^cpWDh={9jrL&`{fB7PJRnp@<dbQRE3%-`@rK1>DHr%#^@
zy=CTi2Cp8|i6jjRjMtz?xi0iVdcVGKTEt(75x&>w&!08FMB6{v!kk4snwu$4wrx2}
z<;lINUY5i9!1pp;`B6}ROgv??BWwj})FQLmKr6?mmVAO|ynaGZQE{<2Y@*WkZ#tzb
zSFE}Vv<N9xgmgr2M%7AS!J}0VuN8zk0el#=x(4HyknW}AL%~FN5mjS*02`<R13K>H
z+nFRWXZs%yT%5~zqP(l%ZRw#~;(_}xcWDizO=SnVhnw7f^ncPeV4-r8BY*b!rr1Mv
zk>}CCvz$c!!_+$Mt--@9C)}_GY3c6vtAL;Ie3Fa4rv<)EQeIUSGNKYbF16oSCfGnH
zR3l#*+A=#Z{^HI^;?dCbaow%wNM0Wcbuy4%nfA01!H8)*kh4R9Y{t8c4gByfkzfMC
zmB7zL3z*`6?r@zwznN(4JyUj2ikl1<-$3y=qBfXAg9ijiwKP0<{W48uo*@~Xf0K-t
z7o0I>LPjVJV%C;v$)Z*fJX!v*<AwY3W&Qa9tbs4`_y^26_)WvL>IMrKYYQT?Jo}o|
z!EZf}+sCOH<Owb|aUg_~?lfAInqm>JU~G-lYAn4H;Pd41-{C+0Iii03>82~6GvP>w
z#-T&^xO{}H<`&b`Wy^Y_Q4v0|tO+-%D0{;Y;v0%a8lYXwbj7U@ns)z@Lm~(N3y)5}
zW@&V`RM?!Xg`yYa<P!NQ6RJFI%#-j(kpTrql_G)RuTVX>0t{P)nN0XBzg!BKpAPSI
z23Ane87^EYYuv3t#ubGNC?2>u$i(AJ+g8mVh$9TYp`KZUFShaPWnU+ck>LZys@TNz
zc{a-I&m?`y9)z!PXWnuTZ<o>)w;S&vLNY~HP;b<6hZw#t;4`BF5yKO7d%BC0qlk+C
zpjII08xewJ+Rv`rh>nIqS|U;iPGbH5z>!aO6eTSK39tpa%%W1WhO24V=y}EYAIoe`
z2YH=4B8nM~`6zsT7QamaSSKSqL-ayE((c`EtLlK+F7<$vT|sBV4!)<+9?aB1tOGZe
zIls8LnPI(X<iQ|jjjvRLTjjPm+F-nE**k*?)n6x977Ac1fC?X{%YRHfidN7ua3QZ$
zo4-hw=LvBUMc24;Wjkb)L|uM_{9|fohnBE0*2wvs8m0MUUNGSbT=44q-?5Q`S~zyh
z7^w+lM0~*ShvuJ>J!2w+MrnNk304x@&}PW~^{ZFSYMi&HdVvAU-31xqZSx4g8>w))
zY%J>Kqz?E!fe--SZFf?2!9Z9@hHaZR?pD-rx9P0!K5(nuaS+h(;=q@4yu>9wdA_mV
zeDl@AVg682DY-J+=JXUFm@v}p{|GuAI8+M4G3ZpJL3FJed4_8%(p{Mr_M*jY*t9L>
za0|1x!8?&Mi3tKjIT4wyTJ_C5A^QH`dXZVPjg4Jbd)o(aT?~U-1>4;DJIc!Tb@wcl
z|G&6={Haf=`iFvo;rE?VpqNtX6HPLzos6b>j?s=ezz;kB44wX*u+oF<?uG~uwy|&~
zDrZ@BjY)n$dBrG<w*+iZ5E$FzNzPSMh(`unLFQF#6o%@Avq$!=bIgf(HM-b^Au8Ba
zpiUDgKT+Q&RQfYuPD$B8nd`7Q@|^A?>s$bi<oWh$S%WcZ6v}@`geiz~32VBXVqTX+
z6S&bqwVSdWqFBfCgbf^O2B*)Q*$UW24r8_J_2sW7u*i!1^YK+wUkb+4AiT{z&Pq+d
z9LoK?y2A`cgk&Av#m!H>JvcIHFx3kXF5FTpr{U{o!~BmtH#F?wXrVV?75w~F(!1Z!
ze~q<EhD(u=GW~Erle&*bWKbg69P0BY{v5DdLU4w`NQ3be6q{kmIlLC<#9vE*;`G+6
zxn@<Ck(FiHykrdmZ^&wT69x(fdWPdI=TmIf6f%YZsF?*ALXl3xDDOJAurMV&`i(D(
z{NUvjK*iQIQ4>oBVPAty+m1W1-AD?eyfHGmXIAn~{^rmyyODKs*j4<|xokEJGSTR!
zTM*cVx6I6_ci+DGlwi0|OE}}I?>3Kt5HI<uI3j)zf6pw6BJU@7w8_oERl6z%h~!Ak
ziO_f#1vhyyxgAck(E3o%&(*F`t4tBMwQEitjMr&9fOQE&16~GN3$N4-*O>t6E>6s0
z9Qk|Xz?&prpDGz%$_-}C70ux<X3C-flL;u>G|xb6Ib!G07(7cT7I4PJ2n+Q}GdgN<
zuZ7eQ<I%49QK2XWmN1oo@gY-gekGW$oEN@1=Y_i}#YRAcT_m+O;E8Pds-M7SPH!L`
zvKC{-98V0A^fC4N#9CEGuU!WXdh9>BEg;+lD*D;8XJ@9s>G0g}>G_W7+Nno2<~hH)
zKDSq@(Turs+f&e7i>oP<G8Z3JC8hynMq<tXFbp1cXLy$A2|v6W+);7$|F{5lP03i6
zvzW|67eqs~^Vz4r`2doaVA%{o9Li0EP?*mwcM1MmtBNvq(!dV}(V-ubJq+w01JRE2
z&}n}7&olUrx#kL(-%wg~c@{KgGQES($?%I4aQ@j&WJ_Ff1Cxxa%{cusY<YVNA3<0w
zpj%9mq=N6zVLGS)1m<wdWnM_x(t}4T*$o^Rnf5^e&|n#zdj##y!H?6mA2ZF4JT}wU
zLs+v8iC4)F95*qj74SCCHF)$PzVuxV2MZm2{MPs5o!gA1pb?4$xxTr9$sFX6keg)H
zgpdwFz{Lg<XhO)8F$IUMW5Ak(=L~EM@DVlY*w6gGMW7TVn`tS|WEByR1Y<N)?>L6O
z#~z-GcplCu2B?PtEAb#xJs#x>^lQ*!sgY^EKg1PB<Wo4LWM|Vd%viTBHf%dTGs}z~
z|L)zpp-N-JGDCYM%G_>@$)ogiOL^n1re~&}TTCM%P*8BcUe6Hl2)7vS_(jOC&6KIC
zVmRBfyND+_pyU!s%IynILA#(s2>MJ(EI`knItz6RrGfC~4jsBpk0!q9G8kakdz*fc
z?Ul50h7?gY<98|3lgp3dWs0GqfHxcvW=$ru2azZsKgrBOB^WPn`(Z`SGr?`hGv0}`
zRV*?SmzO_V2u8X@ocmP4=GJQ!5_tSPhTgi)6V6CEVsW_>9YfVmuIB76T}H3a_uRlW
ze7>So-xIJtWqoPF`PypiQT>JMn!PGCRGnz$EqEYIH%fbZ2F(?@-Yr8XZi1aH><%>e
zPoeXZeHb!;T8k7C#qTGy*v7H1mQkM+-MRF0(VtgoiFoNUlMP|TQLM|6c)O`?JVvEM
zsfEOy4l_`43qUM5nqA}+z;h)#22A7(6`UxE?Xs6c`UF}0iG6bI_vj?{JM|nsXRE4_
z9-w44wv^ly<r}9b?~vsq3gJZ%qZ{SIVNJ%f9GsEWM45}U*~RyxQ`^-uk%K}1Sgk$-
z++prj`&NVzsxp?q5g_G;qODB5lOS3~hk3juesiNnEjr$yal(O|Tuk&^^pkb3521oi
zH8Q$mKTP)YNgP0iu=nAT?rfWtMhN8IvSm>6eyOml-TtevX_C?b)Tv>igL>0(q+<CI
z{&@A66R_g|yg1HW<y(0-m|0Prv1oxvL--c+O*8}OyiWN|(b=jRA;p+D26l;LOEChk
z`l_tGqxTd)OYo_p3UioG_mIsMU5&gsJn`qdrZ_P-HnY%WSlpNB!Q&3pp^kdS-VPBY
zf3R&WCnuh<b_Knvv7MbLctzbLHSNTW9SuyV4AlzLPo-d*#W2uEXvLu4%pun?kmU6r
zIl54{dcq%U;;m3*pj4k^vb&$PYmm-4EMyesBUA{P4w*M1gs;4Ui5n^u1PY)mr}119
zG#>(!$QLi`S*mmko}y$oabi+h+c3pHY_b+!Rml01H<ynHOW%vP6TW2$Gm{>x8wh^F
z(CRCLgfi$yrV!R>(9;>cMa0$6or8hg{*TDlm0^=y8$mfe-^J16ck%qziWtO*!r0VQ
zkGx|1%jGGLMV4gc?-?r{rQiC#0@`$<0Ycn=2W3yo8<Y$3X(?Z1^&jXeN(knY>`PEp
zQ`^dr4Y7bgIt;61{+iVyaPu-9Db8I5Utqp)8#=;I*m&?h&c?*dF^XnAC<;YL=FmIA
zXi1I`pn+RRa{ueWKvNa0w<S`=ez}&d(zNNyG8Gc!MgSbU;0Hb3M}s>sksb;mQR@Ad
z%Ngs~mHlvh#^v~Dq>J6|!mw)xZFVE(m+6Xw6KfFnVJiv>!M2CZ)9-$EfEPsLBFw!3
zHZfq40l~q4C1iJK)k^FwD2VaGzL}XLm!s&lQ*_Q<3Se<W4+!uQUubMEI<Nmc7|NE2
zwcrymGVdVz7*JD2i5qa3VyVz6kPVH!feV)XQ9uSyr>Ov+no#wzJCaiIWk+2>u(!)E
zS~Z4}oJYZ(U|0bfETdgt-{9b2R#N$p+E!aGVc5lqDMUoQU$ZZ<v|**zjrFu0OmZi~
z7NXN()ags*ubE`&zbMn6pNQR$HqOL?0EqaQLd<4$JQxfEk51>G{Xum{ujql20nTD8
zT%sv==f<t-&6!D34ziyk^&E~dQ(!#uc8V-I6R-$}WFO_+(>ppJ@Y5#y8y#TH67G@@
zj*o7`&m^*oe>WQWr2r|*-(3Hf;!3(+dIZ1o=g-Gg4o6W9@p4*i)!UG652fz81{n&x
z;}K)tpIEgp8V+F2AZhSHrb859Mm}u^93#MB_|zB#-%m^HjujI<d*$dq;|SWP&ho(d
zv#=Sq8E-~wm}R4x%a_-|g`p*wKxmm2C4tNh-Tk+_#N;VE^LJ$-IGW?rlEPD)_HYJ(
z7l4@!Jty?*w~H2W1-050vK2>+C?4CZ^rfoMRJe?t!ZN2(kdUkmOn~73YoL++o#iJ8
z4>h7Bg3OsWghHWN$5Hzr5Z5y>S<9S8<X(ch@bg4T5B*Dk8Z0*{WgOyyMDkA9e2OZD
zgH!9^^QXye4uA*@d59c*5t=L?C(}R9ymvbVAq6rCe>y>(mP-~sT$%&7x!d^hoj8mP
z2!**O3nq!tjAQ)v?%msWM{h><zG?NY%LItx)pA1(6*Se0w@jdcYcm$47o7~$;oY_O
zA~rr>n`1w0MM@|4i_G!GQw;G)ah%`kM<fa?g@9Z4JX~_iIU4SI=Ga8&kwQW&DcOS~
zqKp*7;-zMad5vfD3ns4r)4>xjpAhOhZ>n}%t+6l1S(~W{^dH!hj3GFUl7i?W(3P$(
z(<$mAh{n+&gcDMcYz&mY0dNK9;RX*6E_NNIfPugRsHl4pU-2`zX#J+@$f3WF{`<M{
zRG)6kmFhJ<qjEW<-*c-7J#$}+xL~KK%xiD1J%5)~;1PArz%e{J+TS5;b(lqHa7^X9
zyIMMpdnkD?XuR*@-^Y3$e7W>d>E7x`ZuJhT9(XXLSB1xf&K39mM%eq~*YTAT3Wt`i
zd{nq7%BmpD`Ps8)nAkpf`En-?|66Y@;ES9lu9R#S&|(fZixT;MV(LyG<!;>?+wD#9
zW<ATjx(>~htH17;cpJ>QjPI4dskU&_<IK#4iaBiD7&B%J<D}ND9BlrY7@-J&U{W-U
z(>g&s+(C~PWoWN$*?FAL#JXSUIN1BWo};4dP3fj;(25?G%G*2Bp`n8Gu*A%z3;DJF
z*=L_zb5>{Nk=Ms7a%seF@E=9^0CT^yaMnQ&9zTDs9A(uU^y1pJ5vx|OK7H+4TS*=i
z$A@yywwwFH*{EZbRi`WKro1WXlvdh4<jk3)R^gowU(!D?_5IX-+w}GyD4uuPSXXuH
z#v?feD^8y~rwq2F=j7D?wev_yJu7Q#ED~~m{BEVv;C+c25DuMFfz$L5vzrSq-6;JL
z6dvBZQ>RXR4T#RWUN@I{BgiCMB`#0c)Tm|jz4w<)LuTBs9H?@0ZdRSWdTd>oS?HlJ
zRo(rgJ*W0NGSNh1dXFB-@i0i%ndOlb=~-_rymYU$4yF5TzV71}FI1A=?v=*uEp=8K
z9`34Ds}u^pdeu1Ve@~ri8Bo|{|C2tdD!H(VH0;)Y8Gbc7nj_x5V~fdZ2Bi0&8z#^n
zVK|l+U2JAy@eV9?mg{?-Z}%zY=IV567a_K+9rwq!F6P&VV`37Qk(rr?R8+hTX{?B;
zc`IUb*TlQwfN>ou+bFD*Pe1$AI-y;6<?QgCZRS3@q@|)~leqNE0cE=x=B5jWn4d`e
z+`oVS7kDDUjowL6pQL8cz<biwsHi;p_m3qd+tUjV9*pC~69Z(aqgs=OdZ$&168LQa
zSfMZ76%jFjQz`WA6(gI0K*oH&oaRbUf_Kzo;S;5A$JLB7^H1=2e<>B4{`&jED$7-U
z{1)eQX{NkrL}@LL^*00dwDh{FtfZ4B7nuXWx9CCHt5Ey|mmD41A$8{YHa9e;zP3Me
zVpd#WR8qFZ#>%jH3+NO~-a;I29})ZEXgB@!ALh3FH0{-CDZDmfMxIagsj^Mae%Z3k
z5IA+zwpfI1OUZ`~a+7gMfUyoGL%_37UxT{Th)eI}+S=rOk2bVSDc26Vcg0lMyW8TH
zs!uZ;g}l=Bz2IkB+30|o_PMiX?|VKvdtgq-HV2;kHMW<G|6^bw<o)<96p38!4!$iZ
z{i9!RId`s?RDLjoRU0U=YUu0l8rmWGXqXGwN<1(8<{9+spBrl&-~HvOpR#3D!_zvx
zV5Nh9pWM%07qqrFKdn0VUj6#hJ6fA{bth~-IE@;YhyHoZSj`R{E@4C@^OsR>@3xNH
zo#B}dW}RE{tE+F|^&ApC#|ZriQ2(v`GTrmfx~Vle9Tla{ItM-tfDASCcGHlpT~#Q@
zmFWP)DjSQP*!4TsJ|!@lzRWMX=l=8@a*<ueIt`+;Pkc0f_h|m0dr8?g11xX0cH60S
zB4bs2j@#QGhFg$~#y-?c8)5j~t=Gxc2Xb?w9FJ6Qph+}2|Lf_I#2ypt0IL>to~_le
zo|3}LBJ9h`qxX7D8`ZI7Wb%4uJi~?!)Aw{?P+9IY^wy*?>&MR7tg9fo{ZVMex~*f*
z$AIo;!HP+zjc@Q}@Hu3@mXP|<`fb&7pLQoaMx_m%d)8y(+$P($cv@`vtruF_=*Hu{
zjf0Z^np;bUME|Tw5*ntE#wG>E6n2<fPoW$8`DxasK6{-fFdSj1*1^`+R&t5mhB5;i
zk56a4C*0k0&2#gY=eKtjv~)31D`;aD^1XhSpYP7h7}Rb?Yy+3SJN2|ZeRSG}30;^_
z?k*~Fz&X^mWTD}^6|qv0(SQ=He=Ss=q)M!nV2BZ|PoPdxONEh<5kr1bJVF6|E32c`
z|6+iz3V1dN98Yntrm~mbd$%ziTbzGg)yU)8Z?o6mmA-|Ro^jh~y7Ow%x$<iRcNUau
z7yZ!HON@$*oxAZ(R{!?2K25gO%X4kVNh@p8zuOuAw{KtGy62s1Keb0E=ze1O!cavk
z=|kCC03a_A;~!&Yi$+GRZQDRXqhb(?pEeZqoj29&qlBC{>XtT}v8M@&jcreS__LOq
z%~hIA{eFHJnY2^;)CLv=$oc_J_}5<>INmSVc3caU{9RIQaZ>99mG1Sj!}Sk+9y4}q
zfznf1L+m=L**3jy^Xz;0YH@~1LL{VZ#yUOf4AmPbaRbFb-Pt+fxa#4EjV_Lk-dV*4
zjD=|$`?;^9yVxT3BowB-*Prju%4A&69?2d49bOta+wrVv`>dg<GPENYZMa*^^kVsp
z-^vG^E64JsPCX-LD#qW>y<*$GBK~bnMEaFLyzNUW+%F!Qc6LM4(6^-%^DmUXY*0^0
zSUd<@7DZNC&z-xIS1N|bEp|*!%!y*MPpeBP*`HL0Xw_ip(!za*7X_p7+~7F7&(q?{
z@{NRE)AH}V6jaMzDf#<<JTa35J|^VKm7zW3b=V3fnKQ?E)<^Z;JBMT}x@iw3@U&@&
z;p^}>qM2qw(^9ea!@o;quMgLE5sGJ-y;B|4C$Yc>TIKE@l#tMgSSV8hv812@R`Q%R
zZck&1xmy#`H@6G<Vxbk7=o9E8vuCD9!66}|^BMP-PMxGGpgN~Awo?$H&#VuQuR9R{
z(mG2vj*qu4IT#%jRG&o(5Xa>^5Fc2?PDLp$dayt4uKstYk7>@mn7yuBRd(EJ{L5`X
z&HQDKj>T6t<-jgeQXt)PfBp-8ZGEq9+Vf}DRmVOkWriZk^!2vQ8Gd@ZY^*r;6ZP&M
z0MlgW=rqqj|HFq_X{TxzX?e1iBrs*tq-39|J$4@&)p}~&Lv8=FMdeclxQ9A7dOGBM
zRr!y~O`+|VrOks10l2Y%x|?J7Gy84r_PXlyI-*$&E;war^J-i3i~@DoZPT>#-bMqC
zG|kD$`3XyQ4+a-sE`0HQcD5){SO3so8<V;K7aokv_)@&EU4q?SShW(uW&dz`%p8ND
zCtW8_?85BCJFP&ktLna4xzn_jpkRY+y`NNt1V&h5Sfc23tC@*WSDWhz!r5#5%PuUl
z$!<e=X<Rd8h#z+?Beyn|CVb}gTOHc+$1<D0!7{!l#Z&NeP7`IgxM2iGL>R7LA78zZ
zkqx7rL0nIQN+;b}r?%kNGV9(D)mE)~202E=DraV9eqB3mMUo$EuPvDlcifYH+hXK5
zGqLI$jGkLi0+dgs@A&Xxi3m^DGC7)J^ut(RKWOUGrP>@Q@1sXGDPVq#wx`)|+ox||
ziiEGw;qn-A8+KLwxH=0H5OH?d&r4&?&Zv$pxE|Cq_mZAa&={MU^>B{`=GeoP3<Ql&
zX{;64VEWoEKeQ+QntG~H{g7D!SBx4d(BD43n9_FOmag+*Ux|Np89dj#YuV1lm5C|&
z$vgoi#Xl1_ZPe21c!275T<y>N1jEIc;wH-%LIK*IFDmMk&6qi-?%u@&69dUk6j>qH
zKX2)piw4unDtzX>74l%ZQ(H=ilFOmUE!|W%7K}g{oime6n-%wau4VX!;$b&`5Kvq$
zXt?m|X~&p<&XOPXF?#(Lv*a~>TlF*@)$uO4kjd=X-bg*AjxagjrhyVi;TAB_U=RYN
zv`s}`e)Q<k_nV)b88e=b#oIpe`+ZdQmla3G)5TH8fE6)NCgn0~5|SQ(Cf?y4((h<%
z1#avyE_v+ZuqEjuY=;ax#PXY66RY!Diq5hGHk4hdt=k=S7&P@|_8}w-$9ca+0?Mwz
zvK#eVc5T-zFf`QRad#{2EpJOxrVmWm3TtqBVjp%V$d`?&Dhi`(1q}JhuMQwjblw>S
z)tb94!&C<T1<pfpzu1KYv&V}4uvrc~<<pfVH|i1)64Lm~moIAz{4M2RDsEs&Dfnu~
zv1^*~os#E6-fHg}Yh$dVCTR3NjtHRFxa8^k*M?*`_YB;hj7vmBL<=M*lPJ$A7&1Q5
zOF|hnA*xcQ$kOk%7;$|)_O@DiO?=I++S_ar;^%aCpZ&4o*W)KoHt*Tff+Cc>A9&%y
zjLyTt#eoDNYrry6liAHS*YDpgu3QMZUS9W9{XA!6&ZULhs;j_Mwu47}tGcl2Ycpl<
zr4O<eUdm0X`<)~`<v42LwI~4AC(N_4*(C+QtyrDGmFPwmKn+PSJOH4b!ZI8m)X()k
zehLDzF8Onj0t!=)P)G0%Fp+78ZYY`gu}VPUgUjD0AL7A1+O2o`Ms!k*lStg@kr=Pt
zWIi=$$%A$T*XrCx6K~Z0-c;9;vE7Skmo8GIm#yvUt0gK^X4@?gA8<CNeG61=wcFvT
zd0bRfRB(9s9Z)A^8{_af0{jWSa6uLei*YLrOGD|Yz`>QzKYQuu;XiE~{VF&YJv+BF
z#%6#t7J07d1s4FfmS@IM9cb9ND#`0cHb1b%bUdrQ{r5F+VXd52;+hxVhO1h?FZOP4
z1HZ1LI3HX>z$%dk|6j~e$4P+*(Rxa!4Ng5gnBEp->ux?d^!5?kHSOVtzARo|`nImt
z{a2A4ArE;dCB;qqVG9ue{j7}K)PL6RFD{!HuS|$Yo20Kl>?o<^U7+89o8T0!nr!=u
zVN;$)($9tsCdpVTBH_DLHNTz~)4<xg13{MBI<L~>$<wD>7!KA0-QscVv^x|t?K?Mh
z_Xa*<a9G%64jO0rNls3u)nm?@?u2=Z7P}t5^j>=Ul$9%oVp-t}>AWN4?7Ix(^47$!
z_naA)32l~szY!kM)J|Bx9T|96L<Bf7R9idI3XAd3fSBBRwaVSKrnPQ7^{Z9C!}=S(
zTtu273fO=N@sa}OUOw}786Q>X{O_QYcFb_!B@0r#*(};sRlb%ply<)N_oK2BO5$;=
z7CUHESwmrLWz}Sx-e$~$8e=37I_|F<3@DN5KG5&iq*bdY!7CuB<ufXpWNE1(MHH?b
z_v**MXKu$iAqP-NmMg|lJ$?NAS)VpgxG5;rzwgSgvOoNQazOLn>Wvd+ESmBBxm-(_
z4aF;8A)bPWGap!7yas73*!BCB0(lWy^HBg?RBhr+xi-O0F<zrH8~Jq9n!q>Hf=srh
z>kTeAr2<)%I^hNxbHRvXO5b{Hh7#h35k@=Ez|Za{e01(#^i$ScBj7clgUtYq0Z>4O
zDjSRi3*Y!Hs7rF`4{3f+8_%b6BuL`fvo-sRA+J(_Iq!VV&(1l`*a>LD+P}Qwd39sN
z@<}0@U+7`Ym}0FP^o)7y>9c1WlJ5}(6J<{ut5B|@BojhE^W%~;ni^g{I{Woa^f?7Z
zMMcK*=Is@Jf7Xp%`_1~O1K-PhoO@EAV9keM9w@ulaN*noG}e{AiIR(|TBSifP>}N6
zJ^VS0=<hvvur;gr4zU7d4X^S2q)dKVz0C@W18IF1QK(+GI^v_Vb8zzX!DR&lH;pbT
z=2IeXYQwWtQV>S{{r!#m_=CsZAo7ZEFYVd9l@EvG8{xJSmsB0~e(XkL5HNvuZ^4z9
z;w9BI%VuSBg_PZIf4}+h`1R{%=btr>%<=<MiFKKB?4hzkuH0_$=^d8Qty@(l=eY(X
zCU&l@tmKkVe3q_!1z|^^T4u$v4}id|U^<fr591DUav2*^J?eJ#VkIpYKyqQl{R%!L
z4-_LPxOH2B<(L4RrhbHZ3>P=pjEU5<r%(TUX-W$n*i257R9l(Y3`(BvmS48k6_js2
z5nrL8BH>;#!;mW{4?>*PC_a@G)Ul)fY07aQM3$wkix$O>o|=I;xZzr~xw?B2k;&3_
zwYtpqse^A7SKf4v7Go+T^PAzc5i2(S&VKvu(b&>MQ|4%EO<OF*-H*~EZQ_7&zpS*P
zyd(7l@?-esKpFerI=Z-|h*XyCKf^|hNR3>dU^i^!NU5-i1zWo6hFfqnZ!90dqQDOP
z8^9xbN5{z!*O|E$Y+T)Nk}jHO;H#|&Q|PttYk%RL{o-}I2^=7~jl1Nn<IAz(*-fUE
zp?Tj3lP4dr^Cvsa6bj-F!0-6q8=1lSZl=V?;6qLPd2Z$nk|i))WQz$USW#pO&U+Eg
zqskVq!-w0h8<YX&0%kO9`0zid+s3MAr+Tej=Hv5jmYRWdEvr|r=8fM1)qArv8pCCn
zJ8JUEx$H1Z2eu68@2f=y*$!r!$K1zY_Uwbzqdz&=F3TNu)|8#k%_(5%TrbT*pGf%S
z_|e5OGJhTBcb#k_CzA=Fd@02@&7Vhod<+})THE^rpZiY>NIn=zXUzX4w7}R=Qpop&
z&iNJBpnIq0URv1j_xJ96>C^D>TLT(1EzzAkxjtz9_MH#hm;N6YfHGg{<LD<IurYIe
z`$l+Wf7wp8Lm5a3+Q5HW!%3<nQf~EEj5ZoVT;AK}`Y?U{OG~f3e7|v3adC0psxiNh
zkLcTXddV7rOF72n>j@E5{kM=X7~9(J=H@8I=VZm&43IX6m+u^HooH%VbzSc?EdQ-`
zkvsSY!ysDo!^a0HdsJDUV*Q4o$S9o30^4f^s@-TX@XjF}Uy(<aQFV6Upil+Sj#Cfv
zn{Tk)aWJY!@o2Gka%!kRqLX(yZNT(D*j~^vcs?++%g#@%X{`7P=C=qCxV-CFqEXC;
zqmF;R*tD~)i9j4$!LcCi;Vq?NqE-g$MV3dv2A-^?vBSsy#jl!;pegq(KaChU^2cbW
zO*sP;;}`9hxDQ(TU%H@G7fdg$U<-c%hUxgO&}g>b`{d3*AL!qREC{P;MeKH3M_KJo
zO`Z1O)Wj}~ZjSuU?x_`+al`7=!Xj}V!-%K*k4a}ShzHo@YEebz(iYzV+82SDV&9an
zd&bqr8(sg>U}#X+VXoNQ+k53Wr-=01WkY_4^4}AGCb#?w+aMo|ng<3$MIX`#-^~Y|
z=pt%u3QX9)M!BY{t(vzqQG4?G^)9Bi$D^_vehO3%$==2rsH|$<ro|372x??a5jiI_
z>B#+qfl>?XE-7(d_h!FsFOnC3{fgynvIfz@;(2@9zWfWEJ(|O^FfcbakK8(_RjadY
zIm1gtXeT6XNrPLGyJBI)y~QGj%is!WUr)7xSR)b<6f#pMtM?kUf-IHMRi$3yuZJ!R
zA!gDZiyQkz3`*ABKqKtOx%e&g9df>^N&~@%m<cIEqr7QuX}gW{oN7B(wN=j!N7A-c
zRe6jKag5I9+@d|CinxW2!o<R2jiFxDl`Cxs<uCAG6R`IC_cam<zpNn|5-b(tch1F?
z^2*D8^X@eLmJv%Q&r!!PlbW9{Bq1ULn1w#-E-n4nvjaUFgXs>|QP<L%He$wC(-|A^
zqnF(J<khPtBo%%V8<yN3x*|2`N^?V1d!SKk$@hF>{}(qf4KtmAbXICZ@!Bf#a-3IF
zWGM}v;Ie|dIyuSGNl(x5iI4z;;IJf5*VP~9%9BF^Nw0es5NXTy?aHIeJjNv04b)zB
z-sl)Rwm;jIt{>M~kZr&^z16FSOQjo~-QPV*bt)}6xsN<gbSG2u{itpEia>_W(H&5Y
zP-Fw~moHUs(vYOWdcEal=H^hcn{7%aCz&qVa%8(LKT;|QMkU<P?)*Fu0tn8KL}GaD
zgTagh{iZ`WUmaVzp>Qn4UgKTI3GTx^zc|jjMH$%)tverFG;2($8$EtzS;{_`CvV=!
zt~C<eJ5;&<U@I=RjVN}sM|o%k)?#QF8@83#`?F>vc+{JF`WL#YtcY$(SYEm+{xJg{
z^gQz^&!kI0Jp8IS_4QvATkLq}Gtzlc`%XC3lJLUCfV%y$s7Q}?p2I0Exv6PAqISS}
z$9SOJ)1jeFnCQRvkJH!B7)WAxfiZ<TB?r(Wztc57M~3aBGW_WLpvd0$j-0wX3<}xT
zCS5to4G>4qK_p-MV^f`mZn~EvD8m7liHYwIO#cekGB7Nxsa>Q7$%_x`4Ms#3g^GPd
zd&mZ@K)W-wE(|PpOW#bMD__5*>%FOF3jVwl&hcQdEfERurUG>YcO>-RfR*PfznqTz
z<HwKt;t%^FQ%hbxVi8pt)sZfz47||6!QpX6##Ztk9nQMWoB?iF=L;5Lubn0{HWi{6
zgS?7&8d_SKgXAlV!;Mru6yae1dZ(ohXa4PqVY9Ak19^8ONri{zh%8%o+buybNT)eT
zn<TP8!YD$uZ<?<C5x?R*O5!mjJrN5SHDO8l6|8Cz`#TeywR>s#MiXZ^uQHP&%JN=(
z$1v@<E#DVAY7PKk{yWeVdk2R<>KcKYS;iptOs}2YlN#F+YoW8SaJ%ums;5T;SRo`x
z(1w<Kn;}{YTv_K2(%EEq+PEvA763knoJ#8^QN)<08rs6ji_7@pYVyU0^Wnd&nSB1)
zX&!K!UNgMV4b9qE-UeA1IW7-}k^SeB$Sp7Ss<momvgL%+ji^YVU#LqhjWOz`zF-bT
zDMV3&M!ES<lpv>adv*;p)AH4V2>L#3)kJ#f2?sBMI?DV2o2({&ck012U<2wq*xmZv
zc}7EV@$o_5PxNioK#9XkPavZarotxX<_+1-4FtG=0!e6L;<T~u@zuYe;PKUD?IRn`
zx=jQE+a6^j_l~0KR-cY7@|m$vfKGOF>`N&zf5b62)F|C4CKX?yNTx2$NJ~v+&&Eyt
z?K8+cbxF0#HhpoQ0<tl&wT0%#GV%IESVl84IiW5y4ta64N8pRH-!)_>T_dCBa>&?f
z@~4M{(=0(6jo$x81MpxP97$`qH-#V4@Kn0S@ZJ~9`arx{5M@Pt-p)L90d`)pn~GjX
z@)!-od?-V*<4*`g(b=ANH!wb<QxqR2{1&@9Bqj1mZ*2~D)wG~uU^19h=G~=Q!5=DZ
z6R5`PBxjx`x~zHug@?~^7VJ8oJqGH~x>Num+7vk(5{uvsQjyhHY+y2GK!}4~?nied
zM#*FXQrwJ^)+RG>6Ov3heE4v=8tH|8V@|Nj5!wiK8U>rCrsh{3asY7fM#dK^TOJqx
z*kKh8Kidm}8TL+$A!YpxHs!(xp!iP4y0VO*@D8MR6sAVpRF1XHfUNeG)lIf|F)4j;
zVyuQh1<1Twgj;lYHiTk#;q>F%{i&J2m<GYy1nUv1hY<H4ue@7>RBzvTTlj9g<<{!5
z%h#Y=^O)~aB1tqNvQd|%FrF07C$+m4FfAECV5u=DR1NlPJMFZ;r*Co_K9)du0_5W}
zw2QTV0k3-=pLyQ<ZRw7*f}7l$g1;k%{j;(M=fXP~!`EhvhA3s`qoez0l$#bWN~#Ds
zsid?QSC_S6%3M%-CHo7bJg#|>e}?)0^+vZCd;2hY4}cZ1rXep?ts%AOg$9#u7>&%?
zI;Hcv2IM_4x2Rs^!lj~>sQ-9n(~p*k<^KBHr}67O;TUz6*dru=;6s6kqq6LpD&{0^
z7%c=jE+r#kA@|GF|6siSF#G=j*ztBW|DP)pvx@zH-(#neoa{dSHQYVhe;tQuE@0{A
z{{5SQJ;cobK0W`Dv0($~uwJB`<8F&DrWbnO%A*F~7b!&oIhF9rUu4r5%2p}iWI1(Z
zS76p=e67!LD%{Yw(q7){?f^5g0p&rx#|IlE;+*v22Vv!jiu&l>qr73-S|K@7^q#(a
zxwW8Y^4RqowR#d;#D~tT1X<^$H;kmTPb)ZZFq&eQzj#<+kW0TE3p4A(=YP*3G-UXU
z0-A%$9b)JYD`*EQ4!v-T`XAd*?9r}SYC76SDt_SsRQw9abdcYexB+)xBDS%)`8E>V
ze7b;YqshRC_1aulYeU50FNJE8Z5Pq|#CGGJ{i*1NjN_1w@goW^#NkeAIn@?B%9vy3
z(`fYMr7k{VyB)J3sOf9-o7=c1H*CD;@Vlu+ZB~QthzFRU0+;4UB%eMYwgIHVoh|ko
z2Dpye#g!Bz9tJbX6wB|ka-2uT#l?Zs=B^q;SAE)8=TS@_jgM0md~x7I)?-nqGSCIw
z;~!Yc_2brh;F6&ex~%gLr-09K-jP+jjRS2Mw3#dF=630LJ7on>xP&9=$NTnOWQxy|
zVaHx;XYaeSu>I4*SzQk0F0aUraqE8c*fAzh|6^*{&8+P;l;&@e>UNux_oq30tp{Be
zKmF<Rhwbd16RK_7_<)1frd(L7JqI1Qy}wBR%FB+GJ3iWLtee3ur{31ud!8@4>7H28
zNwD!2JL<EfMOsjxP*()}@9+U;YNidAA?$?FAxBbEwwm#!VdWehYRA0qmNet*DweQc
zq66l`jSkKy+DbW`k&zJqTZAY~T@COI<n{m9d-HIt`?p>6ZlJj&Q>MyLB1DuFN*Qic
z$UGLwoCwKGvzbUr$&@K%o@o+NiOf?)k~tzp*yl&jv)=W-YaMGJ$6EW?`=9ObN5AKX
z`}_S2*Kl6vbzbEDjgG>vFMa=^T!TLX;U=A&ny2C_<HK~w{1P*m1LS+X+fS?aXcyDB
zf@DFr@~Ksx{2b9*hHvpnP31@6o;^zgpILwB)3L)ZqP#;`4LvMvZ7YG|?pRnT8}ry8
zdH<XS?_v25*=w8I@-J(I4AyQ)IJa3*=2zjCNyRBq7Ta38-7HVQDgtotMWGPf@&%|M
z5260kl<CPxs?OdIJb&Uytnr?GZn0zKQoQ1If1%=)D5C=AK^#X4SU6VDG?qjg-6rAE
z@4lUOdjgH(_@4Lwc!tovcO1r;4IJ7-a04*U1oHvlgD~hX48nzTfAQxk->ot{z${WU
zWE0g9fP&;%TVsG0>akTYZkN#;j*21#508{f79xp&V$-3E7xyhcjOxj?dW+#Exyzu!
zL9UPhYXl2}8_p8M9f&*qa=G1V2g&<KW=hpljv`|8ez4JO_HGPrHu42n$90R`QYaei
z{xY^@C%Q)(t)`N7A6(#_Izl_y+3h^@lpzHJ_fF2AzU7D8tC9NPNX<aJQqt;y<Ft{r
zyv9D^O0=#jEF<_$4gRKNgLyDP<4|qI=|KWf32Y(|y6vW*6A<to$T`L$P%>NA+&|8o
zW4iQ~AopcNH=xpbz8Zqrf(InT9u!dg(UrZC_|r>_V=Ea~F2@ZcN{@wI=k^~)i!1(+
ztJJh>8A4(6Kf+Sezug@w@^ENjQ4uENJ61JW+p9=jbe}^BP(gT5gu4i1kTU^9Fz2}d
ztAf|TH?&UCp}kE8Hs;fmD*;UjD1*=!Aih+5X1C&BBw*29ISp!Hp$?z>K<ZU&oX_~q
z5}dKUU5_HQE*pI%LeJpqU>WKEM-&W28?-%~hr>kU(P`+y4&U~nkdZzV*W;5Hot&Oo
z<!zUUt5+zt(qg@$!WyzmR862vNNO5Y!m#EMX9Bch#4g>+=nplzmJ~5Sg@}h1<CyEJ
ze2r=PR~6M;R-^i$KCOj315Sd3mx%5_&|d?`i1-c}p?@V8c~1~$0J{J&eh*Y0MMfg5
zMAY=bx9qkPW&XB*o35UoAHYFeWqr=Awqg(Na3eu}0ifgx{4VLgT=04;-V#vkxr*eG
za`yHjm`v#lxe{iibfkbcTG+du)UhZb*>fQuqPT9^=%w%V30<pl^5iDcX%1%yu;RB}
zX~q{+df<x*o`Ev0wszznSccoiw;q7WIX#%;{DiAYUBXa;9XuxD_dkwBR_H6yDSzZ>
zQ^+`4EU7(c0vzKRgs>rDdGe9UGXlQ@u7T3a<Yu6Eh$zT8gwrCS1a3}s@S&}{ADX>E
znE+4>ut3sX8x84u<d0>e2N6py6GVs;w=r;)k>5=s(=W|tgzykTPrMTFDqO67|6*R4
zHh{UOWc+xv*Zl9B>;2{LAd;gXQTeR9qG@D!_$7)!apg&>e5evg4v$k4(x@1l;fM0>
z&w%ME*Rts0W8f6joC({~Focu}eN6M#@OfVn8qItr|4c3<@{FD;abX?yHdgP^WPPK8
z^nqn97v0Kg!{bgDFV@x=23K)UM+xvcb7#J!acz!hOtM9P3=TC~vwUGAiDLs*Xf)!@
z)NglUcH3-_p&e3CSb;KRpMtfubwyX{QT?B!=)=q$k86uCpC~6rzmzgOvrBzAMZP#y
zk}*;$+l?+!Ju1u$6U~+EX6Yl8A{;L7P^!K6x}%)ucJ0TH)J7c>q~=}V#g9W|ap|`<
z%M8c`g{Mj2e8{sQT|jp^snb!s@$yVK_Eh_b?UmTgDzws2$WWCb-J}JySA2Q)`*>25
zRTd=7P6uXYvcTV<4oawAO;-5FmoJZyKuEBoE)%jvU3PL(VK>VaKEDaZ$4N3=w^G%8
z1t)$FEDMs$k4a2JYoH)9K8Pj$0efS@o#t>zW@?1VglpM%*RqL-%+Kd_%9XSSn-41u
z7k8a28tb(1_{P8JtoKLV13C`*bj&@aAys2?k#66<2CI#331a98$cw(&=td4j^^(jj
z-MDk$mykP$a*92gwNIn^gjD^q-U5FFf8v;^N}t&Zh`K~(cdgkO@jBdiVf(F}_1ky?
zj51}0sIB!vw>v7tT+B-sGbNU_n1cQwWWg9PD+!^AX7I*!kzgO2z(7j4eBX)^1R0=w
z3Iou|vGYb?vVzlv0!Rf!gDL=Wn>ZT05p5P3>t27cZHFB9t$Rs3OcMJ;7Zg|(r&l#R
zH3{92r|0p?I&?##`pz959?~DZY0+*-a2e^C6UQf&E+_wXm%gO?;`5W}he0Z0%B#An
z&txO?ZnGiD1OQd+0X!Otb}I_#k8@)b*1dcFa6g1{0SX~z;VKSXS&RJC>&~ad(?A-D
zpC!tqFu#OfNY4Sv#m04;ET85<j$ut49}zJ^Y5TN_U1CxwP;k&8)}t~^qH3_RuPyP&
zK=~Ubrhu6TN{PNLLmcmRyVYCe!MGD1sj~7KZ{c(Lx961##;}L`$5&16$_!r{!5QjN
zr_{9-VOVn0ppiDMa$REDxfMX$e*5X~+~_u=`4Yauv!bG9D6~f3Qo5mca!C!>tzvai
z-5EsQeHnZp?~Ab>%dC4~dp<`IcN9s=KiCUU8xofP<C_Hp(mo3CUpKuMLVM|!$%#r&
zmqe!nA-9&tAN!I%&>$UVm9S*ZiuRZ(3wqt@nYsB3sTK0@f!AS+zLa;~3UwS_ajoB_
zfmw~Aw>i>_hrj@ukf=ggT5#BT)|)K@8Adq!KgFruh54}s#Jt~0KnskD>Zl{O_FT!x
zUbK|{3TT0K(FF#a9J5hwBW4HfMo{bv7Jsj~%xs%?$=AD>oe2BT&_9a^ec2j!I<OKD
z5ko(K;4|kXcI}W(yxu72%jZ{CtcHtIMxP%_)!%7P>t4_38}8^-(gmA}fNkOk{m3l$
z7IpT}Rc6#P(H>IAFS=gbsBWSM@aNQSQ~;~zR>yc=%?_tKnbu@JDYr{1!P6h~8#tgZ
z73Q<gKye1;>YbC*gH7ZNy7Z2FaEAi2-@0;V(XizG<Hx_xxJ$bne6U=B!3#BZMAp15
z6|j{<7AeYiHMmwTb(FmCuC0)IQ{{V=!>o3S#;Sp(mtnJ40k1`$!_n)gRv#5P-=wL1
z8aNl?P7=+cMrL0x$IL6ai#4v=Czn=MX_SoqPEJjI4xyLOCq8ntX~=W_77QI^@<O#I
zly2%+z89(ZC~UlK2;oap6Ys_T_u8`JdLN5JSGCX6_u%9V*m>RTu=vOpRUJ|623iW?
zUjaKsc?G~Tal!DvWkwq?MPng?zFa*fuwfxZ;?M@f3W}A>g~|24k727<U?gkiMrF`|
zMo1oGom2-9|Ej%{0#!z#R!op-YSBEF5DgaxIX56~cCYo!J>1B(Mg3!``o6XF14^AA
zN@OMG{61ALB(`wHC~WuP&g2XlUeGSq(<$yVKx`RbKRL)m%H~KV$vxe*D||K^le$T4
z2&-~(g2@cdum0L9BjW2M%?a%VQSUZ@;|p_RJIow{X%!gc$jFgj=t3v)W~Wn&A+==w
zM8z>(R*lCMo~IJ)<bGbj?$#BGYAf|9UCvt)lKE(4D08y3BF(kQN*vr{+$ww{X8eyB
z40JXyEI}E0P;Z>UquauaMUkueXL0tR#2zz0nM5df^%g3Q5)3;+Yr0VQC7ccF{axWR
z0TIPKH-`PTPI#rHEgi7?V(nUHH2R@H2G`ptvp`#WgL<6AL@Ger%tqsfuwzKKkQ!A#
z6~rMC%g%BleZogCL#&gYV`-%>%n9k-r;wmM1GGZvBoBnkYG}8~d{1iztE-cJhH6l*
zjL74RXk}js03E(O6imq<Iohu>==rc>8lLY4-%n!wngBw8g+=j3ruP@UK-j)@C2Bc1
zDh|eHtn&0c*LnXjm>F&}TC;rVgrH4>H4<n%I&0wBr$2%M)La*AL_m5($RxoYl>FoN
zgK0)nJ_gqJ6AHKE8-w-EK>8DjLrn}LvEUVTW`Tdot1lq~<zw(m7LoTXm@G#XjIjQ}
z0h`pT2>r31_A<paLof8}rp>)YyDAvs#bR<QS!|{2%Y{W>S@g9U2LMzDb&@kYvBH7<
zTlP=Qpa#X(P!ZatK1*zeUgXI<;EYGf56<^^(h(GIDbXM1TU&m*WQfS$t84pNRXt+y
z&`Ca*<;CaSu#*HFho$@GI%4DWO4WqjqHOOSmSsK^aA`oKKrD}ZKP=zKe{YOWvo)}@
zSXGNdwOcL?4`dp#soWP~pVBiH`Kg@=c^%A}VZslH6*EFv6`uL#dQ%<994wFHL)Z4E
z%hF6chU|}Hl4va$BG=2xc83LD01g5zR!*JF^mIY9_OXVK<w3gR=PGveNbFem7<?rd
zQlzmbv*)_Ix&$w7ypS>5tC-$Gs~!ZOjziS_(?i-=z0o?UW%K1gHMAr2In?aGSFe$|
z?u-`zknoOlBna{A2v`J$t}kf>qJY8=>;iw4P((HO`&tr~r9L|qV!VuU@9i~EH-Lcu
zX39r2fiM_|-u_rVk4J#)ICr*!Af4I54b}pZM&v3a6rQDlp~Jzy)z7?t%k>va5Y1ne
ztrvI&I&PWK7IFEQx)mTS23K8pCqjGqFvzl)@oe-Xh$%jIZN15)YZ;x~cmPYHL|g<m
zEW|D)FhGsfX}_pEa<E%9jn>@_{ixjls-;js#X+eAsre%3E%qZ6G_Sn)ERn=wD8!tr
zXPZfO%w789W)Y>@E$Qs>mES;vB39TilLdxt7n5>@2<uu}ybo{>ykU`DX*J87rwlTA
zUDNP#h4xomkurJ#IM_e}NI@%qYW{T=Y7y8^r9|~0wZRonGFiAI7baOCodf#@D+iJ>
zA^JC2xmA91D4OZql33wU$yfFjZ6dHQ$$Mz&O}~@^f(cjR#S=D4ayxXZ@0a%%$u|1&
z+Z}%TI?m`;x*NJ>UvV4^%%jtBo&p*sR$M-e2ec5oi_39)akur%kNT^DwLjnpk%^d-
zWlZVs6dlW)^I77XHY0QV`0Ld^2*%UKogp9+*`K_;sp#NsHLlUOOkjr*P$GQ*`+ttJ
zzo*MHc*<S~H6J)JqgSXXty7+riVC{WkfoQXURgRP-GG08x~%@Y%_e^dNB@kUePPE~
zzq<L_#JN$2qKE1!?^2A|EIM2g@;)Y-!PnwFA|o{y+4C=RjfOJkq~A-LN_32&ZiXX@
zq}M91xX$%hd(5pAqXYW)QSL{>5l$XKS4j@WVk>X_8YTXJzWC(+HB`h{g(uwk#7j*5
z;+K-T%6Af%{r%$Ix!r|WvA<to44X>HZ~XnDDn)#w@Xa&$#r93%k|UX<P<*#e-i?82
zgOgs^7o`HUhZ-aP{a-Es0zUwkh}Uo-v7|^8aa=vuQ9!zZBRHxS-j7l+Bw)~+f(ui(
z(wDMGW8kh}T1yM(|Hbk|Oz`AB(O0T`j`9IW=s%w?tYZOw`zOjmMCr$<fPA8Qn|T0!
z(kcLP+Ul!ZRspz~I9zdj@`4HjmvUK_p!ooBW9<*HVF7tfh$_|!SHb|{iY}qxsv8Me
z&Qz0AZ<HBnyh%>&L0wz(RGUe7eo-|`Vx2<9(1%DFQF+TUsPz5LJ1@~`Ej5`JgbV`&
zEnLykX#R*)?EudT<Dg@APk4za4Q=o)HHuK;1>CuKeYk5gD@ORThff68gE>XFa;>C1
zC=iH!BF)bHeeTc>wOmPb-Aww`_y5?&vi9fZ-nM+F6n4hda*ZrBhb59ExwDOK4f>wi
zedE+_7y0`e1z$aD+&rG#Lc1wm;}Lv7eEnMGvhB7>=4maoC(S3dGv)m4A5Q6gvA2C#
z^`zV+;Nb1`XS~=)T?Iv7`803-%LGG_Y4uf;YTI>Or@q-X&l8sL2+QSXyC0g`8&`##
zIrxq`vKnYj-spkIpKjS4{Rz{Aci+is5%8kP^Tee6aR9g!#K^K8WA*t{I$Qb^2Jq5Y
z7F_%%1%cNaJcFYzK5icayn`*%K5&kw*}B2~TFzVs(zz+F^!TGq)*elpvur95abKFP
zA8_<&ww}ui+@zbiPV80rpg0A`W}^6J#L>LVG-?c$SKRmTrW1ZV#&J-*LEh;_vpe<t
z(eRa|?TFA<mQp~$H8OB+Gg}pOgZ?ylrWQ6f%PFM9!Wy{!*+#We)K5Mme_Bo<L>R=#
z=Y6^3j_N>1a2`rZ%P6F0nKY5$*OR@VxpM@-C*l>L`(;P3AF*fnt$+5JFU_y6G0(_P
zH~1cqSu>pgKPcYNfntOn5uS$O1d)a#ZDJIu3Uf4EKOjGuQP{H@l{vzh1KIB8dh^!t
ziPisJKCU>Pe}ciq>!%iml=|PxcY}|L)bQWS_kVM1(E=t?e^}Ll)Wyh)!^^;L?!}99
zp)yVyi703yp`<n4#WQV083EL;6%-)Nmt~**xR&y&|Ho%c6QH=LD4HP<CZ&$2rb~G-
zw~_XW>^7j-s2=f3vios!7O39?pA6W_5(*F}^nWOk7YBh4PK4r(?gXSE%N34}Pojkp
z2cAMfsqyY7i)&l{lO^$s)ce=s09N~7*&M&WQR(7>#qlI6IZ_zh{wcMPsy>`6p2ZZ1
zzYe4g!j;TqzjD;x;Vb?ubO8l_a8FMC&(QsZj~i2D(0@&47tFm|0E3!Pb>Hn58r0wQ
zv*fc*<l_xJ)`hhP(D?@ePR{qaNJ3vC=H0PSSG;xpsM?jTa|))W+|s|lY$7KNJO&U0
z3xNh+JGc;QWBhncvM_g+Alp#kw>4RO{m=4EFLnA=d^TFJDcK(RdDl>EuI`f2g>m|!
zeE!|8!E-9#gB(Dw_*E(F>4RLXA|K7YM41C!z>=hC4&d-Br|ztPn{jNTq8HI0kJ|p`
z%!0f;q9oZ0i8eB57Sj&7t$26%ammK2R(#y%*W=wAo8#Z)xv$ZwW@c--iDtKX2Q3f4
zu;D40A>UuK@M)wRbbs9S%U%(kBW>k6%&NxnCi;7u?dzFM)ZQJjtePlfG|ti?4;`!%
zK6R?&F=cwe<))cFBTHYEXQ;ej(ubW!V7&KC^wm~EL_%rO<)DcKJJX41iEZ8JvRnM`
zfGD<R{0K5rVKChFx^uwM584(FA-=OdtE|irFd+E_Op7jr5ajP!kPD-dV-`@<ehEI2
zKMGW=L2?N>H~KvHg)hDZy=FIBxalK>jfLi`y#8EADCb&Mg)M;s6)zF;=)rZ}@!;Sk
z^-hSz7P|j-7{|TlY;eDx?U5z)usq;pzRly_IN_J6(bpS-`s@WfP4ed3oR6))$u~6L
z{y46BPR7WjNsF<v-fP%=OKE&7`eQ%~djh=+6bWd-BSixhvM9X7R>U6!32ts~E_{37
zwuqWz7T*Q*-a)sP1G>hf@`axOTo05U;<bbwxs-xVwMx%A2Y!afzCGJGXS+@{^A279
zrKEAG0z@N<x3JgmTErA;Nv93zWKc4F36MO9G@9Y2fj#g-^y7a$T>Sl?EyzB(Wp&rJ
zI&-wqt2HmP%vsvGU@E9(t9fd2%WRQZkUApe2yLS!-pO-sP?I87j^a8KcC)I~bn_;;
z$YjyX<m8ZwinKKA*00L_=MqLQa*Jow^Q-lNa0tyg3T5Jt<>cnqKP!v9XGG~cva^Ex
z%R{Oi&K}r1F5lz1m)?{$9iMBH6Xcn4e%RdiVt97P2ofC%D#_22KM<Uh;!P$Onk}AT
zWsN9>pVSx9$@JF$@kwPMu>$8O-2;V%E_e*?TGf2vW}3#x(AL|5P6m@#kKW$Xk#%z)
z{d7(!%Lj{y8`)>?IZY~GHI-SR%a(|kkwQTM#3SfSB-m{f^lYjc{51qeBQAjj;{D+D
ziA*%lbYXjl;KQc3Z{MPtL+G>w#a?WsymH!qM?|GHn?%dMXASIbe#$2}jp2(iXAMI-
zul?s!#FHJpV(}E^_hjJ|aTJl&w<W88OOu~{HtG{cB07A%h4*G#B+&YQ98=(97`$+V
zBjZw)=ypG$jb@|Mv!sYbc?wCa_OV<Pj6z}x^7G?htU(>+ME8X~nETTm(;L^Rl;-<f
zFeGM4U}N8ia|y6gEy`0gL(ks7?R40I?Z0eJvID2Zf!kH-t$tNA<{eI7{4VOoDBolJ
zG#~R={c~&G@-aS?S16NS`f6Wjnb%)mQMzQk^Wx_CUyQo`;`Bi@pa3@}Rcy5Ku{%jZ
zhG)zC_v?U6(7AMV>q?YhegU}sj`8TR;HRT$hz{s~(BUdsDU-5wYn=_VrEg>&4WA|!
znu10M%!&<?TsP!ff10m@d1&cit)flsD@malS+3WYwvOm<7W8o69*VJT2^H=(lM2}@
z@N`b5E%A9)%&KlRrNlwqJaw6E^+$tNy9se0{`W!6qx|D;05Og&E@p)p1VgCveSa=E
z6XvC7#aQjp<zj+CN2KXG3)*gwJ){Q({BqDulZyZD`G{^V<tcbzv(eX_o_*ULKH~4R
zWXg;u-^`SQ*<)LG-@*6JyMpz{H`Y$o<quUek7aWzE-vFRSw=nc@NAu36mJmpe(n#f
z>Rs|aC+k6N+tp|97VW_}{{=8UftZ*goFKE3J%8rJtSerK1;OI?r=3$im_S}Vmr=2J
z+IkGT;X_G@AJ80{p!OzTdLpFA`CsgEkM6q{O|V<mzj&-<Ogo&;+%VVOps@HvYya8U
zh^VMl6!`BUr|sJsK%b^ikTUz%E?#x!V9`j{N+Rr)nwm;T-fMpcXrdxQd6<)PLGefI
z%dv|d>9t3zBCPN@ko4J0Ax&UMtUlpU*8kT>(R}x!i#&=8j{_b>e7r}u`9;%Nn!<%(
z$A8U$Ub;hnH(E4;CEo&sTS7sL#La)6WJomFhC%x;!1cl*e-Ontifd0&EOpYerKMQK
z(I92}gFX)6onV_TqkyrvH{nv&|Kusu-enO3D1MN>C?;#<uFE+8&}SQR1r>@lB!2FB
zc0g&BuQ!~@G76?gChqA*agFG*Q79k*2LJOcwW7>GO>u^a&<QD}qSL$O3D3n74-W3s
z4)R#n01DE7Y?N*{>5|&bauh=xE4bf2z_I=!@?Q>4b051uIXEwNsLgKf$lyLw`r5Uy
zo%qi)A*-uAx4S5031k#5MLQf?NUI<Vzo7!1q!wVeLwyoyF8EfAdCR_|nb)+^kFgMX
zqSu|ekKDXTyMw92r;%?She{1R1@sJ<KMgbin~3!Gf+$!fxf2Ca(t=8wHI`5a!;au}
z&RIKU9}oJ*W+*qW7-y_Cl}*6Ft<9A4XjGX#z=0R|@X`6dJN(yYbXf~^(ii>)E}1%y
zz48rjDI9E1|Hrz?rEZyB49h(ZR2mo>ax`ZcvQ3qN3;zddRe-w=^unIcekA2)@|chh
z^3v&-YOSmPv~JoN)$jmJdfR-cuI=&s48Pl7kd0@-uAqbP3Osx7gM~?=af`fm&!l)C
ze>ub5>_cHwhuteD?ocMhJTps1B`H$e^g~lK(n)59AuJ#^q8s=2!iMWFLhky8t_tcn
zSF#Qaz;9D+Ikqc~w8iND4Q-s0V-kJ*D`&RY5Zm~RZ4{T<+To_zoR^TTmiMgJ!OHCP
z)d8cX75t|^s@Rh!ub7|8Bn^GU(F4s3*doJ;yZi#ZAJ-iP8cy{uhMGFJi}?IJoW+F7
zw$wn<pR}xl=R$fEfX$PKgW(}0exc%TTqrJ?sm3#({AiN3Zf#ZZ84o(+ETJ)EBXo!A
z(y8Xwd^tjj(dC1~)h#P7%zOP8WBIuDw?TW>!E==#&u5YyQQlQMW3YItT>Lc=m338l
z8vUn)p>@R~&isFvTKq2nf&Uzw|L<oq^IpSUms~Wa+7`y^hHrUXro84#tLDF_s2Ogk
zMYBAfuPHS+@v7cX=F~Goy~-Z>tP>mE8?~&}7?$fUKe|^f|H)<V;aS#|tEgvd-u#)b
z7VoF>9ID<M^=8(|(4d9-HF~V*1pQfzJ#rfrraXKOF7MkI8Dk<!W7U+Jc;^lap`EhI
zc$Px2k8Kd(1%>YOAI}Q#c~emVwuGqBV0&H~`}*ss{^1+^&*US^y#+*Rmiin@zD%KL
z$RM%%04JUbN&4}yS;AB-Mk**E!jHf8X@QIz)GIvynRm;@A_E5r_6U}Z6ay*i<YnYn
zaX75Hiu3Scu5ELulbk8&5O?|hNy%?5ewX4sc$H}##oJ*Tb4FP!yD(?L0v6e!q-$b!
z4uWTQNb8vCN60z!efe^QUjx5!^a&IwrQpEtN;;*2=->p+r9k%mqK&7QJk^YqTSj?k
z(D(H#zmU*zpcVj0iQ+WcwuIa5FtCDj3i46`Up1KqOKG*mF2x^YzJ%$>lSq*i<K&ik
ze}B5xh~mv-2}!P3zVzCI{<G*mW?gfFv<t{_>K8kvVC8rQ5N__wFBG7H%F5@=hQCr6
zKJi+2$wSH5r)wx7JmOZ3itw7ZTZWpzxShh4VJ`WI?p<+{mn$!o8BIZxzG9{+)9+w8
zB5zZ-U*A7R9z+XhG$zyuQ`@2639&BK_1a6w0zEEu(4r(C=li~tvKANVN!F(}P`UHG
z7UsEv(H}H5+;$RgiYZSsNO{aeqx^<*kQY$HDoyIe(VErvV``+MrBxac<0S*~zNtkM
z)&D-}ZlLdTXnxm18HSe%6{(P_UJ=eMQm(n6X>n8aFt!-tO2VwBNHW6`2-EZIk3QqG
zkV-pi6NB}CGmbreI3o}o%}7bzw7Qhy&4n(!JWR2tHlQU7%fLr%Kz1MomI<{10<4eF
zUp*dY>LDEyYS5CVJ?7W1bo}_<e0U<=udR4py3Mb{a+V%)I^x4jIUfqQXa(M5SOW5q
zLvZtZ&RrhDNhNp)lbgYLd-0U51LgE3UFU4v%DS_!;6!<^JlRg3g^fImNg$_`9_V<t
zdCDL>gBrQt&71Y$QV6%X#-6v!K|xL6LHv2Oma{n#A|~W$-EH{tn4;SC-3G|P?V2+U
z@H*;W+>ni`o0#<Q21Byx)#FacJ{pX(L)%z9A6lczn|I<jxil@wCY>XjBxE`9rADrN
z9*K2^%NwZHVMmKDBrWxMYBoKG!xS*0muD5hB}+cAc<}@8tA0WYjx}hxt_bURsMyP}
z1h6-%bl&u+6XAO*&6*4(7{l4KI2Z3*g=*q#@Y{z@C%*l6BMTML3{+qH821^p(h!VK
z+eKxCP$WQbp-aQiS+=zCLJMGI=NMw@SHpf#y@T*)#zEN?_&Znll;5qUG_9gBq8!=?
zT^H~HHP4=Xw2R4~%nXxLP_PDv7=$Z_B7?z}%$Pr(URhP((|5IcLK?zvvQ2Op9n0eR
zF|LCt6J<a|C>qvg|Lw{6;=T*of>~Y!K^#^pO+o3~s67f{x#)g|z+6Uauc@i&6i%Q~
zSopnrY3Q}G0<`tIt}c@2zO;v4#oUOUI3x$CVQ<fVfT)m8Y|C<TClVj;7ob4<=4pi1
z(<PLn0Z>3qL985>fHr6y#DCA^W{HHjyuCB7e(}JgdqV9ycI;3%c2}$(?U5NS?~o8-
z+E02wD9-Oz^>FQX{8T_7yBt>B<-z?l?SPDu&3r|XJ=o*TV0UPJwWfGa9L1sBnsO@N
z7GzndYVQ=cX;nRNaWuT^T!lpiI?(cLJG5lxf5l@YmJ{UOE32xEbB*QYuN6BE#G#qM
zcp{ppOXXa|Iaf95yPx9;YXr+{eX~yHOf}uxJ0kHaEfh*$AS@yU!ZgIl!61h7Sc27G
zx&3t*$RBxV-_hO0WOnwf)wlPt{G&O~F&=yb>gqV84UirSzjKEVf^vWh1|UV-jQE&u
z6qo+tOMWhM8t*~x&865L@<sRP&a*Xt|4BLRw^aC_e_q*%i}Ux3?h!f7i(iNC#KEln
z{rdlZ`2VTZNG75MFiD|Z=UPagpz8ILXL0*rif?tD(}n{&!m%D}{(1?{Iwg6%iVw;@
zgIiVS)^3qDAQy@pGn_(HT$J|PL9`nES1I65Ti|rgz~khhgW>*xi^sznb1hHe2dNnL
zRS!bIwxkcbJy~AAKCkEkh!vIqY35{1n`CsrsiP#b$G+B0^fejAHOU2Lg|TKv7y;RW
z^v3DWFMD|5WW=z<p}S6gi#>dqHU*qsUe-w1T<>Za?Y?t}AFZHA%rk&asRnXoclJxU
ze1C{p8O8KtTR1q<iEupYeuo5%T)Y`H(3o-yPF%CX(*sh9Cz$*b!|QSAwuZ|_<j@|w
zFR6(s_AWQ4E1=ssi6($=ccC=WQw}RlIm)Cb_OMgxXRB#g0zw?HSFcLkRjz&ew&}<Q
zsRT0Bi+3)%7o8vBJkW?%#DZOty9R6Ly_!L2zx8`aoHRZ;eFPsm<~IwML!r~<`yOXt
z+xq0NbxTGBj}+QTB3$q3pN8FSM1Y@(uGF0e4(P+xf=mP@OmsSWB!F0N`Vr!(iV$>I
z(2ujXgRyQ(IjImf3H^@h{vVo(f5@UrXl^ryBy2JC8zwOr@g#Wk?{PK1IG6NI&*D4c
z)~$HtLPI_)IClq)yBsWr*6)`oSK$MD$<~J=MB>wi3|bR-%RE#NWmVJl^7uzOAso;Q
z9pE&zc=;6Q_eXZ^1F04NR|A?1co}(GqoSe^jcFgOBT@)~5PS^F1??TdN6GLr;<p>F
zh?)xqVuf|!)5OXj?7#4cb_1t-n<?T0$_;h~7qxTf5lMlCQVxShc?{C)b&0rI{R+=A
z#AX)mWe$BxricA%ce>v<EOx9%aH4uzgjgH*tNQO)8_PZ1Y&#3p!oaK8Fo21RrNi@!
z8@HZpoIBe3Uj2j)cL$xkx+u0mTSX;cd<w{blA{6Mp)pQw4Hc@bX^ucSR^Qx!W$Y8~
zbZCBv-VU}7H0hM|{PWjVmxg>CYeA_ZiD0xi@mlKYVF+fe8k1vT2Q6RT*a&)%AO7jo
zR#{DEj+mC<yhlnUwG3(Ew3-!k;Cl)i7S8t$g9$fRtz{KUe(4|2h#vnGbfUd#0X)eo
z{Szd>z4`rb?`aRR;AvDvL+|F$wv9NO9^mCoJX#e(&d_Fydb-b$L`cIL525>nx7`bJ
z2CM)X<vz+>9F4|sAYLkFU7?|O`>la&Qx0i&9khDgCo!cZ#FYkd9C@b~AS{qd_)xbs
zLK-<G&L4stJFnlYI)2+}v~xc-vOK?sDC>~cjRfC)w2<lAd6FH^RG6z+r^}my*(7=h
zRCxvve6$dqk1@Z09`DXGwr2<l8nMju5XwGTSaul1|I9ml%IamVQ+}OX7x6R$U>#JB
zmH@fqR=Z8iL->9*kg%Spbw(s$h*<cqRs@3GtwwNbjM}2K2p|TIy1rv&-=WhPvD)?f
z?d6My)@P~RayH}~^|!Yo7RL<Bn?&pj<W8`TDn4eSw+uH+g$#Npx_AN@eA}9ZJloo4
zh54%Rml)G*+9~;U%C!k2WcvF1ncJY8nO2=z0f1ivxVe&hkn$x2G=U@Nc27o27_anR
zX=$Uq_ul@R0b9{52sfR=?B5vOgsHYLrK|^>AH!muc3*%PglZtGxRMy{-^XNHxA8G9
zP<w?JwtyHfgkdRF*s?oBWSQDPhO9%0z=K!1xpzDn$AN_BJeGh_?uuE4CHQilMc7Qt
z2@e&aW;3p+l2}KNL;2iKeT3#8ez^j?)4uP05e68F7-0zzEn%iYV~C<E&Bbeu3z7we
zuwB&3ce%Fhxhc|flPB}+K9;{URa(3&OV56ne*!w5DQTfJV;w?Yq9_SFGY54_x-zgg
z)iy%EF1Q1HpTuS>^M4a&QVBn)u0Dy5=<uRKH+1!`!>prUGccaJnuh%_WQeWf%{44e
zGD~P!&e#6_+My{X{mzl*^m~{v=Gs(gc$3(2#X~;(e4R*)hhl?6l-9Ih=O)2z+diiT
z3uhFxftBw+xmzFEK~GPQ5=Xqo`5^<Oh<y%4JXBo3W;84*i9@3{FJMG@>NHHvp|+zY
z?`AGQ1zHo!OR}JNLpf?Fu8iuO+NZ!%kW(+sn!7IZ=P-i8X0#<nJa~}t`0?X3*0l3+
zx`lNuO}{SL<p6E%bQy+8Z-gd}POANyDi=&K-F(vn+m;7GFx~WK^}v4cMn_IPBUiLf
zK;5Ak5gY_A`9Lrpn-<u&Pa9$Md$7Gt->}ZZ6vyNE@a@=<^s<-ufWHqzEGVL6#!Z{J
zspMD?NeJRUp=ArQVgux;GjI$Hb8oz?8@Fo9`j-p9XjC1yEA0pLJxWN@5;{gQGn8-7
zq*R4N#O9zAW)p$d8K!aG?IXJD!57vX6zMuwkTv2cS<A2m&+Erb-ESRIb4DGrXfbm#
zL)++*9P}dgou%1w&_xyHVWDvyaw;^u7+kmw(Eb$>8XAkGGk{+5zOyS7)e*F#8kq`T
zu#=Zp4k2Oj7m2HGZue!p${WXZ(wn&oLa7bh!3WhJm6sqdCvJu~d8+zl+I6|O@0G7!
z-F8)>N?w{s$b!Ec5e=k)!&LUa5<TUB$lq?)`W;k=6v-&F%aF^{w1vi~`t$j3pljBn
ziL!6X)d_7haAYS)o{ID#!M8h)M9^r>KRL{D)qHm}a>m9iz)VITAt4nyhQvL)aqhW+
zP-HRYQep3v!id@qxfL2xEuI<YHx$6(KYG)P`|_zF^};(ndYm7Di7E77D><_N`w$Ee
zdvSt#Q!1uE&Y-%6_SPhf`h4N?hiG<X<<KCEK4TLp*7m0eip+x;-?OKDv>_4C*U^sw
zNFJ5?qn__ZoQE*hQrp$FIKg-G8VqHEA5l-~nJvN#v31B)$PpjJm?YWvnAa?)&2Ms@
z^u<wwn?c;>G{}JCpZeEkx7i>AN@|QfvZB`wLU1J(70+uN?1PPKP|%0(6z&~DzngVi
zwi#SgD32w6g#5hH!3P~9T1c?-kP}hWVVf|VW_w~nf`OAwCC1?KHxvZ6(c!aWH)ZS}
zY69s~ZlW$sMQ<xUc&n)DC%K#W`B$w%qlf@U+BbKQi6>j>_{Liis$gG5wGP9DJR}Ol
za8`o4*)P9k?L|r@Xi)gwub0Mrl-9ol6KmRG!Q?y|UUL=d+~aXV4BD8so5TozuAF@5
zng_O|_Lm_<eum>LNOUP&%V@e+_BoeKfvdQao4eO<7Gr4!prasGE>tQZr|RI4%_?TG
zl{(>a5Rrn;P1C5|r+u{^>ujSc9Xr;9=MA9Y#}7Qj^?2EO@%_WYa!h`|_6JwR<LFu<
zWeZEd1PSzqYmxxh{vhwh0PV0*)DRsw2-ZTAYK>>Y6kRo*{!&DS;nC6VmA#8W%Uaxw
zR`fE&v5H$+09Mh<wFpHFV>!7NBt&%IH$eHK398dbpp?SUc1fOY*aa<sBKeVm`dV=;
zkWNoRiyf=67rF!yobSv&0Xa-g_wR?PG&g@Gy575VkP~*%5F={Cc+jD5iBO~%E&9p~
zwM|5k(#&Q9%kGDrRV1(W8N8jNM6<`*3pH*_sEM|hT+D>1wf`{N%H<eFnGV+6T#MIS
zBpz~GaFyvnX|vbi^p(euiyu0(n<nFHIgbV#B8B5gKT?D59k`eQbjWzv6!nIEGp@v%
zU^w&(vnsq5xxWEN-C=d)=tpC!+)ymCfV@pu0s$~cOd9F|p)G@i9d{Gbir0;R^U(xb
z>Rvf_JVrVl1H`E6K*<8+I<dB@<*BAfhC2%EwDF35952wdH8yQ5`k>ZO|McJR4K9ff
zBPEe@TVm20Z@Y|$h#Dxh#Wte4Nyj+|wiO12_EeME2Q<iHyD(cxK8|=OD`;;=1{MH$
zL2Hygrj5O2?EUFl#yhXwz!+@)=FVvwi~z_uT7{?hG$^O^POml5uY@>PG_7gG^<*3+
z3&V4oR@m|Do6Q8_#DtDKc$q}ZgA(B9SHeBj#(W1y62b%>7!7H%D(lgK&wVxclFTxG
z;5jp3GVO-h22<$TrG|xvg~g#8ns4vkpMJe3?(Kd1ZOd0jNhfH#Ha*rosXo5PB~u<-
zk5I|E&10#LHw<7V;$%KpR)pYu^>wg~6tk0th)DtUSt0AvD>!!_$xI%Q@%xy$i%wPa
z_N&7TH)fkD1J+Q6?K+7#E5oIG>*6h1Lt_+P0?oCZQc{M1D*wjovFlSh$@D`qN|=|r
zfKGS#%NukG1o&tmm!8r@Bgm>cO8RoCVYPh?r2C}D;&sc(6H>fXau%IXUA-p3ipL9Q
zA|^6f2&qkH7wSUtF}Xz14^Y^)pk#NH{N=&y3DtPQB~nlX6MIFNJeH~<jLKD%j#FX6
z5iL%lSaD4<S`;_P6;LS;<F4~aFoOR+Q8sV|Jkb&<7yLf?+NvUQnfdQODbE=IDgFIg
zd6MBjenRaNSv!1r@0J6m^7rfi(;qfl=R`?%M_QK>Vt;J=Em(dYD!ScYzj^aG<cEeC
z#wwGo1o=JrNQxu23K4l&0-9+clyOxe?sa5%I1j8>bab@pb@c%ZN<Wj+#zO^081Zpz
zgZ<Hn+Os)3w{1I&4F<He!AT<GwMiC+!cfD$jJtq&#MpC#-`K|HvCZClNVJ&;Rrw7a
z9dm)C@=_s`0yLb_IRgewqqQ~pE<e9{L;?mr=%&-yUB&J&tX$4NY@n&O+pgmLXJ05A
zCcj8=#EcJQeqjuunO$M9j=KsvI6^O3o0U6jY9fkRy?p7$sDklw@$3sNpJw+`f$L#z
zhHO#H3#&YS5s|FUsxtbs|8HPF+1Eg8(A3^arwaH2Z=M9UeAL^s0L}zpQ{a_Eqf}zo
zZ-g6a#0Bg;S1|1CmV5*UjZYjfl^^uaNHC5oN5Vy1LZ%X9B_+T1T+4XmDaj}@N3ipf
z;M2A(TMnc9vN0MUP1<9I4|IygZTK?=<j<@n5g}0eJoNcOa;d269P*7?0D$=Zd>*Z4
z$C>=)fYgwb3L$bTVu^3K5n7nziCvcU*o@<=ja`@ow#$zI6GQ#p|M~sH2^KH)M2xgw
zVBk6@7yJybybrTC+V{oLn<r>oP6Fmz(B$PoPg2xdn)Kp1&kCf0cu(ku{FA=I33AAI
z8lVNR9vtj++`2TRI<ck5^}GdeEa0C+%HP%2en7P|1#BukjOkNDwzG3`DaHHAAs^-&
z=YeWALKV;vLc1`=RGLZyG)(5S*-W;rd)Kp$_-EJ2WJFVsFqs!U0Lqx0Rd?`506N$n
zA=9q?eE!OpyLk8vOgbWQ(}!#8*u)?dk(iXE9EPwB%&c_CTP6bzM(&!4K_N%j1u2xL
ziMZli;mFI4=O>sw(2F_o+dXXcIZd%B6)cFRyL%~K5yu^GOf>PGOHkIgeDi%Blh@_S
zZsd5$n7T=-51tsyA%zYrgbnN3#DOgsP~qz)31AGD)j-}GxG$Ce&gTP=p2s*DMc7w$
zYz-e-0{)vzh3S0IZZ-+&Gh*sQ%9&R#*NB&2hZvya&ZT?N(T_<a8HRi0_;XknfpYhq
zA~#f&;#;lr^cdOL(g3^c+`e5N(K5m2#UP-W2er!yVR{Nhk>8?{tVOy}nI1Ot<gM+;
z8WH7nLEctw^j*>2aSU>QjD~z`wCVspiH2Ed#Jtd{fe#8qAB)C7t}@NIqWpnN&nkig
zfK=6t!U|3ja3OKTmWa^a^G$BwzWth{0k+`sPP)<^Y#Vza9l|J@fh++>ooA<vj4`YV
zqr-#LU6PV|n8We9x_YG(U!>#27b>)mDo7cImyy9416wsRjilnA+`D@>mB1oMA8>i<
zc&l*reQ{k9FfXTi#928EDP2H$89^m5HBk%`U=V|)_-5{!#Q!Ab>ej7I#6Zc)`er@t
zMh_P7t0U}ceTaUQ6ctsGf?z|Ba2!trbp+{Bn8xmVG6DhuFE?tIx3=oE+3ZDvDAX9M
zQH*)lC@m<5AtDNF@%2-Umk(UcRwCJfTFE@nw362TJ}4nt0RGW#Qpfx|WsFZo38e^V
zz8GRZb;jT|Y01%=VSqu+@|fN@K_pYv-tH!;YT~J^ahT;e3)*Z3yUR*w*Rq^CxN!B}
zscz5D&uNMbG61y>koT;ZRdgS@y?VPtVb|!5H~IoU^3qbtTYz6*UHW#wk#(-=f`bFn
z31(uUdHOS+P1~v;R(^!F3ZLEoSj$>bdS!m;Bm|~3XLL8J11DPC8c`#@IZv;n+x-Fc
z?7f%%x?B<b9Ca`-LO@L2n)0wvzr@+r5G;&ry>!ugZvjx0VHJo0x(`NYoCqucZc^Wm
z-$48$!oqN0cR~2-R?Qg(H$f)<abE<e>&azBOp*4cHr)~1j{60#9=krV0h&&|f$kVH
zm1P&6wg<NzG{(qUM*B+JoVLff<tQfYk?*<3Q_>~mUrAq|>85_zN2YZmiEH_Vwc!RZ
z#PY$Mf<S0MzEFUia2OefO(I@kW=&`fc0`KT<=kP*%lV;q{2SUbXCNe*Lh3H${Sc&k
zuH%r@Ko--0ha>qQRULTEYX&d5zQ^j%hP1vf8&K}E#fWU2s#~No0`bS*V%Px!n*kgd
zg6%{eS3Z7_MacJ3NmcL7>Yhhn<FEb``m0?~&beDJD$`Ii;=@3A!i_f_SSeJjd1{8H
z<iA`v_0I-9d-?JcM@?T1L|^)oRdOvTTraU=VkkyJcV1Qf*(QAiVu89X=mo+GegG!o
z&mIX=0Kro#hW==w(nym3TQv!Xwr-8AO+=Jv0eL<>s6HY+*nxp_@cf2_=MQ6hfx&8d
zdX^JKc&r(?(s>P;9KcT|p)FC<|6>5c(jQQ+4ID#)r%%FA`2dz!^uz)9KFK>8wS}2E
z37F1I+}Hv3^#qScnavftRpJn(0;C;riOEh=6;U-otVkw}>g5zDH3-Rp+YEwcQjdhj
z38{)Tk|=p2gjK3heBX@{fixdS;NNUEeF=H7BZA&Yt3MN?p|pw(m`67WZkrW=zp&MQ
zC0K!zNY@g4^~G>;{c}xG49K8e&z_7wUesD4g;l_z>4HR6F@#ja0Nu=io#TqiA2|t#
zf3Wti@uEan>$f{WT2V1}0tHFy-PL6jx>z_g<c+*kRzQey!3qirX@IxxBAWqFasWpq
z*WewSd*B9SXYbx+Xq8z3eNRVnVnE(j+t_3AO@O!~7@I+Q4&y6|>{JtXhgnb#zCAJS
zsaphUtBiJ^`9{)fDyI}X$~{WJXUV{O8aV}v9pWz>`3fPN_u#DXGIoZA-@R+aaPjTg
z3!&wEuz#f136B~0d=#z7xwwHGn1c>#yA})?Toh77psvVRPv&7_EOD=ZMJ7QDw8SBK
z#btxXBK1~x=@{6m1X3v^Rn0$JfQoK}#KecfDX9DwSB`!tx?ukH{sC?Hpq&Lv?LHsO
zFaDaIl9J*Ib`W^i7YCpc-r&b5@a4yEeCGnpO4rGf#r%MKwCh+kdZOy#0L%{i6ojL2
zn}}O@yTb*MAwf0kxowf>BHAOU4Pot%wB;yZ>u*DBN&j5`(W(CFSrT%j8x*Mn5kM@H
zXVoOyZG&bZ6cXcCXrhB%9Bz%2EwIicUI5=2FiD2M&+U6)oZ7MzM@xSkFN2OfE86IZ
zQ^TUwj5ycY>>a|c!+^Qse2GJWyn7Po@38Nh^5$f<-mr^4i=~iU?6<>#F@$&ULlH01
zg-4oFb#opO)l|^X4DaqS4?>;)EQ2c&@JCQ9`qr#(*97C3h%)(A;`vkq)jkO(eu`V$
zm&oOsu*BH&cOt?LTLvBAoB{f@bulvF_wvOTc4V0<kb8m&i{vdxY=D^1sOdM{T_WJD
z8iX)7LXlOyYKXP*A9*46f#Z$qU5!@(7r7al<r=o;(<qNBJ&UNnEZ5nye9e<*a@StB
zuYUt;2iP%$XKwUj1&R+a2p_Ovejaa3vMi7XoB8!StnQyMGhCU7BErmDJI{D<wYAC9
zf`MzYtNjP=E%8dlnDi#*^8J({bal-nwJ(W8sf4X&YF~9uA?&%=5TvuJuXEo-yHS%>
z`p+#d8Irz_b(|}RoOPQ$PgnX?@z2}}W}R?`(1qzl-<ZQs`b-jb{`q69R_7DzYMA&`
zpXN%c0&8jkp!IJM965<Y%tI0nFbc3uk{&Z8sR(R3LZ*R&RAYALj0MDp+Ocq<1j;2P
z8&sH|TZ}<ABmy8B^4)JRPYtNcJ0dEgsvkcd1TD~b_-=S|Qt!s7o6y=m^{sU8aVp`i
z#LZqapwzUT>!S{KF0|ArGoiMC%J*mc4ws}8mQVvNg9<!}hKPv)tYkcX2c_{Lh*7#_
z1@C8HF9ni&Vo<-;%a<?1J0wmV9KS;R&)ewzC*gC&wEB&T91_bH=G{n?MNA5B$61ZN
zV1RtN9Li6>h8-qZ&}5Cl;q}+K<90-cxBM_J^qcH|s<<$x5BQU2VS@foNI|yRr}U+=
zP2B66K+wERrcIMeuYTBmAC?NfRPw^t0HnU-XQWR(rF3Bz_3)u!%6ZlFIc)B2%rrKJ
zaneHkDfGL>nQ#F*Yzu166%cT~0XS$LkW{KVriQ?t5`-2SZ5BYN5o_bSz09ouQmvVY
zqpdzY^<n(8X$fU(_VSNUmS1?&!(gNxw`)%;OeT?VBDKhZP6vN*pz76wH90US_ghvK
z=LghA;xXOR0_{5ZgGlD|@@=9po-6~jz+<3gWy0|?51wB_IV64v76U^$l1Xj=VKg-p
zlQ{fv$^$7kRMxb#rPxF?q#9JEkQv~#xV0BMrNNS6dIH0yP4S>vK_d$Hd<ZJ45gMwU
z2vQ4)&I#;T3kx^$A~Hz331*glK?*%H7}KRjilzAUDNMH`rU5Fuw*XM!;~W7=vG|uE
zxiY<N4^;$Zu@6AV!c`DO@{oE{WakcH;S=hIMGA9sE-u6SV^fI5A|)<D_qi=`k<4u*
z{RJjla&&gdG5HAEAuE?xD~Wq&n$rg5q(=m01_O7eX#7QJ(U5e_bKKvU+|6w}CowDb
z^@uZDP8m05_lij`$B@Rd9=%OXYA8G6kdE)$OkG{q?pS27?Vc^O5)J{@Rc`zA(bagf
zH|^_>OZ$l-T!8gNj7Op@*v{T}>oQy1+BQ5_7HD0XghM5mi0F-_105WMm`sc~j)ETM
z2K8tj`yo8WX=&^}eifm%#;_>>)(<FT81%F%!Hg<=arC<PcKfXwh6Zcna`z%qJTfyk
zuSGhJWrmwTXh$wDDo%(6AEOoK2x&9H{Iq!m0O@7p`<ehc6WfP@#@bHZ&rh5}&07@!
zvpCd7kp~}xJ->~efR+istv&d3Wzq;8tN@&^1uX;V0Hy`N2Qu(tig*je?`)3Tmye#G
z`x9`{!2+a{Klyw0w1|K1zMEIq?7up;{sxa$nz)Ja^lF*u1MEQ|JUM(uBpVq(8XftD
z<8s+gHlP)$7nb{%K)XX5`Hwgv_O!dJ+!2hhZjof}M-r3hm&OD4x-qR|B;$Pxnr%{$
z+om96zKe1NAs+Rv>mya&2m$R>F!opvx|~mBbOktr_yVN{?);~l+{<wRHtK%z4+^>s
z-N5RP6|hR(nc5mYNvRhk*f;8bikQ#zP)m?rL3^P^vD6uR+1SEAxPQ0C@7=n;H2s2t
znlVSC3FeP?`}WsoM0X;6$D(~8^>Ttf+m-)*MBHePNNtHiJ4PLYw$yI21%0@K2n27M
z3J5}PtbXpXtkVdRFjs(fPRO&PRJQZ(*dYL0g7~83;+Pf^Lm&kN1SEtsPzK*4<P5Ax
zLZ#>g8lufcZ?X;dJK?v}AnQkbDOh}89Eu#lAriNEe+U;h6a&vXMw-cT5bP9v8&<8;
zz3Iu)(w`F^`!`+5Q5^leopJ8yL3+7~JY~OcoX(o$t|K%)$Qg_bSIg$}?mNP_zrI*_
zUfG3r7yw~Q@CTP|lOZy3kTPM@WIXV}1TI%IFe-6QM=3T9zi5TqL#istFA!Ru=9#$d
zzbhX9ITiUnh<mrO+-I%3F(Ue|gnb+jW`t@+mT`>s^XCvJggan`tw&Udls}@$N)urw
z@}GwQoS`d|g2!+Lk(AK82+KtMu-aEtYwiPDDR&Tlve&v<g)BeJmE{{w48*#EIE;~l
zLl6!d_0TyCVH06u114%_DjF5QISqctZ@J+9zCKHE)xr`Gh)<=yLQkwK<T&%dha|oV
zN@hzdyVB^kY{?|98Qg=9VixZoa%~BCl$10C4bpm8A~3g}(aEo(P@YrIoY*5^4tPNm
zZ$S_yWUeH2=9SDSbxR<(4HR3itYG5E<&|&v;qMBkJluN!)~%DL`%_X8J!1Wl*9pvZ
z5(*7ykd8Bxmc9A$pllI*SSV6?9jjOf4>=Tsl8Ug4m(l&ycQ6r>@xRSxI?!_Xij5(%
z)=5n`o-l;&9=wzd1jz{_Hvm-_A<!WSKPWE{D4Tq<r|ugj#=HmV^W>m}?9RP=wGf?d
zyp)C!;Q`~FC^&G~3$psAuY!DspZiwkRd3Gb;41MlK4xLNSyt*>#x9U;Y@w;?GBu9z
z*sDkSHop;3b-<A;0Lz~6AX0!-vr0N9<E=1BARfWFCO%-WHLF){{}3+6B!un>=tlAv
z|AS1sMl%q?BA+5r)lvI`aD14!`>*prL{5Ob{vIXPz-SpD-+<^z27GPp<~X7^mivq#
zH-Ss-_FJP2NEp80qS9|U*)7!ofZe05N+2~x_j6PX=ulkuBA7?}G^23HNoxbi?m(ad
zZU`^}Kx+%s3`y#PY{?od5Ir1ggkoGNpDuzm6r6$rhFMV2_csnPb=t!&6>#uod3x~j
zbIJs`1vG~yd?pu2ZYMu*pP<hDh*?-$yU*X}(5J8#qb{7_%@unxTiEP9ENu7QOYpT0
z&Xx9*vD?OQ(`&)GM5@PlI>C3T7tj3K)a)x)md_zcuFt8HPD2^r5&)$fr``L9K2ZOV
zF?5WE{#RYUz|Y^qFh+XNSLInODnX`8=+Zo<6I?wr{^a~7U{NF^2OcHP<glK}b)QY^
zLkuiWMqtY-dgcZQ3oJbr@cEiA%jWxRNcG22bBzX#TLQHK$**5m(k^)VIbcevz&lkD
zoxsf_)j&FwifD#kTwD{dHcq$}>;_>z^cE5h72*%&#7axis7VY57=&Ao>^c{T;1tvu
z(3Nq5u=J#42LGY0r-IV7IoWWa0m%?OPermq;Em?TQNR>1>v6~J6g$`Nm)|PS`RA7G
zf6WhIjgZnXoO0i~&7uj4#$(8y2GLmC0-A6vUiIp3B}9>oB9{x;abuqIy+rm=`mzdd
z1-9f7LV`ormJGwC(#4m<(;r7yVIvx5c^Tg<z3K12=j@dh;3PFQHH5f|DFb%wEd_SQ
zQ&{MRUhyaufa>9-28QVyD)8;7*eI!}Gy*<~02vkbD;1yz5+Jig)bkEGVT5g56{njD
zM&0P#{5jT^ljI7}aM%<k0@=RJj>8`RasfOvAAJ67j#yj#Omzs?9@oq1b%+8Ja5>1S
zTih>5^8Vegr)#^@&Lg<?fM$>6`A>&JY3c#2lDZa#%gBZyCZ9ENyCI~TeUzlcps%>p
zo040pvi%nFrbjqXn+qhs-IB*k4D-4rEP*zD3$R=G$cT`Ttub{wP^^QhcpB0Y?8YKo
zgQ;RHWv3Clr{w15hE&>Me^JRL&la}YRpKl7^8gc(4KfQEyhm9{Gq#K!aCvGrK>|?E
zBi0@P>?82*db0>|4Q!J}kUny8aTaZ7DtIeKR}XD{2zH=l^JCq}&LoUYK)PxJbD0sN
zYw4D-Et_0dFhZJecFgwP!`%ba6^GOk#N^|^hl)|zB?)=_oB}GsNKx;EOc&<q>>@^x
zBoC++V4gBUSA=3H%8q=jFwQKU@FQ(dA+KP+=|~h;42#lOa?#EMZh*uUS;@UvG(vhH
z`8*Dv8l1B%fvi{3tS;5whOz@O#fXjxLz(<dSZM@LS}2A#rt9xRb`)1_VyeyE%5958
zh%mX)RFQ_{e-Q!)LSMbcA=wqW!Wl%Dum1-jV!$ThuOM)=>Mkt;in<sKOW6X-!?mz|
z?6SoQVcT6)$Z=nI!H_EGNXN>O;JeR()Qtx}Bn1O0qaReoCdfkWSc%m2SC@fVB#{69
z0CGuj+jg;+AWnL}KI${g4{%$boE3<dXU??NPVHG!+|Frv!vI`m0MJ9Ee2SGbJ)H-_
z61k%&Z8d<bM%onEs@4LMc|Bz{1bPJ7VjQm;DC+h(?8L_c1xM572j~+eLAQ3Xj6zx{
zu0($rXhe5CI1ZW0Os=>y=@_r9`VIU9+4q;3S02FvkVA=$CS*s{$^fBZ`>5*oC7l}c
zZS}!?s!{v}%0<SKWwcjBhm~0|4;Ny4u2j*psR-IomLX&@687-D04pKc!S`x6;f6!%
zP{84OGc3>))X#^YP5c4OyCW>9`Xqx2;Io??AX^N<MEwR$yQBpmI9MB$Z1s~TLBL5w
zUds;;r2dTS?4teOtNnIi0%wnWK1$fU2+BcM1Av-xYU2xJI{0sz+k$}oKOD4|U}xQZ
z%3JMQAj}LWSOE9bz7cxA?Ba(jYUZ>r`xHjNFBG=lbRSbu2}eT)IqEp)u(`O51oIIM
zqADw&L@Xi8G8I8`S_GE&OiWC^oR|1AJ&T|11LBTE9CIbWK0VNQ8q2LgeI`4GNX1`p
zGKg*8aa`@B_iK>sfu6o{5AJk5>g5TrgA4<@_dqPg8BGBK_qwJ=jlcmZ*wI0p(;M;S
zqKKskaEp%zzy1~vqB+E_fb<4%8eCDol3gsHHU0b{Ylk>-ycG~ypzIjpzr~m2J3&^M
ze&up8sBdE@l>jec?SN^MTm{7)82%KDG_gX9U)buIHfc;E)XW!uQHOJK0+c0foMQ5x
zfVnXn1;7`|9!OUZ3`+lf&r5hulAIt;&Qy%Oh&3hw;}k-_sLf?qhh@#GE9eOL4xp79
zMr1CuTSX2@kO~6fLSqT4&lS{;I1_RJcoZa(Ly;jDBWp-0OdUQ_6VjD*3LZ}Ao#G^A
z9JE+Vdj2WIwjn2k6^71}7!+qbFdYlY!g{A<G?>VB`<?GOFVR@x<wY1?m;u*}(2nRw
zq?{7Thc!$?tVH4@6mw9QhfB7+v-Fvg>J1uZzp#X`YZ`Y=x84WUR`wMfkP?*|%7c0%
zo8VF|DyjL##f684s%`77z}!!g*OPJv0%jOEk`RKBioBhiJ75UHR6<9Cr`&e^h1#~P
zKe&ZrT6XWS&j9~Zs`QXZLK80zUU3kZy8Jy__yJTDr7+Qu6q7GqH9Ge{lyvI#dI2$`
z`3n312b>399v(iBiV^3<u1`{jrG=O#H4IK)P3#5_ir@&e0OlBwsyMRoS1Zo_w|aGI
zqw<5&JO>R8ut{Lvvh~({Ek1MR%+bcNmw*(otxafCj}Hh4;L2<xj5{1TLJXpl*BK|3
z8TS43#wENv_wUy|S8#yqNAqj<k<?UF&`=)Z615Ykn@<1CAGTP&PC0dW`rmp?8j3-+
z9-RWKf@kxWSkyg`Y6lpMu$5e7jO#!hC8jP3%SntsEV}5=6UxeIxjpADk+vda*CYTv
zj<6aA3*i9H4D2?sTmUbRaBK(M+`8)BHxJM&Cc!c@YApUtnk-0L6ELjnBvv^u{=E1p
zTQtqaXmMH%a~S9kq(TNyeCJLwIz&j+jz27J-K>FRXnd6cppYql<3C(4H%2}{!x8xM
z$w;{k-A3}lR!dv1TfP*vdC~(yXv@9(QT9y@DrM9R{A+KgfMy!!vanqR$>O<kE0fEb
z6TF$D1$0Y*K&XSqMqV8CAC!*#D^1>X!w#0g7n=^tIDAozQh7utZ{UoGL=&7DjWjJm
zk^*D27K}XxErT@+fE(EC@#GiFDWCkR_0=L6jRHp0VRsQIGq^6HBtMtUu5JgmBQ$t{
z(ZEqQjHE_6MZel}Q1->x3vpbWUN^Xyj5~*X_<O(^CYU-d704ARB67pg0n5N<m_8Gy
z7Nd{73H>W{3I@`oHx~6}QtECmy_5qa4(AQBYA6w>5VAVB=4voEvK;)Bhj+Q`<Pugg
zxw=gaCL&#w$BLgs2`j<3vK7F4DuLdR&$+{K9%9@Sw$u(hB_TI!C;V8ZRVm`m0<(3i
zlC#lO^$?eqz!zW~knekMaR&%Y4T4xQ!fZu;onZ@f%5DoP{KK<jhxJqMB{T3iO%D$*
zBYg{h)36An=upR<1X0`yQmLn9OgB?h#X}3Y8VoYtmmuaKJOuc<`jBl$8llHRr?7@X
zN!Wd^ek&tSBOq&@tV4PE=$(V-C-qJ+HNfox!zT>IUO+YxF)?;;c!-058v<pdw#9qN
zGBo@o!@`pqHA?q5Tf|2rBP19;HbnLMD#}wcJVNL4HD!c54Mr2WUR4m#jKj=Q(!ls0
zsL-Oz43QPPd=V#-xofcb(J?lFwN%k_l|n)Ic$g+^n;*~SLr&o_F`1-Jhs6pbMMQw;
zsC*?P@PM~$1m7>tSptK7Wb?VJ&<-O!YC=@PiC4kk5X|=r$6}MB8r;?pngR%!8$$ey
z1g~sbd&&BnPCDbXR^s>jGFoX<aUbmyAULc2XTN^B7708F$AI9Oh(5j{e+-+?bb0|$
zJ{hZi?d=Mds~xEZg9~;%d^KYglKg*WgBxQGf8W2GK69>rVS#(xoUW!~;d$L}#;FAv
zjZA@U*7e^nF!@(1hIgR8)3$AiW&>IsoDdzM3abSQ4cI?o=QgElrridT8|X2HDhMvg
zHU7rc1o5>5{wfJtJnES7s57v7T)%&O(eEG`w?euZyB?B6Z$k>M9XIO}$6iLDMvst5
z4|bBb+v`hpDB$ZtAmB;(dDe6`QnJEPA`lRwnv-CS@$-LiV6NFxj!C!s*@INCZ&Vi%
z5s6*DEg^MiwknMD<4N9+UC(`nYi=67<>+C6<$LaQ_qB%E>nTjGTI;m(P(%*;#enjI
z`&>KV$Pb7WmJMcm@O|uMAadJMK3z=#y0a}PzXyU!q~VOT2@_@}ah@Q+3{CN)Zw~b^
zQh993v(_d&*2d3EDXIG>nqP-TMd_Y4t*WftyvOEUjLeqMJuaIzZ9?uO0O`J_s-NHL
zj}QQx>==S4);+G;$ZFugQjoN27Ax2Rm_V4ogUF_&`uWXJE-?5pbr~m^Agqi+gJ?H0
z$ggvCB18-%6aUbwfIo~s1Dc7Ou^QYeVR(0~w5DGnLSYy<3JSAQSeCcjC4j`+et&=R
zn!#nLF=f*roj1Q3>U;!#2YA5^@;qv|y|)M$W%@OQ2c(QDm$1|ff{v3WL>LGwpbx~P
z!E1uOjjlkB{SHUbz!ysR;7kU~C>nfN0=3XRr&cpN{`m2OJL4(an+P;Vrg**6O4B3(
zPB5obKxHdpRag+xmQ{=z8ZtsU^u;E3nh0;W`ojV<Te91<Vzq96^pX(xdtyO{BK!ku
zVRvQEg!zRH<zMA`@|ZIiH3}9MG_6AO@HD-~mqyoIiAT35)=nRO3USw~k5GKpMp+8!
ziVzwtwb1vwFFaQ3E($$=AxXA9!Z4=)SW^6y74oXs^_wVkORDvKmLmN}C9IJIT*w}o
zqa2@rhFpeG8;cMc+KowOcW6q>Y;`Hd7mB+c!X)J*+MNsgDt-rM$rRMY7K5z3So8<V
z1r`^HYGTA%Kr&P`{iS%PjE(S5NQfJjfN&Hc8|rbdaW4BD22dcMAv)R4B}dpH)*;Dl
zM<WD5?Zm^EX&LpubrssjKS2A|Y7+B>@SpsI>zSW5hl$w6_6wL|^0F1#@7b#_mtT1e
zt4A_-Xn!Mc`GzC=*OrrhHP8zLQcnFm=QLWT9}Ro)`fLs!fFJ1@j3&PSW@v;@hNoj3
z+zoTYa!KI4rHt?-lhI2*S#(G;#=oBYbbp`SgDel!w~g!<o(nLc`e;l@{#RdWa?mg{
zBtuU*Mg85$rsd2-e6@MAbDn~R#Tv*yV5$^UpNJnuZxR3pf?~qz3hdmOGVBw1`BUDC
z0)fHl7~}<bWlJQM<nZl1*NIV6e7FRdQgi~e-xh>%q#p+PU5TtoVfd+y()0^I*G!8d
z<lfj)yZhI)U7j9y#tO&J2Uc@UjmiAlbK{G3B>&H%iqvg(G#R^ezKt}cMWW*&0fcn6
zHIk$|ix_2`8MNe|LObhjl0?#Vr>gi|B=v1)S=r}|)54`(U0}iT^E?_%<&pWA4Pp<m
zWvRPb$}ymXfG$Y;HBnMUCM<vu>sHO=LR-m?@7rtLz5a~tm~FT}dcd|g>+&uonv_EV
zD;eH%7rt{jE1x;;!d?}B;o7)62S<TDgF<D|t3HQ=djh!mw=yzpJleckXYWo&>U3G+
zvGzyQyHmgCLXAgnIi9w?y*;MIX^QJi*{$4~xzd=rA4Oz58WiYqx^n>D|EO(gb28IA
zQHX+;J;=P1Q&Zda?~f|m4qZuTC}LC@nT7IQX5ohnC)*2)1ofQ>>M1Q-T4JtT{lBPs
z?{KX9KWzL$MhFSnWh6-@Em;X6Evf9xrp%P=y-FELDiKM`CVOv9S|sa2k|bpB-}!cb
zpWkyl$KmLYyZgRf*XQ#duh;oHN8pSblkytrAzckI&g+PUeQoSFQ%GmNS(lOSK|bf@
zz8{zOVxn~%c~qG5P|>|vrrcFnYn7A`!=cJlsO+_71g^Kv-*frT>`dp@P0!yD)6&*9
z7&+G`7iRkdG53f}30Dl;2Odv^cJi;m!Q!8D&?+$^IF|{=TA7(LNW%)LwS*g8O?`b5
zqP8d`$}m#vN7mGs^+9pYJV*PT{7RjS*WeOJoQhYkT^qdY!CPxtSy4fOUnL1+Murmw
zHS5$bITHjnB4=(zpQyZuh)C-9l8lUPzoc9ah^|&uRZaLjfIt`zCBY1IaMEA$jnZuB
zC(Iw~NFra>f6~3NP^}S;*EYM||75|NPSLqbKlEo>fQeF}Nf_<`M97EunW{30oVhu3
zq(VSIAQTB!6ra~;ifkqx)mmCSTew)YPZ;NMczecR9ZTbwu)Co5PZOpgqesOsS~oQ0
zD=IEV^n-y;)tDlX{f}S1tcdTgbQA<S^HV<~YezesV21y*`aJMbNp=hDRkaQuR);5@
zI46MX9q>^_?$Ia-(k2=19d$nsDmJ?J70!vZs6fx3RUT;GJwFeg<Nm3|;=J=Ko4dS=
zJJ=F+cwaSkY`W*Qbg@&rA-DK;!ZqwQgJS0`^rZnDstw%#Vh=QvTx4hD-^^yWT-z#3
zPmFY?8tyTT=HYZi@Lg8zS2uTe5zg!DIj={=#MHmEv@hsp5SNgcb$L~`f%AGMroR-G
zD_3^HA`^>CF07hLUGL`RrlYS<qJd^zp5Qmpu6l30ve4^8!W4#pHIU~IH2)0bu@b{y
zrH6H`C;|K8>@zabcx$?p9Aj$vEHpG!NmfF&<l?U$>-1B{Q3(9;pcE`jwv{R?d+$UC
z24^u^3fj5#JfR}d+LdF9<OvxWWx^Y3d}?a$zu&qNzm-oxpL4tnEOA(ek9A9~6bay1
z`niXOgax*C+oFD}#SNfj&0de;#yUVU3{-=x*{;Q*u9DqHRTcX{%?ZE@Ja&v7?r(U9
zt=2!ttYU~CQN!q#J@%d~$6F*>h7})|^pO1`6^%5d3d^j7{QQ_L1sY5u0XMm^n`9@q
za*yNOW&qoil~_vV-*CZl3Xq#VQjww~NrHt#7naB_-HvZZf*|Ew|B0ic2}BFP3vrK_
zu-h*Y=Taz*WS;BWZ1n8~00paGl#3Bs;;`bi3thVxG~0GfvGHT5C`c$AWxa?9jqVFB
z{7_`%<U)`q>;VZr{3OTwWbhzXaNvXW>{atOh41uM1*e~^M4H$OY&Us%i#vAj2~DM`
zKwm$CL~SbIGDZse{%$at1SV#^k$3BW*3j*eOMuH^GqZI#=SkXU9%r$WMJBE!gyYaT
zj6Wy^jfy-bBDi{}<j#X+DR{af<^WMHFn(NESSWBA6$Q|J6MTa2w%b+kSbi9NEQ#i5
zynOoR5NKJI7`@QLdYvNYJ}HeuHx!VL1sVkSzbXrF!3*s%_Bv5HIG$p*fQXpyY5A=>
zSWt)*<t72X+rn;gH156pg9#3`v$L}t{QNf-7rk)q;HRdxHfs>ocww5sF+mjkNUl?*
zsCalNV%MFK<#UM-etkw2*4QZa!VyVbSueIBfP+)(LZ@~}Kfp=^Zs20<tN?o_Fd~A(
zxfSm!gvb*B%rbwm&Slp;h*vy)*X-1YJo39?EqP1Gw4<|=Exv#Cx^>~0i#3nkj8Lf<
zQzUV+NI;A4N06Oi^hH%w+GPto+54>M;Qn=MEwn+&?u$4)O4uqIO6rYk=H?V~v|~yk
zV^znhCC+Ka7R{lmMM-qsEYc|w^ES)+s}w@uEkS#oM&s7X&kf2hf%If>c_7o62iX7r
z9^Pp$z7hG?XJoOZ2!*RfO{B&OiIyZ==JRK|LKC)I8r8H7c%eA1mp$qX9S#PC-nda^
zo8J+H>rfP{40e|g^4ausA%7DMy$6h_kY1S^u6c+_M7*g9FLM}wZ-6{GP|9y$($Nj^
z6nw|&C=d5}3->oZsz-lF;w66c_071v!nqJ~ZTfU5gR0+q)Gi1>h8MC(q!#i9MYxz(
z^Yb(NS5~CwgK@Af?6L=mHpdCSK2i%t(;VE~^jQY1VZBHMZ~FfI2IvU1{I}{_)bQr8
zvRd{_L|Fi!Bc@@QxX$zGS+|)K_>iO`{F|a^)+q7)o^rbQS8K-R<*B>)Z(Xpn(|{=!
zvgUAmsMbAveWnT%Ch}~;dU31Z1frn;29mjdnBs$#iL0kFMonlfZcm^ObQK>%TVU0{
zb0-&;ykLDI%(7wnO!UFwn%e*8;dJ4GMA9HV!SqNPgd>X^K)ew^eIQ|*R-k;@^oy*#
z%=z>Kbd6)*@@xg-kr07r=k@LEQ2+rF2!o-akb$qdXt}#ofQcTi92|)-pTm<uo+2Ak
z#Uix=$SN%`w6I7F*Od2?MuI)@K0~a%s+yWe)N7<>Vg13snRt2;dtUHL;0XI~!U3nn
zO{?Cso%l7O=-PU5kHhyK8<6DS*%kKPy<}7p-S0X+IjNea*9(x91oI#yfEbpYk9V!3
zYxIeD8-*`f=CPL;G=i_m26$fG-X0E4NJ@oTK^Dg;SpkSU?y65BZ>#RES@L6le}4oV
z)SyJ)=6)K!g%F9-Q;7SfCB~%j!ong}e*vYTy6}gSA%0ZWQ|cYHw6!g}?QCqMF_`$<
z8yU7vG?EFmQZ>pt4&2=j9`GXriU!#kFpkIf_%$$4iOzHsGodW{A#B-*l<lTJ)8GL+
zjXW`y;*Jj=oF+GQAVbpBYv!z0@~4!dMno$q!ZV@z!!^&Fqn6l?@J5C)SD7#~AF8Pp
zn3p0y1!puP$rj3c8VE0)M~mYdfBgK}1ArjeRtqbPraHrDCz6cgkKATKuoU<YOe@6L
z9btd+XB_`!$#zB6x|6ti$g0HYkY#YQtW3E%US!v<=+pMF$mLQsO~#hSu3e*=$N^9W
zX=V{``~I_kC(HK8k@wuTch=xniSe<I=^?BHomqulelJGXH8^f>z~8q7g(HbWf#Sf_
zdtoA_%q&qS1lS!hlyLz`Bg>+if&FHu0*E{T@POdL87J52HD?Pw+n(nc8Nu(*jZHK)
zT)nL&PTms?R!2WSTI9lq`IC+3OrFs+hRnBceEJ&g!4~A3nt$py1tK*BK{U(e&8Wxy
z5qGa6DVW8-UO^!?PT7G<Av5TVThS3&4xrs4W*v3yeq66eAI4VFPQJY*9ThH#HCz{d
zWxiXEZK~agopep=nwm*iP-qZw&jsj6l5hirB@v-HNx_u8|E@(%B7j3S>}HJXAhtaX
z9EYLkpu&s7yuZG_zBk-^UW|B86HOY)_qDP4!`(HEwvL2uPA-Log>4tg-*ZY&kCuXZ
z;D6ziG!E}7!BikgesOWDqP6>)+oEr-!#aREwes<SPzS_g-??)qvw0N)Z(T5>k9_#>
zChAHn!Yugv9Cw)N{CyNcgro6o0OFU0Y_n9HB!8~Bx)#>g9SGStq8N@@A*_S_`dwlA
z+8n-3Ow9k(Y7!2vo%{(0PCj@JBy1Tsw8-w=Jalw)OsiHM@b=y(7p8_R48+3`(^hlP
zU(1BIVu?GRQG93Y8V|0A0)?{dkoXpf&8SD01L?tLg8`^QiR@PIsE7z3zy|0zicdKj
zN%Nf4)61stI=p%h1~Y=Zd9ZII9zHy5L6sBbybe=pA25r1K=+<Tl6vch>j)>a`TlAp
zSPJBvMzvXY>G!cC<X*giFAQep`DYuC#lJfm?tGzWu!n2yHq-I;3Z_vXES@XQ&UoP}
z5>X^d;c}rMZ01otpl<M)6N&rr8>4qNKs6eO^g#UzItpUY*~5CjS@(m5REMXIe>6PU
z&51R)v97u#H$Q(ZIwA@_tYXhq?Dt^6GTjj0?-v;91lWe<8&GMkf}{RrN=l+0`N6Hz
zk>2+8YantS1QAi;V|6?O8SQ%ADY`G(t2}xh!AzSNR^g^4;j6^EdD%rDmfO=XwYY(!
z(h4Iq;##_VfXi-0x!2azRB_a7NR|D=>&+s)B(ELC852JGi<8Xp{p5aIySBcn18P|W
zetxVM6Y9boAoI*?PW1PT7%h-;iM&m;LLus+h!Ymsx|Iznd*y^lGwoJDPR{G2GWJ6Q
z1MrH{)C}j>0Xjf&5PbTV1s_Zf`j(KGm4T|^c`q)fl|*YN<MKf|lK@F&rZ}8b$$#}K
zp{VG=!mGTzqK<m4Q)PH{sc-fm#1^~)uhP4IP7A^2$WOp0J&v~xY?Au-_qOc+!vz4>
zfZR_g#wl{nHSnflgc;ekOt%SZX`@$?RMX92y`<&JK6QWa=$<TtKP}!#@yJ7~81v96
zTT1ta5=~Ch-GL=g)j@FWo#FDqhYcql`IgNd9*y<cBP!}ABntZDx8T1Qf|d5F$xslg
zF*Z+G$tQ+(7fTr#iuCo4yu7?V>Y_9w-U3EGNp5?1HwswU0{cn*^c1%*5MUKtJH`%a
ziHS7WV$-N=w_~2XF8nT9A{QCqpJ}K`Dee6{YAEq2VWSueW#~)Yjfv@nAWUFX_QlB?
zFd%h5E0}bu12S-tZQH^Z%#ieqXBJol!&+WJ!7If_U^kB}o>p296uTIOH@tprv4m8s
zMy2~HKUnkq4YjnGFoz~#&#3B%!ywL5MhZ$S4-DUypBM)6K#H?m-PfbSr<w3bKOn<k
zlJ2+g+~0gjfN65dOq$32On=;0YoW2Z5>9mEoyesnn~p|q@}8-+EOUYW(I5T&?w2nF
z+$6(6q>|)JvU+-YGOb+cdr~S8v86b5O}5qpX7N<|;={qo=?9dI<dfn4L`6V{{z}Kp
zjQzU0UuUN&%6TjN)pGke`Ro6`Q;>>8{ls0D?wj6W$Nom`<KW<+BiJ3!PfB6+j^57D
zzi{EXadgIi2J0*ykVYY+{jakCu5vU|S=wjAdLwS%_5;YrGb^!S!^fkZ`zjm-0YH-k
zKnT9si7|14UJ8wzR1YyxAVxYQ7###);ql7#q@C;8bFdCbcB}dfTDe;6<ZG`7%Yn)N
z*|{Y`y)KFrGPs{5qOPYEY~f(SZ>y5>wgclbA>%3UA<Q%D08Ig<ra`HTcZ(8<_m&2Z
zTX^NGYiN)`P92Ev9!e&-M!oECXvQf>Z~Mwwg|_>1nSr!4(!gPf^lN!f{hDwKSITo5
zA}NK`bxYHX6wL}qZEoVwz%Xz<lFUJXjCmLZ*j9A+?z_2BzlMjeT^hvqGyL}Vq?uVb
zvQ|m)Oc7K_%E7rB7x#X4Mm{A)DX>=#I0<55EP7cObYI+AxP=dioR<8zI6QBHJH-Z2
zsuoZ#LU!pZR*#cJEb`<sEo+(EAz$)373sMPEjk25ZL1k7RYfB+Uqb8hBgb07?DF3P
zUN<gP8IJ4ff1H%=?W3cot;{qXtE|Y~f}(0dYfw6@*F#s1bj|1m*GIi3%_N9HteD}b
zySPk;M7_rvE?Jx%geg{oR>oHKuP75Cw(G$y4@eJ|=?7d~<Y3it<MCtB{OW@AI+Sy*
zPW<ojgfx-zX!3bs&613aj1O=eH5~YoT#w7kA063>HP||V$W4!3H9q2;*Bu&_rl_!e
z(X_M_h2e1o()@1RU`FqQrw(2Z6~4QgUS7(gVq%_3XUNO4!ON<yuBP$Hz0G^wsrslm
z=gbqg7QV+}Na|^`3Lk1OfF{L}rAPpou1!roa7gCy5t}}W#gEi4F!2y7>Wf?ZIQsy1
zXze}N*cy8@p36ym@nR*yClIUlJUv}_vII>fa%F;05K@-b?mU0?Y}TWKf`aX7|E{ib
z@!v&t>{uylR1p&u<&WAc0l{S8AZs2tKu@W6pM<HF1*#vCIf!Oz_g}WF!1>{0zMpTz
zL9jlr^4&Qno%8{*NUi@)M=(F>OL#&zNSe_hdL_$1@J{1tX`W*Gl6aR@>Y|ayoEHI3
zAu=ipv$s^!Qht=A2w5R2thUZO<d~TsttV;Ft*)R=j2b0dp*n!vz}KSRti)wCuHt&A
zCkg&w^DVRc<;K^Fjm1Z{Y8JEcbs<)iE#>PC90L#`1KPt{#NNn0=nh+aa7#QDKUM62
zH1u$sJn1vla3iDn03VE`)1vLj%i9fsgE>?I(;x;8EZ@naR=pOT@)(*@!d!n}Shxzh
z2MzVM$fX2h>~7q61VF-W$dikMJ)EYK(+(shJEBCU{7ocm5JY~~99KiF+K`@uK9CaT
zw>sD87IIvX6Zqxp*TRLrBZ`f}P9Qx{1Oq8j(KqPJ083wa6%*qi7>v_Kabf&3-k-+7
zgDU|kuvBD4ox%O_t<-%F=}#~n6%iG!sux44$a!;Kf=cfLJLD1^DlRylNzsR5{+V42
z`;g%zSb|BJnN62^ohHvpNlL2VkpMwl&%uOlm4wo_n7XfmgF*#LD}tApezuzb7gJsX
z1ClzIA2nvAuE2$Zw~ildL8=@ZF8M$5$kR+=*W@h#qJ|@l6z=Plm4(4jQ?a&|dS$dl
zK`|Lt?HElQ*uS3!LyrHg`FCCYY3M_6WdIt2Z@8Ma_M6PXhjNaz6bJOYZnh?i=U~_w
z5*~g&kPE*zHISV2Xl{|u*>1N(_t$1FW=&R4O`{S-3eiGeQ2B7L6Gcuc<iUdnIHp<~
zn-UD4eeT9IRL*mDC#AkbV%M&(TgH(UoNQVqd^axc7Hr9wCnhFP-dmh%LBRz{&<4V9
z$MsLbAJs7v{SOzoH+>2=i^Qm#RGnQA8cl=ivbX0d$k?I4e7+8TWk^g+oF0ZZm_<K=
z+HjkxNknzy8m=bI3z)TLpA+rMetBz@B;jT!*pHHb&Qk2ywF44S^b;C)pMkz%fkx?D
z&UtnyOh2NMLnZsAr9}-eUCP%U7(Y)#TOGF6)I~&cZ7eR6%YE~%oo^fk_0mnQgC5uo
z4r}$AV~X9*g&!}}Kw`DB;^W*SIbi1&Fgm-Z>?yC%xNH2X-Yk&t{I<Qu+ez>+x;~0v
zqW$H|vMe4uWMxUcDLu}8T|M>m0c2lOkYJRZ$A<VMJS02Y6)y24dITA%h6>k#iStaq
zJSp|^-$VlFqr`iftbCFg2{*zKMRtUW15~xdsy4_prvYg)5F%qUb^!8P93#a(YbgZ8
z0tbwgUF79oy0jUpP)xuh9z6IxdVn$g>rK8L6gdeK`M5B)MiD!VUM+H3n_61xU#izw
zZpXGWV9!MJ_aMam8?MyjMn)(}(U2a%hL>CpufxTNU5U+aiSeS1SSab}ZtZI{*3U4#
zMUVk}c`Yri3cyt~s2hL>61gx|HZjb{D}Vu%_mUTB2rziB2D7d|>Em2@q$+?T)15na
zW>T-cYHH!DVq4`I=WB;NLwuj0Ep?oHd;@xYm$D3+zI|(;?ZQ&Dh_1!rMJ>w(jFg^Y
zv)Zh`$`*8{kn7hIrMkAm?8J{Wt0cr6+^FjicA@R9K9@sY4oP3ND;khUHOgg1N&8<r
zMEN=4GZl4A*&^i*X$E3S($z)x;{)ItD3!?`>@S*?3cOOo-8d$>Wy=B7kAIvM7#JB1
zCo~KUc#DO%r2lk9YeHn2gqbbAVMAhJAs>o>8elg-RtOk{WKMJ}mf-8ZN($mo_50@X
zC_gATxR)S2pOLXmfL<6*XSMaAMwo`6ZpR<z4?fnf0-xa*b(pgEU%CQ!AT**uaZ9N5
z%`uO+8a+9GR9$@qNOzUYiWJRX3a?7l?Cmq!hD_AD2*%=*<hN(ThT61Gdp2!)y=5HN
z7%&_Bq{hW+B=PC>4Hi8O?yx%_o{NYYN?6y*^M`^_&$F^jz0N&|x2VRh6aXp%$=o3i
z4M~q}4EX-F=tPd2nB2q?on7uXXd_zaI^i1mXp1OSe!H}klx4rd+BLlPQkyr^Q1FOb
zOwzJo&-tgbQ~;?62t=wtWsrrI2L9=NH>UC)$vIknITajtiZ)(?gUL_!hx5^zVC5++
zG(=eeNnZ<i18pCkFuFh#)PnUQN=ib2Fw1x+?>TfZt-h}AwXnFBo4Xrb7(%}fAXyy`
zjSP%X-(}OcVC9WM1JaEPx3mWXDmK-`ri@3$&kFKl9Mj(j9DX<i$oF7V?xcg~u>VAk
znIxBAZ$X0RU!>yl3k}oaqPqI~B&p25!Hb{h@sN*)5|<#XKx<m&@9+BFQ<2`!*mezx
zr7H;P1(rkh7>;QckGSsPrY2y3*>4}Ji>}6PY4}S40#(pzic8IaliY(rz|RkHEEF8r
z)h|7Q^|+@dCK6LqS;31Tos>)b`BIOZoSew$=mVuz8ES_Qvl4$MfX`OYG*5#RWwAS!
z#;5czc9fc<<7h$r6sq&leS*V%_hZyWE!J4J1Pc$%k0Ox1cWwDJg3Puo6Y(c#k3iE`
z1v=O6-+idXYavRUTdm(O^c!JZ7NO@@iLg%*L8kh<c3)l~%i5-+h#Q_UzF)0tF@LM7
zt_C6EwO*kYo{J;O3QS!p)d2s&dk>8?K`kRu^<YV_7OdZoNJJ?qX3#vUAdd(D*ZjYg
z<FQ_+Kdra+0zHcrE?mE%hkpkKMt{%!V)Z;dH#<2sWj5(-ne}bU`1^(i6_`>#7g1^)
zL!^#fX#TF<yR$h=MI$Q^ohKB?B<I%)h?-~mX>u}s`-itpO*(le()07LeX*Vx{MywO
z4oM(Cj3NU+Ubs>$jmRH(S)roKKKA{b`hf#T0L2G%U&!)2lBlDSs6%m(HoqWy-$PM!
zn1!94v2jd)ymie5OauU?QV=ue^{wli3~8ebq=oa=$nav65D*+p=O7rl6J_kraE!}5
zE{kWjY<B6pyf>DkZD`WR;N&~tZWwO)AwUSjrwsXadO<T;bPkdv?kOaM$w812O%ZT1
z8WPn33SKbO0hXXq2)i)UMB%&v0=EJ%AO?Ftzx~|KC+Dk+R&L2}?|Ey%5Qi^MRwFoo
zI0E<M+y@+W{PgMfJSK<;uQpqn)h~g;Qng#}YE`C{9=>w+8_L%5@m)u~{_epp79X(?
zA^zFsmKJsvQ)wO+>r0o88ymm<um)_QW5=$sc*0Ht1q~WA0z9l-BawN)<NdiEs2X&P
zjYBb4dZ=*C?i)jyMp7lj2#&a`CV8O{OU+dSK7{&0X@--Y417t>EWx=+Ip6^lO!_d|
z0~Hj*FeK5zfdcycac@$(x4awVE&&S@hYdXi_riKaF;+~c1yEguNr(@uHT`f)&Hvm}
z&NsI0=iR$J$k-L5_yEcr&MwU^oBh4{RepZBODm=d?K~m8mp&*Zeo)S-)R5*;6Xz@(
zqF223SCB|4NIedu4DhiU8ghI&3gY}I<Zcyq|Fhj}mHS~pEIcO0(maz>e;OGW7)V48
zK)A;xC4$f?0%HwA+%bL(r0Y;8jesUuG{Gf!>#&6X`*Q+o8$DAoSmE*Dy;bKpu~(UA
z1uYF05l*JyQ>ygxfT})7#lGjcb4mtCA+7KF3*}P})A*n=seeiJS&FrCIg#+*xIR`T
zjath2p$Sn!&1tp%N1qk{u|?t_i^u2xIuvkR^KH4#Pp5ny0B-+BudnUPmn*F;Vj?2G
zTRKTWhSJ-S`>0!r-!(}D6q7E;%yZ&vrk)UF(gc$O6xJUdckpE8A?}j@`$tP3h>7Rc
z^bs_V)FGh68L$60Rlc<Fviu7%r`s;nA5Go|Qq9Qw01(~t;#vH^h8l^zd-+IY7Df=b
z0JaH?Y&9#Jza8Ulo&C=7p>f3}8D7PIa*Kzt)9@~orobFRUSj5Jx_U<rrWGMKZde94
zfJ_rU*8+E*ET0ye-Ql6zw0%{4KxpoUHR%D&=THH(^O$@doInX*hw71Fv-p}O{}wTW
zPSqPM@ZOdYXdvy(;^M#M2UlpVG<Bj`<8e9Qemfd_nZ-qTdxg~YBzXK~4)mihff6-I
z;B^_33mxX<b!)?VF+>Uk6$PqIMjxxTq-W29bFVt+&9bffg&`uoQ|s({pCo=I+(BfO
zf7*kS4ObFSS8&Z58XFm@->7LRDQZ`*?qcx__9_GNm{eLS1d`ecszFV2ZkpP;b3eIr
z@3QR1XqKeCLrq2y;y>+$QMR$W_U!3Jau^E;oxmOR$UT2$1s@RG%|9oYO_M%`KuUmX
z@#esSn|ns_j&|5huO~+c8J{MVz5o;<=JD_K#_54$-Zy~fn6`Mj%2KaRM&%l%RZnU&
z(gmsl=?IID7p8<IjO|<O-3N>(9D5ja4fN6|^bSARxe^N!-{l*ulnP$D@xFHHSo3B~
z5pv`MpfqA&b7PBJOLH>=LTdq^LtT@QkU#+~=*NbxXE;#+QzWDYFtM|r7YWZ1Iw<4#
zcje?c?g;Yu@!?lcr{A&EX-Al+`;DCAK8bZz|JQw|udKaLhF_uIl=i4%ILbzzS$Y4i
z&G&<kU{HgdmYQLYo*b=x@cXN%n2n8Njj0OXKl&X-WdK24!-o&&5;L~Fysy!l!rYmN
znirKViEKewV&Q3Us{#QH2_Y<7c;(v{elb!mFZ`}~*ye{No5vUH(hu{3TSrzBvf0QP
zVs8@UlivLouU|>{FRgjX@7d2aAtuE9hlBwmm^Hr*j04Pw(c!kL58XAk<D2UK>r&?^
zNG|4i7$&Hp_y>xND?1S11wAt$`!8R<te2K%g^MAb@rwtj*r_Z5uA5idVC6_iFo5D<
z69#e!K*QiuSZP^E-z2kcKq5I~=qHg=U+E=r^kE2YHk6kln>MY)*ZGr|ZYzY-x<jhX
zcDKOb+7kCEJBo^li6DyyjC>-1WsN(TW1fQv$NGKy)}f+6Tectl2Bn^D4BY*sCr?(u
zbwdDC**9VL$QUB%pBw$K&WC7Y3<DHQJRKZHUY7qUy~@Oav15Uw7?}cr?A(iDYQm+U
z?c+yZ$mW<RBN!;H-@bjc=x?}`i_-ZE^uti3211krf<MV`#h@QhD<aj+LkCKMhGUoi
z8*Rwwa_L$Fg^nKB!*m!SBNI>S=uwTcXEzLP9I}c2he(Mq`NF5nrZ1_<<d0RmeN8JB
zzPQBZhh6J@PE|tn4~a%F4nPoyaNg12A<|sCfHYxEclToBUu~V8AAUZI;S>_Og*}7$
zb~d%eaN##>8$JT=`Tg5^(y4IR9}g+N;u^B2P2D>dl5t3)<AKf%z>i9rUh!#(n74!i
z1>_1{-{~o<<Mi83jbP18x7!J^R4QJ>Cp*SQU0ULV)t1d;#Vb|RZ6?(nC~!=#78L)#
zy-xjy;laUbfOHOnY55qN`zkF}R_=eO?8%Rm5C)1l=EW^Se>sP@^WP%PH3~H>WE9;v
z)3DfL@4z5_*ZQgs6>?Rgw*bq83RW=CZo<C-=LWd?$L8jsx;hPbcF_~bd`2(^@HLEe
zpG%@m?jvjcPgi-@kRr}f2(HA%#WR~)W`?!4u-4J}s#Lsmz##b?&=J%knDO-Be=R+3
zYiD<^yzOX5M+b7O>rn7h<Zj3V7dD{A-oJkx3<gxe_>nd%6I5&@P59TZUrDT-QJTi3
zODSy&Y^<zQ%z{XxMyc1h25balstA;Ya63*&F+e2_|F{aE3C9f$uVWes@eiGYjWHg+
zIf~l?lq^|LGk*_TAHY%NqJ(%Y@1<*b!Q%i9gY(xd7GXv~1}_xBs7lx&IDHu`!+)W5
z@SvkYqg{SZLL>S$LOBEXcl+-J<m8ZYs%5)dk1cirNHfyZK38GOEMB0$Hrwywrki!S
zQ*#hj2Ly*8^#RuOYIH2*cyZ5-V@V%Ej3)?En(ze)*F20z`?+q6JlaIhV>I;-OLKD8
zAc4DjE9$0dRkQx`4>vMLRVBGZQN4g)ojT#yYNLUEez4yd;1)o!eBoL=HB_wl9X64>
zFyIAwm8{c8XzM{vvB23+5%f*J6lj0?bX=QrWUZy6v0zxQ+>@tIMK^9_B!h(yAJ`r^
zU&SGUq87v-(449<*7PNCWZ2X%f?fmOCWF#clR{3M>B_>}vCqxM+$!c^2pu(Xp1*O;
z+4%n8p$=iUndlqn%^(&J9xh)J256<?>8V5-sdeibd9Fb)@~I$|hv^(}c%^YS@)zhl
zFa{#Twhex|FKW=7dMIl5NGT`?kijRx)jxms2UYILvuA|hhwIL*RG4I=qEH&aSSb7M
z3Y<B#ClcNnzy9U5IQGa2?J)r^FtP;$2E+yB-&L0|+d;k~iY;MhGB>$2LFHeptFI4_
z)waK`UPWuP5tky+`Y+AR*-h6j%&`I`!SoR4ZRXfR+t&}LVu0hKw9J~x!1D*%1@PD)
zIiu3tJip&n7|0bS#WhgpfU8m8!Z!&yD6sk6)ci<Pr6Zu-k{3kC3ZOK!@hqS!BDBMH
z3|qiw>C2=QL%mvAJ*ece0ee)kSb6+He^i#Hl8aJ(tABo38MNB7lBidja9?7`cmHq@
zDvLAc&-?evNuMs_+Bft1n1|Aj?Q0}UZ{mTDwskG?Ao&e@NELMso-){Xts3>Xi$*Kx
z*r_mjbtv94K}{AeS^QGpX#AYf-eGDx%eZjKPWy>1_TyA~8OwW@cE^5fZM_Z!7bd8W
zOH1Pd%Uk~KMY+FMdVJz+xmkaSCk7&5Rv={oTqvGN`L!pCZz~63b=aLM1NxA<F*irH
z07;CXK#<+6kyQ9&;efjO`aOGMPR&Sc+-P>3I+&s4lA#Xn2*FoiNy{!Uh0I0*lVZ|g
zGU@f4aoh68>~$}ma9c6f9_=;~mN+uvu-gz+^^-K3w4cCbqr|>vK$`b%3Z9*@!K#Ew
z)<ta~e2Kxr%WWb~DO5Tt$-0&1GCuvTtgNi0V27Z)0=_dZbp5b&c;6fQqs!e7m@}kn
z%&j#3)SSGU*?Eehk<OE^*0`+EVRt((HhGn&i#Py!?9zJ{1oRFv(`AaBrkp#_ExkTt
z^f7=NcC$O9sZqtZB27MJsYQa2D*&RS&Raf=AXlM;MU<lgT3{XOZNr=c0)}p!F%&_~
zhP~CDaCdu>ob2kc-^M=l)~(t`ugM+Q6bxh_4JQkhoK{xKGO;gS@j5A+3JQ3!hrQ+$
z)Ilw^aaAI8YvHo_e}^a)xMc|xZ_n<_5$=?C9`qs{Sipx-L}9Q`Pk}wodL^Zfe1GvI
zY)npqT7#xbz!DrLI0E-$>*ChH3POD_ipJ>lgk>n14YNPUza1$V{l>zUm7>|u3d%0e
zy!_J8@8A4crESsUb3**{@^cHbCmiY>1>s?^8iW}*#jT`(KHK!^6WLjMeG0E0WxPLU
z6<vXrnx6Vx+9{|M4+%p?bqruAnz_k2$!fU9`T6<V!}vc9PS~;JA57M5{vd)w3y+c}
zCrJV|GfJ&(1Zc&*!vG?ZMRW+nm>_>^qH84BCR~4}zYepY^v&uAI(L5G9AB(tW9%Mw
zR6Zf0p-yeZB())r>1*^XwrTy9hXE!jD7SEVfi{5KF|+n7B5S`Tsz5LI<lx;$M5sfq
zcFc=#IA9(s28bJ_iCzjfo|zngFX!RTTd2AJo1!cF{7m~{jSCSaEdWo^ZQB}$pp*(i
z>6{Oca)sHUa<gRJy_Rnoy?;*)rhYhb=#Uk6p+-{VscF=%pz7qet-|$(txNvb3fpj(
zU=FeH?%`fS*TMFC{^Es2XwKK>=As492ZtrlM6+RfLPuqKe@coQn4T{{B!-&Dg%%bT
z1!Er+2*zRs)Rhb=6#ke((hSl%dDD)L+W{cCe|@A+-Y;Xu6g9VPd8h;}ecia%i>21+
zB;k_yB=AfzKI=sfRf}#7s4pYF4ay&Ko1l5Yhk`mu)xh9hMzF$D+N_x4XkZAZh5+Uu
z-e7tO{y@{<1b&d`eh*94)U}TrTwq;x)8CZ}e`lv}(f|I~w<Py}xB!&`?Vj&wOBuJl
zH}-rRqvTThI=5@M>37%Qz|<hCXIy25lQ6T1H+yFm3SZPO2H%WNobV3{3bIHO3c9VG
zjC#>WJ6RR00c=~mN%JfNG8KCHQi^mTMMZ8kjT<7hR`{w=P~rmE$EhlYTM%SCkPYMA
zicE1mIpVAaqxzBodoSulat8ud2*I>MxFkWiq*BhFrDmu(h5MaTi^|{1!3m$W4jz%i
zZTakA27|*zkACmyL%GVWkFd8+4I3(Y%}*P!?=U}JF}N2vB$UTHF&Ya)YB~yhra)Ds
zyRXL*NB0EFQ#!AOL9;ox=T=*F$|a$^y#e@_R8{D5|2JR<Lk?&QM4Vah`BNNww<9Dh
z?5WRMt`Xe|0S^J@O2vb0<34K>TvN8uAF6QQ#NcmX@M}=fGq*BTWxWdMZ7iO^1y=*c
z!IhGKRn+6Xb?#%}&X`VOY5AfC)7Nh~Yrj@N;H{Jg87Dt@z)ZnqcZR#{!%X20Dt$#*
z0eI)c1||~J?(g!@y*DZuvO`}18xToyu;UKNe~TjC0aJCW?yA6;RiEir)0c3KF!VSN
zer5VuxS77hR3&A#J3FS3pK%<*#DNyG4eTetnc$;g(4=W(#7`mg;d<c47!_u%eQtIK
zY)BY5NE6a<QIUM!eit0WlV@dMr$(S03?&&Qsr1Xgs4GZS=Gbk25vmaQ+@=1sogdL5
zVt4V(Iy*ZT(N~B~%n4~P<A7Cm`55dJotiINoK`!Ka`c%McNsoQZf<VOs)Aa}?YPe)
z(2<d0{MoZ2ii(1ys<5$92U8e2|7|miPJ+QGz|j9*r!NDmYR$2aR>0k`0DMuJ<~y%L
zsfd9Th(a}(9M#)Kcy5uTCj0E+$T!Z=kZC-<i8?|1Kyvv}j6o^egW-`HK#<wlSx|fd
zgGFLu4R&lmP>?0CS{r}Sh!y|Vo^^LuAkJ2dxNE^hL<1Crc{%vhIJp595uOsV@fe|~
zBg<lNu-k>3zNAv=;8UNNpK0+s$4CBUaS025_|A=dRUfx?aSL7wD4d#`>!iGJJ)!=p
zPT$BAiYd?5xj3QKt1GKIp<Q%FarOQDg2|Ax0(1q2-%|8aK7Rl1*4hS(TD4?d3s|<?
zlRe9f;RjeyBsmp>cI-LU!Ig?9F@A)_@V9WM#lc~h_<EdCneb;N2dX%yZ@=q*(_w-|
z!bqg|=Zpz^p~(l@Azxn_Nbc5!H_YKdRF0KsR&DX#7_HqJ#-H?&SEnsaUmNT`an93|
zV_R{_i8gQje8Se+WL<SEhOY})JFNrY*3LMH`-an+HUf4cTMyO{4hRSW69ND%IidIL
zA(H~^+XBQ_dMcd^XZnGe7y@T*67M!bMR<P>15sv>lfFHXk(cKu^C6%bwFzmo2|h8j
zrpzvzsH^UhCwry@M?ET0t9@9?7i?`0AcFZ6qY_pC-UR7Vb#*x*<Y>Lv-|x4FkCox1
zzJ3yF6YzgfFctlA0v{4{7DD^sF49OM#-VcFi;85K6%WC^jesWRkWvy>K?>^_LaUMh
zDZ&|b6Po@(X9(D;3B4M_m&{>j6qx$;*15A!o;TXln)WDhjksx?II(82YiZ8=!i5Ve
zC~46pg580WB=YucT8ayzf(%oFRMz~W*+wbB)$tA7RTMVsqqQ*`_62hcU=JpFs8$r=
z4FnJ{IwSb^RkFZXbSe(f*pI9zsZmgzaeaJJ%8=e(#><XI1#>Rcr-l=cL8t^J9!%hl
z#nMlA+0y~a7}nZ@lQDO#DqyAQ;KbW8Mf@$m=T(@<qss<21H~Ia5Is+uoYr=!0xL_)
zgk`4Wtv`Qk063jIbqYrd3*Ju!l(ySu3II&rLrsHaCIh>y^3w);#l=^Er~fVcEDPc-
z2AuC}iK}AlO^O86i0f`OIiW|!4@HyyeCDJGgP?+j-z@{_bKLC`QJW+~ov@MAMUU&~
z5nh>IrEQjh;%#GmphtqMR=d<+U(+K7^I6bl(c=caKL;QZ<8vF`7?e7h-FBHS7^VUa
z55<`?K0Xdj3^_VbGoY2hNej2WMPCpKaHPyE55pP?e!i}LdV2HW(`#Fqc|IVQAve3l
zP%7FI5W_t<$+TSsR0Ha0ho9N`w&0J<^0@iHCih2so<9afjqynAy?Y@rQ39xuBUEd(
z3F{OM>b-s5sI2bk;{pVw61~pjmoKdiLP?G>s{9_%?kIv5@S~Q%UHP|c&JJUZL?q9X
z3jX<X-aLxQe^chHWrPTe_fmgEZ0w*uH=#vLf6ed;#Qzm-Hx-0PY%fgPP?3=iF7nZ%
zaCp**Zr}c~rCMl5di2$FZ@NnLtuZ-M^4|*b?yhW}G82tlF*G+kdm7X{#ee+MFE>0h
zV=Td{2W(&yZUq04a831w1{@ZIIN#C|4AcYcj7e5`x=tWd(=K;@xS>2GOYzsAKQ)*w
zq3OZ!juE;P2}W!kLRuMpvpHkB_OUohoD%<^Y2rGdFrq~!<VQdV4#n>-?96s%Pv<Td
z5)u*<{%E;7){yf1Mz35=T^(Q_KH7~NHryKMF~a~6>g77fHkg+AmWIO}|LYZ{%r>HN
z#8r;^F?$S^FG24>1hK%XCKwBfArP@;q(`piRea00Uq=Br@Nvt?M6If#8sclHm@U>q
z+H@U~h#n|BQ}>O5<>UzVG73OOjGIno5>C>aVOrDp{w$bsD9-)K!I?0h`SxL_$Jh}J
zED&n$(jDky^6JLTby!(sHEJYT2soph0D1;u2o5(?FUrCj_*k$-2o+@%W|d$b@<T9&
zxsPwHrSJQ5)p&q_KSzPp3!%r#{0{{3hib(y&Sa}5mI!7;9Uov8PN<~VsyO=J@OWTO
zBrqCb9^Y@+_XqG*`_|Bq5Qj@+)cHM%it|NnN+^Vp!+^)x1HhAp%9IwzlK%o^n=fBj
z6>gRS`oU*NeWNcLNeV@${tpKLL*iHhvw>hcq>yNnHwz}?2MT~mADlfr(;5d3oNjR{
z5(Zs+x^|nFFX11OVKWHeDGvJ_dsdLU7j1w!Y@2bMDhq5xGtfP6?YtKhw5lpqL0F<@
zx9s9&b?J-N2aUPYbt9ngp@HtzvM<q?WmUJ}{+0{#KPh$Aa}&72fAbge8@|46Y}CZt
zT%2k2Mt3X(Jw|tTcV?>%^i;oo{i*;!Cj&p~c4$PFxd9yP2~gD}tpR$CVg_9~=DRq8
za1LV3mNIRN%fZ;V?fGB;<PVS0e}SXppa3H!vAug;4W<Qzggj4I!d<2Q&ZZNjzM+JT
z6sD!!3M@rwf_mC6e|zlvB)u8=P*MVGuzHYN8!{R)zNc`23M}O@_V1qu=k!q^;RG!p
zR9Z4gN9%RgQ@iouUVrpm<_N<fv`8G8CC&Qx!DRh4FGW3F?xhqfzLtxN3cMCPp~{fZ
z38+5qEi>$o+_~Ne2BF|SuXfO8)J1(&6V2Zjzfe-67UFjsc<+ik$_QWuEqn9x(`YtO
zvfvfGlHq`IY&{W(+LDqcHZ~M*GJgy*0HBGjvldSdbUqy0l4~^N407Z-i;PPn0{9X<
z&BgqJ^;vUpTwG15KPqtk5aan@3Re<^&qrrO<-ECukB_n#05G}FfbO>qJ&b<G$R<dk
zGTn&#S_;=H%TObCHP8s5wF^g!MkT_qcqMmeR%j@A;uS#eNO1@J;qW8nI3m^@6$zV)
z;-<%i@miu7Enp37#2C>m05f89y(VN07Dq~02}FfiP>>I^yFKK$9iE=(kqbK_hY`~A
z(5)*T;J>00L84MIrM^n&!wo-lOCYbj7Gu$zUUUC&o!*TvgA>Arw?1{@-S<Emz6x}D
zj74tXsi7l)XaRy$qF5lz5HG0f%sw;Jx^SXi^Qzy)|2~99b3L#^OiDEE?T15@80qfo
zNS^7cKbNytQa_1(dBBWnaAHEI<Y=N!X6?%>@^PB_k0+nSodw*(b_}V*^%hGX^!PA}
z+F#i4Am`d2Dixrw6G9hPVLRfiVyM6x)1SECn=17`Vk4=7Q6wbm4&?p%lv{9Hsw)yK
zaV%CooHO8Sk!>9JHrZ|mu3nyL)4}srp!1;*VzJ4nfwq$%?y!GJ%E)LO%0{(JcnDB2
zc%fb$D}HMh7#!^1Ka=n<pT5_5P^WVDH6X0GDCOhWu*Y#k(NL!k3%syxn&7;yt~`+O
zcMtdfBYMdGuuqFpt#^JyO)NoRps3Ar#N-(c8iwCIk$y+-Q6P|tamWb4$uMye$22bS
z^8kcj<X<Tc!Y8u=z0Cm8M9atCnVZYvv(?OjU?r(QS7!(Hs<Sdk1lGCIBeoYWS}1v6
zmG8)$%WAce=2ZqvJy(oDne+vr?|-|s*S;%lmg)kekT1lC$`8nx%~k*=p(s+=iitAK
z0g7u%m@{)acNu06!Yl@7W%-c%<x7qxMHddBhER{vP<HO#-Ge-NRb0EIWPyj#b}EY?
zlSt?J{5A_2UZ}`G8=%A;t@nDkbWO<{;2eh1HkyCIXl<nJROtot7R=1n0r+X<d`|M*
zTe$V;_VGjE9QKIlmcuD|sS%v6bD+lnVD(AL4qlOtdB;4mw$>8!1|p-#Xi?i5gS?w6
zbD4ysBs##m%A~pi?|qp)*1$`SxH2F<==CuDtex$XpO~5|&UJHOC>A@FGOF14eG}_x
z>J}*}>l$ZvG(j;!k5m&Y;9&G^-u^<gy{+vQ6xHo7>$mW*;Kc9vyHQ*`K&cZq5xSpt
zXQ_$taVjoFIp{Ev;wF20Vi>&_Cb{$X7|XEi;ERXjiH6LPA;a^_O+v+8OGpdQOs)cH
zum-x}!?jxjB32jbHHYyuvni#u3}y^IkBxMuFff;|bP?d^|90`bjB~QzXasZ5I@H;T
zErY?JWtr$~!Ke%)!ZIQ8(E_4tLVHOjS@>M$viA-80$-oOw4x3nYJPABnz?e8@DHp|
zVFXO)gR+33&#J+(OFQ{7SpifD>s>Z*{1QRE1kx|Yz<|qx|JFH!KF<pmBrzAk%|_)}
z#I<&d|7qCCKKys690&CV6viDa&QILm432rERt?)W=LNw}*13dB22`0+pEwN|P<rHu
zwl;rGg}yka#mC172n&IL3urx?A3VnWilA>A<)wM2c9=>)Tn32)UsU~ufa!IRjqZ-c
zE@<lN8s@>{Y5xu7eZH2iB$th*AwHM!#!VbVwvO&kk1My;CiRoWE45C@$`#`7-u*d4
z!zoL9-kkw5&+j`>jW?gS$JoKE%m2pBTey!2Nzw(~KV*SG0MS99l{846B0?_&So*T*
zi$O~ihR1dD7-}nkdq5}8CSJhd3UuSrFx6-VQ^M^%wDvdxNHLn4dPTuZDjM~q<F{il
z)imBvKIj7Dt}3Yr*LdV51RGP^Kx4?ugW;*er3uK+iRu%b;_5RWD7<7j5iTTdEkPf#
zhj9O@1QbjvHQd~Tou=66q);aGJ_w-_pF9Z|V)}exCsF2rMVX0NlC)v=j~m8j9I75T
zW-YMQz$3IvOx;m(b#c&5s9%0XyzI`@s|)DzEcO9-`!n8TJFKGoxWb;clfa-*sqe-6
zx`e-vDJi?u2f(vzV=#|{BDWk{X_pH7Zt!Ao<E%{6V~Kjb0$LZI(AHq4ddL0TBSp%`
z!g~FFx+*vd1{0DjV16R>_@-_GHoo$HGGZ__6(*4)>%yJ?(9g{cbG>U6O6XsQKbg{}
z9^o0U-~p);{gj1*IYxtM;)AyQZEkMfZ5e^rL+1Fvu=B?PRYg#~8h*{l`KpE^4@%Tx
z;d;l^t`VfQw*UPM!s_XFMv`2ZOhHTr)Cw;hkZ4}@MF)q7wAIkFWA+RyCI>*JC=eRX
z$bzGvDb<DFd&%?)rcb1x1;L|^$epn|(fZOs4n~?Y0vTaa3f&6*l?_57i*>n0&F6B!
zt;&h-OYctGky-kHo;M@VWb66L)EyD4SXcyjnu{0;PD)<A2muRxKJFg5Qw}%qc3jml
zC5!MzetR^-th_VHie=}A+Hys>Lb&!hdJLf3`LA7LXdI);g+UA$sYN&-n4QhK-Ba!^
zlC{m8frIT$V6}ouZa5TImYkHT#q|Y6EHM|uSJ;@aXC#jMPo<=yVtb3)$R_aVM;<N2
zaN=oGVndcaFgOJv@Vb;jy#Z<K_;>+**#~b;<^y>0q-e%$>Q6QxN(=6CECvu8&J1;d
zhizEAGQK|%WNIS?;HEhhhLT*_&W=7-rpgdFKu3nQ6LLm8RA7ZF;3H+U^kP7ox|E5B
z;%jebw-rNvm#>M;^^SyI48>{?S`soKhf11}g7;+8RVD-i4>wU#qdh6S6~l4aRk0Bq
zngaUl^ys_~OT`TN>6&YRqBscdImexEndM@zD{V(A=5AM6?BbTfuAHPRR2Cc_2wJPJ
z1rQwlCBSHW0d<TF!0JmpDTPak;xq9R=hNA4tutrD@qIIjmJDUa&YZZ|PfKww?&x**
z@UR8`OipxC^I)<85Ec;Y;dn)&>o1@Enjr%Tj~V;tOJEi&w8L+_(9?GX3xvq1k0vLb
zp|({5Oi4q1{?H-M8r=<<1QTyuFeB$&YOFL-@w_e{luE9w%C)SUka=T{LcoAN41%XT
zRTmjRycmVhES2s-B~kU$9A80LY3Hyu&#r$w5AxkKuU*Tp|EmB_p{>c}`1p#BDIPpg
z84FsFxW1`YR-U;KiaJeRVhU&lmOBf?#b&oOoK8(`h}JgD{jX}>lNyr-zz9$aMkt0O
z3cB}L=ewLAe7tbR>E!*x4EQF2s@tWnL;|UP^Ttw%kiF9fIL-t0z^(uttZv8yBV9lm
zy&!2$8AbYB+J58cGnO}<ztC@>$;Qc3CVZm>6tO(D?=$(-L+PcLGzU*9>?{%D<t<3J
z4SP#p0vK{jg3xSl$-9{h9Qb=Q9k16~>fP!6W2M{-Iw?w_x_Pf{a|Nc}HOIlwCWC+2
zlA>q~)J`Vc&YzPk`bI`K2)UTh3Ki@j!q|DROgny-)@i{%uWg8C;lRb6{Eo=ag2D^&
zm(Ylzkbu+@1rJ!EtA4eZOSVjPS?4l*b-IQt1$`30|LM18i+*7IuEQ-zv<AGaDBtl{
zL`_E&!-Q-|b7J5RYXPOcdtW8bW%)f<q<P-nvlR~mL>EC{;=9<oj~_yPUloo!jUBl!
zq5E?g-!<}#U~d5tk&wUexV(M))~Wa%JV-EYna=~E?|XmC8Cet`m1(u1W7|R38(GS_
z_H*<|^vvSY!-rA17e765KV)3AM>_c{bBK*!g^YA90xlf+goOUybm`Ej9XxPeecNw2
zws-A(Eruh%Z(7mQ9MIv2i1W?9n#gj+zUYbVr*p?X42Lg%u=@1$O5(*L#o?CC8F?Fr
zUVoN&@A_fR<56P_lUX+@h+#+>ZfsXbukx&bGazHVC%QJUJ~;#hS#k1X@Dp53lhb<m
zr3XyXVn(hOcN92u(_Y*OB$fbl=&Yc{#;K~j>)4?qM{40Y90-J!6g%)Tq-`97l^mJ!
z5S}DCZ1K+23=F<e9?>vw$b<|a<*W9JwDo6|&ZhM{cVF^0U4Py%FE(KnJA1H3lCPFH
z>Y2Yh(JDzC8WPn{{q(6oB`Ohv{OQ>+5yy=ppMJ(wJ&E1>lp0gis)Q3GuIJQl?Iz5g
zeM>Wq#AE|fn@Snh|6Fr`)X6~Fqll5cukg<Z+4l4R2HIsB>d}|efEgU2X?dKM_CDq1
z=>~;j;I3I-3$rx+4_9r3iA3Z>%q$Kwo8}rRJwK%)Mt}G0!P;9LcI%>|*JF`kg#ldk
z6@C-3<IX9a>d}BlAB-1R(x?;bUir1`<1V4A?7N(pl@%G%g%QC&S8x3wrX!sJf@Y1e
z2GSaUv+&!Azy|mYfU7I+T|9TaSZLqa>{b(v<meXQ;`;Gb{8q0Vj*}=Y;QFtWPdB`L
z@uHHjBcKOxSy;-F4h4EmaBA?8ezy#F>0hAVmPQ>KDglT)^y}Ap)x6Tu(l&@#uG_3p
zp1N9jP6m5jxtHDTWhVE(d?miOPAwlzDi!{$*hV23CY-$M;mzyWxfJziegz>$q9P1(
zIg>;u<&qJH>arWOcR8<lxrawL8r(eq(hg1@gS)jT^I-;ZM}^%<E732%7-A!I^i&MN
z{eb7Z?ejy0swIw90m+#GI(ndce<@*Z4heoS%%(ofomp}Rg$MoCCXDHMtI}sICHP&N
zX6lUR=n-BquzC5>fj~Pxw<`q!u~J?jjpA}5VgxWmz_v*O(HsMkjyj7hgOU|IjblF+
z_DM=I?ccxujBrD?G!W&3pYzb>WoTSJ{c&)DuxYSTer%}Y4XOlfhp=;i!c7cb8(Jmh
zQXfX;ma?W_iHeT~H05WXuj8C2<uZOlYzdy|$fZ+hK21+bFWK3)uo^ekS$V#=H+(A#
zEGoY}DyTKGJ!<<di^$4yqFK~eQqge>X70RHePL%0A+CeA0p=5HkOd@&5r9YuN2b2M
z7n2s@?R-1^>A6eF=(kZw-v!5VUI0J1sIOmI)+7>6UNxpqblZRI_h&Yha-qS8?GeAv
z+_^4cthq9n*>q_?{qV@z68-9KhXe?N@qr=8oq9RXeYwJ2O`4~A$QjIG5h<w~CyCYv
z9)<S~OAyRIHTB(K$x)7T@rN4Yj&8i3vgS!f(<5ICUW9(fpP5F`Lezzg$FQFW8Iic}
zbkrCfJ<5hUA@bfm7hCUwC3fx-v123fvePx}%zcoTlvF)9mcFhyJ_gL=e~ZJH3mMXS
z^yJn}-pQ(JiEmSky+too>@3}QS~G_#w41M_45u&eJ}*U>y|4R~2)_fMn^i;3Ng$pm
zHu_ueqP7#w?+4ZI=Z~#@zu(0x36OP3+!ZKsYru@9y?d6B;h=z`!A2HFI0mhvE_0x2
zA}g1?eX@?E8fdf{7&W4xEX})?Hk+1_vGHrUTWM!My-RSlFPX)jI(4=9ShB7N>Iihm
z>L}gw{_8izfRj^7WHkc|RwkTq*~G<+g-94{@Tf31<N+f}eHxt#ZY@m2da$U~Q2h|{
zg9qSV^^$n`tIRkJ$@rH%t_Kg~zj^Dv;ppb0xtwc^K0$wezu7huR-%SI`PBV}QL^s!
zgoO2?7nJrrE6lyqU6ILC!u7`C2PqYe(q#B=ne#G8zrDTLoVD8gZ+oGyu;^LdZ1YN5
zLL#q3KnYB9y06H0JIGaJOQU@p{c_@1HzKSyt~pDdh7JRvxx<}?a6O)EYDXn411ZEM
zd69`0uT5NUDn@Mn-HnXwzWV5GA7C-M)!@Rf{ttWqm30Fh$ZW~63tzwxZMWNFaF>2r
z#DK!}WbMj5{*@|q8$mJzNJ>=NphZhWi6|?_H*}(jCo)c){?)MgCJLHtSsBncfdP|B
zhByy^YtvNqfYNihFeGH+tI9k5sO=&>R+1bqU(@q#_BfgH6=EA7-4`qB9iPma6qp%A
zYbD;?`J9@pGwTbl1IW=Z6>7LG@#3WRT?mO4`OxuzXT>(6wW&H1D);KX%iS=$S9frb
zfw2MgiaeSFfIa@;8uLR#Dsp@+7kN4+n=x~9fik;PY6|vlA;3p0!DMh@9(?Tdf4Bfp
zHP3zFxcGeXP@>Mff71?eEEkpY5}=U~vKL72`gIvPGsb3S!#je%3^`Yyk#)6&V+=?y
z|Kq(}SI4@6_QRYvsOBC9WB^s%Zl%0>#Rc7_nCj7xD}Zl_I-ST5fRDyXCvSxs6Q}|t
zH(8KDqc_m~=zfbSIyXK$?^*1#)QLT;)e+i;fvUb#s`)m)LBq3cSKW`@O1=1M{n=Nl
z^xFbIu>+8xO@QBok80xj87XkYoK)%#@a=7P8%%R3?r<4Mj=}Y)^;e+Y@msm~zTGz8
ze4ty5Hr@LalkYdX5Y;k*Dz|dTxfi*J)o|HI`wUzn%duAjGr*qw@`WvM`zf@@GGh+p
z<i!9F6Gb?K{K<{$cZJ`E`!2UNcB-e+eYC{?y=h*>%<PBDpw&i#^F2-@tP}8}@SwUK
zD$eDsc9bNnP*{Yl?vA~4Cjf&rFsBkSGb5N_!Ouy=+2oWU8!0ME?ZEI-y<PsA)tAuI
z@8mBPPixu(z1n@7v1{zB3Qhj*2-V{#C=*?c#?kZWraQ|2dlte<+Oj?YE)3Dy2hw$p
zfs5Lt_T(l71fG>a7R0M4vd)Apm`<^ga4;}tbb+bCz?mTI9m$p&;<OU;HV`a43pg>H
zaYqG3T+rH4kPxZ?h(a9rPevAWCGT|8SIaQ5u%O`K#hjw_FI?@Od}_&gZ474)9O^zF
zew!p#-tnpKUVkJ@kd4t~>)MNTbz_RSQ^BRczp(7Ufq5PoBH)qYOJMqUUH*hrbiw65
zW@24K#zq(wqq-f+a&r=|8Y|J~@>_pNi9S8^bU{Xboug`0mhVp9@zj?SHljaM3h#|<
zmy`_An9`B-3Su^8P7Wv2HxQd~Nr+zpUoG;dHE7zw>Y^)0BsYl#KAQir5uO?t<*dcr
z|HTjBi3C6a7mZls+|f2&K3<vrn*vcPuq=u?GIS&}!7YOVU=0yvK_Em|ghe5njsMYq
z$faM4bA12DPoH`=cm_7EVxU7)g#xWa{vEu6bq6+kaQ(>0JSK8fHN=~u@^~F)9-x1K
zYl=ReOk2<>c0Y}^<9F6QenIbA?k_sCf6b2yb5+myf?$K)J@aeRg}^y_SA_>F<O5}m
zP+F)Ch(TPutiItGkdPLm@d5vx(Z=RZBwV2mVIhygSF)<_GPF~b16Rl#1vD-RwAV`8
zmsh_JAiNFSCTJ1IW|WB76N)Bm%~9ob_Vc;qtI6X)hC&@bK==O@9eX~@{UpO71(rfk
z&%no{Y14jyiIJtSX)^t!|0__mJz&NOV;fXXAej*u8Z%PTM8bpWg&DJBfP~fQ*c+~$
zif7X@k8rdsf{x(&+Nd$M$!D9K953Kmab16&O~)lkHHJlvR|B*88_*7cv4FeQK~Pgd
z1anpps+M14B&qIU2v`SEC_4JrPckq~71(V9Mp6|{e_Ex(<Tc)u2?)x-p{b;ox$^y8
z+#7V+UEp7%nwu^;1{@EwLr-CS?@<$x9e16E4ets?U9QiW)sf~2e{HQUVrM7$$Qy_b
z&L42F+-$EEcMv`bz*K8%>xOz1wx0}+G>mcc`(mkKl29>l<@O%~bQdvi4_x*f{Qa9i
zVe^PRt|Inn&{M(;v*M7VlVCr$K=+|f&R4EfT-fQ>VJ~~1q5WaXzKlKh5C7b7E-=<8
zZIUzJ?+Z?p-=CRPAP6DaL%LZZ4cYkmtzo@&f85ts%UJ!pHs!InFd@wQ+=`jIgiCuJ
z*+0Jr2algPp?0_4rEUkRiP){l&sgsP&|Zz1(N?dSw`M&sr3Dt}gL@DqI>DdGh~<^)
zLvj!-R9)CfJH0<|7HLp)zijHnewefMc_9St7R20hCC8+qd*D1&8g^a=qb9X~a@DH~
zO^(KAWZgJ#m6;WU{&%+@faNE#)AT-|Pow^dG6s1EN(<C5DmAGPVpW>^;7@p)d>$Bo
zc^md^%K!y%Wt>2(0gr>Gh!uw(5+}5HRNlVvsq;l|j2av1Ex>^({kyKAV7A)8Yl%)i
ze3$NvolVtDyjBY{YeQ?j3L6w6Yz8ZR$N|6(OzF;*$g{i*tkpP$_lrjQSnv`?FsUj6
zc0k=(f<_2K2qNJ)B)<w;8KR-8Ck|lWYr@Hs@T}{0pj4bXW};M^(qy(?XAZR%qObZ+
z?HZiP!e)|wD+;j)3;?mXz_(K$mx|WTlqLjHbe6>(N4Ltl{%nztJ1Q4<6ucTBO?1Cf
z^P)gHk&4D0NzP9k8{wy1L?au{`)8$}%8NHkIr(yXsS&QS>2+iWnBoQ_4+lJ>fo4eF
zf&9==fZvM0^i?h7LE%4I$+(URG3WB!43%PhKi<4Q9EewtN)dS_vnuRS<hzJb5`Pa`
z$9+-S+&)|M&_QNO0Pcq1<LA`Z^ZokQUxvft^|$#MxoYm^+MVhkN2_;{3u2>{g(JX}
z8#tG;0<h=b_KB{wWq{op8#?DFsUwBYGKZw>1BWr>pmRGcGPZU)Q6v@}5nSHzC(0Ml
zP;dc6mR4y7Sx}ve0Ziln2C21<sVN(v1>i6^A;5G2H3R-eb}A6%+|fQ;u5Zt5=~{pf
z1rg65XZM6)*&i-j->!Q*>?<01p<>BNj4x74eXDM<$=31(_UYo;&VDPl88KZ+v73N4
zp*VP)mv`6Zge2FKu|a@5w7Yh`krc<f@}BQqH)^&tVMYvhLw;cFR!hgO=mQxB1xjUK
zVF^-32JOYvm!cK;VB3nU*N!HSQfwHV%7Ki_$EDZLA*+0L=Tl$m{bK9gu?yc5z2roh
zOk>~gK6h*+u;pn)uY5CZ1M!}UHT!3;a%ZVl@96$HliT0#zvZSoX-*_18wPRomXzs9
zMWa+AbQ%yqiNz2~SF|e_rXdAahPC(V{x>+WK@`Lc26DA}RCC%->@luK`2k-Ibz)Tm
zJ}8*eVlP!=dgVqGyM|`gu>i9abs2J=yI0x(9Hain*fyTf>(q$}VL4uSHAPG(iL5OF
zGuP|B%8`e1j!jbs6Dbv-K%k_C5u-NStWnzCE0XV_R~TX1lynQVI<W!3{ZI{U))8@g
zfR7}X1Vk-w(pl0iAFzBAH-EUMG;K@(H@#la1O$2txRcD&q~@mL{lU*B#%D;%nkE6K
zGr)un@wD^Yi#ipK?WZ=RjTEXTqWBHaeIY)zGSb!40)#7KBKz^%w-Cb80Dcu*-R%QW
z03be$uTbAY+J$A*(1pS(sv#J|9_%a<K~H>@pyO|F!Q3Mlbva0{gffd;1zgg=!xCft
zkI1&BtWX&_2O?14%T9yh-$(iQQI8M`15WbWBA~k+RIG*pgT|L=qc;t{y>j#5{cVIh
z!wACcBCZ!eG>tUTEciEINFiR4^h8JU&d2v!@%cdk0V-fP+T|zZ<&nOrqpTmCO-fTn
zXwc&KZGozzb*KYvcCQ)zp6l1#@_x+SJv5FZv=$rOU&1L__u`&|gDDU6zWprpd@fK5
zGcYQGd+Nxqg7!*K&VfQ+0;GPFD_h%+9zKk{SPjsU(EjiMbcePunI+`VmZYT-|4NiS
zu;#`UJvQ3`9v27-{Tn=!l9Ip^yqIK%YyCn~2m^DU6w~{3(pv29X5OOG%O|-QhyxOH
zD-c}k$gaTYO;G{!5n->#OJKV0@*`4{R#AJoGSURJ0C$uTBG`v9LM`#%xLOX=XUqpE
z6!b`}c>Q-iYP3A3Gs|vvGS)vX5OTKna$!mAy;D-r${TF;T@^TEHz5Mm5-zdCa*~WC
zh@J7k6~D8Fp!fJ$@V>jZosQh*wU&T7Xo%4++1$Cgk@M!bT&28e$e0w=jwja+p&&RT
zdtZM;_DFu%_3NWW;7EcvZmGVgw24dGobPg_v5Cpy6x&T=L6AX2U5^1)Q!RtOi_+A}
zD-PmLh1ipu^u!aOao>_4p{StRrCiy)qBYZkj+upJ^k=O7r-M``|9)2(*Cc{OsGn}K
zA3Y>2bTD>D!w8&LE==r*8iJCDR~A6_B5NO2K2?p`lB@7CZFu{(8uOB1P)iBF8RJw;
zmdJ<`B{9@i80Yp)IKgJ98eCE+_OBxc<6;xqu$wyr`sT_4Y+6{2%%)ad{X$bMV|shi
zT6Z@>A9fmX2J{-l9U}%6utrotXO3`yKsdKX2HBuftidRHJIE}l23fSkdJ&^2(tqF?
zniM$DfbArn@$HOkB(wGYS!Tn0)#MDIlK27r2xN6v(*Q}vyJtaMJ{XWOvRz*O0T=&q
zNiJ>uAd?XZ2?@3dHC;(MVpjmf2J^yzwjl;k@?M|8twu+Ii{qY5x+ga`A%rXdygM=~
zf@T0BeTPC5VtJ)=@?-!sHz7Ynb%Qoy5`i{-KMF?T;fqn3srS>zkE;gZC2|v4gXCJM
z04M&~#u=^~kN5`0^l)i(0TLiMoATncl?N=7Os7zDP?488FzHm4JrvJ>Y}J)=t~WYZ
zaz!f*_LsE|IfJ~F1Y0NwiJv_e`W<AU@~nMuQ*iUxV0v_J)q>~ge;H7X=CszGiAs(K
z^MIJO6Xp!G?98C?Y&gOM-9B;5QhoNBxYI(^RR^g&W5QTT79=9@H3#QdJd!Ow8#p}B
z9rSz=L3a3VKqkjsEe!7}KEqLt1OJz^1tbZpTLrke52WCXWcJnA$r|y0k&6o#FG4B9
z%p+|;eEUp33#61!9vavOrI!yxlUQ?mf4%$)$pd3pFMNe>Qo<u-n_;n6&A85ES$>UK
zD-IP>I6%<=%|R$ZjK{C;eTQiVa&@Z6G9=2xb2+C!FXC<{SOEbwAgM$5Q3X%N|BtHo
zfakjZ_y18^8Y)_sXeTPAU8F&KDbg^bU6d4(5z(e4O-e%%N(os>rG?amA}flLoy_>(
zUv-`T@0{<g>)gKQoJ&5R_xm-T&&TsI_H8g_h|ozwjEV3%7zhht2O+YMWe~KOLIY)d
z+u~B}BhIzYhd16IGxJQx*PlasISknM<6ESvp03mY?+b2+4M*4Y-I}i|b#?2r?amiG
zber8M%Y{esjfFnK^jM(h46P(x=6xdA8Q0Z8XwqQpqYb#;@bKBXf%0)&RidChTw1v=
znL?^M3nmV*)My_W1gm)SNI(?&JhaYebZ<5yM)<MM--L~1-XiwXP?yaaYVPWLOzTg^
zl#iUjyt6<kV*48Uu|I(T<d@&J`S2q`m5rDg)6HZ|=h-S9E-%^W7x>;|oG^zLU^`rd
zFvGp%kpt_BQJ+X}wrT2XX1`+Bu3Zvy6jmMR8YVHXr|UgPfCEoYa<*n1i!qlZFKdAJ
z&?l#2L?yOVv-aN;YJ~<t=|y9^w3{7zTyxGp!0LagWjm%`W={yfSvVpX#_j97B%o)n
zqwbfsO;zY>T|NBdycb1A!%oknf(=R^M3cmi@r+nkAXk`@4e}o1i>3k}(DDJvGd+5o
zT@Q3DVY70?KgXXBRGNK7SH+uKhBejk=Q0n4g6lmc77cqp!Q^RVzbBpgz50+~vTNm)
z`>Pd_c4s^S&G0Hw_S`?<4TF+wTB3G6eO{-VU%TsDy-`!EE@{{Mi}~{XJQ_~j^$imq
z9JlpOHM#GSG|Njof6p)>JI~w)UL*L68(StYdx|B+3RuQiyet*{Y?&uQAi0VL9T<wU
zV7Pvfk!lhzjxLxAKu=|8#R~Yk{F!+1_?v&48ZQ7WY1empH)+BId8!<9%apEfYgU?W
zOEq$@nRun9bwQ-vxO4Cw>6~ZgSy?Sh$X?1xEbN@YFo+%RLYS5-7K+B9Nz?;7zr2zc
zWI;R17msI7jx(jLDp=2V{<9~gw%~ItzpV}K%gq+*RIAG3kIGKfO_=tu+nL!z?IiP=
z3kXN*9Zcb@m#yp54ml_5RSfJX$$n4o3S2#Hcqi7S00-D}OKC7qjGTyij#9NK*l;v1
zJ~;7WIx%?phcDLP&tW!887>bxNfe_wO=!`e_rQ60rgaSc@*V_~jVb(HnOA98#7~ot
zF?Tq%NLuZg-?D^KtwvQl_gikhmPR_k8@+?$rNh*ouG<-LsLj%juxvUVpW{a9#t`bZ
z-uI!qlJK#RkNK^&hePlW<7v;2l09*~QCy3SNZLSboD(qoXro65;6$-cgt5(NpQCzr
z9E@e}{Sma$u@nJ}o}axYcqJnM>ootZzE(>+R5)8lT{@J9_G*yJ;K73(up)K$jqogs
zQk9R9d|I&M=egD|90N+OX%eK=bSEy@SqP{F>J(c(?qYHzVy^=0s7jP?CpQ@G{}ifr
zE5GbAW#2lSpeNgu$Zob9@6>E33OV?B@l81u`$q4&>T0`9N%`RNN~y>Avfjl<Pl&RR
zh;qJlPIOtn8ahw`q4bGMvA-X<`jn1zXdmvE1XclFSt2uUkPI|*P!KNc;c>Hk)?a<u
zE^$o0y|0C;N!d<UH@B03b7X_g<V3`ZinhDRK@hgd0K+_Q_dH(*1{1#+Oc+~vWXJ5j
zK3lpXf6(~OJ)d?@+9{v`NV=0&qiZh@Pxfogjx+7^eBs+Y$ETId|I|szVeLyf1%*W`
z=X46Pdfrd3vTO5)FE&$m3S)^6_uRBcbP&Q5+|bUA+t?MiBfnp;&-l~}RQA1o_3DL+
zW<>VlmHRuk`gJ|%Ry0P{dqG=iX%<psj^!XHqJe?ZSjO~d?ro%=H07lmH&NOce>SMs
zP*=BJwz60IPJ45wJ}eHowz{%wbZ5bFM*J=u^Mv+b(XbVM&MA3F4lq@dlz;@6UgiYw
zN||X6P%K^tIAY(ozn8a5p=HoDFlZm}zT?^Tz|9OEBqi{nyV`fo8M6ecXR7on>;|sD
z?HS*e1bY~xj~R4(2!|g5LE-h(M)Q;I;T+2(_@Uzci#2B%aZ$mGjuR~KFGl#{@J6Un
zhu2qzUs|&1u_*cv#L;(yVDg8Z#vcg7oV^KDAwmQx1}k8#o5l-=)N@mJt0+3vo2i^q
zmeW$dR2+OJ?chem;DAH5-gC1j99=otlg@h{ANBVh-HiLI&#8?qjSAkHh>CMgarPNA
z8Mp|%1)_J{xh_p9&QOg=y|qs$AqYYh0<wo-4ZySh7T#m!pp&NTV1f9%q%^EYQ1qX1
zEVwSONQe<;1e}p?$*bL~3E!IU*ga0g`_)#@mpc`x08cJ?(h5Q_8&GM4M#%yw%WKxI
z)z#Dcw=6DA#d*Y;lW*?!zyYNr$yGA6D&0%2)?W_~_cqKa9Z|dk9ngoUUBjWF<!<a=
z8p`XLBk&dmaYjaq&n{@+zWr$rK?UmiQ+9;h6CP9nVO2cwnOB$nc|Y6u!b0!LSzU$X
zTh!y`>s}TtgSA^Fq=@MF+oR%{SNmX@yQ*G0yc@&LECEY_We!2CJM9(R*x~oOxE|X*
zSwW0(APD1Yj<i&ibFQw!R|7T;Ym^J7kU1)iiRiGd=&1<($(8%MMQ9g9MhR{QZHp?j
zuO8dr(G=sY-ps{V^yKp0pO3b-+$;|J5ppVH-M!8Y&m;$azcPGE_URb;z{FeEZyeh|
z?Tv}NC*q?5+Q1Lke}Sk<NYEMlu~$X;Bzz6w2@lb~7E43{Tu%N7gNG;IYy=jjUEiyB
zZ^YiXBvg;%YDHPWXb7S5&th4XebO&+>(L(6p|8hGXg2tXU?C2Zb^g_Pw8QiXkBs{C
zC<$toZ@Yv4H^s-M0!jDJb3Kj=;{u`~2AwV<*GmW|gyQL?{ucp42(twzjKl$F3k2!}
zNa3;V*pslPyLay@4kdPTyu=j>XVWGiin6h#3vMiLSX^nhB>%wr&x1MjJWm{cFu=~-
z+#Cu306bOt$HLT31`FnXuW%Y=XFL=Nnt8R=fTEjSYb&P1@S(UFVnq+dw%YvT3nwmu
znC?P6Ld3%{GR$TwxfQp@%T5`S#UV=3CGse$n5-27au<ER#7HB%%mW4{4d|oA`TSnN
z=A$+rIC!x6V{Z`?!U>Uxdito)H#3`s>jZNT4@A%q1XlaG`u>8AWFZgLz%%$9we5t~
zF8c}^;<ZlYHHAtStQx!0*>?ZEpSfB+z6PI3cy@p9$`heR>w^Ps)rN1sW!|+d_OeR{
zLXed7ER5+lad&L(?!!~u)z6ysX_ska+n-c>VRg;fb@lFD*%=$#&VZ;V`XSUSrWn9k
z*xQ-lL61+(JlGucuKCtc%T}+Rug?2bp~?!|k*bl0Hz@!fNV0Sv&qB~=P|Nu{CEnWW
z4risWxVdO15-Kb|sG33K_yf{<xVdaZZo!sngB#23=bNOi#a%tVWwczUXBW**Cpe&N
z6_)lAJK#kNUqg!59)}MYPUzvWMpt(~Ace#X`W`fV2Q7|qlszZkSfqc%5VblzKFLD8
zfLo`X#J-k>V#ldzeSiF>y?&Uq<<zwD$ZKn#t9@xt=PCSb1VBJN2tf;eCkAeVlGCAA
z-W1qh)l01Ek83p-bYn&bl5~*5+TUyaYSub6zOXl-<=y9{+D<u_Es9v94-OT=fd&|#
zZ)oQ`JSjlvk8j-Azkc<*FRQLD8-Sb#d{$y!u{kjJANu5c?@$DhjG%A-D0}@h57QH2
zYJuq!Pa{A4=}MHzL~fKmZRYDdXP7eQ#=d}M9v4oIvFhKYvrap5%Qx;MqD{@Lv_bcS
zW%!yly@C_M?ijSW8<HMov};RrZgJER?JMz%Gv@p=-pyx2W#crhIBao*O)uQe%sq69
zsHMfldEGfV9HlbH-=QBcp{qak6flv@l3Cjha4*W&X##=>V}Cjhq0WbQ4B*_(ZT~!z
zR4;>|=GcA<ZiHR@DQ}sOkl;4);Eld9e}CV35$#-5L&#5Ipwu{QZ1C8ssJid>yEBHC
zV86z%dp5gQ4%;36N;P$7bWZv7j>gVqKJT2x5y-3|?2dHFrRFi6+juW<=a`bVJ0Q0p
zRUAr$UNECgfL-{~+Ue>Pp;lp!iiV4lK=651-QUn<^9h}*=ih9ErD%O-e2Y6w8!<TK
z##|ZXeW?SiJ)G?r17D>NMO47MDJUqQ7Z=3{do^BZ8bW>~WY_yW^{MoJ{YBWjiPIj!
zc0Nf$I%Z}A@oOT00DMhkZ&1%sSks?(pv4lJpP60*8I`JweHpitPVk*fqDhCb?PrZM
zp80R%sfSM%e;B5h_U`B|xmPa+QLfYQ$Z=zcpr=b3_NV09|G<((fE%37_b>tuTVs1=
zOlP;ho<Kse%%MZ(0^Q6kkO$xvsEu%n=Pe1{Q80Ja`2l)0*=2i>kA@D~4<;qNc-Sh#
zvVmF-RW4q#<T)Mh%X=q^?FTt)J`1h1Us@eOk-$i5$k!(bK)w#$r>W`!$+pB{h`X!V
z5W8;ur)p2f&eA`%qPu!m$^4A9R)0(&z+Q!Yq|hbgB`L>rG%oApx3O-QN%Ag_b@9Ar
z&CymI9c5#+9mmLaZR_j3eehw7Ui~@k_C#N{iBMv8%XWM@H8v}c{dIR_X=Z%Y4xJ%A
zPG*L^{!+a+*;Af*;AZPy%9Pw~BC?N<TKmuxz2T4*YgD{X%^5jMzk(Ufeqqqcz<DNU
z4y9=O=1AWu4qXqLYfqB3hH!?aZF<v>aqx9c&bOvqpm^9kZ?MJrbmaf2x_HlwO)RWS
zgEk+RElkVadL1y5xs>?1J|4m=mEY|qJRA1;ivPyGGpE!Rea(yX`U_?a&&K=@7XWjX
zOkq$b$#Me!fP()-fPDA;w+`So37r1PvhLg+*LACV0my4B)`iFyM7Dn%ntVZ`bI^^?
zgUxameEVmXyvfoX3dKW>G{0=k-0tg|u`Am&HjnY~$+=LTg?k@hni<r9$@}WZOfi_s
zFqn{~LkQc}HB^6b%NR##i(}jXthXrV2-Oh_niX=pQhwc_DV{YdLuhTUpQDgDG3zAq
zjwgotlNFCVTZi{n_iq>faQNo-OX}Fcy@T(Qh^trk@5?>ilR*U-41R_>5lv6m-AjET
zoFkBBPj~3sY5nza+XnSx7`<AN+{c#(R;hURSpC&H>V?*DI9D~|gkcLhDVd`O@U{Eg
zkzH5FC0W;?32oP|<my4;ieFtV^(fD`GShmJ+3ntM?N4Q9WqtQas_M<gB=PDQl`{ik
zDPoEARCbCq2MGd!51c!9E~tF=xi>QH-`s0!^!;7b$Xv^aZbxre@^BI<z~*u^GaWFD
zvX-Xu<*(w#ol)Er!bE1ywV9~BvD)`k`du<N*7rvv@jj;y`z@QG@ecmAn6Ssyf0h|E
z4;?vVJRy-dm^kzEXz?}GRTn9uxu>}S;GXwo8@{Lz0{LI5z`GYb(YXsXj9<%Y>wehq
zFN@F5++uYvF)9Wp!ZrCvjRUYE;INu0@R;@k1~PD_PxL$}w9`3{cDoIFv+C?neI`FS
z#it&prK~dkI#KSmp)^hr%?~SD#F27s|J6gAgkvqwH%2>nz>f9mAMZZ8JLVi9b-%P3
zgZp`W{gCdBL6I1|BR2y-n~a~o>dHGa7A!cDmjok7xF3sSfw!ueqPbsbr7RBX42wT+
z-5Ycvebabj{kT~VW-slq*gtw@bEn052X>un;ayRy#r&D6;9xAPsEB*|)M2S>ld3;?
zgUCsOIZq$nn`v{A%5exCQZH+giy#)wR-ZG|81JD|r<+4ydP~gUhacB1{H7#$P#oU?
zQTC2}5J?GmT6?{t;Ww?rhnlOIgMKDseE5Wv4dRIFb#(s2D^y0*R%x3Av0FeupR8Yx
zS`}kgb#C>nPp3Zw<1vJ-RA$jy+_2#s0RRDpKxKAC+KAg8?+v?+_RL)HCB;4-`)!5{
zLPYe%`s9)6W?IZrk-u;#%v-eRsi?ZsHVKOkDkWNi&m71_uIFEWc|uJoLVUOq#c`-Z
zDt6{r2o)WyVrzR5;`GNCpXtrW*Zz>!D6A1l4-(dQIH1ccC5A|UT&A+z5vw^*V!Yo?
z;)(EU=KOqx98=vIL^zRMA_RfLFYH6Kty=dt(4#~xvLql8|HLsk%*X3=6Rj`kfG{<1
zS&}wNPZZ^U3Q$MgKduD=`S%04N4&x`$Bn5_f@#UH1#8!=*@5ZuC8%gKitjBu5#;*w
z_~F9?<FCmcy@i7P?AeU%Z@3J!f9+lQ`+IGpqD*HF&D#@u$*%P6q|=GfZH|gF2?XC1
z&Z~d!Gf{XkyX$FYZJXGm@$&ie^M}2nI9PN>Gf%@KuxL7hsiJo<O9qV|9g<zwAvu}8
zLlH(yV-iBMWz5`%OiS9S{XOO4B(=xat}Sa^3Z5ha?gZ6Z!ewXSa}0E)kPTy>z}}=F
zLxXb=@*d3+{KGUNy@WW92hIu@8})W<79?Y_gGplcgar{+eFX%&om)cC1%`<;k*{*#
zb2MT&iXa{`WD*CuP>+Xcx2_o~Xd?j0EW;hIM@8i0fiHl4A=wk7@ealR1wvqy@ErDW
z$IDBK!~KMlXk%$;61HQzm=f$C8P$YD`8c~dp9m5m(SbZWcbxr*Gfraf3T9c~xB~|c
z<cqYbcQ+Oa0-SJ-y?)^lRXCuyuwmdTqdNcXJN6$11-cCdNYxT08kU5nAJ(1<ztutK
z*c1gZ#XB8}#}o$VFLtJG8t;Ge=JT!kN~txB7qodmuu0lq*`;%A;~ilS@o}!ffo4+^
z9jjhPZ-{zLQ~&lO!_jO=wFcDzcdRSV#KBb>(rMj;kn$_Xw$Y4TCT&qfBuI$nluA`1
zf38+zcU<QruiQ7Uiml6)kua9bgKyOT)L8Y-Zi0*>jOB%iJ$@-qN}Uj%6GkK{!IY@4
zc;!ndm_&I;H70R6_3)bDcCW)abn4VEEVGZh`UhdqNfw3zr+Zrn0T4(C3xbj=vh>#7
zJEvFO>7FEqI2)5yAg*T-tp?Q**wJ2Z{Fur+AF%Zk#<hQE=oA*5voXW&GC4w4etkK7
zUpWX=#!YfzyW6+9mNu{7YT+n~;qlBpdnO96(dhC=GT50u1z<U3hy*iH$me+ag09H3
zu`uCx*W<}lbuq8GT`Q?}^DQg6dQUKYf+<z#+*;>M*z_8e%gSe*#{AE_Zh;93bxJFN
z2rhFGFIJBM<h&c(Ks-b{%f9F;%u>-DdTfismQl=~mKSUV#Nzt+4_&-|{pnWki#e8c
zuaefU?nRKokmqj*<NV`7nNYeDh8P-t7uxj`+d1ou-%Xl6JuY)yTMSX53iwTV&RwEf
zeHFtu;2&xmiyJlul&?UzIJoVg)3dBY(&yJDkK>&-J#^~IPR~p_U+kRA+Y><p5~4>G
z+_D)Pe{G{<cS*T(Nk}hvvcl}@<(Ebh8xe4SAJ}&z>;-`+<2Ff)X-38^Ox{gfwmkU{
zOXO+BAP>uN+9NEV*c|lYRj};3WH~j-B<Gi<z^4*HADqh2eD97tiUTy8#T4aWQQP22
zo7$-*=l37oH^%68VBiyEL@(sGk#0j@Es7$#3U)uM0}d9S4S(*Dy@U>k6N}>mHvALn
z-mwU(LN|1N{NhDEqCsIaE`-v8`v!G@S^JRT@3{-MgrCw_>|rKLdJuch_@B4#9v*{e
zeYhF^5^Y#EoEC%ZJ!cIb4aBGz#2^{ufscS!;cVa}%K5$9p4?2|vXgf%4yehvGlPMB
zt#u;SX(v#M2qMfG_Jq{Lj)FA^N(}7<+Ljn1fkFzUB#(03fc;Cy*(L4q{J4s*KB6Y$
zWXdv&N~z1IoX%b9(FKM$&E31sxsNR>e|aoBL@r;7W$_s@UPOi=NZjt^<bTq)OtAA+
zs*Fkgi`^)X9zBN!2K3q{y++H!Yz%-glyf1ar6VF{4sVLsiORkEMcU7qMxiHp{GZC{
zXkoYOaHi1U9(I<s*kNv7=#sKoPsKac!;0S~ws9}%apv~rQ!k9OJ9s>we|_liL+C$+
z*?Uw4blazZfLr~vhFf&(`lb3Eev{^;woR{Ch$BC(?~=3M?W;w5#o?(rjPxHO5!b){
zm2Gk>zt*GJUf4<qmOf|x{K6pvyB3CL8jIn6;{!EpZ}a<yxb`m5*q`%YbI;BGP@P3_
zCoJrc+6f*mSUrEf_3LW$DxZ0>`bmm#u!#gNb6^?i!`kY1q^+3Si+h`H;@YD}S*lJm
z59tdayU5e54;&HUCZ?eglihM4&=ogub2xvPwNXoANTqn@wDwggAt@Cy6H3PD!wiCV
z(jo{Sec_NJ-urG>MF>QUx2~_1PfWWuw~CVlKJdw#4lreCKP}FiIt!$J=_!n|`_Jb^
zcb1Q7`NQky3;Mbm%qWSj<T;+uzbVBU6jqcxFOB?LE?ebejA)@b^@lu(prajj`dgq@
zPhtX4IFP+X<H(5%HFr;wqa%FH!#@^duSPHE*V@Zb(`q+hv!&&)k3WeHGn-NI&n&kM
zsjp0~cD%$FQ*W4dwm1aPy}Pf~9p4;zlJ%gMc#$9csZ`>pz&=Hr@x9(D7^G4#WFc3&
zbN+JO^crn=2($dU$(t({S9Z0D|LqR?FDU4+Lh_-n4{@GEt3(PLlh{&&^zpcKS9aZ)
zAY=0?MU+Sk+by!8WXdvh5l{^H%E0s+O!@>Ew7J*OXQ=M00&~4zT8Oc7jD3v=dV^!$
z<(~wi=v+Vo|FhpMvfmkHD8{>!75+{RudjprLnAVnzI)_rs1w3?7}IN5{xqw#tCLk1
z$u;jhMHa$N3|0lagRQD@-Pq1<;1Fi?-hvd(s8hS#e047XylI#eex9&;<cJXoa6o2g
zM=*2#>*`#ER#Rb8j?Ier=X0eiKKKst(oc{%cE`SS9?t@4hT5O+Y?Yu&vA>go3=L$C
zs+45OzW!>2n9B4TSz<f#v8haM2onZywo?Ub0k}&wgmn(mZ<ufl0d+XV|0;$uN=whI
zU(H5#Sl=*5Tf1KFrrXZk?|Y*EvTLQg8$v>tH%T*Kl5;ZFwfXpwBN)52J(=(JY!!CV
z3DXwwkeE7%k-3G%^-PNk?`+w7UxjJ6NLDBS_2m@hIM>$G<iP*^O7kXCo!o-acKerf
ztOKSJZao&+O!yBnAL%CiEQkB;bLE3`ru~Jp=iOM&ZTN2IL{f-KuRPnSz-JMAoplPK
z?HDokFMI~zgT~f16Z0$}UnZ1rp#|a}h4N$lYC7p9C5&v~T;Mw(Kx~ZrX9T<_J~Vu3
zx<rdyj4xD}KMBE<Z?VHXN4{Ry(%gXIe-f+qN9}a{SG1eCyX*GGfmQCzbF34xcc!Sc
zi6A~Ks|2B3LK!>V>o77T&H0M2T_Yxy#MEY8b<K2MmatH~x52Y3JMmWN2-b<Hs}Uor
z9xY`i$?8~z7l@vUzTdB6tAuQM$ju)<%(JlKyv0Buv~e7F5$f6*ED%spF|**77iwdD
z{o2O3{tUWQ_N`d4f_WOZcX?gS%D?RB^q-4mAJa0o*{cC?S6NwUnVgG5Rrg-KGAh<R
z{-ZSVQ4hT0TGt++RRj$IDia~mL-!<3`GS*{8xv*n+8Bw1tL=~u##eTI7QrF*qf|Yy
z^8}=j1YE&)qOA$3Mx5{z6wF)*ipD=>(XiRemdP?FBPyI$mqE<tM!1eZy>z`)$}g&K
z6AY#p{n@y*VmYilkp{!vK16Ndt$*da?t)6H;>{*w4RE`ONQnan1`zeC7ur?(NoVhT
zc{D5nJHJO(OX%rSd{z<cCMXPD6(-Mk2pc-zC%*ZHPvmxr2*_Px077-h#G2}mL-{j8
zsb77U7Rdb1gz<>h|9(vi|GXv$JlLW2<|K~Ebfhxi&`P8nD>NRr<=~piM(iVG3k2P1
zUD4h_OjK9~s4igSp;Vh`Z$D+T%cqOZ)D#SE#Jj_42>6-2=Bd|lvZ!GgWPCo}H$A@1
z$>V^Nyx{5MM0@1ec<g__7(tkW4qS**2=ERz1E)SkrD$7(^v$WvODMSJ9?6#%(gC67
zUfxB;=XDjOmFO?#QZKd;IV&av7EcC_9|*4lEQfn{7O!XXIf>>0c8(AkWPA3U9WtGA
zj`OQ6gREAis|)&+99zL;h{EW<{)yh1U|bQ#!T8Ugx|nV+XC08{LD)>IMnXxJd>tlE
z%gGANvDp$c87lS~wdMd7^47<FH#~kTV=PJPf}noo?W;MSSrN)~h$co{0Rm_VIAnxv
z4_ClEl_1Tb;%dbvOBgaT1R?jTwfw!c*8jXP!H!!dH5)#fh_h9gw&%@OT0-G>AY4>|
zo+OTkw9?5IaK8$@sgUAeo?%$~`6BEX{?f5nx`^asw3j4fiRy-?jd|AI;|pXPErH?1
z>CS;^Vf;@A!RW$k3_}6IN`vyc&vHzkb`MhV#1M8doUEdN_Iv+9xs@?zXjJ#lGf*CO
zo+Vw;Ee3~oJ2=8naj7uCdniBKon2qW%i-r9nKWG!$KlCg5Fa=y?I`eKhIN6#@vJN&
zuPfz1L-l$gawm{ZP?D&;q4hFR5wNWWR!%zqP0E%jBVK7g)c<NwK1MDB7pFZ^d}8O@
zK5<M@>*`zM&%U&PhCblqN-%f;_xg{iP98sg{Mf2!ex+^b4B$<XuRslWay!dhNUZ2j
zV$`z<hTQPUx%IcU|0MK9^NlI`cxWRHBV{>eOKYcG7otn((=AVw^e0FWA8|*cs<^Q8
zM)v%+iP<h;gXT^prti|ei$pc1O*6gK5!DW80^yi~BXd}L*61HHj>G4uV-zUrJ<&0T
zMvYAz!^~=!aE^(LoE++R^7WT|d#6<fm9{oECng_c&PgY0QtPY_GX{p(f74(=Ma5nZ
zDo8%vF(>PO<4~66po&(@i0JI;vuDqqUvV6JE~g1QYKE^23SXNMx?}5>E#EU2$l;pm
zVFpS#Kf@~XqSi{|uI*QytyNi6o%G4utPexwE{#qOj-Rgf5`5kw_eop?e%b9^i{q3H
z5M(X^%CAI3jO+!T3SF77>jSe9QYNR0(|<Onf4=lopJpr6=={fGEzE0zCR*CCjU}dD
zJS%L}1@Fz^^zk`^&&$ayes!SJCqxjiD`t}a2-S7lL=rPDFH&xs@S;ZJupWDo9aeMz
z{OR;$>4CE&R9*iixmI+f$b^6!Z@2q8xhs`4yk`q_ANCHh1e@NTO=-KPQcA0+8v_;R
zx>I*xxAQLuLe$5J;6gBTJQry|Qx%q4qOXDS$u}A|t|z4J0lbZ!=}9-lQcW9a_bq0^
zT|mRunS=Xn(r$BnQ=jwQyBJ&ppBJ}xf?NCO<ntnN0a70(X%dNp{jbJ@`5Qj6=?GA9
zeO6{>)P7NefRewg!sbOd!T&c&hc@Zz>N<VGj-yA94z`cuVbXu~t2Q3|M~ez3gV?e(
zgGvo!TubNS%nrxA(O0X4+}EQv|9pcOT86LDyyMi`0-o89D6Cn=+P>Y&q8tYt3!ncn
ze1^2uaI8h1GW`=n`=b9W0)bi%{$m+=HLppm=-iVkcKcIQKlnNR>I)vU<RNx^f>?QO
z!q#3>4lyAS2pIUN7$^!a^S5s$fCYwibG4BVIYBZX#<2|iPGw@IaCi}KO8axCL79kE
zRKNT7>@aXISi3l3_l=mDD|NZ`D6c`F?W<Rx&#XYdQz!J8Zc|6}?doc9h3Sg$x#13&
z`EYai8jJPo#W`MN*O%Fm$?xw5L#_pL3fJX1YjOKxVZ$v8{!@jI${<zMproWfN6UZj
zOOTnwMK0dpa64)IX#PiCn&YX@c`m>sj;4v9+gceg8KXFy?`^et^EZ93WNsa%xx$@O
z<Q4e(cEu@K2&TQgPmauE#Ta{agega~nYKu+t&K{^Syo5^iE)ko@9eCsPO-*boU4+_
z8xMG2^qn+QYU}t5W}}?xPP9w3r&K~HYuIZKq*6tvCj;a@@-@JE;iJ+J+N1>N@HrJr
z&Pa%w*K=Ef>W-FX`)1NyW`P)doCxUAOp#fZaW?^7bY5%ABmoqGXkdukk4YC*0x!UL
zamu4X9ZWsUUL^NS{klowkaEVBaIsx>?6{lRdK<5P;YNqc9fv!~0jZ_e^+|OsIH)vC
z{-xW?9X(=Z6z7$btRP~n7c5zWNFB$-Lp-xn+iC1mHEKGqerSvLsu;KPE)T#x!ww4C
zzF;%|s<#!dfn&0%+vSK~2*ZE37hi_bd9QS+3rUJX&T3ncYgxR3wSbNOP$85LKUUg}
z($YuGWZ08BPPFNMYX`2wZbneFh7Ogc)}R66{C9`tmYm(TD!X?C=gG_iB54s|56Js>
zRgm$xC!f+Wxe5<Y5Bj*4>*KIXpBGiY8+FRg2=NSbmki*Vl8xHepniS^mte!E>9QJi
zI#E;gDy0O2bvBWARGIMGfYlh}!bDjIBR8__VRpL*CVx8UE_BeBWaF?^1!f+e&FNzq
z|9cNf?D5&svS#NEs9;42!-ZBL2+V*Syf}fdsfRG}){m3&^*6ZHEU#|k=IY8B{PJa=
zw<CI;K7ASoCjpQVsv(XI!r1cA1kaBXqNtGQ@+V-VzbEeRZ^<U13>PzEU44C>hl|gv
zOLNB$lp+gN@cQR4v60AU4Luu|sv)u$C?=W5qbO)g*YNYa)*Mll!_s<Zzt!}A6@b@F
zD#;h#AjYY}UFPI!NM67O;zWbQp*wi5G~GprVU+8CNE_WVRMHN{1r7(~4^hnttnHns
zAJ`XvLM`0&>d#`q6cZDZDHDsFiM$iNMzRCCUEFjq?v0VX@4=TAgLJ#O{*?c$)Nat6
z)n^G9(4JHUb=jTsMN<A_ZEx<`Yn?cmIEi|<)@Zf1IsH<5)<Zn65gk=V=lu>@>%Gq8
z>Yx2Ri1W*>k%L%h8bXW5k(IhN<IYUFhmSW`2ECNug3MaQcJZ*DY+k^0H{P*uKe%_V
zJ4hcFhd8uKtQFd$8#=5lg*7PW=&%|6j(aR>i?rSgAcJ}3{tijt1I;w6g8v@tbWlK^
zXj&h)zIR+7*~wT!IG~VOE$A($x8n)~1V6k&p9(}+)roj9iFyG5Zhk)u;u32NLos2o
zg>`7zL+3w>E7{ej$11u>C9z!=(hGNwk>Y?56K)Ydil>xBQMeL^A#_f2W<RqBeU#dt
zT_!G*+bOXWeRN6=Z>o@{VWFU1zX2%^bDbFheL_M)hCPX3uFr@d))TQ%2Tm`crP=f#
z(e#Z`@`hazZ%k%NIE1t<Q~e`8d$Y0J3ctlE8%$LBEOX?|jVhU3h|o;<{}3K1f9Jp@
zqQHlSt0y=}ZE|t*<IEbCpi1diNjD<RF#T7SQ_k*8v-CjyK9ejjEltSoP3OfViydJa
zk&Qa&&<}40FI~Hq_j;B;pC(;+oWUi1TUw;SL_??xSpJVBTQYu|PnI#hvLXY1_dg)W
zt#ezJfHO1w!eDSlafP6HabMs;9aHuEAMqZDYl6h&@9$sv{2wd9c{Oelo%o;n&Gr8Q
z3!a()3p*z7fF**$fA{HAqT6`5tuLV9giW)yM-IFcQLW*Hvm{vS>F3*HryRq20+Ezx
zrZqJ$zu7NI*npkRG;W95`8G%<Y>)1#4BuiormF4}?!S9Z$Gb`>zAy**#}?A>UN<1D
z-$$OHu~iT3Q|ej=b!S!do>3?7vU4cD89_b6kATylU5wSAHE|pd7XY|=JP$><PuOv@
zK0vL)6Sc@WK+Dzk3fA0zGCKGCIG1UTosdvvzz*y8?oAM&gEVRC0x7mu57krYE+?lw
zMD2B$5GV3jr!qDWQOJU%$!&_Z7OX&2#>|sWI*Jv@VIk){^N+8{pC^|TLTP7^H|6pY
z4!8CuM1oC9!&*deFJL2OR)p=m_3c}XH)0lME`Ms*KAk;+%2j!lgwE>KlfH4X!{^pE
z|HB24Kg4Mf74<Rba5hW2Nv=#*Noi@|KF-7L{rhX)O6$7+Qgl|?9<>5hi`Gun5%Edp
z1$JM9swWK^<ey#2%|6hSOrPuHTOhB5=Q(-MfBa9FDt8GBnJyK61rX3<>gCXH$N$!+
z_xTdtUOdr$OF>ztkYpx2BO`YV#9z}Sq&|DwMBy(d`j7eZ4`(`33;;1b=1OW4_Ga_-
z4#pC&HUD7h2bO@FLm!6AA~AKiw{RI3nU@GI?ifY)n(L{spW2^nKxP(|z_phbo&5tW
zjO=7Q)vABL|F_L~$gtWrjmM8oR9e{&x6jtmanG!~t*6qA!Sr*^e*7ugjL%BAp;a7=
zs^Iz~%n47(>HIu^J?N`ShLTm)GQMeDXN3#+IW%bin8?t5E0!-0`F3ezWQq|3dQg>U
zo#{-m7`b<u*V0MX*|i?ZvC82a*IYA~{`~ib-Cbu7^@q|{kh4T50vv^0kg^WBd;va7
z#Ra<FwyjBd8y^li?Kp7AK$(xdhiG?-)c&QjyhFZa?QO56MvX!G+3x~xL~afC(l1@^
zGb_kQF-Y&hNVD$V`W7bZUrKcxIB>>*ritUupRZ{eoOx+j*uqoiE#g8;tJN)rjjOXb
zzsoPRAvbRt<`;r*K&M&|rf4E%+|jPd;c+Pl$`f8e1En6#Ip^p0xq1X)dK(zOJk)(K
z*k*ay*sL2{h&P~MB&yV#ai`zTXW16&V}b3d^D+t@_oN1>NyohHzColCuL<`^FeQ@h
z7){ib!4Dka=QQNj#mxoB5cMVHE`qVorKPXCIaRjtB?qG{u1DqGuS23H*=>EST9c7G
zP|_;(q>F+0fJ5_lOUH~33=CsN*-OoM-<+}T@$p))<1|6f(^48517A<EA`|falUyJc
zCT6J)tsppjrw3<+Giuh9(+jB*z)y`-9My&`JbdWT+lf|4&&(?G6M7qM&*@zp0uzW4
z;$_|1(oW85d!(rd8&|NkB|F+V99wk8!D~|c7-t07c3;nL<Q`H!GRxF-z-*7sB3D%l
zdc2Rk_Wrgpe1oMtOB|oMPF?~Ng)6QA)5vlNTM{5{H~>Z%*Vr`=<HsI9p0T~>1buyo
zE!2syvgb0_z=vI7strBLk^9rF=)QUmTO{0v7015(u~kydrui~O_Uo*+#=3G4F&@;Z
z#Tut_QuLj;tq=7_kwJW;_ph4Dy1HWFeu1Lne%qLz$@8skd5~X&HRns^`VN*16`ym#
z<EAo~OdWlFNL#Cc3fYbo))&3Jx>1FGCAB+l?aJic2g^0Az`Ne9naYiE;@Vs<-#WX`
z#l=KUTv%CvkGkEQwNv!<ZwCdH_NYI5_UyX*A5*6FiZ#Y~($L2#u6Rq?+q|zK5Z;+e
z+TL#5EuBDqnbE{|bh5>LrdSUdAZg{*F#5D{rdnXyrkH%S==yF}UF=Y>oQFsN5rfln
z-@bi)#+VA>G<V0&;>)(izjn=gtST8A!Ge&Kl%$~Y(GOaA4@i7h`J8NF8?B#y)G2jR
zzvCR@WA^-5MgNi-*;uy3acsW3=fmDY;WE+QK;Ij>?&z|AoN9wi##+3PZE2~CuL1r8
z^K|o9rc@FWT?V^D^2;`!E?PKHGW5+_hf{51H0S<ZkKQhIhLTnr4|Y@9z(|k?l%Ute
zblbeG9Zf2aD*T71uBA}{1In0{hi=iYF6PCi@ux0bx)Z(|a9?w4+WEwSd<IEIQh9VT
zGK$e}Hth0I$)hBeYSSimOZoR5lX34$Tcdfb|FpA=ie^aVD(<@7^9u!1+3_K&s-?$N
z7k>T}Ga>&sQ{2eXkZ2L>rLQj~B}LJ(F8}d<!9yjq;s1YJ?P7pwveJu%Z#T4RG@J9y
z%nH`W_U2zYMfNeHXHNR)A1Y10?+n@)mvmU4x3_eZ7_!?aVE*S8-ZB2A#b?ABgJD?|
z`#JjxE;MKuhAZ=zEZIO2pLKI<kIG-I^Q$A~i~Z|6<bUtq*mEwKH)3KgLA<}u5O;Ki
z4;<~W=Fd;~`Q?o&ejj^vQtDca`+1%>Ie969@e@sPIGGvVICk)0H?|FhQ0nRtWD6*(
zsut2Z`TU!OlhrBJkVsb&M<_y=f|Z{9y+T9ZIoBe2l-4uMS}u=0{C%JPHH3SYyjL#?
zIL-bRGv3X97uDA<NbRimQ&j-4Tf#rvOdZ>&bA5fiY_DF=aY$%-lQ8d4+~d?#FO&~w
z`KOn@{uE{aoFCj=mvCi*JI<`?CcGLQT}G`it#AvW0Zbd#kK@KMWY85g_4A8F`__zd
z#)2D%1A}&H|Ni|9s4(&uRyya2rWOjn5vr=@R2!^z3~`<?yLk-47`=O=$v}<6Z`QiY
z&a#VUd@l45B3g}cDm(otXXlQ|f?^7AHkjZr^lnVjK4}_y;Z^$L&6^?MEHf@@DQVa(
z3;J0Wvorktx9fNA9HYzT_K}=H%r(W(nyn&^P);WyGc)3~dGwlTky0j^&Uv|s*B?E)
z07!fvufgp-zfkv|zHosz(h2y)x@*SVs2i>kyV7UdiecE$p*OOLsjYe5DVGOB6)+cH
z6!tu)Tf43b^PpV=1E%N7srl$tE|3n@tDLocdzsFps0y`4Yw6IVYy&7PX1s_QNp0Wf
z<(V|C?b@M3{^8-iSMbd7`~73%5AuTa^|5gmh8Pe!W~tcm>rGZxzKThAyzUxNJWu?s
zfBtIY1=1A}`UVE~DVK0E&->yZ;IDkF--IJ{5hNaD8YPyk^tBP66Uz0Pg$G$SGAY9a
zl(%&0aB59h-5k%Ut2a?MGsXO|{0jycVveYLW><<8FEIQrqPE&o6L-|u{El*sSlT{@
zzPTBv`(cwNMf@}8ize37jvx>At-tFwTUpHwc;B5-3JE<FNkWL7v(kQ;Wavf(<s6+q
z7jNTye;xq9>;^E>8FOEhmX-#M8Y&|rQ%KmakWz{2kvWpTvS5eO1pC3~VNSh%NPeuH
zx>c+VJ5E8|kKSCAxvx9g83`{<ifdddrx&i8vafyzq?h^D*467Z=i0)AT$%YjDP?~J
zQIjc66P%o!*5qG2dv?ycb)W4MKR{EWvrd2yC751l;U48Txfq8IA2@KPj2H%W?AEQ~
zv}!#^;s$Q-$>u%0FJBIpmO_kx1+8E0R(YKSRPu_7OH55oy}ZJ;svo*!uDdtJx@aDL
zD|CSMO-(%&6yBT|I9n)m`%bbnjL`IH8{=z#<;N&-A!REq>CwMyd!aId9^QWHEaG@?
z>2CxNuVF-8gYadnRjvppW+)=kJ4jvR1%8^cFXDttgme2Cb*tS<+}56duQydmE8Dg)
zqZw9Fue&lKXg*d^%)xOUD43U)Q*G!I6nZ&g;`s44A0vZfVwz);M$*MDUw$ehk(aFR
z5Ig#EaY$gGHL5-`ZWb{Y;B@-RY_Ye7pG?P&-<E%IxN7=w#k|MaGI6ghpMH+6dgx-y
zh);u_DyzbPX$VIFlc8!tBzBixJkd7Bv3SPEo<n$75%HIXP<0Eo1(|4FY!rNUPrbZn
zpQ47{7$$jNk(F$Ou_7oK!+OF$7z#<D?opf(FT28}Zkt7J)-+zi6!FZs3ERfx^sN*U
z&>BW*_pHCX8kf?3+_-V)0T)?ymch+GN{fRjTPZ8%1yyR=l)RNT==bdT^HUct1nM_F
zS~L36)^beuMvNMjJ}Zw?6SJ2cxy3v58y>0bd;jj8M)f5wA7kN#Q&z}LG|BUNYw4%K
z0nt$^tMv4154p_M*XNGq$8UfxYH4Y4fzkUGU|&6u@Y}T!)BU)Vb<ZgNR&-F&(S5LZ
z!=)twnU|GkCRM}-G?&*+%&eIH^M}K+0PEi^bpd&f7n7ee_e>=lWxr?qR(`;57k870
z5)x(BM>;7wkNY3r(6ZU%WY_dmzPbEeLu_jH&#w-xQGpgYGrtDM&i>zpPIgbpP*MKT
zO-Vy>3IKY1y`7VcX!fIKi5^iyOUg<mD<aTXa{kW3wT+oST1ula4-_apc!n&t)4sk-
z3=Kco7q1Qs#PN(N(cxctkSEmaa(nmd7w8_#Uy2B7Q<5>IRzz=fz2<N=^}PW?Y`X=R
zl)jvVwVc+K!Xc?WyPknEQ2?Buu;D%oSVkD)zNX==WOm&{PVV2})QuRTqfD5hJF!pa
zHPRc2+cj8n#?jGn=VXPPM}3!?6ciR-bJqx|v$pbo|E;pBDq&0bub>mD>+T6dG}LEp
zZ5EXqrpHL2m7c&GT~BYWFvYs9zCK&Hm~nLR$+?Q(PCG`$$q+%YB%tO{p|MqIbFS6>
zk9&Vgd2RJ9=-8MQ6CM}mOAgc-+znSyn{xV78g78ikT+s7^%J8jZUix9F>(Gq0m_k~
z{TFYIhN9Cs6D|MGE?n9!W;AMt^dN)#_~(vWSwC#Z5aA|0=UAV#pk2>fdR01I3=HIA
zjaitF3t7&W>3Q7XNl6ER>Vmzj_`Pv)am9{`7n8|6KfkhI?egW;2`VNa<m}azU$Aw_
zDm(wiUF^f==H>@4-4AxYomYLfSC_y*bsYEwmnN|QVPTIxedciz!(SD)Ryi$~-uYi`
zRa92<c(Cb2uP(o<hTKD#Cn_82+%0~)CM&E_9?xh7!j4c1i^Le4sgQQ<ex}Xk47o`f
zLeO;`x4Q66B@1gZs8~cEonNiv6zZeA-{rPz-tN8!-R3#J%l3IOM@I^s^KYYrd)xh=
zr(?w#g>IVkb<yK<dPjfiMWP(5JL-F2(1Ss)zD382gXw_Rd|)#ChT3hJe(fdq`A7We
z<U?z#V$bc73F-2AszJj_Jw3z97}8VsuW`2mro_`lvtNB(u}fxbc-m#iB_gcvN@4HB
zwSi9BsSVVP$|@@FhW$XDD0Z!s6spU>i*tt`kacwXxa}*Yp2(s8f_QpH(sKLg98V%$
zDO1E)95%8CRX2eAB9Jk}x_-=N&5B-}v~-f5WfLE#F1m(wtUAz9G4c~$id6DKtEd)J
zDaF<D8TE=-mOeitDFn$l33>Eg1r8p^$QRrACZle#oygI3y!_H)9pHhm`=oF`GIZhD
zP5L*uy9Fy(kS6gNapQ-*w!HpViGlu@E&nRdIo^*%_Y>4=82%myR$`StGAv;)5&;G8
zrva}SNuhfjHEOYpBO;sVw$(gL+%gz^!V`P!ytm=(nKPSGYs!u;0E;W)0SIG&lE_23
zQDQm18ftRr*Gk2nl_?h{EWD`9Aal#%N3wbD@4pq1Lh<hG!qVEkoZFyWq669Vc7Sa+
z-3|jAk1>HmxRD5Md`<s$zvFX86>+yOJrB86)RoA<g3@1c_u7?(X9{yWFqIuOs<QqL
zwchrnBEo*M!ZAh73t)8%oAzPwN0P%JVRu5Pq(9Xfgn*+?pU^0b<IXS3!^WT_V<+LB
zC4LNMaG){&S&|0(;yeZT3VnTP8-<1Vl>{WKLX-*)mX*tSotf$L<Ht4@xQl1HJ&)KZ
zYwA9S^=p{V^BY8>-i*1jC^P)s<2K5ybKJM@zZ$cCY11Jdvfv@@+<EU;jqPSq-GC6Y
zrVqFrblB)BAZKKfv7zno-|usxJin_mZm0$>lo(@Evdp7v|L<SB*P3v6XjVra@!z!2
zh`V8}kHhv)ktgA7nTM~@)t5IhF%exLcO$rgv)8Da@%@eyN*B87Yt{C4ZZos>Uaobh
zY;0(--`>rbf46Lb0|_^oqjsm4*48&RuDiFYal1n&E{wZoi60(pKK*N2`J*kNnWZB}
zkKRCIo#kzB5Pb?0_>kCNSu($>zQ0~H>~!WY6WjiESBf1fAARlPdqwRHW#E^nVqcs=
zDqjrtx^zzu7qE8_Y8dcSmjR1978g%tbpwiH)OrDU&*<xm)@uP@@4Vgl$vXC+ePK-U
z1S>M3jHhg@sj5m_WAn4vysp8nCMJB6myz$Q32M6x96DWIBJz)rEDsqy+LIzR*a$_s
zYV@A6bubixme5{Dhx;3(tWCWXjkPN+2ORAU|K8H2Hy%r?%a{kC_qk<+Us_6{WyA_2
zqhv~VvC~D@z$jJZ`yPGfA0GN+k**x~bYC7S-RVZQ*rN05?v%eUZ>g?ulD$;r`w7=K
zk;?pVyE)YeklR`A5?~6r|AUp4OZ>APB2%&rL%GHrGp!1ze6yJvajD5}k+c;cr+`W#
zi+vp;F|La&e^g4glWpByZKa|elZCnYEP4bubl7w=r2r#>Lqmg?qp8h;J>9NLzo){-
zfw(T|vIEO><wm{DgQ3gb1Rg)ay~8O92g0g8QK>~l6^FlkAnRykaxNkxGjr?ka?ZVt
zv8nn7Mk4bB<%V!({GMFUk@JKToa5mcQ7NL1Xr<?;_iDqy$5V8(3~_|z4Meb!3u_$?
z|A7HyN%er@DPv>Ud3X34`uVT)hrPFz(kW)#@Uq&r?MGiT#~Cw(2N}CvqQ89#ML~Em
z)y0%+R=e=jB<XZLb^MGZ9gIWewqJ`Bw_0l0q^ff+b%eoTOW-XWak8va<|;X!Ap_$u
zHC9=t(kRN_x_9&31J(jwi-2Q1#R3EO?B4R&MUn_6qXiA_vY!PTw3>OzB`$jTgNbQ3
zT5r)#FefAt{JvG57$yr64{zR76sDDr+f*x#Y;=zsW3_vGJyi0YDX)fSPG&Oeuuo&-
z`wt)Na({L8c{lO%^4(5q>U}yJM^rzpZ6v@)+8I_!innp@r_vQpExApaJ7U}0z1wOl
zjf$@Pi<pvv*33w{w~lC&k!v3w<0+!1O1afJRxYolsqT8kOslkSHp)k==Z9*J8FP2x
z(P0Rg5UctmXQ%XfYWXqD9Yvi)fA7K&*^G4Sf^4mYZV6Xv^Dj@_-tX<$fN#wjFoBH%
z_W`pq?kD_m`LbmaHgY%lmZC3Z>JM2)`N+F3n(+ZGEe_`QAK$M!msiAvFz)O@fkldc
zde_$GP8`v5BLM;pOo?s<2LpZu@0~PYfD4hGadoj3Q4WV@bx14j^3TL$kW2kUE-2+^
zdXKkb6RK-#*K@JbqB8*E<%&k9fB$5rwk2QFWOg<*K8Z;WKU2NWEUfhUt8ne82)7U>
z-y}?bx}W>CurQIf*kQebZ`SOnNlXE~wm5(83KUST=$Z}cVCSfFJw84@`e6RbJj=Yo
zj?X>&za~CgNDBbQ-|ht;TnXvvDe7BBq8ka<_W4~$r<fg@eaPs=mzHKoUg3`o^!4%C
zdd7wiZ>5fXw?3tC$7FHCKXK5}<uZBO_;bUC4L$nywVyNdP>`btLwTFFCIFuYyt}Zm
zd%;hpZYtG-#jtyK_(OXHz+8%bIa)SHp-BCHQr@F{OH;vE$xzuG!KeR`7jM$!M*1{o
zal6ia9s|HT-d5;2>yYKCix(GSPAP;5ymjd0ql)zPy#X46^@Z|+v9P$hNl@IpF!q3~
zW|I62Ked=|3MG#w-jF99=+k0dx!@#6=Dd7+Q|R<F=6Z=BL*d{W$6R(CwWC+(z}Vw#
z&^dGXLNCG5_Ivv2Q@OJ|xJg}T3r9PwycPHgag;D&L~$nZ_Lnqw<bZMGPP0+ReH~6w
zY4mFbWq~O7%qsZyQvkAkLN^0(vtbQ)r|WArY~P{kMY&emDYhh@%>+sA-ot;FvTpV}
z0c@ZwxO{Q0@&Vc33wfH%4TM1$YOdrB@7GM^?&;C5-w}*;zS)S50c;nL1JiE9#;ZrC
z2Xm(wE$o^av42VNaraO6E{(zboT;Z@?Zj(oAFf?&1yU+|5+6OWuk~;pU8sFL#?Viu
zuQrX}5?>Qy?h`Q<>xaur#Q({3%%=UiqD8O_E6%bAQQH>gbMD*=L2{>UeFM!eJPuJn
zFeIxJ)K5aA&g4ypzyoQe&|PcWSvhta-=}kl%_GH+rgz2x-r@*O&m$~%);Jp(OmnBS
zB}X{t3*@FPAMTpDt+dD3_`yUO76%*9FFf9oYJ?4#wIP%;mB6yCsk?+L%ARR#l&s9j
z3Z9Dg-SqU+M+IwX49o97#rk^m!2yktI!Alqffus)ozb-3UHgo20l+F7FzkS=f0J;s
z<Le6TBehGfqtaXxTsa`-9v<!ZYxy!l1_i*>Qh89T5-aift^3k8Mmgk8b@;Vyd(R6g
z)9drwY>$c!Dtp|p_tV4ha7m*?&)o6bK8%5;vnM4x-tk@LWp2+0r$dS543(*u(-A%&
zN1wgBapBe4w(Z(&NiXA<827anVfObw;=#1DZrNU=;*22>zE#z#q1#*wRN;k}E56b>
zdhA#cCT;_?wEP6R$tX}y|CwdPeU>XS`sXapmk=2S4f=e^s}OicnwgURv|f8pySPXk
zn{#cgQwx=%AqjSCB#r6P#gUj{Vv8gbSbuKQrz<nv9w^4a3Gb{hx$;<g*4|j-txXdB
zd)H9PcI?t6z5IP{?h-7mBI4(`CFGyYP@*vHv~vo2sT6vdC9FD^hfx=T0*;odc*@Q?
zRDMIblh*HVOT{W@IWp%_=z8$sCx^FJW(uZTadANSmdcu%LZN;TA;3r}%bhyT`Z<Lz
z(N%sfvsDR3Xx}gW-i@DmN1>ps>$bcDx3(n4i9+=*)(zGKXO90fX-j&E;PHx@2!Iml
z^1QJ#l?24f-Fzc_35FwR0k-@T=zZqQrS4CZrtLW<!08Ri{jOY#57}xiauM)JO;xVk
zr7u-i{P1#*$|Ft%iQTlg^gO_-P-B7w(7J%jiH0oGN^t<k$^6jF8fvsLw&jMx#n?C$
ze1eg9OG^uaWV}l~nH(Xn5y5)n`_Zx*oSSBA>9x&rfVT55kQ$@P>1bZqee`{>2${}Z
z-l>$So*5Jx_b5Awy*}&bbwY8Ss26s@^uPNTX$+O`OOLQ$^e90gkB*Rg;al5BS>v>q
z*HWKX6py$6f{CKL-O8l!<300(83CB760|Un>x+#7J=9)D_i)=7|I}&72H})lC$hMH
zX?#x|U5Y2IjaHK=U~-b^i&5KFbAv6%hQeQ&H^`Pi`7(auL~BOujG5VmqT46%jugk2
zLJ#=y)BHo0H<dfl`--Di*zN(QM!zm&!Es-Ap7CphELYz~ojG%eD6iWzvw}=_fnVzD
zPqfN?L#NPp?7FsC%?%klHZ3Pn-e;b!9GwD?yKujnG1oZ<oMY6_QOqU6!d_ZNY>kL!
zxt`z!!pe9&A=B`EY3Q{{cKaS@=jJ-;YM!J|-)lDh$$L&Pu~qRD7YazzF{a^bs_N>7
zj~R0UJ-x2eBu)b#P{W!J=3YXsA;hBF@4rUc%(`Gp2ql@4k`ilu2_+bH@@K?=h2Vr`
znG0_ou&sPxUkZPuo-S)OMHG5qVIs_{<v2^0QHy~#y7B3u?p#E>8y*2q)KF8Nx_mjA
z4(+|Fxwwi(&t(GF`a7>RnSOjVL@Gb$mZ%I9d-x(yXfM+*iOv=vqu^`Wk=NaO^msyP
z8Lru#wIgPYf&1cf*4EKK3NHnUYunCI26l#zg92696J}IoJo_L17pFU9Kkwk5$Ri)2
zrWOE)gsMYvS4f-}^NQVdvhK1^f<$k-F-5`G@z-h;$UCQArflG0&0n~%P$YF)MwpN3
zBZSzDEI<-qAv`7D`fGWC%<LNcXe0(m%-5c&w{yORV{ShXSb?umH#flH5dTDf@;!0l
z@injMd!J`U?QQl|z4LI@$maa^S|!%bE?9b5v(7j9`Dr>#<f<(9^cln^Fg$~^h#+R6
zd`Q7+rU6&p2Z|w-gX<yPccz3|YZdGL@4tGJyjI$@hgifaXW%d+Nuv~|^qOvY^o&e1
zBpx*VoU15b3A`fM)4@vTWQ*T_`t&BgrB)4tK{0f}gpq;&c7solu8gw#ujnvPyW%iM
zuD~Z~$=>f7URP5yTUQP%w-F;p{yX{w$H^EbdLqD!tu987b(QXs%8u)Er_yugI=4Cs
zemCk);RMdk6P*uuxKr^j<>IWO;F{ta60<qcqT9A@y93GGI4~p}wHR=zJ_t<t-Bee@
zg)`Q!(o}qN#^WPYUp97Uk22Gx^8=re>ffLXSWFvYsc4i+olj+ko%7!9r!>|*?I6ZD
z@a!#vRZJ<C#VYz+yC;Z!T6S<~Oe*q_MZ0$0&qzvRBWWtNryCNr)gKfgW+m2qUK-UM
z;&D`C3=KrQef3^6vBFG#>M&DkD}orRVLTO7RHV{h<V>*{IQmv-@>*&{5j-pUvC-dK
zMB*#WZ*fcxZ9i3kO=cx8^^js=ad8q${XE14adbmiJ^%2tXnA59zXx#^jwP#z4$6H=
zz^<A@I!ilr=ulEOGtPhSiyO79yRnv^CQD#XvKsaaq#@oqg!@WA6&H|~V5{fQ5}S4(
z-o11(#7HlV5PxOFPi^`y4yKfKD6+p&XqzwmlxVT>t|Ht)s5SHQlsMo8vbKxa@2?Lv
zen&rUd(JAS7&CUZw5_SFO;@>zj{L#;F}XvaiU1NWTzEA4;JZ>RIL7unE&=g)i{!xw
zc$-|>D@;tBb$?H4>PeY{lHhGxYap-bbs59&V)Ppso?jS=fTVAomJdJ!h{IJtz|pn$
z--H2F)4yTme=hC85JAfo7lh)2M>T)tntytFX~=Mf#y>xpo2Z<R_~q$);r^_o9>#80
zuVvgo3|$&{+^FUkwZ_O^e75(chN>+Rp&EN;&y8Oml>XftvV+DR`#=C9J$+5{+^=1@
z+Gt#b#tW-w|JSch-@2s?7vW`ca=K^t-zO)f8W|ODf176gJv(-9v>P-FT-&pi>Sk$t
zefN}Z%(ko$`awg?>t_%*g+mC84vJp8FBeZuTA}aF<_f(PxIjlo8tDh_j{eG=i3|@R
z$={4VIO2e;R3CMV4q8ngQy&!W1vh8R&O_)aCwE=Hj*SX!@^E0_n~RgY@Px}s9^Cu$
zCyQJaaWO7tuhRN)T_gzwnvKnaJ=yFmA)yf59f7<N1Z$~qTh`~fxl}HV)+VKHebtNZ
z)P38xJ|}#?+p~2d6sF>1LPxJK1k50J)aQjsUaSVe{cCJw4n|uR@H3awxf?rPI0dfQ
zK3|VdiQWYd1kx<|71)R`H!)cX)g~~o8h!g+OrIxt86LWS+CE5JbP?Zb_tKIj2?<Cw
zTzh7N0XZvItVk3AdBQ1Fm>+out<cFU94l!BSLoNy;vFHooR6jSTyLN%;X>_NTUl}-
z0B-KTdt4Y2&C-?IKL2=Vs2X*!uz@I7@Mk;S?=xYGTsMtMmpz6=4{o5w>a3{IhZ`$6
zHu2m00C*FO>~l#w89noYtnaL^!CX1otr(m|0{sdMupYPW`B-J_5mZ~@!zf+~Oe!<4
zk;T3JqwN;0=<e*?djFR*-iS9T7hJ?i*(1kZD(9m7bjkoq$b<c-?b%V8KNDgm*wBWO
zw^t3Ff7fzII0Yo-V3AD8^4-3o2HMYF$GnFy^TpqB<(k^Xv`yEeqE?q)tre#0S;MCD
z*)(2rmAip+=Gf+?H^vdm{x&1yIY%C)91IqWF_?v;ZxCa;AA6#=#SGI}U~)b}Ai1AE
zD<65nX$e=LE7Eog^r$qDN+pwTs;zl&2!n|+K<!`#4>XkzRLW7q-n@Q&(?2U#W9;MT
zC<*;2W7sPc<5}8KyAJQODzpCZ;lm3fist9VM5R1~XtNrf&$T(n1?XH{_H55Yu#(N?
zjv-BU_c^lf6A8ola+LuU>pd`#T`O8v9Bm7}>hkd>`vD@O?V}=oKP#7O9pM>VpeGyd
zZ|+?Q6Q1Hx7^hQj2oI>*(#uvld1<}>ysrb1Hb?x4#@3;20@oH=YT<TLGWn~TN!%qs
z^09k=Zes6pW!cmmaxcjrD_H@c#JG2@i>K3Yw{~5mOfgOuZeArH?$wK_nai*C?fSo-
z?f?-5Yh87^QdT;2*O`b@?tc^}d1~>^>Hte$xU)aAyqlM#R*{JYcok!*XrCt#4nUzh
zzoX&b(mOp{mlkO23|h8D2|&2gkX!t@>6l7>5NXloO`A;XD+>gp3rZSlJCI|TmsgcL
zZMc747(LT-l}8qK?P%|wUF-g5`wSetl%WiRO1z$3|4++9?T+YAtdiLrq&|vqvm)J+
z&d4shG+DRFj-n|mNl=o0Rkb!bw|a5y+_nC)fM)cDV++c&Qtz>(M^;nojRnRhFM$!5
z@)1`kNg#h3E%k*G&Cul^nNBE=nx~)iaTNV51(Ia{<hP~Yze9g1T^{T3@cR}PoA0*^
zTSE#+czd@phGe=EWb)wN!i{wF*(~N2jB?i07IL2pZIs1a^MXB&W`7Q_L**O<Np9Vy
zzo>9-b>M&qNHziZMHS8veTujOnW4T&9xR{x#{At=mP}b;O0*0j+v=x1mrC?^Tq<Vm
zAOX;F+fCgwIu2I*Ufa9H{QfpJ%_kWuN7;}6odiZNO3AmnI*!!q^or~55tY15&K+o<
zRNJDjCs~n^a##fxiCi}93|**rLn1#P89mU7evZFAV^%Ixxf;BC;Bx1uT$NGFRQhi<
zju$78Lq#l7W#<8EYM~3;iEK1qr^d~~<sl4EFx2Yah`oMEZDWMCe`GI~x={F9XH)hj
z!i1y%uc&Si3_;dJ?Wv}mUvdHOZ##|*mU=Xo>zq+^X08G`cT=p3Ui3|mhCCkO(4OP}
z^E5H_6+F?0GjA>lc%L<3cUOBaC9RO{VfN&B3on)6%-gcD!CHSfEgh<I&?5;(S2e;l
zf$0!ZwI>@G#vX;FP;Y#Excy?=SpQ7BMcs@i+bRbtcM>KKT8(L5w?aamZ@52pNB#^i
z8C}Q{lB5KrFVRy=-{|M~ocAgO==i-1ob4ghWjwh~p+Bb7!`@%MdUZz|y=Q_^!gtcI
z07MtH*p$^hW*wS75&{VH5?2h|4+AnY5EFvKfcHd9AAPp&D*6wCAwhirj&LlxfW62C
zWc?R8{vI_A?V$E=X0bTG4@7E!?0v#uW6{Nl$JwZ-INUY`y@BOrhcFQyzZtx8_VohG
zfqT#R?|uB`09~){^5r7Wg>IXhwl1n&IwL{E%fg{V*e(((Kl)o~W~fOju8;dtvN>p8
zk!rbHcl)1uk}>0+jBLxOFTHRNahz|y%ZKV((Y~m=p1zK*oKvNnQYD!CDag%qTA2Dk
z4!^*)U^>R`-*9m_=Ov)d6U0UV7E4e{-oSDT+UDQ3xxS%+c(?_ordNLLrf&Sq{5UBy
zGt+2xZ{zE{%cf2STaWr;P~o4f1Vj`Q@?-(r#KgqSmX^=g-HUXX96Nf<m}Su=T*zKI
zRBVU#n$T1+5T=E<?RF#ceDa3i&tV9$sDu~>#r>&*Zrr<<mNWa?Zk{z|xI}-48If5o
zyW7JJ0Ks9H@vUzX{Jk;IPM?b+H`iEwzyP6{!f<JAQbE@_*G}@+NU*dqZqI2^zz7Jl
zX-nr%GSjU>wakZur+>w8BLHUDv23O$g3aFabV(u{F{1kR^VB)9)uw!M_jG$BjsqC5
zLfxY)2l~sh{6|a+#00`lzz4L&(n_x_BLHHYGP2`Jzgop6Z@6CtmazGB8K~Cu=E~Xh
zGSPbQ9Ny0EQQm2ww$@nF#}X}Y?0v}1If+JEu(BDov}xD*%FUZ|rghY`s~o^mp`^T`
z+)3~jAl&3b?v$rrv2>F6kN>V#vywMA>u2ZFFQ3v;!KP6lPU+^AD=tKxN{9(bl2%Is
zR$q~XlbzSFukKrR=|H4{Wl;Wa$}k9)ucRyLIOd{XrzHBKZ*W2oQ^Qae<NViLD_Ld1
za`0&F^Y^!xI5){kkyE8Owzbep3+pI{l)Y<Fs<9ibB0NLKT`t$!etBBmkT^XNTthfr
zJ8a)Pl)&c1sF+dGru1lTuWg?qb+a!e_2fwRSN5sRYEsGEdeAdo#wzH*e$PB-E2py2
zx%Exm`scObzrJ;JZh}L`ukb>isi%MA!Gkl1ZE)RSyqHK=z`N!@J`d+HJ*ERbPn^^Z
z1FtUC@S7`sU(VaOcdavZ_=M#N*}Ja08VJVNpZ^ijq-qzxSYID^QNX^1Jk4)^NUzM?
zEVzX^k!nBmNtuJRYNK@MTB%%23n>(?^j~C{9t7SAXjFsAR@%S+wzV;ogz}NrBH{3k
zDvL4Fw89#sbu)c#euSu8&CD$f!GxGLSvl2QRGiiz0{lL5cXMxU_!PO*r*@U*gw=Z%
z%U!*8Br-0pWtwJ6%kQR!oC>D^&95loh4?=e8R2b!5x5I<^lbQN44AXF7glsd+e}{s
zAp1%4jB)10_6wHE|G}DTULCeZgVyU-dPc@xqqko!DBBgz_lLtpYgDqzG41m3PCyR@
z!UlOyw2_QGp_<rlv%}GH&@Dl95oNygamtd<FjNMP8Wo)V+<JPh#{BZ+(?z6iVEtE9
z^{;)<Bq&$Ke~7q$*;d&O#TcHUyMMuB{TOX)<&#a_M5VET)2{c(ucf1NHEhdLB}3O{
zJXBj9rakQh(l2ovmHpGX5*s6r3dafu5J<Gh!r7?7ueHHXQOW21x0sS%?Jg(gbRRrN
zBtj|%j}&ZEN(uV9Umteaxl*kPSBTB6DXFOe5ICTvzt770TjMb=y)kPJv-c<jMm?33
zQiV$+RFf&z2RPsXfJOZ-bsje@AVfIf1DFMI7>npNW*D!5P%=vOyZbwfFHGAou(6$v
z?%%QZX*yc$HlFZ;m{1=~OaQzQT`!lm1@jro{0}*^dpVw93878WZmJh;jrx2KPt`!N
zNq8>pOzG}o)pea1_mBSW#8<@R=@ii3uX@WvA+Ofi8cLTecIL`{AUh4>IAYTM-0DPS
zxiV3}U6!E|U{~p{=S6(5y~c?2gwJUM5qJmum66Km(*M9v2gVlWnEUoI@4|G-okNqe
znc)H3@gzjto!m(R;kX?rmv=ZIMX*gfq}6eI3JJ0>v=<OIXD*&+j8pi!A0O>j6_g^r
zC6px_2bpDQ&QUjXX7d5~`z2_e{lf1(NZ*paC(#p>2JkC(>hoQ?`qcjR?odj{fx=Ub
zd{gXZyA-k^&!yXAuF-{y7QI=o=^OBHlZaNSh^guNR(E16C)F-^XCk<WwJzjdT<g$%
zN!YkDDyVwMJO7z$#?L$2ZgX1x_SYH?VLG!nviIoZs|nXywtTt2QD>XQhm7K#kND?T
zR4iqp1=gO`V;g(t=90^=_;9eVzGsjLq&~@NQ`%df;$TOkrM@?9rknkG_^b2iBmQ=^
zTWza;RqUS>=hV>PP?xHtQCdshSQu$=Lcu)e^(8Pa410%#&_`KzA?&L*diK)z_Dl-P
z6s}JaMA@*g<xTb5Vz*ztR`6hJ;{Vs)n@9DW{{R0k#x`RgTV<!n7Ex5FY$asRlFG<d
zy|gGrX<>}Lq(UU3n$g~%q)lThYYQ#3NYbK&q@q;c+ttkH{rSxKo$v3Q?>XN;znz&g
zW`@^tJ+JGr-0%0t{qmy`dnK3eXy0o$S={Mi$p@MjM{uTx>FH$Pi&>h*nVG_{1Yorv
zpi=Iyi;v&G6J?wj^XLtg<3@*@)uDptgQ1LWX(ow{CD^77+v`p`1qded0a3Vq2^#LW
z<>mzN(+c`7uCEgvy?}lwA}aXW?YMte;BzKv!@`6lIo1$LK%N88;fF4gp@V-J(7(S8
zgJr@645YX4d)WMqnQJQ|Tjne?nb-j8k{|9^y1=~htBaMjbt83aX6=k>4bSU_ktVx0
z$!RD#BX1<>6%n2)s!Dnf9du>#4gJk-Z!`iLQd{pL+t|?a{f}-x&?Sf@AxfnT&1QE-
zl%$>eLE?RHh)Yr6i=o@4f2y?lu@|$28W++h?ESNPmqtxs687`kuGc@P+oS(if%-_T
zVuNnUCLdRHEf5TXYm1Wv=syitBIoqxTTH8M_g71B`tG{-=ikG|Gr47oWWT~W2}Bk_
zj-6~!BR6Xvz(HrjUOO>J!A!{}`ZF<FAg;kPxL;fx0F*5-ycwSB1T-ilsi1&YEpo<V
zE8G(*CiWKwvWYiA`nk+606~!Vz#{1SI%!O|-hfV<>R8C8Rs`wvm653|Upr=&!Nv=5
z4P7~*vm0JS`E1ZWFLeU!$Lk4HFM8JPyy|+xzqm;!Y3f@R97d_JviRJUspp>>cQmo?
zq0Y2HSeR2;MsZ%=+U|QSr=L$Vdi(s{E+GY>mFA?T;glVdlgB`baE$m~+}1K2%AK*F
zIo-#lRNKiSq51xeyGE2cSTA$#-J{E0_2V6#yop$?f93ROc8=Q^0mLh?9IC7roO%##
z+6AapI*|$Dx>H!fmK%+as4>uicNn3lxKOGqTqvG%{WXfaEmN_D5&neZauVluE8(()
zq4aB8QHgy=cyWdMM7Y5^P?0Z7cv{}Sb9hde@o8&o>nri`vIub*jzn66?a^5jDs_yr
zcu>!OO8)5LiDuJ$K#HifseiZ>=pGyT9RB_PqyFKLeXY=Hdx(&p+Gi?H^j$;l=D?5M
z`x<{1iD%10?~mW>|3N{wCu28#Ohj1zE}5z2x^Wj{K}x3|i#a8xCIqvCC|PHw@{IFO
zp**KSq3-T!!22H-dP=OYMD2i;W<Ts8%tl6oQ0C<gI(+x?q&YVVY78s)V&8<=2DVOt
zi%xl^L;T&PGUVr_sLq*gVCaKYZK_$&UeR?caU^rn@*3@}i4`~PQUEaxwC}`hYas&Q
zp)%+>1U;5rfl22(b~k6gZo%jI@+8lu8|lf*(|2p_Bhvm}kT%8%K&v0p#T#pCW-&s_
zq&*TcF}rD4AC+Z>4(kvQy+PP+LcaedDw2ud6<01g9V1ASK^IA-6Ern7foK6uJ=<Aq
z^e59(7`PT=z34Z{E_9HvrM84<!8eiv_n&xWMy=OFPUXm)16}$K7;v9>LfMdcX3tL_
zNBmQ=j#;qcCetbL{4GRS82d3l)4lTSUSXE*%wU9(-BVSeacMaF$@Ey&p+l?2_=FY=
zUcYxkam3@2s?ujgfAh7Km0M{l+|rxF!X8m>c4emJM{Vqh1#d|l1bY!S-|heyKszv(
z;`lK$W?X%`G!&YQ*u(R>xmPhc9T1$Td8zbR;FT{^tMu7->`Ht%M0GW2$dHK6#$Zn-
zd551-iJ|9raY@druYF!!kMC*6p)OC1&VzG=g-tX`I?I*`KqiYe>`0(u4`^*zYKrmL
z@f*3$o341ST4GUDjbgOov4LN^bm|1*x;tIpfKhcBHJcGYm>{uYtMuq`j|;j!e)J3o
z2w0>4UPEWstHMx$Hye6h{o{{6hShAYjC$1nTwKDkWlz^%4qD(R{oxe1;6*o|yIEdC
z&tvlPzzn*s2}5dAH_y;+@3(R$R755_zW=<M9UT%@mFsvu@yp@5T)G8*5XL>1tuPiU
zCWde?UUW^5?%k~$qS=*e*XBK+SFaXTH@&j*_N&6e!kP?lKDwcI4Qa`*>l&J-{nWvk
z($?^1H6;nn(h&>@Xn*KFya2pec^(DhD^^LhI-b47X1nPds3@e@z-+Hv8jXdv9+4pn
z%t-4r<dUQt*uX1@vgpHr5pz@G>i(FCj-QgR<Iqk1(s^C-^e#i~fJ;MqieWos9!ldn
zshnG4_`2F@3E7?IJ?5kur;Roc`wwAcyJPy|@b-|=ZSo*co@K6|`-zSjhdy+i(iq1|
z=!uuyYXz?;*1*SSbe-<xk~8}nad`i*$FO81GAY59#kyhDRYV5(;ZNspw#-3XvixNb
zMhQS-$Sm#Bmh%<A%)w7?7Uhc&U2sJP*Y=HIWJr81xa<8tbVYlA>~0B`uzHD+3=siJ
zK7W@Tw%7P}H2;;k7kV8O6RoTIcSfL%e>YwfrKJbh3<C@OmW+v9^e0>4(9>-*KU(-&
zPdm4SK2ZqOIBQ}Xq37ZHZ$IAV;`K-CjE4P<z~1SH1Aq;Z6=kp%L1-Zlh%zW+E$SXk
zh~}ojG<0L}U|d_HUSz~Gad~=P_m9uPcWJLeJKyBKvQ0=x^DTuK8chbV8|mJ-aYu=n
zbDV|<$VwjgpX&_{FpYLXoC7#vYHSdE_Q7fc1vRyG8|I$ow&WtKD^#M?D=^`#HaG9?
zdRM%6I@JgYj53F{U#puk;DV>f#7wzq_bFWQCyfhB!$;&+*XOqwZ-nTC3c#iedj|Vd
z8A2tzE+O6f<VkJBlQhlOuG`Ky`FFuSYds|&Rlqb@!0n8i6XJ}dDa%)IH`K?ZyMkmS
zw~e`P*K#(-tWPpsaCt^Rw80$_z63RpYt;sKOpMBPdLaHA;MBBjPh{fU*Mid*52B&q
z_F{2K!&5lBCm{Nq+fL%~ArN485WKyd)FiYja2V3Zg_IH&u@<uvKoMwTpO3fUy3h3o
zHs-BVTl;%-b!+i0&xc!{8EV0J^DU8KQbYpCGC#@%c+x>SGu=EqJv|%Th`qpoNOaNj
zuHSotk{$?yi<KG0I*63b+)<vy*e$E~x27g<U*9x7!b@suoPxyIb!Fv@HZdzL3^1Pf
ztNQ>@2#R*bk9!tGi{W39=`*ulG*LijPoF`9)GlgXYWj5Bv;$pxmvd>3*VaztUZvxI
zqtigO!uEw25|i64>exEi4bk*6(zFyroKVCy!eo$TaIG9+!{di2sUolC#a<2ME^6uZ
zUdT~9#~?0=7~DJI&Snrcn2yX9M==g#;tMVT9y7YWIETf!u(ZZ8a?J!@lHeTq`)?Bz
zDT~0j2(O`$ky{14Tsl@k!HY=^d?_HCX>03W^%h(%CsL7NW~4Z%P<k+2MD?+cvf_uK
zgS)$;=NFM1-R3mXBvZA8vrMyw%D*0v9LT^NjEHtK_lp<n8yWdGd=$nY_yJrB8_&2s
zi;WOL==2Mfjd+)qlUZh>6aeq`5>gOR&KmRt>VoM)eca`HOAA##CLM<<WsGbF71q_v
zj|dx2=;I}Y1P}SQkgK72D+SYGy3%~WUl;Z^CyYOc;5z2m22_!j9|rd6W24jDnOy+u
z#cB{1O}Uj99lo0z?-2XwH3tvsOi@u$dqsaf;KN6cOdp&a8Qkq&>Zt$H0$jPW_+Yf?
zY3FN!Fp0JmB?QV!(b(MjRz6r@XUxh}V5BUR;}q!n1%vgz=%c|IR#BCREpP`R>rIF;
z7Uqa<t#tv264<W<Tbh<~K$6o%!-#~tW(H+F0HPE46ui&d<CAX*NuS~1-#Gtp@b@@?
zk|XZ!@_?^cJHb~XDj`;H%(feMY%Vn*c^m3-S3IE2iJ>#a#U$d&BmsvB9YK9`wI-H{
zFwGov&{H-E`UrLo3=dNB8?ycrzT5N31R)5F3zrZU#$v_PhKDTzZ=+-F-@iXTXL|bj
zTLm??^Zo#O_jM=kw^x@gX9E>s%DgcXXKI~W>7JFx(<$I7(vI!8%ZLlJaA$|uxxbqc
z;mQ?spI5Y6j)I*^AXL8!6hmq3ad=V<f!aGv%0PiQX_AMSI+DKNJ(BkE4qLZ+!5^WF
zWMgipuw&MGdH1~?aD@zwirU6hZhnQ)pwZ~xWK_)h01;|O=SeLG7`2^S@`qiQU}hHv
z4_=G}0E17T1|z2NHt`y8P^${Q(DEN7^w`nI6Ny>o?74;zLI@{ybs5ETXC@TZ5KLj|
zm6@Zz#IBpjI=!J$Mol=d*AGM;J4QxGaer|od$%4v-X5C}^$W|?W2nX{hzF}S_4c!M
z<dqltk;$&L-CL_uT91p(K*>C(V3jG-aWRC!Zs)$s9)!fVbKQoI5(_03v#FIHWSCRs
z$jFaM;zBq9{D7uhc`ha~s8U2{_*n<FNnjo!aT)@5l(9hZn^2et18Ziq;G!K<TLzOy
zcI*fa3F(ITGEgx;b;pJH=?m;|oISTh#v}E5Yej;XC*ot684N_Jj2!1@^4EP+QjhS?
zhv*kM6Vnz0gIL!)oNV;y#8e@KL(s#qcve{0W7m$qI+*1ofy`4Y@<$Mn<sTbdS@}FS
z-cKFDJ`WX_OMQTfW#5HPjI-7I_|c<c(o0CtP0`E>^GA9pNyEqD4A4HP-BG9p!L}A^
zKJN;e=JbRX{~L`<XiIZ}%kB;roj>bQhT?!o!;U~+R?}IHz?M8)M4OJ4-L!StgMicf
zPMo|>tOAr06Kqgika%j+OU;%SW~@vcLIFnpTgCprTOthZ;tH6uxxZzq`T6kpJ8z*!
z?cj&|FDT*ZzUDH@Ur?BX#Qlj8Ciim2dKB|r(TN#qy{4FA8vzBis}j2Gq_q@*iyH|z
zSk#VU(pHQ<<W~1<KR?NnfoqWMTp;QS$6!JCk@GvCR4PN`A&?7_rqB8F;gL;0Uu3dA
zYc#r!Jv`c2g(#sldbWGeFL1KX^*L(|>#f-10xuBsy394!cjp?I7AEa{7(;bgYG6Rq
z$IQGSX(^uw7Ti1Sle_YAdwmsFpN-45v7wgHPn3s3yLR;GWoj{Mb&C$y!^7hKmXeW~
znXyCB4C{<$Zwoi7k%M*$AOz(LTjE8Eo0b^J;5r11J~|8?kQP<7WgpsR-0k2isY}c!
z`Ci(*IRX3T>-c$}`vm1m`q>1V$|_7D!~D~lqk>V9pfD6d3cUNps2WyClh1#48bL0O
zL#ILEvTE-%ap)o0cKdp}8W=dfIXV`yoiN8k5s1RD*-h@_H}V#Do-<)ip-&G1V?szn
zU6A-XPZ-HK%w*3(A=YfpYXc`(QwyjN^T=?mSs|CB(IcW}L5R7MN$02Ae#o5r78@PH
zIH^DQ)-99n(hdSk+1S>!7T{3ShYue=F2&o1mi1S}Vb?4RrZ2rXE3~MM@-e@5)~s3n
z5TmqLSUng<zad4kWnxYW*o0*e3K3M!)365=!a}g(Sck3TKlqe3;jJ#T3Bq=qR0h;?
zlOvH(1im1q3^QxXKT&?~ewWz&)8FV`$6dcAQ}2=H&Cfp(uixTDhvigjqJ-cQxRV^u
zT>t=~(FgA`$i)S3zvBPM4imcsZ$SbL8I(FoROvhdK=%36EO|zDJrPQZ#+D0{#0X#e
zNJn^R=q3t$2|eFbia_SXsqy3A=Lt0&(g=j6V^O(c7SPZZx4R;1#cA_lX>PGw$6fRO
z^N@b0k!H*L?;ke*m*w|zr*xa{9{KFKQpHU>x!BW}yQ=;4VCTMpzUCW-&a#vy{&}R^
z&5<1gM|4ws5cq83v1>lT*9<HY>vaE!`cZ!5;z^;|jdr~b2cL<%*!;?=xcP19hL1Ts
zpIBNEcHFX+WixIJnrsrmhFOcQES6I2cXxMpMcp1hC3PEjcS;OTtFDndR`ofqIm1@1
zS@VY$4g{Z_l0P@M*lJvO^zdO85sx()@a|*>$q~JQ#}@$eGznS)3|;#2+Eyn<+n;V7
z(Q*I2{5b1qOpikX1Lq>3%41*K6G-ZQm98x7TuN>a#)20PX)(6~KYH+BIg(~XmXqU)
ztXc^48hbun`sNR|5owZaIQ_gnzsc$S)wJvc`|5?eJ5+`K`R8&h-f}1p2r!q1&33AU
zQpsh0dvI(FuJM;|wKhpDixLdTcPr-1*-vmh?dy9*8Wpbtbh7EiPZCKV+_@J(=7wVT
zbcI2cDxq2-EXW(U?%jc$uL;|56@-Gu>^TcE)v3`ZHum4scO`#{DnAp!Zx}JCim0eo
ztqjSR&Bom-a|j7nx>lPrrGq(z(TXE7l6^0HG4^0|M~{*+Pw1~)K-I4qMrz;~1xpZ!
z&qXC$_EeG8TIPmt?06YSd$bX_Y7L(}e#<@fmpvE}FZqt~a{5f4)s6#oZ6*mbn+uvQ
zhP@8YTUjRe*HDdJW}gqf?K2FOWeyyjVU4Nd?eNXf>#oif7rCr2jtM*dxa=8qQAR?G
zsE)30^T4=22t?P6i;yE8b7fukO;^{QBH(gHhf-!x;^k3NZIL;YWD+fYFKorV?>aJz
zt#yWVHu=q;Nd{0<qf*1)id~x{U%570esk9B+_A&fEn7kuaFLtsq#})zNHXWM`SLLL
z3x@Ds$ra23RCwFi97{E6Nt41vyr;(Q4-sTHDPj}0TYiE)C4k}YP!Q(^9iE`1v}VN$
zFKSgDxRB1iV-uEFck=lLJYf$&lr_k6-kt%B7VtfO>eMj-6iXFd3=nY@MZMqby*D@n
zR~Q)-zoFjhx#?9iAUruu=bPeL@o67e#6Bvmt%N?$787IW!IF861c1@IH|;5k<=A(B
zbRD)w8&|0;A@!tO|JLFvd+0LG0Y*ahtCaj@r3V9;zbB1)Fr98#Hmg@hF9YJm8HY&&
zk5g<0KZ-th(Q%At@y9$^4Z!)v34OZpzF}A~Rz3UuIOTHGz@29f-z_<Lw`4g`Y|Ol)
zpg~qq<7AU;mM6h$ohkk}BYMa1W#pMB%0?z8N_VFO4q36~<&Np-Iv>_Jtev@7X&7{l
z!q~Bk!~82sYbU0Itnk0%i$0hg9u@j|kH3#k5Q`knMIA<lx65qJNnl()R+SZWn0>or
z`gyeb!Qv?}$nogOlhsJu4CAaKLBhGl1xCJ#b+=6{TC{K>pYi0=Z9Y%8tui$Yq?=Gd
z+bMgfZ|+>{DQVg*lT4D<AyLo9%V8IS@ds9Mk4>*CDkv;~8p<K#M{@c2oY6Y&A&p&Y
z2+;{3#-~73t~?(Xo-_XV{ZW8mT>B8RJ>w#8e*q7KNTSPQi`eNJtu|#5E6a9Hkl8S+
z-12tP>^)v-dwkBGRZL9&Y;<MMEQiIl7vj3KZS#@$OzY=6d?ogvdMk~}Ex#AZB@`m;
zG4M5ZILBP+_`6fSFe-4iRO}?fM2ASmz=OEMBi8)6&cNde{Yz%(efl_EuYAFYc~323
z{i6Hod1Tz&*TpblS11mxN-4$B+9Q53adGB)g^}?udf+U$z}EJf5Inb3xV2uv@Ib*}
zU`a9H|I>@JH=q>@YxqiRsiHs5qGt*RUsy#TBCbTDvzld$n$IFH>H<swV@e}=>aHiB
zM{byL>)`Vq>@m`mo;TUwEw^Rql~s8rdn4uFF)iNnWwnXzLxu8cK$^o5#+<WRu~tQf
z+$DQlFkpn0)y>OYiK|YO1uCd&?ex0-)^)MUy~_Zbf2IHsg*RwW3TDD3TVH(1SWpNc
z7uFyq3FUrLzx!j=FZY?%Z^X|og@vnU|BzU<zHjY4i`elaM_$(3KH+{?ShE)U?nAG_
z+b$n=X_Q(Ogr^rfB+gc<&`K;)3eU;k5$D289AD`<$0|Zqn!C_joI(_G?{G56WrQ>Q
z`bD3BfM^6%CV*xa#DQln<dId}+6iy5!81evmBoOIyurnu_qyJAs!LpV+_r|*v~Yfc
zZ(pqG{hwSYndPtwS!fbhb?#)(^#y|@l6yBG-HH&lCUB;{AR4ZQ3SC72kvGSj%<#`&
z_5*gG;*Eu+?W8vw3ZLqIH~6K`)i<j?+4s1c$t4X>p4a!hyPU!1Q|Hcwwte+%Gh1;W
z-}LGO)2sKk?vJ+c8f((5eXQ70%dRc%RBy|w2{)71gsVn%PpWNgwN5X%$QU&I+eVs(
zZ1J#B+i-R)@Y$7Px`|EFAb0%w?c?V3mS#-Mi;QWH)JyMlb)pr~kT!2EXhjY!;X><y
zRefg7n>ll4Vy;m*hsS$-Tjfuu$ySTPCXVmwYaOc35#{(<By0I^4y!60kYTqwDWl+a
zK<U-q!C#9U8r**0Q`9i6qHWKI12&d>iK!~+4L#&k7s2%9L9B*T;V$xaZT+^r1=sc*
z2Fb<3z%@?|9NTTa>37{~>rKb|TEFb^Wd8t*zq7sf_`In|=}*ZshP{G6o=d<etNTVz
z5PZv}*#aoRi5v6o{Jc1ZI+R!xP|w8rN&YXRt$Hg#^o*jm`<^>D&ND3t2R5_AKkuYx
zDL354_dSsq`<{U#vb6!V>h9`21h1{zwk--XN=q-ZU7#NL08=)vyCosDZvvl+yxwjM
zmmm8$DH0c<gtpe^9Q0Ni0oCLd2CamtToZe3%k`J1E?$g6^t^tU=hz78x6Z!jv|DIB
z^~vPS{9Ht2Xi&PDcIIjK%gHSb1^TU-MWh3c=4$aKv|06>jPx`&N=rWtdkc?7ky?1R
z(v`Ew4o%-(Q}7dtM(kJ()ED8fk!#jS0YR{N7wQg_tShPe7Wmm?Tl(t->li<%6mtm#
zw5Q#}MZ94}{HAQ08*}#u$%qf&3027RbC{XH!@zo6pKg`BVzygLp4&T`uaKP&4~;ng
z^bgn*bM(SYES%K*_O_pWW#$~>=%)CEo#iI)DG5$EqZeYB*|Sqe{qdi=bnGBv5cuAs
zj0~;qJH-J(DYeZcb2mVMVT_p<r*drQ=X(dl`apk9_3b8&dVEoJH`UIXHVhDBu+^(`
zzeufa|DkX_fWs>81P&T9R3fQ;XnM60N{^6{{DldeN{aljkdRDFE!s;e!ku!)qiFrS
zC}Vr_!0N=4UHm*>pBrs|!N9owQLX;BV}YAw{ATWsqKy6n__SJl#k@L<9_-)!?wh+O
z;BLv3<Jl!Rl|GGLeyGmgZ`$Fzn@ql|y#nH^oBgrw+Z(H+y+Uo<x^~=eQKy|0lF(La
z-=mXf-`a+!(@1xNzH3i2X-pwc-8Vh@sP$uUN{UkA`>8cAuFgY;luH9OTF*A#$O(-_
z45f%T7L;_;Qxs$tG05lKc39f0N2iYdw~wFMN<(LO%Q>F9g<^<Y37K_;%K4=61H3d<
zObD{73>VG@)29WBme(8Z_Wh~=|GDz%+~UcX9ibVo@$WXdHFN-e>ur`TTjnc2bhcAW
zuEVd7$Gxm+&<RME${Z2^v(UZMV>_{VtCJfFLG4xOxeY(|&bwyn&6P=;T5GxJ`AyVc
zgNzI@CNi>$o0(rS-^juu%+X;lO^4CIc#*Ur1#cP|K6$sx&yH)$*RS&(WD&)|+A5=C
zq!ejjZ$2lkV5sln?p=HbbvKbrYA<lGZC_aI)|L$E5k<vD5#iI`ExdBn6{I1ew|NOm
z*4miQ4z)E8O^E88xsG^0evqF>N=|U!zQa6qCntaLcKu$f>o{xJ`iTQM`xC~F)fHo~
zC=>&Sf-WxK>fJ_NZGn1W+a@cbF{ddEVotWKa=FiI*|FPO!^f}l7k8gTZok4E3roDS
z33fX7`%e52wIOaPMN=Nu*>66Zs!Al6b~-h@5A8bKc{%qYKO;|sxg?e&LCXcvu19EU
zpv4@|snM3LY6k{ws^u`5vkl9~&Skz5b>nK<^j$RW3)D6ESf}rsuHe35v;>#8Uv$(_
z@?A`7b3$cm&v=A|DC84IO-XNn)W1-tjM7U~Kxw5}qWCL)wvJN$jC@Zym5|%Yt0#SE
zsLopklWqBVLB2$y?OcWVA2CRi%(j$4<4n{tPZU(djFBZMnD|C1Ed?o|=(V$Y#jSrc
z+TpiCa9*(=hiW>nGl^bH)FtbS3K~b~K-(}ruUZ@h`cr@PRcHfGtzLg`1E&}TZRpMK
zKW(1aePE4(InDC+qKypDLN6`y7V6+Gv6k#FOg)&yc?Xi83#X^}p`&EqnCi^S2R<a#
zEqxlGwGpUSCHlurGKWMPA|^q_BGd!d>!?2~KiqNh-0?&tp+kX14F*WI>(Go1v;D_c
z#$+cCha)_eUi5Xjuwtq>d)3|GUo5!D-g&kX0dX3|)dDU{%^~G0d8UmHwY4ljTl%P?
zF4V29)dJuywjy%<b)6Z7HV2rWa$R%&_L~p%XQFol9$8-4St1!p_xU?xBNog@5z*G-
z^5a-cZS{q&QHo$A|LSJl!pNYFgU^_0FQj-PCsm;?QjXR+hocqBzIOs`GoCJym_jw>
z(FobYD_8&SYvP?(odeFO0@T9BFiyZpGid0jDqMz*MSsUvswdVV@Yxvc;Kp96hs3On
zsOX&cHS*n^RqQq&lE7rWCX}DM*fzOMO)SzbkHzyO5=pi&#bny?k|=f<)$kI7W_5SW
zM6ZUIl|2`!h<>rEIK|Cq@;IH?o+Gzje+haP4Es+K#C*!WDqGLj$o6PcL(SltPwcE8
z1Qab27<G!g9$=fOY+gCn>DL`FQn99iaDpp;8oqo(<Qij}+np28_k3Xk8_XR2$K@V(
zLuyT}Z@c-gAbice3>PjpZJWEO8lm&nW2af}0*0OmuKR^iQa@>9>geNh$_MNQ%8_kv
zXbfX5ekavdQQDn)T()w}nlnsid_m`a@4oC1wkzNhy~JvIn49DcMnM*O)3}z}My8_A
zB0NwK>C8NJWbv|PGxBsbs7(t8xc(>euPc?co`mpsTEB^G65q!I%8rbhO!q*eOB&1S
zrLcpN=L<GX<|LwWOPA(`9MN2oL>76I#Koll4H;_iySdZ-uCzA{3o73?JeoZ1B7<um
zj%EE_$BZ90&a0s2go9dNUq3$!lR@&sJd0MA+0J4mcb<!bR0JH;E75XLT}yp3wZ~|s
zy6LS}ma_5TY!eP~44BFV#;@{SEc0VT@DM_B1$p_^iIczXJ}(cBQMcpu@@oY(fE(Q$
z6D#h>0<qB8V1=WIWgh7oD#(XkL~hW1YInmh%A?shLY6N`41QFfA@td_;8UFTcc8-f
z0C&=JxTcZtN(85f<jmFGLX~QUO3;tgLg_qU#_%`GUoUuI?C^4{_xOx$JJ_KYq}Mxu
zKQIca3Z$m*r}A)CT@7@L$Sw7_H?;>p$-Ti;6l7m1Qw)A8=H8+ssk_Boe+#f%ChD-5
z9WSSH>J}(J`&2qg+HD-aT{Lj4y8)J42%nlPq(Kti$blUfj>f7<%$M--z<%CDM5pi#
zBwje}>=OT*gPtF6cSYow_Bz(cVh7VchF47w0^f&ouyxe)KVACk_{q<EW{LiYu2djN
zu*ORHBl$Lc9s@!V=A+O?ub_t#$JBta$C^KRmp*CaW==@BIQ2A|IW$^9MEiwdMs^6M
zbky&By39*2TCyaYEMd<vf5-G<`C7FO679#rf0cXh?j<en$xCG#eW}V>ETr55wo2y0
z0dcTZEV4t)4iA`4Z8Q*u{1&7|;Lt7-Nw?9ot;g^GT*Ecnl6`W-SbgL}E5UP^fe#$I
z^o5NVDXaZw(wfCe9~SlBBUXi<5!<fL|6#5Lo=8pLmjk+W+J#<*slZ2Nwh!O!q8SR0
ziCG<K7`u{6-&r$&gJ)2q@(pB+!Z_^a&2`9(y|=U)ty>pDo31xifcK|Cl4#3co+$pf
z*4#XlDmyIJm{<aBQP&n`$yLs=7CDQipN}+jcTL&2^vd;D%GmRUf!Kh@O$%6MXgD5u
z64=3Y%Yck4&<1>*+?lDsO&sJ0K-C#s;|I=?W2(|)9&)_x_bYAV!HCyv*1{S~QCWFK
zufsv9G6&)eJd|IUOc6((2tENvEzN*soZxNlhc+pytH0clHl+urZQSV5L5A+CmToqQ
z7RNxz!vX{EysxmkFXKEM%<Ccjw|GBq#`omwZfK9->Q94cnFEc@&YGYSvWGUM8gFpF
z-LiaxJLmM;x{2FnVh6)Nx8Gl9WE8;#X>yw*5~2^(pw6D4t{$Ji=gXd1Kup^`(q7j|
zoua#@&H1wiv5CRz)u1t?w~*lA`K-4$i$^sTshk@V9iA$aRAZ!^o9ns#R^H=Bk3>Up
z_N>{fHN5e3sYUY>$4Oud6c$tsq8rOjT4SFcVs0{F!USoif8Q|2nkOzU<w<ME$Gj65
z8C}4G5*rmqlkVxVSr_wuRqoPgoueL9m2`&22&J1SLmxX%`s&p4Twax&r@`ikkdWor
zjwtAnW(}4uJ+W_HE%EQ+$Hr<P!Fs=D3ypzkpLeEL)S+3|qm$CC*s0Y2TXXygJFh3j
z{rdF-iTVSPBK8{h3R@SJeSBJ?R#Eru?Jf;_E~>X1R$*w^`oTu+ZS9^@G_N#uk4kK^
zo&38Sw3^hfGg{j^EKI%BNc}=a#VW=PDT3!tjW)P2ZLVr`=YW1y-)HllCeeBC3&;Uo
zo=RiOQm>v$*_&*_(>qt-wvC5l6Xh8Bhx%y$Iz#=I7n$FaNjQ&~R^_wjJR@i=BqM-i
z=Fhs><g6KQq{P3H^awc*OKZ>ioJuQ&$UEn#aF<=LciQJ<MupF6^@T2)0hm(*xyqlb
z@n65?lT-4xUuC{)y}0qXY-;b2{n~%+NTaR@30X)D#RW`;?)$a1*=_YFuJVA9w}T^=
z+oii_wZBD1MR=}*zBmSc%6x`tI*_PKS8h*T$IoVQO5g=m8#d3K`|FTiy0J1vR&laH
zX{BDP)xElG{?gpScWi5Mb&J*PnOf<P_|)OoA5H)4W2l-6N_0zS3NN<7@>=gynI*T9
zOsr#Nm`RelD}UAG2(eGg?ZU=vG}yc=-l+bV@(YKYj@F-AaPxY)F2g-*io1n!myO5~
zF6C>oXH`?z)hGY#Uh+gHW5*TFK40hB%N^Uax|Gheo7IJF-eR!1$4_I7t>QMY84qm=
ziHXT}=j*<WkU8+N{&idIeVL?5UoS>XNVTiq{3-r+C*_+hx8q0k>2}s`ZjN%PZTyoP
z2O0y^qYbS3-1$)=ky$T%0bBw)H?G)R9_v@`{yJ%(bIqYAvX{^1{aUAIsobS|^KJDh
z079uU;UB7-o=9)2kM^rGv}qP=p<en#MnRn3kMW0|NI6R6!c75rtiXZt*f8aUBCBbs
zVRnX}DlVOIHZYA@!mb73<N}uxbx*t7%CfT#t)4wF?b%KxnV|coUhDx7Hnv{6WBazu
z+o^O`SLcn|po#rEH(#x4m@(bWGq%*99ZH;3?^UIduBzbXnY}Y=<hQ+l{#9MO2cF=U
zfqSGH=?AB}H993%JSiSLb-Rj!TkjKP&{<9yQv3UR%KtvA@o~!E_Sy{oRe8*qF@8Hc
z*ZChxm(J0DSS&q1ZIo9EUk>Q^`d@o_=vPew{%d3B$PD&L>6*5B@x7l#`Vc>Tf)qb>
z{QH;M?ty=G|NEB`<zpQu|MSaxia#a&^UHtz=4-k$CG$-j930l=*E=QqO}F&pp5E*m
zmK%y3D=sVo3NENw!(NP76Ar$1+TVX7aaw!CB|>)HbUlK&&+~T?4jVpI_*vEMnZ*|2
zuX4K<*A!rU(BrD2rH+pA!0FWkO5^j@?Kbe@TjC0+IG;4nlMUJ=114N<$@N7w4te=Y
zr$mLt<qQmBWzj5nWT;=Sp^{t*lmwO&N+HB#@hn=s$JO<VRvs{ocy}XAc}-E?ZJEK(
zJjFw2gInr*)EYG>{2_B7BQ(-ldMpbl%_{DE983XGS0!XP$BAR*(_pV3htt;**tk<n
zrJ|!78=fQQ&EArx6c<<4uCq^v-P~T!$urN~v*STh?pc}5<|g=;ooz)C{%}eU$bbdn
zCewT6u?eXw6w706p2|)#H^C$=5<+o@=2#>3E^k(;6)%Ix2&NK~*GpiI?zErZ_|Tnc
zC)a7D=j{F<eU}!`6uB;*oM+%m^k)<Wa|PjI3)gQm)8aqfyNQ0>1x4d{Ma4|vqEuF1
zp1QU5)yneNrgD|rj+P(=P6&cnOch)pj9tH#O)@o^nb$Ba+8|vPZOd~Zm}5Ls3OL$S
zW9=4x=2zOP)G0j#l<6JZUFoJamCepmqvy)k;GqSKeo7ulz`bQSooTtO0580=W2I|>
zdNm&P9Z7ky@(Y7mdfsphN;hBss;K+hLV=pcx21Uu9#m$wuc;YC;tN%yo;UGX7;}^z
z>7&OQtJLLE?#!o$bhn*<BpCMXd~+L0>PS2qf`;@R^$rkoMq}IS+Xw8k77SXTK59NW
zQSf*QdH@P5PK=3J7x+&RBQfpY+2o&%d}PIc*ETvUUemIPpM5e5_5(e&|CpYdo!{Y7
z8P8H~rbC2$C@jy46J5M`<r#Kub_#F4+A%m6Mr-69j46+z?tz7HkL53V-|E^&%c=2$
z<%c#Nh_1TarPfx)&vTM&6LKD@Pp;j@iY^kkN}bz1PfeGmvuRO+=2w@SZ%8ftw51qm
zTfCUGC?Ph!4TyDcnc3XOrZR{6){OY)lSz7}rRDtd>wg`{|IO*$8}ozq7X%;{kWh5F
zKE$W&Al=@^F7-Kb8oNSvOy7{5-)p<k*99TX@lcjn;l=`Nhq2@#mo8~juLyXYdFUXN
zvxWi9%%ED}`(l4X3pbZ2Te-QHU$tF+aoD}(K{tCX5~{~=ty?CZ)Uy%7Sf*}ws_R>s
zLzk)4UBHm@uWdapZto#&hHgx8C1@D>6ja>wv>CO%v!m2`w6%xz<@eQpyp8<1RLlE;
z@!%63t^M;8cSjF%`LcqOSa*YK)CEK$O7V+(-uwu}sgqP=lX`IDZwUd$O1@*3`tQlD
z5qyCV#4^j>Q)Y@b()rMxcc%52=bDm|o;RgO(5u=@RwXjq^^4j+^;)d}X>nK4e&)Dt
zf3KSk#25;m`8}c_{MI13<E5H+j`|OH8(Hg-q<ZUMQ1syQaf{sB7LM9&+3AO!X=S#u
z5NXQ5P99=-Nr#x#w_7K7APN}>(iE?N0BdZDX1=jJJWc8Qs39%yZR6e5?aE~sh&;IS
z^IN^U6?L+d0o)fXMwFi9?L{Xdb#8^qb4S*=nLCj#znGsr(x^0IT=U7R!)3Fo#PtWV
zko&0=CR{vx%%V}bVf1$Du2g?@J9j~TG6EY?`>i5iu=rh|o#(Ta(p-M)r#|7qhSI^T
z@^PJwd8gy#y8N@ms(VT^xn;Agu~`)=S)W1oze6HWM>1H&XgVoyp(<LkzLqonca{t)
zsjPs!nEm;|Xt^oB?4!G&)Z79m3pXS1QT!}?<liyo7gpDAS@`L3;|lN0PM5phTejJ&
zZ=|aUVNTv0!%yI;kXmK;mQRT|8`fiYWd@-q2qyof+ZX1S{cBHTRaQ5(?}?wqp4h$B
zU%AIT##?-X<?=2fjZ47wO2Hs%a=*|UcV!2;@-DaDdM|#GH<_y^*C(2tEmuqT+q`7b
zikPOCQ+iA?PFU^zq#?`JrchVS>-w)eRn@esxkENvga&5iy^Wl-aj{nGs@ERVXd1wY
zPs#&ogr=rp2DRM#3ksn55$*eJet1B5)UvxytFxXDNn3B{9;$F`jg}j3-efj331<=<
zYUHjtWX<~jdlx&eST#Egq4ZXgMI+ZX3XrRv|LH4jUuYh7d;96%4K26Sr4StP67!_X
zi27@EbPg#ceVS{G$6rEa;BnS}?!aY9$L_USS88r~ud5T&^o_X{FL}W+*`L^y>bm32
z|7S_d#W7nG`=wN?^D8wfEBThsso{E4ylQDanp<LTInI0OpjLWka$r^lz$tzGY5Cb0
zaN27+fP$50q<Id{gCeMZI5t3W^5kU*w4&{X%KLASZ*ZH}IC8VrRQ?<VY!kDH@_K-?
zTQ7KSeV;bwM02;XwrfttrOx=A`*q;ETUHBSjaP2nZ1B&rQwgsu+BhSuX$-h_z3Zy!
zzyH172?0Os-^XxxBn5U^MMXH2w5n$MyIocIF02BjJ?-T+5+yFTLs-7Nwt!Kr2r%1F
z$GUQJ_;imGolA<^JlYTZ5gje-{1`p&?C6bJqG|jp!{U_wYjM0N6x;V(?mvsOId1KT
z&#&`Gs)jnnzBptw^xC+;543m74gkya3*n>(bLjen(i1Z+Tz2qrG7=&MVl{ooSpnE`
zU5qYQUSuPa$f51mN$dadPJG+o#P&ns;p&QBOhJ!>9a+zbc#*nuc**?qpI9E9S8~Op
zit_3oZcs>diy-T}ZSz-SxIsxz1CA`x6Spnj)!({UJs8_d-@JX#cvWq`J4MB<)mb37
zC@nLgkP9ngkAQS-8Mg3%H1P7yK-KiBd=dJN?9vkIa1GnHM^LTqB7v%g^-OiGOZ)fx
zzs$S*K)nA%Gv5EFq@80B#CY$bC$k7pS}jx`dMha?M(4Kgl~Gw4x8)s+66}EqQ~p?T
z?%CUA(L%tW_Y9rf`Rt`h$II2iEr(BS7(Mz~%XcIt3%v=R=T8+*X*fUQ-&^D1tzHIV
zYn)F!<NE!|UxT-U&_C&%;sz8dH~c#K2R(tPkNW$4?-XSi_`G#(X)@=yj{Nqv3i8!^
ze5r2f0z3bl8|~}-(6yntMPv?nb0)a-EskF;PTKjOAmA038n>|CUa@Y?(c0p||9b9(
zwSa)hdQebV>l=N2eWgesy@R>3Wt8dPJHWqv2mBkdp6e|?e6~}zzoInpMdzxqhNTfv
zE)QOKx-V>~Luo&WI+zO+HKM2PRdR1B_xx9aSr8UI@3crT>oT8tw)H18W24I9j=@VP
znK5_(C+}dP_~ToC{^l&OSdZhsM*lY+J?!(1^Pi2fH>x$bP0=$+aZRS4QXxe-jeSUa
zXn51+axWlR%iFeb1(Cww5DP8u;Kh@ZSCg@Kyx&&Zrs~j^r`m8np?%5H<5R*gIaw%}
z{f!l+GR;?mqFOTkvpoi$Xj%43ZssVj#M~Lyr)Ao%%A44&J-$gsXSIp!idalL^e<?{
zrZy7`!^AMcnw2Y`t#1@+DM05K@xhqnK6#@kyCSf4ZgN}i;hGWRpsXw#yTLxvxPh%1
zPS7d;NF?@C+i7~Cw&6q322ZWlo1!gn9#VBhxj7~3>hNRji))kKvU2O2z4)Ol3QLmP
z34I<(xY4Bx$)!Br(mJeLr~XSigb$pKIdTL#+SR;OgVui%<A#t%El25&A1jnEjdTpZ
z_tjIyHt*1N-7<r|_W!`tvG^nkI+-4Djpwl0Do_U=GRlE^8&_B?DkyR9p3-f**5E@c
zssl@JH-`hLX5!CxNh~q)PYs*b`wFe%1W0&XDPbWZtF-}|E^;oaoBbzcH*588&+R22
zDi~Z_9QNJtS-kfa>ZY9OGu9?2D+e^&{cD%nG}zDk%q|_?^KDemyR!e?oUJ?XpX&Ah
zsMP(x{F}FWI#5fWx&N_IN}doUr)Kl}X+%hU<nya($b^oH(VnUqDc?NsEH_jia>DE|
zo##F^E&vFoyE4NpJpbm-Afb4HrUEY17qo&7M(?Qg6ww;wU+$<KIq}dY86-@SOk3f%
z%HA$fAMWerby*B8K=JoIf`m>I5(NkWe_ikN9w$W=hJn{lJTpbNGi5YQ@V*d59QWW}
zLV4TyGyC%gE-mFPbFh#L0yND&LLWRL7@)o$U!ys6lv91TeQIt278%b_4HP)>d~aQ?
zV1qjD*X}<WoK-tXPTBM~VR}-7gq>5zLFM-&N`jH~ZLf;2LTf(<M-G&kB*}JHkdsLI
zt_2-sMjA*}&%2fDctmfp;)M7HY6WOR!qMTkdC3dOmtf*gjqbRw3er$ALN7nXErkQY
zSdvY8a6Nv`%jvp_=U-CVnp1kZ3T2c&=C|KlP#*TQf=iLqj`dcTW{P>kI+Q?l0UlOe
zC3g<;MYRQ$xw{3QC@%;EW-EUe!y>Ah;-X(6G%Krj5{y|9s&)^jWca<h3b0F^SURkI
zUw?k*Z{W?a+c!JN9=W1klHa=1Z|!@>s7^9Nsa^Dhtk0hQ$D)xoK@=`53Z630ur$c3
zobaP{;mxe?@uYQ+-mHgYr-HY7xO_4bmgO19u*`PNl=M*&UE3J4Ji^eNq;C1>#(@K$
zG=GG<gnV4)9m74{;J+$W_AoiBWI?@W%=0N;dH(X=13qM*$)mikH?_D5^`>BT<6~an
zkdSIWTB5R8ymSs~5D}idcQG_jL~rjM3^!SIaFEGK`KRZm>9Ps+g7SNJ>L_|b-r+P_
zHRihyj<Mkb-}6{r5doZ7?v#ue6!6YCVOJK$8VY)zvm4HuAe^7F<41|g1xrCP1`h2c
zx#VaLcR>C81;0;odOh@(3DkE2U#GbgZ)=vFSBTt4JbxjjrZCUPQb^E7e~O;f?A_dR
z*Rj(TG%LvAp-pGwxSCQh{7$%@P?q0`9Vf|@9~e~iq_gC&Hd-M3#hx7=?rQcbH_Tvy
zAa5axds<C~FojXc#%USUH<+co-k`+m`^m3XmvVM4ELIny4?f0o((3vClO@0XuozQL
z3o_FeTw)9oskyd#jQ;>E*-}0uC5sngSh^k>a+TmiwS5kb-)@-`cDl#ilbd9C?4&WF
z3lV0Zz|o=fnb>taL%%XlR#Wm@C-#kYWb0?&l-12H-#(Z1w7Og4AZ5BoXxlKN7MVr#
zYnOlOD>jN!e|4z=d>Pc6SaPvgl+AvPf*{qm7%aK7UmP%1(R<>AGD1)6IT&385OADa
z+7adyTVfmkWpdl+$>M-gWyEm396A#99m*V+5x0IDwLQ{PFo|%?XI96XL>F20USA-|
zT<w3myF^V0R#E6F=m{xUZ>{gYM9nxu+czfNULDX!6zM1!g@QPr(M&<3Nq$?RXq($a
zb1}t@ie#j8kYw(DPTwUtDmOHW0|pL!wC(yk<gwn2R1#u%Ci0|f1nhaNUF?mkZf&lJ
zeKv?b0GwTD55?sbHWPU5Oz(qy+GwV{ObAxq^8d(g^sV|y<{lMq;wQudpbTf^CmF^p
zB<r7A{q1y|P+NP)P*X9;zgbn&4G;+M+W<A2NZyF#1@q$d<1n$_mqR(2+$z05NGYhk
zXF`;&na~AApa6q}0iD>nuUfI+IJntPa!F6*rC$Uy#f8da^_p5fx<#W^zgT}3<;gqn
zgfsF2S1R}0K8LvfoU^@E0R8FT61X5F(6uqKSXE&cZ+!3Q^0yJ*LL7EoX2_OUL+R}5
zIP2nD;<*y;{1SeY^jRsM$VqwN1#`BGcQB0DG6WFL9-inQF5=@dT2JT(-`d7k?SN~$
zSU0@-EOFtI=AF~myX34XY{AfIx463FjncktV!1d|l|m3npQE71szs?f#un5F7EEy8
zglvXIj!r(nryEf?_ri50mUQs+9hoB25Gc#uzJ)^JuMpHnB4!W;76|hqC)Cv(kh`}$
zB$+Z3v-sofKg!3Z{fBb<-@sD;mvi<10KER+EB{aUFT1US-qQg$(yuQQCsv}jVCnp*
Jxxf7SzX08i;GqBj

diff --git a/reference/link.html b/reference/link.html
index 9bd0f3c8..06f8f0bb 100644
--- a/reference/link.html
+++ b/reference/link.html
@@ -174,11 +174,11 @@ <h2 id="ref-examples">Examples<a class="anchor" aria-label="anchor" href="#ref-e
 <span class="r-in"><span><span class="fu">link</span><span class="op">(</span><span class="va">mod</span>, parameter <span class="op">=</span> <span class="st">"scale"</span><span class="op">)</span></span></span>
 <span class="r-out co"><span class="r-pr">#&gt;</span> function (mu) </span>
 <span class="r-out co"><span class="r-pr">#&gt;</span> log(1/mu - 0.01)</span>
-<span class="r-out co"><span class="r-pr">#&gt;</span> &lt;environment: 0x56401dd37dd0&gt;</span>
+<span class="r-out co"><span class="r-pr">#&gt;</span> &lt;environment: 0x563ac474a800&gt;</span>
 <span class="r-in"><span><span class="fu">inv_link</span><span class="op">(</span><span class="va">mod</span>, parameter <span class="op">=</span> <span class="st">"scale"</span><span class="op">)</span></span></span>
 <span class="r-out co"><span class="r-pr">#&gt;</span> function (eta) </span>
 <span class="r-out co"><span class="r-pr">#&gt;</span> 1/(exp(eta) + 0.01)</span>
-<span class="r-out co"><span class="r-pr">#&gt;</span> &lt;environment: 0x56401dd37dd0&gt;</span>
+<span class="r-out co"><span class="r-pr">#&gt;</span> &lt;environment: 0x563ac474a800&gt;</span>
 <span class="r-in"><span></span></span>
 <span class="r-in"><span><span class="co">## Works with `family` objects too</span></span></span>
 <span class="r-in"><span><span class="fu">link</span><span class="op">(</span><span class="fu">shash</span><span class="op">(</span><span class="op">)</span>, parameter <span class="op">=</span> <span class="st">"skewness"</span><span class="op">)</span></span></span>
diff --git a/search.json b/search.json
index 8b94bff7..f828b857 100644
--- a/search.json
+++ b/search.json
@@ -1 +1 @@
-[{"path":[]},{"path":"https://gavinsimpson.github.io/gratia/CODE_OF_CONDUCT.html","id":"our-pledge","dir":"","previous_headings":"","what":"Our Pledge","title":"Contributor Covenant Code of Conduct","text":"members, contributors, leaders pledge make participation community harassment-free experience everyone, regardless age, body size, visible invisible disability, ethnicity, sex characteristics, gender identity expression, level experience, education, socio-economic status, nationality, personal appearance, race, religion, sexual identity orientation. pledge act interact ways contribute open, welcoming, diverse, inclusive, healthy community.","code":""},{"path":"https://gavinsimpson.github.io/gratia/CODE_OF_CONDUCT.html","id":"our-standards","dir":"","previous_headings":"","what":"Our Standards","title":"Contributor Covenant Code of Conduct","text":"Examples behavior contributes positive environment community include: Demonstrating empathy kindness toward people respectful differing opinions, viewpoints, experiences Giving gracefully accepting constructive feedback Accepting responsibility apologizing affected mistakes, learning experience Focusing best just us individuals, overall community Examples unacceptable behavior include: use sexualized language imagery, sexual attention advances kind Trolling, insulting derogatory comments, personal political attacks Public private harassment Publishing others’ private information, physical email address, without explicit permission conduct reasonably considered inappropriate professional setting","code":""},{"path":"https://gavinsimpson.github.io/gratia/CODE_OF_CONDUCT.html","id":"enforcement-responsibilities","dir":"","previous_headings":"","what":"Enforcement Responsibilities","title":"Contributor Covenant Code of Conduct","text":"Community leaders responsible clarifying enforcing standards acceptable behavior take appropriate fair corrective action response behavior deem inappropriate, threatening, offensive, harmful. Community leaders right responsibility remove, edit, reject comments, commits, code, wiki edits, issues, contributions aligned Code Conduct, communicate reasons moderation decisions appropriate.","code":""},{"path":"https://gavinsimpson.github.io/gratia/CODE_OF_CONDUCT.html","id":"scope","dir":"","previous_headings":"","what":"Scope","title":"Contributor Covenant Code of Conduct","text":"Code Conduct applies within community spaces, also applies individual officially representing community public spaces. Examples representing community include using official e-mail address, posting via official social media account, acting appointed representative online offline event.","code":""},{"path":"https://gavinsimpson.github.io/gratia/CODE_OF_CONDUCT.html","id":"enforcement","dir":"","previous_headings":"","what":"Enforcement","title":"Contributor Covenant Code of Conduct","text":"Instances abusive, harassing, otherwise unacceptable behavior may reported package maintainer, Gavin Simpson (see email address CRAN package page) . complaints reviewed investigated promptly fairly. community leaders obligated respect privacy security reporter incident.","code":""},{"path":"https://gavinsimpson.github.io/gratia/CODE_OF_CONDUCT.html","id":"enforcement-guidelines","dir":"","previous_headings":"","what":"Enforcement Guidelines","title":"Contributor Covenant Code of Conduct","text":"Community leaders follow Community Impact Guidelines determining consequences action deem violation Code Conduct:","code":""},{"path":"https://gavinsimpson.github.io/gratia/CODE_OF_CONDUCT.html","id":"id_1-correction","dir":"","previous_headings":"Enforcement Guidelines","what":"1. Correction","title":"Contributor Covenant Code of Conduct","text":"Community Impact: Use inappropriate language behavior deemed unprofessional unwelcome community. Consequence: private, written warning community leaders, providing clarity around nature violation explanation behavior inappropriate. public apology may requested.","code":""},{"path":"https://gavinsimpson.github.io/gratia/CODE_OF_CONDUCT.html","id":"id_2-warning","dir":"","previous_headings":"Enforcement Guidelines","what":"2. Warning","title":"Contributor Covenant Code of Conduct","text":"Community Impact: violation single incident series actions. Consequence: warning consequences continued behavior. interaction people involved, including unsolicited interaction enforcing Code Conduct, specified period time. includes avoiding interactions community spaces well external channels like social media. Violating terms may lead temporary permanent ban.","code":""},{"path":"https://gavinsimpson.github.io/gratia/CODE_OF_CONDUCT.html","id":"id_3-temporary-ban","dir":"","previous_headings":"Enforcement Guidelines","what":"3. Temporary Ban","title":"Contributor Covenant Code of Conduct","text":"Community Impact: serious violation community standards, including sustained inappropriate behavior. Consequence: temporary ban sort interaction public communication community specified period time. public private interaction people involved, including unsolicited interaction enforcing Code Conduct, allowed period. Violating terms may lead permanent ban.","code":""},{"path":"https://gavinsimpson.github.io/gratia/CODE_OF_CONDUCT.html","id":"id_4-permanent-ban","dir":"","previous_headings":"Enforcement Guidelines","what":"4. Permanent Ban","title":"Contributor Covenant Code of Conduct","text":"Community Impact: Demonstrating pattern violation community standards, including sustained inappropriate behavior, harassment individual, aggression toward disparagement classes individuals. Consequence: permanent ban sort public interaction within community.","code":""},{"path":"https://gavinsimpson.github.io/gratia/CODE_OF_CONDUCT.html","id":"attribution","dir":"","previous_headings":"","what":"Attribution","title":"Contributor Covenant Code of Conduct","text":"Code Conduct adapted Contributor Covenant, version 2.0, available https://www.contributor-covenant.org/version/2/0/code_of_conduct.html. Community Impact Guidelines inspired Mozilla’s code conduct enforcement ladder. answers common questions code conduct, see FAQ https://www.contributor-covenant.org/faq. Translations available https://www.contributor-covenant.org/translations.","code":""},{"path":[]},{"path":"https://gavinsimpson.github.io/gratia/CONTRIBUTING.html","id":"something-isnt-working-right-or-generating-an-error","dir":"","previous_headings":"","what":"Something isn’t working right or generating an error","title":"Contributing","text":"something isn’t working, either might expect/want contrary documentation, probably bug missing feature. ’re getting error running {gratia}, ’s also likely bug, opportunity catch use-case wasn’t expecting. First, check issue hasn’t already fixed -development version Github. Install current development version {gratia} R-universe using issue remains, please file issue via Issues page. ’s OK report issue even ’re sure ’s problem, problem better described question, use Discussions page (see ). Feature requests welcome! problem {gratia} hit top TODO list quickly cen provide reproducible example demonstrating problem.","code":"install.packages(\"gratia\", repos = c(   \"https://gavinsimpson.r-universe.dev\",   \"https://cloud.r-project.org\" ))"},{"path":"https://gavinsimpson.github.io/gratia/CONTRIBUTING.html","id":"got-a-question-want-to-show-how-to-do-something-with-gratia","dir":"","previous_headings":"","what":"Got a question? Want to show how to do something with {gratia}?","title":"Contributing","text":"issue best described question, want know something {gratia}, cool example using {gratia} want share , please consider using Discussions page. ’re sure, can always ask issue Question (Use Q&category) , really bug, can easily create Issue discussion.","code":""},{"path":"https://gavinsimpson.github.io/gratia/CONTRIBUTING.html","id":"code-contributions","dir":"","previous_headings":"","what":"Code contributions","title":"Contributing","text":"Code contributions form Pull Requests always appreciated. suitable workflow: Fork repo Github account Clone version account machine account, e.g,. git clone https://github.com/<yourgithubusername>/gratia.git Make sure track progress upstream (.e., version gratis gavinsimpson/gratia) git remote add upstream https://github.com/gavinsimpson/gratia.git. making changes make sure pull changes upstream either git fetch upstream merge later git pull upstream fetch merge one step Make changes (bonus points making changes new feature branch) Push account Submit pull request home base gavinsimpson/gratia","code":""},{"path":"https://gavinsimpson.github.io/gratia/CONTRIBUTING.html","id":"development-tools--paradigm--ethos","dir":"","previous_headings":"Code contributions","what":"Development tools / paradigm / ethos","title":"Contributing","text":"Please note following contributing code: {gratia} tightly aligned tidyverse use {dplyr} related packages lot internally developing package. plan replacing code lower-level code using {vctrs}, right now development focus filling functionality package premature optimisation. original aim {gratia} provide {ggplot2} plotting smooths; please stick principle use plotting paradigm. aim {mgcv}-feature complete; {mgcv} can something terms plotting smooths, handling specialists smooths, etc, principle {gratia} support . aim general compatibility {mgcv}; {gratia} deviates {mgcv} things, needs good justification; example {gratia} deviates multivariate isotropic smooths fitted s() plotted, Dave Miller (@dill) argued convincingly way. Don’t add dependencies! Unless accompanied strong justification, want reduce dependencies increase number. ’m using {styler} style code, using 2 spaces indent. code written change, however. Respect 80 character line length limit. contribute functionality fix bug, please add test using {testthat} framework insure new things works correctly bug stays fixed. contributions must result new NOTES, WARNINGS, ERRORS running R CMD check ---cran; please check contributions Winbuilder example.","code":""},{"path":"https://gavinsimpson.github.io/gratia/CONTRIBUTING.html","id":"email","dir":"","previous_headings":"","what":"Email","title":"Contributing","text":"hate email! can email GMail address — ’s like can stop :-) — unless capture attention immediately label message {gratia}-related, quickly get swamped never seen . Even label , ’s guarantee ever get round replying; usually happens ’ll forget don’t remember check gratia label often. end result email , get response , tardy. infinitely better use Discussions Issues pages Github ask questions package report problems. Please don’t email work (academic) address (email address might find ) {gratia}. question GAMs, much better ask CrossValidated StackOverflow depending whether question statistical programming related. relates {gratia} can use Discussions page. Asking question public allows others reply public, contributes body knowledge easily available others. circumstances send email multiple addresses; quickest way get message trash.","code":""},{"path":"https://gavinsimpson.github.io/gratia/LICENSE.html","id":null,"dir":"","previous_headings":"","what":"The MIT License (MIT)","title":"The MIT License (MIT)","text":"Copyright (c) 2013-2024 Gavin L. Simpson Permission hereby granted, free charge, person obtaining copy software associated documentation files (“Software”), deal Software without restriction, including without limitation rights use, copy, modify, merge, publish, distribute, sublicense, /sell copies Software, permit persons Software furnished , subject following conditions: copyright notice permission notice shall included copies substantial portions Software. SOFTWARE PROVIDED “”, WITHOUT WARRANTY KIND, EXPRESS IMPLIED, INCLUDING LIMITED WARRANTIES MERCHANTABILITY, FITNESS PARTICULAR PURPOSE NONINFRINGEMENT. EVENT SHALL AUTHORS COPYRIGHT HOLDERS LIABLE CLAIM, DAMAGES LIABILITY, WHETHER ACTION CONTRACT, TORT OTHERWISE, ARISING , CONNECTION SOFTWARE USE DEALINGS SOFTWARE.","code":""},{"path":"https://gavinsimpson.github.io/gratia/articles/custom-plotting.html","id":"background","dir":"Articles","previous_headings":"","what":"Background","title":"Customizing plots","text":"draw() function {gratia} envisaged ggplot-based alternative mgcv:::plot.gam(). , never intended allow sorts customization possible ggplot() packages use ggplot() plotting layer. largely due decision produce multiple separate ggplot() plots GAMs multiple smooths, subsequently combined single figure device, initially using {cowplot} recently {patchwork}. things way evident consider might represent smooths 3 4 variables (common might think; consider space-time models via te(x, y, time, d = c(2,1)) space-depth-time models [think ocean atmospheric data space depth (height), observed time] via te(x, y, depth, time, d = c(2, 1, 1))), require facets top produce small multiples, means can’t use facets plot separate smooths. Additional complications arise consider complex smooth types, splines sphere, might want us different coordinate systems geoms best represent underlying smooth. gone root combining multiple ggplot objects single figure, problem customizing plots quickly rears head. vignette presents solutions problem modifying adding plots produced draw() culminating example illustrating use {gratia}’s utility functions produce plots lower-lever components.","code":""},{"path":"https://gavinsimpson.github.io/gratia/articles/custom-plotting.html","id":"adding-layers-to-plots-with-the-operator","dir":"Articles","previous_headings":"","what":"Adding layers to plots with the & operator","title":"Customizing plots","text":"start simulating data fitting GAM four smooth functions default plot produced draw()  want change theme plots, can’t append theme() layer p affects last plot patchwork1  One way apply theme plots patchwork & operator.","code":"library(\"gratia\") library(\"mgcv\") #> Loading required package: nlme #> This is mgcv 1.9-1. For overview type 'help(\"mgcv-package\")'. library(\"ggplot2\") library(\"dplyr\") #>  #> Attaching package: 'dplyr' #> The following object is masked from 'package:nlme': #>  #>     collapse #> The following objects are masked from 'package:stats': #>  #>     filter, lag #> The following objects are masked from 'package:base': #>  #>     intersect, setdiff, setequal, union library(\"patchwork\")  # simulate data n <- 400 eg1 <- data_sim(\"eg1\", n = n, seed = 1)  # fit model m <- gam(y ~ s(x0) + s(x1) + s(x2) + s(x3),   data = eg1, method = \"REML\") p <- draw(m) p p + theme_bw() p & theme_bw()"},{"path":"https://gavinsimpson.github.io/gratia/articles/custom-plotting.html","id":"combining-individual-plots-produced-by-draw","dir":"Articles","previous_headings":"","what":"Combining individual plots produced by draw()","title":"Customizing plots","text":"draw() methods like draw.gam() return object created patchwork::wrap_plots(), result isn’t straightforward combine objects new patchwork avoid error, need use patchwork::plot_layout() set dimensions want  achieved directly via draw()  instructive know combine outputs draw() need arise, want create patchwork plots different models.","code":"p1 <- draw(m, select = \"s(x0)\") p2 <- draw(m, select = \"s(x1)\") p3 <- draw(m, select = \"s(x2)\") p1 + p2 + p3 #> Error in `wrap_dims()`: #> ! Need 3 panels, but together `nrow` and `ncol` only provide 1. #> ℹ Please increase `ncol` and/or `nrow`. p1 + p2 + p3 + plot_layout(ncol = 3) draw(m, select = c(\"s(x0)\", \"s(x1)\", \"s(x2)\"), ncol = 3)"},{"path":"https://gavinsimpson.github.io/gratia/articles/custom-plotting.html","id":"building-your-own-plot-by-hand","dir":"Articles","previous_headings":"","what":"Building your own plot by hand","title":"Customizing plots","text":"{gratia} provides high-level functions like draw() get good graphical overview fitted model, little option customisation — isn’t possible desirable allow possible customisation options fatures {ggplot2} single function. Think many arguments require! Instead, {gratia} also exports lower-level functions used draw() can create plot using whatever {ggplot2} functions make sense. next code blocks ’ll see plot created draw(m) can recreated hand using lower-level building blocks. main thing need evaluate smooths values covariates. done using smooth_estimates(). also need add credible interval evaluations, can done tidyverse-style via add_confint() default draw.gam() add partial residuals partial effects plots. achieve effect, need add partial residuals data used fit model. can done via add_partial_residuals() will2 add columns names \"s(x0)\". \"s(x1)\", etc. data. Now everything need recreate plots created draw.gam(). code block filter sm focus specific smooth, f(x2)f(x2) (\"s(x2)\"), add rug plot observed values x2, credible interval around estimated smooth, partial residuals point layer, estimated smooth line layer, annotation  Assuming repeat steps smooths, creating plot objects p_sx0, p_sx1, p_sx2, p_sx3 (code shown), can complete plot creating patchwork desired number rows columns  real benefit complete control data plotted can use power {ggplot2} map additional variables plot aesthetics. example, let’s assume factor variable original data want colour partial residuals according levels factor. Let’s create factor Now can modify plotting code map fac colour aesthetic plot partial residuals. save typing, ’ll reorder layers plot add partial residuals last  can also simple model checking plotting smooth partial residuals coloured according one covariates (also plotting actual residuals covariates). code chunk , map covariate x1 colour size aesthetics (note deleted cex = 1.5 allow mapping size)  resulting plot doesn’t show particular problems model way data simulated, hopefully illustrates can possible use low-level functions provided {gratia}.","code":"# evaluate the smooths sm <- smooth_estimates(m) |>   add_confint() sm #> # A tibble: 400 × 11 #>    .smooth .type .by   .estimate   .se     x0    x1    x2    x3 .lower_ci #>    <chr>   <chr> <chr>     <dbl> <dbl>  <dbl> <dbl> <dbl> <dbl>     <dbl> #>  1 s(x0)   TPRS  NA       -0.929 0.422 0.0131    NA    NA    NA    -1.76  #>  2 s(x0)   TPRS  NA       -0.881 0.396 0.0230    NA    NA    NA    -1.66  #>  3 s(x0)   TPRS  NA       -0.834 0.372 0.0329    NA    NA    NA    -1.56  #>  4 s(x0)   TPRS  NA       -0.786 0.348 0.0429    NA    NA    NA    -1.47  #>  5 s(x0)   TPRS  NA       -0.738 0.326 0.0528    NA    NA    NA    -1.38  #>  6 s(x0)   TPRS  NA       -0.690 0.305 0.0627    NA    NA    NA    -1.29  #>  7 s(x0)   TPRS  NA       -0.643 0.287 0.0727    NA    NA    NA    -1.20  #>  8 s(x0)   TPRS  NA       -0.595 0.270 0.0826    NA    NA    NA    -1.12  #>  9 s(x0)   TPRS  NA       -0.548 0.255 0.0925    NA    NA    NA    -1.05  #> 10 s(x0)   TPRS  NA       -0.501 0.242 0.102     NA    NA    NA    -0.975 #> # ℹ 390 more rows #> # ℹ 1 more variable: .upper_ci <dbl> # add partial residuals to data eg1 <- eg1 |>   add_partial_residuals(m) names(eg1) #>  [1] \"y\"     \"x0\"    \"x1\"    \"x2\"    \"x3\"    \"f\"     \"f0\"    \"f1\"    \"f2\"    #> [10] \"f3\"    \"s(x0)\" \"s(x1)\" \"s(x2)\" \"s(x3)\" p_sx2 <- sm |>   filter(.smooth == \"s(x2)\") |>   ggplot() +   geom_rug(aes(x = x2),     data = eg1,     sides = \"b\", length = grid::unit(0.02, \"npc\")   ) +   geom_ribbon(aes(ymin = .lower_ci, ymax = .upper_ci, x = x2),     alpha = 0.2   ) +   geom_point(aes(x = x2, y = `s(x2)`),     data = eg1, cex = 1.5, colour = \"steelblue3\"   ) +   geom_line(aes(x = x2, y = .estimate), lwd = 1.2) +   labs(y = \"Partial effect\", title = \"s(x2)\") p_sx2 p_sx0 + p_sx1 + p_sx2 + p_sx3 + plot_layout(ncol = 2) set.seed(12) eg1 <- eg1 |>   mutate(fac = sample(letters[1:4], n(), replace = TRUE)) plt <- sm |>   filter(.smooth == \"s(x2)\") |>   ggplot() +   geom_rug(aes(x = x2),     data = eg1,     sides = \"b\", length = grid::unit(0.02, \"npc\")   ) +   geom_ribbon(aes(ymin = .lower_ci, ymax = .upper_ci, x = x2),     alpha = 0.2   ) +   geom_line(aes(x = x2, y = .estimate), lwd = 1.2) +   labs(y = \"Partial effect\", title = \"s(x2)\")  plt +   geom_point(     aes(       x = x2, y = `s(x2)`,       colour = fac     ), # <-- map fac to colour aesthetic     data = eg1, cex = 1.5   ) plt +   geom_point(     aes(       x = x2, y = `s(x2)`,       colour = x1, size = x1     ), # <-- map fac to colour aesthetic     data = eg1, alpha = 0.3   ) + # <-- deleted cex!!   scale_colour_viridis_c(option = \"plasma\")"},{"path":"https://gavinsimpson.github.io/gratia/articles/data-slices.html","id":"carbon-dioxide-uptake-in-grass-plants","dir":"Articles","previous_headings":"","what":"Carbon Dioxide Uptake in Grass Plants","title":"Data slices","text":"first example uses small data set experimental study cold tolerance grass Echinochloa crusgalli. data data frame CO2 provided {datasets} package ships R.  One way model data allow different smooths combinations treatment type covariates can look fitted smooths using draw()  might want compare model fitted values treatment types (origins), ignoring random effect component. want evaluate model range values covariate conc combinations factors. data slice covariate space, can create using data_slice(). create data slice conc Quebec type chilled treatment use Notice data_slice() filled something remaining covariates didn’t mention? case, data_slice() doesn’t know tt created, chosen modal level tt factor, correct choice case. Instead, need specify correct level explicitly tt created data slice, can predict model using combination covariate values specified slice. use predict.gam() , fitted_values() function {gratia} easier use, especially non-Gaussian models Notice excluded random effect term; even though specify something plant covariate can ignore term model using exclude argument. fitted_values() creates credible interval scale link function back-transforms response scale scale = \"response\", also default. Plotting fitted values data slice now requires simple {ggplot2} knowledge  Next, let’s compare fitted effects treatment Mississippi origin plants , replaced automatically-generated tt variable correctly specified call fct_cross(), retaining levels type treatment factors. insures correct combinations corresponding treatment type factors also preserve original levels tt covariate created. can visualise fitted values data slice  creating data slices, used helper functions specify covariate values slice. {gratia} provides several helper functions: evenly(x, n = 100) — creates n evenly spaced values range covariate, evenly(x, = 5 — creates evenly spaced values range covariate increments 5, evenly(x, ..., lower = 5, upper = 10) — either two uses evenly() shown use lower upper limits vector x. Arguments lower upper can used change one upper lower bounds. evenly(fct) — produces new factor containing level specified factor fct just , ref_level(fct) — creates new factor containing just reference level specified factor covariate fct, level(fct, \"level\") — creates factor requested \"level\" factor fct. cases involving factors, helper functions set levels factor match original model fit2. second argument data_slice() ... ... argument allows provide expressions create covariate values want data slice. Expressions passed ... evaluated within model frame fitted model (argument object) data (supplied). restricted either using helper functions provide {gratia}; R function used long makes sense context model frame, returns something can combined using tidyr::expand_grid().","code":"## data load and prep data(CO2, package = \"datasets\") plant <- CO2 |>   as_tibble() |>   rename(plant = Plant, type = Type, treatment = Treatment) |>   mutate(plant = factor(plant, ordered = FALSE)) plant_ylab <- expression(CO[2] ~ uptake ~ (mu * mol ~ m^{-3})) plant_xlab <- expression(CO[2] ~ concentration ~ (mL ~ L^{-1}))  plant |>   ggplot(aes(x = conc, y = uptake, group = plant, colour = treatment)) +   geom_point() +   geom_line() +   facet_wrap(~type) +   labs(y = plant_ylab, x = plant_xlab, colour = \"Treatment\") plant <- plant |>   mutate(tt = fct_cross(treatment, type)) m_plant <- gam(uptake ~ treatment * type + s(conc, by = tt, k = 6) +     s(plant, bs = \"re\"),   data = plant, method = \"REML\", family = Gamma(link = \"log\") ) overview(m_plant) #>  #> Generalized Additive Model with 8 terms #>  #>   term                             type              k   edf statistic p.value   #>   <chr>                            <chr>         <dbl> <dbl>     <dbl> <chr>     #> 1 treatment                        parametric       NA  1         1.59 0.2124864 #> 2 type                             parametric       NA  1        11.2  0.0014830 #> 3 treatment:type                   parametric       NA  1         7.45 0.0085489 #> 4 s(conc):ttnonchilled:Quebec      TPRS              5  4.72     69.7  < 0.001   #> 5 s(conc):ttchilled:Quebec         TPRS              5  4.71     86.5  < 0.001   #> 6 s(conc):ttnonchilled:Mississippi TPRS              5  4.62     74.1  < 0.001   #> 7 s(conc):ttchilled:Mississippi    TPRS              5  4.39     25.3  < 0.001   #> 8 s(plant)                         Random effect    12  7.40     12.8  < 0.001 draw(m_plant, residuals = TRUE, scales = \"fixed\") ds1 <- data_slice(m_plant,   conc = evenly(conc, n = 100),   type = level(type, \"Quebec\"), treatment = level(treatment, \"chilled\") ) ds1 #> # A tibble: 100 × 5 #>     conc type   treatment tt                plant #>    <dbl> <fct>  <fct>     <fct>             <fct> #>  1   95  Quebec chilled   nonchilled:Quebec Qn1   #>  2  104. Quebec chilled   nonchilled:Quebec Qn1   #>  3  113. Quebec chilled   nonchilled:Quebec Qn1   #>  4  122. Quebec chilled   nonchilled:Quebec Qn1   #>  5  132. Quebec chilled   nonchilled:Quebec Qn1   #>  6  141. Quebec chilled   nonchilled:Quebec Qn1   #>  7  150. Quebec chilled   nonchilled:Quebec Qn1   #>  8  159. Quebec chilled   nonchilled:Quebec Qn1   #>  9  168. Quebec chilled   nonchilled:Quebec Qn1   #> 10  177. Quebec chilled   nonchilled:Quebec Qn1   #> # ℹ 90 more rows ds1 <- data_slice(m_plant,   conc = evenly(conc, n = 100),   treatment = level(treatment, \"chilled\"), type = level(type, \"Quebec\"),   tt = level(tt, \"chilled:Quebec\") ) ds1 #> # A tibble: 100 × 5 #>     conc treatment type   tt             plant #>    <dbl> <fct>     <fct>  <fct>          <fct> #>  1   95  chilled   Quebec chilled:Quebec Qn1   #>  2  104. chilled   Quebec chilled:Quebec Qn1   #>  3  113. chilled   Quebec chilled:Quebec Qn1   #>  4  122. chilled   Quebec chilled:Quebec Qn1   #>  5  132. chilled   Quebec chilled:Quebec Qn1   #>  6  141. chilled   Quebec chilled:Quebec Qn1   #>  7  150. chilled   Quebec chilled:Quebec Qn1   #>  8  159. chilled   Quebec chilled:Quebec Qn1   #>  9  168. chilled   Quebec chilled:Quebec Qn1   #> 10  177. chilled   Quebec chilled:Quebec Qn1   #> # ℹ 90 more rows fv1 <- fitted_values(m_plant, data = ds1, scale = \"response\", exclude = \"s(plant)\") fv1 #> # A tibble: 100 × 10 #>     .row  conc treatment type   tt      plant .fitted    .se .lower_ci .upper_ci #>    <int> <dbl> <fct>     <fct>  <fct>   <fct>   <dbl>  <dbl>     <dbl>     <dbl> #>  1     1   95  chilled   Quebec chille… Qn1      13.0 0.0783      11.2      15.2 #>  2     2  104. chilled   Quebec chille… Qn1      14.1 0.0757      12.1      16.3 #>  3     3  113. chilled   Quebec chille… Qn1      15.2 0.0737      13.1      17.5 #>  4     4  122. chilled   Quebec chille… Qn1      16.3 0.0722      14.2      18.8 #>  5     5  132. chilled   Quebec chille… Qn1      17.6 0.0714      15.3      20.2 #>  6     6  141. chilled   Quebec chille… Qn1      18.9 0.0711      16.4      21.7 #>  7     7  150. chilled   Quebec chille… Qn1      20.2 0.0712      17.6      23.3 #>  8     8  159. chilled   Quebec chille… Qn1      21.6 0.0716      18.8      24.9 #>  9     9  168. chilled   Quebec chille… Qn1      23.0 0.0721      20.0      26.5 #> 10    10  177. chilled   Quebec chille… Qn1      24.4 0.0726      21.2      28.1 #> # ℹ 90 more rows fv1 |>   ggplot(aes(x = conc, y = .fitted)) +   geom_point(     data = plant |>       filter(type == \"Quebec\", treatment == \"chilled\"),     mapping = aes(y = uptake),     alpha = 0.8, colour = \"steelblue\"   ) +   geom_ribbon(aes(ymin = .lower_ci, ymax = .upper_ci), alpha = 0.2) +   geom_line() +   labs(     x = plant_xlab, y = plant_ylab,     title = expression(Estimated ~ CO[2] ~ uptake),     subtitle = \"Chilled plants of the Quebec type\"   ) ds2 <- data_slice(m_plant,   conc = evenly(conc, n = 100),   treatment = evenly(treatment), type = level(type, \"Mississippi\") ) |>   mutate(tt = fct_cross(treatment, type, keep_empty = TRUE)) ds2 #> # A tibble: 200 × 5 #>     conc treatment  type        tt                     plant #>    <dbl> <fct>      <fct>       <fct>                  <fct> #>  1   95  nonchilled Mississippi nonchilled:Mississippi Qn1   #>  2   95  chilled    Mississippi chilled:Mississippi    Qn1   #>  3  104. nonchilled Mississippi nonchilled:Mississippi Qn1   #>  4  104. chilled    Mississippi chilled:Mississippi    Qn1   #>  5  113. nonchilled Mississippi nonchilled:Mississippi Qn1   #>  6  113. chilled    Mississippi chilled:Mississippi    Qn1   #>  7  122. nonchilled Mississippi nonchilled:Mississippi Qn1   #>  8  122. chilled    Mississippi chilled:Mississippi    Qn1   #>  9  132. nonchilled Mississippi nonchilled:Mississippi Qn1   #> 10  132. chilled    Mississippi chilled:Mississippi    Qn1   #> # ℹ 190 more rows fitted_values(m_plant,   data = ds2, scale = \"response\",   exclude = \"s(plant)\" ) |>   ggplot(aes(x = conc, y = .fitted, group = treatment)) +   geom_point(     data = plant |> filter(type == \"Mississippi\"),     mapping = aes(y = uptake, colour = treatment),     alpha = 0.8   ) +   geom_ribbon(aes(ymin = .lower_ci, ymax = .upper_ci, fill = treatment),     alpha = 0.2   ) +   geom_line(aes(colour = treatment)) +   labs(     x = plant_xlab, y = plant_ylab,     title = expression(Estimated ~ CO[2] ~ uptake),     subtitle = \"Comparison of treatment in plants of the Mississippi type\",     colour = \"Treatment\", fill = \"Treatment\"   ) args(gratia:::data_slice.gam) #> function (object, ..., data = NULL, envir = NULL)  #> NULL"},{"path":"https://gavinsimpson.github.io/gratia/articles/data-slices.html","id":"slices-through-a-2d-smooth","dir":"Articles","previous_headings":"","what":"Slices through a 2D smooth","title":"Data slices","text":"second example, ’ll use bivariate example data set {mgcv} fit tensor product covariates x z aim example create univariate data slice 2D smooth user-specified values x holding z one fixed values. visualise effect smooth level, using smooth_estimates(), response level, using fitted_values().","code":"# simulate data from the bivariate surface df <- data_sim(\"eg2\", n = 1000, scale = 0.25, seed = 2)  # fit the GAM m_biv <- gam(y ~ te(x, z), data = df, method = \"REML\")"},{"path":"https://gavinsimpson.github.io/gratia/articles/data-slices.html","id":"using-smooth_estimates","dir":"Articles","previous_headings":"Slices through a 2D smooth","what":"Using smooth_estimates()","title":"Data slices","text":"begin creating slice data space. also create label point nice axis label. evaluate smooth desired values add confidence interval can plot sm using {ggplot2}  Note interval Marra Wood (2012) interval. doesn’t include uncertainty model constant term moment, unless smooth close linear shouldn’t make much difference. extends multiple slices asking several discrete z","code":"ds3 <- data_slice(m_biv,   x = evenly(x, n = 100),   z = quantile(z, probs = 0.25) )  z_val <- with(ds3, round(quantile(z, probs = 0.25), 2)) ylab <- bquote(hat(f)(x, .(z_val))) sm <- smooth_estimates(m_biv, select = \"te(x,z)\", data = ds3) |>   add_confint() sm #> # A tibble: 100 × 9 #>    .smooth .type        .by   .estimate    .se       x     z .lower_ci .upper_ci #>    <chr>   <chr>        <chr>     <dbl>  <dbl>   <dbl> <dbl>     <dbl>     <dbl> #>  1 te(x,z) Tensor prod… NA        0.103 0.0583 6.63e-4 0.245   -0.0107     0.218 #>  2 te(x,z) Tensor prod… NA        0.122 0.0548 1.08e-2 0.245    0.0148     0.230 #>  3 te(x,z) Tensor prod… NA        0.141 0.0514 2.08e-2 0.245    0.0400     0.242 #>  4 te(x,z) Tensor prod… NA        0.159 0.0482 3.09e-2 0.245    0.0648     0.254 #>  5 te(x,z) Tensor prod… NA        0.177 0.0451 4.10e-2 0.245    0.0890     0.266 #>  6 te(x,z) Tensor prod… NA        0.195 0.0422 5.11e-2 0.245    0.113      0.278 #>  7 te(x,z) Tensor prod… NA        0.213 0.0396 6.12e-2 0.245    0.135      0.291 #>  8 te(x,z) Tensor prod… NA        0.230 0.0372 7.13e-2 0.245    0.157      0.303 #>  9 te(x,z) Tensor prod… NA        0.247 0.0351 8.14e-2 0.245    0.178      0.316 #> 10 te(x,z) Tensor prod… NA        0.263 0.0333 9.14e-2 0.245    0.198      0.328 #> # ℹ 90 more rows sm |>   ggplot(aes(x = x, y = .estimate)) +   geom_ribbon(aes(ymin = .lower_ci, ymax = .upper_ci), alpha = 0.2) +   geom_line() +   labs(     title = \"Evaluation of smooth te(x,z) at fixed z\",     y = ylab   ) ds4 <- data_slice(m_biv,   x = evenly(x, n = 100),   z = round(quantile(z, probs = c(0.25, 0.5, 0.75)), 2) )  sm <- smooth_estimates(m_biv, select = \"te(x,z)\", data = ds4) |>   add_confint() |>   mutate(fz = factor(z))  sm |>   ggplot(aes(x = x, y = .estimate, colour = fz, group = fz)) +   geom_ribbon(aes(ymin = .lower_ci, ymax = .upper_ci, fill = fz, colour = NULL),     alpha = 0.2   ) +   geom_line() +   labs(     title = \"Evaluation of smooth te(x,z) at fixed z\",     y = expression(hat(f)(x, z)), colour = \"z\", fill = \"z\"   )"},{"path":"https://gavinsimpson.github.io/gratia/articles/data-slices.html","id":"using-fitted_samples","dir":"Articles","previous_headings":"Slices through a 2D smooth","what":"Using fitted_samples()","title":"Data slices","text":"want evaluate surface x fixed z conditional upon values covariates (model predicted fitted values) fitted_samples() tidy wrapper predict.gam(). single z  multiple z difference now model constant included well uncertainty.","code":"fitted_values(m_biv, data = ds3) |> # default is response scale, not link   ggplot(aes(x = x, y = .fitted)) +   geom_ribbon(aes(ymin = .lower_ci, ymax = .upper_ci), alpha = 0.2) +   geom_line() +   labs(     title = \"Fitted values from model\",     y = expression(hat(y))   ) fitted_values(m_biv, data = ds4) |>   mutate(fz = factor(z)) |>   ggplot(aes(x = x, y = .fitted, colour = fz, group = fz)) +   geom_ribbon(aes(ymin = .lower_ci, ymax = .upper_ci, fill = fz, colour = NULL),     alpha = 0.2   ) +   geom_line() +   labs(     title = \"Fitted values from model\",     y = expression(hat(y)), colour = \"z\", fill = \"z\"   )"},{"path":"https://gavinsimpson.github.io/gratia/articles/gratia.html","id":"plotting","dir":"Articles","previous_headings":"","what":"Plotting","title":"Getting started with gratia","text":"gratia provides draw() function produce plots using ggplot2 📦. plot estimated smooths GAM fitted , use  plots produced partial effect plots, show component contributions, link scale, model term linear predictor. y axis plots typically centred around 0 due smooths sum--zero identifiability constraint applied . constraint allows model include multiple smooths remain identifiable. plots allow read contributions smooth fitted response (link scale); show link-scale predictions response smooth, conditional upon terms model, including parametric effects intercept, zero contribution. parlance marginaleffects package (Arel-Bundock, Greifer, Heiss Forthcoming), plots show adjusted predictions, just adjustment includes setting contribution model terms predicted value zero. partial derivatives (marginaleffects call marginal effect slope), gratia provides derivatives(). resulting plot intended reasonable overview estimated model, offers limited option modify resulting plot. want full control, can obtain data used create plot smooth_estimates() evaluate smooths values evenly spaced range covariate(s). want evaluate selected smooths, can specify via smooth argument. takes smooth labels names smooths known mgcv. list labels smooths use evaluate f(x2)f(x_2) use can generate plot using ggplot2 package, example","code":"draw(m) sm <- smooth_estimates(m) sm #> # A tibble: 400 × 9 #>    .smooth .type .by   .estimate   .se       x0    x1    x2    x3 #>    <chr>   <chr> <chr>     <dbl> <dbl>    <dbl> <dbl> <dbl> <dbl> #>  1 s(x0)   TPRS  NA       -1.32  0.390 0.000239    NA    NA    NA #>  2 s(x0)   TPRS  NA       -1.24  0.365 0.0103      NA    NA    NA #>  3 s(x0)   TPRS  NA       -1.17  0.340 0.0204      NA    NA    NA #>  4 s(x0)   TPRS  NA       -1.09  0.318 0.0304      NA    NA    NA #>  5 s(x0)   TPRS  NA       -1.02  0.297 0.0405      NA    NA    NA #>  6 s(x0)   TPRS  NA       -0.947 0.279 0.0506      NA    NA    NA #>  7 s(x0)   TPRS  NA       -0.875 0.263 0.0606      NA    NA    NA #>  8 s(x0)   TPRS  NA       -0.803 0.249 0.0707      NA    NA    NA #>  9 s(x0)   TPRS  NA       -0.732 0.237 0.0807      NA    NA    NA #> 10 s(x0)   TPRS  NA       -0.662 0.228 0.0908      NA    NA    NA #> # ℹ 390 more rows smooths(m) #> [1] \"s(x0)\" \"s(x1)\" \"s(x2)\" \"s(x3)\" sm <- smooth_estimates(m, smooth = \"s(x2)\") #> Warning: The `smooth` argument of `smooth_estimates()` is deprecated as of gratia #> 0.8.9.9. #> ℹ Please use the `select` argument instead. #> This warning is displayed once every 8 hours. #> Call `lifecycle::last_lifecycle_warnings()` to see where this warning was #> generated. sm #> # A tibble: 100 × 6 #>    .smooth .type .by   .estimate   .se      x2 #>    <chr>   <chr> <chr>     <dbl> <dbl>   <dbl> #>  1 s(x2)   TPRS  NA      -4.47   0.476 0.00359 #>  2 s(x2)   TPRS  NA      -4.00   0.406 0.0136  #>  3 s(x2)   TPRS  NA      -3.53   0.345 0.0237  #>  4 s(x2)   TPRS  NA      -3.06   0.295 0.0338  #>  5 s(x2)   TPRS  NA      -2.58   0.263 0.0438  #>  6 s(x2)   TPRS  NA      -2.09   0.250 0.0539  #>  7 s(x2)   TPRS  NA      -1.59   0.253 0.0639  #>  8 s(x2)   TPRS  NA      -1.08   0.264 0.0740  #>  9 s(x2)   TPRS  NA      -0.564  0.278 0.0841  #> 10 s(x2)   TPRS  NA      -0.0364 0.289 0.0941  #> # ℹ 90 more rows library(\"ggplot2\") library(\"dplyr\") #>  #> Attaching package: 'dplyr' #> The following object is masked from 'package:nlme': #>  #>     collapse #> The following objects are masked from 'package:stats': #>  #>     filter, lag #> The following objects are masked from 'package:base': #>  #>     intersect, setdiff, setequal, union sm |>   add_confint() |>   ggplot(aes(y = .estimate, x = x2)) +   geom_ribbon(aes(ymin = .lower_ci, ymax = .upper_ci),     alpha = 0.2, fill = \"forestgreen\"   ) +   geom_line(colour = \"forestgreen\", linewidth = 1.5) +   labs(     y = \"Partial effect\",     title = expression(\"Partial effect of\" ~ f(x[2])),     x = expression(x[2])   )"},{"path":"https://gavinsimpson.github.io/gratia/articles/gratia.html","id":"model-diagnostics","dir":"Articles","previous_headings":"","what":"Model diagnostics","title":"Getting started with gratia","text":"appraise() function provides standard diagnostic plots GAMs  plots produced (left--right, top--bottom), quantile-quantile (QQ) plot deviance residuals, scatterplot deviance residuals linear predictor, histogram deviance residuals, scatterplot observed vs fitted values. Adding partial residuals partial effect plots produced draw() can also help diagnose problems model, oversmoothing","code":"appraise(m) draw(m, residuals = TRUE)"},{"path":"https://gavinsimpson.github.io/gratia/articles/gratia.html","id":"want-to-learn-more","dir":"Articles","previous_headings":"","what":"Want to learn more?","title":"Getting started with gratia","text":"gratia active development area development currently lacking documentation. find package, look help pages package look examples code help get going.","code":""},{"path":[]},{"path":"https://gavinsimpson.github.io/gratia/articles/posterior-simulation.html","id":"what-are-we-simulating","dir":"Articles","previous_headings":"","what":"What are we simulating?","title":"Posterior Simulation","text":"Posterior simulation involves randomly sampling MVN(𝛃̂,𝐕b)\\text{MVN}(\\hat{\\boldsymbol{\\beta}}, \\mathbf{V}_{\\text{b}}) EF(μi,ϕ)\\text{EF}(\\mu_i, \\phi), . might simulate posterior distribution single estimated smooth function see uncertainty estimate function. simulate just subset βj⋅\\beta_{j \\cdot} associated fjf_j interest. Instead, might interested uncertainty expectation (expected value) model given values covariates, case can simulate 𝛃\\boldsymbol{\\beta} sample posterior 𝔼(yi)\\mathbb{E}(y_i), fitted values model. might want generate new values response variable via draws conditional distribution response, simulating new response data 𝕪*\\mathbb{y}^{\\ast}, either observed 𝐱\\mathbf{x} new values $^{}, yi*|𝛈,𝐱∼EF(μî,ϕ)y^{\\ast}_i | \\boldsymbol{\\eta}, \\mathbf{x} \\sim \\text{EF}(\\hat{\\mu_i}, \\phi). Finally, can combine posterior simulation distributions generate posterior draws new data 𝕪*\\mathbb{y}^{\\ast} also include uncertainty expected values. gratia functionality options following functions smooth_samples() generates draws posterior distribution single estimated smooth functions, fitted_samples() generates draws posterior distribution 𝔼(yi|𝐗i=xi)\\mathbb{E}(y_i | \\mathbf{X}_i = x_i), expected value responss, predicted_samples(), generates new response data given supplied values covariates yi*|𝐗i=xi*y^{\\ast}_i | \\mathbf{X}_i = x^{\\ast}_i posterior_samples(), generates draws posterior distribution model, including uncertainty estimated parameters model. simpler terms, fitted_samples() generates predictions “average” expected value response values covariates. predictions include uncertainty estimated values model coefficients. contrast, posterior_samples() generates predictions actual values response might expect observe (model correct) given values covariates. predicted values include variance sampling distribution (error term). predicted_samples() lies somewhere two; predicted values include variation sampling distribution, take model fixed, known. worth reminding posterior draws conditional upon selected values smoothing parameter(s) λj\\lambda_j. act wiggliness estimated smooths known, actual fact estimated (selected perhaps better description) wiglinesses data model fitting. estimated GAM fitted method argument \"REML\", \"ML\", version 𝐕b\\mathbf{V}_{\\text{b}} corrected selected smoothing parameters, 𝐕c\\mathbf{V}_{\\text{c}}, generally available. allows, extent, posterior simulation account additional source uncertainty chosen values 𝛌\\boldsymbol{\\lambda}. two additional functions available gratia posterior simulation: simulate(), derivative_samples(). gratia provides simulate() methods models estimated using gam(), bam(), gamm(), well fitted via scam() scam package. simulate() base R convention thing predicted_samples(), just non-tidy way (pejorative; returns simulated response values matrix, arguably useful math statistical computation.) derivative_samples() provides draws posterior distribution derivative response variable small change focal covariate value. derivative_samples() less general version fitted_samples(); achieve thing two separate calls fitted_samples(). ’ll reserve discussion derivative_samples() separate vignette focused estimating derivatives GAMs. following sections ’ll look four main posterior simulation functions turn.","code":""},{"path":"https://gavinsimpson.github.io/gratia/articles/posterior-simulation.html","id":"posterior-smooths-and-smooth_samples","dir":"Articles","previous_headings":"","what":"Posterior smooths and smooth_samples()","title":"Posterior Simulation","text":"can sample posterior distribution coefficients particular smooth β̂j\\hat{\\beta}_j given values smoothing parameters 𝛌̂\\hat{\\boldsymbol{\\lambda}}. generate posterior samples smooths sampling 𝛃j⋆∼N(β̂j,𝐕β̂j)\\boldsymbol{\\beta}_{j\\star} \\sim N(\\hat{\\beta}_j, \\mathbf{V}_{\\hat{\\beta}_j}) forming 𝐗𝛃̂j𝛃j⋆𝖳\\mathbf{X}_{\\hat{\\boldsymbol{\\beta}}_j} \\boldsymbol{\\beta}_{j\\star}^{\\mathsf{T}}. sampling can done using smooth_samples(). illustrate , ’ll simulate data Gu & Wabha’s 4 smooth example, fit GAM simulated data simulating posterior distribution estimated smooth, sampling coefficients particular smooth. model, coefficients smooth f(x0)f(x_0) stored elements 2 10 coefficients vector. sample posterior distribution coefficients use smooth_samples() choosing particular smooth ’re interested using select argument; want sample smooths posteriors smooths model, select can left default value. Typically ’re bothered particular values covariate evaluate posterior smooths; ask 100 evenly spaced values x0 using n_vals, can provide covariates values via data argument. number posterior smooths sampled controlled argument n; ask 100 samples. Objects returned smooth_samples() draw() method available draw posterior smooths can set n_samples randomly select many smooths draw (seed can provided via argument seed make set chosen smooths repeatable.) credible interval smooth contain smooths. standard 95% credible interval, sampled smooths exceed limits interval.  Following Marra & Wood (2012), blue credible interval contain average 95% grey lines (posterior smooths) given value x0x_0. across function frequentist interpretation credible interval implies values x0x_0 coverage less 95% values greater 95%.","code":"ss_df <- data_sim(\"eg1\", seed = 42) m_ss <- gam(y ~ s(x0) + s(x1) + s(x2) + s(x3), data = ss_df, method = \"REML\") s_x0 <- get_smooth(m_ss, \"s(x0)\") smooth_coef_indices(s_x0) #> [1]  2  3  4  5  6  7  8  9 10 sm_samp <- smooth_samples(m_ss, select = \"s(x0)\", n_vals = 100, n = 100,   seed = 21) sm_samp |>   draw(alpha = 0.3) # evaluate the fitted smooth over x0 and add on a credible interval sm_est <- smooth_estimates(m_ss, select = \"s(x0)\") |>   add_confint() # plot the smooth, credible interval, and posterior smooths sm_est |>   ggplot(aes(x = x0)) +   geom_lineribbon(aes(ymin = .lower_ci, ymax = .upper_ci),     orientation = \"vertical\", fill = \"#56B4E9\", alpha = 0.5   ) +   geom_line(     data = sm_samp,     aes(y = .value, group = .draw), alpha = 0.2   ) +   geom_line(aes(y = .estimate), linewidth = 1, colour = \"#E69F00\") +   labs(y = smooth_label(s_x0))"},{"path":"https://gavinsimpson.github.io/gratia/articles/posterior-simulation.html","id":"posterior-fitted-values-via-fitted_samples","dir":"Articles","previous_headings":"","what":"Posterior fitted values via fitted_samples()","title":"Posterior Simulation","text":"Posterior fitted values draws posterior distribution mean expected value response. expectations returned use predict() estimated GAM, except fitted_samples() includes uncertainty estimated model coefficients, whereas predict() just uses estimated coefficients. example, using data_sim() simulate data example 6 Luo & Wahba (1997) sin(2⋅(4x−2))+2⋅exp(−256⋅(x−0.5)2) \\sin(2 \\cdot (4x - 2)) + 2 \\cdot \\exp(-256 \\cdot (x - 0.5)^2)  data fit adaptive smoother Next create data slice 200 values interval (0,1) ’ll predict model generate posterior fitted values compute fitted values new data posterior fitted values drawn fitted_samples() using Gaussian approximation posterior. just take 10 draws posterior observation new_df merge posterior draws data Adding posterior fitted samples plot data, superimposing Bayesian credible interval fitted values  see posterior draws largely contained credible interval. difference smooth_samples() now ’re including effects model terms. simple model single smooth identity link, difference model constant term uncertainty included samples.","code":"f <- function(x) {   sin(2 * ((4 * x) - 2)) + (2 * exp(-256 * (x - 0.5)^2)) } df <- data_sim(\"lwf6\", dist = \"normal\", scale = 0.3, seed = 2) plt <- df |>   ggplot(aes(x = x, y = y)) +   geom_point(alpha = 0.5) +   geom_function(fun = f) plt m <- gam(y ~ s(x, k = 25, bs = \"ad\"), data = df, method = \"REML\") new_df <- data_slice(m, x = evenly(x, lower = 0, upper = 1, n = 200)) |>   mutate(.row = row_number()) fv <- fitted_values(m, data = new_df) fs <- fitted_samples(m, data = new_df, n = 10, seed = 4) |>   left_join(new_df |> select(.row, x), by = join_by(.row == .row)) plt +   geom_ribbon(data = fv, aes(y = .fitted, ymin = .lower_ci, ymax = .upper_ci),     fill = \"red\", alpha = 0.3) +   geom_line(data = fs, aes(group = .draw, x = x, y = .fitted),     colour = \"yellow\", alpha = 0.4)"},{"path":[]},{"path":"https://gavinsimpson.github.io/gratia/articles/posterior-simulation.html","id":"distributional-gams","dir":"Articles","previous_headings":"Additional examples","what":"Distributional GAMs","title":"Posterior Simulation","text":"possible, predicted_samples() posterior_samples() also work distributional GAMs. possible suitable random number generator available family object stored within model. illustrate, reuse example Matteo Fasiolo, author mgcViz package, based classic mcycle data set MASS package. loading data adding row number variable use late data fit standard Gaussian GAM conditional mean accel. Next, simulate n_sim new response values observed data using predicted_samples() comments briefly indicate dplyr code . Now can plot observed simulated data resulting plot shown left-hand panel figure . clearly problem ; simulated data don’t look much like observations 15ms immediately impact ~45ms impact. due model fitted conditional mean accel. Instead, model conditional mean conditional variance data, linear predictors parameters Gaussian distribution Simulating new data follows using code earlier benefit data wrangling now realised can replace data plot created earlier simulations distribution GAM, plot plot simulated response data distributional GAM shown right hand panel plot. Now, much less disagreement observed data can produce fitted mdoel.","code":"data(mcycle, package = \"MASS\") mcycle <- mcycle |>   mutate(     .row = row_number()   ) |>   relocate(.row, .before = 1L) m_gau <- gam(accel ~ s(times, k = 20),   data = mcycle, method = \"REML\" ) n_sim <- 10 n_data <- nrow(mcycle)  sim_gau <- predicted_samples(m_gau, n = n_sim, seed = 10) |>   left_join(mcycle |> select(-accel), # join on the observed data for times     by = \".row\"   ) |>   rename(accel = .response) |>        # rename   bind_rows(mcycle |>     relocate(.after = last_col())) |> # bind on observer data   mutate(                             # add indicator: simulated or observed     type = rep(c(\"simulated\", \"observed\"),       times = c(n_sim * n_data, n_data)     ),     .alpha = rep(                     # set alpha values for sims & observed       c(0.2, 1), time = c(n_sim * n_data, n_data)     )   ) plt_labs <- labs(   x = \"Time after impact [ms]\",   y = \"Acceleration [g]\" )  plt_gau <- sim_gau |>   ggplot(aes(x = times)) +   geom_point(aes(y = accel, colour = type, alpha = .alpha)) +   plt_labs +   scale_colour_okabe_ito(order = c(6, 5)) +   scale_alpha_identity() m_gaulss <- gam(   list(accel ~ s(times, k = 20, bs = \"tp\"),     ~ s(times, bs = \"tp\")   ), data = mcycle, family = gaulss() ) sim_gaulss <- predicted_samples(m_gaulss, n = n_sim, seed = 20) |>   left_join(mcycle |> select(-accel),     by = \".row\"   ) |>   rename(accel = .response) |>   bind_rows(mcycle |> relocate(.after = last_col())) |>   mutate(     type = rep(c(\"simulated\", \"observed\"),       times = c(n_sim * n_data, n_data)     ),     .alpha = rep(c(0.2, 1), time = c(n_sim * n_data, n_data))   ) plt_gaulss <- plt_gau %+% sim_gaulss  plt_gau + plt_gaulss +   plot_annotation(tag_levels = \"a\") +   plot_layout(guides = \"collect\", ncol = 2)"},{"path":"https://gavinsimpson.github.io/gratia/articles/posterior-simulation.html","id":"prediction-intervals","dir":"Articles","previous_headings":"Additional examples","what":"Prediction intervals","title":"Posterior Simulation","text":"One use posterior simulation generate prediction intervals fitted model. Prediction intervals include two sources uncertainty; estimated model , plus sampling uncertainty error arises drawing observations conditional distribution response. example, Gaussian GAM, first source uncertainty comes uncertainty estimates βj\\beta_j, model coefficients. uncertainty mean expected value response. second source uncertainty stems error term, estimated variance response. two parameters define conditional distribution YiY_i. value covariate(s) 𝐗\\mathbf{X}, estimated model defines entire distribution response values might expect observe covariate values. illustrate, ’ll fit simple GAM single smooth function data simulate Gu & Wabha’s function f2f_2 using data_sim(). simulate 400 values Gaussian distribution variance σ2=1\\sigma^2 = 1. simulated data, true function generated shown  GAM data contains single smooth function x consider new value covariate x, x*=0.5x^{\\ast} = 0.5, expected value response given model, 𝔼(y*|x=x*)\\mathbb{E}(y^{*} | x = x^{*}), ~2.92, obtain using predict() value mean Gaussian distribution , model correct description data, describes distribution values YY might take x=0.5x = 0.5. Gaussian distribution defined two parameters; mean, μ\\mu, describes middle distribution, variance, σ2\\sigma^2, describes spread distribution mean. fully describe Gaussian distribution response x=0.5x = 0.5, need estimate variance. didn’t model explicitly GAM, get estimate model’s scale parameter, ϕ\\phi. stored element scale model object can visualise distribution looks like magic ggdist package  orange region shows expected density response values x*=0.5x^{\\ast} = 0.5 model predicts expect observe. region assumes uncertainty estimate mean variance. Prediction intervals take account variation expected value, plus uncertainty expected value. fitted_values() conveniently returns uncertainty us, default 95% credible interval .se column standard error (standard deviation) estimated value (.fitted), .lower_ci .upper_ci lower upper uncertainty bounds (95% level) estimated value respectively. GAMs fitted mgcv don’t corresponding estimate uncertainty scale parameter, ϕ\\phi, model estimated standard deviation σ̂\\hat{\\sigma}. pretty easy compute upper lower tail quantiles fitted Gaussian distribution range values x get prediction interval, ’d ignoring uncertainty model estimates mean. Posterior simulation provides simple convenient way generate prediction interval includes model uncertainty, works principle families available mgcv (although practice, families currently supported gratia). compute prediction interval x GAM, creating set data evenly range x observed data used fit model added variable .row used later match posterior simulated values row prediction data set ds. also compute fitted values new observations using fitted_values(). step isn’t required order posterior simulation gratia, ’ll use fitted values later show model estimated values uncertainty contrast prediction interval. use posterior_samples() generate new response data new x values ds use join add prediction data draw asked 10000 posterior draws new value x. Ideally ’d generate least three four times many draws get precise estimate prediction interval, keep number low vignette avoid excessive computation time. ’re also using smoothness parameter selection corrected version Bayesian covariance matrix; matrix adjusted account us knowing value smoothing parameter f(xi)f(x_i). ps tibble, n * nrow(ds) rows. .draw variable groups simulated values posterior draw, .row groups posterior draws value x. summarise posterior draws using {dplyr} need function compute quantiles posterior distribution value x (.row). following function simple wrapper around quantile() function base R, arranges output quantile() data frame. apply function set posterior draws, grouping .row summarise separately posterior distribution new value x. reframe() used summarise posterior using quantile_fun() function. ease use, pivot resulting summary long wide format add covariate values joining .row variable 95% prediction interval shown first 10 rows prediction data. column labelled .q50 median posterior distribution. can now use various objects produced plot fitted values model (uncertainties), well prediction intervals just generated. add observed data used fit model black points, summarise posterior samples (ps) using hexagonal binning (avoid plotting 2 million posterior samples)  outermost pair blue lines plot prediction interval created. interval encloses, expected, almost observe data points. also encloses, design, posterior samples, indicated filled hexagonal bins, warmer colours indicating larger counts posterior draws.","code":"df <- data_sim(\"gwf2\", n = 400, scale = 1, dist = \"normal\", seed = 8) df |>   ggplot(aes(x = x, y = y)) +   geom_point() +   geom_function(fun = gw_f2, colour = \"#0072B2\", linewidth = 1.5) m <- gam(y ~ s(x), data = df, method = \"REML\", family = gaussian()) mu <- predict(m, newdata = data.frame(x = 0.5)) mu #>        1  #> 2.919094 sigma <- m$scale sigma #> [1] 1.019426 df |>   ggplot(aes(x = x, y = y)) +   stat_halfeye(aes(ydist = dist_normal(mean = mu, sd = sigma)),     x = 0.5, scale = 0.2, slab_fill = \"#E69F00\", slab_alpha = 0.7   ) +   geom_point() +   geom_function(fun = gw_f2, colour = \"#0072B2\", linewidth = 1.5) +   geom_point(x = 0.5, y = mu, colour = \"red\") fitted_values(m, data = data.frame(x = 0.5)) #> # A tibble: 1 × 6 #>    .row     x .fitted   .se .lower_ci .upper_ci #>   <int> <dbl>   <dbl> <dbl>     <dbl>     <dbl> #> 1     1   0.5    2.92 0.161      2.60      3.23 ds <- data_slice(m, x = evenly(x, n = 200)) |>   mutate(.row = row_number()) fv <- fitted_values(m, data = ds) ps <- posterior_samples(m, n = 10000, data = ds, seed = 24,   unconditional = TRUE) |>   left_join(ds, by = join_by(.row == .row)) ps #> # A tibble: 2,000,000 × 4 #>     .row .draw .response       x #>    <int> <int>     <dbl>   <dbl> #>  1     1     1   -1.34   0.00129 #>  2     2     1   -0.0495 0.00629 #>  3     3     1    0.0308 0.0113  #>  4     4     1   -0.783  0.0163  #>  5     5     1    0.861  0.0213  #>  6     6     1    0.475  0.0263  #>  7     7     1    0.858  0.0313  #>  8     8     1    0.143  0.0363  #>  9     9     1   -0.0344 0.0413  #> 10    10     1    1.04   0.0463  #> # ℹ 1,999,990 more rows quantile_fun <- function(x, probs = c(0.025, 0.5, 0.975), ...) {   tibble::tibble(     .value = quantile(x, probs = probs, ...),     .q = probs * 100   ) } p_int <- ps |>   group_by(.row) |>   reframe(quantile_fun(.response)) |>   pivot_wider(     id_cols = .row, names_from = .q, values_from = .value,     names_prefix = \".q\"   ) |>   left_join(ds, by = join_by(.row == .row)) p_int #> # A tibble: 200 × 5 #>     .row  .q2.5    .q50 .q97.5       x #>    <int>  <dbl>   <dbl>  <dbl>   <dbl> #>  1     1 -2.84  -0.847    1.25 0.00129 #>  2     2 -2.70  -0.651    1.41 0.00629 #>  3     3 -2.50  -0.434    1.62 0.0113  #>  4     4 -2.24  -0.207    1.83 0.0163  #>  5     5 -2.04  -0.0197   2.01 0.0213  #>  6     6 -1.81   0.191    2.25 0.0263  #>  7     7 -1.59   0.391    2.41 0.0313  #>  8     8 -1.38   0.594    2.60 0.0363  #>  9     9 -1.20   0.831    2.84 0.0413  #> 10    10 -0.935  1.06     3.04 0.0463  #> # ℹ 190 more rows fv |>   ggplot(aes(x = x, y = .fitted)) +   # summarise the posterior samples   geom_hex(     data = ps, aes(x = x, y = .response, fill = after_stat(count)),     bins = 50, alpha = 0.7   ) +   # add the lower and upper prediction intervals   geom_line(data = p_int, aes(y = .q2.5), colour = \"#56B4E9\",     linewidth = 1.5) +   geom_line(data = p_int, aes(y = .q97.5), colour = \"#56B4E9\",     linewidth = 1.5) +   # add the lower and upper credible intervals   geom_line(aes(y = .lower_ci), colour = \"#56B4E9\", linewidth = 1) +   geom_line(aes(y = .upper_ci), colour = \"#56B4E9\", linewidth = 1) +   # add the fitted model   geom_line() +   # add the observed data   geom_point(data = df, aes(x = x, y = y)) +   scale_fill_viridis_c(option = \"plasma\") +   theme(legend.position = \"none\") +   labs(y = \"Response\")"},{"path":"https://gavinsimpson.github.io/gratia/articles/posterior-simulation.html","id":"metropolis-hastings-sampler","dir":"Articles","previous_headings":"Additional examples","what":"Metropolis Hastings sampler","title":"Posterior Simulation","text":"cases, Gaussian approximation posterior distribution model coefficients can fail. Simon Wood shows example just failure ?gam.mh help page, Gaussian approximation basically useless binomial GAM large numbers zeroes. mgcv::gam.mh() implements simple Metropolis Hastings sampler, alternates proposals Gaussian t distribution approximation posterior random walk proposals based shrunken approximate posterior covariance matrix. section, rework Simon’s example failure Gaussian approximation ?gam.mh show use gratia generate posterior draws using Metropolis Hastings sampler provided gam.mh(). begin defining function simulate data example. use simulate data set plot  Note zeroes large parts covariate space response zeroes. fit binomial (logistic) GAM data generate sample posterior distribution using default Gaussian approximation subsequently using simpler Metropolis Hastings sampler. method argument used select Metropolis Hastings sampler, specify two additional arguments: thin, controls many draws skipped retained sample, rw_scale, scaling factor posterior covariance matrix shrunk random walk proposals. leave two important arguments defaults: burnin = 1000, number samples discard prior sampling, t_df = 40, degrees freedom t proposals. degrees freedom t proposals large, ’re effectively Gaussian approximation default, alternating proposals random walk proposals. collected posterior draws, summarise set 50%, 80%, 95% intervals using ggdist::median_qi(), add data locations left join First plot intervals Gaussian approximation posterior, repeat plot using intervals derived Metropolis Hastings sampler, arranging two plots using patchwork  Gaussian approximation-based intervals shown left figure, range x largely useless, covering entire range response, despite fact observed zeroes large parts covariate space. Contrast intervals ones obtained using Metropolis Hastings sampler; intervals much better reflect uncertainty estimated response function x data zeroes.","code":"ga_fail <- function(seed) {   df <- tibble(y = c(     rep(0, 89), 1, 0, 1, 0, 0, 1, rep(0, 13), 1, 0, 0, 1,     rep(0, 10), 1, 0, 0, 1, 1, 0, 1, rep(0, 4), 1, rep(0, 3),     1, rep(0, 3), 1, rep(0, 10), 1, rep(0, 4), 1, 0, 1, 0, 0,     rep(1, 4), 0, rep(1, 5), rep(0, 4), 1, 1, rep(0, 46)   )) |>     mutate(       x = withr::with_seed(         seed,         sort(c(0:10 * 5, rnorm(length(y) - 11) * 20 + 100))       ),       .row = row_number()     ) |>     relocate(.row, .before = 1L)   df } df <- ga_fail(3)  df |>   ggplot(aes(x = x, y = y)) +   geom_point() m_logit <- gam(y ~ s(x, k = 15), data = df, method = \"REML\", family = binomial) fs_ga <- fitted_samples(m_logit, n = 2000, seed = 2) fs_mh <- fitted_samples(m_logit,   n = 2000, seed = 2, method = \"mh\", thin = 2,   rw_scale = 0.4 ) excl_col <- c(\".draw\", \".parameter\", \".row\") int_ga <- fs_ga |>   group_by(.row) |>   median_qi(.width = c(0.5, 0.8, 0.95), .exclude = excl_col) |>   left_join(df, by = join_by(.row == .row))  int_mh <- fs_mh |>   group_by(.row) |>   median_qi(.width = c(0.5, 0.8, 0.95), .exclude = excl_col) |>   left_join(df, by = join_by(.row == .row)) plt_ga <- df |>   ggplot(aes(x = x, y = y)) +   geom_point() +   geom_lineribbon(     data = int_ga,     aes(x = x, y = .fitted, ymin = .lower, ymax = .upper)   ) +   scale_fill_brewer() +   labs(title = \"Gaussian approximation\")  plt_mh <- df |>   ggplot(aes(x = x, y = y)) +   geom_point() +   geom_lineribbon(     data = int_mh,     aes(x = x, y = .fitted, ymin = .lower, ymax = .upper)   ) +   scale_fill_brewer() +   labs(title = \"Metropolis Hastings sampler\")  plt_ga + plt_mh + plot_layout(guides = \"collect\")"},{"path":[]},{"path":"https://gavinsimpson.github.io/gratia/authors.html","id":null,"dir":"","previous_headings":"","what":"Authors","title":"Authors and Citation","text":"Gavin L. Simpson. Author, maintainer, copyright holder. Henrik Singmann. Contributor.","code":""},{"path":"https://gavinsimpson.github.io/gratia/authors.html","id":"citation","dir":"","previous_headings":"","what":"Citation","title":"Authors and Citation","text":"Simpson G (????). gratia: Graceful ggplot-Based Graphics Functions GAMs Fitted using mgcv. R package version 0.9.2.9013, https://gavinsimpson.github.io/gratia/.","code":"@Manual{,   title = {{gratia}: Graceful {ggplot}-Based Graphics and Other Functions                   for {GAM}s Fitted using {mgcv}},   author = {Gavin L. Simpson},   abstract = {Graceful ggplot-based graphics and utility functions for              working with generalized additive models (GAMs) fitted using the              mgcv package. Provides a reimplementation of the plot() method for              GAMs that mgcv provides, as well as tidyverse-compatible              representations of estimated smooths.},   note = {R package version 0.9.2.9013},   url = {https://gavinsimpson.github.io/gratia/}, }"},{"path":[]},{"path":"https://gavinsimpson.github.io/gratia/index.html","id":"overview","dir":"","previous_headings":"","what":"Overview","title":"An R package for working with generalized additive models","text":"Working GAMs within ‘tidyverse’ can tedious even difficult without good understanding GAMs model returned ‘mgcv’ model objects contain. ‘gratia’ designed help . ‘gratia’ provides ‘ggplot’-based graphics utility functions working generalized additive models (GAMs) fitted using ‘mgcv’ package, via reimplementation plot() method GAMs ‘mgcv’ provides, well ‘tidyverse’ compatible representations estimated smooths.","code":""},{"path":"https://gavinsimpson.github.io/gratia/index.html","id":"features","dir":"","previous_headings":"","what":"Features","title":"An R package for working with generalized additive models","text":"main features gratia currently ggplot2-based replacement mgcv:::plot.gam(): draw.gam(). example, classic four term additive example Gu & Wahba:  Estimated smooths GAM bivariate smooth:  Estimated smooths GAM Note specialist smoothers (bs %% c(\"mrf\",\"sw\", \"sf\")) currently supported, univariate, factor continuous -variable smooths, simple random effect smooths (bs = 're'), factor-smooth interaction smooths (bs = \"fs\"), constrained factor smooths (bs = \"sz\"), full soap film smooths (bs = \"\"), bivariate, trivariate, quadvariate TPRS tensor product smooths supported, Estimation derivatives fitted smoothers: derivatives(), Estimation point-wise across--function confidence intervals simultaneous intervals smooths: confint.gam(). Model diagnostics via appraise()  Model diagnostics figure","code":""},{"path":"https://gavinsimpson.github.io/gratia/index.html","id":"installing-gratia","dir":"","previous_headings":"","what":"Installing gratia","title":"An R package for working with generalized additive models","text":"gratia now available CRAN, can installed however gratia active development may wish install development version github. easiest way via install_github() function package remotes. Make sure remotes installed, run install package. Alternatively, binary packages development version available rOpenSci’s R Universe service:","code":"install.packages(\"gratia\") remotes::install_github(\"gavinsimpson/gratia\") # Install gratia in R install.packages(\"gratia\", repos = c(   \"https://gavinsimpson.r-universe.dev\",   \"https://cloud.r-project.org\" ))"},{"path":"https://gavinsimpson.github.io/gratia/index.html","id":"history","dir":"","previous_headings":"","what":"History","title":"An R package for working with generalized additive models","text":"gratia grew earlier package, schoenberg, development earlier package tsgam, originally intended used GAMs fitted time series. developing tsgam however became clear package used generally name “tsgam” longer appropriate. avoid breaking blog posts written using tsgam decided copy git repo history new repo package name schoenberg. later date someone released another package called schoenberg CRAN, scuppered idea. Now ’m calling package gratia. Hopefully won’t change …","code":""},{"path":"https://gavinsimpson.github.io/gratia/index.html","id":"why-gratia","dir":"","previous_headings":"","what":"Why gratia?","title":"An R package for working with generalized additive models","text":"naming greta package, Nick Golding observed recent phenomena naming statistical modelling software, Stan Edward, individuals played prominent role development field. lead Nick name Tensor Flow-based package greta Grete Hermann. spirit, gratia named recognition contributions Grace Wahba, pioneering work penalised spline models foundation way GAMs estimated mgcv. wanted name package grace, explicitly recognise Grace’s contributions, unfortunately already package named Grace CRAN. looked elsewhere inspiration. English word “grace” derives Latin gratia, meaning “favor, charm, thanks” (according Merriam Webster). chair Grace Wabha currently holds named Isaac J Schoenberg, former University Madison-Wisconsin Professor Mathematics, 1946 paper provided first mathematical reference “splines”. (Hence previous name package.)","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/add_confint.html","id":null,"dir":"Reference","previous_headings":"","what":"Add a confidence interval to an existing object — add_confint","title":"Add a confidence interval to an existing object — add_confint","text":"Add confidence interval existing object","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/add_confint.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Add a confidence interval to an existing object — add_confint","text":"","code":"add_confint(object, coverage = 0.95, ...)  # S3 method for class 'smooth_estimates' add_confint(object, coverage = 0.95, ...)  # S3 method for class 'parametric_effects' add_confint(object, coverage = 0.95, ...)  # Default S3 method add_confint(object, coverage = 0.95, ...)"},{"path":"https://gavinsimpson.github.io/gratia/reference/add_confint.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Add a confidence interval to an existing object — add_confint","text":"object R object. coverage numeric; coverage interval. Must range 0 < coverage < 1. ... arguments passed methods.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/add_constant.html","id":null,"dir":"Reference","previous_headings":"","what":"Add a constant to estimated values — add_constant","title":"Add a constant to estimated values — add_constant","text":"Add constant estimated values","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/add_constant.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Add a constant to estimated values — add_constant","text":"","code":"add_constant(object, constant = NULL, ...)  # S3 method for class 'smooth_estimates' add_constant(object, constant = NULL, ...)  # S3 method for class 'smooth_samples' add_constant(object, constant = NULL, ...)  # S3 method for class 'mgcv_smooth' add_constant(object, constant = NULL, ...)  # S3 method for class 'parametric_effects' add_constant(object, constant = NULL, ...)  # S3 method for class 'tbl_df' add_constant(object, constant = NULL, column = NULL, ...)  # S3 method for class 'evaluated_parametric_term' add_constant(object, constant = NULL, ...)"},{"path":"https://gavinsimpson.github.io/gratia/reference/add_constant.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Add a constant to estimated values — add_constant","text":"object object add constant . constant constant add. ... additional arguments passed methods. column character; \"tbl_df\" method, column add constant .","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/add_constant.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Add a constant to estimated values — add_constant","text":"Returns object estimate shifted addition supplied constant.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/add_constant.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Add a constant to estimated values — add_constant","text":"Gavin L. Simpson","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/add_fitted.gam.html","id":null,"dir":"Reference","previous_headings":"","what":"Add fitted values from a GAM to a data frame — add_fitted.gam","title":"Add fitted values from a GAM to a data frame — add_fitted.gam","text":"Add fitted values GAM data frame","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/add_fitted.gam.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Add fitted values from a GAM to a data frame — add_fitted.gam","text":"","code":"# S3 method for class 'gam' add_fitted(data, model, value = \".fitted\", type = \"response\", ...)"},{"path":"https://gavinsimpson.github.io/gratia/reference/add_fitted.gam.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Add fitted values from a GAM to a data frame — add_fitted.gam","text":"data data frame containing values variables used fit model. Passed stats::predict() newdata. model fitted model stats::predict() method available. S3 method dispatch performed model argument. value character; name variable model predictions stored. type character; type predictions return. See mgcv::predict.gam() options. ... additional arguments passed mgcv::predict.gam().","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/add_fitted.gam.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Add fitted values from a GAM to a data frame — add_fitted.gam","text":"data frame (tibble) formed data predictions model.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/add_fitted.gam.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Add fitted values from a GAM to a data frame — add_fitted.gam","text":"","code":"load_mgcv() df <- data_sim(\"eg1\", seed = 1) df <- df[, c(\"y\", \"x0\", \"x1\", \"x2\", \"x3\")] m <- gam(y ~ s(x0) + s(x1) + s(x2) + s(x3), data = df, method = \"REML\")  # add fitted values to our data add_fitted(df, m) #> # A tibble: 400 x 6 #>          y     x0     x1     x2    x3 .fitted #>      <dbl>  <dbl>  <dbl>  <dbl> <dbl>   <dbl> #>  1  3.34   0.266  0.659  0.859  0.367    5.90 #>  2 -0.0758 0.372  0.185  0.0344 0.741    3.15 #>  3 10.7    0.573  0.954  0.971  0.934    8.28 #>  4  8.73   0.908  0.898  0.745  0.673    8.65 #>  5 15.0    0.202  0.944  0.273  0.701   15.7  #>  6  7.67   0.898  0.724  0.677  0.848    8.38 #>  7  7.58   0.945  0.370  0.348  0.706    7.84 #>  8  8.51   0.661  0.781  0.947  0.859    6.74 #>  9 10.6    0.629  0.0111 0.339  0.446    9.14 #> 10  3.72   0.0618 0.940  0.0317 0.677    7.04 #> # i 390 more rows  # with type = \"terms\" or \"iterms\" add_fitted(df, m, type = \"terms\") #> # A tibble: 400 x 10 #>          y     x0     x1     x2    x3 .constant `s(x0)` `s(x1)` `s(x2)` `s(x3)` #>      <dbl>  <dbl>  <dbl>  <dbl> <dbl>     <dbl>   <dbl>   <dbl>   <dbl>   <dbl> #>  1  3.34   0.266  0.659  0.859  0.367      7.94  0.175    0.559  -2.81   0.0351 #>  2 -0.0758 0.372  0.185  0.0344 0.741      7.94  0.435   -1.92   -3.23  -0.0687 #>  3 10.7    0.573  0.954  0.971  0.934      7.94  0.593    3.35   -3.47  -0.122  #>  4  8.73   0.908  0.898  0.745  0.673      7.94 -0.812    2.77   -1.19  -0.0498 #>  5 15.0    0.202  0.944  0.273  0.701      7.94 -0.0589   3.23    4.63  -0.0576 #>  6  7.67   0.898  0.724  0.677  0.848      7.94 -0.745    1.15    0.146 -0.0981 #>  7  7.58   0.945  0.370  0.348  0.706      7.94 -1.07    -1.31    2.34  -0.0589 #>  8  8.51   0.661  0.781  0.947  0.859      7.94  0.434    1.67   -3.20  -0.101  #>  9 10.6    0.629  0.0111 0.339  0.446      7.94  0.512   -1.95    2.63   0.0132 #> 10  3.72   0.0618 0.940  0.0317 0.677      7.94 -0.695    3.20   -3.35  -0.0508 #> # i 390 more rows"},{"path":"https://gavinsimpson.github.io/gratia/reference/add_fitted.html","id":null,"dir":"Reference","previous_headings":"","what":"Add fitted values from a model to a data frame — add_fitted","title":"Add fitted values from a model to a data frame — add_fitted","text":"Add fitted values model data frame","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/add_fitted.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Add fitted values from a model to a data frame — add_fitted","text":"","code":"add_fitted(data, model, value = \".value\", ...)"},{"path":"https://gavinsimpson.github.io/gratia/reference/add_fitted.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Add fitted values from a model to a data frame — add_fitted","text":"data data frame containing values variables used fit model. Passed stats::predict() newdata. model fitted model stats::predict() method available. S3 method dispatch performed model argument. value character; name variable model predictions stored. ... additional arguments passed methods.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/add_fitted.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Add fitted values from a model to a data frame — add_fitted","text":"data frame (tibble) formed data fitted values model.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/add_fitted_samples.html","id":null,"dir":"Reference","previous_headings":"","what":"Add posterior draws from a model to a data object — add_fitted_samples","title":"Add posterior draws from a model to a data object — add_fitted_samples","text":"Adds draws posterior distribution model data object using one fitted_samples(), predicted_samples(), posterior_samples().","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/add_fitted_samples.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Add posterior draws from a model to a data object — add_fitted_samples","text":"","code":"add_fitted_samples(object, model, n = 1, seed = NULL, ...)  add_predicted_samples(object, model, n = 1, seed = NULL, ...)  add_posterior_samples(object, model, n = 1, seed = NULL, ...)  add_smooth_samples(object, model, n = 1, seed = NULL, select = NULL, ...)"},{"path":"https://gavinsimpson.github.io/gratia/reference/add_fitted_samples.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Add posterior draws from a model to a data object — add_fitted_samples","text":"object data frame tibble posterior draws added. model fitted GAM (GAM-like) object posterior draw method exists. n integer; number posterior draws add. seed numeric; value seed random number generator. ... arguments passed posterior draw function, currently one fitted_samples(), predicted_samples(), posterior_samples(). n seed already specified arguments also passed posterior sampling function. select character; select smooth's posterior draw . default, NULL, means posteriors smooths model wil sampled individually. supplied, character vector requested smooth terms.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/add_fitted_samples.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Add posterior draws from a model to a data object — add_fitted_samples","text":"","code":"load_mgcv()  df <- data_sim(\"eg1\", n = 400, seed = 42)  m <- gam(y ~ s(x0) + s(x1) + s(x2) + s(x3), data = df, method = \"REML\")  # add fitted samples (posterior draws of the expected value of the response) # note that there are 800 rows in the output: 400 data by `n = 2` samples. df |>   add_fitted_samples(m, n = 2, seed = 84) #> # A tibble: 800 × 14 #>        y    x0     x1    x2    x3     f    f0    f1     f2    f3  .row .draw #>    <dbl> <dbl>  <dbl> <dbl> <dbl> <dbl> <dbl> <dbl>  <dbl> <dbl> <int> <int> #>  1  2.99 0.915 0.0227 0.909 0.402  1.62 0.529  1.05 0.0397     0     1     1 #>  2  2.99 0.915 0.0227 0.909 0.402  1.62 0.529  1.05 0.0397     0     1     2 #>  3  4.70 0.937 0.513  0.900 0.432  3.25 0.393  2.79 0.0630     0     2     1 #>  4  4.70 0.937 0.513  0.900 0.432  3.25 0.393  2.79 0.0630     0     2     2 #>  5 13.9  0.286 0.631  0.192 0.664 13.5  1.57   3.53 8.41       0     3     1 #>  6 13.9  0.286 0.631  0.192 0.664 13.5  1.57   3.53 8.41       0     3     2 #>  7  5.71 0.830 0.419  0.532 0.182  6.12 1.02   2.31 2.79       0     4     1 #>  8  5.71 0.830 0.419  0.532 0.182  6.12 1.02   2.31 2.79       0     4     2 #>  9  7.63 0.642 0.879  0.522 0.838 10.4  1.80   5.80 2.76       0     5     1 #> 10  7.63 0.642 0.879  0.522 0.838 10.4  1.80   5.80 2.76       0     5     2 #> # ℹ 790 more rows #> # ℹ 2 more variables: .parameter <chr>, .fitted <dbl>  # add posterior draws from smooth s(x2) df |>   add_smooth_samples(m, n = 2, seed = 2, select= \"s(x2)\") #> # A tibble: 800 × 15 #>        y    x0     x1    x2    x3     f    f0    f1     f2    f3  .row .smooth #>    <dbl> <dbl>  <dbl> <dbl> <dbl> <dbl> <dbl> <dbl>  <dbl> <dbl> <int> <chr>   #>  1  2.99 0.915 0.0227 0.909 0.402  1.62 0.529  1.05 0.0397     0     1 s(x2)   #>  2  2.99 0.915 0.0227 0.909 0.402  1.62 0.529  1.05 0.0397     0     1 s(x2)   #>  3  4.70 0.937 0.513  0.900 0.432  3.25 0.393  2.79 0.0630     0     2 s(x2)   #>  4  4.70 0.937 0.513  0.900 0.432  3.25 0.393  2.79 0.0630     0     2 s(x2)   #>  5 13.9  0.286 0.631  0.192 0.664 13.5  1.57   3.53 8.41       0     3 s(x2)   #>  6 13.9  0.286 0.631  0.192 0.664 13.5  1.57   3.53 8.41       0     3 s(x2)   #>  7  5.71 0.830 0.419  0.532 0.182  6.12 1.02   2.31 2.79       0     4 s(x2)   #>  8  5.71 0.830 0.419  0.532 0.182  6.12 1.02   2.31 2.79       0     4 s(x2)   #>  9  7.63 0.642 0.879  0.522 0.838 10.4  1.80   5.80 2.76       0     5 s(x2)   #> 10  7.63 0.642 0.879  0.522 0.838 10.4  1.80   5.80 2.76       0     5 s(x2)   #> # ℹ 790 more rows #> # ℹ 3 more variables: .term <chr>, .draw <int>, .value <dbl>"},{"path":"https://gavinsimpson.github.io/gratia/reference/add_partial_residuals.html","id":null,"dir":"Reference","previous_headings":"","what":"Add partial residuals — add_partial_residuals","title":"Add partial residuals — add_partial_residuals","text":"Add partial residuals","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/add_partial_residuals.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Add partial residuals — add_partial_residuals","text":"","code":"add_partial_residuals(data, model, ...)  # S3 method for class 'gam' add_partial_residuals(data, model, select = NULL, partial_match = FALSE, ...)"},{"path":"https://gavinsimpson.github.io/gratia/reference/add_partial_residuals.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Add partial residuals — add_partial_residuals","text":"data data frame containing values variables used fit model. Passed stats::residuals() newdata. model fitted model stats::residuals() method available. S3 method dispatch performed model argument. ... arguments passed methods. select character, logical, numeric; smooths plot. NULL, default, model smooths drawn. Numeric select indexes smooths order specified formula stored object. Character select matches labels smooths shown example output summary(object). Logical select operates per numeric select order smooths stored. partial_match logical; smooths selected partial matches select? TRUE, select can single string match .","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/add_partial_residuals.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Add partial residuals — add_partial_residuals","text":"","code":"load_mgcv() df <- data_sim(\"eg1\", seed = 1) df <- df[, c(\"y\", \"x0\", \"x1\", \"x2\", \"x3\")] m <- gam(y ~ s(x0) + s(x1) + s(x2) + s(x3), data = df, method = \"REML\")  ## add partial residuals add_partial_residuals(df, m) #> # A tibble: 400 x 9 #>          y     x0     x1     x2    x3 `s(x0)` `s(x1)` `s(x2)` `s(x3)` #>      <dbl>  <dbl>  <dbl>  <dbl> <dbl>   <dbl>   <dbl>   <dbl>   <dbl> #>  1  3.34   0.266  0.659  0.859  0.367  -2.38   -2.00   -5.36  -2.52   #>  2 -0.0758 0.372  0.185  0.0344 0.741  -2.79   -5.15   -6.45  -3.29   #>  3 10.7    0.573  0.954  0.971  0.934   2.99    5.75   -1.07   2.28   #>  4  8.73   0.908  0.898  0.745  0.673  -0.734   2.84   -1.11   0.0287 #>  5 15.0    0.202  0.944  0.273  0.701  -0.752   2.54    3.94  -0.750  #>  6  7.67   0.898  0.724  0.677  0.848  -1.46    0.432  -0.567 -0.812  #>  7  7.58   0.945  0.370  0.348  0.706  -1.33   -1.57    2.08  -0.318  #>  8  8.51   0.661  0.781  0.947  0.859   2.21    3.44   -1.42   1.68   #>  9 10.6    0.629  0.0111 0.339  0.446   2.01   -0.445   4.13   1.51   #> 10  3.72   0.0618 0.940  0.0317 0.677  -4.02   -0.123  -6.67  -3.37   #> # i 390 more rows  ## add partial residuals for selected smooths add_partial_residuals(df, m, select = \"s(x0)\") #> # A tibble: 400 x 6 #>          y     x0     x1     x2    x3 `s(x0)` #>      <dbl>  <dbl>  <dbl>  <dbl> <dbl>   <dbl> #>  1  3.34   0.266  0.659  0.859  0.367  -2.38  #>  2 -0.0758 0.372  0.185  0.0344 0.741  -2.79  #>  3 10.7    0.573  0.954  0.971  0.934   2.99  #>  4  8.73   0.908  0.898  0.745  0.673  -0.734 #>  5 15.0    0.202  0.944  0.273  0.701  -0.752 #>  6  7.67   0.898  0.724  0.677  0.848  -1.46  #>  7  7.58   0.945  0.370  0.348  0.706  -1.33  #>  8  8.51   0.661  0.781  0.947  0.859   2.21  #>  9 10.6    0.629  0.0111 0.339  0.446   2.01  #> 10  3.72   0.0618 0.940  0.0317 0.677  -4.02  #> # i 390 more rows"},{"path":"https://gavinsimpson.github.io/gratia/reference/add_residuals.gam.html","id":null,"dir":"Reference","previous_headings":"","what":"Add residuals from a GAM to a data frame — add_residuals.gam","title":"Add residuals from a GAM to a data frame — add_residuals.gam","text":"Add residuals GAM data frame","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/add_residuals.gam.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Add residuals from a GAM to a data frame — add_residuals.gam","text":"","code":"# S3 method for class 'gam' add_residuals(data, model, value = \".residual\", type = \"deviance\", ...)"},{"path":"https://gavinsimpson.github.io/gratia/reference/add_residuals.gam.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Add residuals from a GAM to a data frame — add_residuals.gam","text":"data data frame containing values variables used fit model. Passed stats::predict() newdata. model fitted model stats::predict() method available. S3 method dispatch performed model argument. value character; name variable model predictions stored. type character; type residuals return. See mgcv::residuals.gam() options. ... additional arguments passed mgcv::residuals.gam().","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/add_residuals.gam.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Add residuals from a GAM to a data frame — add_residuals.gam","text":"data frame (tibble) formed data residuals model.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/add_residuals.gam.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Add residuals from a GAM to a data frame — add_residuals.gam","text":"","code":"load_mgcv() df <- data_sim(\"eg1\", seed = 1) df <- df[, c(\"y\", \"x0\", \"x1\", \"x2\", \"x3\")] m <- gam(y ~ s(x0) + s(x1) + s(x2) + s(x3), data = df, method = \"REML\")  ## add_residuals(df, m) #> # A tibble: 400 x 6 #>          y     x0     x1     x2    x3 .residual #>      <dbl>  <dbl>  <dbl>  <dbl> <dbl>     <dbl> #>  1  3.34   0.266  0.659  0.859  0.367   -2.56   #>  2 -0.0758 0.372  0.185  0.0344 0.741   -3.22   #>  3 10.7    0.573  0.954  0.971  0.934    2.40   #>  4  8.73   0.908  0.898  0.745  0.673    0.0785 #>  5 15.0    0.202  0.944  0.273  0.701   -0.693  #>  6  7.67   0.898  0.724  0.677  0.848   -0.714  #>  7  7.58   0.945  0.370  0.348  0.706   -0.259  #>  8  8.51   0.661  0.781  0.947  0.859    1.78   #>  9 10.6    0.629  0.0111 0.339  0.446    1.50   #> 10  3.72   0.0618 0.940  0.0317 0.677   -3.32   #> # i 390 more rows"},{"path":"https://gavinsimpson.github.io/gratia/reference/add_residuals.html","id":null,"dir":"Reference","previous_headings":"","what":"Add residuals from a model to a data frame — add_residuals","title":"Add residuals from a model to a data frame — add_residuals","text":"Add residuals model data frame","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/add_residuals.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Add residuals from a model to a data frame — add_residuals","text":"","code":"add_residuals(data, model, value = \".residual\", ...)"},{"path":"https://gavinsimpson.github.io/gratia/reference/add_residuals.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Add residuals from a model to a data frame — add_residuals","text":"data data frame containing values variables used fit model. Passed stats::residuals() newdata. model fitted model stats::residuals() method available. S3 method dispatch performed model argument. value character; name variable model residuals stored. ... additional arguments passed methods.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/add_residuals.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Add residuals from a model to a data frame — add_residuals","text":"data frame (tibble) formed data residuals model.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/add_sizer.html","id":null,"dir":"Reference","previous_headings":"","what":"Add indicators of significant change after SiZeR — add_sizer","title":"Add indicators of significant change after SiZeR — add_sizer","text":"Add indicators significant change SiZeR","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/add_sizer.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Add indicators of significant change after SiZeR — add_sizer","text":"","code":"add_sizer(object, type = c(\"change\", \"sizer\"), ...)  # S3 method for class 'derivatives' add_sizer(object, type = c(\"change\", \"sizer\"), ...)  # S3 method for class 'smooth_estimates' add_sizer(object, type = c(\"change\", \"sizer\"), derivatives = NULL, ...)"},{"path":"https://gavinsimpson.github.io/gratia/reference/add_sizer.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Add indicators of significant change after SiZeR — add_sizer","text":"object R object. Currently supported methods classes \"derivatives\". type character; \"change\" adds single variable object indicating credible interval derivative excludes 0. \"sizer\" adds two variables indicating whether derivative postive negative. ... arguments passed methods derivatives object class \"derivatives\", resulting call derivatives().","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/add_sizer.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Add indicators of significant change after SiZeR — add_sizer","text":"","code":"load_mgcv() df <- data_sim(\"eg1\", n = 400, dist = \"normal\", scale = 2, seed = 42) m <- gam(y ~ s(x0) + s(x1) + s(x2) + s(x3), data = df, method = \"REML\")  ## first derivatives of all smooths using central finite differences d <- derivatives(m, type = \"central\") |>   add_sizer()  # default adds a .change column names(d) #>  [1] \".smooth\"     \".by\"         \".fs\"         \".derivative\" \".se\"         #>  [6] \".crit\"       \".lower_ci\"   \".upper_ci\"   \".change\"     \"x0\"          #> [11] \"x1\"          \"x2\"          \"x3\""},{"path":"https://gavinsimpson.github.io/gratia/reference/appraise.html","id":null,"dir":"Reference","previous_headings":"","what":"Model diagnostic plots — appraise","title":"Model diagnostic plots — appraise","text":"Model diagnostic plots","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/appraise.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Model diagnostic plots — appraise","text":"","code":"appraise(model, ...)  # S3 method for class 'gam' appraise(   model,   method = c(\"uniform\", \"simulate\", \"normal\", \"direct\"),   use_worm = FALSE,   n_uniform = 10,   n_simulate = 50,   seed = NULL,   type = c(\"deviance\", \"pearson\", \"response\"),   n_bins = c(\"sturges\", \"scott\", \"fd\"),   ncol = NULL,   nrow = NULL,   guides = \"keep\",   level = 0.9,   ci_col = \"black\",   ci_alpha = 0.2,   point_col = \"black\",   point_alpha = 1,   line_col = \"red\",   ... )  # S3 method for class 'lm' appraise(model, ...)"},{"path":"https://gavinsimpson.github.io/gratia/reference/appraise.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Model diagnostic plots — appraise","text":"model fitted model. Currently models inheriting class \"gam\", well classes \"glm\" \"lm\" calls stats::glm stats::lm supported. ... arguments passed patchwork::wrap_plots(). method character; method used generate theoretical quantiles. default \"uniform\", generates reference quantiles using random draws uniform distribution inverse cummulative distribution function (CDF) fitted values. reference quantiles averaged n_uniform draws. \"simulate\" generates reference quantiles simulating new response data model observed values covariates, residualised generate reference quantiles, using n_simulate simulated data sets. \"normal\" generates reference quantiles using standard normal distribution. \"uniform\" computationally efficient, \"simulate\" allows reference bands drawn QQ-plot. \"normal\" avoided used fall back random number generator (\"simulate\") inverse CDF (\"uniform\"``) available family` used model fitting. Note method = \"direct\" deprecated favour method = \"uniform\". use_worm logical; worm plot drawn place QQ plot? n_uniform numeric; number times randomize uniform quantiles direct computation method (method = \"direct\") QQ plots. n_simulate numeric; number data sets simulate estimated model using simulation method (method = \"simulate\") QQ plots. seed numeric; random number seed use method = \"simulate\" method = \"uniform\". type character; type residuals use. \"deviance\", \"response\", \"pearson\" residuals allowed. n_bins character numeric; either number bins string indicating calculate number bins. ncol, nrow numeric; numbers rows columns spread plots. guides character; one \"keep\" (default), \"collect\", \"auto\". Passed patchwork::plot_layout() level numeric; coverage level QQ plot reference intervals. Must strictly 0 < level < 1. used method = \"simulate\". ci_alpha, ci_col colour transparency used draw QQ plot reference interval method = \"simulate\". point_col, point_alpha colour transparency used draw points plots. See graphics::par() section Color Specification. passed individual plotting functions, therefore affects points plots. line_col colour specification 1:1 line QQ plot reference line residuals vs linear predictor plot.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/appraise.html","id":"note","dir":"Reference","previous_headings":"","what":"Note","title":"Model diagnostic plots — appraise","text":"wording used mgcv::qq.gam() uses direct reference simulated residuals method (method = \"simulated\"). avoid confusion, method = \"direct\" deprecated favour method = \"uniform\".","code":""},{"path":[]},{"path":"https://gavinsimpson.github.io/gratia/reference/appraise.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Model diagnostic plots — appraise","text":"","code":"load_mgcv() ## simulate some data... dat <- data_sim(\"eg1\", n = 400, dist = \"normal\", scale = 2, seed = 2) mod <- gam(y ~ s(x0) + s(x1) + s(x2) + s(x3), data = dat) ## run some basic model checks appraise(mod, point_col = \"steelblue\", point_alpha = 0.4)   ## To change the theme for all panels use the & operator, for example to ## change the ggplot theme for all panels library(\"ggplot2\") appraise(mod, seed = 42,   point_col = \"steelblue\", point_alpha = 0.4,   line_col = \"black\" ) & theme_minimal()"},{"path":"https://gavinsimpson.github.io/gratia/reference/basis.html","id":null,"dir":"Reference","previous_headings":"","what":"Basis expansions for smooths — basis","title":"Basis expansions for smooths — basis","text":"Creates basis expansion definition smoother using syntax mgcv's smooths via mgcv::s()., mgcv::te(), mgcv::ti(), mgcv::t2(), fitted GAM(M).","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/basis.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Basis expansions for smooths — basis","text":"","code":"basis(object, ...)  # S3 method for class 'gam' basis(   object,   select = NULL,   term = deprecated(),   data = NULL,   n = 100,   n_2d = 50,   n_3d = 16,   n_4d = 4,   partial_match = FALSE,   ... )  # S3 method for class 'scam' basis(   object,   select = NULL,   term = deprecated(),   data = NULL,   n = 100,   n_2d = 50,   n_3d = 16,   n_4d = 4,   partial_match = FALSE,   ... )  # S3 method for class 'gamm' basis(   object,   select = NULL,   term = deprecated(),   data = NULL,   n = 100,   n_2d = 50,   n_3d = 16,   n_4d = 4,   partial_match = FALSE,   ... )  # S3 method for class 'list' basis(   object,   select = NULL,   term = deprecated(),   data = NULL,   n = 100,   n_2d = 50,   n_3d = 16,   n_4d = 4,   partial_match = FALSE,   ... )  # Default S3 method basis(   object,   data,   knots = NULL,   constraints = FALSE,   at = NULL,   diagonalize = FALSE,   ... )"},{"path":"https://gavinsimpson.github.io/gratia/reference/basis.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Basis expansions for smooths — basis","text":"object smooth specification, result call one mgcv::s()., mgcv::te(), mgcv::ti(), mgcv::t2(), fitted GAM(M) model. ... arguments passed mgcv::smoothCon(). select character; select smooths fitted model term argument renamed select data data frame containing variables used smooth. n numeric; number points range covariate evaluate smooth. n_2d numeric; number new observations dimension bivariate smooth. currently used; n used dimensions. n_3d numeric; number new observations generate third dimension 3D smooth. n_4d numeric; number new observations generate dimensions higher 2 (!) kD smooth (k >= 4). example, smooth 4D smooth, dimensions 3 4 get n_4d new observations. partial_match logical; case character select, select match partially smooths? partial_match = TRUE, select must single string, character vector length 1. knots list data frame named components containing knots locations. Names must match covariates basis required. See mgcv::smoothCon(). constraints logical; identifiability constraints applied smooth basis. See argument absorb.cons mgcv::smoothCon(). data frame containing values smooth covariate(s) basis evaluated. diagonalize logical; TRUE, reparameterises smooth associated penalty identity matrix. effect turning last diagonal elements penalty zero, highlights penalty null space.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/basis.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Basis expansions for smooths — basis","text":"tibble.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/basis.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Basis expansions for smooths — basis","text":"Gavin L. Simpson","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/basis.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Basis expansions for smooths — basis","text":"","code":"load_mgcv() df <- data_sim(\"eg4\", n = 400, seed = 42)  bf <- basis(s(x0), data = df) bf <- basis(s(x2, by = fac, bs = \"bs\"), data = df, constraints = TRUE)"},{"path":"https://gavinsimpson.github.io/gratia/reference/basis_size.html","id":null,"dir":"Reference","previous_headings":"","what":"Extract basis dimension of a smooth — basis_size","title":"Extract basis dimension of a smooth — basis_size","text":"Extract basis dimension smooth","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/basis_size.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Extract basis dimension of a smooth — basis_size","text":"","code":"basis_size(object, ...)  # S3 method for class 'mgcv.smooth' basis_size(object, ...)  # S3 method for class 'gam' basis_size(object, ...)  # S3 method for class 'gamm' basis_size(object, ...)"},{"path":"https://gavinsimpson.github.io/gratia/reference/basis_size.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Extract basis dimension of a smooth — basis_size","text":"object fitted GAM(M). Currently mgcv::gam() (anything inherits \"gam\" class, e.g. mgcv::bam()) mgcv::gamm() supported. ... Arguments passed methods.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/basis_size.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Extract basis dimension of a smooth — basis_size","text":"","code":"load_mgcv()  df <- data_sim(\"eg1\", n = 200, seed = 1) m <- bam(y ~ s(x0) + s(x1) + s(x2) + s(x3), data = df)  basis_size(m) #> s(x0) s(x1) s(x2) s(x3)  #>     9     9     9     9"},{"path":"https://gavinsimpson.github.io/gratia/reference/bird_move.html","id":null,"dir":"Reference","previous_headings":"","what":"Simulated bird migration data — bird_move","title":"Simulated bird migration data — bird_move","text":"Data generated hypothetical study bird movement along migration corridor, sampled throughout year. dataset consists simulated sample records numbers observed locations 100 tagged individuals six species bird, ten locations along latitudinal gradient, one observation taken every four weeks. Counts simulated randomly species location week creating species-specific migration curve gave probability finding individual given species given location, simulated distribution individuals across sites using multinomial distribution, subsampling using binomial distribution simulation observation error (.e. every bird present location detected). data set (bird_move) consists variables count, latitude, week species.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/bird_move.html","id":"format","dir":"Reference","previous_headings":"","what":"Format","title":"Simulated bird migration data — bird_move","text":"data frame","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/bird_move.html","id":"source","dir":"Reference","previous_headings":"","what":"Source","title":"Simulated bird migration data — bird_move","text":"Pedersen EJ, Miller DL, Simpson GL, Ross N. 2018. Hierarchical generalized additive models: introduction mgcv. PeerJ Preprints 6:e27320v1 doi:10.7287/peerj.preprints.27320v1 .","code":""},{"path":[]},{"path":"https://gavinsimpson.github.io/gratia/reference/boundary.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Extract the boundary of a soap film smooth — boundary","text":"","code":"boundary(x, ...)  # S3 method for class 'soap.film' boundary(x, ...)  # S3 method for class 'gam' boundary(x, select, ...)"},{"path":"https://gavinsimpson.github.io/gratia/reference/boundary.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Extract the boundary of a soap film smooth — boundary","text":"x R object. Currently objects inherit classes \"soap.film\" \"gam\". ... arguments passed methods. select character; label soap film smooth extract boundary.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/boundary.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Extract the boundary of a soap film smooth — boundary","text":"list lists data frames specifying loops define boundary soap film smooth.","code":""},{"path":[]},{"path":"https://gavinsimpson.github.io/gratia/reference/check_user_select_smooths.html","id":null,"dir":"Reference","previous_headings":"","what":"Select smooths based on user's choices — check_user_select_smooths","title":"Select smooths based on user's choices — check_user_select_smooths","text":"Given vector indexing smooths GAM, returns logical vector selecting requested smooths.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/check_user_select_smooths.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Select smooths based on user's choices — check_user_select_smooths","text":"","code":"check_user_select_smooths(   smooths,   select = NULL,   partial_match = FALSE,   model_name = NULL )"},{"path":"https://gavinsimpson.github.io/gratia/reference/check_user_select_smooths.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Select smooths based on user's choices — check_user_select_smooths","text":"smooths character; vector smooth labels. select numeric, logical, character vector selected smooths. partial_match logical; case character select, select match partially smooths? partial_match = TRUE, select must single string, character vector length 1. model_name character; model name used error messages.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/check_user_select_smooths.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Select smooths based on user's choices — check_user_select_smooths","text":"logical vector length length(smooths) indicating smooths selected.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/check_user_select_smooths.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Select smooths based on user's choices — check_user_select_smooths","text":"Gavin L. Simpson","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/coef.scam.html","id":null,"dir":"Reference","previous_headings":"","what":"Extract coefficients from a fitted scam model. — coef.scam","title":"Extract coefficients from a fitted scam model. — coef.scam","text":"Extract coefficients fitted scam model.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/coef.scam.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Extract coefficients from a fitted scam model. — coef.scam","text":"","code":"# S3 method for class 'scam' coef(object, parametrized = TRUE, ...)"},{"path":"https://gavinsimpson.github.io/gratia/reference/coef.scam.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Extract coefficients from a fitted scam model. — coef.scam","text":"object model object fitted scam() parametrized logical; extract parametrized coefficients, respect linear inequality constraints model. ... arguments.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/compare_smooths.html","id":null,"dir":"Reference","previous_headings":"","what":"Compare smooths across models — compare_smooths","title":"Compare smooths across models — compare_smooths","text":"Compare smooths across models","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/compare_smooths.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Compare smooths across models — compare_smooths","text":"","code":"compare_smooths(   model,   ...,   select = NULL,   smooths = deprecated(),   n = 100,   data = NULL,   unconditional = FALSE,   overall_uncertainty = TRUE,   partial_match = FALSE )"},{"path":"https://gavinsimpson.github.io/gratia/reference/compare_smooths.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Compare smooths across models — compare_smooths","text":"model Primary model comparison. ... Additional models compare smooths model. select character; select smooths compare. default (NULL) means smooths model compared. Numeric select indexes smooths order specified formula stored model. Character select matches labels smooths shown example output summary(object). Logical select operates per numeric select order smooths stored. smooths Use select instead. n numeric; number points range covariate evaluate smooth. data data frame covariate values evaluate smooth. unconditional logical; confidence intervals include uncertainty due smoothness selection? TRUE, corrected Bayesian covariance matrix used. overall_uncertainty logical; uncertainty model constant term included standard error evaluate values smooth? partial_match logical; smooths selected partial matches select? TRUE, select can single string match .","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/compare_smooths.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Compare smooths across models — compare_smooths","text":"","code":"load_mgcv() dat <- data_sim(\"eg1\", seed = 2)  ## models to compare smooths across - artificially create differences m1 <- gam(y ~ s(x0, k = 5) + s(x1, k = 5) + s(x2, k = 5) + s(x3, k = 5),   data = dat, method = \"REML\" ) m2 <- gam(y ~ s(x0, bs = \"ts\") + s(x1, bs = \"ts\") + s(x2, bs = \"ts\") +   s(x3, bs = \"ts\"), data = dat, method = \"REML\")  ## build comparisons comp <- compare_smooths(m1, m2) comp #> # A tibble: 8 x 5 #>   .model .smooth .type         .by   data               #>   <chr>  <chr>   <chr>         <chr> <list>             #> 1 m1     s(x0)   TPRS          NA    <tibble [100 x 3]> #> 2 m2     s(x0)   TPRS (shrink) NA    <tibble [100 x 3]> #> 3 m1     s(x1)   TPRS          NA    <tibble [100 x 3]> #> 4 m2     s(x1)   TPRS (shrink) NA    <tibble [100 x 3]> #> 5 m1     s(x2)   TPRS          NA    <tibble [100 x 3]> #> 6 m2     s(x2)   TPRS (shrink) NA    <tibble [100 x 3]> #> 7 m1     s(x3)   TPRS          NA    <tibble [100 x 3]> #> 8 m2     s(x3)   TPRS (shrink) NA    <tibble [100 x 3]> ## notice that the result is a nested tibble  draw(comp)"},{"path":"https://gavinsimpson.github.io/gratia/reference/conditional_values.html","id":null,"dir":"Reference","previous_headings":"","what":"Conditional predictions from a GAM — conditional_values","title":"Conditional predictions from a GAM — conditional_values","text":"Generate predicted values GAM, conditional upon supplied values covariates. conditional_values() modelled marginaleffects::plot_predictions(), intentionally simpler, restrictive functionality. intended use case quickly visualizing predicted values fitted GAM response scale. complex model predictions, strongly encouraged use marginaleffects::plot_predictions().","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/conditional_values.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Conditional predictions from a GAM — conditional_values","text":"","code":"conditional_values(   model,   condition = NULL,   data = NULL,   scale = c(\"response\", \"link\", \"linear_predictor\"),   ... )  # S3 method for class 'gam' conditional_values(   model,   condition = NULL,   data = NULL,   scale = c(\"response\", \"link\", \"linear_predictor\"),   n_vals = 100,   ci_level = 0.95,   ... )"},{"path":"https://gavinsimpson.github.io/gratia/reference/conditional_values.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Conditional predictions from a GAM — conditional_values","text":"model fitted GAM object. condition either character vector list supplying names covariates, possibly values, condition . order values determines plotted via draw.conditional_values() method; first element mapped x channel, second element colour channel, third ggplot2::facet_wrap() fourth element present, present, fourth element mapped rows third element mapped columns ggplot2::facet_grid(). data data frame values predict. supplied overrides values supplied condition. scale character; scale predictions returned ? ... arguments passed fitted_values(). n_vals numeric; number values generate numeric variables named condition. ci_level numeric; number interval (0,1) giving coverage credible intervals.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/conditional_values.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Conditional predictions from a GAM — conditional_values","text":"data frame (tibble) class \"conditional_values\".","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/conditional_values.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Conditional predictions from a GAM — conditional_values","text":"Gavin L. Simpson","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/conditional_values.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Conditional predictions from a GAM — conditional_values","text":"","code":"load_mgcv() df <- data_sim(\"eg1\", seed = 2) m1 <- gam(y ~ s(x0) + s(x1) + s(x2) + s(x3), data = df, method = \"REML\")  # predictions conditional on values evenly spaced over x2, all other # variables in model are held at representative values cv <- conditional_values(   m1,   condition = \"x2\" ) # plot cv |> draw()   # as above but condition on `x1` also. When plotted, `x1` is mapped to the # colour channel, noting that it has been summarised using fivenum() cv <- conditional_values(   m1,   condition = c(\"x2\", \"x1\") ) # plot cv |> draw()   # can pass `condition` a list, allowing for greater flexibility # For example, here we condition on all four variables in the model, # summarising: #   * `x1` at its five number summary, #   * `x0  at its quartiles #   * `x3` at its mean a d mean +/- sd cv <- conditional_values(   m1,   condition = list(\"x2\", x1 = \"fivenum\", x0 = \"quartile\", x3 = \"threenum\") ) # plot cv |> draw()   # some model terms can be exclude from the conditional predictions using the # `exclude` mechanism of `predict.gam`. Here we exclude the effects of # `s(x0)` and `s(x3)` from the conditional predictions. This, in effect, # treats these smooths as having **0** effect on the conditional predictions # of the response, even though the two smooths conditioned on (`s(x2)` and # `s(x1)`) were estimated given the two excluded smooths were in the model cv <- conditional_values(   m1,   condition = list(\"x2\", x1 = \"minmax\"),   exclude = c(\"s(x0)\", \"s(x3)\") ) # plot cv |> draw()   # categorical conditions are also handled df <- data_sim(\"eg4\", seed = 2) m2 <- gam(y ~ fac + s(x2, by = fac) + s(x0), data = df, method = \"REML\") cv <- conditional_values(   m2,   condition = list(\"fac\", x2 = \"fivenum\") ) # plot - we see a discrete x axis cv |> draw()   # in this example we condition on `x2` and `fac %in% c(2,3)` cv <- conditional_values(   m2,   condition = list(\"x2\", fac = 2:3) ) # plot - smooths of `x2` for `fac == 2` and `fac == 3` cv |> draw()"},{"path":"https://gavinsimpson.github.io/gratia/reference/confint.fderiv.html","id":null,"dir":"Reference","previous_headings":"","what":"Point-wise and simultaneous confidence intervals for derivatives of smooths — confint.fderiv","title":"Point-wise and simultaneous confidence intervals for derivatives of smooths — confint.fderiv","text":"Calculates point-wise confidence simultaneous intervals first derivatives smooth terms fitted GAM.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/confint.fderiv.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Point-wise and simultaneous confidence intervals for derivatives of smooths — confint.fderiv","text":"","code":"# S3 method for class 'fderiv' confint(   object,   parm,   level = 0.95,   type = c(\"confidence\", \"simultaneous\"),   nsim = 10000,   ncores = 1L,   ... )"},{"path":"https://gavinsimpson.github.io/gratia/reference/confint.fderiv.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Point-wise and simultaneous confidence intervals for derivatives of smooths — confint.fderiv","text":"object object class \"fderiv\" containing estimated derivatives. parm parameters (smooth terms) given intervals vector terms. missing, parameters considered. level numeric, 0 < level < 1; confidence level point-wise simultaneous interval. default 0.95 95% interval. type character; type interval compute. One \"confidence\" point-wise intervals, \"simultaneous\" simultaneous intervals. nsim integer; number simulations used computing simultaneous intervals. ncores number cores generating random variables multivariate normal distribution. Passed mvnfast::rmvn(). Parallelization take place OpenMP supported (appears work Windows current R). ... additional arguments methods","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/confint.fderiv.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Point-wise and simultaneous confidence intervals for derivatives of smooths — confint.fderiv","text":"data frame components: term; factor indicating term row relates, lower; lower limit confidence simultaneous interval, est; estimated derivative upper; upper limit confidence simultaneous interval.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/confint.fderiv.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Point-wise and simultaneous confidence intervals for derivatives of smooths — confint.fderiv","text":"Gavin L. Simpson","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/confint.fderiv.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Point-wise and simultaneous confidence intervals for derivatives of smooths — confint.fderiv","text":"","code":"load_mgcv() dat <- data_sim(\"eg1\", n = 1000, dist = \"normal\", scale = 2, seed = 2) mod <- gam(y ~ s(x0) + s(x1) + s(x2) + s(x3), data = dat, method = \"REML\")  # new data to evaluate the derivatives at, say over the middle 50% of range # of each covariate middle <- function(x, n = 25, coverage = 0.5) {   v <- (1 - coverage) / 2   q <- quantile(x, prob = c(0 + v, 1 - v), type = 8)   seq(q[1], q[2], length = n) } new_data <- sapply(dat[c(\"x0\", \"x1\", \"x2\", \"x3\")], middle) new_data <- data.frame(new_data) ## first derivatives of all smooths... fd <- fderiv(mod, newdata = new_data) #> Warning: `fderiv()` was deprecated in gratia 0.7.0. #> i Please use `derivatives()` instead.  ## point-wise interval ci <- confint(fd, type = \"confidence\") ci #> # A tibble: 100 x 4 #>    term    lower   est upper #>    <chr>   <dbl> <dbl> <dbl> #>  1 s(x0)  1.7     4.1    6.6 #>  2 s(x0)  1.3     3.8    6.3 #>  3 s(x0)  0.99    3.5    6.0 #>  4 s(x0)  0.68    3.1    5.6 #>  5 s(x0)  0.37    2.8    5.2 #>  6 s(x0)  0.0049  2.4    4.8 #>  7 s(x0) -0.40    2.0    4.5 #>  8 s(x0) -0.79    1.7    4.2 #>  9 s(x0) -1.1     1.3    3.8 #> 10 s(x0) -1.4     0.99   3.4 #> # i 90 more rows  ## simultaneous interval for smooth term of x2 x2_sint <- confint(fd,   parm = \"x2\", type = \"simultaneous\",   nsim = 10000, ncores = 2 ) # \\donttest{ x2_sint #> # A tibble: 25 x 4 #>    term  lower   est upper #>    <chr> <dbl> <dbl> <dbl> #>  1 s(x2)  -24.  -15.  -5.6 #>  2 s(x2)  -35.  -26. -16.  #>  3 s(x2)  -41.  -33. -24.  #>  4 s(x2)  -44.  -36. -29.  #>  5 s(x2)  -44.  -36. -28.  #>  6 s(x2)  -42.  -34. -25.  #>  7 s(x2)  -38.  -30. -21.  #>  8 s(x2)  -33.  -24. -16.  #>  9 s(x2)  -27.  -19. -11.  #> 10 s(x2)  -22.  -14.  -5.8 #> # i 15 more rows # }"},{"path":"https://gavinsimpson.github.io/gratia/reference/confint.gam.html","id":null,"dir":"Reference","previous_headings":"","what":"Point-wise and simultaneous confidence intervals for smooths — confint.gam","title":"Point-wise and simultaneous confidence intervals for smooths — confint.gam","text":"Calculates point-wise confidence simultaneous intervals smooth terms fitted GAM.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/confint.gam.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Point-wise and simultaneous confidence intervals for smooths — confint.gam","text":"","code":"# S3 method for class 'gam' confint(   object,   parm,   level = 0.95,   data = newdata,   n = 100,   type = c(\"confidence\", \"simultaneous\"),   nsim = 10000,   shift = FALSE,   transform = FALSE,   unconditional = FALSE,   ncores = 1,   partial_match = FALSE,   ...,   newdata = NULL )  # S3 method for class 'gamm' confint(object, ...)  # S3 method for class 'list' confint(object, ...)"},{"path":"https://gavinsimpson.github.io/gratia/reference/confint.gam.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Point-wise and simultaneous confidence intervals for smooths — confint.gam","text":"object object class \"gam\" \"gamm\". parm parameters (smooth terms) given intervals vector terms. missing, parameters considered, although currently implemented. level numeric, 0 < level < 1; confidence level point-wise simultaneous interval. default 0.95 95% interval. data data frame; new values covariates used model fit. selected smooth(s) wil evaluated supplied values. n numeric; number points evaluate smooths . type character; type interval compute. One \"confidence\" point-wise intervals, \"simultaneous\" simultaneous intervals. nsim integer; number simulations used computing simultaneous intervals. shift logical; constant term add smooth? transform logical; smooth evaluated transformed scale? generalised models, involves applying inverse link function used fit model. Alternatively, name , actual, function can supplied transform smooth confidence interval. unconditional logical; TRUE (freq == FALSE) Bayesian smoothing parameter uncertainty corrected covariance matrix returned, available. ncores number cores generating random variables multivariate normal distribution. Passed mvnfast::rmvn(). Parallelization take place OpenMP supported (appears work Windows current R). partial_match logical; matching parm use partial match exact match? Can used length(parm) 1. ... additional arguments methods newdata DEPRECATED! data frame; containing new values covariates used model fit. selected smooth(s) wil evaluated supplied values.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/confint.gam.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Point-wise and simultaneous confidence intervals for smooths — confint.gam","text":"tibble components: .smooth; character indicating term row relates, .type; type smooth, .name variable smooth, NA otherwise, one vectors values smooth evaluated, named per variables smooth, zero variables containing values variable, .estimate; estimated value smooth, .se; standard error estimated value smooth, .crit; critical value 100 * level% confidence interval. .lower_ci; lower limit confidence simultaneous interval, .upper_ci; upper limit confidence simultaneous interval,","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/confint.gam.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Point-wise and simultaneous confidence intervals for smooths — confint.gam","text":"Gavin L. Simpson","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/confint.gam.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Point-wise and simultaneous confidence intervals for smooths — confint.gam","text":"","code":"load_mgcv() dat <- data_sim(\"eg1\", n = 1000, dist = \"normal\", scale = 2, seed = 2) mod <- gam(y ~ s(x0) + s(x1) + s(x2) + s(x3), data = dat, method = \"REML\")  # new data to evaluate the smooths at, say over the middle 50% of range # of each covariate middle <- function(x, n = 50, coverage = 0.5) {   v <- (1 - coverage) / 2   q <- quantile(x, prob = c(0 + v, 1 - v), type = 8)   seq(q[1], q[2], length = n) } new_data <- sapply(dat[c(\"x0\", \"x1\", \"x2\", \"x3\")], middle) new_data <- data.frame(new_data)  ## point-wise interval for smooth of x2 ci <- confint(mod, parm = \"s(x2)\", type = \"confidence\", data = new_data) ci #> # A tibble: 50 x 9 #>    .smooth .type .by      x2 .estimate   .se .crit .lower_ci .upper_ci #>    <chr>   <chr> <chr> <dbl>     <dbl> <dbl> <dbl>     <dbl>     <dbl> #>  1 s(x2)   TPRS  NA     0.26       5.3  0.18   2.0       5.0       5.7 #>  2 s(x2)   TPRS  NA     0.27       5.1  0.18   2.0       4.8       5.5 #>  3 s(x2)   TPRS  NA     0.28       4.9  0.18   2.0       4.6       5.3 #>  4 s(x2)   TPRS  NA     0.29       4.6  0.18   2.0       4.3       5.0 #>  5 s(x2)   TPRS  NA     0.30       4.3  0.19   2.0       3.9       4.7 #>  6 s(x2)   TPRS  NA     0.32       4.0  0.19   2.0       3.6       4.3 #>  7 s(x2)   TPRS  NA     0.33       3.6  0.20   2.0       3.2       4.0 #>  8 s(x2)   TPRS  NA     0.34       3.2  0.20   2.0       2.9       3.6 #>  9 s(x2)   TPRS  NA     0.35       2.9  0.20   2.0       2.5       3.3 #> 10 s(x2)   TPRS  NA     0.36       2.5  0.19   2.0       2.1       2.9 #> # i 40 more rows"},{"path":"https://gavinsimpson.github.io/gratia/reference/data_combos.html","id":null,"dir":"Reference","previous_headings":"","what":"All combinations of factor levels plus typical values of continuous variables — data_combos","title":"All combinations of factor levels plus typical values of continuous variables — data_combos","text":"combinations factor levels plus typical values continuous variables","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/data_combos.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"All combinations of factor levels plus typical values of continuous variables — data_combos","text":"","code":"data_combos(object, ...)  # S3 method for class 'gam' data_combos(   object,   vars = everything(),   complete = TRUE,   envir = environment(formula(object)),   data = NULL,   ... )"},{"path":"https://gavinsimpson.github.io/gratia/reference/data_combos.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"All combinations of factor levels plus typical values of continuous variables — data_combos","text":"object fitted model object. ... arguments passed methods. vars terms include exclude returned object. Uses tidyselect principles. complete logical; combinations factor levels returned? FALSE, combinations levels observed model retained. envir environment within recreate data used fit object. data optional data frame data used fit mdoel reconstruction data model work.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/data_sim.html","id":null,"dir":"Reference","previous_headings":"","what":"Simulate example data for fitting GAMs — data_sim","title":"Simulate example data for fitting GAMs — data_sim","text":"tidy reimplementation functions implemented mgcv::gamSim() can used fit GAMs. new feature sampling distribution can applied example types.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/data_sim.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Simulate example data for fitting GAMs — data_sim","text":"","code":"data_sim(   model = \"eg1\",   n = 400,   scale = NULL,   theta = 3,   power = 1.5,   dist = c(\"normal\", \"poisson\", \"binary\", \"negbin\", \"tweedie\", \"gamma\", \"ocat\",     \"ordered categorical\"),   n_cat = 4,   cuts = c(-1, 0, 5),   seed = NULL,   gfam_families = c(\"binary\", \"tweedie\", \"normal\") )"},{"path":"https://gavinsimpson.github.io/gratia/reference/data_sim.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Simulate example data for fitting GAMs — data_sim","text":"model character; either \"egX\" X integer 1:7, name model. See Details possible options. n numeric; number observations simulate. scale numeric; level noise use. theta numeric; dispersion parameter \\(\\theta\\) use. default entirely arbitrary, chosen provide simulated data exhibits extra dispersion beyond assumed Poisson. power numeric; Tweedie power parameter. dist character; sampling distribution response variable. \"ordered categorical\" synonym \"ocat\". n_cat integer; number categories categorical response. Currently used distr %% c(\"ocat\", \"ordered categorical\"). cuts numeric; vector cut points latent variable, excluding end points -Inf Inf. Must one fewer number categories: length(cuts) == n_cat - 1. seed numeric; seed random number generator. Passed base::set.seed(). gfam_families character; vector distributions use generating data grouped families use family = gfam(). allowed distributions per dist.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/data_sim.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Simulate example data for fitting GAMs — data_sim","text":"data_sim() can simulate data several underlying models known true functions. available options currently : \"eg1\": four term additive true model. classic Gu & Wahba four univariate term test model. See gw_functions details underlying four functions. \"eg2\": bivariate smooth true model. \"eg3\": example containing continuous smooth (varying coefficient) true model. model \\(\\hat{y}_i = f_2(x_{1i})x_{2i}\\) function \\(f_2()\\) \\(f_2(x) = 0.2 * x^{11} * (10 * (1 - x))^6 + 10 * (10 * x)^3 * (1 - x)^{10}\\). \"eg4\": factor smooth true model. true model contains factor 3 levels, response nth level follows nth Gu & Wabha function (\\(n \\{1, 2, 3}\\)). \"eg5\": additive plus factor true model. response linear combination Gu & Wabha functions 2, 3, 4 (latter null function) plus factor term four levels. \"eg6\": additive plus random effect term true model. ´\"eg7\": version model \"eg1\"`, covariates correlated. \"gwf2\": model response Gu & Wabha's \\(f_2(x_i)\\) plus noise. \"lwf6\": model response Luo & Wabha's \"example 6\" function \\(sin(2(4x-2)) + 2 exp(-256(x-0.5)^2)\\) plus noise. \"gfam\": simulates data use GAMs family = gfam(families). See example mgcv::gfam(). model specified dist ignored gfam_families used specify distributions included simulated data. Can vector families allowed dist. \"ocat\" %% gfam_families (\"ordered categorical\"), 4 classes assumed, changed. Link functions used \"identity\" \"normal\", \"logit\" \"binary\", \"ocat\", \"ordered categorical\", \"exp\" elsewhere. random component providing noise sampling variation can follow one distributions, specified via argument dist \"normal\": Gaussian, \"poisson\": Poisson, \"binary\": Bernoulli, \"negbin\": Negative binomial, \"tweedie\": Tweedie, \"gamma\": gamma , \"ordered categorical\": ordered categorical arguments provide parameters distribution.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/data_sim.html","id":"references","dir":"Reference","previous_headings":"","what":"References","title":"Simulate example data for fitting GAMs — data_sim","text":"Gu, C., Wahba, G., (1993). Smoothing Spline ANOVA Component-Wise Bayesian \"Confidence Intervals.\" J. Comput. Graph. Stat. 2, 97–117. Luo, Z., Wahba, G., (1997). Hybrid adaptive splines. J. . Stat. Assoc. 92, 107–116.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/data_sim.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Simulate example data for fitting GAMs — data_sim","text":"","code":"data_sim(\"eg1\", n = 100, seed = 1) #> # A tibble: 100 x 10 #>          y       x0      x1      x2       x3       f      f0     f1     f2    f3 #>      <dbl>    <dbl>   <dbl>   <dbl>    <dbl>   <dbl>   <dbl>  <dbl>  <dbl> <dbl> #>  1 14.532  0.26551  0.65472 0.26751 0.67371  13.713  1.4814  3.7041 8.5277     0 #>  2 16.113  0.37212  0.35320 0.21865 0.094858 12.735  1.8408  2.0267 8.8680     0 #>  3  9.5835 0.57285  0.27026 0.51680 0.49260   6.4103 1.9478  1.7169 2.7456     0 #>  4 15.687  0.90821  0.99268 0.26895 0.46155  16.349  0.56879 7.2817 8.4980     0 #>  5  8.2216 0.20168  0.63349 0.18117 0.37522  12.792  1.1841  3.5501 8.0578     0 #>  6  9.9034 0.89839  0.21321 0.51858 0.99110   4.9081 0.62765 1.5318 2.7487     0 #>  7  5.9362 0.94468  0.12937 0.56278 0.17635   4.6020 0.34587 1.2953 2.9609     0 #>  8 10.839  0.66080  0.47812 0.12916 0.81344   9.7565 1.7502  2.6019 5.4045     0 #>  9 16.883  0.62911  0.92407 0.25637 0.068447 16.909  1.8377  6.3481 8.7237     0 #> 10  7.3603 0.061786 0.59876 0.71794 0.40045   6.3401 0.38578 3.3119 2.6424     0 #> # i 90 more rows  # an ordered categorical response data_sim(\"eg1\", n = 100, dist = \"ocat\", n_cat = 4, cuts = c(-1, 0, 5)) #> # A tibble: 100 x 11 #>        y      x0        x1      x2       x3         f      f0     f1         f2 #>    <int>   <dbl>     <dbl>   <dbl>    <dbl>     <dbl>   <dbl>  <dbl>      <dbl> #>  1     1 0.93708 0.21716   0.51711 0.44457  -3.5517   0.39280 1.5439 2.7461     #>  2     1 0.28614 0.21657   0.85193 0.060386 -4.7654   1.5653  1.5421 0.36166    #>  3     1 0.83045 0.38895   0.44280 0.32751  -1.7693   1.0157  2.1769 3.2727     #>  4     4 0.64175 0.94246   0.15788 0.87843   7.2150   1.8050  6.5858 7.0588     #>  5     3 0.51910 0.96261   0.44232 0.93060   3.8994   1.9964  6.8566 3.2808     #>  6     1 0.73659 0.73986   0.96773 0.39218  -2.3701   1.4725  4.3917 0.00015734 #>  7     1 0.13467 0.73325   0.48459 0.15885  -0.27657  0.82112 4.3340 2.8028     #>  8     3 0.65699 0.53576   0.25246 0.31995   5.2247   1.7616  2.9198 8.7777     #>  9     3 0.70506 0.0022730 0.25969 0.30697   3.0408   1.5991  1.0046 8.6716     #> 10     2 0.45774 0.60894   0.54202 0.10781  -0.036524 1.9824  3.3800 2.8356     #> # i 90 more rows #> # i 2 more variables: f3 <dbl>, latent <dbl>"},{"path":"https://gavinsimpson.github.io/gratia/reference/data_slice.html","id":null,"dir":"Reference","previous_headings":"","what":"Prepare a data slice through model covariates — data_slice","title":"Prepare a data slice through model covariates — data_slice","text":"Prepare data slice model covariates","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/data_slice.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Prepare a data slice through model covariates — data_slice","text":"","code":"data_slice(object, ...)  # Default S3 method data_slice(object, ...)  # S3 method for class 'data.frame' data_slice(object, ...)  # S3 method for class 'gam' data_slice(object, ..., data = NULL, envir = NULL)  # S3 method for class 'gamm' data_slice(object, ...)  # S3 method for class 'list' data_slice(object, ...)  # S3 method for class 'scam' data_slice(object, ...)"},{"path":"https://gavinsimpson.github.io/gratia/reference/data_slice.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Prepare a data slice through model covariates — data_slice","text":"object R model object. ... <dynamic-dots> User supplied variables defining data slice. Arguments passed via ... need named. data alternative data frame values containing variables needed fit model. NULL, default, data used fit model recovered using model.frame. User-supplied expressions passed ... evaluated data. envir environment within recreate data used fit object.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/data_slice.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Prepare a data slice through model covariates — data_slice","text":"data slice data set results one (covariates) varied systematically () range specified subset values interest, remaining covariates model held fixed, representative values. known reference grid package emmeans data grid marginaleffects package. GAMs, covariates specified via ... take representative values determined data used fit model follows: numeric covariates, value fitting data closest median value used, factor covariates, modal (frequently observed) level used, first level (sorted per vector returned base::levels() several levels observed number times. values already computed calling gam() bam() example can found var.summary component fitted model. Function typical_values() extract values interested. Convenience functions evenly(), ref_level(), level() provided help users specify data slices. ref_level(), level() also ensure factor covariates correct levels, needed mgcv::predict.gam() example. extended discussion data_slice() examples, see vignette(\"data-slices\", package = \"gratia\").","code":""},{"path":[]},{"path":"https://gavinsimpson.github.io/gratia/reference/data_slice.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Prepare a data slice through model covariates — data_slice","text":"","code":"load_mgcv()  # simulate some Gaussian data df <- data_sim(\"eg1\", n = 50, seed = 2)  # fit a GAM with 1 smooth and 1 linear term m <- gam(y ~ s(x2, k = 7) + x1, data = df, method = \"REML\")  # Want to predict over f(x2) while holding `x1` at some value. # Default will use the observation closest to the median for unspecified # variables. ds <- data_slice(m, x2 = evenly(x2, n = 50)) ds #> # A tibble: 50 x 2 #>        x2    x1 #>     <dbl> <dbl> #>  1 0.0228 0.403 #>  2 0.0424 0.403 #>  3 0.0619 0.403 #>  4 0.0815 0.403 #>  5 0.101  0.403 #>  6 0.121  0.403 #>  7 0.140  0.403 #>  8 0.160  0.403 #>  9 0.179  0.403 #> 10 0.199  0.403 #> # i 40 more rows  # for full control, specify the values you want ds <- data_slice(m, x2 = evenly(x2, n = 50), x1 = 0.3)  # or provide an expression (function call) which will be evaluated in the # data frame passed to `data` or `model.frame(object)` ds <- data_slice(m, x2 = evenly(x2, n = 50), x1 = mean(x1))"},{"path":"https://gavinsimpson.github.io/gratia/reference/datagen.html","id":null,"dir":"Reference","previous_headings":"","what":"Generate data over the range of variables used in smooths — datagen","title":"Generate data over the range of variables used in smooths — datagen","text":"smooth GAM, generate new data range variables involved smooth. function deprecated useful narrow use-case. Use data_slice() instead.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/datagen.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Generate data over the range of variables used in smooths — datagen","text":"","code":"datagen(x, ...)  # S3 method for class 'mgcv.smooth' datagen(x, n = 100, data, ...)  # S3 method for class 'fs.interaction' datagen(x, n = 100, data, ...)  # S3 method for class 'gam' datagen(x, smooth = NULL, n = 200, ...)  # S3 method for class 'gamm' datagen(x, ...)  # S3 method for class 'list' datagen(x, ...)"},{"path":"https://gavinsimpson.github.io/gratia/reference/datagen.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Generate data over the range of variables used in smooths — datagen","text":"x object new data required. Currently objects classes \"gam\", \"gamm\" supported, smooths mgcv inheriting class \"mgcv.smooth\". ... arguments passed methods n numeric; number data values generate per term smooth. data data frame; \"mgcv.smooth\" objects, data used fit GAM need supplied.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/datagen.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Generate data over the range of variables used in smooths — datagen","text":"data frame new values spread range observed values.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/datagen.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Generate data over the range of variables used in smooths — datagen","text":"Gavin L. Simpson","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/derivative_samples.html","id":null,"dir":"Reference","previous_headings":"","what":"Posterior expectations of derivatives from an estimated model — derivative_samples","title":"Posterior expectations of derivatives from an estimated model — derivative_samples","text":"Posterior expectations derivatives estimated model","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/derivative_samples.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Posterior expectations of derivatives from an estimated model — derivative_samples","text":"","code":"derivative_samples(object, ...)  # Default S3 method derivative_samples(object, ...)  # S3 method for class 'gamm' derivative_samples(object, ...)  # S3 method for class 'gam' derivative_samples(   object,   focal = NULL,   data = NULL,   order = 1L,   type = c(\"forward\", \"backward\", \"central\"),   scale = c(\"response\", \"linear_predictor\"),   method = c(\"gaussian\", \"mh\", \"inla\", \"user\"),   n = 100,   eps = 1e-07,   n_sim = 10000,   level = lifecycle::deprecated(),   seed = NULL,   envir = environment(formula(object)),   draws = NULL,   mvn_method = c(\"mvnfast\", \"mgcv\"),   ... )"},{"path":"https://gavinsimpson.github.io/gratia/reference/derivative_samples.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Posterior expectations of derivatives from an estimated model — derivative_samples","text":"object R object compute derivatives ... arguments passed methods fitted_samples() focal character; name focal variable. response derivative response respect variable returned. variables involved model held constant values. can missing supplying data, case, focal variable identified one variable constant. data data frame containing values model covariates evaluate first derivatives smooths. supplied, one variable must held constant value. order numeric; order derivative. type character; type finite difference used. One \"forward\", \"backward\", \"central\". scale character; derivative estimated response linear predictor (link) scale? One \"response\" (default), \"linear predictor\". method character; method used draw samples posterior distribution. \"gaussian\" uses Gaussian (Laplace) approximation posterior. \"mh\" uses Metropolis Hastings sample alternates t proposals proposals based shrunken version posterior covariance matrix. \"inla\" uses variant Integrated Nested Laplace Approximation due Wood (2019), (currently implemented). \"user\" allows user-supplied posterior draws (currently implemented). n numeric; number points evaluate derivative (data supplied). eps numeric; finite difference. n_sim integer; number simulations used computing simultaneous intervals. level  seed numeric; random seed simulations. envir environment within recreate data used fit object. draws matrix; user supplied posterior draws used method = \"user\". mvn_method character; one \"mvnfast\" \"mgcv\". default uses mvnfast::rmvn(), can considerably faster generate large numbers MVN random values mgcv::rmvn(), might work marginal fits, covariance matrix close singular.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/derivative_samples.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Posterior expectations of derivatives from an estimated model — derivative_samples","text":"tibble, currently following variables: .derivative: estimated partial derivative, additional columns containing covariate values derivative evaluated.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/derivative_samples.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Posterior expectations of derivatives from an estimated model — derivative_samples","text":"Gavin L. Simpson","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/derivative_samples.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Posterior expectations of derivatives from an estimated model — derivative_samples","text":"","code":"load_mgcv() df <- data_sim(\"eg1\", dist = \"negbin\", scale = 0.25, seed = 42)  # fit the GAM (note: for execution time reasons using bam()) m <- bam(y ~ s(x0) + s(x1) + s(x2) + s(x3),   data = df, family = nb(), method = \"fREML\")  # data slice through data along x2 - all other covariates will be set to # typical values (value closest to median) ds <- data_slice(m, x2 = evenly(x2, n = 200))  # samples from posterior of derivatives fd_samp <- derivative_samples(m,   data = ds, type = \"central\",   focal = \"x2\", eps = 0.01, seed = 21, n_sim = 100 )  # plot the first 20 posterior draws if (requireNamespace(\"ggplot2\") && requireNamespace(\"dplyr\")) {   library(\"ggplot2\")   fd_samp |>     dplyr::filter(.draw <= 20) |>     ggplot(aes(x = x2, y = .derivative, group = .draw)) +     geom_line(alpha = 0.5) }"},{"path":"https://gavinsimpson.github.io/gratia/reference/derivatives.html","id":null,"dir":"Reference","previous_headings":"","what":"Derivatives of estimated smooths via finite differences — derivatives","title":"Derivatives of estimated smooths via finite differences — derivatives","text":"Derivatives estimated smooths via finite differences","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/derivatives.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Derivatives of estimated smooths via finite differences — derivatives","text":"","code":"derivatives(object, ...)  # Default S3 method derivatives(object, ...)  # S3 method for class 'gamm' derivatives(object, ...)  # S3 method for class 'gam' derivatives(   object,   select = NULL,   term = deprecated(),   data = newdata,   order = 1L,   type = c(\"forward\", \"backward\", \"central\"),   n = 100,   eps = 1e-07,   interval = c(\"confidence\", \"simultaneous\"),   n_sim = 10000,   level = 0.95,   unconditional = FALSE,   frequentist = FALSE,   offset = NULL,   ncores = 1,   partial_match = FALSE,   ...,   newdata = NULL )"},{"path":"https://gavinsimpson.github.io/gratia/reference/derivatives.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Derivatives of estimated smooths via finite differences — derivatives","text":"object R object compute derivatives . ... arguments passed methods. select character; select smooth's posterior draw . default (NULL) means posteriors smooths model wil sampled . supplied, character vector requested terms. Can partial match smooth term; see argument partial_match . term Use select instead. data data frame containing values model covariates evaluate first derivatives smooths. order numeric; order derivative. type character; type finite difference used. One \"forward\", \"backward\", \"central\". n numeric; number points evaluate derivative . eps numeric; finite difference. interval character; type interval compute. One \"confidence\" point-wise intervals, \"simultaneous\" simultaneous intervals. n_sim integer; number simulations used computing simultaneous intervals. level numeric; 0 < level < 1; confidence level point-wise simultaneous interval. default 0.95 95% interval. unconditional logical; use smoothness selection-corrected Bayesian covariance matrix? frequentist logical; use frequentist covariance matrix? offset numeric; value use offset term ncores number cores generating random variables multivariate normal distribution. Passed mvnfast::rmvn(). Parallelization take place OpenMP supported (appears work Windows current R). partial_match logical; smooths selected partial matches term? TRUE, term can single string match . newdata Deprecated: use data instead.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/derivatives.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Derivatives of estimated smooths via finite differences — derivatives","text":"tibble, currently following variables: smooth: smooth row refers , var: name variable involved smooth, data: values var derivative evaluated, derivative: estimated derivative, se: standard error estimated derivative, crit: critical value derivative ± (crit * se) gives upper lower bounds requested confidence simultaneous interval (given level), lower: lower bound confidence simultaneous interval, upper: upper bound confidence simultaneous interval.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/derivatives.html","id":"note","dir":"Reference","previous_headings":"","what":"Note","title":"Derivatives of estimated smooths via finite differences — derivatives","text":"derivatives() ignore random effect smooths encounters object.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/derivatives.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Derivatives of estimated smooths via finite differences — derivatives","text":"Gavin L. Simpson","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/derivatives.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Derivatives of estimated smooths via finite differences — derivatives","text":"","code":"load_mgcv() dat <- data_sim(\"eg1\", n = 400, dist = \"normal\", scale = 2, seed = 42) mod <- gam(y ~ s(x0) + s(x1) + s(x2) + s(x3), data = dat, method = \"REML\")  ## first derivatives of all smooths using central finite differences derivatives(mod, type = \"central\") #> # A tibble: 400 x 12 #>    .smooth .by   .fs   .derivative   .se .crit .lower_ci .upper_ci      x0    x1 #>    <chr>   <chr> <chr>       <dbl> <dbl> <dbl>     <dbl>     <dbl>   <dbl> <dbl> #>  1 s(x0)   NA    NA           7.41  3.33  1.96     0.874      13.9 2.39e-4    NA #>  2 s(x0)   NA    NA           7.40  3.33  1.96     0.884      13.9 1.03e-2    NA #>  3 s(x0)   NA    NA           7.39  3.30  1.96     0.929      13.8 2.04e-2    NA #>  4 s(x0)   NA    NA           7.36  3.24  1.96     1.01       13.7 3.04e-2    NA #>  5 s(x0)   NA    NA           7.32  3.15  1.96     1.14       13.5 4.05e-2    NA #>  6 s(x0)   NA    NA           7.26  3.04  1.96     1.30       13.2 5.06e-2    NA #>  7 s(x0)   NA    NA           7.18  2.90  1.96     1.49       12.9 6.06e-2    NA #>  8 s(x0)   NA    NA           7.09  2.76  1.96     1.69       12.5 7.07e-2    NA #>  9 s(x0)   NA    NA           6.99  2.61  1.96     1.87       12.1 8.07e-2    NA #> 10 s(x0)   NA    NA           6.87  2.47  1.96     2.03       11.7 9.08e-2    NA #> # i 390 more rows #> # i 2 more variables: x2 <dbl>, x3 <dbl>  ## derivatives for a selected smooth derivatives(mod, type = \"central\", select = \"s(x1)\") #> # A tibble: 100 x 9 #>    .smooth .by   .fs   .derivative   .se .crit .lower_ci .upper_ci       x1 #>    <chr>   <chr> <chr>       <dbl> <dbl> <dbl>     <dbl>     <dbl>    <dbl> #>  1 s(x1)   NA    NA         -0.907  3.12  1.96     -7.02      5.20 0.000405 #>  2 s(x1)   NA    NA         -0.906  3.11  1.96     -7.01      5.20 0.0105   #>  3 s(x1)   NA    NA         -0.898  3.10  1.96     -6.97      5.17 0.0205   #>  4 s(x1)   NA    NA         -0.880  3.06  1.96     -6.88      5.12 0.0306   #>  5 s(x1)   NA    NA         -0.849  3.00  1.96     -6.73      5.03 0.0406   #>  6 s(x1)   NA    NA         -0.803  2.92  1.96     -6.52      4.92 0.0507   #>  7 s(x1)   NA    NA         -0.740  2.81  1.96     -6.25      4.77 0.0607   #>  8 s(x1)   NA    NA         -0.659  2.69  1.96     -5.93      4.61 0.0708   #>  9 s(x1)   NA    NA         -0.557  2.56  1.96     -5.57      4.46 0.0809   #> 10 s(x1)   NA    NA         -0.436  2.42  1.96     -5.19      4.32 0.0909   #> # i 90 more rows ## or via a partial match derivatives(mod, type = \"central\", select = \"x1\", partial_match = TRUE) #> # A tibble: 100 x 9 #>    .smooth .by   .fs   .derivative   .se .crit .lower_ci .upper_ci       x1 #>    <chr>   <chr> <chr>       <dbl> <dbl> <dbl>     <dbl>     <dbl>    <dbl> #>  1 s(x1)   NA    NA         -0.907  3.12  1.96     -7.02      5.20 0.000405 #>  2 s(x1)   NA    NA         -0.906  3.11  1.96     -7.01      5.20 0.0105   #>  3 s(x1)   NA    NA         -0.898  3.10  1.96     -6.97      5.17 0.0205   #>  4 s(x1)   NA    NA         -0.880  3.06  1.96     -6.88      5.12 0.0306   #>  5 s(x1)   NA    NA         -0.849  3.00  1.96     -6.73      5.03 0.0406   #>  6 s(x1)   NA    NA         -0.803  2.92  1.96     -6.52      4.92 0.0507   #>  7 s(x1)   NA    NA         -0.740  2.81  1.96     -6.25      4.77 0.0607   #>  8 s(x1)   NA    NA         -0.659  2.69  1.96     -5.93      4.61 0.0708   #>  9 s(x1)   NA    NA         -0.557  2.56  1.96     -5.57      4.46 0.0809   #> 10 s(x1)   NA    NA         -0.436  2.42  1.96     -5.19      4.32 0.0909   #> # i 90 more rows"},{"path":"https://gavinsimpson.github.io/gratia/reference/difference_smooths.html","id":null,"dir":"Reference","previous_headings":"","what":"Differences of factor smooth interactions — difference_smooths","title":"Differences of factor smooth interactions — difference_smooths","text":"Estimates pairwise differences (comparisons) factor smooth interactions (smooths factor argument) pairs groups defined factor. group means can optionally included difference.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/difference_smooths.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Differences of factor smooth interactions — difference_smooths","text":"","code":"difference_smooths(model, ...)  # S3 method for class 'gam' difference_smooths(   model,   select = NULL,   smooth = deprecated(),   n = 100,   ci_level = 0.95,   data = NULL,   group_means = FALSE,   partial_match = TRUE,   unconditional = FALSE,   frequentist = FALSE,   ... )"},{"path":"https://gavinsimpson.github.io/gratia/reference/difference_smooths.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Differences of factor smooth interactions — difference_smooths","text":"model fitted model. ... arguments passed methods. currently used. select character, logical, numeric; smooths plot. NULL, default, model smooths drawn. Numeric select indexes smooths order specified formula stored object. Character select matches labels smooths shown example output summary(object). Logical select operates per numeric select order smooths stored. smooth Use select instead. n numeric; number points evaluate difference pairs smooths. ci_level numeric 0 1; coverage credible interval. data data frame locations evaluate difference smooths. group_means logical; group means included difference? partial_match logical; smooth match partially smooths? partial_match = TRUE, smooth must single string, character vector length 1. Unlike similar functions, default TRUE intention users matching factor-smooth labels. unconditional logical; account smoothness selection model? frequentist logical; use frequentist covariance matrix?","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/difference_smooths.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Differences of factor smooth interactions — difference_smooths","text":"","code":"load_mgcv() df <- data_sim(\"eg4\", seed = 42) m <- gam(y ~ fac + s(x2, by = fac) + s(x0), data = df, method = \"REML\")  sm_dif <- difference_smooths(m, select = \"s(x2)\") sm_dif #> # A tibble: 300 x 9 #>    .smooth .by   .level_1 .level_2 .diff   .se .lower_ci .upper_ci      x2 #>    <chr>   <chr> <chr>    <chr>    <dbl> <dbl>     <dbl>     <dbl>   <dbl> #>  1 s(x2)   fac   1        2        0.386 0.618  -0.824        1.60 0.00359 #>  2 s(x2)   fac   1        2        0.479 0.574  -0.646        1.60 0.0136  #>  3 s(x2)   fac   1        2        0.572 0.534  -0.474        1.62 0.0237  #>  4 s(x2)   fac   1        2        0.665 0.497  -0.308        1.64 0.0338  #>  5 s(x2)   fac   1        2        0.758 0.464  -0.151        1.67 0.0438  #>  6 s(x2)   fac   1        2        0.850 0.435  -0.00342      1.70 0.0539  #>  7 s(x2)   fac   1        2        0.941 0.412   0.134        1.75 0.0639  #>  8 s(x2)   fac   1        2        1.03  0.393   0.262        1.80 0.0740  #>  9 s(x2)   fac   1        2        1.12  0.378   0.380        1.86 0.0841  #> 10 s(x2)   fac   1        2        1.21  0.367   0.489        1.93 0.0941  #> # i 290 more rows  draw(sm_dif)   # include the groups means for `fac` in the difference sm_dif2 <- difference_smooths(m, select = \"s(x2)\", group_means = TRUE) draw(sm_dif2)"},{"path":[]},{"path":"https://gavinsimpson.github.io/gratia/reference/dispersion.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Dispersion parameter for fitted model — dispersion","text":"","code":"dispersion(model, ...)  # S3 method for class 'gam' dispersion(model, ...)  # S3 method for class 'glm' dispersion(model, ...)"},{"path":"https://gavinsimpson.github.io/gratia/reference/dispersion.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Dispersion parameter for fitted model — dispersion","text":"model fitted model. ... arguments passed methods.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.basis.html","id":null,"dir":"Reference","previous_headings":"","what":"Plot basis functions — draw.basis","title":"Plot basis functions — draw.basis","text":"Plots basis functions using ggplot2","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.basis.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Plot basis functions — draw.basis","text":"","code":"# S3 method for class 'basis' draw(   object,   legend = FALSE,   labeller = NULL,   ylab = NULL,   title = NULL,   subtitle = NULL,   caption = NULL,   ncol = NULL,   nrow = NULL,   angle = NULL,   guides = \"keep\",   contour = FALSE,   n_contour = 10,   contour_col = \"black\",   ... )"},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.basis.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Plot basis functions — draw.basis","text":"object object, result call basis(). legend logical; legend drawn indicate basis functions? labeller labeller function label facets. default use ggplot2::label_both(). ylab character expression; label y axis. supplied, suitable label generated object. title character expression; title plot. See ggplot2::labs(). subtitle character expression; subtitle plot. See ggplot2::labs(). caption character expression; plot caption. See ggplot2::labs(). ncol, nrow numeric; numbers rows columns spread plots angle numeric; angle x axis tick labels drawn passed angle argument ggplot2::guide_axis(). guides character; one \"keep\" (default), \"collect\", \"auto\". Passed patchwork::plot_layout() contour logical; contours draw plot using ggplot2::geom_contour(). n_contour numeric; number contour bins. result n_contour - 1 contour lines drawn. See ggplot2::geom_contour(). contour_col colour specification contour lines. ... arguments passed methods. used method.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.basis.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Plot basis functions — draw.basis","text":"patchwork object.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.basis.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Plot basis functions — draw.basis","text":"Gavin L. Simpson","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.basis.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Plot basis functions — draw.basis","text":"","code":"load_mgcv() df <- data_sim(\"eg1\", n = 400, seed = 42) m <- gam(y ~ s(x0) + s(x1) + s(x2) + s(x3), data = df, method = \"REML\")  bf <- basis(m) draw(bf)   bf <- basis(m, \"s(x2)\") draw(bf)"},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.compare_smooths.html","id":null,"dir":"Reference","previous_headings":"","what":"Plot comparisons of smooths — draw.compare_smooths","title":"Plot comparisons of smooths — draw.compare_smooths","text":"Plot comparisons smooths","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.compare_smooths.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Plot comparisons of smooths — draw.compare_smooths","text":"","code":"# S3 method for class 'compare_smooths' draw(object, ncol = NULL, nrow = NULL, guides = \"collect\", ...)"},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.compare_smooths.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Plot comparisons of smooths — draw.compare_smooths","text":"object class \"compare_smooths\", result call compare_smooths(). ncol, nrow numeric; numbers rows columns spread plots guides character; one \"keep\" (default), \"collect\", \"auto\". Passed patchwork::plot_layout() ... additional arguments passed patchwork::wrap_plots().","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.conditional_values.html","id":null,"dir":"Reference","previous_headings":"","what":"Plot conditional predictions — draw.conditional_values","title":"Plot conditional predictions — draw.conditional_values","text":"Plot conditional predictions","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.conditional_values.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Plot conditional predictions — draw.conditional_values","text":"","code":"# S3 method for class 'conditional_values' draw(   object,   facet_scales = \"fixed\",   discrete_colour = NULL,   discrete_fill = NULL,   ... )"},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.conditional_values.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Plot conditional predictions — draw.conditional_values","text":"object object class \"conditional_values\", result call conditional_values(). facet_scales character; facets axis scales across facets? See ggplot2::facet_wrap() details. Options : \"fixed\" (default), \"free_x\", \"free_y\", \"free\". discrete_colour suitable colour scale used plotting discrete variables. discrete_fill suitable fill scale used plotting discrete variables. ... additional arguments passed patchwork::wrap_plots().","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.derivatives.html","id":null,"dir":"Reference","previous_headings":"","what":"Plot derivatives of smooths — draw.derivatives","title":"Plot derivatives of smooths — draw.derivatives","text":"Plot derivatives smooths","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.derivatives.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Plot derivatives of smooths — draw.derivatives","text":"","code":"# S3 method for class 'derivatives' draw(   object,   select = NULL,   scales = c(\"free\", \"fixed\"),   add_change = FALSE,   change_type = c(\"change\", \"sizer\"),   alpha = 0.2,   change_col = \"black\",   decrease_col = \"#56B4E9\",   increase_col = \"#E69F00\",   lwd_change = 1.5,   ncol = NULL,   nrow = NULL,   guides = \"keep\",   angle = NULL,   ... )  # S3 method for class 'partial_derivatives' draw(   object,   select = NULL,   scales = c(\"free\", \"fixed\"),   alpha = 0.2,   ncol = NULL,   nrow = NULL,   guides = \"keep\",   angle = NULL,   ... )"},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.derivatives.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Plot derivatives of smooths — draw.derivatives","text":"object fitted GAM, result call mgcv::gam(). select character, logical, numeric; smooths plot. NULL, default, model smooths drawn. Numeric select indexes smooths order specified formula stored object. Character select matches labels smooths shown example output summary(object). Logical select operates per numeric select order smooths stored. scales character; univariate smooths plotted y-axis scale? scales = \"free\", default, univariate smooth y-axis scale. scales = \"fixed\", common y axis scale used univariate smooths. Currently affect y-axis scale plots parametric terms. add_change logical; periods significant change highlighted plot? change_type character; type change indicate. \"change\", differentiation made periods significant increase decrease. \"sizer\", periods increase decrease differentiated resulting plot. alpha numeric; alpha transparency confidence simultaneous interval. change_col, decrease_col, increase_col colour specifications use indicating periods change. col_change used change_type = \"change\", col_decrease col_increase used `change_type = \"sizer\"“. lwd_change numeric; linewidth use change indicators. ncol, nrow numeric; numbers rows columns spread plots guides character; one \"keep\" (default), \"collect\", \"auto\". Passed patchwork::plot_layout() angle numeric; angle x axis tick labels drawn passed angle argument ggplot2::guide_axis(). ... additional arguments passed patchwork::wrap_plots().","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.derivatives.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Plot derivatives of smooths — draw.derivatives","text":"","code":"load_mgcv() dat <- data_sim(\"eg1\", n = 800, dist = \"normal\", scale = 2, seed = 42) mod <- gam(y ~ s(x0) + s(x1) + s(x2) + s(x3), data = dat, method = \"REML\")  ## first derivative of all smooths df <- derivatives(mod, type = \"central\") draw(df)  ## fixed axis scales draw(df, scales = \"fixed\")"},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.difference_smooth.html","id":null,"dir":"Reference","previous_headings":"","what":"Plot differences of smooths — draw.difference_smooth","title":"Plot differences of smooths — draw.difference_smooth","text":"Plot differences smooths","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.difference_smooth.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Plot differences of smooths — draw.difference_smooth","text":"","code":"# S3 method for class 'difference_smooth' draw(   object,   select = NULL,   rug = FALSE,   ref_line = FALSE,   contour = FALSE,   contour_col = \"black\",   n_contour = NULL,   ci_alpha = 0.2,   ci_col = \"black\",   smooth_col = \"black\",   line_col = \"red\",   scales = c(\"free\", \"fixed\"),   ncol = NULL,   nrow = NULL,   guides = \"keep\",   xlab = NULL,   ylab = NULL,   title = NULL,   subtitle = NULL,   caption = NULL,   angle = NULL,   ... )"},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.difference_smooth.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Plot differences of smooths — draw.difference_smooth","text":"object fitted GAM, result call mgcv::gam(). select character, logical, numeric; smooths plot. NULL, default, model smooths drawn. Numeric select indexes smooths order specified formula stored object. Character select matches labels smooths shown example output summary(object). Logical select operates per numeric select order smooths stored. rug logical; ref_line logical; contour logical; contour lines added smooth surfaces? contour_col colour specification contour lines. n_contour numeric; number contour bins. result n_contour - 1 contour lines drawn. See ggplot2::geom_contour(). ci_alpha numeric; alpha transparency confidence simultaneous interval. ci_col colour specification confidence/credible intervals band. Affects fill interval. smooth_col colour specification smooth difference line. line_col colour specification drawing reference lines scales character; univariate smooths plotted y-axis scale? scales = \"free\", default, univariate smooth y-axis scale. scales = \"fixed\", common y axis scale used univariate smooths. Currently affect y-axis scale plots parametric terms. ncol, nrow numeric; numbers rows columns spread plots guides character; one \"keep\" (default), \"collect\", \"auto\". Passed patchwork::plot_layout() xlab, ylab, title, subtitle, caption character; labels annotate plots angle numeric; angle x axis tick labels drawn passed angle argument ggplot2::guide_axis(). ... additional arguments passed patchwork::wrap_plots().","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.difference_smooth.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Plot differences of smooths — draw.difference_smooth","text":"","code":"load_mgcv() # simulate some data; a factor smooth example df <- data_sim(\"eg4\", seed = 42) # fit GAM m <- gam(y ~ fac + s(x2, by = fac) + s(x0), data = df, method = \"REML\")  # calculate the differences between pairs of smooths the f_j(x2) term diffs <- difference_smooths(m, select = \"s(x2)\") draw(diffs)"},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.evaluated_parametric_term.html","id":null,"dir":"Reference","previous_headings":"","what":"Plot estimated parametric effects — draw.evaluated_parametric_term","title":"Plot estimated parametric effects — draw.evaluated_parametric_term","text":"Plots estimated univariate bivariate smooths using ggplot2.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.evaluated_parametric_term.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Plot estimated parametric effects — draw.evaluated_parametric_term","text":"","code":"# S3 method for class 'evaluated_parametric_term' draw(   object,   ci_level = 0.95,   constant = NULL,   fun = NULL,   xlab,   ylab,   title = NULL,   subtitle = NULL,   caption = NULL,   rug = TRUE,   position = \"identity\",   response_range = NULL,   ... )"},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.evaluated_parametric_term.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Plot estimated parametric effects — draw.evaluated_parametric_term","text":"object object, result call evaluate_parametric_term(). ci_level numeric 0 1; coverage credible interval. constant numeric; constant add estimated values smooth. constant, supplied, added estimated value confidence band computed. fun function; function applied estimated values confidence interval plotting. Can function name function. Function fun applied adding constant, provided. xlab character expression; label x axis. supplied, suitable label generated object. ylab character expression; label y axis. supplied, suitable label generated object. title character expression; title plot. See ggplot2::labs(). subtitle character expression; subtitle plot. See ggplot2::labs(). caption character expression; plot caption. See ggplot2::labs(). rug evaluate_parametric_terms(), logical indicate rug plot drawn. position Position adjustment, either string, result call position adjustment function. response_range numeric; vector two values giving range response data guide. Used fix plots common scale/range. Ignored show set \"se\". ... arguments passed methods.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.evaluated_parametric_term.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Plot estimated parametric effects — draw.evaluated_parametric_term","text":"ggplot2::ggplot() object.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.evaluated_parametric_term.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Plot estimated parametric effects — draw.evaluated_parametric_term","text":"Gavin L. Simpson","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.gam.html","id":null,"dir":"Reference","previous_headings":"","what":"Plot estimated smooths from a fitted GAM — draw.gam","title":"Plot estimated smooths from a fitted GAM — draw.gam","text":"Plots estimated smooths fitted GAM model similar way mgcv::plot.gam() instead using base graphics, ggplot2::ggplot() used instead.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.gam.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Plot estimated smooths from a fitted GAM — draw.gam","text":"","code":"# S3 method for class 'gam' draw(   object,   data = NULL,   select = NULL,   parametric = FALSE,   terms = NULL,   residuals = FALSE,   scales = c(\"free\", \"fixed\"),   ci_level = 0.95,   n = 100,   n_3d = 16,   n_4d = 4,   unconditional = FALSE,   overall_uncertainty = TRUE,   constant = NULL,   fun = NULL,   dist = 0.1,   rug = TRUE,   contour = TRUE,   grouped_by = FALSE,   ci_alpha = 0.2,   ci_col = \"black\",   smooth_col = \"black\",   resid_col = \"steelblue3\",   contour_col = \"black\",   n_contour = NULL,   partial_match = FALSE,   discrete_colour = NULL,   discrete_fill = NULL,   continuous_colour = NULL,   continuous_fill = NULL,   position = \"identity\",   angle = NULL,   ncol = NULL,   nrow = NULL,   guides = \"keep\",   widths = NULL,   heights = NULL,   crs = NULL,   default_crs = NULL,   lims_method = \"cross\",   wrap = TRUE,   caption = TRUE,   envir = environment(formula(object)),   ... )"},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.gam.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Plot estimated smooths from a fitted GAM — draw.gam","text":"object fitted GAM, result call mgcv::gam(). data optional data frame used supply data smooths evaluated plotted. usually needed, option need fine control exactly data used plotting. select character, logical, numeric; smooths plot. NULL, default, model smooths drawn. Numeric select indexes smooths order specified formula stored object. Character select matches labels smooths shown example output summary(object). Logical select operates per numeric select order smooths stored. parametric logical; plot parametric terms also? Note select used selecting smooths plot. terms argument used select parametric effects plotted. default, mgcv::plot.gam(), draw parametric effects. terms character; model parametric terms drawn? Default NULL plot parametric terms can drawn. residuals logical; partial residuals smooth drawn? Ignored anything simple univariate smooth. scales character; univariate smooths plotted y-axis scale? scales = \"free\", default, univariate smooth y-axis scale. scales = \"fixed\", common y axis scale used univariate smooths. Currently affect y-axis scale plots parametric terms. ci_level numeric 0 1; coverage credible interval. n numeric; number points range covariate evaluate smooth. n_3d numeric; number new observations generate third dimension 3D smooth. n_4d numeric; number new observations generate dimensions higher 2 (!) kD smooth (k >= 4). example, smooth 4D smooth, dimensions 3 4 get n_4d new observations. unconditional logical; confidence intervals include uncertainty due smoothness selection? TRUE, corrected Bayesian covariance matrix used. overall_uncertainty logical; uncertainty model constant term included standard error evaluate values smooth? constant numeric; constant add estimated values smooth. constant, supplied, added estimated value confidence band computed. fun function; function applied estimated values confidence interval plotting. Can function name function. Function fun applied adding constant, provided. dist numeric; greater 0, used determine location far data plotted plotting 2-D smooths. data scaled unit square deciding exclude, dist distance within unit square. See mgcv::exclude..far() details. rug logical; draw rug plot bottom plot 1-D smooths plot locations data higher dimensions. contour logical; contours draw plot using ggplot2::geom_contour(). grouped_by logical; factor smooths drawn one panel per level factor (FALSE, default), individual smooths combined single panel containing levels (TRUE)? ci_alpha numeric; alpha transparency confidence simultaneous interval. ci_col colour specification confidence/credible intervals band. Affects fill interval. smooth_col colour specification smooth line. resid_col colour specification partial residuals. contour_col colour specification contour lines. n_contour numeric; number contour bins. result n_contour - 1 contour lines drawn. See ggplot2::geom_contour(). partial_match logical; smooths selected partial matches select? TRUE, select can single string match . discrete_colour suitable colour scale used plotting discrete variables. discrete_fill suitable fill scale used plotting discrete variables. continuous_colour suitable colour scale used plotting continuous variables. continuous_fill suitable fill scale used plotting continuous variables. position Position adjustment, either string, result call position adjustment function. angle numeric; angle x axis tick labels drawn passed angle argument ggplot2::guide_axis(). ncol, nrow numeric; numbers rows columns spread plots guides character; one \"keep\" (default), \"collect\", \"auto\". Passed patchwork::plot_layout() widths, heights relative widths heights column row grid. get repeated match dimensions grid. 1 plot widths = NULL, value widths set internally widths = 1 accommodate plots smooths use fixed aspect ratio. crs coordinate reference system (CRS) use plot. data projected CRS. See ggplot2::coord_sf() details. default_crs coordinate reference system (CRS) use non-sf layers plot. left default NULL, CRS used 4326 (WGS84), appropriate spline---sphere smooths, parameterized terms latitude longitude coordinates. See ggplot2::coord_sf() details. lims_method character; affects axis limits determined. See ggplot2::coord_sf(). careful; testing examples, changing \"orthogonal\" example chlorophyll-example Simon Wood's GAM book quickly used RAM test system OS killed R. incorrect usage part; right now grid points SOS smooths evaluated (supplied user) can produce invalid coordinates corners tiles grid generated tile centres without respect spacing tiles. wrap logical; wrap plots patchwork? FALSE, list ggplot objects returned, 1 per term plotted. caption logical; show smooth type caption plot? envir environment look data within. ... additional arguments passed patchwork::wrap_plots().","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.gam.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Plot estimated smooths from a fitted GAM — draw.gam","text":"object returned created patchwork::wrap_plots().","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.gam.html","id":"note","dir":"Reference","previous_headings":"","what":"Note","title":"Plot estimated smooths from a fitted GAM — draw.gam","text":"Internally, plots smooth created using ggplot2::ggplot() composed single plot using patchwork::wrap_plots(). result, possible use + add plots way one might typically work ggplot() plots. Instead, use & operator; see examples.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.gam.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Plot estimated smooths from a fitted GAM — draw.gam","text":"Gavin L. Simpson","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.gam.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Plot estimated smooths from a fitted GAM — draw.gam","text":"","code":"load_mgcv()  # simulate some data df1 <- data_sim(\"eg1\", n = 400, dist = \"normal\", scale = 2, seed = 2) # fit GAM m1 <- gam(y ~ s(x0) + s(x1) + s(x2) + s(x3), data = df1, method = \"REML\")  # plot all smooths draw(m1)   # can add partial residuals draw(m1, residuals = TRUE)   df2 <- data_sim(\"eg2\", n = 1000, dist = \"normal\", scale = 1, seed = 2) m2 <- gam(y ~ s(x, z, k = 40), data = df2, method = \"REML\") draw(m2, contour = FALSE, n = 50)   # See https://gavinsimpson.github.io/gratia/articles/custom-plotting.html # for more examples and for details on how to modify the theme of all the # plots produced by draw(). To modify all panels, for example to change the # theme, use the & operator"},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.gamlss.html","id":null,"dir":"Reference","previous_headings":"","what":"Plot smooths of a GAMLSS model estimated by GJRM::gamlss — draw.gamlss","title":"Plot smooths of a GAMLSS model estimated by GJRM::gamlss — draw.gamlss","text":"Provides draw() method GAMLSS (distributional GAMs) fitted GJRM::gamlss().","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.gamlss.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Plot smooths of a GAMLSS model estimated by GJRM::gamlss — draw.gamlss","text":"","code":"# S3 method for class 'gamlss' draw(   object,   scales = c(\"free\", \"fixed\"),   ncol = NULL,   nrow = NULL,   guides = \"keep\",   widths = NULL,   heights = NULL,   ... )"},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.gamlss.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Plot smooths of a GAMLSS model estimated by GJRM::gamlss — draw.gamlss","text":"object model, fitted GJRM::gamlss() scales character; univariate smooths plotted y-axis scale? scales = \"free\", default, univariate smooth y-axis scale. scales = \"fixed\", common y axis scale used univariate smooths. Currently affect y-axis scale plots parametric terms. ncol, nrow numeric; numbers rows columns spread plots guides character; one \"keep\" (default), \"collect\", \"auto\". Passed patchwork::plot_layout() widths, heights relative widths heights column row grid. get repeated match dimensions grid. 1 plot widths = NULL, value widths set internally widths = 1 accommodate plots smooths use fixed aspect ratio. ... arguments passed draw.gam()","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.gamlss.html","id":"note","dir":"Reference","previous_headings":"","what":"Note","title":"Plot smooths of a GAMLSS model estimated by GJRM::gamlss — draw.gamlss","text":"Plots smooths labelled linear predictor belong.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.gamlss.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Plot smooths of a GAMLSS model estimated by GJRM::gamlss — draw.gamlss","text":"","code":"if (require(\"GJRM\", quietly = TRUE)) {   # follow example from ?GJRM::gamlss   load_mgcv()   suppressPackageStartupMessages(library(\"GJRM\"))   set.seed(0)   n <- 100   x1 <- round(runif(n))   x2 <- runif(n)   x3 <- runif(n)   f1 <- function(x) cos(pi * 2 * x) + sin(pi * x)   y1 <- -1.55 + 2 * x1 + f1(x2) + rnorm(n)   dataSim <- data.frame(y1, x1, x2, x3)    eq_mu <- y1 ~ x1 + s(x2)   eq_s <- ~ s(x3, k = 6)   fl <- list(eq_mu, eq_s)   m <- gamlss(fl, data = dataSim)    draw(m) }"},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.html","id":null,"dir":"Reference","previous_headings":"","what":"Generic plotting via ggplot2 — draw","title":"Generic plotting via ggplot2 — draw","text":"Generic plotting via ggplot2","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Generic plotting via ggplot2 — draw","text":"","code":"draw(object, ...)"},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Generic plotting via ggplot2 — draw","text":"object R object plot. ... arguments passed methods.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Generic plotting via ggplot2 — draw","text":"ggplot2::ggplot() object.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Generic plotting via ggplot2 — draw","text":"Generic function plotting R objects uses ggplot2 package.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Generic plotting via ggplot2 — draw","text":"Gavin L. Simpson","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.mgcv_smooth.html","id":null,"dir":"Reference","previous_headings":"","what":"Plot basis functions — draw.mgcv_smooth","title":"Plot basis functions — draw.mgcv_smooth","text":"Plots basis functions using ggplot2","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.mgcv_smooth.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Plot basis functions — draw.mgcv_smooth","text":"","code":"# S3 method for class 'mgcv_smooth' draw(   object,   legend = FALSE,   use_facets = TRUE,   labeller = NULL,   xlab,   ylab,   title = NULL,   subtitle = NULL,   caption = NULL,   angle = NULL,   ... )"},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.mgcv_smooth.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Plot basis functions — draw.mgcv_smooth","text":"object object, result call basis(). legend logical; legend drawn indicate basis functions? use_facets logical; factor smooths, use facets show basis functions level factor? FALSE, separate ggplot object created level combined using patchwork::wrap_plots(). Currently ignored. labeller labeller function label facets. default use ggplot2::label_both(). xlab character expression; label x axis. supplied, suitable label generated object. ylab character expression; label y axis. supplied, suitable label generated object. title character expression; title plot. See ggplot2::labs(). subtitle character expression; subtitle plot. See ggplot2::labs(). caption character expression; plot caption. See ggplot2::labs(). angle numeric; angle x axis tick labels drawn passed angle argument ggplot2::guide_axis(). ... arguments passed methods. used method.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.mgcv_smooth.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Plot basis functions — draw.mgcv_smooth","text":"ggplot2::ggplot() object.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.mgcv_smooth.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Plot basis functions — draw.mgcv_smooth","text":"Gavin L. Simpson","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.mgcv_smooth.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Plot basis functions — draw.mgcv_smooth","text":"","code":"load_mgcv() df <- data_sim(\"eg4\", n = 400, seed = 42)  bf <- basis(s(x0), data = df) draw(bf)   bf <- basis(s(x2, by = fac, bs = \"bs\"), data = df) draw(bf)"},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.pairwise_concurvity.html","id":null,"dir":"Reference","previous_headings":"","what":"Plot concurvity measures — draw.pairwise_concurvity","title":"Plot concurvity measures — draw.pairwise_concurvity","text":"Plot concurvity measures","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.pairwise_concurvity.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Plot concurvity measures — draw.pairwise_concurvity","text":"","code":"# S3 method for class 'pairwise_concurvity' draw(   object,   title = \"Smooth-wise concurvity\",   subtitle = NULL,   caption = NULL,   x_lab = \"Term\",   y_lab = \"With\",   fill_lab = \"Concurvity\",   continuous_colour = NULL,   ... )  # S3 method for class 'overall_concurvity' draw(   object,   title = \"Overall concurvity\",   subtitle = NULL,   caption = NULL,   y_lab = \"Concurvity\",   x_lab = NULL,   bar_col = \"steelblue\",   bar_fill = \"steelblue\",   ... )"},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.pairwise_concurvity.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Plot concurvity measures — draw.pairwise_concurvity","text":"object object inheriting class \"concurvity\", usually result call model_concurvity() abbreviated form concrvity(). title character; plot title. subtitle character; plot subtitle. caption character; plot caption x_lab character; label x axis. y_lab character; label y axis. fill_lab character; label use fill guide. continuous_colour function; continuous colour (fill) scale use. ... arguments passed methods. bar_col colour specification bar colour. bar_fill colour specification bar fill","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.parametric_effects.html","id":null,"dir":"Reference","previous_headings":"","what":"Plot estimated effects for model parametric terms — draw.parametric_effects","title":"Plot estimated effects for model parametric terms — draw.parametric_effects","text":"Plot estimated effects model parametric terms","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.parametric_effects.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Plot estimated effects for model parametric terms — draw.parametric_effects","text":"","code":"# S3 method for class 'parametric_effects' draw(   object,   scales = c(\"free\", \"fixed\"),   ci_level = 0.95,   ci_col = \"black\",   ci_alpha = 0.2,   line_col = \"black\",   constant = NULL,   fun = NULL,   rug = TRUE,   position = \"identity\",   angle = NULL,   ...,   ncol = NULL,   nrow = NULL,   guides = \"keep\" )"},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.parametric_effects.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Plot estimated effects for model parametric terms — draw.parametric_effects","text":"object fitted GAM, result call mgcv::gam(). scales character; univariate smooths plotted y-axis scale? scales = \"free\", default, univariate smooth y-axis scale. scales = \"fixed\", common y axis scale used univariate smooths. Currently affect y-axis scale plots parametric terms. ci_level numeric 0 1; coverage credible interval. ci_col colour specification confidence/credible intervals band. Affects fill interval. ci_alpha numeric; alpha transparency confidence simultaneous interval. line_col colour specification used regression lines linear continuous terms. constant numeric; constant add estimated values smooth. constant, supplied, added estimated value confidence band computed. fun function; function applied estimated values confidence interval plotting. Can function name function. Function fun applied adding constant, provided. rug logical; draw rug plot bottom plot 1-D smooths plot locations data higher dimensions. position Position adjustment, either string, result call position adjustment function. angle numeric; angle x axis tick labels drawn passed angle argument ggplot2::guide_axis(). ... additional arguments passed patchwork::wrap_plots(). ncol, nrow numeric; numbers rows columns spread plots guides character; one \"keep\" (default), \"collect\", \"auto\". Passed patchwork::plot_layout()","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.penalty_df.html","id":null,"dir":"Reference","previous_headings":"","what":"Display penalty matrices of smooths using ggplot — draw.penalty_df","title":"Display penalty matrices of smooths using ggplot — draw.penalty_df","text":"Displays penalty matrices smooths heatmap using ggplot","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.penalty_df.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Display penalty matrices of smooths using ggplot — draw.penalty_df","text":"","code":"# S3 method for class 'penalty_df' draw(   object,   normalize = FALSE,   as_matrix = TRUE,   continuous_fill = NULL,   xlab = NULL,   ylab = NULL,   title = NULL,   subtitle = NULL,   caption = NULL,   ncol = NULL,   nrow = NULL,   guides = \"keep\",   ... )"},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.penalty_df.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Display penalty matrices of smooths using ggplot — draw.penalty_df","text":"object fitted GAM, result call mgcv::gam(). normalize logical; normalize penalty range -1, 1? as_matrix logical; plotted penalty matrix oriented? TRUE row 1, column 1 penalty matrix draw upper left, whereas, FALSE drawn lower left plot. continuous_fill suitable fill scale used plotting continuous variables. xlab character expression; label x axis. supplied, axis label drawn. May vector, one per penalty. ylab character expression; label y axis.  supplied, axis label drawn. May vector, one per penalty. title character expression; title plot. See ggplot2::labs(). May vector, one per penalty. subtitle character expression; subtitle plot. See ggplot2::labs(). May vector, one per penalty. caption character expression; plot caption. See ggplot2::labs(). May vector, one per penalty. ncol, nrow numeric; numbers rows columns spread plots. guides character; one \"keep\" (default), \"collect\", \"auto\". Passed patchwork::plot_layout() ... additional arguments passed patchwork::wrap_plots().","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.penalty_df.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Display penalty matrices of smooths using ggplot — draw.penalty_df","text":"","code":"load_mgcv() dat <- data_sim(\"eg4\", n = 400, seed = 42) m <- gam(y ~ s(x0) + s(x1, bs = \"cr\") + s(x2, bs = \"bs\", by = fac),   data = dat, method = \"REML\" )  ## produce a multi-panel plot of all penalties draw(penalty(m))   # for a specific smooth draw(penalty(m, select = \"s(x2):fac1\"))"},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.rootogram.html","id":null,"dir":"Reference","previous_headings":"","what":"Draw a rootogram — draw.rootogram","title":"Draw a rootogram — draw.rootogram","text":"rootogram model diagnostic tool assesses goodness fit statistical model. observed values response compared expected fitted model. discrete, count responses, frequency count (0, 1, 2, etc) observed data expected conditional distribution response implied model compared. continuous variables, observed expected frequencies obtained grouping data bins. rootogram drawn using ggplot2::ggplot() graphics. design closely follows Kleiber & Zeileis (2016).","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.rootogram.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Draw a rootogram — draw.rootogram","text":"","code":"# S3 method for class 'rootogram' draw(   object,   type = c(\"hanging\", \"standing\", \"suspended\"),   sqrt = TRUE,   ref_line = TRUE,   warn_limits = TRUE,   fitted_colour = \"steelblue\",   bar_colour = NA,   bar_fill = \"grey\",   ref_line_colour = \"black\",   warn_line_colour = \"black\",   ylab = NULL,   xlab = NULL,   ... )"},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.rootogram.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Draw a rootogram — draw.rootogram","text":"object R object plot. type character; type rootogram draw. sqrt logical; show observed fitted frequencies ref_line logical; draw reference line zero? warn_limits logical; draw Tukey's warning limit lines +/- 1? fitted_colour, bar_colour, bar_fill, ref_line_colour, warn_line_colour colours used draw respective element rootogram. xlab, ylab character; labels x y axis rootogram. May missing (NULL), case suitable labels used. ' ... arguments passed methods.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.rootogram.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Draw a rootogram — draw.rootogram","text":"'ggplot' object.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.rootogram.html","id":"references","dir":"Reference","previous_headings":"","what":"References","title":"Draw a rootogram — draw.rootogram","text":"Kleiber, C., Zeileis, ., (2016) Visualizing Count Data Regressions Using Rootograms. . Stat. 70, 296–303. doi:10.1080/00031305.2016.1173590","code":""},{"path":[]},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.rootogram.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Draw a rootogram — draw.rootogram","text":"","code":"load_mgcv() df <- data_sim(\"eg1\", n = 1000, dist = \"poisson\", scale = 0.1, seed = 6)  # A poisson example m <- gam(y ~ s(x0, bs = \"cr\") + s(x1, bs = \"cr\") + s(x2, bs = \"cr\") +   s(x3, bs = \"cr\"), family = poisson(), data = df, method = \"REML\") rg <- rootogram(m)  # plot the rootogram draw(rg)   # change the type of rootogram draw(rg, type = \"suspended\")"},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.smooth_estimates.html","id":null,"dir":"Reference","previous_headings":"","what":"Plot the result of a call to smooth_estimates() — draw.smooth_estimates","title":"Plot the result of a call to smooth_estimates() — draw.smooth_estimates","text":"Plot result call smooth_estimates()","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.smooth_estimates.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Plot the result of a call to smooth_estimates() — draw.smooth_estimates","text":"","code":"# S3 method for class 'smooth_estimates' draw(   object,   constant = NULL,   fun = NULL,   contour = TRUE,   grouped_by = FALSE,   contour_col = \"black\",   n_contour = NULL,   ci_alpha = 0.2,   ci_col = \"black\",   smooth_col = \"black\",   resid_col = \"steelblue3\",   decrease_col = \"#56B4E9\",   increase_col = \"#E69F00\",   change_lwd = 1.75,   partial_match = FALSE,   discrete_colour = NULL,   discrete_fill = NULL,   continuous_colour = NULL,   continuous_fill = NULL,   angle = NULL,   ylim = NULL,   crs = NULL,   default_crs = NULL,   lims_method = \"cross\",   caption = TRUE,   ... )"},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.smooth_estimates.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Plot the result of a call to smooth_estimates() — draw.smooth_estimates","text":"object fitted GAM, result call mgcv::gam(). constant numeric; constant add estimated values smooth. constant, supplied, added estimated value confidence band computed. fun function; function applied estimated values confidence interval plotting. Can function name function. Function fun applied adding constant, provided. contour logical; contours draw plot using ggplot2::geom_contour(). grouped_by logical; factor smooths drawn one panel per level factor (FALSE, default), individual smooths combined single panel containing levels (TRUE)? contour_col colour specification contour lines. n_contour numeric; number contour bins. result n_contour - 1 contour lines drawn. See ggplot2::geom_contour(). ci_alpha numeric; alpha transparency confidence simultaneous interval. ci_col colour specification confidence/credible intervals band. Affects fill interval. smooth_col colour specification smooth line. resid_col colour specification partial residuals. decrease_col, increase_col colour specifications use indicating periods change. col_change used change_type = \"change\", col_decrease col_increase used `change_type = \"sizer\"“. change_lwd numeric; value set linewidth ggplot2::geom_line(), used represent periods change. partial_match logical; smooths selected partial matches select? TRUE, select can single string match . discrete_colour suitable colour scale used plotting discrete variables. discrete_fill suitable fill scale used plotting discrete variables. continuous_colour suitable colour scale used plotting continuous variables. continuous_fill suitable fill scale used plotting continuous variables. angle numeric; angle x axis tick labels drawn passed angle argument ggplot2::guide_axis(). ylim numeric; vector y axis limits use panels drawn. crs coordinate reference system (CRS) use plot. data projected CRS. See ggplot2::coord_sf() details. default_crs coordinate reference system (CRS) use non-sf layers plot. left default NULL, CRS used 4326 (WGS84), appropriate spline---sphere smooths, parameterized terms latitude longitude coordinates. See ggplot2::coord_sf() details. lims_method character; affects axis limits determined. See ggplot2::coord_sf(). careful; testing examples, changing \"orthogonal\" example chlorophyll-example Simon Wood's GAM book quickly used RAM test system OS killed R. incorrect usage part; right now grid points SOS smooths evaluated (supplied user) can produce invalid coordinates corners tiles grid generated tile centres without respect spacing tiles. caption logical; show smooth type caption plot? ... additional arguments passed patchwork::wrap_plots().","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.smooth_estimates.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Plot the result of a call to smooth_estimates() — draw.smooth_estimates","text":"","code":"load_mgcv() # example data df <- data_sim(\"eg1\", seed = 21) # fit GAM m <- gam(y ~ s(x0) + s(x1) + s(x2) + s(x3), data = df, method = \"REML\") # plot all of the estimated smooths sm <- smooth_estimates(m) draw(sm)  # evaluate smooth of `x2` sm <- smooth_estimates(m, select = \"s(x2)\") # plot it draw(sm)   # customising some plot elements draw(sm, ci_col = \"steelblue\", smooth_col = \"forestgreen\", ci_alpha = 0.3)   # Add a constant to the plotted smooth draw(sm, constant = coef(m)[1])   # Adding change indicators to smooths based on derivatives of the smooth d <- derivatives(m, n = 100) # n to match smooth_estimates()  smooth_estimates(m) |>   add_sizer(derivatives = d, type = \"sizer\") |>   draw()"},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.smooth_samples.html","id":null,"dir":"Reference","previous_headings":"","what":"Plot posterior smooths — draw.smooth_samples","title":"Plot posterior smooths — draw.smooth_samples","text":"Plot posterior smooths","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.smooth_samples.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Plot posterior smooths — draw.smooth_samples","text":"","code":"# S3 method for class 'smooth_samples' draw(   object,   select = NULL,   n_samples = NULL,   seed = NULL,   xlab = NULL,   ylab = NULL,   title = NULL,   subtitle = NULL,   caption = NULL,   alpha = 1,   colour = \"black\",   contour = FALSE,   contour_col = \"black\",   n_contour = NULL,   scales = c(\"free\", \"fixed\"),   rug = TRUE,   partial_match = FALSE,   angle = NULL,   ncol = NULL,   nrow = NULL,   guides = \"keep\",   ... )"},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.smooth_samples.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Plot posterior smooths — draw.smooth_samples","text":"object fitted GAM, result call mgcv::gam(). select character, logical, numeric; smooths plot. NULL, default, model smooths drawn. Numeric select indexes smooths order specified formula stored object. Character select matches labels smooths shown example output summary(object). Logical select operates per numeric select order smooths stored. n_samples numeric; NULL, sample n_samples posterior draws plotting. seed numeric; random seed used sampling draws. xlab character expression; label x axis. supplied, suitable label generated object. ylab character expression; label y axis. supplied, suitable label generated object. title character expression; title plot. See ggplot2::labs(). subtitle character expression; subtitle plot. See ggplot2::labs(). caption character expression; plot caption. See ggplot2::labs(). alpha numeric; alpha transparency confidence simultaneous interval. colour colour use draw posterior smooths. Passed ggplot2::geom_line() argument colour. contour logical; contour lines added smooth surfaces? contour_col colour specification contour lines. n_contour numeric; number contour bins. result n_contour - 1 contour lines drawn. See ggplot2::geom_contour(). scales character; univariate smooths plotted y-axis scale? scales = \"free\", default, univariate smooth y-axis scale. scales = \"fixed\", common y axis scale used univariate smooths. Currently affect y-axis scale plots parametric terms. rug logical; draw rug plot bottom plot 1-D smooths plot locations data higher dimensions. partial_match logical; smooths selected partial matches select? TRUE, select can single string match . angle numeric; angle x axis tick labels drawn passed angle argument ggplot2::guide_axis(). ncol, nrow numeric; numbers rows columns spread plots guides character; one \"keep\" (default), \"collect\", \"auto\". Passed patchwork::plot_layout() ... arguments passed patchwork::wrap_plots().","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.smooth_samples.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Plot posterior smooths — draw.smooth_samples","text":"Gavin L. Simpson","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.smooth_samples.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Plot posterior smooths — draw.smooth_samples","text":"","code":"load_mgcv() dat1 <- data_sim(\"eg1\", n = 400, dist = \"normal\", scale = 1, seed = 1) ## a single smooth GAM m1 <- gam(y ~ s(x0) + s(x1) + s(x2) + s(x3), data = dat1, method = \"REML\") ## posterior smooths from m1 sm1 <- smooth_samples(m1, n = 15, seed = 23478) ## plot draw(sm1, alpha = 0.7)  ## plot only 5 randomly smapled draws draw(sm1, n_samples = 5, alpha = 0.7)   ## A factor-by smooth example dat2 <- data_sim(\"eg4\", n = 400, dist = \"normal\", scale = 1, seed = 1) ## a multi-smooth GAM with a factor-by smooth m2 <- gam(y ~ fac + s(x2, by = fac) + s(x0), data = dat2, method = \"REML\") ## posterior smooths from m1 sm2 <- smooth_samples(m2, n = 15, seed = 23478) ## plot, this time selecting only the factor-by smooth draw(sm2, select = \"s(x2)\", partial_match = TRUE, alpha = 0.7)   # \\donttest{ ## A 2D smooth example dat3 <- data_sim(\"eg2\", n = 400, dist = \"normal\", scale = 1, seed = 1) ## fit a 2D smooth m3 <- gam(y ~ te(x, z), data = dat3, method = \"REML\") ## get samples sm3 <- smooth_samples(m3, n = 10) ## plot just 6 of the draws, with contour line overlays draw(sm3, n_samples = 6, contour = TRUE, seed = 42)  # }"},{"path":"https://gavinsimpson.github.io/gratia/reference/draw_parametric_effect.html","id":null,"dir":"Reference","previous_headings":"","what":"Internal function to draw an individual parametric effect — draw_parametric_effect","title":"Internal function to draw an individual parametric effect — draw_parametric_effect","text":"Internal function draw individual parametric effect","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/draw_parametric_effect.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Internal function to draw an individual parametric effect — draw_parametric_effect","text":"","code":"draw_parametric_effect(   object,   ci_level = 0.95,   ci_col = \"black\",   ci_alpha = 0.2,   line_col = \"black\",   constant = NULL,   fun = NULL,   xlab = NULL,   ylab = NULL,   title = NULL,   subtitle = NULL,   caption = NULL,   rug = TRUE,   position = \"identity\",   ylim = NULL,   angle = NULL,   factor_levels = NULL,   ... )"},{"path":"https://gavinsimpson.github.io/gratia/reference/draw_parametric_effect.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Internal function to draw an individual parametric effect — draw_parametric_effect","text":"object fitted GAM, result call mgcv::gam(). ci_level numeric 0 1; coverage credible interval. ci_col colour specification confidence/credible intervals band. Affects fill interval. ci_alpha numeric; alpha transparency confidence simultaneous interval. constant numeric; constant add estimated values smooth. constant, supplied, added estimated value confidence band computed. fun function; function applied estimated values confidence interval plotting. Can function name function. Function fun applied adding constant, provided. xlab character expression; label x axis. supplied, suitable label generated object. ylab character expression; label y axis. supplied, suitable label generated object. title character expression; title plot. See ggplot2::labs(). subtitle character expression; subtitle plot. See ggplot2::labs(). caption character expression; plot caption. See ggplot2::labs(). rug logical; draw rug plot bottom plot 1-D smooths plot locations data higher dimensions. position Position adjustment, either string, result call position adjustment function. angle numeric; angle x axis tick labels drawn passed angle argument ggplot2::guide_axis(). factor_levels list; named list factor levels ... additional arguments passed patchwork::wrap_plots().","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/edf.html","id":null,"dir":"Reference","previous_headings":"","what":"Effective degrees of freedom for smooths and GAMs — edf","title":"Effective degrees of freedom for smooths and GAMs — edf","text":"Extracts effective degrees freedom (EDF) model smooth terms overall EDF fitted GAMs","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/edf.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Effective degrees of freedom for smooths and GAMs — edf","text":"","code":"edf(object, ...)  # S3 method for class 'gam' edf(   object,   select = NULL,   smooth = deprecated(),   type = c(\"default\", \"unconditional\", \"alternative\"),   partial_match = FALSE,   ... )  model_edf(object, ..., type = c(\"default\", \"unconditional\", \"alternative\"))"},{"path":"https://gavinsimpson.github.io/gratia/reference/edf.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Effective degrees of freedom for smooths and GAMs — edf","text":"object fitted model extract smooth-specific EDFs. ... arguments passed methods. select character, logical, numeric; smooths plot. NULL, default, model smooths drawn. Numeric select indexes smooths order specified formula stored object. Character select matches labels smooths shown example output summary(object). Logical select operates per numeric select order smooths stored. smooth Use select instead. extracted. NULL, default, EDFs smooths returned. type character: type EDF return. \"default\" returns standard EDF; \"unconditional\" selects EDF corrected smoothness parameter selection, available; \"alternative\" returns alternative formulation EDF Wood (2017, pp. 252) partial_match logical; smooths selected partial matches select? TRUE, select can single string match .","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/edf.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Effective degrees of freedom for smooths and GAMs — edf","text":"Multiple formulations effective degrees freedom available. additional uncertainty due selection smoothness parameters can taken account computing EDF smooths. form EDF available type = \"unconditional\". Wood (2017; pp. 252) describes alternative EDF  model $$\\mathrm{EDF} = 2\\mathrm{tr}(\\mathbf{F}) -   \\mathrm{tr}(\\mathbf{FF}),$$ \\(\\mathrm{tr}\\) matrix trace \\(\\mathbf{F}\\) matrix mapping un-penalized coefficient estimates penalized coefficient estimates.  trace \\(\\mathbf{F}\\) effectively average shrinkage coefficients multipled number coefficients (Wood, 2017). Smooth-specific EDFs obtained summing relevent elements \\(\\mathrm{diag}(2\\mathbf{F} - \\mathbf{FF})\\).","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/edf.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Effective degrees of freedom for smooths and GAMs — edf","text":"","code":"load_mgcv() df <- data_sim(\"eg1\", n = 400, seed = 42) m <- gam(y ~ s(x0) + s(x1) + s(x2) + s(x3), data = df, method = \"REML\")  # extract the EDFs for all smooths edf(m) #> # A tibble: 4 x 2 #>   .smooth   .edf #>   <chr>    <dbl> #> 1 s(x0)   3.4248 #> 2 s(x1)   3.2213 #> 3 s(x2)   7.9049 #> 4 s(x3)   1.8847  # or selected smooths edf(m, select = c(\"s(x0)\", \"s(x2)\")) #> # A tibble: 2 x 2 #>   .smooth   .edf #>   <chr>    <dbl> #> 1 s(x0)   3.4248 #> 2 s(x2)   7.9049  # accounting for smoothness parameter uncertainty edf(m, type = \"unconditional\") #> # A tibble: 4 x 2 #>   .smooth   .edf #>   <chr>    <dbl> #> 1 s(x0)   3.7697 #> 2 s(x1)   3.8728 #> 3 s(x2)   8.0670 #> 4 s(x3)   2.8834  # over EDF of the model, including the intercept model_edf(m) #> # A tibble: 1 x 2 #>   .model   .edf #>   <chr>   <dbl> #> 1 m      17.436  # can get model EDF for multiple models m2 <- gam(y ~ s(x0) + s(x1) + s(x3), data = df, method = \"REML\") model_edf(m, m2) #> # A tibble: 2 x 2 #>   .model    .edf #>   <chr>    <dbl> #> 1 m      17.436  #> 2 m2      7.5777"},{"path":"https://gavinsimpson.github.io/gratia/reference/eval_smooth.html","id":null,"dir":"Reference","previous_headings":"","what":"S3 methods to evaluate individual smooths — eval_smooth","title":"S3 methods to evaluate individual smooths — eval_smooth","text":"S3 methods evaluate individual smooths","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/eval_smooth.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"S3 methods to evaluate individual smooths — eval_smooth","text":"","code":"eval_smooth(smooth, ...)  # S3 method for class 'mgcv.smooth' eval_smooth(   smooth,   model,   n = 100,   n_3d = NULL,   n_4d = NULL,   data = NULL,   unconditional = FALSE,   overall_uncertainty = TRUE,   dist = NULL,   ... )  # S3 method for class 'soap.film' eval_smooth(   smooth,   model,   n = 100,   n_3d = NULL,   n_4d = NULL,   data = NULL,   unconditional = FALSE,   overall_uncertainty = TRUE,   ... )  # S3 method for class 'scam_smooth' eval_smooth(   smooth,   model,   n = 100,   n_3d = NULL,   n_4d = NULL,   data = NULL,   unconditional = FALSE,   overall_uncertainty = TRUE,   dist = NULL,   ... )  # S3 method for class 'fs.interaction' eval_smooth(   smooth,   model,   n = 100,   data = NULL,   unconditional = FALSE,   overall_uncertainty = TRUE,   ... )  # S3 method for class 'sz.interaction' eval_smooth(   smooth,   model,   n = 100,   data = NULL,   unconditional = FALSE,   overall_uncertainty = TRUE,   ... )  # S3 method for class 'random.effect' eval_smooth(   smooth,   model,   n = 100,   data = NULL,   unconditional = FALSE,   overall_uncertainty = TRUE,   ... )  # S3 method for class 'mrf.smooth' eval_smooth(   smooth,   model,   n = 100,   data = NULL,   unconditional = FALSE,   overall_uncertainty = TRUE,   ... )  # S3 method for class 't2.smooth' eval_smooth(   smooth,   model,   n = 100,   n_3d = NULL,   n_4d = NULL,   data = NULL,   unconditional = FALSE,   overall_uncertainty = TRUE,   dist = NULL,   ... )  # S3 method for class 'tensor.smooth' eval_smooth(   smooth,   model,   n = 100,   n_3d = NULL,   n_4d = NULL,   data = NULL,   unconditional = FALSE,   overall_uncertainty = TRUE,   dist = NULL,   ... )"},{"path":"https://gavinsimpson.github.io/gratia/reference/eval_smooth.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"S3 methods to evaluate individual smooths — eval_smooth","text":"smooth currently object inherits class mgcv.smooth. ... arguments passed methods model fitted model; currently mgcv::gam() mgcv::bam() models suported. n numeric; number points range covariate evaluate smooth. n_3d, n_4d numeric; number points range last covariate 3D 4D smooth. default NULL achieves standard behaviour using n points range covariate, resulting n^d evaluation points, d dimension smooth. d > 2 can result many evaluation points slow performance. smooths d > 4, value n_4d used dimensions > 4, unless NULL, case default behaviour (using n dimensions) observed. data optional data frame values evaluate smooth . unconditional logical; confidence intervals include uncertainty due smoothness selection? TRUE, corrected Bayesian covariance matrix used. overall_uncertainty logical; uncertainty model constant term included standard error evaluate values smooth? dist numeric; greater 0, used determine location far data plotted plotting 2-D smooths. data scaled unit square deciding exclude, dist distance within unit square. See mgcv::exclude..far() details.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/evaluate_parametric_term.html","id":null,"dir":"Reference","previous_headings":"","what":"Evaluate parametric model terms — evaluate_parametric_term","title":"Evaluate parametric model terms — evaluate_parametric_term","text":"Returns values parametric model terms values factor terms grid covariate values linear parametric terms. function now deprecated favour parametric_effects().","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/evaluate_parametric_term.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Evaluate parametric model terms — evaluate_parametric_term","text":"","code":"evaluate_parametric_term(object, ...)  # S3 method for class 'gam' evaluate_parametric_term(object, term, unconditional = FALSE, ...)"},{"path":"https://gavinsimpson.github.io/gratia/reference/evaluate_parametric_term.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Evaluate parametric model terms — evaluate_parametric_term","text":"object object class \"gam\" \"gamm\". ... arguments passed methods. term character; parametric term whose effects evaluated unconditional logical; confidence intervals include uncertainty due smoothness selection? TRUE, corrected Bayesian covariance matrix used.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/evaluate_smooth.html","id":null,"dir":"Reference","previous_headings":"","what":"Evaluate a smooth — evaluate_smooth","title":"Evaluate a smooth — evaluate_smooth","text":"Evaluate smooth grid evenly spaced value range covariate associated smooth. Alternatively, set points smooth evaluated can supplied.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/evaluate_smooth.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Evaluate a smooth — evaluate_smooth","text":"","code":"evaluate_smooth(object, ...)"},{"path":"https://gavinsimpson.github.io/gratia/reference/evaluate_smooth.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Evaluate a smooth — evaluate_smooth","text":"object object class \"gam\" \"gamm\". ... arguments passed methods.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/evaluate_smooth.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Evaluate a smooth — evaluate_smooth","text":"data frame, class \"evaluated_1d_smooth\" evaluated_2d_smooth, inherit classes \"evaluated_smooth\" \"data.frame\".","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/evaluate_smooth.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Evaluate a smooth — evaluate_smooth","text":"evaluate_smooth() deprecated favour smooth_estimates(), provides cleaner way evaluate smooth range covariate values. smooth_estimates() can handle much wider range models evaluate_smooth() capable smooth_estimates() much easier extend handle new smooth types. code uses evaluate_smooth() work simply changing function call smooth_estimates(). However, differences: newdata argument becomes data","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/evenly.html","id":null,"dir":"Reference","previous_headings":"","what":"Create a sequence of evenly-spaced values — evenly","title":"Create a sequence of evenly-spaced values — evenly","text":"continuous vector x, evenly seq_min_max() create sequence n evenly-spaced values range lower – upper. default, lower defined min(x) upper max(x), excluding NAs. factor x, function returns levels(x).","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/evenly.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Create a sequence of evenly-spaced values — evenly","text":"","code":"evenly(x, n = 100, by = NULL, lower = NULL, upper = NULL)  seq_min_max(x, n, by = NULL, lower = NULL, upper = NULL)"},{"path":"https://gavinsimpson.github.io/gratia/reference/evenly.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Create a sequence of evenly-spaced values — evenly","text":"x numeric; vector evenly-spaced values returned n numeric; number evenly-spaced values return. default 100 used convenience typically used evaluating smooth. numeric; increment sequence. specified, argument n ignored sequence returned min(x) max(x) increments . lower numeric; lower bound interval. upper numeric; upper bound interval.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/evenly.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Create a sequence of evenly-spaced values — evenly","text":"numeric vector length n.","code":""},{"path":[]},{"path":"https://gavinsimpson.github.io/gratia/reference/evenly.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Create a sequence of evenly-spaced values — evenly","text":"","code":"x <- rnorm(10) n <- 10L  # 10 values evenly over the range of `x` evenly(x, n = n) #>  [1] -0.83562861 -0.56552757 -0.29542652 -0.02532547  0.24477557  0.51487662 #>  [7]  0.78497766  1.05507871  1.32517976  1.59528080  # evenly spaced values, incrementing by 0.2 evenly(x, by = 0.2) #>  [1] -0.83562861 -0.63562861 -0.43562861 -0.23562861 -0.03562861  0.16437139 #>  [7]  0.36437139  0.56437139  0.76437139  0.96437139  1.16437139  1.36437139 #> [13]  1.56437139  # evenly spaced values, incrementing by 0.2, starting at -2 evenly(x, by = 0.2, lower = -2) #>  [1] -2.0 -1.8 -1.6 -1.4 -1.2 -1.0 -0.8 -0.6 -0.4 -0.2  0.0  0.2  0.4  0.6  0.8 #> [16]  1.0  1.2  1.4"},{"path":"https://gavinsimpson.github.io/gratia/reference/factor_combos.html","id":null,"dir":"Reference","previous_headings":"","what":"All combinations of factor levels — factor_combos","title":"All combinations of factor levels — factor_combos","text":"combinations factor levels","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/factor_combos.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"All combinations of factor levels — factor_combos","text":"","code":"factor_combos(object, ...)  # S3 method for class 'gam' factor_combos(object, vars = everything(), complete = TRUE, ...)"},{"path":"https://gavinsimpson.github.io/gratia/reference/factor_combos.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"All combinations of factor levels — factor_combos","text":"object fitted model object. ... arguments passed methods. vars terms include exclude returned object. Uses tidyselect principles. complete logical; combinations factor levels returned? FALSE, combinations levels observed model retained.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/family.gam.html","id":null,"dir":"Reference","previous_headings":"","what":"Extract family objects from models — family.gam","title":"Extract family objects from models — family.gam","text":"Provides stats::family() method range GAM objects.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/family.gam.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Extract family objects from models — family.gam","text":"","code":"# S3 method for class 'gam' family(object, ...)  # S3 method for class 'gamm' family(object, ...)  # S3 method for class 'bam' family(object, ...)  # S3 method for class 'list' family(object, ...)"},{"path":"https://gavinsimpson.github.io/gratia/reference/family.gam.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Extract family objects from models — family.gam","text":"object fitted model. Models fitted mgcv::gam(), mgcv::bam(), mgcv::gamm(), gamm4::gamm4() currently supported. ... arguments passed methods.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/family_name.html","id":null,"dir":"Reference","previous_headings":"","what":"Name of family used to fit model — family_name","title":"Name of family used to fit model — family_name","text":"Extracts name family used fit supplied model.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/family_name.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Name of family used to fit model — family_name","text":"","code":"family_name(object, ...)"},{"path":"https://gavinsimpson.github.io/gratia/reference/family_name.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Name of family used to fit model — family_name","text":"object R object. ... arguments passed methods.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/family_name.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Name of family used to fit model — family_name","text":"character vector containing family name.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/family_type.html","id":null,"dir":"Reference","previous_headings":"","what":"Extracts the type of family in a consistent way — family_type","title":"Extracts the type of family in a consistent way — family_type","text":"Extracts type family consistent way","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/family_type.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Extracts the type of family in a consistent way — family_type","text":"","code":"family_type(object, ...)  # S3 method for class 'family' family_type(object, ...)  # Default S3 method family_type(object, ...)"},{"path":"https://gavinsimpson.github.io/gratia/reference/family_type.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Extracts the type of family in a consistent way — family_type","text":"object R object. Currently family() objects anything family() method. ... arguments passed methods.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/fderiv.html","id":null,"dir":"Reference","previous_headings":"","what":"First derivatives of fitted GAM functions — fderiv","title":"First derivatives of fitted GAM functions — fderiv","text":"function deprecated limited first order forward finite differences derivatives , improved offer needed functionality without breaking backwards compatability papers blog posts already used fderiv(). replacement, derivatives(), now available recommended new analyses.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/fderiv.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"First derivatives of fitted GAM functions — fderiv","text":"","code":"fderiv(model, ...)  # S3 method for class 'gam' fderiv(   model,   newdata,   term,   n = 200,   eps = 1e-07,   unconditional = FALSE,   offset = NULL,   ... )  # S3 method for class 'gamm' fderiv(model, ...)"},{"path":"https://gavinsimpson.github.io/gratia/reference/fderiv.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"First derivatives of fitted GAM functions — fderiv","text":"model fitted GAM. Currently models fitted mgcv::gam() mgcv::gamm() supported. ... Arguments passed methods. newdata data frame containing values model covariates evaluate first derivatives smooths. term character; vector one terms derivatives required. missing, derivatives smooth terms returned. n integer; newdata missing original data can reconstructed model n controls number values range covariate populate newdata. eps numeric; value finite difference used approximate first derivative. unconditional logical; TRUE, smoothing parameter uncertainty corrected covariance matrix used, available, otherwise uncorrected Bayesian posterior covariance matrix used. offset numeric; value offset use generating predictions.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/fderiv.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"First derivatives of fitted GAM functions — fderiv","text":"object class \"fderiv\" returned.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/fderiv.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"First derivatives of fitted GAM functions — fderiv","text":"Gavin L. Simpson","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/fderiv.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"First derivatives of fitted GAM functions — fderiv","text":"","code":"load_mgcv() dat <- data_sim(\"eg1\", seed = 2) mod <- gam(y ~ s(x0) + s(x1) + s(x2) + s(x3), data = dat, method = \"REML\")  ## first derivatives of all smooths... fd <- fderiv(mod) ## now use --> fd <- derivatives(mod)  ## ...and a selected smooth fd2 <- fderiv(mod, term = \"x1\") ## now use --> fd2 <- derivatives(mod, select = \"s(x1)\")  ## Models with factors dat <- data_sim(\"eg4\", n = 400, dist = \"normal\", scale = 2, seed = 2) mod <- gam(y ~ s(x0) + s(x1) + fac, data = dat, method = \"REML\")  ## first derivatives of all smooths... fd <- fderiv(mod) ## now use --> fd <- derivatives(mod)  ## ...and a selected smooth fd2 <- fderiv(mod, term = \"x1\") ## now use --> fd2 <- derivatives(mod, select = \"s(x1)\")"},{"path":"https://gavinsimpson.github.io/gratia/reference/fitted_samples.html","id":null,"dir":"Reference","previous_headings":"","what":"Draw fitted values from the posterior distribution — fitted_samples","title":"Draw fitted values from the posterior distribution — fitted_samples","text":"Expectations (fitted values) response drawn posterior distribution fitted model using Gaussian approximation posterior simple Metropolis Hastings sampler.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/fitted_samples.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Draw fitted values from the posterior distribution — fitted_samples","text":"","code":"fitted_samples(model, ...)  # S3 method for class 'gam' fitted_samples(   model,   n = 1,   data = newdata,   seed = NULL,   scale = c(\"response\", \"linear_predictor\"),   method = c(\"gaussian\", \"mh\", \"inla\", \"user\"),   n_cores = 1,   burnin = 1000,   thin = 1,   t_df = 40,   rw_scale = 0.25,   freq = FALSE,   unconditional = FALSE,   draws = NULL,   mvn_method = c(\"mvnfast\", \"mgcv\"),   ...,   newdata = NULL,   ncores = NULL )"},{"path":"https://gavinsimpson.github.io/gratia/reference/fitted_samples.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Draw fitted values from the posterior distribution — fitted_samples","text":"model fitted model supported types ... arguments passed methods. fitted_samples(), passed mgcv::predict.gam(). posterior_samples() passed fitted_samples(). predicted_samples() passed relevant simulate() method. n numeric; number posterior samples return. data data frame; new observations posterior draws model evaluated. supplied, data used fit model used data, available model. seed numeric; random seed simulations. scale character; scale fitted values returned ? \"linear predictor\" synonym \"link\" prefer terminology. method character; method used draw samples posterior distribution. \"gaussian\" uses Gaussian (Laplace) approximation posterior. \"mh\" uses Metropolis Hastings sampler alternates t proposals proposals based shrunken version posterior covariance matrix. \"inla\" uses variant Integrated Nested Laplace Approximation due Wood (2019), (currently implemented). \"user\" allows user-supplied posterior draws (currently implemented). n_cores number cores generating random variables multivariate normal distribution. Passed mvnfast::rmvn(). Parallelization take place OpenMP supported (appears work Windows current R). burnin numeric; number samples discard burnin draws. used method = \"mh\". thin numeric; number samples skip taking n draws. Results thin * n draws posterior taken. used method = \"mh\". t_df numeric; degrees freedom t distribution proposals. used method = \"mh\". rw_scale numeric; Factor scale posterior covariance matrix generating random walk proposals. Negative non finite skip random walk step. used method = \"mh\". freq logical; TRUE use frequentist covariance matrix parameter estimators, FALSE use Bayesian posterior covariance matrix parameters. unconditional logical; TRUE (freq == FALSE) Bayesian smoothing parameter uncertainty corrected covariance matrix used, available. draws matrix; user supplied posterior draws used method = \"user\". mvn_method character; one \"mvnfast\" \"mgcv\". default uses mvnfast::rmvn(), can considerably faster generate large numbers MVN random values mgcv::rmvn(), might work marginal fits, covariance matrix close singular. newdata Deprecated: use data instead. ncores Deprecated; use n_cores instead. number cores generating random variables multivariate normal distribution. Passed mvnfast::rmvn(). Parallelization take place OpenMP supported (appears work Windows current R).","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/fitted_samples.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Draw fitted values from the posterior distribution — fitted_samples","text":"tibble (data frame) 3 columns containing posterior predicted values long format. columns row (integer) row data posterior draw relates , draw (integer) index, range 1:n, indicating draw row relates , response (numeric) predicted response indicated row data.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/fitted_samples.html","id":"note","dir":"Reference","previous_headings":"","what":"Note","title":"Draw fitted values from the posterior distribution — fitted_samples","text":"Models offset terms supplied via offset argument mgcv::gam() etc. ignored mgcv::predict.gam(). , kind offset term also ignored posterior_samples(). Offset terms included model formula supplied mgcv::gam() etc ignored posterior samples produced reflect offset term values. side effect requiring new data values provided posterior_samples() via data argument must include offset variable.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/fitted_samples.html","id":"references","dir":"Reference","previous_headings":"","what":"References","title":"Draw fitted values from the posterior distribution — fitted_samples","text":"Wood, S.N., (2020). Simplified integrated nested Laplace approximation. Biometrika 107, 223–230. doi:10.1093/biomet/asz044","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/fitted_samples.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Draw fitted values from the posterior distribution — fitted_samples","text":"Gavin L. Simpson","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/fitted_samples.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Draw fitted values from the posterior distribution — fitted_samples","text":"","code":"load_mgcv() dat <- data_sim(\"eg1\", n = 1000, dist = \"normal\", scale = 2, seed = 2) m1 <- gam(y ~ s(x0) + s(x1) + s(x2) + s(x3), data = dat, method = \"REML\")  fs <- fitted_samples(m1, n = 5, seed = 42) # \\donttest{ fs #> # A tibble: 5,000 x 4 #>     .row .draw .parameter .fitted #>    <int> <int> <chr>        <dbl> #>  1     1     1 location      6.34 #>  2     2     1 location      5.08 #>  3     3     1 location      6.84 #>  4     4     1 location      7.71 #>  5     5     1 location      9.23 #>  6     6     1 location      8.03 #>  7     7     1 location      6.19 #>  8     8     1 location      7.28 #>  9     9     1 location     14.0  #> 10    10     1 location     12.7  #> # i 4,990 more rows # }  # can generate own set of draws and use them drws <- generate_draws(m1, n = 2, seed = 24) fs2 <- fitted_samples(m1, method = \"user\", draws = drws) # \\donttest{ fs2 #> # A tibble: 2,000 x 4 #>     .row .draw .parameter .fitted #>    <int> <int> <chr>        <dbl> #>  1     1     1 location      6.30 #>  2     2     1 location      5.12 #>  3     3     1 location      7.40 #>  4     4     1 location      7.42 #>  5     5     1 location      9.40 #>  6     6     1 location      8.04 #>  7     7     1 location      5.83 #>  8     8     1 location      7.30 #>  9     9     1 location     14.3  #> 10    10     1 location     13.1  #> # i 1,990 more rows # }"},{"path":"https://gavinsimpson.github.io/gratia/reference/fitted_values.html","id":null,"dir":"Reference","previous_headings":"","what":"Generate fitted values from a estimated GAM — fitted_values","title":"Generate fitted values from a estimated GAM — fitted_values","text":"Generate fitted values estimated GAM","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/fitted_values.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Generate fitted values from a estimated GAM — fitted_values","text":"","code":"fitted_values(object, ...)  # S3 method for class 'gam' fitted_values(   object,   data = NULL,   scale = c(\"response\", \"link\", \"linear predictor\"),   ci_level = 0.95,   ... )  # S3 method for class 'gamm' fitted_values(object, ...)  # S3 method for class 'scam' fitted_values(object, ...)"},{"path":"https://gavinsimpson.github.io/gratia/reference/fitted_values.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Generate fitted values from a estimated GAM — fitted_values","text":"object fitted model. Currently models fitted mgcv::gam() mgcv::bam() supported. ... arguments passed mgcv::predict.gam(). Note type, newdata, se.fit already used passed mgcv::predict.gam(). data optional data frame covariate values fitted values returned. scale character; scale fitted values returned ? \"linear predictor\" synonym \"link\" prefer terminology. ci_level numeric; value 0 1 indicating coverage credible interval.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/fitted_values.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Generate fitted values from a estimated GAM — fitted_values","text":"tibble (data frame) whose first m columns contain either data used fit model (data NULL), variables supplied data. Four columns added: fitted: fitted values specified scale, se: standard error fitted values (always link scale), lower, upper: limits credible interval fitted values, specified scale. Models fitted certain families include additional variables mgcv::ocat() models: scale = \"repsonse\", returned object contain row column category column, indicate row data row returned object belongs. Additionally, nrow(data) * n_categories rows returned object; row predicted probability single category response.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/fitted_values.html","id":"note","dir":"Reference","previous_headings":"","what":"Note","title":"Generate fitted values from a estimated GAM — fitted_values","text":"families, regardless scale fitted values returned, se component returned object link (linear predictor) scale, response scale. exception mgcv::ocat() family, se response scale scale = \"response\".","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/fitted_values.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Generate fitted values from a estimated GAM — fitted_values","text":"","code":"load_mgcv() sim_df <- data_sim(\"eg1\", n = 400, dist = \"normal\", scale = 2, seed = 2) m <- gam(y ~ s(x0) + s(x1) + s(x2) + s(x3), data = sim_df, method = \"REML\") fv <- fitted_values(m) fv #> # A tibble: 400 x 9 #>     .row       x0        x1         x2       x3  .fitted      .se .lower_ci #>    <int>    <dbl>     <dbl>      <dbl>    <dbl>    <dbl>    <dbl>     <dbl> #>  1     1 0.184882 0.617142  0.415244   0.132410  8.73875 0.354677   8.04360 #>  2     2 0.702374 0.569064  0.531439   0.365331  7.62581 0.337779   6.96378 #>  3     3 0.573326 0.153970  0.00324621 0.454532  3.12106 0.591862   1.96103 #>  4     4 0.168052 0.0348332 0.252100   0.537114 11.1124  0.402378  10.3237  #>  5     5 0.943839 0.997953  0.155229   0.185495 14.0533  0.452947  13.1655  #>  6     6 0.943475 0.835574  0.878840   0.449276  6.13080 0.364521   5.41635 #>  7     7 0.129159 0.586562  0.203511   0.256527 12.4838  0.355808  11.7864  #>  8     8 0.833449 0.339117  0.583528   0.618458  6.25215 0.344700   5.57655 #>  9     9 0.468019 0.166883  0.804473   0.880744  4.21463 0.372003   3.48552 #> 10    10 0.549984 0.807410  0.264717   0.317747 15.5283  0.369999  14.8031  #> # i 390 more rows #> # i 1 more variable: .upper_ci <dbl>"},{"path":"https://gavinsimpson.github.io/gratia/reference/fix_offset.html","id":null,"dir":"Reference","previous_headings":"","what":"Fix the names of a data frame containing an offset variable. — fix_offset","title":"Fix the names of a data frame containing an offset variable. — fix_offset","text":"Identifies variable, , model offset, fixed name offset(foo(var)) converted var, possibly sets values variable offset_val.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/fix_offset.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Fix the names of a data frame containing an offset variable. — fix_offset","text":"","code":"fix_offset(model, newdata, offset_val = NULL)"},{"path":"https://gavinsimpson.github.io/gratia/reference/fix_offset.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Fix the names of a data frame containing an offset variable. — fix_offset","text":"model fitted GAM. newdata data frame; new values predict . offset_val numeric, optional; provided, offset variable newdata set constant value returning newdata","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/fix_offset.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Fix the names of a data frame containing an offset variable. — fix_offset","text":"original newdata returned fixed names possibly modified offset variable.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/fix_offset.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Fix the names of a data frame containing an offset variable. — fix_offset","text":"Gavin L. Simpson","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/fix_offset.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Fix the names of a data frame containing an offset variable. — fix_offset","text":"","code":"load_mgcv() df <- data_sim(\"eg1\", n = 400, dist = \"normal\", seed = 2) m <- gam(y ~ s(x0) + s(x1) + offset(x2), data = df, method = \"REML\") names(model.frame(m)) #> [1] \"y\"          \"offset(x2)\" \"x0\"         \"x1\"         names(fix_offset(m, model.frame(m), offset_val = 1L)) #> [1] \"y\"  \"x2\" \"x0\" \"x1\""},{"path":"https://gavinsimpson.github.io/gratia/reference/fixef.gam.html","id":null,"dir":"Reference","previous_headings":"","what":"Extract fixed effects estimates from a fitted GAM — fixef.gam","title":"Extract fixed effects estimates from a fitted GAM — fixef.gam","text":"Extract fixed effects estimates fitted GAM","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/fixef.gam.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Extract fixed effects estimates from a fitted GAM — fixef.gam","text":"","code":"# S3 method for class 'gam' fixef(object, ...)  # S3 method for class 'gamm' fixef(object, ...)  # S3 method for class 'lm' fixef(object, ...)  # S3 method for class 'glm' fixef(object, ...)  fixed_effects(object, ...)  # Default S3 method fixed_effects(object, ...)"},{"path":"https://gavinsimpson.github.io/gratia/reference/fixef.gam.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Extract fixed effects estimates from a fitted GAM — fixef.gam","text":"object fitted GAM ... arguments passed methods","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/fixef.gam.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Extract fixed effects estimates from a fitted GAM — fixef.gam","text":"","code":"load_mgcv()  # run example if lme4 is available if (require(\"lme4\")) {   data(sleepstudy, package = \"lme4\")   m <- gam(     Reaction ~ Days + s(Subject, bs = \"re\") +       s(Days, Subject, bs = \"re\"),     data = sleepstudy, method = \"REML\"   )   fixef(m) } #> Loading required package: lme4 #> Loading required package: Matrix #> (Intercept)        Days  #>   251.40510    10.46729"},{"path":"https://gavinsimpson.github.io/gratia/reference/fixef.html","id":null,"dir":"Reference","previous_headings":"","what":"Extract fixed effects estimates — fixef","title":"Extract fixed effects estimates — fixef","text":"Extract fixed effects estimates","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/fixef.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Extract fixed effects estimates — fixef","text":"object fitted GAM ... arguments passed methods","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/gaussian_draws.html","id":null,"dir":"Reference","previous_headings":"","what":"Posterior samples using a simple Metropolis Hastings sampler — gaussian_draws","title":"Posterior samples using a simple Metropolis Hastings sampler — gaussian_draws","text":"Posterior samples using simple Metropolis Hastings sampler","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/gaussian_draws.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Posterior samples using a simple Metropolis Hastings sampler — gaussian_draws","text":"","code":"gaussian_draws(model, ...)  # S3 method for class 'gam' gaussian_draws(   model,   n,   n_cores = 1L,   index = NULL,   frequentist = FALSE,   unconditional = FALSE,   mvn_method = \"mvnfast\",   ... )  # S3 method for class 'scam' gaussian_draws(   model,   n,   n_cores = 1L,   index = NULL,   frequentist = FALSE,   parametrized = TRUE,   mvn_method = \"mvnfast\",   ... )"},{"path":"https://gavinsimpson.github.io/gratia/reference/gaussian_draws.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Posterior samples using a simple Metropolis Hastings sampler — gaussian_draws","text":"model fitted R model. Currently models fitted mgcv::gam() mgcv::bam(), return object inherits objects supported. , \"inherits\" used loose fashion; models fitted scam::scam() support even though models strictly inherit class \"gam\" far inherits() concerned. ... arguments passed methods. n numeric; number posterior draws take. n_cores integer; number CPU cores use generating multivariate normal distributed random values. used mvn_method = \"mvnfast\" method = \"gaussian\". index numeric; vector indices coefficients use. Can used subset mean vector covariance matrix extracted model. frequentist logical; TRUE, frequentist covariance matrix parameter estimates used. FALSE, Bayesian posterior covariance matrix parameters used. See mgcv::vcov.gam(). unconditional logical; TRUE  Bayesian smoothing parameter uncertainty corrected covariance matrix used, available model. See mgcv::vcov.gam(). mvn_method character; one \"mvnfast\" \"mgcv\". default uses mvnfast::rmvn(), can considerably faster generate large numbers MVN random values mgcv::rmvn(), might work marginal fits, covariance matrix close singular. parametrized logical; use parametrized coefficients covariance matrix, respect linear inequality constraints model. scam::scam() model fits.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/get_by_smooth.html","id":null,"dir":"Reference","previous_headings":"","what":"Extract an factor-by smooth by name — get_by_smooth","title":"Extract an factor-by smooth by name — get_by_smooth","text":"Extract factor-smooth name","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/get_by_smooth.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Extract an factor-by smooth by name — get_by_smooth","text":"","code":"get_by_smooth(object, term, level)"},{"path":"https://gavinsimpson.github.io/gratia/reference/get_by_smooth.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Extract an factor-by smooth by name — get_by_smooth","text":"object fitted GAM model object. term character; name smooth term extract. level character; level factor exrtact smooth .","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/get_by_smooth.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Extract an factor-by smooth by name — get_by_smooth","text":"single smooth object, list smooths several match named term.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/get_smooth.html","id":null,"dir":"Reference","previous_headings":"","what":"Extract an mgcv smooth by name — get_smooth","title":"Extract an mgcv smooth by name — get_smooth","text":"Extract mgcv smooth name","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/get_smooth.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Extract an mgcv smooth by name — get_smooth","text":"","code":"get_smooth(object, term)"},{"path":"https://gavinsimpson.github.io/gratia/reference/get_smooth.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Extract an mgcv smooth by name — get_smooth","text":"object fitted GAM model object. term character; name smooth term extract","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/get_smooth.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Extract an mgcv smooth by name — get_smooth","text":"single smooth object, list smooths several match named term.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/get_smooths_by_id.html","id":null,"dir":"Reference","previous_headings":"","what":"Extract an mgcv smooth given its position in the model object — get_smooths_by_id","title":"Extract an mgcv smooth given its position in the model object — get_smooths_by_id","text":"Extract mgcv smooth given position model object","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/get_smooths_by_id.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Extract an mgcv smooth given its position in the model object — get_smooths_by_id","text":"","code":"get_smooths_by_id(object, id)  # S3 method for class 'gam' get_smooths_by_id(object, id)  # S3 method for class 'scam' get_smooths_by_id(object, id)  # S3 method for class 'gamm' get_smooths_by_id(object, id)  # S3 method for class 'gamm4' get_smooths_by_id(object, id)  # S3 method for class 'list' get_smooths_by_id(object, id)"},{"path":"https://gavinsimpson.github.io/gratia/reference/get_smooths_by_id.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Extract an mgcv smooth given its position in the model object — get_smooths_by_id","text":"object fitted GAM model object. id numeric; position smooth model object.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/gratia-package.html","id":null,"dir":"Reference","previous_headings":"","what":"gratia: Graceful 'ggplot'-Based Graphics and Other Functions for GAMs Fitted Using 'mgcv' — gratia-package","title":"gratia: Graceful 'ggplot'-Based Graphics and Other Functions for GAMs Fitted Using 'mgcv' — gratia-package","text":"Graceful 'ggplot'-based graphics utility functions working generalized additive models (GAMs) fitted using 'mgcv' package. Provides reimplementation plot() method GAMs 'mgcv' provides, well 'tidyverse' compatible representations estimated smooths.","code":""},{"path":[]},{"path":"https://gavinsimpson.github.io/gratia/reference/gratia-package.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"gratia: Graceful 'ggplot'-Based Graphics and Other Functions for GAMs Fitted Using 'mgcv' — gratia-package","text":"Maintainer: Gavin L. Simpson ucfagls@gmail.com (ORCID) [copyright holder] contributors: Henrik Singmann (ORCID) [contributor]","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/gss_vocab.html","id":null,"dir":"Reference","previous_headings":"","what":"Data from the General Social Survey (GSS) from the National Opinion Research Center of the University of Chicago — gss_vocab","title":"Data from the General Social Survey (GSS) from the National Opinion Research Center of the University of Chicago — gss_vocab","text":"subset data carData::GSSvocab dataset carData package, containing observations 2016 .","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/gss_vocab.html","id":"format","dir":"Reference","previous_headings":"","what":"Format","title":"Data from the General Social Survey (GSS) from the National Opinion Research Center of the University of Chicago — gss_vocab","text":"data frame 1858 rows 3 variables: vocab: numeric; number words 10 correct vocabulary test. nativeBorn: factor; respondent born US? factor levels yes. ageGroup: factor; grouped age respondent levels 18-29 30-39, 40-49, 50-59, 60+.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/gw_functions.html","id":null,"dir":"Reference","previous_headings":"","what":"Gu and Wabha test functions — gw_f0","title":"Gu and Wabha test functions — gw_f0","text":"Gu Wabha test functions","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/gw_functions.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Gu and Wabha test functions — gw_f0","text":"","code":"gw_f0(x, ...)  gw_f1(x, ...)  gw_f2(x, ...)  gw_f3(x, ...)"},{"path":"https://gavinsimpson.github.io/gratia/reference/gw_functions.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Gu and Wabha test functions — gw_f0","text":"x numeric; vector points evaluate function , interval (0,1) ... arguments passed methods, ignored.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/gw_functions.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Gu and Wabha test functions — gw_f0","text":"","code":"x <- seq(0, 1, length = 6) gw_f0(x) #> [1] 0.000e+00 1.176e+00 1.902e+00 1.902e+00 1.176e+00 2.449e-16 gw_f1(x) #> [1] 1.000 1.492 2.226 3.320 4.953 7.389 gw_f2(x) #> [1] 0.000 8.591 4.261 3.199 1.100 0.000 gw_f3(x) # should be constant 0 #> [1] 0 0 0 0 0 0"},{"path":"https://gavinsimpson.github.io/gratia/reference/has_theta.html","id":null,"dir":"Reference","previous_headings":"","what":"Are additional parameters available for a GAM? — has_theta","title":"Are additional parameters available for a GAM? — has_theta","text":"additional parameters available GAM?","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/has_theta.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Are additional parameters available for a GAM? — has_theta","text":"","code":"has_theta(object)"},{"path":"https://gavinsimpson.github.io/gratia/reference/has_theta.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Are additional parameters available for a GAM? — has_theta","text":"object R object, either family() object object whose class family() method.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/has_theta.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Are additional parameters available for a GAM? — has_theta","text":"logical; TRUE additional parameters available, FALSE otherwise.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/has_theta.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Are additional parameters available for a GAM? — has_theta","text":"","code":"load_mgcv() df <- data_sim(\"eg1\", dist = \"poisson\", seed = 42, scale = 1 / 5) m <- gam(y ~ s(x0) + s(x1) + s(x2) + s(x3),   data = df, method = \"REML\",   family = nb() ) has_theta(m) #> [1] TRUE p <- theta(m)"},{"path":"https://gavinsimpson.github.io/gratia/reference/is_by_smooth.html","id":null,"dir":"Reference","previous_headings":"","what":"Tests for by variable smooths — is_by_smooth","title":"Tests for by variable smooths — is_by_smooth","text":"Functions check smooth -variable one test type -variable smooth factor-smooth continous-smooth interaction.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/is_by_smooth.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Tests for by variable smooths — is_by_smooth","text":"","code":"is_by_smooth(smooth)  is_factor_by_smooth(smooth)  is_continuous_by_smooth(smooth)  by_variable(smooth)  by_level(smooth)"},{"path":"https://gavinsimpson.github.io/gratia/reference/is_by_smooth.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Tests for by variable smooths — is_by_smooth","text":"smooth object class \"mgcv.smooth\"","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/is_by_smooth.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Tests for by variable smooths — is_by_smooth","text":"logical vector.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/is_by_smooth.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Tests for by variable smooths — is_by_smooth","text":"Gavin L. Simpson","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/is_factor_term.html","id":null,"dir":"Reference","previous_headings":"","what":"Is a model term a factor (categorical)? — is_factor_term","title":"Is a model term a factor (categorical)? — is_factor_term","text":"Given name (term label) term model, identify term factor term numeric. useful considering interactions, terms like fac1:fac2 num1:fac1 may requested user. terms type fac1:fac2 function return TRUE.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/is_factor_term.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Is a model term a factor (categorical)? — is_factor_term","text":"","code":"is_factor_term(object, term, ...)  # S3 method for class 'terms' is_factor_term(object, term, ...)  # S3 method for class 'gam' is_factor_term(object, term, ...)  # S3 method for class 'bam' is_factor_term(object, term, ...)  # S3 method for class 'gamm' is_factor_term(object, term, ...)  # S3 method for class 'list' is_factor_term(object, term, ...)"},{"path":"https://gavinsimpson.github.io/gratia/reference/is_factor_term.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Is a model term a factor (categorical)? — is_factor_term","text":"object R object method dispatch performed term character; name model term, sense attr(terms(object), \"term.labels\"). Currently checked see term exists model. ... arguments passed methods.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/is_factor_term.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Is a model term a factor (categorical)? — is_factor_term","text":"logical: TRUE variables involved term factors, otherwise FALSE.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/is_mgcv_smooth.html","id":null,"dir":"Reference","previous_headings":"","what":"Check if objects are smooths or are a particular type of smooth — is_mgcv_smooth","title":"Check if objects are smooths or are a particular type of smooth — is_mgcv_smooth","text":"Check objects smooths particular type smooth","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/is_mgcv_smooth.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Check if objects are smooths or are a particular type of smooth — is_mgcv_smooth","text":"","code":"is_mgcv_smooth(smooth)  stop_if_not_mgcv_smooth(smooth)  check_is_mgcv_smooth(smooth)  is_mrf_smooth(smooth)"},{"path":"https://gavinsimpson.github.io/gratia/reference/is_mgcv_smooth.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Check if objects are smooths or are a particular type of smooth — is_mgcv_smooth","text":"smooth R object, typically list","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/is_mgcv_smooth.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Check if objects are smooths or are a particular type of smooth — is_mgcv_smooth","text":"Check smooth inherits class \"mgcv.smooth\". stop_if_not_mgcv_smooth() wrapper around is_mgcv_smooth(), useful programming checking supplied object one mgcv's smooths, throwing consistent error . check_is_mgcv_smooth() similar stop_if_not_mgcv_smooth() returns result is_mgcv_smooth() invisibly.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/is_offset.html","id":null,"dir":"Reference","previous_headings":"","what":"Is a model term an offset? — is_offset","title":"Is a model term an offset? — is_offset","text":"Given character vector model terms, checks see , , model offset.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/is_offset.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Is a model term an offset? — is_offset","text":"","code":"is_offset(terms)"},{"path":"https://gavinsimpson.github.io/gratia/reference/is_offset.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Is a model term an offset? — is_offset","text":"terms character vector model terms.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/is_offset.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Is a model term an offset? — is_offset","text":"logical vector length terms.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/is_offset.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Is a model term an offset? — is_offset","text":"Gavin L. Simpson","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/is_offset.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Is a model term an offset? — is_offset","text":"","code":"load_mgcv() df <- data_sim(\"eg1\", n = 400, dist = \"normal\") m <- gam(y ~ s(x0) + s(x1) + offset(x0), data = df, method = \"REML\") nm <- names(model.frame(m)) nm #> [1] \"y\"          \"offset(x0)\" \"x0\"         \"x1\"         is_offset(nm) #> [1] FALSE  TRUE FALSE FALSE"},{"path":"https://gavinsimpson.github.io/gratia/reference/link.html","id":null,"dir":"Reference","previous_headings":"","what":"Extract link and inverse link functions from models — link","title":"Extract link and inverse link functions from models — link","text":"Returns link inverse estimated model, provides simple way extract functions complex models multiple links, location scale models.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/link.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Extract link and inverse link functions from models — link","text":"","code":"link(object, ...)  # S3 method for class 'family' link(object, parameter = NULL, which_eta = NULL, ...)  # S3 method for class 'gam' link(object, parameter = NULL, which_eta = NULL, ...)  # S3 method for class 'bam' link(object, parameter = NULL, which_eta = NULL, ...)  # S3 method for class 'gamm' link(object, ...)  # S3 method for class 'glm' link(object, ...)  # S3 method for class 'list' link(object, ...)  inv_link(object, ...)  # S3 method for class 'family' inv_link(object, parameter = NULL, which_eta = NULL, ...)  # S3 method for class 'gam' inv_link(object, parameter = NULL, which_eta = NULL, ...)  # S3 method for class 'bam' inv_link(object, parameter = NULL, which_eta = NULL, ...)  # S3 method for class 'gamm' inv_link(object, ...)  # S3 method for class 'list' inv_link(object, ...)  # S3 method for class 'glm' inv_link(object, ...)  extract_link(family, ...)  # S3 method for class 'family' extract_link(family, inverse = FALSE, ...)  # S3 method for class 'general.family' extract_link(family, parameter, inverse = FALSE, which_eta = NULL, ...)"},{"path":"https://gavinsimpson.github.io/gratia/reference/link.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Extract link and inverse link functions from models — link","text":"object family object fitted model extract family object.  Models fitted stats::glm(), mgcv::gam(), mgcv::bam(), mgcv::gamm(), gamm4::gamm4() currently supported. ... arguments passed methods. parameter character; parameter distribution. Usually \"location\" \"scale\" \"shape\" may provided location scale models. options include \"mu\" synonym \"location\", \"sigma\" scale parameter mgcv::gaulss(), \"pi\" zero-inflation term mgcv::ziplss(), \"power\" mgcv::twlss() power parameter, \"xi\", shape parameter mgcv::gevlss(), \"epsilon\" \"skewness\" skewness \"delta\" \"kurtosis\" kurtosis parameter mgcv::shash(), \"phi\" scale parameter mgcv::gammals() & mgcv::twlss(). which_eta numeric; linear predictor extract families mgcv::mvn() mgcv::multinom(). family family object, result call family(). inverse logical; return inverse link function?","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/link.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Extract link and inverse link functions from models — link","text":"Gavin L. Simpson","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/link.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Extract link and inverse link functions from models — link","text":"","code":"load_mgcv()  link(gaussian()) #> function (mu)  #> mu #> <environment: namespace:stats> link(nb()) #> function (mu)  #> log(mu) #> <environment: namespace:stats>  inv_link(nb()) #> function (eta)  #> pmax(exp(eta), .Machine$double.eps) #> <environment: namespace:stats>  dat <- data_sim(\"eg1\", seed = 4234) mod <- gam(list(y ~ s(x0) + s(x1) + s(x2) + s(x3), ~1),   data = dat,   family = gaulss )  link(mod, parameter = \"scale\") #> function (mu)  #> log(1/mu - 0.01) #> <environment: 0x56401dd37dd0> inv_link(mod, parameter = \"scale\") #> function (eta)  #> 1/(exp(eta) + 0.01) #> <environment: 0x56401dd37dd0>  ## Works with `family` objects too link(shash(), parameter = \"skewness\") #> function (mu)  #> mu #> <environment: namespace:stats>"},{"path":"https://gavinsimpson.github.io/gratia/reference/load_mgcv.html","id":null,"dir":"Reference","previous_headings":"","what":"Load mgcv quietly — load_mgcv","title":"Load mgcv quietly — load_mgcv","text":"Simple function loads mgcv package whilst suppressing startup messages prints console.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/load_mgcv.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Load mgcv quietly — load_mgcv","text":"","code":"load_mgcv()"},{"path":"https://gavinsimpson.github.io/gratia/reference/load_mgcv.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Load mgcv quietly — load_mgcv","text":"Returns logical vectors invisibly, indicating whether package loaded .","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/lp_matrix.html","id":null,"dir":"Reference","previous_headings":"","what":"Return the linear prediction matrix of a fitted GAM — lp_matrix","title":"Return the linear prediction matrix of a fitted GAM — lp_matrix","text":"lp_matrix() wrapper predict(..., type = \"lpmatrix\") returning linear predictor matrix model training data (data = NULL), user-specified data values supplied via data.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/lp_matrix.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Return the linear prediction matrix of a fitted GAM — lp_matrix","text":"","code":"lp_matrix(model, ...)  # S3 method for class 'gam' lp_matrix(model, data = NULL, ...)"},{"path":"https://gavinsimpson.github.io/gratia/reference/lp_matrix.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Return the linear prediction matrix of a fitted GAM — lp_matrix","text":"model fitted model ... arguments passed methods predict methods including mgcv::predict.gam() mgcv::predict.bam() data data frame values return linear prediction matrix.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/lp_matrix.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Return the linear prediction matrix of a fitted GAM — lp_matrix","text":"linear prediction matrix returned matrix. object returned class \"lp_matrix\", inherits classes \"matrix\" \"array\". special class allows printing matrix controlled, printing matrix tibble.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/lp_matrix.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Return the linear prediction matrix of a fitted GAM — lp_matrix","text":"linear prediction matrix \\(\\mathbf{X}_p\\) matrix maps values parameters \\(\\hat{\\mathbf{\\beta}}_p\\) values linear predictor model \\(\\hat{\\eta}_p = \\mathbf{X}_p \\hat{\\mathbf{\\beta}}_p\\). \\(\\mathbf{X}_p\\) model matrix spline covariates replaced values basis functions evaluated respective covariates. Parametric covariates also included.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/lp_matrix.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Return the linear prediction matrix of a fitted GAM — lp_matrix","text":"","code":"load_mgcv() df <- data_sim(\"eg1\", seed = 1) m <- gam(y ~ s(x0) + s(x1) + s(x2) + s(x3), data = df)  # linear prediction matrix for observed data xp <- lp_matrix(m) ## IGNORE_RDIFF_BEGIN xp #> Linear prediction matrix (400 x 37) #>   `(Intercept)` `s(x0).1` `s(x0).2` `s(x0).3` `s(x0).4` `s(x0).5` `s(x0).6` #>           <dbl>     <dbl>     <dbl>     <dbl>     <dbl>     <dbl>     <dbl> #> 1             1     0.961     0.227    0.706     -0.135     0.457  -0.146   #> 2             1     0.651    -0.241    0.0684    -0.308     0.394  -0.00994 #> 3             1    -0.385    -0.549    0.0660    -0.204    -0.416  -0.247   #> 4             1    -1.27      0.156   -1.53       0.222    -1.60    0.198   #> 5             1     1.05      0.420    1.11       0.214     0.893   0.0351  #> # i 395 more rows ## IGNORE_RDIFF_END  # the object `xp` *is* a matrix class(xp) #> [1] \"lp_matrix\" \"matrix\"    \"array\"     # but we print like a tibble to avoid spamming the R console  # linear predictor matrix for new data set ds <- data_slice(m, x2 = evenly(x2)) xp <- lp_matrix(m, data = ds) ## IGNORE_RDIFF_BEGIN xp #> Linear prediction matrix (100 x 37) #>   `(Intercept)` `s(x0).1` `s(x0).2` `s(x0).3` `s(x0).4` `s(x0).5` `s(x0).6` #>           <dbl>     <dbl>     <dbl>     <dbl>     <dbl>     <dbl>     <dbl> #> 1             1     0.170    -0.542   -0.0371   -0.0534     0.144    -0.353 #> 2             1     0.170    -0.542   -0.0371   -0.0534     0.144    -0.353 #> 3             1     0.170    -0.542   -0.0371   -0.0534     0.144    -0.353 #> 4             1     0.170    -0.542   -0.0371   -0.0534     0.144    -0.353 #> 5             1     0.170    -0.542   -0.0371   -0.0534     0.144    -0.353 #> # i 95 more rows ## IGNORE_RDIFF_END"},{"path":"https://gavinsimpson.github.io/gratia/reference/lss_parameters.html","id":null,"dir":"Reference","previous_headings":"","what":"General names of LSS parameters for each GAM family — lss_parameters","title":"General names of LSS parameters for each GAM family — lss_parameters","text":"General names LSS parameters GAM family","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/lss_parameters.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"General names of LSS parameters for each GAM family — lss_parameters","text":"","code":"lss_parameters(object)"},{"path":"https://gavinsimpson.github.io/gratia/reference/mh_draws.html","id":null,"dir":"Reference","previous_headings":"","what":"Posterior samples using a Gaussian approximation to the posterior distribution — mh_draws","title":"Posterior samples using a Gaussian approximation to the posterior distribution — mh_draws","text":"Posterior samples using Gaussian approximation posterior distribution","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/mh_draws.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Posterior samples using a Gaussian approximation to the posterior distribution — mh_draws","text":"","code":"mh_draws(model, ...)  # S3 method for class 'gam' mh_draws(   model,   n,   burnin = 1000,   thin = 1,   t_df = 40,   rw_scale = 0.25,   index = NULL,   ... )"},{"path":"https://gavinsimpson.github.io/gratia/reference/mh_draws.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Posterior samples using a Gaussian approximation to the posterior distribution — mh_draws","text":"model fitted R model. Currently models fitted mgcv::gam() mgcv::bam(), return object inherits objects supported. , \"inherits\" used loose fashion; models fitted scam::scam() support even though models strictly inherit class \"gam\" far inherits() concerned. ... arguments passed methods. n numeric; number posterior draws take. burnin numeric; length initial burn period discard. See mgcv::gam.mh(). thin numeric; retain thin samples. See mgcv::gam.mh(). t_df numeric; degrees freedom static multivariate t proposal. See mgcv::gam.mh(). rw_scale numeric; factor scale posterior covariance matrix generating random walk proposals. See mgcv::gam.mh(). index numeric; vector indices coefficients use. Can used subset mean vector covariance matrix extracted model.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/model_concurvity.html","id":null,"dir":"Reference","previous_headings":"","what":"Concurvity of an estimated GAM — model_concurvity","title":"Concurvity of an estimated GAM — model_concurvity","text":"Concurvity estimated GAM","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/model_concurvity.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Concurvity of an estimated GAM — model_concurvity","text":"","code":"model_concurvity(model, ...)  # S3 method for class 'gam' model_concurvity(   model,   terms = everything(),   type = c(\"all\", \"estimate\", \"observed\", \"worst\"),   pairwise = FALSE,   ... )  concrvity(   model,   terms = everything(),   type = c(\"all\", \"estimate\", \"observed\", \"worst\"),   pairwise = FALSE,   ... )"},{"path":"https://gavinsimpson.github.io/gratia/reference/model_concurvity.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Concurvity of an estimated GAM — model_concurvity","text":"model fitted GAM. Currently objects class \"gam\" supported ... arguents passed methods. terms currently ignored type character; pairwise logical; extract pairwise concurvity model terms?","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/model_concurvity.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Concurvity of an estimated GAM — model_concurvity","text":"","code":"## simulate data with concurvity... library(\"tibble\") load_mgcv() set.seed(8) n <- 200 df <- tibble(   t = sort(runif(n)),   x = gw_f2(t) + rnorm(n) * 3,   y = sin(4 * pi * t) + exp(x / 20) + rnorm(n) * 0.3 )  ## fit model m <- gam(y ~ s(t, k = 15) + s(x, k = 15), data = df, method = \"REML\")  ## overall concurvity o_conc <- concrvity(m) draw(o_conc)   ## pairwise concurvity p_conc <- concrvity(m, pairwise = TRUE) draw(p_conc)"},{"path":"https://gavinsimpson.github.io/gratia/reference/model_constant.html","id":null,"dir":"Reference","previous_headings":"","what":"Extract the model constant term — model_constant","title":"Extract the model constant term — model_constant","text":"Extracts model constant term(s), model intercept, fitted model object.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/model_constant.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Extract the model constant term — model_constant","text":"","code":"model_constant(model, ...)  # S3 method for class 'gam' model_constant(model, lp = NULL, ...)  # S3 method for class 'gamlss' model_constant(model, ...)"},{"path":"https://gavinsimpson.github.io/gratia/reference/model_constant.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Extract the model constant term — model_constant","text":"model fitted model coef() method exists. ... arguments passed methods. lp numeric; linear predictors extract constant terms .","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/model_constant.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Extract the model constant term — model_constant","text":"","code":"load_mgcv()  # simulate a small example df <- data_sim(\"eg1\", seed = 42)  # fit the GAM m <- gam(y ~ s(x0) + s(x1) + s(x2) + s(x3), data = df, method = \"REML\")  # extract the estimate of the constant term model_constant(m) #> [1] 7.495 #> attr(,\"par_names\") #> [1] \"location\" # same as coef(m)[1L] coef(m)[1L] #> (Intercept)  #>       7.495"},{"path":"https://gavinsimpson.github.io/gratia/reference/model_vars.html","id":null,"dir":"Reference","previous_headings":"","what":"List the variables involved in a model fitted with a formula — model_vars","title":"List the variables involved in a model fitted with a formula — model_vars","text":"List variables involved model fitted formula","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/model_vars.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"List the variables involved in a model fitted with a formula — model_vars","text":"","code":"model_vars(model, ...)  # S3 method for class 'gam' model_vars(model, ...)  # Default S3 method model_vars(model, ...)  # S3 method for class 'bam' model_vars(model, ...)  # S3 method for class 'gamm' model_vars(model, ...)  # S3 method for class 'gamm4' model_vars(model, ...)  # S3 method for class 'list' model_vars(model, ...)"},{"path":"https://gavinsimpson.github.io/gratia/reference/model_vars.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"List the variables involved in a model fitted with a formula — model_vars","text":"model fitted model object $pred.formula, $terms component \"terms\" attribute ... Arguments passed methods. Currently ignored.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/model_vars.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"List the variables involved in a model fitted with a formula — model_vars","text":"","code":"load_mgcv()  # simulate some Gaussian data df <- data_sim(\"eg1\", n = 50, seed = 2)  # fit a GAM with 1 smooth and 1 linear term m1 <- gam(y ~ s(x2, k = 7) + x1, data = df, method = \"REML\") model_vars(m1) #> [1] \"x1\" \"x2\"  # fit a lm with two linear terms m2 <- lm(y ~ x2 + x1, data = df) model_vars(m2) #> [1] \"x2\" \"x1\""},{"path":"https://gavinsimpson.github.io/gratia/reference/n_eta.html","id":null,"dir":"Reference","previous_headings":"","what":"The Number of linear predictors in model — n_eta","title":"The Number of linear predictors in model — n_eta","text":"Extracts number linear predictors fitted model.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/n_eta.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"The Number of linear predictors in model — n_eta","text":"","code":"n_eta(model, ...)  # S3 method for class 'gam' n_eta(model, ...)"},{"path":"https://gavinsimpson.github.io/gratia/reference/n_eta.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"The Number of linear predictors in model — n_eta","text":"model fitted model. Currently, models inheriting class \"gam\" supported. ... arguments passed methods.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/n_eta.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"The Number of linear predictors in model — n_eta","text":"integer vector length 1 containing number linear predictors model.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/n_smooths.html","id":null,"dir":"Reference","previous_headings":"","what":"How many smooths in a fitted model — n_smooths","title":"How many smooths in a fitted model — n_smooths","text":"many smooths fitted model","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/n_smooths.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"How many smooths in a fitted model — n_smooths","text":"","code":"n_smooths(object)  # Default S3 method n_smooths(object)  # S3 method for class 'gam' n_smooths(object)  # S3 method for class 'gamm' n_smooths(object)  # S3 method for class 'bam' n_smooths(object)"},{"path":"https://gavinsimpson.github.io/gratia/reference/n_smooths.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"How many smooths in a fitted model — n_smooths","text":"object fitted GAM related model. Typically result call mgcv::gam(), mgcv::bam(), mgcv::gamm().","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/nb_theta.html","id":null,"dir":"Reference","previous_headings":"","what":"Negative binomial parameter theta — nb_theta","title":"Negative binomial parameter theta — nb_theta","text":"Negative binomial parameter theta","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/nb_theta.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Negative binomial parameter theta — nb_theta","text":"","code":"nb_theta(model)  # S3 method for class 'gam' nb_theta(model)"},{"path":"https://gavinsimpson.github.io/gratia/reference/nb_theta.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Negative binomial parameter theta — nb_theta","text":"model fitted model.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/nb_theta.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Negative binomial parameter theta — nb_theta","text":"numeric vector length 1 containing estimated value theta.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/nb_theta.html","id":"methods-by-class-","dir":"Reference","previous_headings":"","what":"Methods (by class)","title":"Negative binomial parameter theta — nb_theta","text":"nb_theta(gam): Method class \"gam\"","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/nb_theta.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Negative binomial parameter theta — nb_theta","text":"","code":"load_mgcv() df <- data_sim(\"eg1\", n = 500, dist = \"poisson\", scale = 0.1, seed = 6)  m <- gam(y ~ s(x0, bs = \"cr\") + s(x1, bs = \"cr\") + s(x2, bs = \"cr\") +   s(x3, bs = \"cr\"), family = nb, data = df, method = \"REML\") ## IGNORE_RDIFF_BEGIN nb_theta(m) #> [1] 239333.8 ## IGNORE_RDIFF_END"},{"path":"https://gavinsimpson.github.io/gratia/reference/nested_partial_residuals.html","id":null,"dir":"Reference","previous_headings":"","what":"Partial residuals in nested form — nested_partial_residuals","title":"Partial residuals in nested form — nested_partial_residuals","text":"Computes partial residuals smooth terms, formats long/tidy format, nests partial_residual column result nested data frame one row per smooth.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/nested_partial_residuals.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Partial residuals in nested form — nested_partial_residuals","text":"","code":"nested_partial_residuals(object, terms = NULL, data = NULL)"},{"path":"https://gavinsimpson.github.io/gratia/reference/nested_partial_residuals.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Partial residuals in nested form — nested_partial_residuals","text":"object fitted GAM model terms vector terms include partial residuals . Passed argument terms mgcv::predict.gam()]. data optional data frame","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/nested_partial_residuals.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Partial residuals in nested form — nested_partial_residuals","text":"nested tibble (data frame) one row per smooth term. Contains two columns: smooth - label indicating smooth term partial_residual - list column containing tibble (data frame) 1 column partial_residual containing requested partial residuals indicated smooth.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/nested_rug_values.html","id":null,"dir":"Reference","previous_headings":"","what":"Values for rug plot in nested form — nested_rug_values","title":"Values for rug plot in nested form — nested_rug_values","text":"Extracts original data smooth terms, formats long/tidy format, nests data column(s) result nested data frame one row per smooth.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/nested_rug_values.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Values for rug plot in nested form — nested_rug_values","text":"","code":"nested_rug_values(object, terms = NULL, data = NULL)"},{"path":"https://gavinsimpson.github.io/gratia/reference/nested_rug_values.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Values for rug plot in nested form — nested_rug_values","text":"object fitted GAM model terms vector terms include original data . Passed argument terms mgcv::predict.gam()]. data optional data frame","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/nested_rug_values.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Values for rug plot in nested form — nested_rug_values","text":"nested tibble (data frame) one row per smooth term.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/null_deviance.html","id":null,"dir":"Reference","previous_headings":"","what":"Extract the null deviance of a fitted model — null_deviance","title":"Extract the null deviance of a fitted model — null_deviance","text":"Extract null deviance fitted model","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/null_deviance.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Extract the null deviance of a fitted model — null_deviance","text":"","code":"null_deviance(model, ...)  # Default S3 method null_deviance(model, ...)"},{"path":"https://gavinsimpson.github.io/gratia/reference/null_deviance.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Extract the null deviance of a fitted model — null_deviance","text":"model fitted model ... arguments passed methods","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/observed_fitted_plot.html","id":null,"dir":"Reference","previous_headings":"","what":"Plot of fitted against observed response values — observed_fitted_plot","title":"Plot of fitted against observed response values — observed_fitted_plot","text":"Plot fitted observed response values","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/observed_fitted_plot.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Plot of fitted against observed response values — observed_fitted_plot","text":"","code":"observed_fitted_plot(   model,   ylab = NULL,   xlab = NULL,   title = NULL,   subtitle = NULL,   caption = NULL,   point_col = \"black\",   point_alpha = 1 )"},{"path":"https://gavinsimpson.github.io/gratia/reference/observed_fitted_plot.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Plot of fitted against observed response values — observed_fitted_plot","text":"model fitted model. Currently class \"gam\". ylab character expression; label y axis. supplied, suitable label generated. xlab character expression; label y axis. supplied, suitable label generated. title character expression; title plot. See ggplot2::labs(). subtitle character expression; subtitle plot. See ggplot2::labs(). caption character expression; plot caption. See ggplot2::labs(). point_col colour used draw points plots. See graphics::par() section Color Specification. passed individual plotting functions, therefore affects points plots. point_alpha numeric; alpha transparency points plots.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/overview.html","id":null,"dir":"Reference","previous_headings":"","what":"Provides an overview of a model and the terms in that model — overview","title":"Provides an overview of a model and the terms in that model — overview","text":"Provides overview model terms model","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/overview.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Provides an overview of a model and the terms in that model — overview","text":"","code":"overview(model, ...)  # S3 method for class 'gam' overview(   model,   parametric = TRUE,   random_effects = TRUE,   dispersion = NULL,   frequentist = FALSE,   accuracy = 0.001,   stars = FALSE,   ... )"},{"path":"https://gavinsimpson.github.io/gratia/reference/overview.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Provides an overview of a model and the terms in that model — overview","text":"model fitted model object overview. ... arguments passed methods. parametric logical; include model parametric terms overview? random_effects tests fully penalized smooth terms (zero-dimensional null space, e.g. random effects) computationally expensive large data sets producing p values can take long time. random_effects = FALSE, tests expensive terms skipped. dispersion numeric; known value dispersion parameter. default NULL implies estimated value default value (1 Poisson distribution example) specified used instead. frequentist logical; default Bayesian estimated covariance matrix parameter estimates used calculate p values parametric terms. frequentist = FALSE, frequentist covariance matrix parameter estimates used. accuracy numeric; accuracy report p values, p values value displayed \"< accuracy\". stars logical; significance stars added output?","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/overview.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Provides an overview of a model and the terms in that model — overview","text":"","code":"load_mgcv() df <- data_sim(n = 400, seed = 2) m <- gam(y ~ x3 + s(x0) + s(x1, bs = \"bs\") + s(x2, bs = \"ts\"),   data = df, method = \"REML\" ) overview(m) #>  #> Generalized Additive Model with 4 terms #>  #>   term  type              k   edf statistic p.value #>   <chr> <chr>         <dbl> <dbl>     <dbl> <chr>   #> 1 x3    parametric       NA  1         4.28 0.03926 #> 2 s(x0) TPRS              9  3.02      6.25 < 0.001 #> 3 s(x1) B spline          9  2.81     71.0  < 0.001 #> 4 s(x2) TPRS (shrink)     9  7.91     83.8  < 0.001"},{"path":"https://gavinsimpson.github.io/gratia/reference/parametric_effects.html","id":null,"dir":"Reference","previous_headings":"","what":"Estimated values for parametric model terms — parametric_effects","title":"Estimated values for parametric model terms — parametric_effects","text":"Estimated values parametric model terms","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/parametric_effects.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Estimated values for parametric model terms — parametric_effects","text":"","code":"parametric_effects(object, ...)  # S3 method for class 'gam' parametric_effects(   object,   terms = NULL,   data = NULL,   unconditional = FALSE,   unnest = TRUE,   ci_level = 0.95,   envir = environment(formula(object)),   transform = FALSE,   ... )"},{"path":"https://gavinsimpson.github.io/gratia/reference/parametric_effects.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Estimated values for parametric model terms — parametric_effects","text":"object fitted model object. ... arguments passed methods. terms character; model parametric terms drawn? Default NULL plot parametric terms can drawn. data optional data frame may may used? FIXME! unconditional logical; confidence intervals include uncertainty due smoothness selection? TRUE, corrected Bayesian covariance matrix used. unnest logical; unnest parametric effect objects? ci_level numeric; coverage required confidence interval. Currently ignored. envir environment look data within. transform logical; TRUE, parametric effect plotted transformed scale result effect straight line. FALSE, effect plotted raw data (.e. log10(x), poly(z), x-axis plot x z respectively.)","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/parametric_terms.html","id":null,"dir":"Reference","previous_headings":"","what":"Names of any parametric terms in a GAM — parametric_terms","title":"Names of any parametric terms in a GAM — parametric_terms","text":"Names parametric terms GAM","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/parametric_terms.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Names of any parametric terms in a GAM — parametric_terms","text":"","code":"parametric_terms(model, ...)  # Default S3 method parametric_terms(model, ...)  # S3 method for class 'gam' parametric_terms(model, ...)"},{"path":"https://gavinsimpson.github.io/gratia/reference/parametric_terms.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Names of any parametric terms in a GAM — parametric_terms","text":"model fitted model. ... arguments passed methods.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/partial_derivatives.html","id":null,"dir":"Reference","previous_headings":"","what":"Partial derivatives of estimated multivariate smooths via finite differences — partial_derivatives","title":"Partial derivatives of estimated multivariate smooths via finite differences — partial_derivatives","text":"Partial derivatives estimated multivariate smooths via finite differences","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/partial_derivatives.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Partial derivatives of estimated multivariate smooths via finite differences — partial_derivatives","text":"","code":"partial_derivatives(object, ...)  # Default S3 method partial_derivatives(object, ...)  # S3 method for class 'gamm' partial_derivatives(object, ...)  # S3 method for class 'gam' partial_derivatives(   object,   select = NULL,   term = deprecated(),   focal = NULL,   data = newdata,   order = 1L,   type = c(\"forward\", \"backward\", \"central\"),   n = 100,   eps = 1e-07,   interval = c(\"confidence\", \"simultaneous\"),   n_sim = 10000,   level = 0.95,   unconditional = FALSE,   frequentist = FALSE,   offset = NULL,   ncores = 1,   partial_match = FALSE,   seed = NULL,   ...,   newdata = NULL )"},{"path":"https://gavinsimpson.github.io/gratia/reference/partial_derivatives.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Partial derivatives of estimated multivariate smooths via finite differences — partial_derivatives","text":"object R object compute derivatives . ... arguments passed methods. select character; vector one smooth terms derivatives required. missing, derivatives smooth terms returned. Can partial match smooth term; see argument partial_match . term Use select instead. focal character; name focal variable. partial derivative estimated smooth respect variable returned. variables involved smooth held constant. can missing supplying data, case, focal variable identified one variable constant. data data frame containing values model covariates evaluate first derivatives smooths. supplied, one variable must held constant value. order numeric; order derivative. type character; type finite difference used. One \"forward\", \"backward\", \"central\". n numeric; number points evaluate derivative . eps numeric; finite difference. interval character; type interval compute. One \"confidence\" point-wise intervals, \"simultaneous\" simultaneous intervals. n_sim integer; number simulations used computing simultaneous intervals. level numeric; 0 < level < 1; confidence level point-wise simultaneous interval. default 0.95 95% interval. unconditional logical; use smoothness selection-corrected Bayesian covariance matrix? frequentist logical; use frequentist covariance matrix? offset numeric; value use offset term ncores number cores generating random variables multivariate normal distribution. Passed mvnfast::rmvn(). Parallelization take place OpenMP supported (appears work Windows current R). partial_match logical; smooths selected partial matches term? TRUE, term can single string match . seed numeric; RNG seed use. newdata Deprecated: use data instead.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/partial_derivatives.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Partial derivatives of estimated multivariate smooths via finite differences — partial_derivatives","text":"tibble, currently following variables: .smooth: smooth row refers , .partial_deriv: estimated partial derivative, .se: standard error estimated partial derivative, .crit: critical value derivative ± (crit * se) gives upper lower bounds requested confidence simultaneous interval (given level), .lower_ci: lower bound confidence simultaneous interval, .upper_ci: upper bound confidence simultaneous interval.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/partial_derivatives.html","id":"note","dir":"Reference","previous_headings":"","what":"Note","title":"Partial derivatives of estimated multivariate smooths via finite differences — partial_derivatives","text":"partial_derivatives() ignore random effect smooths encounters object.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/partial_derivatives.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Partial derivatives of estimated multivariate smooths via finite differences — partial_derivatives","text":"Gavin L. Simpson","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/partial_derivatives.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Partial derivatives of estimated multivariate smooths via finite differences — partial_derivatives","text":"","code":"library(\"ggplot2\") library(\"patchwork\") load_mgcv() df <- data_sim(\"eg2\", n = 2000, dist = \"normal\", scale = 0.5, seed = 42)  # fit the GAM (note: for execution time reasons, k is set articifially low) m <- gam(y ~ te(x, z, k = c(5, 5)), data = df, method = \"REML\")  # data slice through te(x,z) holding z == 0.4 ds <- data_slice(m, x = evenly(x, n = 100), z = 0.4)  # evaluate te(x,z) at values of x & z sm <- smooth_estimates(m, select = \"te(x,z)\", data = ds) |>   add_confint()  # partial derivatives pd_x <- partial_derivatives(m, data = ds, type = \"central\", focal = \"x\")  # draw te(x,z) p1 <- draw(m, rug = FALSE) &   geom_hline(yintercept = 0.4, linewidth = 1) p1   # draw te(x,z) along slice cap <- expression(z == 0.4) p2 <- sm |>   ggplot(aes(x = x, y = .estimate)) +   geom_ribbon(aes(ymin = .lower_ci, ymax = .upper_ci), alpha = 0.2) +   geom_line() +   labs(     x = \"x\", y = \"Partial effect\", title = \"te(x,z)\",     caption = cap   ) p2   # draw partial derivs p3 <- pd_x |>   draw() +   labs(caption = cap) p3   # draw all three panels p1 + p2 + p3 + plot_layout(ncol = 3)"},{"path":"https://gavinsimpson.github.io/gratia/reference/partial_residuals.html","id":null,"dir":"Reference","previous_headings":"","what":"Partial residuals — partial_residuals","title":"Partial residuals — partial_residuals","text":"Partial residuals","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/partial_residuals.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Partial residuals — partial_residuals","text":"","code":"partial_residuals(object, ...)  # S3 method for class 'gam' partial_residuals(object, select = NULL, partial_match = FALSE, ...)"},{"path":"https://gavinsimpson.github.io/gratia/reference/partial_residuals.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Partial residuals — partial_residuals","text":"object R object, typically model. Currently objects class \"gam\" (inherit class) supported. ... arguments passed methods. select character, logical, numeric; smooths plot. NULL, default, model smooths drawn. Numeric select indexes smooths order specified formula stored object. Character select matches labels smooths shown example output summary(object). Logical select operates per numeric select order smooths stored. partial_match logical; smooths selected partial matches select? TRUE, select can single string match .","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/partial_residuals.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Partial residuals — partial_residuals","text":"","code":"## load mgcv load_mgcv()  ## example data - Gu & Wabha four term model df <- data_sim(\"eg1\", n = 400, seed = 42) ## fit the model m <- gam(y ~ s(x0) + s(x1) + s(x2) + s(x3), data = df, method = \"REML\")  ## extract partial residuals partial_residuals(m) #> # A tibble: 400 x 4 #>    `s(x0)`  `s(x1)` `s(x2)`  `s(x3)` #>      <dbl>    <dbl>   <dbl>    <dbl> #>  1 -0.3527 -1.321   -2.180   0.6077  #>  2 -0.1233  0.5013  -1.775   0.9613  #>  3  1.429   1.515    5.609   0.9910  #>  4 -1.110  -1.700   -0.8882 -0.6593  #>  5 -2.120  -0.01378 -2.733  -3.012   #>  6  1.254  -1.224    3.915   0.07275 #>  7 -0.5220  3.023   -0.8197 -1.019   #>  8  1.398   0.2184   7.055   1.897   #>  9  2.797   0.4969   7.329   2.498   #> 10  1.151  -0.2267   0.7202  0.7437  #> # i 390 more rows  ## and for a select term partial_residuals(m, select = \"s(x2)\") #> # A tibble: 400 x 1 #>    `s(x2)` #>      <dbl> #>  1 -2.180  #>  2 -1.775  #>  3  5.609  #>  4 -0.8882 #>  5 -2.733  #>  6  3.915  #>  7 -0.8197 #>  8  7.055  #>  9  7.329  #> 10  0.7202 #> # i 390 more rows  ## or with partial matching partial_residuals(m, select = \"x\", partial_match = TRUE) # returns all #> # A tibble: 400 x 4 #>    `s(x0)`  `s(x1)` `s(x2)`  `s(x3)` #>      <dbl>    <dbl>   <dbl>    <dbl> #>  1 -0.3527 -1.321   -2.180   0.6077  #>  2 -0.1233  0.5013  -1.775   0.9613  #>  3  1.429   1.515    5.609   0.9910  #>  4 -1.110  -1.700   -0.8882 -0.6593  #>  5 -2.120  -0.01378 -2.733  -3.012   #>  6  1.254  -1.224    3.915   0.07275 #>  7 -0.5220  3.023   -0.8197 -1.019   #>  8  1.398   0.2184   7.055   1.897   #>  9  2.797   0.4969   7.329   2.498   #> 10  1.151  -0.2267   0.7202  0.7437  #> # i 390 more rows"},{"path":"https://gavinsimpson.github.io/gratia/reference/penalty.html","id":null,"dir":"Reference","previous_headings":"","what":"Extract and tidy penalty matrices — penalty","title":"Extract and tidy penalty matrices — penalty","text":"Extract tidy penalty matrices","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/penalty.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Extract and tidy penalty matrices — penalty","text":"","code":"penalty(object, ...)  # Default S3 method penalty(   object,   rescale = FALSE,   data,   knots = NULL,   constraints = FALSE,   diagonalize = FALSE,   ... )  # S3 method for class 'gam' penalty(   object,   select = NULL,   smooth = deprecated(),   rescale = FALSE,   partial_match = FALSE,   ... )  # S3 method for class 'mgcv.smooth' penalty(object, rescale = FALSE, ...)  # S3 method for class 'tensor.smooth' penalty(object, margins = FALSE, ...)  # S3 method for class 't2.smooth' penalty(object, margins = FALSE, ...)  # S3 method for class 're.smooth.spec' penalty(object, data, ...)"},{"path":"https://gavinsimpson.github.io/gratia/reference/penalty.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Extract and tidy penalty matrices — penalty","text":"object fitted GAM smooth. ... additional arguments passed methods. rescale logical; default, mgcv scale penalty matrix better performance mgcv::gamm(). rescale TRUE, scaling undone put penalty matrix back original scale. data data frame; data frame values terms mentioned smooth specification. knots list data frame named components containing knots locations. Names must match covariates basis required. See mgcv::smoothCon(). constraints logical; identifiability constraints applied smooth basis. See argument absorb.cons mgcv::smoothCon(). diagonalize logical; TRUE, reparameterises smooth associated penalty identity matrix. effect turning last diagonal elements penalty zero, highlights penalty null space. select character, logical, numeric; smooths extract penalties . NULL, default, penalties model smooths drawn. Numeric select indexes smooths order specified formula stored object. Character select matches labels smooths shown example output summary(object). Logical select operates per numeric select order smooths stored. smooth Use select instead. partial_match logical; smooths selected partial matches select? TRUE, select can single string match . margins logical; extract penalty matrices tensor product marginal smooths tensor product?","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/penalty.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Extract and tidy penalty matrices — penalty","text":"'tibble' (data frame) class penalty_df inheriting tbl_df, following components: .smooth - character; label mgcv uses refer smooth, .type - character; type smooth, .penalty - character; label specific penalty. smooths multiple penalty matrices, penalty component identifies particular penalty matrix uses labelling mgcv uses internally, .row - character; label form fn n integer nth basis function, referencing columns penalty matrix, .col - character; label form fn n integer nth basis function, referencing columns penalty matrix, .value - double; value penalty matrix combination row col,","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/penalty.html","id":"note","dir":"Reference","previous_headings":"","what":"Note","title":"Extract and tidy penalty matrices — penalty","text":"print() method uses base::zapsmall() turn small numbers 0s display purposes ; underlying values penalty matrix matrices changed. smooths subject eigendecomposition (e.g. default thin plate regression splines, bs = \"tp\"), signs eigenvectors defined can expect differences across systems penalties smooths system-, OS-, CPU architecture- specific.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/penalty.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Extract and tidy penalty matrices — penalty","text":"Gavin L. Simpson","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/penalty.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Extract and tidy penalty matrices — penalty","text":"","code":"load_mgcv() dat <- data_sim(\"eg4\", n = 400, seed = 42) m <- gam(   y ~ s(x0, bs = \"cr\") + s(x1, bs = \"cr\") +     s(x2, by = fac, bs = \"cr\"),   data = dat, method = \"REML\" )  # penalties for all smooths penalty(m) #> # A tibble: 405 x 6 #>    .smooth .type .penalty .row  .col   .value #>    <chr>   <chr> <chr>    <chr> <chr>   <dbl> #>  1 s(x0)   CRS   s(x0)    F1    F1     0.783  #>  2 s(x0)   CRS   s(x0)    F1    F2    -0.635  #>  3 s(x0)   CRS   s(x0)    F1    F3     0.265  #>  4 s(x0)   CRS   s(x0)    F1    F4    -0.0203 #>  5 s(x0)   CRS   s(x0)    F1    F5     0.0441 #>  6 s(x0)   CRS   s(x0)    F1    F6     0.0378 #>  7 s(x0)   CRS   s(x0)    F1    F7     0.0482 #>  8 s(x0)   CRS   s(x0)    F1    F8     0.0216 #>  9 s(x0)   CRS   s(x0)    F1    F9     0.0247 #> 10 s(x0)   CRS   s(x0)    F2    F1    -0.635  #> # i 395 more rows  # for a specific smooth penalty(m, select = \"s(x2):fac1\") #> # A tibble: 81 x 6 #>    .smooth    .type .penalty   .row  .col   .value #>    <chr>      <chr> <chr>      <chr> <chr>   <dbl> #>  1 s(x2):fac1 CRS   s(x2):fac1 F1    F1     1.66   #>  2 s(x2):fac1 CRS   s(x2):fac1 F1    F2    -0.755  #>  3 s(x2):fac1 CRS   s(x2):fac1 F1    F3     0.430  #>  4 s(x2):fac1 CRS   s(x2):fac1 F1    F4     0.0846 #>  5 s(x2):fac1 CRS   s(x2):fac1 F1    F5     0.192  #>  6 s(x2):fac1 CRS   s(x2):fac1 F1    F6     0.152  #>  7 s(x2):fac1 CRS   s(x2):fac1 F1    F7     0.188  #>  8 s(x2):fac1 CRS   s(x2):fac1 F1    F8     0.164  #>  9 s(x2):fac1 CRS   s(x2):fac1 F1    F9     0.0597 #> 10 s(x2):fac1 CRS   s(x2):fac1 F2    F1    -0.755  #> # i 71 more rows"},{"path":"https://gavinsimpson.github.io/gratia/reference/post_draws.html","id":null,"dir":"Reference","previous_headings":"","what":"Low-level Functions to generate draws from the posterior distribution of model coefficients — post_draws","title":"Low-level Functions to generate draws from the posterior distribution of model coefficients — post_draws","text":"Low-level Functions generate draws posterior distribution model coefficients Generate posterior draws fitted model","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/post_draws.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Low-level Functions to generate draws from the posterior distribution of model coefficients — post_draws","text":"","code":"post_draws(model, ...)  # Default S3 method post_draws(   model,   n,   method = c(\"gaussian\", \"mh\", \"inla\", \"user\"),   mu = NULL,   sigma = NULL,   n_cores = 1L,   burnin = 1000,   thin = 1,   t_df = 40,   rw_scale = 0.25,   index = NULL,   frequentist = FALSE,   unconditional = FALSE,   parametrized = TRUE,   mvn_method = c(\"mvnfast\", \"mgcv\"),   draws = NULL,   seed = NULL,   ... )  generate_draws(model, ...)  # S3 method for class 'gam' generate_draws(   model,   n,   method = c(\"gaussian\", \"mh\", \"inla\"),   mu = NULL,   sigma = NULL,   n_cores = 1L,   burnin = 1000,   thin = 1,   t_df = 40,   rw_scale = 0.25,   index = NULL,   frequentist = FALSE,   unconditional = FALSE,   mvn_method = c(\"mvnfast\", \"mgcv\"),   seed = NULL,   ... )"},{"path":"https://gavinsimpson.github.io/gratia/reference/post_draws.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Low-level Functions to generate draws from the posterior distribution of model coefficients — post_draws","text":"model fitted R model. Currently models fitted mgcv::gam() mgcv::bam(), return object inherits objects supported. , \"inherits\" used loose fashion; models fitted scam::scam() support even though models strictly inherit class \"gam\" far inherits() concerned. ... arguments passed methods. n numeric; number posterior draws take. method character; algorithm use sample posterior. Currently implemented methods : \"gaussian\" \"mh\". \"gaussian\" calls gaussian_draws() uses Gaussian approximation posterior distribution. \"mh\" uses simple Metropolis Hasting sampler alternates static proposals based Gaussian approximation posterior, random walk proposals. Note, setting t_df low value result heavier-tailed statistic proposals. See mgcv::gam.mh() details. mu numeric; user-supplied mean vector (vector model coefficients). Currently ignored. sigma matrix; user-supplied covariance matrix mu. Currently ignored. n_cores integer; number CPU cores use generating multivariate normal distributed random values. used mvn_method = \"mvnfast\" method = \"gaussian\". burnin numeric; length initial burn period discard. See mgcv::gam.mh(). thin numeric; retain thin samples. See mgcv::gam.mh(). t_df numeric; degrees freedom static multivariate t proposal. See mgcv::gam.mh(). rw_scale numeric; factor scale posterior covariance matrix generating random walk proposals. See mgcv::gam.mh(). index numeric; vector indices coefficients use. Can used subset mean vector covariance matrix extracted model. frequentist logical; TRUE, frequentist covariance matrix parameter estimates used. FALSE, Bayesian posterior covariance matrix parameters used. See mgcv::vcov.gam(). unconditional logical; TRUE  Bayesian smoothing parameter uncertainty corrected covariance matrix used, available model. See mgcv::vcov.gam(). parametrized logical; use parametrized coefficients covariance matrix, respect linear inequality constraints model. scam::scam() model fits. mvn_method character; one \"mvnfast\" \"mgcv\". default uses mvnfast::rmvn(), can considerably faster generate large numbers MVN random values mgcv::rmvn(), might work marginal fits, covariance matrix close singular. draws matrix; user supplied posterior draws used method = \"user\". seed numeric; random seed use. NULL, random seed generated without affecting current state R's RNG.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/post_link_funs.html","id":null,"dir":"Reference","previous_headings":"","what":"A list of transformation functions named for LSS parameters in a GAMLSS — post_link_funs","title":"A list of transformation functions named for LSS parameters in a GAMLSS — post_link_funs","text":"list transformation functions named LSS parameters GAMLSS","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/post_link_funs.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"A list of transformation functions named for LSS parameters in a GAMLSS — post_link_funs","text":"","code":"post_link_funs(   location = identity_fun,   scale = identity_fun,   shape = identity_fun,   skewness = identity_fun,   kurtosis = identity_fun,   power = identity_fun,   pi = identity_fun )"},{"path":"https://gavinsimpson.github.io/gratia/reference/posterior_samples.html","id":null,"dir":"Reference","previous_headings":"","what":"Draw samples from the posterior distribution of an estimated model — posterior_samples","title":"Draw samples from the posterior distribution of an estimated model — posterior_samples","text":"Draw samples posterior distribution estimated model","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/posterior_samples.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Draw samples from the posterior distribution of an estimated model — posterior_samples","text":"","code":"posterior_samples(model, ...)  # S3 method for class 'gam' posterior_samples(   model,   n = 1,   data = newdata,   seed = NULL,   method = c(\"gaussian\", \"mh\", \"inla\", \"user\"),   n_cores = 1,   burnin = 1000,   thin = 1,   t_df = 40,   rw_scale = 0.25,   freq = FALSE,   unconditional = FALSE,   weights = NULL,   draws = NULL,   mvn_method = c(\"mvnfast\", \"mgcv\"),   ...,   newdata = NULL,   ncores = NULL )"},{"path":"https://gavinsimpson.github.io/gratia/reference/posterior_samples.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Draw samples from the posterior distribution of an estimated model — posterior_samples","text":"model fitted model supported types ... arguments passed methods. fitted_samples(), passed mgcv::predict.gam(). posterior_samples() passed fitted_samples(). predicted_samples() passed relevant simulate() method. n numeric; number posterior samples return. data data frame; new observations posterior draws model evaluated. supplied, data used fit model used data, available model. seed numeric; random seed simulations. method character; method used draw samples posterior distribution. \"gaussian\" uses Gaussian (Laplace) approximation posterior. \"mh\" uses Metropolis Hastings sampler alternates t proposals proposals based shrunken version posterior covariance matrix. \"inla\" uses variant Integrated Nested Laplace Approximation due Wood (2019), (currently implemented). \"user\" allows user-supplied posterior draws (currently implemented). n_cores number cores generating random variables multivariate normal distribution. Passed mvnfast::rmvn(). Parallelization take place OpenMP supported (appears work Windows current R). burnin numeric; number samples discard burnin draws. used method = \"mh\". thin numeric; number samples skip taking n draws. Results thin * n draws posterior taken. used method = \"mh\". t_df numeric; degrees freedom t distribution proposals. used method = \"mh\". rw_scale numeric; Factor scale posterior covariance matrix generating random walk proposals. Negative non finite skip random walk step. used method = \"mh\". freq logical; TRUE use frequentist covariance matrix parameter estimators, FALSE use Bayesian posterior covariance matrix parameters. unconditional logical; TRUE (freq == FALSE) Bayesian smoothing parameter uncertainty corrected covariance matrix used, available. weights numeric; vector prior weights. data null defaults object[[\"prior.weights\"]], otherwise vector ones. draws matrix; user supplied posterior draws used method = \"user\". mvn_method character; one \"mvnfast\" \"mgcv\". default uses mvnfast::rmvn(), can considerably faster generate large numbers MVN random values mgcv::rmvn(), might work marginal fits, covariance matrix close singular. newdata Deprecated: use data instead. ncores Deprecated; use n_cores instead. number cores generating random variables multivariate normal distribution. Passed mvnfast::rmvn(). Parallelization take place OpenMP supported (appears work Windows current R).","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/posterior_samples.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Draw samples from the posterior distribution of an estimated model — posterior_samples","text":"tibble (data frame) 3 columns containing posterior predicted values long format. columns row (integer) row data posterior draw relates , draw (integer) index, range 1:n, indicating draw row relates , response (numeric) predicted response indicated row data.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/posterior_samples.html","id":"note","dir":"Reference","previous_headings":"","what":"Note","title":"Draw samples from the posterior distribution of an estimated model — posterior_samples","text":"Models offset terms supplied via offset argument mgcv::gam() etc. ignored mgcv::predict.gam(). , kind offset term also ignored posterior_samples(). Offset terms included model formula supplied mgcv::gam() etc ignored posterior samples produced reflect offset term values. side effect requiring new data values provided posterior_samples() via data argument must include offset variable.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/posterior_samples.html","id":"references","dir":"Reference","previous_headings":"","what":"References","title":"Draw samples from the posterior distribution of an estimated model — posterior_samples","text":"Wood, S.N., (2020). Simplified integrated nested Laplace approximation. Biometrika 107, 223–230. doi:10.1093/biomet/asz044","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/posterior_samples.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Draw samples from the posterior distribution of an estimated model — posterior_samples","text":"Gavin L. Simpson","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/predicted_samples.html","id":null,"dir":"Reference","previous_headings":"","what":"Draw new response values from the conditional distribution of the response — predicted_samples","title":"Draw new response values from the conditional distribution of the response — predicted_samples","text":"Predicted values response (new response data) drawn fitted model, created via simulate() (e.g. simulate.gam()) returned tidy, long, format. predicted values include uncertainty estimated model; simply draws conditional distribution response.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/predicted_samples.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Draw new response values from the conditional distribution of the response — predicted_samples","text":"","code":"predicted_samples(model, ...)  # S3 method for class 'gam' predicted_samples(   model,   n = 1,   data = newdata,   seed = NULL,   weights = NULL,   ...,   newdata = NULL )"},{"path":"https://gavinsimpson.github.io/gratia/reference/predicted_samples.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Draw new response values from the conditional distribution of the response — predicted_samples","text":"model fitted model supported types ... arguments passed methods. fitted_samples(), passed mgcv::predict.gam(). posterior_samples() passed fitted_samples(). predicted_samples() passed relevant simulate() method. n numeric; number posterior samples return. data data frame; new observations posterior draws model evaluated. supplied, data used fit model used data, available model. seed numeric; random seed simulations. weights numeric; vector prior weights. data null defaults object[[\"prior.weights\"]], otherwise vector ones. newdata Deprecated: use data instead.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/predicted_samples.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Draw new response values from the conditional distribution of the response — predicted_samples","text":"tibble (data frame) 3 columns containing posterior predicted values long format. columns row (integer) row data posterior draw relates , draw (integer) index, range 1:n, indicating draw row relates , response (numeric) predicted response indicated row data.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/predicted_samples.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Draw new response values from the conditional distribution of the response — predicted_samples","text":"Gavin L. Simpson","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/predicted_samples.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Draw new response values from the conditional distribution of the response — predicted_samples","text":"","code":"load_mgcv() dat <- data_sim(\"eg1\", n = 1000, dist = \"normal\", scale = 2, seed = 2) m <- gam(y ~ s(x0) + s(x1) + s(x2) + s(x3), data = dat, method = \"REML\")  predicted_samples(m, n = 5, seed = 42) #> # A tibble: 5,000 x 3 #>     .row .draw .response #>    <int> <int>     <dbl> #>  1     1     1      8.93 #>  2     2     1      4.23 #>  3     3     1      7.71 #>  4     4     1      8.51 #>  5     5     1     10.1  #>  6     6     1      8.20 #>  7     7     1      8.95 #>  8     8     1      7.20 #>  9     9     1     18.1  #> 10    10     1     12.7  #> # i 4,990 more rows  ## Can pass arguments to predict.gam() newd <- data.frame(   x0 = runif(10), x1 = runif(10), x2 = runif(10),   x3 = runif(10) )  ## Exclude s(x2) predicted_samples(m, n = 5, newd, exclude = \"s(x2)\", seed = 25) #> # A tibble: 50 x 3 #>     .row .draw .response #>    <int> <int>     <dbl> #>  1     1     1      9.42 #>  2     2     1      6.97 #>  3     3     1      8.10 #>  4     4     1      9.95 #>  5     5     1      6.75 #>  6     6     1     10.3  #>  7     7     1     10.8  #>  8     8     1     10.5  #>  9     9     1      8.43 #> 10    10     1     12.2  #> # i 40 more rows  ## Exclude s(x1) predicted_samples(m, n = 5, newd, exclude = \"s(x1)\", seed = 25) #> # A tibble: 50 x 3 #>     .row .draw .response #>    <int> <int>     <dbl> #>  1     1     1      6.05 #>  2     2     1      5.28 #>  3     3     1      5.96 #>  4     4     1     13.7  #>  5     5     1      4.36 #>  6     6     1      5.11 #>  7     7     1     12.5  #>  8     8     1      5.66 #>  9     9     1     12.6  #> 10    10     1      8.38 #> # i 40 more rows  ## Select which terms --- result should be the same as previous ## but note that we have to include any parametric terms, including the ## constant term predicted_samples(m,   n = 5, newd, seed = 25,   terms = c(\"Intercept\", \"s(x0)\", \"s(x2)\", \"s(x3)\") ) #> # A tibble: 50 x 3 #>     .row .draw .response #>    <int> <int>     <dbl> #>  1     1     1    -1.94  #>  2     2     1    -2.71  #>  3     3     1    -2.03  #>  4     4     1     5.73  #>  5     5     1    -3.63  #>  6     6     1    -2.87  #>  7     7     1     4.48  #>  8     8     1    -2.33  #>  9     9     1     4.65  #> 10    10     1     0.395 #> # i 40 more rows"},{"path":"https://gavinsimpson.github.io/gratia/reference/qq_plot.html","id":null,"dir":"Reference","previous_headings":"","what":"Quantile-quantile plot of model residuals — qq_plot","title":"Quantile-quantile plot of model residuals — qq_plot","text":"Quantile-quantile plots (QQ-plots) GAMs using reference quantiles Augustin et al (2012).","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/qq_plot.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Quantile-quantile plot of model residuals — qq_plot","text":"","code":"qq_plot(model, ...)  # Default S3 method qq_plot(model, ...)  # S3 method for class 'gam' qq_plot(   model,   method = c(\"uniform\", \"simulate\", \"normal\", \"direct\"),   type = c(\"deviance\", \"response\", \"pearson\"),   n_uniform = 10,   n_simulate = 50,   seed = NULL,   level = 0.9,   ylab = NULL,   xlab = NULL,   title = NULL,   subtitle = NULL,   caption = NULL,   ci_col = \"black\",   ci_alpha = 0.2,   point_col = \"black\",   point_alpha = 1,   line_col = \"red\",   ... )  # S3 method for class 'glm' qq_plot(model, ...)  # S3 method for class 'lm' qq_plot(model, ...)"},{"path":"https://gavinsimpson.github.io/gratia/reference/qq_plot.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Quantile-quantile plot of model residuals — qq_plot","text":"model fitted model. Currently models inheriting class \"gam\", well classes \"glm\" \"lm\" calls stats::glm stats::lm supported. ... arguments passed ot methods. method character; method used generate theoretical quantiles. default \"uniform\", generates reference quantiles using random draws uniform distribution inverse cummulative distribution function (CDF) fitted values. reference quantiles averaged n_uniform draws. \"simulate\" generates reference quantiles simulating new response data model observed values covariates, residualised generate reference quantiles, using n_simulate simulated data sets. \"normal\" generates reference quantiles using standard normal distribution. \"uniform\" computationally efficient, \"simulate\" allows reference bands drawn QQ-plot. \"normal\" avoided used fall back random number generator (\"simulate\") inverse CDF available family used model fitting (`\"uniform\"“). Note method = \"direct\" deprecated favour method = \"uniform\". type character; type residuals use. \"deviance\", \"response\", \"pearson\" residuals allowed. n_uniform numeric; number times randomize uniform quantiles direct computation method (method = \"uniform\"). n_simulate numeric; number data sets simulate estimated model using simulation method (method = \"simulate\"). seed numeric; random number seed use method = \"simulate\" method = \"uniform\". level numeric; coverage level reference intervals. Must strictly 0 < level < 1. used method = \"simulate\". ylab character expression; label y axis. supplied, suitable label generated. xlab character expression; label y axis. supplied, suitable label generated. title character expression; title plot. See ggplot2::labs(). May vector, one per penalty. subtitle character expression; subtitle plot. See ggplot2::labs(). May vector, one per penalty. caption character expression; plot caption. See ggplot2::labs(). May vector, one per penalty. ci_col fill colour reference interval method = \"simulate\". ci_alpha alpha transparency reference interval method = \"simulate\". point_col colour points QQ plot. point_alpha alpha transparency points QQ plot. line_col colour used draw reference line.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/qq_plot.html","id":"note","dir":"Reference","previous_headings":"","what":"Note","title":"Quantile-quantile plot of model residuals — qq_plot","text":"wording used mgcv::qq.gam() uses direct reference simulated residuals method (method = \"simulated\"). avoid confusion, method = \"direct\" deprecated favour method = \"uniform\".","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/qq_plot.html","id":"references","dir":"Reference","previous_headings":"","what":"References","title":"Quantile-quantile plot of model residuals — qq_plot","text":"underlying methodology used method \"simulate\" \"uniform\" described Augustin et al (2012): Augustin, N.H., Sauleau, E.-., Wood, S.N., (2012) quantile quantile plots generalized linear models. Computational Statatistics Data Analysis 56, 2404-2409 doi:10.1016/j.csda.2012.01.026 .","code":""},{"path":[]},{"path":"https://gavinsimpson.github.io/gratia/reference/qq_plot.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Quantile-quantile plot of model residuals — qq_plot","text":"","code":"load_mgcv() ## simulate binomial data... dat <- data_sim(\"eg1\", n = 200, dist = \"binary\", scale = .33, seed = 0) p <- binomial()$linkinv(dat$f) # binomial p n <- sample(c(1, 3), 200, replace = TRUE) # binomial n dat <- transform(dat, y = rbinom(n, n, p), n = n) m <- gam(y / n ~ s(x0) + s(x1) + s(x2) + s(x3),   family = binomial, data = dat, weights = n,   method = \"REML\" )  ## Q-Q plot; default using direct randomization of uniform quantiles qq_plot(m)   ## Alternatively use simulate new data from the model, which ## allows construction of reference intervals for the Q-Q plot qq_plot(m,   method = \"simulate\",   seed = 42,   point_col = \"steelblue\",   point_alpha = 0.4 )   ## ... or use the usual normality assumption qq_plot(m, method = \"normal\")"},{"path":"https://gavinsimpson.github.io/gratia/reference/ref_level.html","id":null,"dir":"Reference","previous_headings":"","what":"Return the reference or specific level of a factor — ref_level","title":"Return the reference or specific level of a factor — ref_level","text":"Extracts reference specific level supplied factor, returning factor levels one supplied.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/ref_level.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Return the reference or specific level of a factor — ref_level","text":"","code":"ref_level(fct)  level(fct, level)"},{"path":"https://gavinsimpson.github.io/gratia/reference/ref_level.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Return the reference or specific level of a factor — ref_level","text":"fct factor; factor reference specific level extracted. level character; specific level extract case level().","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/ref_level.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Return the reference or specific level of a factor — ref_level","text":"length 1 factor levels supplied factor fct.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/ref_level.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Return the reference or specific level of a factor — ref_level","text":"","code":"f <- factor(sample(letters[1:5], 100, replace = TRUE))  # the reference level ref_level(f) #> [1] a #> Levels: a b c d e  # a specific level level(f, level = \"b\") #> [1] b #> Levels: a b c d e  # note that the levels will always match the input factor identical(levels(f), levels(ref_level(f))) #> [1] TRUE identical(levels(f), levels(level(f, \"c\"))) #> [1] TRUE"},{"path":"https://gavinsimpson.github.io/gratia/reference/ref_sims.html","id":null,"dir":"Reference","previous_headings":"","what":"Reference simulation data — ref_sims","title":"Reference simulation data — ref_sims","text":"set reference objects testing data_sim().","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/ref_sims.html","id":"format","dir":"Reference","previous_headings":"","what":"Format","title":"Reference simulation data — ref_sims","text":"named list simulated data sets created data_sim().","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/reorder_fs_smooth_terms.html","id":null,"dir":"Reference","previous_headings":"","what":"Reorder random factor smooth terms to place factor last — reorder_fs_smooth_terms","title":"Reorder random factor smooth terms to place factor last — reorder_fs_smooth_terms","text":"Reorder random factor smooth terms place factor last","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/reorder_fs_smooth_terms.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Reorder random factor smooth terms to place factor last — reorder_fs_smooth_terms","text":"","code":"reorder_fs_smooth_terms(smooth)"},{"path":"https://gavinsimpson.github.io/gratia/reference/reorder_fs_smooth_terms.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Reorder random factor smooth terms to place factor last — reorder_fs_smooth_terms","text":"smooth mgcv smooth object","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/reorder_tensor_smooth_terms.html","id":null,"dir":"Reference","previous_headings":"","what":"Reorder tensor product terms for nicer plotting — reorder_tensor_smooth_terms","title":"Reorder tensor product terms for nicer plotting — reorder_tensor_smooth_terms","text":"tensor product smooth 3 terms contains 2d marginal smooth, get nicer output smooth_estimates() hence nicer plot draw.smooth_estimates() method reorder terms smooth vary terms 2d marginal first, terms vary slowly generate data evaluate smooth . results automatically generated data focuses (first one) 2d marginal smooth, end result smooth_estimates() shows 2d smooth changes terms involved smooth.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/reorder_tensor_smooth_terms.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Reorder tensor product terms for nicer plotting — reorder_tensor_smooth_terms","text":"","code":"reorder_tensor_smooth_terms(smooth)"},{"path":"https://gavinsimpson.github.io/gratia/reference/reorder_tensor_smooth_terms.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Reorder tensor product terms for nicer plotting — reorder_tensor_smooth_terms","text":"smooth mgcv smooth object","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/rep_first_factor_value.html","id":null,"dir":"Reference","previous_headings":"","what":"Repeat the first level of a factor n times — rep_first_factor_value","title":"Repeat the first level of a factor n times — rep_first_factor_value","text":"Function repeat first level factor n times return vector factor original levels intact","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/rep_first_factor_value.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Repeat the first level of a factor n times — rep_first_factor_value","text":"","code":"rep_first_factor_value(f, n)"},{"path":"https://gavinsimpson.github.io/gratia/reference/rep_first_factor_value.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Repeat the first level of a factor n times — rep_first_factor_value","text":"f factor n numeric; number times repeat first level f","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/rep_first_factor_value.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Repeat the first level of a factor n times — rep_first_factor_value","text":"factor length n levels f, whose elements first level f.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/residuals_hist_plot.html","id":null,"dir":"Reference","previous_headings":"","what":"Histogram of model residuals — residuals_hist_plot","title":"Histogram of model residuals — residuals_hist_plot","text":"Histogram model residuals","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/residuals_hist_plot.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Histogram of model residuals — residuals_hist_plot","text":"","code":"residuals_hist_plot(   model,   type = c(\"deviance\", \"pearson\", \"response\"),   n_bins = c(\"sturges\", \"scott\", \"fd\"),   ylab = NULL,   xlab = NULL,   title = NULL,   subtitle = NULL,   caption = NULL )"},{"path":"https://gavinsimpson.github.io/gratia/reference/residuals_hist_plot.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Histogram of model residuals — residuals_hist_plot","text":"model fitted model. Currently class \"gam\". type character; type residuals use. \"deviance\", \"response\", \"pearson\" residuals allowed. n_bins character numeric; either number bins string indicating calculate number bins. ylab character expression; label y axis. supplied, suitable label generated. xlab character expression; label y axis. supplied, suitable label generated. title character expression; title plot. See ggplot2::labs(). subtitle character expression; subtitle plot. See ggplot2::labs(). caption character expression; plot caption. See ggplot2::labs().","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/residuals_linpred_plot.html","id":null,"dir":"Reference","previous_headings":"","what":"Plot of residuals versus linear predictor values — residuals_linpred_plot","title":"Plot of residuals versus linear predictor values — residuals_linpred_plot","text":"Plot residuals versus linear predictor values","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/residuals_linpred_plot.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Plot of residuals versus linear predictor values — residuals_linpred_plot","text":"","code":"residuals_linpred_plot(   model,   type = c(\"deviance\", \"pearson\", \"response\"),   ylab = NULL,   xlab = NULL,   title = NULL,   subtitle = NULL,   caption = NULL,   point_col = \"black\",   point_alpha = 1,   line_col = \"red\" )"},{"path":"https://gavinsimpson.github.io/gratia/reference/residuals_linpred_plot.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Plot of residuals versus linear predictor values — residuals_linpred_plot","text":"model fitted model. Currently class \"gam\". type character; type residuals use. \"deviance\", \"response\", \"pearson\" residuals allowed. ylab character expression; label y axis. supplied, suitable label generated. xlab character expression; label y axis. supplied, suitable label generated. title character expression; title plot. See ggplot2::labs(). subtitle character expression; subtitle plot. See ggplot2::labs(). caption character expression; plot caption. See ggplot2::labs(). point_col colour used draw points plots. See graphics::par() section Color Specification. passed individual plotting functions, therefore affects points plots. point_alpha numeric; alpha transparency points plots. line_col colour specification 1:1 line.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/response_derivatives.html","id":null,"dir":"Reference","previous_headings":"","what":"Derivatives on the response scale from an estimated GAM — response_derivatives","title":"Derivatives on the response scale from an estimated GAM — response_derivatives","text":"Derivatives response scale estimated GAM","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/response_derivatives.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Derivatives on the response scale from an estimated GAM — response_derivatives","text":"","code":"response_derivatives(object, ...)  # Default S3 method response_derivatives(object, ...)  # S3 method for class 'gamm' response_derivatives(object, ...)  # S3 method for class 'gam' response_derivatives(   object,   focal = NULL,   data = NULL,   order = 1L,   type = c(\"forward\", \"backward\", \"central\"),   scale = c(\"response\", \"linear_predictor\"),   method = c(\"gaussian\", \"mh\", \"inla\", \"user\"),   n = 100,   eps = 1e-07,   n_sim = 10000,   level = 0.95,   seed = NULL,   mvn_method = c(\"mvnfast\", \"mgcv\"),   ... )"},{"path":"https://gavinsimpson.github.io/gratia/reference/response_derivatives.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Derivatives on the response scale from an estimated GAM — response_derivatives","text":"object R object compute derivatives . ... arguments passed methods fitted_samples() focal character; name focal variable. response derivative response respect variable returned. variables involved model held constant values. can missing supplying data, case, focal variable identified one variable constant. data data frame containing values model covariates evaluate first derivatives smooths. supplied, one variable must held constant value. order numeric; order derivative. type character; type finite difference used. One \"forward\", \"backward\", \"central\". scale character; derivative estimated response linear predictor (link) scale? One \"response\" (default), \"linear predictor\". method character; method used draw samples posterior distribution. \"gaussian\" uses Gaussian (Laplace) approximation posterior. \"mh\" uses Metropolis Hastings sample alternates t proposals proposals based shrunken version posterior covariance matrix. \"inla\" uses variant Integrated Nested Laplace Approximation due Wood (2019), (currently implemented). \"user\" allows user-supplied posterior draws (currently implemented). n numeric; number points evaluate derivative (data supplied). eps numeric; finite difference. n_sim integer; number simulations used computing simultaneous intervals. level numeric; 0 < level < 1; coverage level credible interval. default 0.95 95% interval. seed numeric; random seed simulations. mvn_method character; one \"mvnfast\" \"mgcv\". default uses mvnfast::rmvn(), can considerably faster generate large numbers MVN random values mgcv::rmvn(), might work marginal fits, covariance matrix close singular.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/response_derivatives.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Derivatives on the response scale from an estimated GAM — response_derivatives","text":"tibble, currently following variables: .row: integer, indexing row data row output represents .focal: name variable partial derivative evaluated, .derivative: estimated partial derivative, .lower_ci: lower bound confidence simultaneous interval, .upper_ci: upper bound confidence simultaneous interval, additional columns containing covariate values derivative evaluated.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/response_derivatives.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Derivatives on the response scale from an estimated GAM — response_derivatives","text":"Gavin L. Simpson","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/response_derivatives.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Derivatives on the response scale from an estimated GAM — response_derivatives","text":"","code":"library(\"ggplot2\") library(\"patchwork\") load_mgcv() df <- data_sim(\"eg1\", dist = \"negbin\", scale = 0.25, seed = 42)  # fit the GAM (note: for execution time reasons using bam()) m <- bam(y ~ s(x0) + s(x1) + s(x2) + s(x3),   data = df, family = nb(), method = \"fREML\" )  # data slice through data along x2 - all other covariates will be set to # typical values (value closest to median) ds <- data_slice(m, x2 = evenly(x2, n = 100))  # fitted values along x2 fv <- fitted_values(m, data = ds)  # response derivatives - ideally n_sim = >10000 y_d <- response_derivatives(m,   data = ds, type = \"central\", focal = \"x2\",   eps = 0.01, seed = 21, n_sim = 1000 )  # draw fitted values along x2 p1 <- fv |>   ggplot(aes(x = x2, y = .fitted)) +   geom_ribbon(aes(ymin = .lower_ci, ymax = .upper_ci, y = NULL),     alpha = 0.2   ) +   geom_line() +   labs(     title = \"Estimated count as a function of x2\",     y = \"Estimated count\"   )  # draw response derivatives p2 <- y_d |>   ggplot(aes(x = x2, y = .derivative)) +   geom_ribbon(aes(ymin = .lower_ci, ymax = .upper_ci), alpha = 0.2) +   geom_line() +   labs(     title = \"Estimated 1st derivative of estimated count\",     y = \"First derivative\"   )  # draw both panels p1 + p2 + plot_layout(nrow = 2)"},{"path":"https://gavinsimpson.github.io/gratia/reference/rootogram.html","id":null,"dir":"Reference","previous_headings":"","what":"Rootograms to assess goodness of model fit — rootogram","title":"Rootograms to assess goodness of model fit — rootogram","text":"rootogram model diagnostic tool assesses goodness fit statistical model. observed values response compared expected fitted model. discrete, count responses, frequency count (0, 1, 2, etc) observed data expected conditional distribution response implied model compared. continuous variables, observed expected frequencies obtained grouping data bins. rootogram drawn using ggplot2::ggplot() graphics. design closely follows Kleiber & Zeileis (2016).","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/rootogram.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Rootograms to assess goodness of model fit — rootogram","text":"","code":"rootogram(object, ...)  # S3 method for class 'gam' rootogram(object, max_count = NULL, breaks = \"Sturges\", ...)"},{"path":"https://gavinsimpson.github.io/gratia/reference/rootogram.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Rootograms to assess goodness of model fit — rootogram","text":"object R object ... arguments passed methods max_count integer; largest count consider breaks continuous responses, group response. Can anything acceptable breaks argument graphics::hist.default()","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/rootogram.html","id":"references","dir":"Reference","previous_headings":"","what":"References","title":"Rootograms to assess goodness of model fit — rootogram","text":"Kleiber, C., Zeileis, ., (2016) Visualizing Count Data Regressions Using Rootograms. . Stat. 70, 296–303. doi:10.1080/00031305.2016.1173590","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/rootogram.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Rootograms to assess goodness of model fit — rootogram","text":"","code":"load_mgcv() df <- data_sim(\"eg1\", n = 1000, dist = \"poisson\", scale = 0.1, seed = 6)  # A poisson example m <- gam(y ~ s(x0, bs = \"cr\") + s(x1, bs = \"cr\") + s(x2, bs = \"cr\") +   s(x3, bs = \"cr\"), family = poisson(), data = df, method = \"REML\") rg <- rootogram(m) rg #> # A tibble: 21 x 3 #>     .bin .observed   .fitted #>    <dbl>     <int>     <dbl> #>  1     0       113 116.640   #>  2     1       236 227.869   #>  3     2       230 239.168   #>  4     3       200 181.679   #>  5     4        94 113.432   #>  6     5        68  62.4881  #>  7     6        27  31.6795  #>  8     7        22  15.1323  #>  9     8         4   6.88637 #> 10     9         3   2.99628 #> # i 11 more rows draw(rg) # plot the rootogram   # A Gaussian example df <- data_sim(\"eg1\", dist = \"normal\", seed = 2) m <- gam(y ~ s(x0) + s(x1) + s(x2) + s(x3), data = df, method = \"REML\") draw(rootogram(m, breaks = \"FD\"), type = \"suspended\")"},{"path":"https://gavinsimpson.github.io/gratia/reference/seq_min_max_eps.html","id":null,"dir":"Reference","previous_headings":"","what":"Create a sequence of evenly-spaced values adjusted to accommodate a small adjustment — seq_min_max_eps","title":"Create a sequence of evenly-spaced values adjusted to accommodate a small adjustment — seq_min_max_eps","text":"Creates sequence n evenly-spaced values range min(x) – max(x), minimum maximum adjusted always contained within range x x may shifted forwards backwards amount related eps. particularly useful computing derivatives via finite differences without adjustment may predicting values outside range data hence conmstraints penalty.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/seq_min_max_eps.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Create a sequence of evenly-spaced values adjusted to accommodate a small adjustment — seq_min_max_eps","text":"","code":"seq_min_max_eps(x, n, order, type = c(\"forward\", \"backward\", \"central\"), eps)"},{"path":"https://gavinsimpson.github.io/gratia/reference/seq_min_max_eps.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Create a sequence of evenly-spaced values adjusted to accommodate a small adjustment — seq_min_max_eps","text":"x numeric; vector evenly-spaced values returned n numeric; number evenly-spaced values return order integer; order derivative. Either 1 2 first second order derivatives type character; type finite difference used. One \"forward\", \"backward\", \"central\" eps numeric; finite difference","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/seq_min_max_eps.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Create a sequence of evenly-spaced values adjusted to accommodate a small adjustment — seq_min_max_eps","text":"numeric vector length n.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/shift_values.html","id":null,"dir":"Reference","previous_headings":"","what":"Shift numeric values in a data frame by an amount eps — shift_values","title":"Shift numeric values in a data frame by an amount eps — shift_values","text":"Shift numeric values data frame amount eps","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/shift_values.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Shift numeric values in a data frame by an amount eps — shift_values","text":"","code":"shift_values(df, h, i, FUN = `+`, focal = NULL)"},{"path":"https://gavinsimpson.github.io/gratia/reference/shift_values.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Shift numeric values in a data frame by an amount eps — shift_values","text":"df data frame tibble. h numeric; amount shift values df . logical; vector indexing columns df included shift. FUN function; function applut shift. Typically + -. focal character; focal variable computing partial derivatives. allows shifting focal variable eps.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/simulate.html","id":null,"dir":"Reference","previous_headings":"","what":"Simulate from the posterior distribution of a GAM — simulate.gam","title":"Simulate from the posterior distribution of a GAM — simulate.gam","text":"Simulations posterior distribution fitted GAM model involve computing predicted values observation data simulated data required, generating random draws probability distribution used fitting model.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/simulate.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Simulate from the posterior distribution of a GAM — simulate.gam","text":"","code":"# S3 method for class 'gam' simulate(   object,   nsim = 1,   seed = NULL,   data = newdata,   weights = NULL,   ...,   newdata = NULL )  # S3 method for class 'gamm' simulate(   object,   nsim = 1,   seed = NULL,   data = newdata,   weights = NULL,   ...,   newdata = NULL )  # S3 method for class 'scam' simulate(   object,   nsim = 1,   seed = NULL,   data = newdata,   weights = NULL,   ...,   newdata = NULL )"},{"path":"https://gavinsimpson.github.io/gratia/reference/simulate.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Simulate from the posterior distribution of a GAM — simulate.gam","text":"object fitted GAM, typically result call mgcv::gam` mgcv::gamm(). nsim numeric; number posterior simulations return. seed numeric; random seed simulations. data data frame; new observations posterior draws model evaluated. supplied, data used fit model used newdata, available object. weights numeric; vector prior weights. newdata null defaults object[[\"prior.weights\"]], otherwise vector ones. ... arguments passed methods. simulate.gam() simulate.scam() pass ... predict.gam(). can pass additional arguments terms, exclude, select model terms included predictions. may useful, example, excluding effects random effect terms. newdata Deprecated. Use data instead.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/simulate.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Simulate from the posterior distribution of a GAM — simulate.gam","text":"(Currently) matrix nsim columns.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/simulate.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Simulate from the posterior distribution of a GAM — simulate.gam","text":"simulate.gam() function, family component fitted model must contain, updateable contain, required random number generator. See mgcv::fix.family.rd().","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/simulate.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Simulate from the posterior distribution of a GAM — simulate.gam","text":"Gavin L. Simpson","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/simulate.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Simulate from the posterior distribution of a GAM — simulate.gam","text":"","code":"load_mgcv() dat <- data_sim(\"eg1\", n = 400, dist = \"normal\", scale = 2, seed = 2) m1 <- gam(y ~ s(x0) + s(x1) + s(x2) + s(x3), data = dat, method = \"REML\")  sims <- simulate(m1, nsim = 5, seed = 42) head(sims) #>           [,1]      [,2]      [,3]      [,4]      [,5] #> [1,] 11.445470 11.374304 10.098681  7.264881  8.796630 #> [2,]  6.510912  5.909584  9.057362  7.698084 11.444781 #> [3,]  3.837995  3.230610  3.550240  3.759380  4.774581 #> [4,] 12.361830 11.209226 10.714215 11.861957 10.746417 #> [5,] 14.851461 12.911440 11.356984 15.783913 15.106270 #> [6,]  5.921276  4.158963  5.520856  7.973614  9.654888"},{"path":"https://gavinsimpson.github.io/gratia/reference/smallAges.html","id":null,"dir":"Reference","previous_headings":"","what":"Lead-210 age-depth measurements for Small Water — smallAges","title":"Lead-210 age-depth measurements for Small Water — smallAges","text":"dataset containing lead-210 based age depth measurements SMALL1 core Small Water.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/smallAges.html","id":"format","dir":"Reference","previous_headings":"","what":"Format","title":"Lead-210 age-depth measurements for Small Water — smallAges","text":"data frame 12 rows 7 variables.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/smallAges.html","id":"source","dir":"Reference","previous_headings":"","what":"Source","title":"Lead-210 age-depth measurements for Small Water — smallAges","text":"Simpson, G.L. (Unpublished data).","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/smallAges.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Lead-210 age-depth measurements for Small Water — smallAges","text":"variables follows: Depth Drymass Date Age Error SedAccRate SedPerCentChange","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/smooth_coef_indices.html","id":null,"dir":"Reference","previous_headings":"","what":"Indices of the parametric terms for a particular smooth — smooth_coef_indices","title":"Indices of the parametric terms for a particular smooth — smooth_coef_indices","text":"Returns vector indices parametric terms represent supplied smooth. Useful extracting model coefficients columns covariance matrix.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/smooth_coef_indices.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Indices of the parametric terms for a particular smooth — smooth_coef_indices","text":"","code":"smooth_coef_indices(smooth)"},{"path":"https://gavinsimpson.github.io/gratia/reference/smooth_coef_indices.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Indices of the parametric terms for a particular smooth — smooth_coef_indices","text":"smooth object inherits class mgcv.smooth","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/smooth_coef_indices.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Indices of the parametric terms for a particular smooth — smooth_coef_indices","text":"numeric vector indices.","code":""},{"path":[]},{"path":"https://gavinsimpson.github.io/gratia/reference/smooth_coef_indices.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Indices of the parametric terms for a particular smooth — smooth_coef_indices","text":"Gavin L. Simpson","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/smooth_coefs.html","id":null,"dir":"Reference","previous_headings":"","what":"Coefficients for a particular smooth — smooth_coefs","title":"Coefficients for a particular smooth — smooth_coefs","text":"Returns vector model coefficients parametric terms represent supplied smooth.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/smooth_coefs.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Coefficients for a particular smooth — smooth_coefs","text":"","code":"smooth_coefs(object, ...)  # S3 method for class 'gam' smooth_coefs(object, select, term = deprecated(), ...)  # S3 method for class 'bam' smooth_coefs(object, select, term = deprecated(), ...)  # S3 method for class 'gamm' smooth_coefs(object, select, term = deprecated(), ...)  # S3 method for class 'gamm4' smooth_coefs(object, select, term = deprecated(), ...)  # S3 method for class 'list' smooth_coefs(object, select, term = deprecated(), ...)  # S3 method for class 'mgcv.smooth' smooth_coefs(object, model, ...)  # S3 method for class 'scam' smooth_coefs(object, select, term = deprecated(), ...)"},{"path":"https://gavinsimpson.github.io/gratia/reference/smooth_coefs.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Coefficients for a particular smooth — smooth_coefs","text":"object fitted GAM(M) object, , \"mgcv.smooth\" method, object inherits class mgcv.smooth. ... arguments passed methods. select character; label smooth whose coefficients returned. term Use select instead. model fitted GAM(M) object.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/smooth_coefs.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Coefficients for a particular smooth — smooth_coefs","text":"numeric vector model coefficients.","code":""},{"path":[]},{"path":"https://gavinsimpson.github.io/gratia/reference/smooth_coefs.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Coefficients for a particular smooth — smooth_coefs","text":"Gavin L. Simpson","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/smooth_coefs.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Coefficients for a particular smooth — smooth_coefs","text":"","code":"load_mgcv() df <- data_sim(\"eg1\", seed = 2) m <- gam(y ~ s(x0) + s(x1) + s(x2) + s(x3), data = df, method = \"REML\")  ## IGNORE_RDIFF_BEGIN smooth_coefs(m, select = \"s(x2)\") #>   s(x2).1   s(x2).2   s(x2).3   s(x2).4   s(x2).5   s(x2).6   s(x2).7   s(x2).8  #> -6.533373  9.694277  2.194078 -1.967280 -2.374874  1.207638 -1.572586  9.269744  #>   s(x2).9  #>  5.622738  ## IGNORE_RDIFF_END"},{"path":"https://gavinsimpson.github.io/gratia/reference/smooth_data.html","id":null,"dir":"Reference","previous_headings":"","what":"Generate regular data over the covariates of a smooth — smooth_data","title":"Generate regular data over the covariates of a smooth — smooth_data","text":"Generate regular data covariates smooth","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/smooth_data.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Generate regular data over the covariates of a smooth — smooth_data","text":"","code":"smooth_data(   model,   id,   n = 100,   n_2d = NULL,   n_3d = NULL,   n_4d = NULL,   offset = NULL,   include_all = FALSE,   var_order = NULL )"},{"path":"https://gavinsimpson.github.io/gratia/reference/smooth_data.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Generate regular data over the covariates of a smooth — smooth_data","text":"model fitted model id number ID smooth within model process. n numeric; number new observations generate. n_2d numeric; number new observations generate second dimension 2D smooth. Currently ignored. n_3d numeric; number new observations generate third dimension 3D smooth. n_4d numeric; number new observations generate dimensions higher 2 (!) kD smooth (k >= 4). example, smooth 4D smooth, dimensions 3 4 get n_4d new observations. offset numeric; value model offset use. include_all logical; include covariates involved smooth? FALSE, covariates involved smooth included returned data frame. TRUE, representative value included covariates model actually used smooth. can useful want pass returned data frame mgcv::PredictMat(). var_order character; order terms smooth processed. useful tensor products least one 2d marginal smooth.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/smooth_data.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Generate regular data over the covariates of a smooth — smooth_data","text":"","code":"load_mgcv() df <- data_sim(\"eg1\", seed = 42) m <- bam(y ~ s(x0) + s(x1) + s(x2) + s(x3), data = df)  # generate data over range of x1 for smooth s(x1) smooth_data(m, id = 2) #> # A tibble: 100 x 1 #>           x1 #>        <dbl> #>  1 0.0004050 #>  2 0.01046   #>  3 0.02052   #>  4 0.03057   #>  5 0.04063   #>  6 0.05069   #>  7 0.06074   #>  8 0.07080   #>  9 0.08086   #> 10 0.09091   #> # i 90 more rows  # generate data over range of x1 for smooth s(x1), with typical value for # other covariates in the model smooth_data(m, id = 2, include_all = TRUE) #> # A tibble: 100 x 4 #>           x1     x0     x2     x3 #>        <dbl>  <dbl>  <dbl>  <dbl> #>  1 0.0004050 0.4883 0.4708 0.4879 #>  2 0.01046   0.4883 0.4708 0.4879 #>  3 0.02052   0.4883 0.4708 0.4879 #>  4 0.03057   0.4883 0.4708 0.4879 #>  5 0.04063   0.4883 0.4708 0.4879 #>  6 0.05069   0.4883 0.4708 0.4879 #>  7 0.06074   0.4883 0.4708 0.4879 #>  8 0.07080   0.4883 0.4708 0.4879 #>  9 0.08086   0.4883 0.4708 0.4879 #> 10 0.09091   0.4883 0.4708 0.4879 #> # i 90 more rows"},{"path":"https://gavinsimpson.github.io/gratia/reference/smooth_dim.html","id":null,"dir":"Reference","previous_headings":"","what":"Dimension of a smooth — smooth_dim","title":"Dimension of a smooth — smooth_dim","text":"Extracts dimension estimated smooth.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/smooth_dim.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Dimension of a smooth — smooth_dim","text":"","code":"smooth_dim(object)  # S3 method for class 'gam' smooth_dim(object)  # S3 method for class 'gamm' smooth_dim(object)  # S3 method for class 'mgcv.smooth' smooth_dim(object)"},{"path":"https://gavinsimpson.github.io/gratia/reference/smooth_dim.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Dimension of a smooth — smooth_dim","text":"object R object. See Details list supported objects.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/smooth_dim.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Dimension of a smooth — smooth_dim","text":"numeric vector dimensions smooth.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/smooth_dim.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Dimension of a smooth — smooth_dim","text":"generic function methods objects class \"gam\", \"gamm\", \"mgcv.smooth\".","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/smooth_dim.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Dimension of a smooth — smooth_dim","text":"Gavin L. Simpson","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/smooth_estimates.html","id":null,"dir":"Reference","previous_headings":"","what":"Evaluate smooths at covariate values — smooth_estimates","title":"Evaluate smooths at covariate values — smooth_estimates","text":"Evaluate smooth grid evenly spaced value range covariate associated smooth. Alternatively, set points smooth evaluated can supplied. smooth_estimates() new implementation evaluate_smooth(), replaces function, removed package.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/smooth_estimates.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Evaluate smooths at covariate values — smooth_estimates","text":"","code":"smooth_estimates(object, ...)  # S3 method for class 'gam' smooth_estimates(   object,   select = NULL,   smooth = deprecated(),   n = 100,   n_3d = 16,   n_4d = 4,   data = NULL,   unconditional = FALSE,   overall_uncertainty = TRUE,   dist = NULL,   unnest = TRUE,   partial_match = FALSE,   ... )"},{"path":"https://gavinsimpson.github.io/gratia/reference/smooth_estimates.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Evaluate smooths at covariate values — smooth_estimates","text":"object object class \"gam\" \"gamm\". ... arguments passed methods. select character; select smooth's posterior draw . default (NULL) means posteriors smooths model wil sampled . supplied, character vector requested terms. smooth Use select instead. n numeric; number points range covariate evaluate smooth. n_3d, n_4d numeric; number points range last covariate 3D 4D smooth. default NULL achieves standard behaviour using n points range covariate, resulting n^d evaluation points, d dimension smooth. d > 2 can result many evaluation points slow performance. smooths d > 4, value n_4d used dimensions > 4, unless NULL, case default behaviour (using n dimensions) observed. data data frame covariate values evaluate smooth. unconditional logical; confidence intervals include uncertainty due smoothness selection? TRUE, corrected Bayesian covariance matrix used. overall_uncertainty logical; uncertainty model constant term included standard error evaluate values smooth? dist numeric; greater 0, used determine location far data plotted plotting 2-D smooths. data scaled unit square deciding exclude, dist distance within unit square. See mgcv::exclude..far() details. unnest logical; unnest smooth objects? partial_match logical; case character select, select match partially smooths? partial_match = TRUE, select must single string, character vector length 1.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/smooth_estimates.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Evaluate smooths at covariate values — smooth_estimates","text":"data frame (tibble), class \"smooth_estimates\".","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/smooth_estimates.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Evaluate smooths at covariate values — smooth_estimates","text":"","code":"load_mgcv() dat <- data_sim(\"eg1\", n = 400, dist = \"normal\", scale = 2, seed = 2) m1 <- gam(y ~ s(x0) + s(x1) + s(x2) + s(x3), data = dat, method = \"REML\")  ## evaluate all smooths smooth_estimates(m1) #> # A tibble: 400 x 9 #>    .smooth .type .by   .estimate      .se         x0    x1    x2    x3 #>    <chr>   <chr> <chr>     <dbl>    <dbl>      <dbl> <dbl> <dbl> <dbl> #>  1 s(x0)   TPRS  NA    -0.966542 0.316118 0.00710904    NA    NA    NA #>  2 s(x0)   TPRS  NA    -0.925391 0.297170 0.0171157     NA    NA    NA #>  3 s(x0)   TPRS  NA    -0.884233 0.279256 0.0271224     NA    NA    NA #>  4 s(x0)   TPRS  NA    -0.843050 0.262594 0.0371291     NA    NA    NA #>  5 s(x0)   TPRS  NA    -0.801824 0.247376 0.0471358     NA    NA    NA #>  6 s(x0)   TPRS  NA    -0.760536 0.233728 0.0571425     NA    NA    NA #>  7 s(x0)   TPRS  NA    -0.719175 0.221701 0.0671492     NA    NA    NA #>  8 s(x0)   TPRS  NA    -0.677736 0.211261 0.0771559     NA    NA    NA #>  9 s(x0)   TPRS  NA    -0.636220 0.202303 0.0871626     NA    NA    NA #> 10 s(x0)   TPRS  NA    -0.594641 0.194685 0.0971693     NA    NA    NA #> # i 390 more rows  ## or selected smooths smooth_estimates(m1, select = c(\"s(x0)\", \"s(x1)\")) #> # A tibble: 200 x 7 #>    .smooth .type .by   .estimate      .se         x0    x1 #>    <chr>   <chr> <chr>     <dbl>    <dbl>      <dbl> <dbl> #>  1 s(x0)   TPRS  NA    -0.966542 0.316118 0.00710904    NA #>  2 s(x0)   TPRS  NA    -0.925391 0.297170 0.0171157     NA #>  3 s(x0)   TPRS  NA    -0.884233 0.279256 0.0271224     NA #>  4 s(x0)   TPRS  NA    -0.843050 0.262594 0.0371291     NA #>  5 s(x0)   TPRS  NA    -0.801824 0.247376 0.0471358     NA #>  6 s(x0)   TPRS  NA    -0.760536 0.233728 0.0571425     NA #>  7 s(x0)   TPRS  NA    -0.719175 0.221701 0.0671492     NA #>  8 s(x0)   TPRS  NA    -0.677736 0.211261 0.0771559     NA #>  9 s(x0)   TPRS  NA    -0.636220 0.202303 0.0871626     NA #> 10 s(x0)   TPRS  NA    -0.594641 0.194685 0.0971693     NA #> # i 190 more rows"},{"path":"https://gavinsimpson.github.io/gratia/reference/smooth_label.html","id":null,"dir":"Reference","previous_headings":"","what":"Extract the label for a smooth used by 'mgcv' — smooth_label","title":"Extract the label for a smooth used by 'mgcv' — smooth_label","text":"label 'mgcv' uses smooths useful many contexts, including selecting smooths labelling plots. smooth_label() extracts label 'mgcv' smooth object, .e. object inherits class \"mgcv.smooth\". typically found $smooth component GAM fitted mgcv::gam() mgcv::bam(), related functions.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/smooth_label.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Extract the label for a smooth used by 'mgcv' — smooth_label","text":"","code":"smooth_label(object, ...)  # S3 method for class 'gam' smooth_label(object, id, ...)  # S3 method for class 'mgcv.smooth' smooth_label(object, ...)"},{"path":"https://gavinsimpson.github.io/gratia/reference/smooth_label.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Extract the label for a smooth used by 'mgcv' — smooth_label","text":"object R object. Currently, methods class \"gam\" mgcv smooth objects inheriting class \"mgcv.smooth\" supported. ... arguments passed methods. id numeric; indices smooths whose labels extracted. missing, labels smooths model returned.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/smooth_label.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Extract the label for a smooth used by 'mgcv' — smooth_label","text":"character vector.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/smooth_label.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Extract the label for a smooth used by 'mgcv' — smooth_label","text":"","code":"load_mgcv() df <- data_sim(\"gwf2\", n = 100) m <- gam(y ~ s(x), data = df, method = \"REML\")  # extract the smooth sm <- get_smooths_by_id(m, id = 1)[[1]]  # extract the label smooth_label(sm) #> [1] \"s(x)\"  # or directly on the fitted GAM smooth_label(m$smooth[[1]]) #> [1] \"s(x)\"  # or extract labels by idex/position smooth_label(m, id = 1) #> [1] \"s(x)\""},{"path":"https://gavinsimpson.github.io/gratia/reference/smooth_samples.html","id":null,"dir":"Reference","previous_headings":"","what":"Posterior draws for individual smooths — smooth_samples","title":"Posterior draws for individual smooths — smooth_samples","text":"Returns draws posterior distributions smooth functions GAM. Useful, example, visualising uncertainty individual estimated functions.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/smooth_samples.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Posterior draws for individual smooths — smooth_samples","text":"","code":"smooth_samples(model, ...)  # S3 method for class 'gam' smooth_samples(   model,   select = NULL,   term = deprecated(),   n = 1,   data = newdata,   method = c(\"gaussian\", \"mh\", \"inla\", \"user\"),   seed = NULL,   freq = FALSE,   unconditional = FALSE,   n_cores = 1L,   n_vals = 200,   burnin = 1000,   thin = 1,   t_df = 40,   rw_scale = 0.25,   rng_per_smooth = FALSE,   draws = NULL,   partial_match = NULL,   mvn_method = c(\"mvnfast\", \"mgcv\"),   ...,   newdata = NULL,   ncores = NULL )"},{"path":"https://gavinsimpson.github.io/gratia/reference/smooth_samples.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Posterior draws for individual smooths — smooth_samples","text":"model fitted model supported types ... arguments passed methods. fitted_samples(), passed mgcv::predict.gam(). posterior_samples() passed fitted_samples(). predicted_samples() passed relevant simulate() method. select character; select smooth's posterior draw . default (NULL) means posteriors smooths model wil sampled . supplied, character vector requested terms. term Use select instead. n numeric; number posterior samples return. data data frame; new observations posterior draws model evaluated. supplied, data used fit model used data, available model. method character; method used draw samples posterior distribution. \"gaussian\" uses Gaussian (Laplace) approximation posterior. \"mh\" uses Metropolis Hastings sampler alternates t proposals proposals based shrunken version posterior covariance matrix. \"inla\" uses variant Integrated Nested Laplace Approximation due Wood (2019), (currently implemented). \"user\" allows user-supplied posterior draws (currently implemented). seed numeric; random seed simulations. freq logical; TRUE use frequentist covariance matrix parameter estimators, FALSE use Bayesian posterior covariance matrix parameters. unconditional logical; TRUE (freq == FALSE) Bayesian smoothing parameter uncertainty corrected covariance matrix used, available. n_cores number cores generating random variables multivariate normal distribution. Passed mvnfast::rmvn(). Parallelization take place OpenMP supported (appears work Windows current R). n_vals numeric; many locations evaluate smooth data supplied burnin numeric; number samples discard burnin draws. used method = \"mh\". thin numeric; number samples skip taking n draws. Results thin * n draws posterior taken. used method = \"mh\". t_df numeric; degrees freedom t distribution proposals. used method = \"mh\". rw_scale numeric; Factor scale posterior covariance matrix generating random walk proposals. Negative non finite skip random walk step. used method = \"mh\". rng_per_smooth logical; TRUE, behaviour gratia version 0.8.1 earlier used, whereby separate call random number generator (RNG) performed smooth. FALSE, single call RNG performed model parameters draws matrix; user supplied posterior draws used method = \"user\". partial_match logical; smooths selected partial matches select? TRUE, select can single string match . mvn_method character; one \"mvnfast\" \"mgcv\". default uses mvnfast::rmvn(), can considerably faster generate large numbers MVN random values mgcv::rmvn(), might work marginal fits, covariance matrix close singular. newdata Deprecated: use data instead. ncores Deprecated; use n_cores instead. number cores generating random variables multivariate normal distribution. Passed mvnfast::rmvn(). Parallelization take place OpenMP supported (appears work Windows current R).","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/smooth_samples.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Posterior draws for individual smooths — smooth_samples","text":"tibble additional classes \"smooth_samples\" `\"posterior_samples\". \"gam\" method, columns currently returned (order) : .smooth; character vector. Indicates smooth function particular draw, .term; character vector. Similar smooth, contain full label smooth, differentiate factor-smooths example. .; character vector. smooth involves term, variable named , NA_character_ otherwise. .row; integer. vector values seq_len(n_vals), repeated n > 1L. Indexes row data particular draw. .draw; integer. vector integer values indexing particular posterior draw row belongs . .value; numeric. value smooth function posterior draw covariate combination. xxx; numeric. series one columns containing data required smooth, named per variables involved respective smooth. Additional columns present case factor smooths, contain level factor named by_variable particular posterior draw.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/smooth_samples.html","id":"warning","dir":"Reference","previous_headings":"","what":"Warning","title":"Posterior draws for individual smooths — smooth_samples","text":"set variables returned order tibble subject change future versions. rely position.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/smooth_samples.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Posterior draws for individual smooths — smooth_samples","text":"Gavin L. Simpson","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/smooth_samples.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Posterior draws for individual smooths — smooth_samples","text":"","code":"load_mgcv() dat <- data_sim(\"eg1\", n = 400, seed = 2) m1 <- gam(y ~ s(x0) + s(x1) + s(x2) + s(x3), data = dat, method = \"REML\")  sms <- smooth_samples(m1, select = \"s(x0)\", n = 5, seed = 42) # \\donttest{ sms #> # A tibble: 1,000 x 8 #>    .smooth .term .type .by    .row .draw .value      x0 #>    <chr>   <chr> <chr> <chr> <int> <int>  <dbl>   <dbl> #>  1 s(x0)   s(x0) TPRS  NA        1     1 -0.357 0.00711 #>  2 s(x0)   s(x0) TPRS  NA        1     2 -0.465 0.00711 #>  3 s(x0)   s(x0) TPRS  NA        1     3 -0.720 0.00711 #>  4 s(x0)   s(x0) TPRS  NA        1     4 -1.27  0.00711 #>  5 s(x0)   s(x0) TPRS  NA        1     5 -1.18  0.00711 #>  6 s(x0)   s(x0) TPRS  NA        2     1 -0.365 0.0121  #>  7 s(x0)   s(x0) TPRS  NA        2     2 -0.464 0.0121  #>  8 s(x0)   s(x0) TPRS  NA        2     3 -0.708 0.0121  #>  9 s(x0)   s(x0) TPRS  NA        2     4 -1.24  0.0121  #> 10 s(x0)   s(x0) TPRS  NA        2     5 -1.16  0.0121  #> # i 990 more rows # }  ## A factor by example (with a spurious covariate x0) dat <- data_sim(\"eg4\", n = 1000, seed = 2)  ## fit model... m2 <- gam(y ~ fac + s(x2, by = fac) + s(x0), data = dat) sms <- smooth_samples(m2, n = 5, seed = 42) draw(sms)"},{"path":[]},{"path":"https://gavinsimpson.github.io/gratia/reference/smooth_terms.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"List the variables involved in smooths — smooth_terms","text":"","code":"smooth_terms(object, ...)"},{"path":"https://gavinsimpson.github.io/gratia/reference/smooth_terms.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"List the variables involved in smooths — smooth_terms","text":"object R object result call mgcv::gam(), mgcv::bam(), mgcv::gamm(), inherits classes \"gam\" \"mgcv.smooth\", \"fs.interaction\". ... arguments passed methods. Currently unused.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/smooth_type.html","id":null,"dir":"Reference","previous_headings":"","what":"Determine the type of smooth and return it n a human readable form — smooth_type","title":"Determine the type of smooth and return it n a human readable form — smooth_type","text":"Determine type smooth return n human readable form","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/smooth_type.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Determine the type of smooth and return it n a human readable form — smooth_type","text":"","code":"smooth_type(smooth)  # Default S3 method smooth_type(smooth)  # S3 method for class 'tprs.smooth' smooth_type(smooth)  # S3 method for class 'ts.smooth' smooth_type(smooth)  # S3 method for class 'cr.smooth' smooth_type(smooth)  # S3 method for class 'cs.smooth' smooth_type(smooth)  # S3 method for class 'cyclic.smooth' smooth_type(smooth)  # S3 method for class 'pspline.smooth' smooth_type(smooth)  # S3 method for class 'cpspline.smooth' smooth_type(smooth)  # S3 method for class 'Bspline.smooth' smooth_type(smooth)  # S3 method for class 'duchon.spline' smooth_type(smooth)  # S3 method for class 'fs.interaction' smooth_type(smooth)  # S3 method for class 'sz.interaction' smooth_type(smooth)  # S3 method for class 'gp.smooth' smooth_type(smooth)  # S3 method for class 'mrf.smooth' smooth_type(smooth)  # S3 method for class 'random.effect' smooth_type(smooth)  # S3 method for class 'sw' smooth_type(smooth)  # S3 method for class 'sf' smooth_type(smooth)  # S3 method for class 'soap.film' smooth_type(smooth)  # S3 method for class 't2.smooth' smooth_type(smooth)  # S3 method for class 'sos.smooth' smooth_type(smooth)  # S3 method for class 'tensor.smooth' smooth_type(smooth)  # S3 method for class 'mpi.smooth' smooth_type(smooth)  # S3 method for class 'mpd.smooth' smooth_type(smooth)  # S3 method for class 'cx.smooth' smooth_type(smooth)  # S3 method for class 'cv.smooth' smooth_type(smooth)  # S3 method for class 'micx.smooth' smooth_type(smooth)  # S3 method for class 'micv.smooth' smooth_type(smooth)  # S3 method for class 'mdcx.smooth' smooth_type(smooth)  # S3 method for class 'mdcv.smooth' smooth_type(smooth)  # S3 method for class 'miso.smooth' smooth_type(smooth)  # S3 method for class 'mifo.smooth' smooth_type(smooth)"},{"path":"https://gavinsimpson.github.io/gratia/reference/smooth_type.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Determine the type of smooth and return it n a human readable form — smooth_type","text":"smooth object inheriting class mgcv.smooth.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/smooths.html","id":null,"dir":"Reference","previous_headings":"","what":"Names of smooths in a GAM — smooths","title":"Names of smooths in a GAM — smooths","text":"Names smooths GAM","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/smooths.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Names of smooths in a GAM — smooths","text":"","code":"smooths(object)  # Default S3 method smooths(object)  # S3 method for class 'gamm' smooths(object)"},{"path":"https://gavinsimpson.github.io/gratia/reference/smooths.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Names of smooths in a GAM — smooths","text":"object fitted GAM related model. Typically result call mgcv::gam(), mgcv::bam(), mgcv::gamm().","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/spline_values.html","id":null,"dir":"Reference","previous_headings":"","what":"Evaluate a spline at provided covariate values — spline_values","title":"Evaluate a spline at provided covariate values — spline_values","text":"Evaluate spline provided covariate values","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/spline_values.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Evaluate a spline at provided covariate values — spline_values","text":"","code":"spline_values(   smooth,   data,   model,   unconditional,   overall_uncertainty = TRUE,   frequentist = FALSE )"},{"path":"https://gavinsimpson.github.io/gratia/reference/spline_values.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Evaluate a spline at provided covariate values — spline_values","text":"smooth currently object inherits class mgcv.smooth. data data frame values evaluate smooth . model fitted model; currently mgcv::gam() mgcv::bam() models suported. unconditional logical; confidence intervals include uncertainty due smoothness selection? TRUE, corrected Bayesian covariance matrix used. overall_uncertainty logical; uncertainty model constant term included standard error evaluate values smooth? frequentist logical; use frequentist covariance matrix?","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/spline_values2.html","id":null,"dir":"Reference","previous_headings":"","what":"Evaluate a spline at provided covariate values — spline_values2","title":"Evaluate a spline at provided covariate values — spline_values2","text":"function spline_values2() renamed spline_values() version 0.9.0. allowed following removal evaluate_smooth(), function using spline_values(). spline_values2() renamed spline_values().","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/spline_values2.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Evaluate a spline at provided covariate values — spline_values2","text":"","code":"spline_values2(   smooth,   data,   model,   unconditional,   overall_uncertainty = TRUE,   frequentist = FALSE )"},{"path":"https://gavinsimpson.github.io/gratia/reference/spline_values2.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Evaluate a spline at provided covariate values — spline_values2","text":"smooth currently object inherits class mgcv.smooth. data optional data frame values evaluate smooth . model fitted model; currently mgcv::gam() mgcv::bam() models suported. unconditional logical; confidence intervals include uncertainty due smoothness selection? TRUE, corrected Bayesian covariance matrix used. overall_uncertainty logical; uncertainty model constant term included standard error evaluate values smooth? frequentist logical; use frequentist covariance matrix?","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/term_names.html","id":null,"dir":"Reference","previous_headings":"","what":"Extract names of all variables needed to fit a GAM or a smooth — term_names","title":"Extract names of all variables needed to fit a GAM or a smooth — term_names","text":"Extract names variables needed fit GAM smooth","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/term_names.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Extract names of all variables needed to fit a GAM or a smooth — term_names","text":"","code":"term_names(object, ...)  # S3 method for class 'gam' term_names(object, ...)  # S3 method for class 'mgcv.smooth' term_names(object, ...)  # S3 method for class 'gamm' term_names(object, ...)"},{"path":"https://gavinsimpson.github.io/gratia/reference/term_names.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Extract names of all variables needed to fit a GAM or a smooth — term_names","text":"object fitted GAM object (inheriting class \"gam\" mgcv::smooth.construct smooth object, inheriting class \"mgcv.smooth\". ... arguments passed methods. currently used.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/term_names.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Extract names of all variables needed to fit a GAM or a smooth — term_names","text":"vector variable names required terms model","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/term_variables.html","id":null,"dir":"Reference","previous_headings":"","what":"Names of variables involved in a specified model term — term_variables","title":"Names of variables involved in a specified model term — term_variables","text":"Given name (term label) term model, returns names variables involved term.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/term_variables.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Names of variables involved in a specified model term — term_variables","text":"","code":"term_variables(object, term, ...)  # S3 method for class 'terms' term_variables(object, term, ...)  # S3 method for class 'gam' term_variables(object, term, ...)  # S3 method for class 'bam' term_variables(object, term, ...)"},{"path":"https://gavinsimpson.github.io/gratia/reference/term_variables.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Names of variables involved in a specified model term — term_variables","text":"object R object method dispatch performed term character; name model term, sense attr(terms(object), \"term.labels\"). Currently checked see term exists model. ... arguments passed methods.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/term_variables.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Names of variables involved in a specified model term — term_variables","text":"character vector variable names.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/theta.html","id":null,"dir":"Reference","previous_headings":"","what":"General extractor for additional parameters in mgcv models — theta","title":"General extractor for additional parameters in mgcv models — theta","text":"General extractor additional parameters mgcv models","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/theta.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"General extractor for additional parameters in mgcv models — theta","text":"","code":"theta(object, ...)  # S3 method for class 'gam' theta(object, transform = TRUE, ...)"},{"path":"https://gavinsimpson.github.io/gratia/reference/theta.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"General extractor for additional parameters in mgcv models — theta","text":"object fitted model ... arguments passed methods. transform logical; transform natural scale parameter","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/theta.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"General extractor for additional parameters in mgcv models — theta","text":"Returns numeric vector additional parameters","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/theta.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"General extractor for additional parameters in mgcv models — theta","text":"","code":"load_mgcv() df <- data_sim(\"eg1\", dist = \"poisson\", seed = 42, scale = 1 / 5) m <- gam(y ~ s(x0) + s(x1) + s(x2) + s(x3),   data = df, method = \"REML\",   family = nb() ) p <- theta(m)"},{"path":"https://gavinsimpson.github.io/gratia/reference/tidy_basis.html","id":null,"dir":"Reference","previous_headings":"","what":"A tidy basis representation of a smooth object — tidy_basis","title":"A tidy basis representation of a smooth object — tidy_basis","text":"Takes object class mgcv.smooth returns tidy representation basis.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/tidy_basis.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"A tidy basis representation of a smooth object — tidy_basis","text":"","code":"tidy_basis(smooth, data = NULL, at = NULL, coefs = NULL, p_ident = NULL)"},{"path":"https://gavinsimpson.github.io/gratia/reference/tidy_basis.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"A tidy basis representation of a smooth object — tidy_basis","text":"smooth smooth object inheriting class \"mgcv.smooth\". Typically, objects returned part fitted GAM GAMM $smooth component model object $gam$smooth component model fitted mgcv::gamm() gamm4::gamm4(). data data frame containing variables used smooth. data frame containing values smooth covariate(s) basis evaluated. coefs numeric; optional vector coefficients smooth p_ident logical vector; used handling scam::scam() smooths.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/tidy_basis.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"A tidy basis representation of a smooth object — tidy_basis","text":"tibble.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/tidy_basis.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"A tidy basis representation of a smooth object — tidy_basis","text":"Gavin L. Simpson","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/tidy_basis.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"A tidy basis representation of a smooth object — tidy_basis","text":"","code":"load_mgcv() df <- data_sim(\"eg1\", n = 400, seed = 42)  # fit model m <- gam(y ~ s(x0) + s(x1) + s(x2) + s(x3), data = df, method = \"REML\")  # tidy representaition of a basis for a smooth definition # extract the smooth sm <- get_smooth(m, \"s(x2)\") # get the tidy basis - need to pass where we want it to be evaluated bf <- tidy_basis(sm, at = df)  # can weight the basis by the model coefficients for this smooth bf <- tidy_basis(sm, at = df, coefs = smooth_coefs(sm, model = m))"},{"path":"https://gavinsimpson.github.io/gratia/reference/to_na.html","id":null,"dir":"Reference","previous_headings":"","what":"Sets the elements of vector to NA — to_na","title":"Sets the elements of vector to NA — to_na","text":"Given vector indexing elements x, sets selected elements x NA.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/to_na.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Sets the elements of vector to NA — to_na","text":"","code":"to_na(x, i)"},{"path":"https://gavinsimpson.github.io/gratia/reference/to_na.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Sets the elements of vector to NA — to_na","text":"x vector values vector values used subset x","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/to_na.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Sets the elements of vector to NA — to_na","text":"Returns x possibly elements set NA","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/too_far.html","id":null,"dir":"Reference","previous_headings":"","what":"Exclude values that lie too far from the support of data — too_far","title":"Exclude values that lie too far from the support of data — too_far","text":"Identifies pairs covariate values lie far original data. function currently basic wrapper around mgcv::exclude..far().","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/too_far.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Exclude values that lie too far from the support of data — too_far","text":"","code":"too_far(x, y, ref_1, ref_2, dist = NULL)"},{"path":"https://gavinsimpson.github.io/gratia/reference/too_far.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Exclude values that lie too far from the support of data — too_far","text":"x, y numeric; vector values covariates compare observed data ref_1, ref_2 numeric; vectors covariate values represent reference x1 x2` compared dist supplied, numeric vector length 1 representing distance data beyond observation excluded. example, want exclude values lie observation 10% range observed data, use 0.1.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/too_far.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Exclude values that lie too far from the support of data — too_far","text":"Returns logical vector length x1.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/too_far_to_na.html","id":null,"dir":"Reference","previous_headings":"","what":"Set rows of data to NA if the lie too far from a reference set of values — too_far_to_na","title":"Set rows of data to NA if the lie too far from a reference set of values — too_far_to_na","text":"Set rows data NA lie far reference set values","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/too_far_to_na.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Set rows of data to NA if the lie too far from a reference set of values — too_far_to_na","text":"","code":"too_far_to_na(smooth, input, reference, cols, dist = NULL)"},{"path":"https://gavinsimpson.github.io/gratia/reference/too_far_to_na.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Set rows of data to NA if the lie too far from a reference set of values — too_far_to_na","text":"smooth mgcv smooth object input data frame containing input observations columns set NA reference data frame containing reference values cols character vector columns whose elements set NA data lies far reference set dist numeric, distance reference set beyond elements input set NA","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/transform_fun.html","id":null,"dir":"Reference","previous_headings":"","what":"Transform estimated values and confidence intervals by applying a function — transform_fun","title":"Transform estimated values and confidence intervals by applying a function — transform_fun","text":"Transform estimated values confidence intervals applying function","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/transform_fun.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Transform estimated values and confidence intervals by applying a function — transform_fun","text":"","code":"transform_fun(object, fun = NULL, ...)  # S3 method for class 'smooth_estimates' transform_fun(object, fun = NULL, constant = NULL, ...)  # S3 method for class 'smooth_samples' transform_fun(object, fun = NULL, constant = NULL, ...)  # S3 method for class 'mgcv_smooth' transform_fun(object, fun = NULL, constant = NULL, ...)  # S3 method for class 'evaluated_parametric_term' transform_fun(object, fun = NULL, constant = NULL, ...)  # S3 method for class 'parametric_effects' transform_fun(object, fun = NULL, constant = NULL, ...)  # S3 method for class 'tbl_df' transform_fun(object, fun = NULL, column = NULL, constant = NULL, ...)"},{"path":"https://gavinsimpson.github.io/gratia/reference/transform_fun.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Transform estimated values and confidence intervals by applying a function — transform_fun","text":"object object apply transform function . fun function apply. ... additional arguments passed methods. constant numeric; constant apply transformation. column character; \"tbl_df\" method, column transform.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/transform_fun.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Transform estimated values and confidence intervals by applying a function — transform_fun","text":"Returns object estimate upper lower values confidence interval transformed via function.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/transform_fun.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Transform estimated values and confidence intervals by applying a function — transform_fun","text":"Gavin L. Simpson","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/typical_values.html","id":null,"dir":"Reference","previous_headings":"","what":"Typical values of model covariates — typical_values","title":"Typical values of model covariates — typical_values","text":"Typical values model covariates","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/typical_values.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Typical values of model covariates — typical_values","text":"","code":"typical_values(object, ...)  # S3 method for class 'gam' typical_values(   object,   vars = everything(),   envir = environment(formula(object)),   data = NULL,   ... )  # S3 method for class 'data.frame' typical_values(object, vars = everything(), ...)"},{"path":"https://gavinsimpson.github.io/gratia/reference/typical_values.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Typical values of model covariates — typical_values","text":"object fitted GAM(M) model. ... arguments passed methods. vars terms include exclude returned object. Uses tidyselect principles. envir environment within recreate data used fit object. data optional data frame data used fit model reconstruction data model work.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/user_draws.html","id":null,"dir":"Reference","previous_headings":"","what":"Handle user-supplied posterior draws — user_draws","title":"Handle user-supplied posterior draws — user_draws","text":"Handle user-supplied posterior draws","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/user_draws.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Handle user-supplied posterior draws — user_draws","text":"","code":"user_draws(model, draws, ...)  # S3 method for class 'gam' user_draws(model, draws, index = NULL, ...)"},{"path":"https://gavinsimpson.github.io/gratia/reference/user_draws.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Handle user-supplied posterior draws — user_draws","text":"model fitted R model. Currently models fitted mgcv::gam() mgcv::bam(), return object inherits objects supported. , \"inherits\" used loose fashion; models fitted scam::scam() support even though models strictly inherit class \"gam\" far inherits() concerned. draws matrix; user supplied posterior draws used method = \"user\". ... arguments passed methods. index vector index (subset) columns draws.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/user_draws.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Handle user-supplied posterior draws — user_draws","text":"supplied draws must matrix (currently), 1 column per model coefficient, 1 row per posterior draw. \"gam\" method argument index, can used subset (select) coefficients (columns) draws. index can valid way selecting (indexing) columns matrix. index useful set posterior draws entire model (say mgcv::gam.mh()) wish use draws individual smooth, via smooth_samples().","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/variance_comp.html","id":null,"dir":"Reference","previous_headings":"","what":"Variance components of smooths from smoothness estimates — variance_comp","title":"Variance components of smooths from smoothness estimates — variance_comp","text":"wrapper mgcv::gam.vcomp() returns smoothing parameters expressed variance components.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/variance_comp.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Variance components of smooths from smoothness estimates — variance_comp","text":"","code":"variance_comp(object, ...)  # S3 method for class 'gam' variance_comp(object, rescale = TRUE, coverage = 0.95, ...)"},{"path":"https://gavinsimpson.github.io/gratia/reference/variance_comp.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Variance components of smooths from smoothness estimates — variance_comp","text":"object R object. Currently models fitted mgcv::gam() mgcv::bam() supported. ... arguments passed methods rescale logical; numerical stability reasons penalty matrices smooths rescaled fitting. rescale = TRUE, rescaling undone, resulting variance components original scale. needed comparing mixed model software, lmer(). coverage numeric; value 0 1 indicating (approximate) coverage confidence interval returned.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/variance_comp.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Variance components of smooths from smoothness estimates — variance_comp","text":"function wrapper mgcv::gam.vcomp() performs three additional services suppresses annoying text output mgcv::gam.vcomp() prints terminal, returns variance smooth well standard deviation, returns variance components tibble.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/vars_from_label.html","id":null,"dir":"Reference","previous_headings":"","what":"Returns names of variables from a smooth label — vars_from_label","title":"Returns names of variables from a smooth label — vars_from_label","text":"Returns names variables smooth label","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/vars_from_label.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Returns names of variables from a smooth label — vars_from_label","text":"","code":"vars_from_label(label)"},{"path":"https://gavinsimpson.github.io/gratia/reference/vars_from_label.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Returns names of variables from a smooth label — vars_from_label","text":"label character; length 1 character vector containing label smooth.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/vars_from_label.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Returns names of variables from a smooth label — vars_from_label","text":"","code":"vars_from_label(\"s(x1)\") #> [1] \"x1\" vars_from_label(\"t2(x1,x2,x3)\") #> [1] \"x1\" \"x2\" \"x3\""},{"path":"https://gavinsimpson.github.io/gratia/reference/which_smooths.html","id":null,"dir":"Reference","previous_headings":"","what":"Identify a smooth term by its label — which_smooths","title":"Identify a smooth term by its label — which_smooths","text":"Identify smooth term label","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/which_smooths.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Identify a smooth term by its label — which_smooths","text":"","code":"which_smooths(object, ...)  # Default S3 method which_smooths(object, ...)  # S3 method for class 'gam' which_smooths(object, terms, ...)  # S3 method for class 'bam' which_smooths(object, terms, ...)  # S3 method for class 'gamm' which_smooths(object, terms, ...)"},{"path":"https://gavinsimpson.github.io/gratia/reference/which_smooths.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Identify a smooth term by its label — which_smooths","text":"object fitted GAM. ... arguments passed methods. terms character; one (partial) term labels identify required smooths.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/worm_plot.html","id":null,"dir":"Reference","previous_headings":"","what":"Worm plot of model residuals — worm_plot","title":"Worm plot of model residuals — worm_plot","text":"Worm plot model residuals","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/worm_plot.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Worm plot of model residuals — worm_plot","text":"","code":"worm_plot(model, ...)  # S3 method for class 'gam' worm_plot(   model,   method = c(\"uniform\", \"simulate\", \"normal\", \"direct\"),   type = c(\"deviance\", \"response\", \"pearson\"),   n_uniform = 10,   n_simulate = 50,   level = 0.9,   ylab = NULL,   xlab = NULL,   title = NULL,   subtitle = NULL,   caption = NULL,   ci_col = \"black\",   ci_alpha = 0.2,   point_col = \"black\",   point_alpha = 1,   line_col = \"red\",   ... )  # S3 method for class 'glm' worm_plot(model, ...)  # S3 method for class 'lm' worm_plot(model, ...)"},{"path":"https://gavinsimpson.github.io/gratia/reference/worm_plot.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Worm plot of model residuals — worm_plot","text":"model fitted model. Currently models inheriting class \"gam\", well classes \"glm\" \"lm\" calls stats::glm stats::lm supported. ... arguments passed ot methods. method character; method used generate theoretical quantiles. default \"uniform\", generates reference quantiles using random draws uniform distribution inverse cummulative distribution function (CDF) fitted values. reference quantiles averaged n_uniform draws. \"simulate\" generates reference quantiles simulating new response data model observed values covariates, residualised generate reference quantiles, using n_simulate simulated data sets. \"normal\" generates reference quantiles using standard normal distribution. \"uniform\" computationally efficient, \"simulate\" allows reference bands drawn QQ-plot. \"normal\" avoided used fall back random number generator (\"simulate\") inverse CDF available family used model fitting (`\"uniform\"“). Note method = \"direct\" deprecated favour method = \"uniform\". type character; type residuals use. \"deviance\", \"response\", \"pearson\" residuals allowed. n_uniform numeric; number times randomize uniform quantiles direct computation method (method = \"uniform\"). n_simulate numeric; number data sets simulate estimated model using simulation method (method = \"simulate\"). level numeric; coverage level reference intervals. Must strictly 0 < level < 1. used method = \"simulate\". ylab character expression; label y axis. supplied, suitable label generated. xlab character expression; label y axis. supplied, suitable label generated. title character expression; title plot. See ggplot2::labs(). May vector, one per penalty. subtitle character expression; subtitle plot. See ggplot2::labs(). May vector, one per penalty. caption character expression; plot caption. See ggplot2::labs(). May vector, one per penalty. ci_col fill colour reference interval method = \"simulate\". ci_alpha alpha transparency reference interval method = \"simulate\". point_col colour points QQ plot. point_alpha alpha transparency points QQ plot. line_col colour used draw reference line.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/worm_plot.html","id":"note","dir":"Reference","previous_headings":"","what":"Note","title":"Worm plot of model residuals — worm_plot","text":"wording used mgcv::qq.gam() uses direct reference simulated residuals method (method = \"simulated\"). avoid confusion, method = \"direct\" deprecated favour method = \"uniform\".","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/worm_plot.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Worm plot of model residuals — worm_plot","text":"","code":"load_mgcv() ## simulate binomial data... dat <- data_sim(\"eg1\", n = 200, dist = \"binary\", scale = .33, seed = 0) p <- binomial()$linkinv(dat$f) # binomial p n <- sample(c(1, 3), 200, replace = TRUE) # binomial n dat <- transform(dat, y = rbinom(n, n, p), n = n) m <- gam(y / n ~ s(x0) + s(x1) + s(x2) + s(x3),   family = binomial, data = dat, weights = n,   method = \"REML\" )  ## Worm plot; default using direct randomization of uniform quantiles ## Note no reference bands are drawn with this method. worm_plot(m)   ## Alternatively use simulate new data from the model, which ## allows construction of reference intervals for the Q-Q plot worm_plot(m,   method = \"simulate\", point_col = \"steelblue\",   point_alpha = 0.4 )   ## ... or use the usual normality assumption worm_plot(m, method = \"normal\")"},{"path":"https://gavinsimpson.github.io/gratia/reference/zooplankton.html","id":null,"dir":"Reference","previous_headings":"","what":"Madison lakes zooplankton data — zooplankton","title":"Madison lakes zooplankton data — zooplankton","text":"Madison lake zooplankton data long-term study seasonal dynamics zooplankton, collected Richard Lathrop. data collected chain lakes Wisconsin (Mendota, Monona, Kegnonsa, Waubesa) approximately bi-weekly 1976 1994. consist samples zooplankton communities, taken deepest point lake via vertical tow. data provided Wisconsin Department Natural Resources collection processing fully described Lathrop (2000).","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/zooplankton.html","id":"format","dir":"Reference","previous_headings":"","what":"Format","title":"Madison lakes zooplankton data — zooplankton","text":"data frame","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/zooplankton.html","id":"source","dir":"Reference","previous_headings":"","what":"Source","title":"Madison lakes zooplankton data — zooplankton","text":"Pedersen EJ, Miller DL, Simpson GL, Ross N. 2018. Hierarchical generalized additive models: introduction mgcv. PeerJ Preprints 6:e27320v1 doi:10.7287/peerj.preprints.27320v1 .","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/zooplankton.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Madison lakes zooplankton data — zooplankton","text":"record consists counts given zooplankton taxon taken subsample single vertical net tow, scaled account relative volume subsample versus whole net sample area net tow rounded nearest 1000 give estimated population density per m2 taxon point time sampled lake.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/zooplankton.html","id":"references","dir":"Reference","previous_headings":"","what":"References","title":"Madison lakes zooplankton data — zooplankton","text":"Lathrop RC. (2000). Madison Wisonsin Lakes Zooplankton 1976–1994. Environmental Data Initiative.","code":""},{"path":[]},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"new-features-development-version","dir":"Changelog","previous_headings":"","what":"New features","title":"gratia (development version)","text":"conditional_values() draw() method compute plot predictions fitted GAM conditional one covariates. function wrapper around fitted_values() allows user simple ways specify covariates condition values covariates take. provides similar functionality marginaleffects::plot_predictions(), simpler. See #300. penalty() basis() can now allow smooth reparameterized resulting basis identity matrix. clearly highlights penalty null space, functions penalty effect . draw.gam() draw.smooth_estimates() gain argument caption, , set FALSE plot smooth basis type caption plot. #307 appraise() qq_plot.gam() now allow user set random seed used generating reference quantiles method = \"uniform\" method = \"simulate\".","code":""},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"bug-fixes-development-version","dir":"Changelog","previous_headings":"","what":"Bug fixes","title":"gratia (development version)","text":"derivative_samples() ignoring scale argument. #293 Reported @jonathonmellor Argument level derivative_samples() included accidentally. v0.9.2.9002 argument deprecated using now generate warning. #291 draw() plotting cyclic P spline smooths. Reported @Zuckerbrot #297 derivatives() fail \"fs\" smooths parametric effects model. Reported @mahonmb #301 Partial residuals partial_residuals() draw.gam() wrong GAMs fitted family = binomial() weights argument contained binomial sample sizes prior weights used form weighted working residuals. Now working weights used instead. Reported @emchuron #273 Internal function gammals_link() expecting \"theta\" synonym scale parameter master table \"phi\" coded synonym. Now work expected. level() assumed level single value even though handle multiple levels. #321","code":""},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"gratia-092","dir":"Changelog","previous_headings":"","what":"gratia 0.9.2","title":"gratia 0.9.2","text":"CRAN release: 2024-06-25","code":""},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"breaking-changes-0-9-2","dir":"Changelog","previous_headings":"","what":"Breaking changes","title":"gratia 0.9.2","text":"parametric_effects() slightly escaped great renaming happened 0.9.0. Columns type term gain prefix .. now rectified two columns now .type .term.","code":""},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"user-visible-changes-0-9-2","dir":"Changelog","previous_headings":"","what":"User visible changes","title":"gratia 0.9.2","text":"Plots random effects now labelled smooth label. Previously, title taken fro variable involved smooth, doesn’t work terms like s(subject, continuous_var, bs = \"re\") random slopes, previsouly title \"subject\". Now terms title \"s(subject,continuous_var)\". Simple random intercept terms, s(subject, bs = \"re\"), now titled \"s(subject)\". #287 vignettes custom-plotting.Rmd, posterior-simulation.Rmd moved vignettes/articles thus longer available package vignettes. Instead, accessible Articles package website: https://gavinsimpson.github.io/gratia/","code":""},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"new-features-0-9-2","dir":"Changelog","previous_headings":"","what":"New features","title":"gratia 0.9.2","text":"fitted_samples() now works gam() models multiple linear predictors, currently location parameter supported. parameter indicated new variable .parameter returned object.","code":""},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"bug-fixes-0-9-2","dir":"Changelog","previous_headings":"","what":"Bug fixes","title":"gratia 0.9.2","text":"partial_residuals() computing partial residuals deviance residuals. compatibility mgcv::plot.gam(), partial residuals now computed working residuals. Reported @wStockhausen #273 appraise() passing ci_col argument qq_plot() worm_plot(). Reported Sate Ahmed. Couldn’t pass mvn_method posterior sampling functions user facing functions fitted_samples(), posterior_samples(), smooth_samples(), derivative_samples(), repsonse_derivatives(). Reported @stefgehrig #279 fitted_values() works quantile GAMs fitted qgam(). confint.gam() applying shift estimate upper lower interval. #280 reported @TIMAVID & @rbentham parametric_effects() draw.parametric_effects() forget levels factors (intentionally), lead problems ordered factors ordering levels preserved. Now, parametric_effects() returns named list factor levels attribute \"factor_levels\" containing required information order levels preserved plotting. #284 Reported @mhpob parametric_effects() fail parametric terms model interaction terms (don’t currently handle). #282","code":""},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"gratia-090","dir":"Changelog","previous_headings":"","what":"gratia 0.9.0","title":"gratia 0.9.0","text":"CRAN release: 2024-03-27","code":""},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"breaking-changes-0-9-0","dir":"Changelog","previous_headings":"","what":"Breaking changes","title":"gratia 0.9.0","text":"Many functions now return objects different named variables. order avoid clashes variable names used user’s models data, period (.) now used prefix generated variable names. functions whose names changed : smooth_estimates(), fitted_values(), fitted_samples(), posterior_samples(), derivatives(), partial_derivatives(), derivative_samples(). addition, add_confint() also adds newly-named variables. derivatives() partial_derivatives() now work like smooth_estimates(); place var data columns, gratia now stores data variables derivatives evaluated columns object actual variable names. way spline---sphere (SOS) smooths (bs = \"sos\") plotted changed use ggplot2::coord_sf() instead previously-used ggplot2::coord_map(). changed made result coord_map() soft-deprecated (“superseded”) minor versions ggplot2 now already, changes guides system version 3.5.0 ggplot2. axes plots created coord_map() never really worked correctly changing angle tick labels never worked. coord_map() superseded, didn’t receive updates guides system side effect changes, code plotted SOS smooths producing warning release ggplot2 version 3.5.0. projection settings used draw SOS smooths previously controlled via arguments projection orientation. arguments affect ggplot2::coord_sf(), Instead projection used controlled new argument crs, takes PROJ string detailing projection use integer refers known coordinate reference system (CRS). default projection used +proj=ortho +lat_0=20 +lon_0=XX XX mean longitude coordinates data points.","code":"1. `est` is now `.estimate`, 2. `lower` and `upper` are now `.lower_ci` and `.upper_ci`, 3. `draw` and `row` and now `.draw` and `.row` respectively, 4. `fitted`, `se`, `crit` are now `.fitted`, `.se`, `.crit`, respectively 5. `smooth`, `by`, and `type` in `smooth_estimates()` are now `.smooth`,    `.by`, `.type`, respectively."},{"path":[]},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"defunct-0-9-0","dir":"Changelog","previous_headings":"Breaking changes > Defunct and deprecated functions and arguments","what":"Defunct","title":"gratia 0.9.0","text":"evaluate_smooth() deprecated gratia version 0.7.0. function ’s methods removed package. Use smooth_estimates() instead.","code":""},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"deprecated-functions-0-9-0","dir":"Changelog","previous_headings":"Breaking changes > Defunct and deprecated functions and arguments","what":"Deprecated functions","title":"gratia 0.9.0","text":"following functions deprecated version 0.9.0 gratia. eventually removed package part clean ahead eventual 1.0.0 release. functions become defunct version 0.11.0 1.0.0, whichever released soonest. evaluate_parametric_term() deprecated. Use parametric_effects() instead. datagen() deprecated. never really originally designed , replaced data_slice().","code":""},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"deprecated-arguments-0-9-0","dir":"Changelog","previous_headings":"Breaking changes > Defunct and deprecated functions and arguments","what":"Deprecated arguments","title":"gratia 0.9.0","text":"make functions package consistent, arguments select, term, smooth used thing hence latter two deprecated favour select. deprecated argument used, warning issued value assigned argument assigned select function continue.","code":""},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"user-visible-changes-0-9-0","dir":"Changelog","previous_headings":"","what":"User visible changes","title":"gratia 0.9.0","text":"smooth_samples() now uses single call RNG generate draws posterior smooths. Previous version 0.9.0, smooth_samples() separate call mvnfast::rmvn() smooth. result, result call smooth_samples() model multiple smooths now produce different results generated previously. regain old behaviour, add rng_per_smooth = TRUE smooth_samples() call. Note, however, using per-smooth RNG calls method = \"mh\" inefficient , method, posterior draws coefficients model sampled . , use rng_per_smooth = TRUE method = \"gaussian\". output smooth_estimates() draw() method changed tensor product smooths involve one 2D marginal smooths. Now, covariate values supplied via data argument, smooth_estimates() identifies one marginals 2d surface allows covariates involved surface vary fastest, ahead terms marginals. change made provides better default nothing provided data. also affects draw.gam(). fitted_values() now level support location, scale, shape families. Supported families mgcv::gaulss(), mgcv::gammals(), mgcv::gumbls(), mgcv::gevlss(), mgcv::shash(), mgcv::twlss(), mgcv::ziplss(). gratia now requires dplyr versions >= 1.1.0 tidyselect >= 1.2.0. new vignette Posterior Simulation available, describes posterior simulation fitted GAMs using {gratia}.","code":""},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"new-features-0-9-0","dir":"Changelog","previous_headings":"","what":"New features","title":"gratia 0.9.0","text":"Soap film smooths using basis bs = \"\" now handled draw(), smooth_estimates() etc. #8 response_derivatives() new function computing derivatives response respect (continuous) focal variable. First second order derivatives can computed using forward, backward, central finite differences. uncertainty estimated derivative determined using posterior sampling via fitted_samples(), hence can derived Gaussian approximation posterior using Metropolis Hastings sampler (see .) derivative_samples() work horse function behind response_derivatives(), computes returns posterior draws derivatives additive combination model terms. Requested @jonathanmellor #237 data_sim() can now simulate response data gamma, Tweedie ordered categorical distributions. data_sim() gains two new example models \"gwf2\", simulating data Gu & Wabha’s f2 function, \"lwf6\", example function 6 Luo & Wabha (1997 JASA 92(437), 107-116). data_sim() can also simulate data use GAMs fitted using family = gfam() grouped families different types data response handled. #266 part #265 fitted_samples() smooth_samples() can now use Metropolis Hastings sampler mgcv::gam.mh(), instead Gaussian approximation, sample posterior distribution model specific smooths respectively. posterior_samples() new function family fitted_samples() smooth_samples(). posterior_samples() returns draws posterior distribution response, combining uncertainty estimated expected value response dispersion response distribution. difference posterior_samples() predicted_samples() latter includes variation due drawing samples conditional distribution response (uncertainty expected values ignored), former includes sources uncertainty. fitted_samples() can new use matrix user-supplied posterior draws. Related #120 add_fitted_samples(), add_predicted_samples(), add_posterior_samples(), add_smooth_samples() new utility functions add respective draws posterior distribution existing data object covariate values object: obj |> add_posterior_draws(model). #50 basis_size() new function extract basis dimension (number basis functions) smooths. Methods available objects inherit classes \"gam\", \"gamm\", \"mgcv.smooth\" (individual smooths). data_slice() gains method data frames tibbles. typical_values() gains method data frames tibbles. fitted_values() now works models fitted using mgcv::ocat() family. predicted probability category returned, alongside Wald interval created using standard error (SE) estimated probability. SE estimated probabilities transformed logit (linear predictor) scale, Wald credible interval formed, back-transformed response (probability) scale. fitted_values() now works GAMMs fitted using mgcv::gamm(). Fitted (predicted) values use GAM part model, thus exclude random effects. link() inv_link() work models fitted using cnorm() family. worm plot can now drawn place QQ plot appraise() via new argument use_worm = TRUE. #62 smooths() now works models fitted mgcv::gamm(). overview() now returns basis dimension smooth gains argument stars TRUE add significance stars output plus legend printed tibble footer. Part wish @noamross #214 New add_constant() transform_fun() methods smooth_samples(). evenly() gains arguments lower upper modify lower / upper bound interval evenly spaced values generated. add_sizer() new function add information whether derivative smooth significantly changing (credible interval excludes 0). Currently, methods derivatives() smooth_estimates() objects implemented. Part request @asanders11 #117 draw.derivatives() gains arguments add_change change_type allow derivatives smooths plotted indicators credible interval derivative excludes 0. Options allow periods decrease increase differentiated via change_type = \"sizer\" instead default change_type = \"change\", emphasises either type change way. Part wish @asanders11 #117 draw.gam() can now group factor smooths given factor single panel, rather plotting smooths level separate panels. achieved via new argument grouped_by. Requested @RPanczak #89 draw.smooth_estimates() can now also group factor smooths given factor single panel. underlying plotting code used draw_smooth_estimates() univariate smooths can now add change indicators plots smooths change indicators added object created smooth_estimates() using add_sizer(). See example ?draw.smooth_estimates. smooth_estimates() can, evaluating 3D 4D tensor product smooth, identify one 2D smooths marginal tensor product. users provide covariate values evaluate smooths, smooth_estimates() focus 2D marginal smooth (first one involved tensor product), instead following ordering terms definition tensor product. #191 example, te(z, x, y, bs = c(cr, ds), d = c(1, 2)), second marginal smooth 2D Duchon spline covariates x y. Previously, smooth_estimates() generated n values z x n_3d values y, evaluated tensor product combinations generated values. ignore structure implicit tensor product, likely want know surface estimated Duchon spline x y smoothly varies z. Previously smooth_estimates() generate surfaces z x, varying y. Now, smooth_estimates() correctly identifies one marginal smooths tensor product 2D surface focus surface varying terms tensor product. improved behaviour needed bam() models always possible obvious thing reorder smooths defining tensor product te(x, y, z, bs = c(ds, cr), d = c(2, 1)). discrete = TRUE used bam() terms tensor product may get rearranged model setup maximum efficiency (See Details ?mgcv::bam). Additionally, draw.gam() now also works way. New function null_deviance() extracts null deviance fitted model. draw(), smooth_estimates(), fitted_values(), data_slice(), smooth_samples() now work models fitted scam::scam(). matters, current support extends univariate smooths. generate_draws() new low-level function generating posterior draws fitted model coefficients. generate_daws() S3 generic function extensible users. Currently provides simple interface simple Gaussian approximation sampler (gaussian_draws()) simple Metropolis Hasting sample (mh_draws()) available via mgcv::gam.mh(). #211 smooth_label() new function extracting labels ‘mgcv’ creates smooths smooth object . penalty() default method works s(), te(), t2(), ti(), create smooth specification. transform_fun() gains argument constant allow addition constant value objects (e.g. estimate confidence interval). enables single obj |> transform_fun(fun = exp, constant = 5) instead separate calls add_constant() transform_fun(). Part discussion #79 model_constant() new function simply extracts first coefficient estimated model.","code":""},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"bug-fixes-0-9-0","dir":"Changelog","previous_headings":"","what":"Bug fixes","title":"gratia 0.9.0","text":"link(), inv_link(), related family functions ocat() weren’t correctly identifying family name result throw error even passed object correct family. link() inv_link() now work correctly betar() family fitted GAM. print() method lp_matrix() now converts matrix data frame conversion tibble. makes sense results typical behaviour columns printed object doubles. Constrained factor smooths (bs = \"sz\") factor first variable mentioned smooth (.e. s(x, f, bs = \"sz\") continuous x factor f) now plotable draw(). #208 parametric_effects() unable handle special parametric terms like poly(x) log(x) formulas. Reported @fhui28 #212 parametric_effects() now works better location, scale, shape models. Reported @pboesu #45 parametric_effects now works missing values one variables used fitted GAM. #219 response_derivatives() incorrectly using .data tidyselect selectors. typical_values() handle logical variables GAM fit mgcv stores numerics var.summary. affected evenly() data_slice(). #222 parametric_effects() fail two ordered factors model. Reported @dsmi31 #221 Continuous smooths evaluated median value variable instead value 1. #224 fitted_samples() (hence posterior_samples()) now handles models offset terms formula. Offset terms supplied via offset argument ignored mgcv:::predict.gam() hence ignored also gratia. Reported @jonathonmellor #231 #233 smooth_estimates() fail \"fs\" smooth multivariate base smoother used factor last variable specified definition smooth: s(x1, x2, f, bs = \"fs\", xt = list(bs = \"ds\")) work, s(f, x1, x2, bs = \"fs\", xt = list(bs = \"ds\")) (ordering variables places factor last) emit obscure error. ordering terms involved smooth now doesn’t matter. Reported @chrisaak #249. draw.gam() fail plotting multivariate base smoother used \"sz\" smooth. Now, use case identified message printed indicating (currently) gratia doesn’t know plot smooth. Reported @chrisaak #249. draw.gam() fail plotting multivariate base smoother used \"fs\" smooth. Now, use case identified message printed indicating (currently) gratia doesn’t know plot smooth. Reported @chrisaak #249. derivative_samples() fail order = 2 computing forward finite differences, regardless type order = 1. Partly reported @samlipworth #251. draw() method penalty() normalizing penalty range 0–1, claimed documented -1–1 argument normalize = TRUE. now fixed. smooth_samples() failing data supplied contained variables used smooth sampled. Hence generally fail unless single smooth sampled model contained single smooth. function never intended retain variables data written way fail relocating data columns end posterior sampling object. #255 draw.gam() draw.smooth_estimates() fail plotting univariate tensor product smooth (e.g. te(x), ti(x), t2()). Reported @wStockhausen #260 plot.smooth() printing factor level subtitles ordered factor smooths.","code":""},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"gratia-082","dir":"Changelog","previous_headings":"","what":"gratia 0.8.2","title":"gratia 0.8.2","text":"CRAN release: 2024-01-09 Small fixes CRAN.","code":""},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"gratia-081","dir":"Changelog","previous_headings":"","what":"gratia 0.8.1","title":"gratia 0.8.1","text":"CRAN release: 2023-02-02","code":""},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"user-visible-changes-0-8-1","dir":"Changelog","previous_headings":"","what":"User visible changes","title":"gratia 0.8.1","text":"smooth_samples() now returns objects variables involved smooths correct name. Previously variables named .x1, .x2, etc. Fixing #126 improving compatibility compare_smooths() smooth_estimates() allowed variables named correctly. gratia now depends version 1.8-41 later mgcv package.","code":""},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"new-features-0-8-1","dir":"Changelog","previous_headings":"","what":"New features","title":"gratia 0.8.1","text":"draw.gam() can now handle tensor products include marginal random effect smooth. Beware plotting smooths many levels, however, separate surface plot produced level.","code":""},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"bug-fixes-0-8-1","dir":"Changelog","previous_headings":"","what":"Bug fixes","title":"gratia 0.8.1","text":"Additional fixes changes dplyr 1.1.0. smooth_samples() now works sampling posteriors multiple smooths different dimension. #126 reported @Aariq","code":""},{"path":[]},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"user-visible-changes-0-8-0","dir":"Changelog","previous_headings":"","what":"User visible changes","title":"gratia 0.8.0","text":"{gratia} now depends R version 4.1 later. new vignette “Data slices” supplied {gratia}. Functions {gratia} harmonised use argument named data instead newdata passing new data evaluate features smooths. message printed newdata used now . Existing code need changed data takes value newdata. Note due way ... handled R, R script uses data argument, run versions gratia prior 8.0 (released; 0.7.3.8 using development version) user-supplied data silently ignored. , scripts using data check installed version gratia >= 0.8 package developers update depend versions >= 0.8 using gratia (>= 0.8) DESCRIPTION. order plots smooths changed draw.gam() match order smooths specified model formula. See Bug Fixes detail #154.","code":""},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"new-features-0-8-0","dir":"Changelog","previous_headings":"","what":"New features","title":"gratia 0.8.0","text":"Added basic support GAMLSS (distributional GAMs) fitted gamlss() function package GJRM. Support currently restricted draw() method. difference_smooths() can now include group means difference, many users expected. include group means use group_means = TRUE function call, e.g. difference_smooths(model, smooth = \"s(x)\", group_means = TRUE). Note: function still differs plot_diff() package itsadug, essentially computes differences model predictions. main practical difference effects beyond factor smooth, including random effects, may included plot_diff(). implements main wish #108 (@dinga92) #143 (@mbolyanatz) despite protestations complicated cases (isn’t; complexity just cancels .) data_slice() totally revised. Now, user provides values variables want slice variables model specified held typical values (.e. value observation closest median numeric variables, modal factor level.) Data slices now produced passing name = value pairs variables values want appear slice. example value pair can expression looked (evaluated) data argument model frame fitted model (default). example, resulting slice data frame 100 observations, comprising x1, vector 100 values spread evenly range x1, constant value mean x2 x2 variable, constant factor level, model class fac, fac variable model. partial_derivatives() new function computing partial derivatives multivariate smooths (e.g. s(x,z), te(x,z)) respect one margins smooth. Multivariate smooths dimension handled, one dimensions allowed vary. Partial derivatives estimated using method finite differences, forward, backward, central finite differences available. Requested @noamross #101 overview() provides simple overview model terms fitted GAMs. new bs = \"sz\" basis released mgcv version 1.18-41 now supported smooth_estimates(), draw.gam(), draw.smooth_estimates() basis unique plotting method. #202 basis() now method fitted GAM(M)s can extract estimated basis model plot , using estimated coefficients smooth weight basis. #137 also new draw.basis() method plotting results call basis(). method can now also handle bivariate bases. tidy_basis() lower level function heavy lifting basis(), now exported. tidy_basis() returns tidy representation basis supplied object inheriting class \"mgcv.smooth\". objects returned $smooth component fitted GAM(M) model. lp_matrix() new utility function quickly return linear predictor matrix estimated model. wrapper predict(..., type = \"lpmatrix\") evenly() synonym seq_min_max() preferred going forward. Gains argument produce sequences covariate increment units . ref_level() level() new utility functions extracting reference specific level factor respectively. useful specifying covariate values condition data slice. model_vars() new, public facing way returning vector variables used model. difference_smooths() now use user-supplied data points evaluate pair smooths. Also note argument newdata renamed data. #175 draw() method difference_smooths() now uses better labels plot titles avoid long labels even modest factor levels. derivatives() now works factor-smooth interaction (\"fs\") smooths. draw() methods now allow angle tick labels x axis plots rotated using argument angle. Requested @tamas-ferenci #87 draw.gam() related functions (draw.parametric_effects(), draw.smooth_estimates()) now add basis plot using caption. #155 smooth_coefs() new utility function extracting coefficients particular smooth fitted model. smooth_coef_indices() associated function returns indices (positions) vector model coefficients (returned coef(gam_model)) coefficients pertain stated smooth. draw.gam() now better handles patchworks plots one plots fixed aspect ratios. #190","code":"m <- gam(y ~ s(x1) + x2 + fac) data_slice(model, x1 = evenly(x1, n = 100), x2 = mean(x2))"},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"bug-fixes-0-8-0","dir":"Changelog","previous_headings":"","what":"Bug fixes","title":"gratia 0.8.0","text":"draw.posterior_smooths now plots posterior samples fixed aspect ratio smooth isotropic. #148 derivatives() now ignores random effect smooths (derivatives don’t make sense anyway). #168 confint.gam(...., method = \"simultaneous\") now works factor smooths parm passed full name specific smooth s(x)faclevel. order plots produced gratia::draw.gam() matches order smooths entered model formula. Recent changes internals gratia::draw.gam() switch smooth_estimates() undertaken lead change behaviour resulting use dplyr::group_split(), ’s coercion internally character vector factor. factor now created explicitly, levels set correct order. #154 Setting dist argument set response smooth values NA lay far support data multivariate smooths, lead incorrect scale response guide. now fixed. #193 Argument fun draw.gam() applied parametric terms. Reported @grasshoppermouse #195 draw.gam() adding uncertainty linear predictors smooths overall_uncertainty = TRUE used. Now draw.gam() includes uncertainty linear predictors smooth takes part. #158 partial_derivatives() works provided single data point evaluate derivative. #199 transform_fun.smooth_estimates() addressing wrong variable names trying transform confidence interval. #201 data_slice() doesn’t fail error used model contains offset term. #198 confint.gam() longer uses evaluate_smooth(), soft deprecated. #167 qq_plot() worm_plot() compute wrong deviance residuals used generate theoretical quantiles exotic families (distributions) available mgcv. also affected appraise() QQ plot; residuals shown plots deviance residuals shown y-axis QQ plot correct. generation reference intervals/quantiles affected.","code":""},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"gratia-073","dir":"Changelog","previous_headings":"","what":"gratia 0.7.3","title":"gratia 0.7.3","text":"CRAN release: 2022-05-09","code":""},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"user-visible-changes-0-7-3","dir":"Changelog","previous_headings":"","what":"User visible changes","title":"gratia 0.7.3","text":"Plots smooths now use “Partial effect” y-axis label place “Effect”, better indicate displayed.","code":""},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"new-features-0-7-3","dir":"Changelog","previous_headings":"","what":"New features","title":"gratia 0.7.3","text":"confint.fderiv() confint.gam() now return results tibble instead common--garden data frame. latter mostly already . Examples confint.fderiv() confint.gam() reworked, part remove inconsistent output examples run M1 macs.","code":""},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"bug-fixes-0-7-3","dir":"Changelog","previous_headings":"","what":"Bug fixes","title":"gratia 0.7.3","text":"compare_smooths() failed passed non-standard model “names” like compare_smooths(m_gam, m_gamm$gam) compare_smooths(l[[1]], l[[2]]) even evaluated objects valid GAM(M) models. Reported Andrew Irwin #150","code":""},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"gratia-072","dir":"Changelog","previous_headings":"","what":"gratia 0.7.2","title":"gratia 0.7.2","text":"CRAN release: 2022-03-17","code":""},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"new-features-0-7-2","dir":"Changelog","previous_headings":"","what":"New features","title":"gratia 0.7.2","text":"draw.gam() draw.smooth_estimates() can now handle splines sphere (s(lat, long, bs = \"sos\")) special plotting methods using ggplot2::coord_map() handle projection spherical coordinates. orthographic projection used default, essentially arbitrary (northern hemisphere-centric) default orientation view. fitted_values() insures data (hence returned object) tibble rather common garden data frame.","code":""},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"bug-fixes-0-7-2","dir":"Changelog","previous_headings":"","what":"Bug fixes","title":"gratia 0.7.2","text":"draw.posterior_smooths() redundantly plotting duplicate data rug plot. Now unique set covariate values used drawing rug. data_sim() passing scale argument bivariate example setting (\"eg2\"). draw() methods gamm() gamm4::gamm4() fits passing arguments draw.gam(). draw.smooth_estimates() produce subtitle data continuous smooth factor smooth. Now subtitle contains name continuous variable.","code":""},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"gratia-071","dir":"Changelog","previous_headings":"","what":"gratia 0.7.1","title":"gratia 0.7.1","text":"Due issue size package source tarball, wasn’t discovered submission CRAN, 0.7.1 never released.","code":""},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"new-features-0-7-1","dir":"Changelog","previous_headings":"","what":"New features","title":"gratia 0.7.1","text":"draw.gam() draw.smooth_estimates(): {gratia} can now handle smooths 3 4 covariates plotting. smooths 3 covariates, third covariate handled ggplot2::facet_wrap() set (default n = 16) small multiples drawn, 2d surface evaluated specified value third covariate. smooths 4 covariates, ggplot2::facet_grid() used draw small multiples, default producing 4 rows 4 columns plots specific values third fourth covariates. number small multiples produced controlled new arguments n_3d (default = n_3d = 16) n_4d (default n_4d = 4, yielding n_4d * n_4d = 16 facets) respectively. affects plotting; smooth_estimates() able handle smooths number covariates . handling higher-dimensional smooths, actually drawing plots default device can slow, especially default value n = 100 (3D 4D smooths result 160,000 data points plotted). recommended reduce n smaller value: n = 50 reasonable compromise resolution speed. model_concurvity() returns concurvity measures mgcv::concurvity() estimated GAMs tidy format. synonym concrvity() also provided. draw() method provided produces bar plot heatmap concurvity values depending whether overall concurvity smooth pairwise concurvity smooth model requested. draw.gam() gains argument resid_col = \"steelblue3\" allows colour partial residuals (plotted) changed.","code":""},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"bug-fixes-0-7-1","dir":"Changelog","previous_headings":"","what":"Bug fixes","title":"gratia 0.7.1","text":"model_edf() using type argument. result ever returned default EDF type. add_constant() methods weren’t applying constant required variables. draw.gam(), draw.parametric_effects() now actually work model parametric effects. #142 Reported @Nelson-Gon parametric_effects() fail model parametric terms predict.gam() returns empty arrays passed exclude = character(0).","code":""},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"gratia-070","dir":"Changelog","previous_headings":"","what":"gratia 0.7.0","title":"gratia 0.7.0","text":"CRAN release: 2022-02-07","code":""},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"major-changes-0-7-0","dir":"Changelog","previous_headings":"","what":"Major changes","title":"gratia 0.7.0","text":"draw.gam() now uses smooth_estimates() internally consequently uses draw() method underlying plotting code. simplified code compared evaluate_smooth() methods, allow future development addition features easily evaluate_smooth() retained. Similarly, evaluate_parametric_terms() now deprecated favour parametric_effects(), also used internally draw.gam() parametric terms present model (parametric = TRUE). lot code reused differences plots result change minimal, corner cases may missed. File Issue notice something changed think shouldn’t. draw.gam() now plots 2D isotropic smooths (TPRS Duchon splines) equally-scaled x y coordinates using coord_equal(ratio = 1). Alignment plots little different now plotting models multiple smooths. See Issue #81.","code":""},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"deprecated-functions-0-7-0","dir":"Changelog","previous_headings":"Major changes","what":"Deprecated functions","title":"gratia 0.7.0","text":"version 0.7.0, following functions considered deprecated use discouraged: fderiv() soft-deprecated favour derivatives(), evaluate_smooth() soft-deprecated favour smooth_estimates(), evaluate_parametric_term() soft-deprecated favour parametric_effects(). first call one functions generate warning, pointing newer, alternative, function. safe ignore warnings, deprecated functions longer receive updates thus risk removed package future date. newer alternatives can handle types models smooths, especially case smooth_estimates().","code":""},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"new-features-0-7-0","dir":"Changelog","previous_headings":"","what":"New features","title":"gratia 0.7.0","text":"fitted_values() provides tidy wrapper around predict.gam() generating fitted values model. New covariate values can provided via argument data. credible interval fitted values returned, values can link (linear predictor) response scale. Note function returns expected values response. Hence, “fitted values” used instead “predictions” case new covariate values differentiate values case generating new response values fitted model. rootogram() draw() method produce rootograms diagnostic plots fitted models. Currently models fitted poisson(), nb(), negbin(), gaussian() families. New helper functions typical_values(), factor_combos() data_combos() quickly creating data sets producing predictions fitted models covariatess fixed come typical representative values. typical_values() new helper function return typical values covariates fitted model. returns value observation closest median numerical covariates modal level factor preserving levels factor. typical_values() useful preparing data slices scenarios fitted values estimated model required. factor_combos() extracts returns combinations levels factors found data used fit model. Unlike typical_values(), factor_combos() returns combinations factor levels observed data, just modal level. Optionally, combinations factor levels can returned, just observed data. data_combos() combines returns factor data factor_combos() plus typical values numerical covariates. useful want generate predictions model combination factor terms holding continuous covariates median values. nb_theta() new extractor function returns theta parameter fitted negative binomial GAM (families nb() negbin()). Additionally, theta() has_theta() provide additional functionality. theta() experimental function extracting additional parameters model family. has_theta() useful checking additional parameters available family model. edf() extracts effective degrees freedom (EDF) fitted model specific smooth model. Various forms EDF can extracted. model_edf() returns EDF overall model. supplied multiple models, EDFs model returned comparison. draw.gam() can now show “rug” plot bivariate smooth drawing small points high transparency smooth surface data coordinates. addition, rugs plots factor smooths now show locations covariate values specific level factor levels. better reflects data used estimate smooth, even though basis smooth set using covariate locations. draw.gam() draw.smooth_estimates() now allow aspects plot changed: fill (colour) alpha attributes credible interval, line colour smooth can now specified using arguments ci_col, ci_alpha, smooth_col respectively. Partial residuals can now plotted factor smooths. allow , partial residuals filtered residuals associated particular level’s smooth drawn plot smooth. smooth_estimates() uses check_user_select_smooths() handle user-specified selection smooth terms. flexible previously, allows easier selection smooths evaluate. fixef() now imported (re-exported) nlme package, methods models fitted gam() gamm(), extract fixed effects estimates fitted models. fixed_effects() alias fixef(). draw() method smooth_samples() can now handle 2D smooths. Additionally, number posterior draws plot can now specified plotting using new argument n_samples, result n_samples draws selected random set draws plotting. New argument seed allows selection draws repeatable.","code":""},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"bug-fixes-0-7-0","dir":"Changelog","previous_headings":"","what":"Bug fixes","title":"gratia 0.7.0","text":"smooth_estimates() filtering user-supplied data level specific smooth used factor smooths. result smooth evaluated rows user-supplied data, therefore result nrow(user_data) * nlevels(by_variable) rows returned object instead nrow(user_data) rows. add_confint() method smooth_estimates() upper lower intervals reversed. #107 Reported @Aariq draw.gam() smooth_estimates() ignoring dist argument allows covariate values lie far support data excluded returning estimated values smooth plotting . #111 Reported @Aariq smooth_samples() factor GAM return samples first factor level . Reported @rroyaute discussion #121 smooth_samples() fail model contained random effect “smooths”. now ignored message running smooth_samples(). Reported @isabellaghement #121 link(), inv_link() failing models fitted family = scat(). Reported @Aariq #130","code":""},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"gratia-060","dir":"Changelog","previous_headings":"","what":"gratia 0.6.0","title":"gratia 0.6.0","text":"CRAN release: 2021-04-18","code":""},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"major-changes-0-6-0","dir":"Changelog","previous_headings":"","what":"Major changes","title":"gratia 0.6.0","text":"{cowplot} package replaced {patchwork} package producing multi-panel figures draw() appraise(). shouldn’t affect code used {gratia} , passed additional arguments cowplot::plot_grid() used align axis arguments draw() appraise(), ’ll need adapt code accordingly. Typically, can simply delete align axis arguments {patchwork} just work align plots nicely. arguments passed via ... cowplot::plot_grid() just ignored patchwork::wrap_plots() unless passed arguments match arguments patchwork::wrap_plots().","code":""},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"new-features-0-6-0","dir":"Changelog","previous_headings":"","what":"New features","title":"gratia 0.6.0","text":"{patchwork} package now used multi-panel figures. , {gratia} longer Imports {cowplot} package. Worm plot diagnostic plots available via new function worm_plot(). Worm plots detrended Q-Q plots, deviation Q-Q reference line emphasized deviations around line occupy full height plot. worm_plot() methods available models classes \"gam\", \"glm\", \"lm\". (#62) Smooths can now compared across models using compare_smooths(), comparisons visualised associated draw() method. (#85 @dill) feature bit experimental; returned object uses nested lists may change future users find confusing. reference line qq_plot() method = \"normal\" previously drawn line intercept 0 slope 1, match methods. inconsistent stats::qqplot() drew line 1st 3rd quartiles. qq_plot() method = \"normal\" now uses robust reference line. Reference lines methods remain drawn slope 1 intercept 0. qq_plot() method = \"normal\" now draws point-wise reference band using standard error order statistic. draw() method penalty() now plots penalty matrix heatmaps -logical orientation, match matrices might written printed R console. link(), inv_link() now work models fitted gumbls() shash() families. (#84) extract_link() lower level utility function related link() inv_link(), now exported.","code":""},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"user-visible-changes-0-6-0","dir":"Changelog","previous_headings":"","what":"User visible changes","title":"gratia 0.6.0","text":"default method name generating reference quantiles qq_plot() changed \"direct\" \"uniform\", avoid confusion mgcv::qq.gam() help page description methods. Accordingly using method = \"direct\" deprecated message effect displayed used. way smooths/terms selected derivatives() switched use mechanism draw.gam()’s select argument. get partial match term, now need also specify partial_match = TRUE call derivatives().","code":""},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"bug-fixes-0-6-0","dir":"Changelog","previous_headings":"","what":"Bug fixes","title":"gratia 0.6.0","text":"transform_fun() copy paste bug definition generic. (#96 @Aariq) derivatives() user-supplied newdata fail factor smooths interval = \"simultaneous\" introduce rows derivative == 0 interval = \"confidence\" didn’t subset rows newdata specific level factor computing derivatives. (#102 @sambweber) evaluate_smooth() can now handle random effect smooths defined using ordered factor. (#99 @StefanoMezzini)","code":""},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"gratia-051","dir":"Changelog","previous_headings":"","what":"gratia 0.5.1","title":"gratia 0.5.1","text":"CRAN release: 2021-01-23","code":""},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"new-features-0-5-1","dir":"Changelog","previous_headings":"","what":"New features","title":"gratia 0.5.1","text":"smooth_estimates() can now handle bivariate multivariate thinplate regression spline smooths, e.g.  s(x, z, ), tensor product smooths (te(), t2(), & ti()), e.g. te(x, z, ) factor smooth interactions, e.g. s(x, f, bs = \"fs\") random effect smooths, e.g. s(f, bs = \"re\") penalty() provides tidy representation penalty matrices smooths. tidy representation suitable plotting ggplot(). draw() method provided, represents penalty matrix heatmap.","code":""},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"user-visible-changes-0-5-1","dir":"Changelog","previous_headings":"","what":"User visible changes","title":"gratia 0.5.1","text":"newdata argument smooth_estimates() changed data originally intended.","code":""},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"gratia-050","dir":"Changelog","previous_headings":"","what":"gratia 0.5.0","title":"gratia 0.5.0","text":"CRAN release: 2021-01-10","code":""},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"new-features-0-5-0","dir":"Changelog","previous_headings":"","what":"New features","title":"gratia 0.5.0","text":"Partial residuals models can computed partial_residuals(). partial residuals weighted residuals model added contribution smooth term (returned predict(model,   type = \"terms\"). Wish #76 (@noamross) Also, new function add_partial_residuals() can used add partial residuals data frames. Users can now control extent colour fill scales used plotting smooths draw() methods use . useful change fill scale plotting 2D smooths, change discrete colour scale used plotting random factor smooths (bs = \"fs\"). user can pass scales via arguments discrete_colour continuous_fill. effects certain smooths can excluded data simulated model using simulate.gam() predicted_samples() passing exclude terms predict.gam(). allows excluding random effects, example, model predicted values used simulate new data conditional distribution. See example predicted_samples(). Wish #74 (@hgoldspiel) draw.gam() related functions gain arguments constant fun allow user-defined constants transformations smooth estimates confidence intervals applied. Part wish Wish #79. confint.gam() now works 2D smooths also. smooth_estimates() early version code replace (likely supersede) evaluate_smooth(). smooth_estimates() can currently handle 1D smooths standard types.","code":""},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"user-visible-changes-0-5-0","dir":"Changelog","previous_headings":"","what":"User visible changes","title":"gratia 0.5.0","text":"meaning parm confint.gam changed. argument now requires smooth label match smooth. vector labels can provided, partial matching smooth label works single parm value. default behaviour remains unchanged however; parm NULL smooths evaluated returned confidence intervals. data_class() longer exported; ever intended internal function.","code":""},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"bug-fixes-0-5-0","dir":"Changelog","previous_headings":"","what":"Bug Fixes","title":"gratia 0.5.0","text":"confint.gam() failing tensor product smooth due matching issues. Reported @tamas-ferenci #88 also fixes #80 related issue selecting specific smooth. vdiffr package now used conditionally package tests. Reported Brian Ripley #93","code":""},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"gratia-041","dir":"Changelog","previous_headings":"","what":"gratia 0.4.1","title":"gratia 0.4.1","text":"CRAN release: 2020-05-30","code":""},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"user-visible-changes-0-4-1","dir":"Changelog","previous_headings":"","what":"User visible changes","title":"gratia 0.4.1","text":"draw.gam() scales = \"fixed\" now applies terms can plotted, including 2d smooths. Reported @StefanoMezzini #73","code":""},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"bug-fixes-0-4-1","dir":"Changelog","previous_headings":"","what":"Bug fixes","title":"gratia 0.4.1","text":"dplyr::combine() deprecated. Switch vctrs::vec_c(). draw.gam() scales = \"fixed\" wasn’t using fixed scales 2d smooths model. Reported @StefanoMezzini #73","code":""},{"path":[]},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"new-features-0-4-0","dir":"Changelog","previous_headings":"","what":"New features","title":"gratia 0.4.0","text":"draw.gam() can include partial residuals drawing univariate smooths. Use residuals = TRUE add partial residuals univariate smooth drawn. feature available smooths one variable, smooths, factor-smooth interactions (bs = \"fs\"). coverage credible confidence intervals drawn draw.gam() can specified via argument ci_level. default arbitrarily 0.95 reason (rough) compatibility plot.gam(). change effect making intervals slightly narrower previous versions gratia; intervals drawn ± 2 × standard error. default intervals now drawn ± ~1.96 × standard error. New function difference_smooths() computing differences factor smooth interactions. Methods available gam(), bam(), gamm() gamm4::gamm4(). Also draw() method, can handle differences 1D 2D smooths currently (handling 3D 4D smooths planned). New functions add_fitted() add_residuals() add fitted values (expectations) model residuals existing data frame. Currently methods available objects fitted gam() bam(). data_sim() tidy reimplementation mgcv::gamSim() added ability use sampling distributions Gaussian models implemented. Currently Gaussian, Poisson, Bernoulli sampling distributions available. smooth_samples() can handle continuous variable smooths varying coefficient models. link() inv_link() now work families available mgcv, including location, scale, shape families, specialised families described ?mgcv::family.mgcv. evaluate_smooth(), data_slice(), family(), link(), inv_link() methods models fitted using gamm4() gamm4 package. data_slice() can generate data 1-d slice (single variable varying). colour points, reference lines, simulation band appraise() can now specified via arguments point_col, point_alpha, ci_col ci_alpha line_col passed qq_plot(), observed_fitted_plot(), residuals_linpred_plot(), residuals_hist_plot(), also now take new arguments applicable. Added utility functions is_factor_term() term_variables() working models. is_factor_term() identifies named term factor using information terms() object fitted model. term_variables() returns character vector variable names involved model term. strictly working parametric terms models. appraise() now works models fitted glm() lm(), underlying functions calls, especially qq_plot. appraise() also works models fitted family gaulss(). location scale models models fitted extended family functions supported upcoming releases.","code":""},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"user-visible-changes-0-4-0","dir":"Changelog","previous_headings":"","what":"User visible changes","title":"gratia 0.4.0","text":"datagen() now internal function longer exported. Use data_slice() instead. evaluate_parametric_term() now much stricter can evaluate main effect terms, .e. whose order, stored terms object model 1.","code":""},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"bug-fixes-0-4-0","dir":"Changelog","previous_headings":"","what":"Bug fixes","title":"gratia 0.4.0","text":"draw() method derivatives() getting x-axis label factor smooths correctly, instead using NA second subsequent levels factor. datagen() method class \"gam\" couldn’t possibly worked anything simplest models fail even simple factor smooths. issues fixed, behaviour datagen() changed, function now intended use users. Fixed issue models terms form factor1:factor2 incorrectly identified numeric parametric terms. #68","code":""},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"gratia-031","dir":"Changelog","previous_headings":"","what":"gratia 0.3.1","title":"gratia 0.3.1","text":"CRAN release: 2020-03-29","code":""},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"new-features-0-3-1","dir":"Changelog","previous_headings":"","what":"New features","title":"gratia 0.3.1","text":"New functions link() inv_link() access link function inverse fitted models family functions. Methods classes: \"glm\", \"gam\", \"bam\", \"gamm\" currently. #58 Adds explicit family() methods objects classes \"gam\", \"bam\", \"gamm\". derivatives() now handles non-numeric creating shifted data finite differences. Fixes problem stringsAsFactors = FALSE default R-devel. #64","code":""},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"bug-fixes-0-3-1","dir":"Changelog","previous_headings":"","what":"Bug fixes","title":"gratia 0.3.1","text":"Updated gratia work tibble versions >= 3.0","code":""},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"gratia-030","dir":"Changelog","previous_headings":"","what":"gratia 0.3.0","title":"gratia 0.3.0","text":"CRAN release: 2020-01-19","code":""},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"new-features-0-3-0","dir":"Changelog","previous_headings":"","what":"New features","title":"gratia 0.3.0","text":"gratia now uses mvnfast package random draws multivariate normal distribution (mvnfast::rmvn()). Contributed Henrik Singmann #28 New function basis() generating tidy representations basis expansions mgcv-like definition smooth, e.g. s(), te(), ti(), t2(). basic smooth types also simple draw() method plotting basis. basis() simple wrapper around mgcv::smoothCon() post processing basis model matrix tidy format. #42 New function smooth_samples() draw samples entire smooth functions posterior distribution. Also draw() method plotting posterior samples.","code":""},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"bug-fixes-0-3-0","dir":"Changelog","previous_headings":"","what":"Bug fixes","title":"gratia 0.3.0","text":"draw.gam() produce empty plots panels parametric terms 2 parametric terms model. Reported @sklayn #39. derivatives() now works factor smooths, including ordered factor smooths. function also now works correctly complex models multiple covariates/smooths. #47 derivatives() also now handles 'fs' smooths. Reported @tomand-uio #57. evaluate_parametric_term() hence draw.gam() fail ziplss() model ) gratia didn’t handle parametric terms models multiple linear predictors correctly, ii) gratia didn’t convert naming convention mgcv terms higher linear predictors. Reported @pboesu #45","code":""}]
+[{"path":[]},{"path":"https://gavinsimpson.github.io/gratia/CODE_OF_CONDUCT.html","id":"our-pledge","dir":"","previous_headings":"","what":"Our Pledge","title":"Contributor Covenant Code of Conduct","text":"members, contributors, leaders pledge make participation community harassment-free experience everyone, regardless age, body size, visible invisible disability, ethnicity, sex characteristics, gender identity expression, level experience, education, socio-economic status, nationality, personal appearance, race, religion, sexual identity orientation. pledge act interact ways contribute open, welcoming, diverse, inclusive, healthy community.","code":""},{"path":"https://gavinsimpson.github.io/gratia/CODE_OF_CONDUCT.html","id":"our-standards","dir":"","previous_headings":"","what":"Our Standards","title":"Contributor Covenant Code of Conduct","text":"Examples behavior contributes positive environment community include: Demonstrating empathy kindness toward people respectful differing opinions, viewpoints, experiences Giving gracefully accepting constructive feedback Accepting responsibility apologizing affected mistakes, learning experience Focusing best just us individuals, overall community Examples unacceptable behavior include: use sexualized language imagery, sexual attention advances kind Trolling, insulting derogatory comments, personal political attacks Public private harassment Publishing others’ private information, physical email address, without explicit permission conduct reasonably considered inappropriate professional setting","code":""},{"path":"https://gavinsimpson.github.io/gratia/CODE_OF_CONDUCT.html","id":"enforcement-responsibilities","dir":"","previous_headings":"","what":"Enforcement Responsibilities","title":"Contributor Covenant Code of Conduct","text":"Community leaders responsible clarifying enforcing standards acceptable behavior take appropriate fair corrective action response behavior deem inappropriate, threatening, offensive, harmful. Community leaders right responsibility remove, edit, reject comments, commits, code, wiki edits, issues, contributions aligned Code Conduct, communicate reasons moderation decisions appropriate.","code":""},{"path":"https://gavinsimpson.github.io/gratia/CODE_OF_CONDUCT.html","id":"scope","dir":"","previous_headings":"","what":"Scope","title":"Contributor Covenant Code of Conduct","text":"Code Conduct applies within community spaces, also applies individual officially representing community public spaces. Examples representing community include using official e-mail address, posting via official social media account, acting appointed representative online offline event.","code":""},{"path":"https://gavinsimpson.github.io/gratia/CODE_OF_CONDUCT.html","id":"enforcement","dir":"","previous_headings":"","what":"Enforcement","title":"Contributor Covenant Code of Conduct","text":"Instances abusive, harassing, otherwise unacceptable behavior may reported package maintainer, Gavin Simpson (see email address CRAN package page) . complaints reviewed investigated promptly fairly. community leaders obligated respect privacy security reporter incident.","code":""},{"path":"https://gavinsimpson.github.io/gratia/CODE_OF_CONDUCT.html","id":"enforcement-guidelines","dir":"","previous_headings":"","what":"Enforcement Guidelines","title":"Contributor Covenant Code of Conduct","text":"Community leaders follow Community Impact Guidelines determining consequences action deem violation Code Conduct:","code":""},{"path":"https://gavinsimpson.github.io/gratia/CODE_OF_CONDUCT.html","id":"id_1-correction","dir":"","previous_headings":"Enforcement Guidelines","what":"1. Correction","title":"Contributor Covenant Code of Conduct","text":"Community Impact: Use inappropriate language behavior deemed unprofessional unwelcome community. Consequence: private, written warning community leaders, providing clarity around nature violation explanation behavior inappropriate. public apology may requested.","code":""},{"path":"https://gavinsimpson.github.io/gratia/CODE_OF_CONDUCT.html","id":"id_2-warning","dir":"","previous_headings":"Enforcement Guidelines","what":"2. Warning","title":"Contributor Covenant Code of Conduct","text":"Community Impact: violation single incident series actions. Consequence: warning consequences continued behavior. interaction people involved, including unsolicited interaction enforcing Code Conduct, specified period time. includes avoiding interactions community spaces well external channels like social media. Violating terms may lead temporary permanent ban.","code":""},{"path":"https://gavinsimpson.github.io/gratia/CODE_OF_CONDUCT.html","id":"id_3-temporary-ban","dir":"","previous_headings":"Enforcement Guidelines","what":"3. Temporary Ban","title":"Contributor Covenant Code of Conduct","text":"Community Impact: serious violation community standards, including sustained inappropriate behavior. Consequence: temporary ban sort interaction public communication community specified period time. public private interaction people involved, including unsolicited interaction enforcing Code Conduct, allowed period. Violating terms may lead permanent ban.","code":""},{"path":"https://gavinsimpson.github.io/gratia/CODE_OF_CONDUCT.html","id":"id_4-permanent-ban","dir":"","previous_headings":"Enforcement Guidelines","what":"4. Permanent Ban","title":"Contributor Covenant Code of Conduct","text":"Community Impact: Demonstrating pattern violation community standards, including sustained inappropriate behavior, harassment individual, aggression toward disparagement classes individuals. Consequence: permanent ban sort public interaction within community.","code":""},{"path":"https://gavinsimpson.github.io/gratia/CODE_OF_CONDUCT.html","id":"attribution","dir":"","previous_headings":"","what":"Attribution","title":"Contributor Covenant Code of Conduct","text":"Code Conduct adapted Contributor Covenant, version 2.0, available https://www.contributor-covenant.org/version/2/0/code_of_conduct.html. Community Impact Guidelines inspired Mozilla’s code conduct enforcement ladder. answers common questions code conduct, see FAQ https://www.contributor-covenant.org/faq. Translations available https://www.contributor-covenant.org/translations.","code":""},{"path":[]},{"path":"https://gavinsimpson.github.io/gratia/CONTRIBUTING.html","id":"something-isnt-working-right-or-generating-an-error","dir":"","previous_headings":"","what":"Something isn’t working right or generating an error","title":"Contributing","text":"something isn’t working, either might expect/want contrary documentation, probably bug missing feature. ’re getting error running {gratia}, ’s also likely bug, opportunity catch use-case wasn’t expecting. First, check issue hasn’t already fixed -development version Github. Install current development version {gratia} R-universe using issue remains, please file issue via Issues page. ’s OK report issue even ’re sure ’s problem, problem better described question, use Discussions page (see ). Feature requests welcome! problem {gratia} hit top TODO list quickly cen provide reproducible example demonstrating problem.","code":"install.packages(\"gratia\", repos = c(   \"https://gavinsimpson.r-universe.dev\",   \"https://cloud.r-project.org\" ))"},{"path":"https://gavinsimpson.github.io/gratia/CONTRIBUTING.html","id":"got-a-question-want-to-show-how-to-do-something-with-gratia","dir":"","previous_headings":"","what":"Got a question? Want to show how to do something with {gratia}?","title":"Contributing","text":"issue best described question, want know something {gratia}, cool example using {gratia} want share , please consider using Discussions page. ’re sure, can always ask issue Question (Use Q&category) , really bug, can easily create Issue discussion.","code":""},{"path":"https://gavinsimpson.github.io/gratia/CONTRIBUTING.html","id":"code-contributions","dir":"","previous_headings":"","what":"Code contributions","title":"Contributing","text":"Code contributions form Pull Requests always appreciated. suitable workflow: Fork repo Github account Clone version account machine account, e.g,. git clone https://github.com/<yourgithubusername>/gratia.git Make sure track progress upstream (.e., version gratis gavinsimpson/gratia) git remote add upstream https://github.com/gavinsimpson/gratia.git. making changes make sure pull changes upstream either git fetch upstream merge later git pull upstream fetch merge one step Make changes (bonus points making changes new feature branch) Push account Submit pull request home base gavinsimpson/gratia","code":""},{"path":"https://gavinsimpson.github.io/gratia/CONTRIBUTING.html","id":"development-tools--paradigm--ethos","dir":"","previous_headings":"Code contributions","what":"Development tools / paradigm / ethos","title":"Contributing","text":"Please note following contributing code: {gratia} tightly aligned tidyverse use {dplyr} related packages lot internally developing package. plan replacing code lower-level code using {vctrs}, right now development focus filling functionality package premature optimisation. original aim {gratia} provide {ggplot2} plotting smooths; please stick principle use plotting paradigm. aim {mgcv}-feature complete; {mgcv} can something terms plotting smooths, handling specialists smooths, etc, principle {gratia} support . aim general compatibility {mgcv}; {gratia} deviates {mgcv} things, needs good justification; example {gratia} deviates multivariate isotropic smooths fitted s() plotted, Dave Miller (@dill) argued convincingly way. Don’t add dependencies! Unless accompanied strong justification, want reduce dependencies increase number. ’m using {styler} style code, using 2 spaces indent. code written change, however. Respect 80 character line length limit. contribute functionality fix bug, please add test using {testthat} framework insure new things works correctly bug stays fixed. contributions must result new NOTES, WARNINGS, ERRORS running R CMD check ---cran; please check contributions Winbuilder example.","code":""},{"path":"https://gavinsimpson.github.io/gratia/CONTRIBUTING.html","id":"email","dir":"","previous_headings":"","what":"Email","title":"Contributing","text":"hate email! can email GMail address — ’s like can stop :-) — unless capture attention immediately label message {gratia}-related, quickly get swamped never seen . Even label , ’s guarantee ever get round replying; usually happens ’ll forget don’t remember check gratia label often. end result email , get response , tardy. infinitely better use Discussions Issues pages Github ask questions package report problems. Please don’t email work (academic) address (email address might find ) {gratia}. question GAMs, much better ask CrossValidated StackOverflow depending whether question statistical programming related. relates {gratia} can use Discussions page. Asking question public allows others reply public, contributes body knowledge easily available others. circumstances send email multiple addresses; quickest way get message trash.","code":""},{"path":"https://gavinsimpson.github.io/gratia/LICENSE.html","id":null,"dir":"","previous_headings":"","what":"The MIT License (MIT)","title":"The MIT License (MIT)","text":"Copyright (c) 2013-2024 Gavin L. Simpson Permission hereby granted, free charge, person obtaining copy software associated documentation files (“Software”), deal Software without restriction, including without limitation rights use, copy, modify, merge, publish, distribute, sublicense, /sell copies Software, permit persons Software furnished , subject following conditions: copyright notice permission notice shall included copies substantial portions Software. SOFTWARE PROVIDED “”, WITHOUT WARRANTY KIND, EXPRESS IMPLIED, INCLUDING LIMITED WARRANTIES MERCHANTABILITY, FITNESS PARTICULAR PURPOSE NONINFRINGEMENT. EVENT SHALL AUTHORS COPYRIGHT HOLDERS LIABLE CLAIM, DAMAGES LIABILITY, WHETHER ACTION CONTRACT, TORT OTHERWISE, ARISING , CONNECTION SOFTWARE USE DEALINGS SOFTWARE.","code":""},{"path":"https://gavinsimpson.github.io/gratia/articles/custom-plotting.html","id":"background","dir":"Articles","previous_headings":"","what":"Background","title":"Customizing plots","text":"draw() function {gratia} envisaged ggplot-based alternative mgcv:::plot.gam(). , never intended allow sorts customization possible ggplot() packages use ggplot() plotting layer. largely due decision produce multiple separate ggplot() plots GAMs multiple smooths, subsequently combined single figure device, initially using {cowplot} recently {patchwork}. things way evident consider might represent smooths 3 4 variables (common might think; consider space-time models via te(x, y, time, d = c(2,1)) space-depth-time models [think ocean atmospheric data space depth (height), observed time] via te(x, y, depth, time, d = c(2, 1, 1))), require facets top produce small multiples, means can’t use facets plot separate smooths. Additional complications arise consider complex smooth types, splines sphere, might want us different coordinate systems geoms best represent underlying smooth. gone root combining multiple ggplot objects single figure, problem customizing plots quickly rears head. vignette presents solutions problem modifying adding plots produced draw() culminating example illustrating use {gratia}’s utility functions produce plots lower-lever components.","code":""},{"path":"https://gavinsimpson.github.io/gratia/articles/custom-plotting.html","id":"adding-layers-to-plots-with-the-operator","dir":"Articles","previous_headings":"","what":"Adding layers to plots with the & operator","title":"Customizing plots","text":"start simulating data fitting GAM four smooth functions default plot produced draw()  want change theme plots, can’t append theme() layer p affects last plot patchwork1  One way apply theme plots patchwork & operator.","code":"library(\"gratia\") library(\"mgcv\") #> Loading required package: nlme #> This is mgcv 1.9-1. For overview type 'help(\"mgcv-package\")'. library(\"ggplot2\") library(\"dplyr\") #>  #> Attaching package: 'dplyr' #> The following object is masked from 'package:nlme': #>  #>     collapse #> The following objects are masked from 'package:stats': #>  #>     filter, lag #> The following objects are masked from 'package:base': #>  #>     intersect, setdiff, setequal, union library(\"patchwork\")  # simulate data n <- 400 eg1 <- data_sim(\"eg1\", n = n, seed = 1)  # fit model m <- gam(y ~ s(x0) + s(x1) + s(x2) + s(x3),   data = eg1, method = \"REML\") p <- draw(m) p p + theme_bw() p & theme_bw()"},{"path":"https://gavinsimpson.github.io/gratia/articles/custom-plotting.html","id":"combining-individual-plots-produced-by-draw","dir":"Articles","previous_headings":"","what":"Combining individual plots produced by draw()","title":"Customizing plots","text":"draw() methods like draw.gam() return object created patchwork::wrap_plots(), result isn’t straightforward combine objects new patchwork avoid error, need use patchwork::plot_layout() set dimensions want  achieved directly via draw()  instructive know combine outputs draw() need arise, want create patchwork plots different models.","code":"p1 <- draw(m, select = \"s(x0)\") p2 <- draw(m, select = \"s(x1)\") p3 <- draw(m, select = \"s(x2)\") p1 + p2 + p3 #> Error in `wrap_dims()`: #> ! Need 3 panels, but together `nrow` and `ncol` only provide 1. #> ℹ Please increase `ncol` and/or `nrow`. p1 + p2 + p3 + plot_layout(ncol = 3) draw(m, select = c(\"s(x0)\", \"s(x1)\", \"s(x2)\"), ncol = 3)"},{"path":"https://gavinsimpson.github.io/gratia/articles/custom-plotting.html","id":"building-your-own-plot-by-hand","dir":"Articles","previous_headings":"","what":"Building your own plot by hand","title":"Customizing plots","text":"{gratia} provides high-level functions like draw() get good graphical overview fitted model, little option customisation — isn’t possible desirable allow possible customisation options fatures {ggplot2} single function. Think many arguments require! Instead, {gratia} also exports lower-level functions used draw() can create plot using whatever {ggplot2} functions make sense. next code blocks ’ll see plot created draw(m) can recreated hand using lower-level building blocks. main thing need evaluate smooths values covariates. done using smooth_estimates(). also need add credible interval evaluations, can done tidyverse-style via add_confint() default draw.gam() add partial residuals partial effects plots. achieve effect, need add partial residuals data used fit model. can done via add_partial_residuals() will2 add columns names \"s(x0)\". \"s(x1)\", etc. data. Now everything need recreate plots created draw.gam(). code block filter sm focus specific smooth, f(x2)f(x2) (\"s(x2)\"), add rug plot observed values x2, credible interval around estimated smooth, partial residuals point layer, estimated smooth line layer, annotation  Assuming repeat steps smooths, creating plot objects p_sx0, p_sx1, p_sx2, p_sx3 (code shown), can complete plot creating patchwork desired number rows columns  real benefit complete control data plotted can use power {ggplot2} map additional variables plot aesthetics. example, let’s assume factor variable original data want colour partial residuals according levels factor. Let’s create factor Now can modify plotting code map fac colour aesthetic plot partial residuals. save typing, ’ll reorder layers plot add partial residuals last  can also simple model checking plotting smooth partial residuals coloured according one covariates (also plotting actual residuals covariates). code chunk , map covariate x1 colour size aesthetics (note deleted cex = 1.5 allow mapping size)  resulting plot doesn’t show particular problems model way data simulated, hopefully illustrates can possible use low-level functions provided {gratia}.","code":"# evaluate the smooths sm <- smooth_estimates(m) |>   add_confint() sm #> # A tibble: 400 × 11 #>    .smooth .type .by   .estimate   .se     x0    x1    x2    x3 .lower_ci #>    <chr>   <chr> <chr>     <dbl> <dbl>  <dbl> <dbl> <dbl> <dbl>     <dbl> #>  1 s(x0)   TPRS  NA       -0.929 0.422 0.0131    NA    NA    NA    -1.76  #>  2 s(x0)   TPRS  NA       -0.881 0.396 0.0230    NA    NA    NA    -1.66  #>  3 s(x0)   TPRS  NA       -0.834 0.372 0.0329    NA    NA    NA    -1.56  #>  4 s(x0)   TPRS  NA       -0.786 0.348 0.0429    NA    NA    NA    -1.47  #>  5 s(x0)   TPRS  NA       -0.738 0.326 0.0528    NA    NA    NA    -1.38  #>  6 s(x0)   TPRS  NA       -0.690 0.305 0.0627    NA    NA    NA    -1.29  #>  7 s(x0)   TPRS  NA       -0.643 0.287 0.0727    NA    NA    NA    -1.20  #>  8 s(x0)   TPRS  NA       -0.595 0.270 0.0826    NA    NA    NA    -1.12  #>  9 s(x0)   TPRS  NA       -0.548 0.255 0.0925    NA    NA    NA    -1.05  #> 10 s(x0)   TPRS  NA       -0.501 0.242 0.102     NA    NA    NA    -0.975 #> # ℹ 390 more rows #> # ℹ 1 more variable: .upper_ci <dbl> # add partial residuals to data eg1 <- eg1 |>   add_partial_residuals(m) names(eg1) #>  [1] \"y\"     \"x0\"    \"x1\"    \"x2\"    \"x3\"    \"f\"     \"f0\"    \"f1\"    \"f2\"    #> [10] \"f3\"    \"s(x0)\" \"s(x1)\" \"s(x2)\" \"s(x3)\" p_sx2 <- sm |>   filter(.smooth == \"s(x2)\") |>   ggplot() +   geom_rug(aes(x = x2),     data = eg1,     sides = \"b\", length = grid::unit(0.02, \"npc\")   ) +   geom_ribbon(aes(ymin = .lower_ci, ymax = .upper_ci, x = x2),     alpha = 0.2   ) +   geom_point(aes(x = x2, y = `s(x2)`),     data = eg1, cex = 1.5, colour = \"steelblue3\"   ) +   geom_line(aes(x = x2, y = .estimate), lwd = 1.2) +   labs(y = \"Partial effect\", title = \"s(x2)\") p_sx2 p_sx0 + p_sx1 + p_sx2 + p_sx3 + plot_layout(ncol = 2) set.seed(12) eg1 <- eg1 |>   mutate(fac = sample(letters[1:4], n(), replace = TRUE)) plt <- sm |>   filter(.smooth == \"s(x2)\") |>   ggplot() +   geom_rug(aes(x = x2),     data = eg1,     sides = \"b\", length = grid::unit(0.02, \"npc\")   ) +   geom_ribbon(aes(ymin = .lower_ci, ymax = .upper_ci, x = x2),     alpha = 0.2   ) +   geom_line(aes(x = x2, y = .estimate), lwd = 1.2) +   labs(y = \"Partial effect\", title = \"s(x2)\")  plt +   geom_point(     aes(       x = x2, y = `s(x2)`,       colour = fac     ), # <-- map fac to colour aesthetic     data = eg1, cex = 1.5   ) plt +   geom_point(     aes(       x = x2, y = `s(x2)`,       colour = x1, size = x1     ), # <-- map fac to colour aesthetic     data = eg1, alpha = 0.3   ) + # <-- deleted cex!!   scale_colour_viridis_c(option = \"plasma\")"},{"path":"https://gavinsimpson.github.io/gratia/articles/data-slices.html","id":"carbon-dioxide-uptake-in-grass-plants","dir":"Articles","previous_headings":"","what":"Carbon Dioxide Uptake in Grass Plants","title":"Data slices","text":"first example uses small data set experimental study cold tolerance grass Echinochloa crusgalli. data data frame CO2 provided {datasets} package ships R.  One way model data allow different smooths combinations treatment type covariates can look fitted smooths using draw()  might want compare model fitted values treatment types (origins), ignoring random effect component. want evaluate model range values covariate conc combinations factors. data slice covariate space, can create using data_slice(). create data slice conc Quebec type chilled treatment use Notice data_slice() filled something remaining covariates didn’t mention? case, data_slice() doesn’t know tt created, chosen modal level tt factor, correct choice case. Instead, need specify correct level explicitly tt created data slice, can predict model using combination covariate values specified slice. use predict.gam() , fitted_values() function {gratia} easier use, especially non-Gaussian models Notice excluded random effect term; even though specify something plant covariate can ignore term model using exclude argument. fitted_values() creates credible interval scale link function back-transforms response scale scale = \"response\", also default. Plotting fitted values data slice now requires simple {ggplot2} knowledge  Next, let’s compare fitted effects treatment Mississippi origin plants , replaced automatically-generated tt variable correctly specified call fct_cross(), retaining levels type treatment factors. insures correct combinations corresponding treatment type factors also preserve original levels tt covariate created. can visualise fitted values data slice  creating data slices, used helper functions specify covariate values slice. {gratia} provides several helper functions: evenly(x, n = 100) — creates n evenly spaced values range covariate, evenly(x, = 5 — creates evenly spaced values range covariate increments 5, evenly(x, ..., lower = 5, upper = 10) — either two uses evenly() shown use lower upper limits vector x. Arguments lower upper can used change one upper lower bounds. evenly(fct) — produces new factor containing level specified factor fct just , ref_level(fct) — creates new factor containing just reference level specified factor covariate fct, level(fct, \"level\") — creates factor requested \"level\" factor fct. cases involving factors, helper functions set levels factor match original model fit2. second argument data_slice() ... ... argument allows provide expressions create covariate values want data slice. Expressions passed ... evaluated within model frame fitted model (argument object) data (supplied). restricted either using helper functions provide {gratia}; R function used long makes sense context model frame, returns something can combined using tidyr::expand_grid().","code":"## data load and prep data(CO2, package = \"datasets\") plant <- CO2 |>   as_tibble() |>   rename(plant = Plant, type = Type, treatment = Treatment) |>   mutate(plant = factor(plant, ordered = FALSE)) plant_ylab <- expression(CO[2] ~ uptake ~ (mu * mol ~ m^{-3})) plant_xlab <- expression(CO[2] ~ concentration ~ (mL ~ L^{-1}))  plant |>   ggplot(aes(x = conc, y = uptake, group = plant, colour = treatment)) +   geom_point() +   geom_line() +   facet_wrap(~type) +   labs(y = plant_ylab, x = plant_xlab, colour = \"Treatment\") plant <- plant |>   mutate(tt = fct_cross(treatment, type)) m_plant <- gam(uptake ~ treatment * type + s(conc, by = tt, k = 6) +     s(plant, bs = \"re\"),   data = plant, method = \"REML\", family = Gamma(link = \"log\") ) overview(m_plant) #>  #> Generalized Additive Model with 8 terms #>  #>   term                             type              k   edf statistic p.value   #>   <chr>                            <chr>         <dbl> <dbl>     <dbl> <chr>     #> 1 treatment                        parametric       NA  1         1.59 0.2124864 #> 2 type                             parametric       NA  1        11.2  0.0014830 #> 3 treatment:type                   parametric       NA  1         7.45 0.0085489 #> 4 s(conc):ttnonchilled:Quebec      TPRS              5  4.72     69.7  < 0.001   #> 5 s(conc):ttchilled:Quebec         TPRS              5  4.71     86.5  < 0.001   #> 6 s(conc):ttnonchilled:Mississippi TPRS              5  4.62     74.1  < 0.001   #> 7 s(conc):ttchilled:Mississippi    TPRS              5  4.39     25.3  < 0.001   #> 8 s(plant)                         Random effect    12  7.40     12.8  < 0.001 draw(m_plant, residuals = TRUE, scales = \"fixed\") ds1 <- data_slice(m_plant,   conc = evenly(conc, n = 100),   type = level(type, \"Quebec\"), treatment = level(treatment, \"chilled\") ) ds1 #> # A tibble: 100 × 5 #>     conc type   treatment tt                plant #>    <dbl> <fct>  <fct>     <fct>             <fct> #>  1   95  Quebec chilled   nonchilled:Quebec Qn1   #>  2  104. Quebec chilled   nonchilled:Quebec Qn1   #>  3  113. Quebec chilled   nonchilled:Quebec Qn1   #>  4  122. Quebec chilled   nonchilled:Quebec Qn1   #>  5  132. Quebec chilled   nonchilled:Quebec Qn1   #>  6  141. Quebec chilled   nonchilled:Quebec Qn1   #>  7  150. Quebec chilled   nonchilled:Quebec Qn1   #>  8  159. Quebec chilled   nonchilled:Quebec Qn1   #>  9  168. Quebec chilled   nonchilled:Quebec Qn1   #> 10  177. Quebec chilled   nonchilled:Quebec Qn1   #> # ℹ 90 more rows ds1 <- data_slice(m_plant,   conc = evenly(conc, n = 100),   treatment = level(treatment, \"chilled\"), type = level(type, \"Quebec\"),   tt = level(tt, \"chilled:Quebec\") ) ds1 #> # A tibble: 100 × 5 #>     conc treatment type   tt             plant #>    <dbl> <fct>     <fct>  <fct>          <fct> #>  1   95  chilled   Quebec chilled:Quebec Qn1   #>  2  104. chilled   Quebec chilled:Quebec Qn1   #>  3  113. chilled   Quebec chilled:Quebec Qn1   #>  4  122. chilled   Quebec chilled:Quebec Qn1   #>  5  132. chilled   Quebec chilled:Quebec Qn1   #>  6  141. chilled   Quebec chilled:Quebec Qn1   #>  7  150. chilled   Quebec chilled:Quebec Qn1   #>  8  159. chilled   Quebec chilled:Quebec Qn1   #>  9  168. chilled   Quebec chilled:Quebec Qn1   #> 10  177. chilled   Quebec chilled:Quebec Qn1   #> # ℹ 90 more rows fv1 <- fitted_values(m_plant, data = ds1, scale = \"response\", exclude = \"s(plant)\") fv1 #> # A tibble: 100 × 10 #>     .row  conc treatment type   tt      plant .fitted    .se .lower_ci .upper_ci #>    <int> <dbl> <fct>     <fct>  <fct>   <fct>   <dbl>  <dbl>     <dbl>     <dbl> #>  1     1   95  chilled   Quebec chille… Qn1      13.0 0.0783      11.2      15.2 #>  2     2  104. chilled   Quebec chille… Qn1      14.1 0.0757      12.1      16.3 #>  3     3  113. chilled   Quebec chille… Qn1      15.2 0.0737      13.1      17.5 #>  4     4  122. chilled   Quebec chille… Qn1      16.3 0.0722      14.2      18.8 #>  5     5  132. chilled   Quebec chille… Qn1      17.6 0.0714      15.3      20.2 #>  6     6  141. chilled   Quebec chille… Qn1      18.9 0.0711      16.4      21.7 #>  7     7  150. chilled   Quebec chille… Qn1      20.2 0.0712      17.6      23.3 #>  8     8  159. chilled   Quebec chille… Qn1      21.6 0.0716      18.8      24.9 #>  9     9  168. chilled   Quebec chille… Qn1      23.0 0.0721      20.0      26.5 #> 10    10  177. chilled   Quebec chille… Qn1      24.4 0.0726      21.2      28.1 #> # ℹ 90 more rows fv1 |>   ggplot(aes(x = conc, y = .fitted)) +   geom_point(     data = plant |>       filter(type == \"Quebec\", treatment == \"chilled\"),     mapping = aes(y = uptake),     alpha = 0.8, colour = \"steelblue\"   ) +   geom_ribbon(aes(ymin = .lower_ci, ymax = .upper_ci), alpha = 0.2) +   geom_line() +   labs(     x = plant_xlab, y = plant_ylab,     title = expression(Estimated ~ CO[2] ~ uptake),     subtitle = \"Chilled plants of the Quebec type\"   ) ds2 <- data_slice(m_plant,   conc = evenly(conc, n = 100),   treatment = evenly(treatment), type = level(type, \"Mississippi\") ) |>   mutate(tt = fct_cross(treatment, type, keep_empty = TRUE)) ds2 #> # A tibble: 200 × 5 #>     conc treatment  type        tt                     plant #>    <dbl> <fct>      <fct>       <fct>                  <fct> #>  1   95  nonchilled Mississippi nonchilled:Mississippi Qn1   #>  2   95  chilled    Mississippi chilled:Mississippi    Qn1   #>  3  104. nonchilled Mississippi nonchilled:Mississippi Qn1   #>  4  104. chilled    Mississippi chilled:Mississippi    Qn1   #>  5  113. nonchilled Mississippi nonchilled:Mississippi Qn1   #>  6  113. chilled    Mississippi chilled:Mississippi    Qn1   #>  7  122. nonchilled Mississippi nonchilled:Mississippi Qn1   #>  8  122. chilled    Mississippi chilled:Mississippi    Qn1   #>  9  132. nonchilled Mississippi nonchilled:Mississippi Qn1   #> 10  132. chilled    Mississippi chilled:Mississippi    Qn1   #> # ℹ 190 more rows fitted_values(m_plant,   data = ds2, scale = \"response\",   exclude = \"s(plant)\" ) |>   ggplot(aes(x = conc, y = .fitted, group = treatment)) +   geom_point(     data = plant |> filter(type == \"Mississippi\"),     mapping = aes(y = uptake, colour = treatment),     alpha = 0.8   ) +   geom_ribbon(aes(ymin = .lower_ci, ymax = .upper_ci, fill = treatment),     alpha = 0.2   ) +   geom_line(aes(colour = treatment)) +   labs(     x = plant_xlab, y = plant_ylab,     title = expression(Estimated ~ CO[2] ~ uptake),     subtitle = \"Comparison of treatment in plants of the Mississippi type\",     colour = \"Treatment\", fill = \"Treatment\"   ) args(gratia:::data_slice.gam) #> function (object, ..., data = NULL, envir = NULL)  #> NULL"},{"path":"https://gavinsimpson.github.io/gratia/articles/data-slices.html","id":"slices-through-a-2d-smooth","dir":"Articles","previous_headings":"","what":"Slices through a 2D smooth","title":"Data slices","text":"second example, ’ll use bivariate example data set {mgcv} fit tensor product covariates x z aim example create univariate data slice 2D smooth user-specified values x holding z one fixed values. visualise effect smooth level, using smooth_estimates(), response level, using fitted_values().","code":"# simulate data from the bivariate surface df <- data_sim(\"eg2\", n = 1000, scale = 0.25, seed = 2)  # fit the GAM m_biv <- gam(y ~ te(x, z), data = df, method = \"REML\")"},{"path":"https://gavinsimpson.github.io/gratia/articles/data-slices.html","id":"using-smooth_estimates","dir":"Articles","previous_headings":"Slices through a 2D smooth","what":"Using smooth_estimates()","title":"Data slices","text":"begin creating slice data space. also create label point nice axis label. evaluate smooth desired values add confidence interval can plot sm using {ggplot2}  Note interval Marra Wood (2012) interval. doesn’t include uncertainty model constant term moment, unless smooth close linear shouldn’t make much difference. extends multiple slices asking several discrete z","code":"ds3 <- data_slice(m_biv,   x = evenly(x, n = 100),   z = quantile(z, probs = 0.25) )  z_val <- with(ds3, round(quantile(z, probs = 0.25), 2)) ylab <- bquote(hat(f)(x, .(z_val))) sm <- smooth_estimates(m_biv, select = \"te(x,z)\", data = ds3) |>   add_confint() sm #> # A tibble: 100 × 9 #>    .smooth .type        .by   .estimate    .se       x     z .lower_ci .upper_ci #>    <chr>   <chr>        <chr>     <dbl>  <dbl>   <dbl> <dbl>     <dbl>     <dbl> #>  1 te(x,z) Tensor prod… NA        0.103 0.0583 6.63e-4 0.245   -0.0107     0.218 #>  2 te(x,z) Tensor prod… NA        0.122 0.0548 1.08e-2 0.245    0.0148     0.230 #>  3 te(x,z) Tensor prod… NA        0.141 0.0514 2.08e-2 0.245    0.0400     0.242 #>  4 te(x,z) Tensor prod… NA        0.159 0.0482 3.09e-2 0.245    0.0648     0.254 #>  5 te(x,z) Tensor prod… NA        0.177 0.0451 4.10e-2 0.245    0.0890     0.266 #>  6 te(x,z) Tensor prod… NA        0.195 0.0422 5.11e-2 0.245    0.113      0.278 #>  7 te(x,z) Tensor prod… NA        0.213 0.0396 6.12e-2 0.245    0.135      0.291 #>  8 te(x,z) Tensor prod… NA        0.230 0.0372 7.13e-2 0.245    0.157      0.303 #>  9 te(x,z) Tensor prod… NA        0.247 0.0351 8.14e-2 0.245    0.178      0.316 #> 10 te(x,z) Tensor prod… NA        0.263 0.0333 9.14e-2 0.245    0.198      0.328 #> # ℹ 90 more rows sm |>   ggplot(aes(x = x, y = .estimate)) +   geom_ribbon(aes(ymin = .lower_ci, ymax = .upper_ci), alpha = 0.2) +   geom_line() +   labs(     title = \"Evaluation of smooth te(x,z) at fixed z\",     y = ylab   ) ds4 <- data_slice(m_biv,   x = evenly(x, n = 100),   z = round(quantile(z, probs = c(0.25, 0.5, 0.75)), 2) )  sm <- smooth_estimates(m_biv, select = \"te(x,z)\", data = ds4) |>   add_confint() |>   mutate(fz = factor(z))  sm |>   ggplot(aes(x = x, y = .estimate, colour = fz, group = fz)) +   geom_ribbon(aes(ymin = .lower_ci, ymax = .upper_ci, fill = fz, colour = NULL),     alpha = 0.2   ) +   geom_line() +   labs(     title = \"Evaluation of smooth te(x,z) at fixed z\",     y = expression(hat(f)(x, z)), colour = \"z\", fill = \"z\"   )"},{"path":"https://gavinsimpson.github.io/gratia/articles/data-slices.html","id":"using-fitted_samples","dir":"Articles","previous_headings":"Slices through a 2D smooth","what":"Using fitted_samples()","title":"Data slices","text":"want evaluate surface x fixed z conditional upon values covariates (model predicted fitted values) fitted_samples() tidy wrapper predict.gam(). single z  multiple z difference now model constant included well uncertainty.","code":"fitted_values(m_biv, data = ds3) |> # default is response scale, not link   ggplot(aes(x = x, y = .fitted)) +   geom_ribbon(aes(ymin = .lower_ci, ymax = .upper_ci), alpha = 0.2) +   geom_line() +   labs(     title = \"Fitted values from model\",     y = expression(hat(y))   ) fitted_values(m_biv, data = ds4) |>   mutate(fz = factor(z)) |>   ggplot(aes(x = x, y = .fitted, colour = fz, group = fz)) +   geom_ribbon(aes(ymin = .lower_ci, ymax = .upper_ci, fill = fz, colour = NULL),     alpha = 0.2   ) +   geom_line() +   labs(     title = \"Fitted values from model\",     y = expression(hat(y)), colour = \"z\", fill = \"z\"   )"},{"path":"https://gavinsimpson.github.io/gratia/articles/gratia.html","id":"plotting","dir":"Articles","previous_headings":"","what":"Plotting","title":"Getting started with gratia","text":"gratia provides draw() function produce plots using ggplot2 📦. plot estimated smooths GAM fitted , use  plots produced partial effect plots, show component contributions, link scale, model term linear predictor. y axis plots typically centred around 0 due smooths sum--zero identifiability constraint applied . constraint allows model include multiple smooths remain identifiable. plots allow read contributions smooth fitted response (link scale); show link-scale predictions response smooth, conditional upon terms model, including parametric effects intercept, zero contribution. parlance marginaleffects package (Arel-Bundock, Greifer, Heiss Forthcoming), plots show adjusted predictions, just adjustment includes setting contribution model terms predicted value zero. partial derivatives (marginaleffects call marginal effect slope), gratia provides derivatives(). resulting plot intended reasonable overview estimated model, offers limited option modify resulting plot. want full control, can obtain data used create plot smooth_estimates() evaluate smooths values evenly spaced range covariate(s). want evaluate selected smooths, can specify via smooth argument. takes smooth labels names smooths known mgcv. list labels smooths use evaluate f(x2)f(x_2) use can generate plot using ggplot2 package, example","code":"draw(m) sm <- smooth_estimates(m) sm #> # A tibble: 400 × 9 #>    .smooth .type .by   .estimate   .se       x0    x1    x2    x3 #>    <chr>   <chr> <chr>     <dbl> <dbl>    <dbl> <dbl> <dbl> <dbl> #>  1 s(x0)   TPRS  NA       -1.32  0.390 0.000239    NA    NA    NA #>  2 s(x0)   TPRS  NA       -1.24  0.365 0.0103      NA    NA    NA #>  3 s(x0)   TPRS  NA       -1.17  0.340 0.0204      NA    NA    NA #>  4 s(x0)   TPRS  NA       -1.09  0.318 0.0304      NA    NA    NA #>  5 s(x0)   TPRS  NA       -1.02  0.297 0.0405      NA    NA    NA #>  6 s(x0)   TPRS  NA       -0.947 0.279 0.0506      NA    NA    NA #>  7 s(x0)   TPRS  NA       -0.875 0.263 0.0606      NA    NA    NA #>  8 s(x0)   TPRS  NA       -0.803 0.249 0.0707      NA    NA    NA #>  9 s(x0)   TPRS  NA       -0.732 0.237 0.0807      NA    NA    NA #> 10 s(x0)   TPRS  NA       -0.662 0.228 0.0908      NA    NA    NA #> # ℹ 390 more rows smooths(m) #> [1] \"s(x0)\" \"s(x1)\" \"s(x2)\" \"s(x3)\" sm <- smooth_estimates(m, smooth = \"s(x2)\") #> Warning: The `smooth` argument of `smooth_estimates()` is deprecated as of gratia #> 0.8.9.9. #> ℹ Please use the `select` argument instead. #> This warning is displayed once every 8 hours. #> Call `lifecycle::last_lifecycle_warnings()` to see where this warning was #> generated. sm #> # A tibble: 100 × 6 #>    .smooth .type .by   .estimate   .se      x2 #>    <chr>   <chr> <chr>     <dbl> <dbl>   <dbl> #>  1 s(x2)   TPRS  NA      -4.47   0.476 0.00359 #>  2 s(x2)   TPRS  NA      -4.00   0.406 0.0136  #>  3 s(x2)   TPRS  NA      -3.53   0.345 0.0237  #>  4 s(x2)   TPRS  NA      -3.06   0.295 0.0338  #>  5 s(x2)   TPRS  NA      -2.58   0.263 0.0438  #>  6 s(x2)   TPRS  NA      -2.09   0.250 0.0539  #>  7 s(x2)   TPRS  NA      -1.59   0.253 0.0639  #>  8 s(x2)   TPRS  NA      -1.08   0.264 0.0740  #>  9 s(x2)   TPRS  NA      -0.564  0.278 0.0841  #> 10 s(x2)   TPRS  NA      -0.0364 0.289 0.0941  #> # ℹ 90 more rows library(\"ggplot2\") library(\"dplyr\") #>  #> Attaching package: 'dplyr' #> The following object is masked from 'package:nlme': #>  #>     collapse #> The following objects are masked from 'package:stats': #>  #>     filter, lag #> The following objects are masked from 'package:base': #>  #>     intersect, setdiff, setequal, union sm |>   add_confint() |>   ggplot(aes(y = .estimate, x = x2)) +   geom_ribbon(aes(ymin = .lower_ci, ymax = .upper_ci),     alpha = 0.2, fill = \"forestgreen\"   ) +   geom_line(colour = \"forestgreen\", linewidth = 1.5) +   labs(     y = \"Partial effect\",     title = expression(\"Partial effect of\" ~ f(x[2])),     x = expression(x[2])   )"},{"path":"https://gavinsimpson.github.io/gratia/articles/gratia.html","id":"model-diagnostics","dir":"Articles","previous_headings":"","what":"Model diagnostics","title":"Getting started with gratia","text":"appraise() function provides standard diagnostic plots GAMs  plots produced (left--right, top--bottom), quantile-quantile (QQ) plot deviance residuals, scatterplot deviance residuals linear predictor, histogram deviance residuals, scatterplot observed vs fitted values. Adding partial residuals partial effect plots produced draw() can also help diagnose problems model, oversmoothing","code":"appraise(m) draw(m, residuals = TRUE)"},{"path":"https://gavinsimpson.github.io/gratia/articles/gratia.html","id":"want-to-learn-more","dir":"Articles","previous_headings":"","what":"Want to learn more?","title":"Getting started with gratia","text":"gratia active development area development currently lacking documentation. find package, look help pages package look examples code help get going.","code":""},{"path":[]},{"path":"https://gavinsimpson.github.io/gratia/articles/posterior-simulation.html","id":"what-are-we-simulating","dir":"Articles","previous_headings":"","what":"What are we simulating?","title":"Posterior Simulation","text":"Posterior simulation involves randomly sampling MVN(𝛃̂,𝐕b)\\text{MVN}(\\hat{\\boldsymbol{\\beta}}, \\mathbf{V}_{\\text{b}}) EF(μi,ϕ)\\text{EF}(\\mu_i, \\phi), . might simulate posterior distribution single estimated smooth function see uncertainty estimate function. simulate just subset βj⋅\\beta_{j \\cdot} associated fjf_j interest. Instead, might interested uncertainty expectation (expected value) model given values covariates, case can simulate 𝛃\\boldsymbol{\\beta} sample posterior 𝔼(yi)\\mathbb{E}(y_i), fitted values model. might want generate new values response variable via draws conditional distribution response, simulating new response data 𝕪*\\mathbb{y}^{\\ast}, either observed 𝐱\\mathbf{x} new values $^{}, yi*|𝛈,𝐱∼EF(μî,ϕ)y^{\\ast}_i | \\boldsymbol{\\eta}, \\mathbf{x} \\sim \\text{EF}(\\hat{\\mu_i}, \\phi). Finally, can combine posterior simulation distributions generate posterior draws new data 𝕪*\\mathbb{y}^{\\ast} also include uncertainty expected values. gratia functionality options following functions smooth_samples() generates draws posterior distribution single estimated smooth functions, fitted_samples() generates draws posterior distribution 𝔼(yi|𝐗i=xi)\\mathbb{E}(y_i | \\mathbf{X}_i = x_i), expected value responss, predicted_samples(), generates new response data given supplied values covariates yi*|𝐗i=xi*y^{\\ast}_i | \\mathbf{X}_i = x^{\\ast}_i posterior_samples(), generates draws posterior distribution model, including uncertainty estimated parameters model. simpler terms, fitted_samples() generates predictions “average” expected value response values covariates. predictions include uncertainty estimated values model coefficients. contrast, posterior_samples() generates predictions actual values response might expect observe (model correct) given values covariates. predicted values include variance sampling distribution (error term). predicted_samples() lies somewhere two; predicted values include variation sampling distribution, take model fixed, known. worth reminding posterior draws conditional upon selected values smoothing parameter(s) λj\\lambda_j. act wiggliness estimated smooths known, actual fact estimated (selected perhaps better description) wiglinesses data model fitting. estimated GAM fitted method argument \"REML\", \"ML\", version 𝐕b\\mathbf{V}_{\\text{b}} corrected selected smoothing parameters, 𝐕c\\mathbf{V}_{\\text{c}}, generally available. allows, extent, posterior simulation account additional source uncertainty chosen values 𝛌\\boldsymbol{\\lambda}. two additional functions available gratia posterior simulation: simulate(), derivative_samples(). gratia provides simulate() methods models estimated using gam(), bam(), gamm(), well fitted via scam() scam package. simulate() base R convention thing predicted_samples(), just non-tidy way (pejorative; returns simulated response values matrix, arguably useful math statistical computation.) derivative_samples() provides draws posterior distribution derivative response variable small change focal covariate value. derivative_samples() less general version fitted_samples(); achieve thing two separate calls fitted_samples(). ’ll reserve discussion derivative_samples() separate vignette focused estimating derivatives GAMs. following sections ’ll look four main posterior simulation functions turn.","code":""},{"path":"https://gavinsimpson.github.io/gratia/articles/posterior-simulation.html","id":"posterior-smooths-and-smooth_samples","dir":"Articles","previous_headings":"","what":"Posterior smooths and smooth_samples()","title":"Posterior Simulation","text":"can sample posterior distribution coefficients particular smooth β̂j\\hat{\\beta}_j given values smoothing parameters 𝛌̂\\hat{\\boldsymbol{\\lambda}}. generate posterior samples smooths sampling 𝛃j⋆∼N(β̂j,𝐕β̂j)\\boldsymbol{\\beta}_{j\\star} \\sim N(\\hat{\\beta}_j, \\mathbf{V}_{\\hat{\\beta}_j}) forming 𝐗𝛃̂j𝛃j⋆𝖳\\mathbf{X}_{\\hat{\\boldsymbol{\\beta}}_j} \\boldsymbol{\\beta}_{j\\star}^{\\mathsf{T}}. sampling can done using smooth_samples(). illustrate , ’ll simulate data Gu & Wabha’s 4 smooth example, fit GAM simulated data simulating posterior distribution estimated smooth, sampling coefficients particular smooth. model, coefficients smooth f(x0)f(x_0) stored elements 2 10 coefficients vector. sample posterior distribution coefficients use smooth_samples() choosing particular smooth ’re interested using select argument; want sample smooths posteriors smooths model, select can left default value. Typically ’re bothered particular values covariate evaluate posterior smooths; ask 100 evenly spaced values x0 using n_vals, can provide covariates values via data argument. number posterior smooths sampled controlled argument n; ask 100 samples. Objects returned smooth_samples() draw() method available draw posterior smooths can set n_samples randomly select many smooths draw (seed can provided via argument seed make set chosen smooths repeatable.) credible interval smooth contain smooths. standard 95% credible interval, sampled smooths exceed limits interval.  Following Marra & Wood (2012), blue credible interval contain average 95% grey lines (posterior smooths) given value x0x_0. across function frequentist interpretation credible interval implies values x0x_0 coverage less 95% values greater 95%.","code":"ss_df <- data_sim(\"eg1\", seed = 42) m_ss <- gam(y ~ s(x0) + s(x1) + s(x2) + s(x3), data = ss_df, method = \"REML\") s_x0 <- get_smooth(m_ss, \"s(x0)\") smooth_coef_indices(s_x0) #> [1]  2  3  4  5  6  7  8  9 10 sm_samp <- smooth_samples(m_ss, select = \"s(x0)\", n_vals = 100, n = 100,   seed = 21) sm_samp |>   draw(alpha = 0.3) # evaluate the fitted smooth over x0 and add on a credible interval sm_est <- smooth_estimates(m_ss, select = \"s(x0)\") |>   add_confint() # plot the smooth, credible interval, and posterior smooths sm_est |>   ggplot(aes(x = x0)) +   geom_lineribbon(aes(ymin = .lower_ci, ymax = .upper_ci),     orientation = \"vertical\", fill = \"#56B4E9\", alpha = 0.5   ) +   geom_line(     data = sm_samp,     aes(y = .value, group = .draw), alpha = 0.2   ) +   geom_line(aes(y = .estimate), linewidth = 1, colour = \"#E69F00\") +   labs(y = smooth_label(s_x0))"},{"path":"https://gavinsimpson.github.io/gratia/articles/posterior-simulation.html","id":"posterior-fitted-values-via-fitted_samples","dir":"Articles","previous_headings":"","what":"Posterior fitted values via fitted_samples()","title":"Posterior Simulation","text":"Posterior fitted values draws posterior distribution mean expected value response. expectations returned use predict() estimated GAM, except fitted_samples() includes uncertainty estimated model coefficients, whereas predict() just uses estimated coefficients. example, using data_sim() simulate data example 6 Luo & Wahba (1997) sin(2⋅(4x−2))+2⋅exp(−256⋅(x−0.5)2) \\sin(2 \\cdot (4x - 2)) + 2 \\cdot \\exp(-256 \\cdot (x - 0.5)^2)  data fit adaptive smoother Next create data slice 200 values interval (0,1) ’ll predict model generate posterior fitted values compute fitted values new data posterior fitted values drawn fitted_samples() using Gaussian approximation posterior. just take 10 draws posterior observation new_df merge posterior draws data Adding posterior fitted samples plot data, superimposing Bayesian credible interval fitted values  see posterior draws largely contained credible interval. difference smooth_samples() now ’re including effects model terms. simple model single smooth identity link, difference model constant term uncertainty included samples.","code":"f <- function(x) {   sin(2 * ((4 * x) - 2)) + (2 * exp(-256 * (x - 0.5)^2)) } df <- data_sim(\"lwf6\", dist = \"normal\", scale = 0.3, seed = 2) plt <- df |>   ggplot(aes(x = x, y = y)) +   geom_point(alpha = 0.5) +   geom_function(fun = f) plt m <- gam(y ~ s(x, k = 25, bs = \"ad\"), data = df, method = \"REML\") new_df <- data_slice(m, x = evenly(x, lower = 0, upper = 1, n = 200)) |>   mutate(.row = row_number()) fv <- fitted_values(m, data = new_df) fs <- fitted_samples(m, data = new_df, n = 10, seed = 4) |>   left_join(new_df |> select(.row, x), by = join_by(.row == .row)) plt +   geom_ribbon(data = fv, aes(y = .fitted, ymin = .lower_ci, ymax = .upper_ci),     fill = \"red\", alpha = 0.3) +   geom_line(data = fs, aes(group = .draw, x = x, y = .fitted),     colour = \"yellow\", alpha = 0.4)"},{"path":[]},{"path":"https://gavinsimpson.github.io/gratia/articles/posterior-simulation.html","id":"distributional-gams","dir":"Articles","previous_headings":"Additional examples","what":"Distributional GAMs","title":"Posterior Simulation","text":"possible, predicted_samples() posterior_samples() also work distributional GAMs. possible suitable random number generator available family object stored within model. illustrate, reuse example Matteo Fasiolo, author mgcViz package, based classic mcycle data set MASS package. loading data adding row number variable use late data fit standard Gaussian GAM conditional mean accel. Next, simulate n_sim new response values observed data using predicted_samples() comments briefly indicate dplyr code . Now can plot observed simulated data resulting plot shown left-hand panel figure . clearly problem ; simulated data don’t look much like observations 15ms immediately impact ~45ms impact. due model fitted conditional mean accel. Instead, model conditional mean conditional variance data, linear predictors parameters Gaussian distribution Simulating new data follows using code earlier benefit data wrangling now realised can replace data plot created earlier simulations distribution GAM, plot  plot simulated response data distributional GAM shown right hand panel plot. Now, much less disagreement observed data can produce fitted mdoel.","code":"data(mcycle, package = \"MASS\") mcycle <- mcycle |>   mutate(     .row = row_number()   ) |>   relocate(.row, .before = 1L) m_gau <- gam(accel ~ s(times, k = 20),   data = mcycle, method = \"REML\" ) n_sim <- 10 n_data <- nrow(mcycle)  sim_gau <- predicted_samples(m_gau, n = n_sim, seed = 10) |>   left_join(mcycle |> select(-accel), # join on the observed data for times     by = \".row\"   ) |>   rename(accel = .response) |>        # rename   bind_rows(mcycle |>     relocate(.after = last_col())) |> # bind on observer data   mutate(                             # add indicator: simulated or observed     type = rep(c(\"simulated\", \"observed\"),       times = c(n_sim * n_data, n_data)     ),     .alpha = rep(                     # set alpha values for sims & observed       c(0.2, 1), time = c(n_sim * n_data, n_data)     )   ) plt_labs <- labs(   x = \"Time after impact [ms]\",   y = \"Acceleration [g]\" )  plt_gau <- sim_gau |>   ggplot(aes(x = times)) +   geom_point(aes(y = accel, colour = type, alpha = .alpha)) +   plt_labs +   scale_colour_okabe_ito(order = c(6, 5)) +   scale_alpha_identity() +   theme(legend.position = \"bottom\") m_gaulss <- gam(   list(accel ~ s(times, k = 20, bs = \"tp\"),     ~ s(times, bs = \"tp\")   ), data = mcycle, family = gaulss() ) sim_gaulss <- predicted_samples(m_gaulss, n = n_sim, seed = 20) |>   left_join(mcycle |> select(-accel),     by = \".row\"   ) |>   rename(accel = .response) |>   bind_rows(mcycle |> relocate(.after = last_col())) |>   mutate(     type = rep(c(\"simulated\", \"observed\"),       times = c(n_sim * n_data, n_data)     ),     .alpha = rep(c(0.2, 1), time = c(n_sim * n_data, n_data))   ) plt_gaulss <- plt_gau %+% sim_gaulss  plt_gau + plt_gaulss +   plot_annotation(tag_levels = \"a\") +   plot_layout(guides = \"collect\", ncol = 2)"},{"path":"https://gavinsimpson.github.io/gratia/articles/posterior-simulation.html","id":"prediction-intervals","dir":"Articles","previous_headings":"Additional examples","what":"Prediction intervals","title":"Posterior Simulation","text":"One use posterior simulation generate prediction intervals fitted model. Prediction intervals include two sources uncertainty; estimated model , plus sampling uncertainty error arises drawing observations conditional distribution response. example, Gaussian GAM, first source uncertainty comes uncertainty estimates βj\\beta_j, model coefficients. uncertainty mean expected value response. second source uncertainty stems error term, estimated variance response. two parameters define conditional distribution YiY_i. value covariate(s) 𝐗\\mathbf{X}, estimated model defines entire distribution response values might expect observe covariate values. illustrate, ’ll fit simple GAM single smooth function data simulate Gu & Wabha’s function f2f_2 using data_sim(). simulate 400 values Gaussian distribution variance σ2=1\\sigma^2 = 1. simulated data, true function generated shown  GAM data contains single smooth function x consider new value covariate x, x*=0.5x^{\\ast} = 0.5, expected value response given model, 𝔼(y*|x=x*)\\mathbb{E}(y^{*} | x = x^{*}), ~2.92, obtain using predict() value mean Gaussian distribution , model correct description data, describes distribution values YY might take x=0.5x = 0.5. Gaussian distribution defined two parameters; mean, μ\\mu, describes middle distribution, variance, σ2\\sigma^2, describes spread distribution mean. fully describe Gaussian distribution response x=0.5x = 0.5, need estimate variance. didn’t model explicitly GAM, get estimate model’s scale parameter, ϕ\\phi. stored element scale model object can visualise distribution looks like magic ggdist package  orange region shows expected density response values x*=0.5x^{\\ast} = 0.5 model predicts expect observe. region assumes uncertainty estimate mean variance. Prediction intervals take account variation expected value, plus uncertainty expected value. fitted_values() conveniently returns uncertainty us, default 95% credible interval .se column standard error (standard deviation) estimated value (.fitted), .lower_ci .upper_ci lower upper uncertainty bounds (95% level) estimated value respectively. GAMs fitted mgcv don’t corresponding estimate uncertainty scale parameter, ϕ\\phi, model estimated standard deviation σ̂\\hat{\\sigma}. pretty easy compute upper lower tail quantiles fitted Gaussian distribution range values x get prediction interval, ’d ignoring uncertainty model estimates mean. Posterior simulation provides simple convenient way generate prediction interval includes model uncertainty, works principle families available mgcv (although practice, families currently supported gratia). compute prediction interval x GAM, creating set data evenly range x observed data used fit model added variable .row used later match posterior simulated values row prediction data set ds. also compute fitted values new observations using fitted_values(). step isn’t required order posterior simulation gratia, ’ll use fitted values later show model estimated values uncertainty contrast prediction interval. use posterior_samples() generate new response data new x values ds use join add prediction data draw asked 10000 posterior draws new value x. Ideally ’d generate least three four times many draws get precise estimate prediction interval, keep number low vignette avoid excessive computation time. ’re also using smoothness parameter selection corrected version Bayesian covariance matrix; matrix adjusted account us knowing value smoothing parameter f(xi)f(x_i). ps tibble, n * nrow(ds) rows. .draw variable groups simulated values posterior draw, .row groups posterior draws value x. summarise posterior draws using {dplyr} need function compute quantiles posterior distribution value x (.row). following function simple wrapper around quantile() function base R, arranges output quantile() data frame. apply function set posterior draws, grouping .row summarise separately posterior distribution new value x. reframe() used summarise posterior using quantile_fun() function. ease use, pivot resulting summary long wide format add covariate values joining .row variable 95% prediction interval shown first 10 rows prediction data. column labelled .q50 median posterior distribution. can now use various objects produced plot fitted values model (uncertainties), well prediction intervals just generated. add observed data used fit model black points, summarise posterior samples (ps) using hexagonal binning (avoid plotting 2 million posterior samples)  outermost pair blue lines plot prediction interval created. interval encloses, expected, almost observe data points. also encloses, design, posterior samples, indicated filled hexagonal bins, warmer colours indicating larger counts posterior draws.","code":"df <- data_sim(\"gwf2\", n = 400, scale = 1, dist = \"normal\", seed = 8) df |>   ggplot(aes(x = x, y = y)) +   geom_point() +   geom_function(fun = gw_f2, colour = \"#0072B2\", linewidth = 1.5) m <- gam(y ~ s(x), data = df, method = \"REML\", family = gaussian()) mu <- predict(m, newdata = data.frame(x = 0.5)) mu #>        1  #> 2.919094 sigma <- m$scale sigma #> [1] 1.019426 df |>   ggplot(aes(x = x, y = y)) +   stat_halfeye(aes(ydist = dist_normal(mean = mu, sd = sigma)),     x = 0.5, scale = 0.2, slab_fill = \"#E69F00\", slab_alpha = 0.7   ) +   geom_point() +   geom_function(fun = gw_f2, colour = \"#0072B2\", linewidth = 1.5) +   geom_point(x = 0.5, y = mu, colour = \"red\") fitted_values(m, data = data.frame(x = 0.5)) #> # A tibble: 1 × 6 #>    .row     x .fitted   .se .lower_ci .upper_ci #>   <int> <dbl>   <dbl> <dbl>     <dbl>     <dbl> #> 1     1   0.5    2.92 0.161      2.60      3.23 ds <- data_slice(m, x = evenly(x, n = 200)) |>   mutate(.row = row_number()) fv <- fitted_values(m, data = ds) ps <- posterior_samples(m, n = 10000, data = ds, seed = 24,   unconditional = TRUE) |>   left_join(ds, by = join_by(.row == .row)) ps #> # A tibble: 2,000,000 × 4 #>     .row .draw .response       x #>    <int> <int>     <dbl>   <dbl> #>  1     1     1   -1.34   0.00129 #>  2     2     1   -0.0495 0.00629 #>  3     3     1    0.0308 0.0113  #>  4     4     1   -0.783  0.0163  #>  5     5     1    0.861  0.0213  #>  6     6     1    0.475  0.0263  #>  7     7     1    0.858  0.0313  #>  8     8     1    0.143  0.0363  #>  9     9     1   -0.0344 0.0413  #> 10    10     1    1.04   0.0463  #> # ℹ 1,999,990 more rows quantile_fun <- function(x, probs = c(0.025, 0.5, 0.975), ...) {   tibble::tibble(     .value = quantile(x, probs = probs, ...),     .q = probs * 100   ) } p_int <- ps |>   group_by(.row) |>   reframe(quantile_fun(.response)) |>   pivot_wider(     id_cols = .row, names_from = .q, values_from = .value,     names_prefix = \".q\"   ) |>   left_join(ds, by = join_by(.row == .row)) p_int #> # A tibble: 200 × 5 #>     .row  .q2.5    .q50 .q97.5       x #>    <int>  <dbl>   <dbl>  <dbl>   <dbl> #>  1     1 -2.84  -0.847    1.25 0.00129 #>  2     2 -2.70  -0.651    1.41 0.00629 #>  3     3 -2.50  -0.434    1.62 0.0113  #>  4     4 -2.24  -0.207    1.83 0.0163  #>  5     5 -2.04  -0.0197   2.01 0.0213  #>  6     6 -1.81   0.191    2.25 0.0263  #>  7     7 -1.59   0.391    2.41 0.0313  #>  8     8 -1.38   0.594    2.60 0.0363  #>  9     9 -1.20   0.831    2.84 0.0413  #> 10    10 -0.935  1.06     3.04 0.0463  #> # ℹ 190 more rows fv |>   ggplot(aes(x = x, y = .fitted)) +   # summarise the posterior samples   geom_hex(     data = ps, aes(x = x, y = .response, fill = after_stat(count)),     bins = 50, alpha = 0.7   ) +   # add the lower and upper prediction intervals   geom_line(data = p_int, aes(y = .q2.5), colour = \"#56B4E9\",     linewidth = 1.5) +   geom_line(data = p_int, aes(y = .q97.5), colour = \"#56B4E9\",     linewidth = 1.5) +   # add the lower and upper credible intervals   geom_line(aes(y = .lower_ci), colour = \"#56B4E9\", linewidth = 1) +   geom_line(aes(y = .upper_ci), colour = \"#56B4E9\", linewidth = 1) +   # add the fitted model   geom_line() +   # add the observed data   geom_point(data = df, aes(x = x, y = y)) +   scale_fill_viridis_c(option = \"plasma\") +   theme(legend.position = \"none\") +   labs(y = \"Response\")"},{"path":"https://gavinsimpson.github.io/gratia/articles/posterior-simulation.html","id":"metropolis-hastings-sampler","dir":"Articles","previous_headings":"Additional examples","what":"Metropolis Hastings sampler","title":"Posterior Simulation","text":"cases, Gaussian approximation posterior distribution model coefficients can fail. Simon Wood shows example just failure ?gam.mh help page, Gaussian approximation basically useless binomial GAM large numbers zeroes. mgcv::gam.mh() implements simple Metropolis Hastings sampler, alternates proposals Gaussian t distribution approximation posterior random walk proposals based shrunken approximate posterior covariance matrix. section, rework Simon’s example failure Gaussian approximation ?gam.mh show use gratia generate posterior draws using Metropolis Hastings sampler provided gam.mh(). begin defining function simulate data example. use simulate data set plot  Note zeroes large parts covariate space response zeroes. fit binomial (logistic) GAM data generate sample posterior distribution using default Gaussian approximation subsequently using simpler Metropolis Hastings sampler. method argument used select Metropolis Hastings sampler, specify two additional arguments: thin, controls many draws skipped retained sample, rw_scale, scaling factor posterior covariance matrix shrunk random walk proposals. leave two important arguments defaults: burnin = 1000, number samples discard prior sampling, t_df = 40, degrees freedom t proposals. degrees freedom t proposals large, ’re effectively Gaussian approximation default, alternating proposals random walk proposals. collected posterior draws, summarise set 50%, 80%, 95% intervals using ggdist::median_qi(), add data locations left join First plot intervals Gaussian approximation posterior, repeat plot using intervals derived Metropolis Hastings sampler, arranging two plots using patchwork  Gaussian approximation-based intervals shown left figure, range x largely useless, covering entire range response, despite fact observed zeroes large parts covariate space. Contrast intervals ones obtained using Metropolis Hastings sampler; intervals much better reflect uncertainty estimated response function x data zeroes.","code":"ga_fail <- function(seed) {   df <- tibble(y = c(     rep(0, 89), 1, 0, 1, 0, 0, 1, rep(0, 13), 1, 0, 0, 1,     rep(0, 10), 1, 0, 0, 1, 1, 0, 1, rep(0, 4), 1, rep(0, 3),     1, rep(0, 3), 1, rep(0, 10), 1, rep(0, 4), 1, 0, 1, 0, 0,     rep(1, 4), 0, rep(1, 5), rep(0, 4), 1, 1, rep(0, 46)   )) |>     mutate(       x = withr::with_seed(         seed,         sort(c(0:10 * 5, rnorm(length(y) - 11) * 20 + 100))       ),       .row = row_number()     ) |>     relocate(.row, .before = 1L)   df } df <- ga_fail(3)  df |>   ggplot(aes(x = x, y = y)) +   geom_point() m_logit <- gam(y ~ s(x, k = 15), data = df, method = \"REML\", family = binomial) fs_ga <- fitted_samples(m_logit, n = 2000, seed = 2) fs_mh <- fitted_samples(m_logit,   n = 2000, seed = 2, method = \"mh\", thin = 2,   rw_scale = 0.4 ) excl_col <- c(\".draw\", \".parameter\", \".row\") int_ga <- fs_ga |>   group_by(.row) |>   median_qi(.width = c(0.5, 0.8, 0.95), .exclude = excl_col) |>   left_join(df, by = join_by(.row == .row))  int_mh <- fs_mh |>   group_by(.row) |>   median_qi(.width = c(0.5, 0.8, 0.95), .exclude = excl_col) |>   left_join(df, by = join_by(.row == .row)) plt_ga <- df |>   ggplot(aes(x = x, y = y)) +   geom_point() +   geom_lineribbon(     data = int_ga,     aes(x = x, y = .fitted, ymin = .lower, ymax = .upper)   ) +   scale_fill_brewer() +   labs(title = \"Gaussian approximation\")  plt_mh <- df |>   ggplot(aes(x = x, y = y)) +   geom_point() +   geom_lineribbon(     data = int_mh,     aes(x = x, y = .fitted, ymin = .lower, ymax = .upper)   ) +   scale_fill_brewer() +   labs(title = \"Metropolis Hastings sampler\")  plt_ga + plt_mh + plot_layout(guides = \"collect\")"},{"path":[]},{"path":"https://gavinsimpson.github.io/gratia/authors.html","id":null,"dir":"","previous_headings":"","what":"Authors","title":"Authors and Citation","text":"Gavin L. Simpson. Author, maintainer, copyright holder. Henrik Singmann. Contributor.","code":""},{"path":"https://gavinsimpson.github.io/gratia/authors.html","id":"citation","dir":"","previous_headings":"","what":"Citation","title":"Authors and Citation","text":"Simpson G (????). gratia: Graceful ggplot-Based Graphics Functions GAMs Fitted using mgcv. R package version 0.9.2.9013, https://gavinsimpson.github.io/gratia/.","code":"@Manual{,   title = {{gratia}: Graceful {ggplot}-Based Graphics and Other Functions                   for {GAM}s Fitted using {mgcv}},   author = {Gavin L. Simpson},   abstract = {Graceful ggplot-based graphics and utility functions for              working with generalized additive models (GAMs) fitted using the              mgcv package. Provides a reimplementation of the plot() method for              GAMs that mgcv provides, as well as tidyverse-compatible              representations of estimated smooths.},   note = {R package version 0.9.2.9013},   url = {https://gavinsimpson.github.io/gratia/}, }"},{"path":[]},{"path":"https://gavinsimpson.github.io/gratia/index.html","id":"overview","dir":"","previous_headings":"","what":"Overview","title":"An R package for working with generalized additive models","text":"Working GAMs within ‘tidyverse’ can tedious even difficult without good understanding GAMs model returned ‘mgcv’ model objects contain. ‘gratia’ designed help . ‘gratia’ provides ‘ggplot’-based graphics utility functions working generalized additive models (GAMs) fitted using ‘mgcv’ package, via reimplementation plot() method GAMs ‘mgcv’ provides, well ‘tidyverse’ compatible representations estimated smooths.","code":""},{"path":"https://gavinsimpson.github.io/gratia/index.html","id":"features","dir":"","previous_headings":"","what":"Features","title":"An R package for working with generalized additive models","text":"main features gratia currently ggplot2-based replacement mgcv:::plot.gam(): draw.gam(). example, classic four term additive example Gu & Wahba:  Estimated smooths GAM bivariate smooth:  Estimated smooths GAM Note specialist smoothers (bs %% c(\"mrf\",\"sw\", \"sf\")) currently supported, univariate, factor continuous -variable smooths, simple random effect smooths (bs = 're'), factor-smooth interaction smooths (bs = \"fs\"), constrained factor smooths (bs = \"sz\"), full soap film smooths (bs = \"\"), bivariate, trivariate, quadvariate TPRS tensor product smooths supported, Estimation derivatives fitted smoothers: derivatives(), Estimation point-wise across--function confidence intervals simultaneous intervals smooths: confint.gam(). Model diagnostics via appraise()  Model diagnostics figure","code":""},{"path":"https://gavinsimpson.github.io/gratia/index.html","id":"installing-gratia","dir":"","previous_headings":"","what":"Installing gratia","title":"An R package for working with generalized additive models","text":"gratia now available CRAN, can installed however gratia active development may wish install development version github. easiest way via install_github() function package remotes. Make sure remotes installed, run install package. Alternatively, binary packages development version available rOpenSci’s R Universe service:","code":"install.packages(\"gratia\") remotes::install_github(\"gavinsimpson/gratia\") # Install gratia in R install.packages(\"gratia\", repos = c(   \"https://gavinsimpson.r-universe.dev\",   \"https://cloud.r-project.org\" ))"},{"path":"https://gavinsimpson.github.io/gratia/index.html","id":"history","dir":"","previous_headings":"","what":"History","title":"An R package for working with generalized additive models","text":"gratia grew earlier package, schoenberg, development earlier package tsgam, originally intended used GAMs fitted time series. developing tsgam however became clear package used generally name “tsgam” longer appropriate. avoid breaking blog posts written using tsgam decided copy git repo history new repo package name schoenberg. later date someone released another package called schoenberg CRAN, scuppered idea. Now ’m calling package gratia. Hopefully won’t change …","code":""},{"path":"https://gavinsimpson.github.io/gratia/index.html","id":"why-gratia","dir":"","previous_headings":"","what":"Why gratia?","title":"An R package for working with generalized additive models","text":"naming greta package, Nick Golding observed recent phenomena naming statistical modelling software, Stan Edward, individuals played prominent role development field. lead Nick name Tensor Flow-based package greta Grete Hermann. spirit, gratia named recognition contributions Grace Wahba, pioneering work penalised spline models foundation way GAMs estimated mgcv. wanted name package grace, explicitly recognise Grace’s contributions, unfortunately already package named Grace CRAN. looked elsewhere inspiration. English word “grace” derives Latin gratia, meaning “favor, charm, thanks” (according Merriam Webster). chair Grace Wabha currently holds named Isaac J Schoenberg, former University Madison-Wisconsin Professor Mathematics, 1946 paper provided first mathematical reference “splines”. (Hence previous name package.)","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/add_confint.html","id":null,"dir":"Reference","previous_headings":"","what":"Add a confidence interval to an existing object — add_confint","title":"Add a confidence interval to an existing object — add_confint","text":"Add confidence interval existing object","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/add_confint.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Add a confidence interval to an existing object — add_confint","text":"","code":"add_confint(object, coverage = 0.95, ...)  # S3 method for class 'smooth_estimates' add_confint(object, coverage = 0.95, ...)  # S3 method for class 'parametric_effects' add_confint(object, coverage = 0.95, ...)  # Default S3 method add_confint(object, coverage = 0.95, ...)"},{"path":"https://gavinsimpson.github.io/gratia/reference/add_confint.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Add a confidence interval to an existing object — add_confint","text":"object R object. coverage numeric; coverage interval. Must range 0 < coverage < 1. ... arguments passed methods.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/add_constant.html","id":null,"dir":"Reference","previous_headings":"","what":"Add a constant to estimated values — add_constant","title":"Add a constant to estimated values — add_constant","text":"Add constant estimated values","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/add_constant.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Add a constant to estimated values — add_constant","text":"","code":"add_constant(object, constant = NULL, ...)  # S3 method for class 'smooth_estimates' add_constant(object, constant = NULL, ...)  # S3 method for class 'smooth_samples' add_constant(object, constant = NULL, ...)  # S3 method for class 'mgcv_smooth' add_constant(object, constant = NULL, ...)  # S3 method for class 'parametric_effects' add_constant(object, constant = NULL, ...)  # S3 method for class 'tbl_df' add_constant(object, constant = NULL, column = NULL, ...)  # S3 method for class 'evaluated_parametric_term' add_constant(object, constant = NULL, ...)"},{"path":"https://gavinsimpson.github.io/gratia/reference/add_constant.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Add a constant to estimated values — add_constant","text":"object object add constant . constant constant add. ... additional arguments passed methods. column character; \"tbl_df\" method, column add constant .","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/add_constant.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Add a constant to estimated values — add_constant","text":"Returns object estimate shifted addition supplied constant.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/add_constant.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Add a constant to estimated values — add_constant","text":"Gavin L. Simpson","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/add_fitted.gam.html","id":null,"dir":"Reference","previous_headings":"","what":"Add fitted values from a GAM to a data frame — add_fitted.gam","title":"Add fitted values from a GAM to a data frame — add_fitted.gam","text":"Add fitted values GAM data frame","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/add_fitted.gam.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Add fitted values from a GAM to a data frame — add_fitted.gam","text":"","code":"# S3 method for class 'gam' add_fitted(data, model, value = \".fitted\", type = \"response\", ...)"},{"path":"https://gavinsimpson.github.io/gratia/reference/add_fitted.gam.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Add fitted values from a GAM to a data frame — add_fitted.gam","text":"data data frame containing values variables used fit model. Passed stats::predict() newdata. model fitted model stats::predict() method available. S3 method dispatch performed model argument. value character; name variable model predictions stored. type character; type predictions return. See mgcv::predict.gam() options. ... additional arguments passed mgcv::predict.gam().","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/add_fitted.gam.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Add fitted values from a GAM to a data frame — add_fitted.gam","text":"data frame (tibble) formed data predictions model.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/add_fitted.gam.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Add fitted values from a GAM to a data frame — add_fitted.gam","text":"","code":"load_mgcv() df <- data_sim(\"eg1\", seed = 1) df <- df[, c(\"y\", \"x0\", \"x1\", \"x2\", \"x3\")] m <- gam(y ~ s(x0) + s(x1) + s(x2) + s(x3), data = df, method = \"REML\")  # add fitted values to our data add_fitted(df, m) #> # A tibble: 400 x 6 #>          y     x0     x1     x2    x3 .fitted #>      <dbl>  <dbl>  <dbl>  <dbl> <dbl>   <dbl> #>  1  3.34   0.266  0.659  0.859  0.367    5.90 #>  2 -0.0758 0.372  0.185  0.0344 0.741    3.15 #>  3 10.7    0.573  0.954  0.971  0.934    8.28 #>  4  8.73   0.908  0.898  0.745  0.673    8.65 #>  5 15.0    0.202  0.944  0.273  0.701   15.7  #>  6  7.67   0.898  0.724  0.677  0.848    8.38 #>  7  7.58   0.945  0.370  0.348  0.706    7.84 #>  8  8.51   0.661  0.781  0.947  0.859    6.74 #>  9 10.6    0.629  0.0111 0.339  0.446    9.14 #> 10  3.72   0.0618 0.940  0.0317 0.677    7.04 #> # i 390 more rows  # with type = \"terms\" or \"iterms\" add_fitted(df, m, type = \"terms\") #> # A tibble: 400 x 10 #>          y     x0     x1     x2    x3 .constant `s(x0)` `s(x1)` `s(x2)` `s(x3)` #>      <dbl>  <dbl>  <dbl>  <dbl> <dbl>     <dbl>   <dbl>   <dbl>   <dbl>   <dbl> #>  1  3.34   0.266  0.659  0.859  0.367      7.94  0.175    0.559  -2.81   0.0351 #>  2 -0.0758 0.372  0.185  0.0344 0.741      7.94  0.435   -1.92   -3.23  -0.0687 #>  3 10.7    0.573  0.954  0.971  0.934      7.94  0.593    3.35   -3.47  -0.122  #>  4  8.73   0.908  0.898  0.745  0.673      7.94 -0.812    2.77   -1.19  -0.0498 #>  5 15.0    0.202  0.944  0.273  0.701      7.94 -0.0589   3.23    4.63  -0.0576 #>  6  7.67   0.898  0.724  0.677  0.848      7.94 -0.745    1.15    0.146 -0.0981 #>  7  7.58   0.945  0.370  0.348  0.706      7.94 -1.07    -1.31    2.34  -0.0589 #>  8  8.51   0.661  0.781  0.947  0.859      7.94  0.434    1.67   -3.20  -0.101  #>  9 10.6    0.629  0.0111 0.339  0.446      7.94  0.512   -1.95    2.63   0.0132 #> 10  3.72   0.0618 0.940  0.0317 0.677      7.94 -0.695    3.20   -3.35  -0.0508 #> # i 390 more rows"},{"path":"https://gavinsimpson.github.io/gratia/reference/add_fitted.html","id":null,"dir":"Reference","previous_headings":"","what":"Add fitted values from a model to a data frame — add_fitted","title":"Add fitted values from a model to a data frame — add_fitted","text":"Add fitted values model data frame","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/add_fitted.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Add fitted values from a model to a data frame — add_fitted","text":"","code":"add_fitted(data, model, value = \".value\", ...)"},{"path":"https://gavinsimpson.github.io/gratia/reference/add_fitted.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Add fitted values from a model to a data frame — add_fitted","text":"data data frame containing values variables used fit model. Passed stats::predict() newdata. model fitted model stats::predict() method available. S3 method dispatch performed model argument. value character; name variable model predictions stored. ... additional arguments passed methods.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/add_fitted.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Add fitted values from a model to a data frame — add_fitted","text":"data frame (tibble) formed data fitted values model.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/add_fitted_samples.html","id":null,"dir":"Reference","previous_headings":"","what":"Add posterior draws from a model to a data object — add_fitted_samples","title":"Add posterior draws from a model to a data object — add_fitted_samples","text":"Adds draws posterior distribution model data object using one fitted_samples(), predicted_samples(), posterior_samples().","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/add_fitted_samples.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Add posterior draws from a model to a data object — add_fitted_samples","text":"","code":"add_fitted_samples(object, model, n = 1, seed = NULL, ...)  add_predicted_samples(object, model, n = 1, seed = NULL, ...)  add_posterior_samples(object, model, n = 1, seed = NULL, ...)  add_smooth_samples(object, model, n = 1, seed = NULL, select = NULL, ...)"},{"path":"https://gavinsimpson.github.io/gratia/reference/add_fitted_samples.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Add posterior draws from a model to a data object — add_fitted_samples","text":"object data frame tibble posterior draws added. model fitted GAM (GAM-like) object posterior draw method exists. n integer; number posterior draws add. seed numeric; value seed random number generator. ... arguments passed posterior draw function, currently one fitted_samples(), predicted_samples(), posterior_samples(). n seed already specified arguments also passed posterior sampling function. select character; select smooth's posterior draw . default, NULL, means posteriors smooths model wil sampled individually. supplied, character vector requested smooth terms.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/add_fitted_samples.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Add posterior draws from a model to a data object — add_fitted_samples","text":"","code":"load_mgcv()  df <- data_sim(\"eg1\", n = 400, seed = 42)  m <- gam(y ~ s(x0) + s(x1) + s(x2) + s(x3), data = df, method = \"REML\")  # add fitted samples (posterior draws of the expected value of the response) # note that there are 800 rows in the output: 400 data by `n = 2` samples. df |>   add_fitted_samples(m, n = 2, seed = 84) #> # A tibble: 800 × 14 #>        y    x0     x1    x2    x3     f    f0    f1     f2    f3  .row .draw #>    <dbl> <dbl>  <dbl> <dbl> <dbl> <dbl> <dbl> <dbl>  <dbl> <dbl> <int> <int> #>  1  2.99 0.915 0.0227 0.909 0.402  1.62 0.529  1.05 0.0397     0     1     1 #>  2  2.99 0.915 0.0227 0.909 0.402  1.62 0.529  1.05 0.0397     0     1     2 #>  3  4.70 0.937 0.513  0.900 0.432  3.25 0.393  2.79 0.0630     0     2     1 #>  4  4.70 0.937 0.513  0.900 0.432  3.25 0.393  2.79 0.0630     0     2     2 #>  5 13.9  0.286 0.631  0.192 0.664 13.5  1.57   3.53 8.41       0     3     1 #>  6 13.9  0.286 0.631  0.192 0.664 13.5  1.57   3.53 8.41       0     3     2 #>  7  5.71 0.830 0.419  0.532 0.182  6.12 1.02   2.31 2.79       0     4     1 #>  8  5.71 0.830 0.419  0.532 0.182  6.12 1.02   2.31 2.79       0     4     2 #>  9  7.63 0.642 0.879  0.522 0.838 10.4  1.80   5.80 2.76       0     5     1 #> 10  7.63 0.642 0.879  0.522 0.838 10.4  1.80   5.80 2.76       0     5     2 #> # ℹ 790 more rows #> # ℹ 2 more variables: .parameter <chr>, .fitted <dbl>  # add posterior draws from smooth s(x2) df |>   add_smooth_samples(m, n = 2, seed = 2, select= \"s(x2)\") #> # A tibble: 800 × 15 #>        y    x0     x1    x2    x3     f    f0    f1     f2    f3  .row .smooth #>    <dbl> <dbl>  <dbl> <dbl> <dbl> <dbl> <dbl> <dbl>  <dbl> <dbl> <int> <chr>   #>  1  2.99 0.915 0.0227 0.909 0.402  1.62 0.529  1.05 0.0397     0     1 s(x2)   #>  2  2.99 0.915 0.0227 0.909 0.402  1.62 0.529  1.05 0.0397     0     1 s(x2)   #>  3  4.70 0.937 0.513  0.900 0.432  3.25 0.393  2.79 0.0630     0     2 s(x2)   #>  4  4.70 0.937 0.513  0.900 0.432  3.25 0.393  2.79 0.0630     0     2 s(x2)   #>  5 13.9  0.286 0.631  0.192 0.664 13.5  1.57   3.53 8.41       0     3 s(x2)   #>  6 13.9  0.286 0.631  0.192 0.664 13.5  1.57   3.53 8.41       0     3 s(x2)   #>  7  5.71 0.830 0.419  0.532 0.182  6.12 1.02   2.31 2.79       0     4 s(x2)   #>  8  5.71 0.830 0.419  0.532 0.182  6.12 1.02   2.31 2.79       0     4 s(x2)   #>  9  7.63 0.642 0.879  0.522 0.838 10.4  1.80   5.80 2.76       0     5 s(x2)   #> 10  7.63 0.642 0.879  0.522 0.838 10.4  1.80   5.80 2.76       0     5 s(x2)   #> # ℹ 790 more rows #> # ℹ 3 more variables: .term <chr>, .draw <int>, .value <dbl>"},{"path":"https://gavinsimpson.github.io/gratia/reference/add_partial_residuals.html","id":null,"dir":"Reference","previous_headings":"","what":"Add partial residuals — add_partial_residuals","title":"Add partial residuals — add_partial_residuals","text":"Add partial residuals","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/add_partial_residuals.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Add partial residuals — add_partial_residuals","text":"","code":"add_partial_residuals(data, model, ...)  # S3 method for class 'gam' add_partial_residuals(data, model, select = NULL, partial_match = FALSE, ...)"},{"path":"https://gavinsimpson.github.io/gratia/reference/add_partial_residuals.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Add partial residuals — add_partial_residuals","text":"data data frame containing values variables used fit model. Passed stats::residuals() newdata. model fitted model stats::residuals() method available. S3 method dispatch performed model argument. ... arguments passed methods. select character, logical, numeric; smooths plot. NULL, default, model smooths drawn. Numeric select indexes smooths order specified formula stored object. Character select matches labels smooths shown example output summary(object). Logical select operates per numeric select order smooths stored. partial_match logical; smooths selected partial matches select? TRUE, select can single string match .","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/add_partial_residuals.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Add partial residuals — add_partial_residuals","text":"","code":"load_mgcv() df <- data_sim(\"eg1\", seed = 1) df <- df[, c(\"y\", \"x0\", \"x1\", \"x2\", \"x3\")] m <- gam(y ~ s(x0) + s(x1) + s(x2) + s(x3), data = df, method = \"REML\")  ## add partial residuals add_partial_residuals(df, m) #> # A tibble: 400 x 9 #>          y     x0     x1     x2    x3 `s(x0)` `s(x1)` `s(x2)` `s(x3)` #>      <dbl>  <dbl>  <dbl>  <dbl> <dbl>   <dbl>   <dbl>   <dbl>   <dbl> #>  1  3.34   0.266  0.659  0.859  0.367  -2.38   -2.00   -5.36  -2.52   #>  2 -0.0758 0.372  0.185  0.0344 0.741  -2.79   -5.15   -6.45  -3.29   #>  3 10.7    0.573  0.954  0.971  0.934   2.99    5.75   -1.07   2.28   #>  4  8.73   0.908  0.898  0.745  0.673  -0.734   2.84   -1.11   0.0287 #>  5 15.0    0.202  0.944  0.273  0.701  -0.752   2.54    3.94  -0.750  #>  6  7.67   0.898  0.724  0.677  0.848  -1.46    0.432  -0.567 -0.812  #>  7  7.58   0.945  0.370  0.348  0.706  -1.33   -1.57    2.08  -0.318  #>  8  8.51   0.661  0.781  0.947  0.859   2.21    3.44   -1.42   1.68   #>  9 10.6    0.629  0.0111 0.339  0.446   2.01   -0.445   4.13   1.51   #> 10  3.72   0.0618 0.940  0.0317 0.677  -4.02   -0.123  -6.67  -3.37   #> # i 390 more rows  ## add partial residuals for selected smooths add_partial_residuals(df, m, select = \"s(x0)\") #> # A tibble: 400 x 6 #>          y     x0     x1     x2    x3 `s(x0)` #>      <dbl>  <dbl>  <dbl>  <dbl> <dbl>   <dbl> #>  1  3.34   0.266  0.659  0.859  0.367  -2.38  #>  2 -0.0758 0.372  0.185  0.0344 0.741  -2.79  #>  3 10.7    0.573  0.954  0.971  0.934   2.99  #>  4  8.73   0.908  0.898  0.745  0.673  -0.734 #>  5 15.0    0.202  0.944  0.273  0.701  -0.752 #>  6  7.67   0.898  0.724  0.677  0.848  -1.46  #>  7  7.58   0.945  0.370  0.348  0.706  -1.33  #>  8  8.51   0.661  0.781  0.947  0.859   2.21  #>  9 10.6    0.629  0.0111 0.339  0.446   2.01  #> 10  3.72   0.0618 0.940  0.0317 0.677  -4.02  #> # i 390 more rows"},{"path":"https://gavinsimpson.github.io/gratia/reference/add_residuals.gam.html","id":null,"dir":"Reference","previous_headings":"","what":"Add residuals from a GAM to a data frame — add_residuals.gam","title":"Add residuals from a GAM to a data frame — add_residuals.gam","text":"Add residuals GAM data frame","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/add_residuals.gam.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Add residuals from a GAM to a data frame — add_residuals.gam","text":"","code":"# S3 method for class 'gam' add_residuals(data, model, value = \".residual\", type = \"deviance\", ...)"},{"path":"https://gavinsimpson.github.io/gratia/reference/add_residuals.gam.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Add residuals from a GAM to a data frame — add_residuals.gam","text":"data data frame containing values variables used fit model. Passed stats::predict() newdata. model fitted model stats::predict() method available. S3 method dispatch performed model argument. value character; name variable model predictions stored. type character; type residuals return. See mgcv::residuals.gam() options. ... additional arguments passed mgcv::residuals.gam().","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/add_residuals.gam.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Add residuals from a GAM to a data frame — add_residuals.gam","text":"data frame (tibble) formed data residuals model.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/add_residuals.gam.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Add residuals from a GAM to a data frame — add_residuals.gam","text":"","code":"load_mgcv() df <- data_sim(\"eg1\", seed = 1) df <- df[, c(\"y\", \"x0\", \"x1\", \"x2\", \"x3\")] m <- gam(y ~ s(x0) + s(x1) + s(x2) + s(x3), data = df, method = \"REML\")  ## add_residuals(df, m) #> # A tibble: 400 x 6 #>          y     x0     x1     x2    x3 .residual #>      <dbl>  <dbl>  <dbl>  <dbl> <dbl>     <dbl> #>  1  3.34   0.266  0.659  0.859  0.367   -2.56   #>  2 -0.0758 0.372  0.185  0.0344 0.741   -3.22   #>  3 10.7    0.573  0.954  0.971  0.934    2.40   #>  4  8.73   0.908  0.898  0.745  0.673    0.0785 #>  5 15.0    0.202  0.944  0.273  0.701   -0.693  #>  6  7.67   0.898  0.724  0.677  0.848   -0.714  #>  7  7.58   0.945  0.370  0.348  0.706   -0.259  #>  8  8.51   0.661  0.781  0.947  0.859    1.78   #>  9 10.6    0.629  0.0111 0.339  0.446    1.50   #> 10  3.72   0.0618 0.940  0.0317 0.677   -3.32   #> # i 390 more rows"},{"path":"https://gavinsimpson.github.io/gratia/reference/add_residuals.html","id":null,"dir":"Reference","previous_headings":"","what":"Add residuals from a model to a data frame — add_residuals","title":"Add residuals from a model to a data frame — add_residuals","text":"Add residuals model data frame","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/add_residuals.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Add residuals from a model to a data frame — add_residuals","text":"","code":"add_residuals(data, model, value = \".residual\", ...)"},{"path":"https://gavinsimpson.github.io/gratia/reference/add_residuals.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Add residuals from a model to a data frame — add_residuals","text":"data data frame containing values variables used fit model. Passed stats::residuals() newdata. model fitted model stats::residuals() method available. S3 method dispatch performed model argument. value character; name variable model residuals stored. ... additional arguments passed methods.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/add_residuals.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Add residuals from a model to a data frame — add_residuals","text":"data frame (tibble) formed data residuals model.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/add_sizer.html","id":null,"dir":"Reference","previous_headings":"","what":"Add indicators of significant change after SiZeR — add_sizer","title":"Add indicators of significant change after SiZeR — add_sizer","text":"Add indicators significant change SiZeR","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/add_sizer.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Add indicators of significant change after SiZeR — add_sizer","text":"","code":"add_sizer(object, type = c(\"change\", \"sizer\"), ...)  # S3 method for class 'derivatives' add_sizer(object, type = c(\"change\", \"sizer\"), ...)  # S3 method for class 'smooth_estimates' add_sizer(object, type = c(\"change\", \"sizer\"), derivatives = NULL, ...)"},{"path":"https://gavinsimpson.github.io/gratia/reference/add_sizer.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Add indicators of significant change after SiZeR — add_sizer","text":"object R object. Currently supported methods classes \"derivatives\". type character; \"change\" adds single variable object indicating credible interval derivative excludes 0. \"sizer\" adds two variables indicating whether derivative postive negative. ... arguments passed methods derivatives object class \"derivatives\", resulting call derivatives().","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/add_sizer.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Add indicators of significant change after SiZeR — add_sizer","text":"","code":"load_mgcv() df <- data_sim(\"eg1\", n = 400, dist = \"normal\", scale = 2, seed = 42) m <- gam(y ~ s(x0) + s(x1) + s(x2) + s(x3), data = df, method = \"REML\")  ## first derivatives of all smooths using central finite differences d <- derivatives(m, type = \"central\") |>   add_sizer()  # default adds a .change column names(d) #>  [1] \".smooth\"     \".by\"         \".fs\"         \".derivative\" \".se\"         #>  [6] \".crit\"       \".lower_ci\"   \".upper_ci\"   \".change\"     \"x0\"          #> [11] \"x1\"          \"x2\"          \"x3\""},{"path":"https://gavinsimpson.github.io/gratia/reference/appraise.html","id":null,"dir":"Reference","previous_headings":"","what":"Model diagnostic plots — appraise","title":"Model diagnostic plots — appraise","text":"Model diagnostic plots","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/appraise.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Model diagnostic plots — appraise","text":"","code":"appraise(model, ...)  # S3 method for class 'gam' appraise(   model,   method = c(\"uniform\", \"simulate\", \"normal\", \"direct\"),   use_worm = FALSE,   n_uniform = 10,   n_simulate = 50,   seed = NULL,   type = c(\"deviance\", \"pearson\", \"response\"),   n_bins = c(\"sturges\", \"scott\", \"fd\"),   ncol = NULL,   nrow = NULL,   guides = \"keep\",   level = 0.9,   ci_col = \"black\",   ci_alpha = 0.2,   point_col = \"black\",   point_alpha = 1,   line_col = \"red\",   ... )  # S3 method for class 'lm' appraise(model, ...)"},{"path":"https://gavinsimpson.github.io/gratia/reference/appraise.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Model diagnostic plots — appraise","text":"model fitted model. Currently models inheriting class \"gam\", well classes \"glm\" \"lm\" calls stats::glm stats::lm supported. ... arguments passed patchwork::wrap_plots(). method character; method used generate theoretical quantiles. default \"uniform\", generates reference quantiles using random draws uniform distribution inverse cummulative distribution function (CDF) fitted values. reference quantiles averaged n_uniform draws. \"simulate\" generates reference quantiles simulating new response data model observed values covariates, residualised generate reference quantiles, using n_simulate simulated data sets. \"normal\" generates reference quantiles using standard normal distribution. \"uniform\" computationally efficient, \"simulate\" allows reference bands drawn QQ-plot. \"normal\" avoided used fall back random number generator (\"simulate\") inverse CDF (\"uniform\"``) available family` used model fitting. Note method = \"direct\" deprecated favour method = \"uniform\". use_worm logical; worm plot drawn place QQ plot? n_uniform numeric; number times randomize uniform quantiles direct computation method (method = \"direct\") QQ plots. n_simulate numeric; number data sets simulate estimated model using simulation method (method = \"simulate\") QQ plots. seed numeric; random number seed use method = \"simulate\" method = \"uniform\". type character; type residuals use. \"deviance\", \"response\", \"pearson\" residuals allowed. n_bins character numeric; either number bins string indicating calculate number bins. ncol, nrow numeric; numbers rows columns spread plots. guides character; one \"keep\" (default), \"collect\", \"auto\". Passed patchwork::plot_layout() level numeric; coverage level QQ plot reference intervals. Must strictly 0 < level < 1. used method = \"simulate\". ci_alpha, ci_col colour transparency used draw QQ plot reference interval method = \"simulate\". point_col, point_alpha colour transparency used draw points plots. See graphics::par() section Color Specification. passed individual plotting functions, therefore affects points plots. line_col colour specification 1:1 line QQ plot reference line residuals vs linear predictor plot.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/appraise.html","id":"note","dir":"Reference","previous_headings":"","what":"Note","title":"Model diagnostic plots — appraise","text":"wording used mgcv::qq.gam() uses direct reference simulated residuals method (method = \"simulated\"). avoid confusion, method = \"direct\" deprecated favour method = \"uniform\".","code":""},{"path":[]},{"path":"https://gavinsimpson.github.io/gratia/reference/appraise.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Model diagnostic plots — appraise","text":"","code":"load_mgcv() ## simulate some data... dat <- data_sim(\"eg1\", n = 400, dist = \"normal\", scale = 2, seed = 2) mod <- gam(y ~ s(x0) + s(x1) + s(x2) + s(x3), data = dat) ## run some basic model checks appraise(mod, point_col = \"steelblue\", point_alpha = 0.4)   ## To change the theme for all panels use the & operator, for example to ## change the ggplot theme for all panels library(\"ggplot2\") appraise(mod, seed = 42,   point_col = \"steelblue\", point_alpha = 0.4,   line_col = \"black\" ) & theme_minimal()"},{"path":"https://gavinsimpson.github.io/gratia/reference/basis.html","id":null,"dir":"Reference","previous_headings":"","what":"Basis expansions for smooths — basis","title":"Basis expansions for smooths — basis","text":"Creates basis expansion definition smoother using syntax mgcv's smooths via mgcv::s()., mgcv::te(), mgcv::ti(), mgcv::t2(), fitted GAM(M).","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/basis.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Basis expansions for smooths — basis","text":"","code":"basis(object, ...)  # S3 method for class 'gam' basis(   object,   select = NULL,   term = deprecated(),   data = NULL,   n = 100,   n_2d = 50,   n_3d = 16,   n_4d = 4,   partial_match = FALSE,   ... )  # S3 method for class 'scam' basis(   object,   select = NULL,   term = deprecated(),   data = NULL,   n = 100,   n_2d = 50,   n_3d = 16,   n_4d = 4,   partial_match = FALSE,   ... )  # S3 method for class 'gamm' basis(   object,   select = NULL,   term = deprecated(),   data = NULL,   n = 100,   n_2d = 50,   n_3d = 16,   n_4d = 4,   partial_match = FALSE,   ... )  # S3 method for class 'list' basis(   object,   select = NULL,   term = deprecated(),   data = NULL,   n = 100,   n_2d = 50,   n_3d = 16,   n_4d = 4,   partial_match = FALSE,   ... )  # Default S3 method basis(   object,   data,   knots = NULL,   constraints = FALSE,   at = NULL,   diagonalize = FALSE,   ... )"},{"path":"https://gavinsimpson.github.io/gratia/reference/basis.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Basis expansions for smooths — basis","text":"object smooth specification, result call one mgcv::s()., mgcv::te(), mgcv::ti(), mgcv::t2(), fitted GAM(M) model. ... arguments passed mgcv::smoothCon(). select character; select smooths fitted model term argument renamed select data data frame containing variables used smooth. n numeric; number points range covariate evaluate smooth. n_2d numeric; number new observations dimension bivariate smooth. currently used; n used dimensions. n_3d numeric; number new observations generate third dimension 3D smooth. n_4d numeric; number new observations generate dimensions higher 2 (!) kD smooth (k >= 4). example, smooth 4D smooth, dimensions 3 4 get n_4d new observations. partial_match logical; case character select, select match partially smooths? partial_match = TRUE, select must single string, character vector length 1. knots list data frame named components containing knots locations. Names must match covariates basis required. See mgcv::smoothCon(). constraints logical; identifiability constraints applied smooth basis. See argument absorb.cons mgcv::smoothCon(). data frame containing values smooth covariate(s) basis evaluated. diagonalize logical; TRUE, reparameterises smooth associated penalty identity matrix. effect turning last diagonal elements penalty zero, highlights penalty null space.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/basis.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Basis expansions for smooths — basis","text":"tibble.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/basis.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Basis expansions for smooths — basis","text":"Gavin L. Simpson","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/basis.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Basis expansions for smooths — basis","text":"","code":"load_mgcv() df <- data_sim(\"eg4\", n = 400, seed = 42)  bf <- basis(s(x0), data = df) bf <- basis(s(x2, by = fac, bs = \"bs\"), data = df, constraints = TRUE)"},{"path":"https://gavinsimpson.github.io/gratia/reference/basis_size.html","id":null,"dir":"Reference","previous_headings":"","what":"Extract basis dimension of a smooth — basis_size","title":"Extract basis dimension of a smooth — basis_size","text":"Extract basis dimension smooth","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/basis_size.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Extract basis dimension of a smooth — basis_size","text":"","code":"basis_size(object, ...)  # S3 method for class 'mgcv.smooth' basis_size(object, ...)  # S3 method for class 'gam' basis_size(object, ...)  # S3 method for class 'gamm' basis_size(object, ...)"},{"path":"https://gavinsimpson.github.io/gratia/reference/basis_size.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Extract basis dimension of a smooth — basis_size","text":"object fitted GAM(M). Currently mgcv::gam() (anything inherits \"gam\" class, e.g. mgcv::bam()) mgcv::gamm() supported. ... Arguments passed methods.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/basis_size.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Extract basis dimension of a smooth — basis_size","text":"","code":"load_mgcv()  df <- data_sim(\"eg1\", n = 200, seed = 1) m <- bam(y ~ s(x0) + s(x1) + s(x2) + s(x3), data = df)  basis_size(m) #> s(x0) s(x1) s(x2) s(x3)  #>     9     9     9     9"},{"path":"https://gavinsimpson.github.io/gratia/reference/bird_move.html","id":null,"dir":"Reference","previous_headings":"","what":"Simulated bird migration data — bird_move","title":"Simulated bird migration data — bird_move","text":"Data generated hypothetical study bird movement along migration corridor, sampled throughout year. dataset consists simulated sample records numbers observed locations 100 tagged individuals six species bird, ten locations along latitudinal gradient, one observation taken every four weeks. Counts simulated randomly species location week creating species-specific migration curve gave probability finding individual given species given location, simulated distribution individuals across sites using multinomial distribution, subsampling using binomial distribution simulation observation error (.e. every bird present location detected). data set (bird_move) consists variables count, latitude, week species.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/bird_move.html","id":"format","dir":"Reference","previous_headings":"","what":"Format","title":"Simulated bird migration data — bird_move","text":"data frame","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/bird_move.html","id":"source","dir":"Reference","previous_headings":"","what":"Source","title":"Simulated bird migration data — bird_move","text":"Pedersen EJ, Miller DL, Simpson GL, Ross N. 2018. Hierarchical generalized additive models: introduction mgcv. PeerJ Preprints 6:e27320v1 doi:10.7287/peerj.preprints.27320v1 .","code":""},{"path":[]},{"path":"https://gavinsimpson.github.io/gratia/reference/boundary.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Extract the boundary of a soap film smooth — boundary","text":"","code":"boundary(x, ...)  # S3 method for class 'soap.film' boundary(x, ...)  # S3 method for class 'gam' boundary(x, select, ...)"},{"path":"https://gavinsimpson.github.io/gratia/reference/boundary.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Extract the boundary of a soap film smooth — boundary","text":"x R object. Currently objects inherit classes \"soap.film\" \"gam\". ... arguments passed methods. select character; label soap film smooth extract boundary.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/boundary.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Extract the boundary of a soap film smooth — boundary","text":"list lists data frames specifying loops define boundary soap film smooth.","code":""},{"path":[]},{"path":"https://gavinsimpson.github.io/gratia/reference/check_user_select_smooths.html","id":null,"dir":"Reference","previous_headings":"","what":"Select smooths based on user's choices — check_user_select_smooths","title":"Select smooths based on user's choices — check_user_select_smooths","text":"Given vector indexing smooths GAM, returns logical vector selecting requested smooths.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/check_user_select_smooths.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Select smooths based on user's choices — check_user_select_smooths","text":"","code":"check_user_select_smooths(   smooths,   select = NULL,   partial_match = FALSE,   model_name = NULL )"},{"path":"https://gavinsimpson.github.io/gratia/reference/check_user_select_smooths.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Select smooths based on user's choices — check_user_select_smooths","text":"smooths character; vector smooth labels. select numeric, logical, character vector selected smooths. partial_match logical; case character select, select match partially smooths? partial_match = TRUE, select must single string, character vector length 1. model_name character; model name used error messages.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/check_user_select_smooths.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Select smooths based on user's choices — check_user_select_smooths","text":"logical vector length length(smooths) indicating smooths selected.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/check_user_select_smooths.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Select smooths based on user's choices — check_user_select_smooths","text":"Gavin L. Simpson","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/coef.scam.html","id":null,"dir":"Reference","previous_headings":"","what":"Extract coefficients from a fitted scam model. — coef.scam","title":"Extract coefficients from a fitted scam model. — coef.scam","text":"Extract coefficients fitted scam model.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/coef.scam.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Extract coefficients from a fitted scam model. — coef.scam","text":"","code":"# S3 method for class 'scam' coef(object, parametrized = TRUE, ...)"},{"path":"https://gavinsimpson.github.io/gratia/reference/coef.scam.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Extract coefficients from a fitted scam model. — coef.scam","text":"object model object fitted scam() parametrized logical; extract parametrized coefficients, respect linear inequality constraints model. ... arguments.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/compare_smooths.html","id":null,"dir":"Reference","previous_headings":"","what":"Compare smooths across models — compare_smooths","title":"Compare smooths across models — compare_smooths","text":"Compare smooths across models","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/compare_smooths.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Compare smooths across models — compare_smooths","text":"","code":"compare_smooths(   model,   ...,   select = NULL,   smooths = deprecated(),   n = 100,   data = NULL,   unconditional = FALSE,   overall_uncertainty = TRUE,   partial_match = FALSE )"},{"path":"https://gavinsimpson.github.io/gratia/reference/compare_smooths.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Compare smooths across models — compare_smooths","text":"model Primary model comparison. ... Additional models compare smooths model. select character; select smooths compare. default (NULL) means smooths model compared. Numeric select indexes smooths order specified formula stored model. Character select matches labels smooths shown example output summary(object). Logical select operates per numeric select order smooths stored. smooths Use select instead. n numeric; number points range covariate evaluate smooth. data data frame covariate values evaluate smooth. unconditional logical; confidence intervals include uncertainty due smoothness selection? TRUE, corrected Bayesian covariance matrix used. overall_uncertainty logical; uncertainty model constant term included standard error evaluate values smooth? partial_match logical; smooths selected partial matches select? TRUE, select can single string match .","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/compare_smooths.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Compare smooths across models — compare_smooths","text":"","code":"load_mgcv() dat <- data_sim(\"eg1\", seed = 2)  ## models to compare smooths across - artificially create differences m1 <- gam(y ~ s(x0, k = 5) + s(x1, k = 5) + s(x2, k = 5) + s(x3, k = 5),   data = dat, method = \"REML\" ) m2 <- gam(y ~ s(x0, bs = \"ts\") + s(x1, bs = \"ts\") + s(x2, bs = \"ts\") +   s(x3, bs = \"ts\"), data = dat, method = \"REML\")  ## build comparisons comp <- compare_smooths(m1, m2) comp #> # A tibble: 8 x 5 #>   .model .smooth .type         .by   data               #>   <chr>  <chr>   <chr>         <chr> <list>             #> 1 m1     s(x0)   TPRS          NA    <tibble [100 x 3]> #> 2 m2     s(x0)   TPRS (shrink) NA    <tibble [100 x 3]> #> 3 m1     s(x1)   TPRS          NA    <tibble [100 x 3]> #> 4 m2     s(x1)   TPRS (shrink) NA    <tibble [100 x 3]> #> 5 m1     s(x2)   TPRS          NA    <tibble [100 x 3]> #> 6 m2     s(x2)   TPRS (shrink) NA    <tibble [100 x 3]> #> 7 m1     s(x3)   TPRS          NA    <tibble [100 x 3]> #> 8 m2     s(x3)   TPRS (shrink) NA    <tibble [100 x 3]> ## notice that the result is a nested tibble  draw(comp)"},{"path":"https://gavinsimpson.github.io/gratia/reference/conditional_values.html","id":null,"dir":"Reference","previous_headings":"","what":"Conditional predictions from a GAM — conditional_values","title":"Conditional predictions from a GAM — conditional_values","text":"Generate predicted values GAM, conditional upon supplied values covariates. conditional_values() modelled marginaleffects::plot_predictions(), intentionally simpler, restrictive functionality. intended use case quickly visualizing predicted values fitted GAM response scale. complex model predictions, strongly encouraged use marginaleffects::plot_predictions().","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/conditional_values.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Conditional predictions from a GAM — conditional_values","text":"","code":"conditional_values(   model,   condition = NULL,   data = NULL,   scale = c(\"response\", \"link\", \"linear_predictor\"),   ... )  # S3 method for class 'gam' conditional_values(   model,   condition = NULL,   data = NULL,   scale = c(\"response\", \"link\", \"linear_predictor\"),   n_vals = 100,   ci_level = 0.95,   ... )"},{"path":"https://gavinsimpson.github.io/gratia/reference/conditional_values.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Conditional predictions from a GAM — conditional_values","text":"model fitted GAM object. condition either character vector list supplying names covariates, possibly values, condition . order values determines plotted via draw.conditional_values() method; first element mapped x channel, second element colour channel, third ggplot2::facet_wrap() fourth element present, present, fourth element mapped rows third element mapped columns ggplot2::facet_grid(). data data frame values predict. supplied overrides values supplied condition. scale character; scale predictions returned ? ... arguments passed fitted_values(). n_vals numeric; number values generate numeric variables named condition. ci_level numeric; number interval (0,1) giving coverage credible intervals.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/conditional_values.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Conditional predictions from a GAM — conditional_values","text":"data frame (tibble) class \"conditional_values\".","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/conditional_values.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Conditional predictions from a GAM — conditional_values","text":"Gavin L. Simpson","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/conditional_values.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Conditional predictions from a GAM — conditional_values","text":"","code":"load_mgcv() df <- data_sim(\"eg1\", seed = 2) m1 <- gam(y ~ s(x0) + s(x1) + s(x2) + s(x3), data = df, method = \"REML\")  # predictions conditional on values evenly spaced over x2, all other # variables in model are held at representative values cv <- conditional_values(   m1,   condition = \"x2\" ) # plot cv |> draw()   # as above but condition on `x1` also. When plotted, `x1` is mapped to the # colour channel, noting that it has been summarised using fivenum() cv <- conditional_values(   m1,   condition = c(\"x2\", \"x1\") ) # plot cv |> draw()   # can pass `condition` a list, allowing for greater flexibility # For example, here we condition on all four variables in the model, # summarising: #   * `x1` at its five number summary, #   * `x0  at its quartiles #   * `x3` at its mean a d mean +/- sd cv <- conditional_values(   m1,   condition = list(\"x2\", x1 = \"fivenum\", x0 = \"quartile\", x3 = \"threenum\") ) # plot cv |> draw()   # some model terms can be exclude from the conditional predictions using the # `exclude` mechanism of `predict.gam`. Here we exclude the effects of # `s(x0)` and `s(x3)` from the conditional predictions. This, in effect, # treats these smooths as having **0** effect on the conditional predictions # of the response, even though the two smooths conditioned on (`s(x2)` and # `s(x1)`) were estimated given the two excluded smooths were in the model cv <- conditional_values(   m1,   condition = list(\"x2\", x1 = \"minmax\"),   exclude = c(\"s(x0)\", \"s(x3)\") ) # plot cv |> draw()   # categorical conditions are also handled df <- data_sim(\"eg4\", seed = 2) m2 <- gam(y ~ fac + s(x2, by = fac) + s(x0), data = df, method = \"REML\") cv <- conditional_values(   m2,   condition = list(\"fac\", x2 = \"fivenum\") ) # plot - we see a discrete x axis cv |> draw()   # in this example we condition on `x2` and `fac %in% c(2,3)` cv <- conditional_values(   m2,   condition = list(\"x2\", fac = 2:3) ) # plot - smooths of `x2` for `fac == 2` and `fac == 3` cv |> draw()"},{"path":"https://gavinsimpson.github.io/gratia/reference/confint.fderiv.html","id":null,"dir":"Reference","previous_headings":"","what":"Point-wise and simultaneous confidence intervals for derivatives of smooths — confint.fderiv","title":"Point-wise and simultaneous confidence intervals for derivatives of smooths — confint.fderiv","text":"Calculates point-wise confidence simultaneous intervals first derivatives smooth terms fitted GAM.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/confint.fderiv.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Point-wise and simultaneous confidence intervals for derivatives of smooths — confint.fderiv","text":"","code":"# S3 method for class 'fderiv' confint(   object,   parm,   level = 0.95,   type = c(\"confidence\", \"simultaneous\"),   nsim = 10000,   ncores = 1L,   ... )"},{"path":"https://gavinsimpson.github.io/gratia/reference/confint.fderiv.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Point-wise and simultaneous confidence intervals for derivatives of smooths — confint.fderiv","text":"object object class \"fderiv\" containing estimated derivatives. parm parameters (smooth terms) given intervals vector terms. missing, parameters considered. level numeric, 0 < level < 1; confidence level point-wise simultaneous interval. default 0.95 95% interval. type character; type interval compute. One \"confidence\" point-wise intervals, \"simultaneous\" simultaneous intervals. nsim integer; number simulations used computing simultaneous intervals. ncores number cores generating random variables multivariate normal distribution. Passed mvnfast::rmvn(). Parallelization take place OpenMP supported (appears work Windows current R). ... additional arguments methods","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/confint.fderiv.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Point-wise and simultaneous confidence intervals for derivatives of smooths — confint.fderiv","text":"data frame components: term; factor indicating term row relates, lower; lower limit confidence simultaneous interval, est; estimated derivative upper; upper limit confidence simultaneous interval.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/confint.fderiv.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Point-wise and simultaneous confidence intervals for derivatives of smooths — confint.fderiv","text":"Gavin L. Simpson","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/confint.fderiv.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Point-wise and simultaneous confidence intervals for derivatives of smooths — confint.fderiv","text":"","code":"load_mgcv() dat <- data_sim(\"eg1\", n = 1000, dist = \"normal\", scale = 2, seed = 2) mod <- gam(y ~ s(x0) + s(x1) + s(x2) + s(x3), data = dat, method = \"REML\")  # new data to evaluate the derivatives at, say over the middle 50% of range # of each covariate middle <- function(x, n = 25, coverage = 0.5) {   v <- (1 - coverage) / 2   q <- quantile(x, prob = c(0 + v, 1 - v), type = 8)   seq(q[1], q[2], length = n) } new_data <- sapply(dat[c(\"x0\", \"x1\", \"x2\", \"x3\")], middle) new_data <- data.frame(new_data) ## first derivatives of all smooths... fd <- fderiv(mod, newdata = new_data) #> Warning: `fderiv()` was deprecated in gratia 0.7.0. #> i Please use `derivatives()` instead.  ## point-wise interval ci <- confint(fd, type = \"confidence\") ci #> # A tibble: 100 x 4 #>    term    lower   est upper #>    <chr>   <dbl> <dbl> <dbl> #>  1 s(x0)  1.7     4.1    6.6 #>  2 s(x0)  1.3     3.8    6.3 #>  3 s(x0)  0.99    3.5    6.0 #>  4 s(x0)  0.68    3.1    5.6 #>  5 s(x0)  0.37    2.8    5.2 #>  6 s(x0)  0.0049  2.4    4.8 #>  7 s(x0) -0.40    2.0    4.5 #>  8 s(x0) -0.79    1.7    4.2 #>  9 s(x0) -1.1     1.3    3.8 #> 10 s(x0) -1.4     0.99   3.4 #> # i 90 more rows  ## simultaneous interval for smooth term of x2 x2_sint <- confint(fd,   parm = \"x2\", type = \"simultaneous\",   nsim = 10000, ncores = 2 ) # \\donttest{ x2_sint #> # A tibble: 25 x 4 #>    term  lower   est upper #>    <chr> <dbl> <dbl> <dbl> #>  1 s(x2)  -24.  -15.  -5.6 #>  2 s(x2)  -35.  -26. -16.  #>  3 s(x2)  -41.  -33. -24.  #>  4 s(x2)  -44.  -36. -29.  #>  5 s(x2)  -44.  -36. -28.  #>  6 s(x2)  -42.  -34. -25.  #>  7 s(x2)  -38.  -30. -21.  #>  8 s(x2)  -33.  -24. -16.  #>  9 s(x2)  -27.  -19. -11.  #> 10 s(x2)  -22.  -14.  -5.8 #> # i 15 more rows # }"},{"path":"https://gavinsimpson.github.io/gratia/reference/confint.gam.html","id":null,"dir":"Reference","previous_headings":"","what":"Point-wise and simultaneous confidence intervals for smooths — confint.gam","title":"Point-wise and simultaneous confidence intervals for smooths — confint.gam","text":"Calculates point-wise confidence simultaneous intervals smooth terms fitted GAM.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/confint.gam.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Point-wise and simultaneous confidence intervals for smooths — confint.gam","text":"","code":"# S3 method for class 'gam' confint(   object,   parm,   level = 0.95,   data = newdata,   n = 100,   type = c(\"confidence\", \"simultaneous\"),   nsim = 10000,   shift = FALSE,   transform = FALSE,   unconditional = FALSE,   ncores = 1,   partial_match = FALSE,   ...,   newdata = NULL )  # S3 method for class 'gamm' confint(object, ...)  # S3 method for class 'list' confint(object, ...)"},{"path":"https://gavinsimpson.github.io/gratia/reference/confint.gam.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Point-wise and simultaneous confidence intervals for smooths — confint.gam","text":"object object class \"gam\" \"gamm\". parm parameters (smooth terms) given intervals vector terms. missing, parameters considered, although currently implemented. level numeric, 0 < level < 1; confidence level point-wise simultaneous interval. default 0.95 95% interval. data data frame; new values covariates used model fit. selected smooth(s) wil evaluated supplied values. n numeric; number points evaluate smooths . type character; type interval compute. One \"confidence\" point-wise intervals, \"simultaneous\" simultaneous intervals. nsim integer; number simulations used computing simultaneous intervals. shift logical; constant term add smooth? transform logical; smooth evaluated transformed scale? generalised models, involves applying inverse link function used fit model. Alternatively, name , actual, function can supplied transform smooth confidence interval. unconditional logical; TRUE (freq == FALSE) Bayesian smoothing parameter uncertainty corrected covariance matrix returned, available. ncores number cores generating random variables multivariate normal distribution. Passed mvnfast::rmvn(). Parallelization take place OpenMP supported (appears work Windows current R). partial_match logical; matching parm use partial match exact match? Can used length(parm) 1. ... additional arguments methods newdata DEPRECATED! data frame; containing new values covariates used model fit. selected smooth(s) wil evaluated supplied values.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/confint.gam.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Point-wise and simultaneous confidence intervals for smooths — confint.gam","text":"tibble components: .smooth; character indicating term row relates, .type; type smooth, .name variable smooth, NA otherwise, one vectors values smooth evaluated, named per variables smooth, zero variables containing values variable, .estimate; estimated value smooth, .se; standard error estimated value smooth, .crit; critical value 100 * level% confidence interval. .lower_ci; lower limit confidence simultaneous interval, .upper_ci; upper limit confidence simultaneous interval,","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/confint.gam.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Point-wise and simultaneous confidence intervals for smooths — confint.gam","text":"Gavin L. Simpson","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/confint.gam.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Point-wise and simultaneous confidence intervals for smooths — confint.gam","text":"","code":"load_mgcv() dat <- data_sim(\"eg1\", n = 1000, dist = \"normal\", scale = 2, seed = 2) mod <- gam(y ~ s(x0) + s(x1) + s(x2) + s(x3), data = dat, method = \"REML\")  # new data to evaluate the smooths at, say over the middle 50% of range # of each covariate middle <- function(x, n = 50, coverage = 0.5) {   v <- (1 - coverage) / 2   q <- quantile(x, prob = c(0 + v, 1 - v), type = 8)   seq(q[1], q[2], length = n) } new_data <- sapply(dat[c(\"x0\", \"x1\", \"x2\", \"x3\")], middle) new_data <- data.frame(new_data)  ## point-wise interval for smooth of x2 ci <- confint(mod, parm = \"s(x2)\", type = \"confidence\", data = new_data) ci #> # A tibble: 50 x 9 #>    .smooth .type .by      x2 .estimate   .se .crit .lower_ci .upper_ci #>    <chr>   <chr> <chr> <dbl>     <dbl> <dbl> <dbl>     <dbl>     <dbl> #>  1 s(x2)   TPRS  NA     0.26       5.3  0.18   2.0       5.0       5.7 #>  2 s(x2)   TPRS  NA     0.27       5.1  0.18   2.0       4.8       5.5 #>  3 s(x2)   TPRS  NA     0.28       4.9  0.18   2.0       4.6       5.3 #>  4 s(x2)   TPRS  NA     0.29       4.6  0.18   2.0       4.3       5.0 #>  5 s(x2)   TPRS  NA     0.30       4.3  0.19   2.0       3.9       4.7 #>  6 s(x2)   TPRS  NA     0.32       4.0  0.19   2.0       3.6       4.3 #>  7 s(x2)   TPRS  NA     0.33       3.6  0.20   2.0       3.2       4.0 #>  8 s(x2)   TPRS  NA     0.34       3.2  0.20   2.0       2.9       3.6 #>  9 s(x2)   TPRS  NA     0.35       2.9  0.20   2.0       2.5       3.3 #> 10 s(x2)   TPRS  NA     0.36       2.5  0.19   2.0       2.1       2.9 #> # i 40 more rows"},{"path":"https://gavinsimpson.github.io/gratia/reference/data_combos.html","id":null,"dir":"Reference","previous_headings":"","what":"All combinations of factor levels plus typical values of continuous variables — data_combos","title":"All combinations of factor levels plus typical values of continuous variables — data_combos","text":"combinations factor levels plus typical values continuous variables","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/data_combos.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"All combinations of factor levels plus typical values of continuous variables — data_combos","text":"","code":"data_combos(object, ...)  # S3 method for class 'gam' data_combos(   object,   vars = everything(),   complete = TRUE,   envir = environment(formula(object)),   data = NULL,   ... )"},{"path":"https://gavinsimpson.github.io/gratia/reference/data_combos.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"All combinations of factor levels plus typical values of continuous variables — data_combos","text":"object fitted model object. ... arguments passed methods. vars terms include exclude returned object. Uses tidyselect principles. complete logical; combinations factor levels returned? FALSE, combinations levels observed model retained. envir environment within recreate data used fit object. data optional data frame data used fit mdoel reconstruction data model work.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/data_sim.html","id":null,"dir":"Reference","previous_headings":"","what":"Simulate example data for fitting GAMs — data_sim","title":"Simulate example data for fitting GAMs — data_sim","text":"tidy reimplementation functions implemented mgcv::gamSim() can used fit GAMs. new feature sampling distribution can applied example types.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/data_sim.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Simulate example data for fitting GAMs — data_sim","text":"","code":"data_sim(   model = \"eg1\",   n = 400,   scale = NULL,   theta = 3,   power = 1.5,   dist = c(\"normal\", \"poisson\", \"binary\", \"negbin\", \"tweedie\", \"gamma\", \"ocat\",     \"ordered categorical\"),   n_cat = 4,   cuts = c(-1, 0, 5),   seed = NULL,   gfam_families = c(\"binary\", \"tweedie\", \"normal\") )"},{"path":"https://gavinsimpson.github.io/gratia/reference/data_sim.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Simulate example data for fitting GAMs — data_sim","text":"model character; either \"egX\" X integer 1:7, name model. See Details possible options. n numeric; number observations simulate. scale numeric; level noise use. theta numeric; dispersion parameter \\(\\theta\\) use. default entirely arbitrary, chosen provide simulated data exhibits extra dispersion beyond assumed Poisson. power numeric; Tweedie power parameter. dist character; sampling distribution response variable. \"ordered categorical\" synonym \"ocat\". n_cat integer; number categories categorical response. Currently used distr %% c(\"ocat\", \"ordered categorical\"). cuts numeric; vector cut points latent variable, excluding end points -Inf Inf. Must one fewer number categories: length(cuts) == n_cat - 1. seed numeric; seed random number generator. Passed base::set.seed(). gfam_families character; vector distributions use generating data grouped families use family = gfam(). allowed distributions per dist.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/data_sim.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Simulate example data for fitting GAMs — data_sim","text":"data_sim() can simulate data several underlying models known true functions. available options currently : \"eg1\": four term additive true model. classic Gu & Wahba four univariate term test model. See gw_functions details underlying four functions. \"eg2\": bivariate smooth true model. \"eg3\": example containing continuous smooth (varying coefficient) true model. model \\(\\hat{y}_i = f_2(x_{1i})x_{2i}\\) function \\(f_2()\\) \\(f_2(x) = 0.2 * x^{11} * (10 * (1 - x))^6 + 10 * (10 * x)^3 * (1 - x)^{10}\\). \"eg4\": factor smooth true model. true model contains factor 3 levels, response nth level follows nth Gu & Wabha function (\\(n \\{1, 2, 3}\\)). \"eg5\": additive plus factor true model. response linear combination Gu & Wabha functions 2, 3, 4 (latter null function) plus factor term four levels. \"eg6\": additive plus random effect term true model. ´\"eg7\": version model \"eg1\"`, covariates correlated. \"gwf2\": model response Gu & Wabha's \\(f_2(x_i)\\) plus noise. \"lwf6\": model response Luo & Wabha's \"example 6\" function \\(sin(2(4x-2)) + 2 exp(-256(x-0.5)^2)\\) plus noise. \"gfam\": simulates data use GAMs family = gfam(families). See example mgcv::gfam(). model specified dist ignored gfam_families used specify distributions included simulated data. Can vector families allowed dist. \"ocat\" %% gfam_families (\"ordered categorical\"), 4 classes assumed, changed. Link functions used \"identity\" \"normal\", \"logit\" \"binary\", \"ocat\", \"ordered categorical\", \"exp\" elsewhere. random component providing noise sampling variation can follow one distributions, specified via argument dist \"normal\": Gaussian, \"poisson\": Poisson, \"binary\": Bernoulli, \"negbin\": Negative binomial, \"tweedie\": Tweedie, \"gamma\": gamma , \"ordered categorical\": ordered categorical arguments provide parameters distribution.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/data_sim.html","id":"references","dir":"Reference","previous_headings":"","what":"References","title":"Simulate example data for fitting GAMs — data_sim","text":"Gu, C., Wahba, G., (1993). Smoothing Spline ANOVA Component-Wise Bayesian \"Confidence Intervals.\" J. Comput. Graph. Stat. 2, 97–117. Luo, Z., Wahba, G., (1997). Hybrid adaptive splines. J. . Stat. Assoc. 92, 107–116.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/data_sim.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Simulate example data for fitting GAMs — data_sim","text":"","code":"data_sim(\"eg1\", n = 100, seed = 1) #> # A tibble: 100 x 10 #>          y       x0      x1      x2       x3       f      f0     f1     f2    f3 #>      <dbl>    <dbl>   <dbl>   <dbl>    <dbl>   <dbl>   <dbl>  <dbl>  <dbl> <dbl> #>  1 14.532  0.26551  0.65472 0.26751 0.67371  13.713  1.4814  3.7041 8.5277     0 #>  2 16.113  0.37212  0.35320 0.21865 0.094858 12.735  1.8408  2.0267 8.8680     0 #>  3  9.5835 0.57285  0.27026 0.51680 0.49260   6.4103 1.9478  1.7169 2.7456     0 #>  4 15.687  0.90821  0.99268 0.26895 0.46155  16.349  0.56879 7.2817 8.4980     0 #>  5  8.2216 0.20168  0.63349 0.18117 0.37522  12.792  1.1841  3.5501 8.0578     0 #>  6  9.9034 0.89839  0.21321 0.51858 0.99110   4.9081 0.62765 1.5318 2.7487     0 #>  7  5.9362 0.94468  0.12937 0.56278 0.17635   4.6020 0.34587 1.2953 2.9609     0 #>  8 10.839  0.66080  0.47812 0.12916 0.81344   9.7565 1.7502  2.6019 5.4045     0 #>  9 16.883  0.62911  0.92407 0.25637 0.068447 16.909  1.8377  6.3481 8.7237     0 #> 10  7.3603 0.061786 0.59876 0.71794 0.40045   6.3401 0.38578 3.3119 2.6424     0 #> # i 90 more rows  # an ordered categorical response data_sim(\"eg1\", n = 100, dist = \"ocat\", n_cat = 4, cuts = c(-1, 0, 5)) #> # A tibble: 100 x 11 #>        y      x0        x1      x2       x3         f      f0     f1         f2 #>    <int>   <dbl>     <dbl>   <dbl>    <dbl>     <dbl>   <dbl>  <dbl>      <dbl> #>  1     1 0.93708 0.21716   0.51711 0.44457  -3.5517   0.39280 1.5439 2.7461     #>  2     1 0.28614 0.21657   0.85193 0.060386 -4.7654   1.5653  1.5421 0.36166    #>  3     1 0.83045 0.38895   0.44280 0.32751  -1.7693   1.0157  2.1769 3.2727     #>  4     4 0.64175 0.94246   0.15788 0.87843   7.2150   1.8050  6.5858 7.0588     #>  5     3 0.51910 0.96261   0.44232 0.93060   3.8994   1.9964  6.8566 3.2808     #>  6     1 0.73659 0.73986   0.96773 0.39218  -2.3701   1.4725  4.3917 0.00015734 #>  7     1 0.13467 0.73325   0.48459 0.15885  -0.27657  0.82112 4.3340 2.8028     #>  8     3 0.65699 0.53576   0.25246 0.31995   5.2247   1.7616  2.9198 8.7777     #>  9     3 0.70506 0.0022730 0.25969 0.30697   3.0408   1.5991  1.0046 8.6716     #> 10     2 0.45774 0.60894   0.54202 0.10781  -0.036524 1.9824  3.3800 2.8356     #> # i 90 more rows #> # i 2 more variables: f3 <dbl>, latent <dbl>"},{"path":"https://gavinsimpson.github.io/gratia/reference/data_slice.html","id":null,"dir":"Reference","previous_headings":"","what":"Prepare a data slice through model covariates — data_slice","title":"Prepare a data slice through model covariates — data_slice","text":"Prepare data slice model covariates","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/data_slice.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Prepare a data slice through model covariates — data_slice","text":"","code":"data_slice(object, ...)  # Default S3 method data_slice(object, ...)  # S3 method for class 'data.frame' data_slice(object, ...)  # S3 method for class 'gam' data_slice(object, ..., data = NULL, envir = NULL)  # S3 method for class 'gamm' data_slice(object, ...)  # S3 method for class 'list' data_slice(object, ...)  # S3 method for class 'scam' data_slice(object, ...)"},{"path":"https://gavinsimpson.github.io/gratia/reference/data_slice.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Prepare a data slice through model covariates — data_slice","text":"object R model object. ... <dynamic-dots> User supplied variables defining data slice. Arguments passed via ... need named. data alternative data frame values containing variables needed fit model. NULL, default, data used fit model recovered using model.frame. User-supplied expressions passed ... evaluated data. envir environment within recreate data used fit object.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/data_slice.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Prepare a data slice through model covariates — data_slice","text":"data slice data set results one (covariates) varied systematically () range specified subset values interest, remaining covariates model held fixed, representative values. known reference grid package emmeans data grid marginaleffects package. GAMs, covariates specified via ... take representative values determined data used fit model follows: numeric covariates, value fitting data closest median value used, factor covariates, modal (frequently observed) level used, first level (sorted per vector returned base::levels() several levels observed number times. values already computed calling gam() bam() example can found var.summary component fitted model. Function typical_values() extract values interested. Convenience functions evenly(), ref_level(), level() provided help users specify data slices. ref_level(), level() also ensure factor covariates correct levels, needed mgcv::predict.gam() example. extended discussion data_slice() examples, see vignette(\"data-slices\", package = \"gratia\").","code":""},{"path":[]},{"path":"https://gavinsimpson.github.io/gratia/reference/data_slice.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Prepare a data slice through model covariates — data_slice","text":"","code":"load_mgcv()  # simulate some Gaussian data df <- data_sim(\"eg1\", n = 50, seed = 2)  # fit a GAM with 1 smooth and 1 linear term m <- gam(y ~ s(x2, k = 7) + x1, data = df, method = \"REML\")  # Want to predict over f(x2) while holding `x1` at some value. # Default will use the observation closest to the median for unspecified # variables. ds <- data_slice(m, x2 = evenly(x2, n = 50)) ds #> # A tibble: 50 x 2 #>        x2    x1 #>     <dbl> <dbl> #>  1 0.0228 0.403 #>  2 0.0424 0.403 #>  3 0.0619 0.403 #>  4 0.0815 0.403 #>  5 0.101  0.403 #>  6 0.121  0.403 #>  7 0.140  0.403 #>  8 0.160  0.403 #>  9 0.179  0.403 #> 10 0.199  0.403 #> # i 40 more rows  # for full control, specify the values you want ds <- data_slice(m, x2 = evenly(x2, n = 50), x1 = 0.3)  # or provide an expression (function call) which will be evaluated in the # data frame passed to `data` or `model.frame(object)` ds <- data_slice(m, x2 = evenly(x2, n = 50), x1 = mean(x1))"},{"path":"https://gavinsimpson.github.io/gratia/reference/datagen.html","id":null,"dir":"Reference","previous_headings":"","what":"Generate data over the range of variables used in smooths — datagen","title":"Generate data over the range of variables used in smooths — datagen","text":"smooth GAM, generate new data range variables involved smooth. function deprecated useful narrow use-case. Use data_slice() instead.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/datagen.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Generate data over the range of variables used in smooths — datagen","text":"","code":"datagen(x, ...)  # S3 method for class 'mgcv.smooth' datagen(x, n = 100, data, ...)  # S3 method for class 'fs.interaction' datagen(x, n = 100, data, ...)  # S3 method for class 'gam' datagen(x, smooth = NULL, n = 200, ...)  # S3 method for class 'gamm' datagen(x, ...)  # S3 method for class 'list' datagen(x, ...)"},{"path":"https://gavinsimpson.github.io/gratia/reference/datagen.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Generate data over the range of variables used in smooths — datagen","text":"x object new data required. Currently objects classes \"gam\", \"gamm\" supported, smooths mgcv inheriting class \"mgcv.smooth\". ... arguments passed methods n numeric; number data values generate per term smooth. data data frame; \"mgcv.smooth\" objects, data used fit GAM need supplied.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/datagen.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Generate data over the range of variables used in smooths — datagen","text":"data frame new values spread range observed values.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/datagen.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Generate data over the range of variables used in smooths — datagen","text":"Gavin L. Simpson","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/derivative_samples.html","id":null,"dir":"Reference","previous_headings":"","what":"Posterior expectations of derivatives from an estimated model — derivative_samples","title":"Posterior expectations of derivatives from an estimated model — derivative_samples","text":"Posterior expectations derivatives estimated model","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/derivative_samples.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Posterior expectations of derivatives from an estimated model — derivative_samples","text":"","code":"derivative_samples(object, ...)  # Default S3 method derivative_samples(object, ...)  # S3 method for class 'gamm' derivative_samples(object, ...)  # S3 method for class 'gam' derivative_samples(   object,   focal = NULL,   data = NULL,   order = 1L,   type = c(\"forward\", \"backward\", \"central\"),   scale = c(\"response\", \"linear_predictor\"),   method = c(\"gaussian\", \"mh\", \"inla\", \"user\"),   n = 100,   eps = 1e-07,   n_sim = 10000,   level = lifecycle::deprecated(),   seed = NULL,   envir = environment(formula(object)),   draws = NULL,   mvn_method = c(\"mvnfast\", \"mgcv\"),   ... )"},{"path":"https://gavinsimpson.github.io/gratia/reference/derivative_samples.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Posterior expectations of derivatives from an estimated model — derivative_samples","text":"object R object compute derivatives ... arguments passed methods fitted_samples() focal character; name focal variable. response derivative response respect variable returned. variables involved model held constant values. can missing supplying data, case, focal variable identified one variable constant. data data frame containing values model covariates evaluate first derivatives smooths. supplied, one variable must held constant value. order numeric; order derivative. type character; type finite difference used. One \"forward\", \"backward\", \"central\". scale character; derivative estimated response linear predictor (link) scale? One \"response\" (default), \"linear predictor\". method character; method used draw samples posterior distribution. \"gaussian\" uses Gaussian (Laplace) approximation posterior. \"mh\" uses Metropolis Hastings sample alternates t proposals proposals based shrunken version posterior covariance matrix. \"inla\" uses variant Integrated Nested Laplace Approximation due Wood (2019), (currently implemented). \"user\" allows user-supplied posterior draws (currently implemented). n numeric; number points evaluate derivative (data supplied). eps numeric; finite difference. n_sim integer; number simulations used computing simultaneous intervals. level  seed numeric; random seed simulations. envir environment within recreate data used fit object. draws matrix; user supplied posterior draws used method = \"user\". mvn_method character; one \"mvnfast\" \"mgcv\". default uses mvnfast::rmvn(), can considerably faster generate large numbers MVN random values mgcv::rmvn(), might work marginal fits, covariance matrix close singular.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/derivative_samples.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Posterior expectations of derivatives from an estimated model — derivative_samples","text":"tibble, currently following variables: .derivative: estimated partial derivative, additional columns containing covariate values derivative evaluated.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/derivative_samples.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Posterior expectations of derivatives from an estimated model — derivative_samples","text":"Gavin L. Simpson","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/derivative_samples.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Posterior expectations of derivatives from an estimated model — derivative_samples","text":"","code":"load_mgcv() df <- data_sim(\"eg1\", dist = \"negbin\", scale = 0.25, seed = 42)  # fit the GAM (note: for execution time reasons using bam()) m <- bam(y ~ s(x0) + s(x1) + s(x2) + s(x3),   data = df, family = nb(), method = \"fREML\")  # data slice through data along x2 - all other covariates will be set to # typical values (value closest to median) ds <- data_slice(m, x2 = evenly(x2, n = 200))  # samples from posterior of derivatives fd_samp <- derivative_samples(m,   data = ds, type = \"central\",   focal = \"x2\", eps = 0.01, seed = 21, n_sim = 100 )  # plot the first 20 posterior draws if (requireNamespace(\"ggplot2\") && requireNamespace(\"dplyr\")) {   library(\"ggplot2\")   fd_samp |>     dplyr::filter(.draw <= 20) |>     ggplot(aes(x = x2, y = .derivative, group = .draw)) +     geom_line(alpha = 0.5) }"},{"path":"https://gavinsimpson.github.io/gratia/reference/derivatives.html","id":null,"dir":"Reference","previous_headings":"","what":"Derivatives of estimated smooths via finite differences — derivatives","title":"Derivatives of estimated smooths via finite differences — derivatives","text":"Derivatives estimated smooths via finite differences","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/derivatives.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Derivatives of estimated smooths via finite differences — derivatives","text":"","code":"derivatives(object, ...)  # Default S3 method derivatives(object, ...)  # S3 method for class 'gamm' derivatives(object, ...)  # S3 method for class 'gam' derivatives(   object,   select = NULL,   term = deprecated(),   data = newdata,   order = 1L,   type = c(\"forward\", \"backward\", \"central\"),   n = 100,   eps = 1e-07,   interval = c(\"confidence\", \"simultaneous\"),   n_sim = 10000,   level = 0.95,   unconditional = FALSE,   frequentist = FALSE,   offset = NULL,   ncores = 1,   partial_match = FALSE,   ...,   newdata = NULL )"},{"path":"https://gavinsimpson.github.io/gratia/reference/derivatives.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Derivatives of estimated smooths via finite differences — derivatives","text":"object R object compute derivatives . ... arguments passed methods. select character; select smooth's posterior draw . default (NULL) means posteriors smooths model wil sampled . supplied, character vector requested terms. Can partial match smooth term; see argument partial_match . term Use select instead. data data frame containing values model covariates evaluate first derivatives smooths. order numeric; order derivative. type character; type finite difference used. One \"forward\", \"backward\", \"central\". n numeric; number points evaluate derivative . eps numeric; finite difference. interval character; type interval compute. One \"confidence\" point-wise intervals, \"simultaneous\" simultaneous intervals. n_sim integer; number simulations used computing simultaneous intervals. level numeric; 0 < level < 1; confidence level point-wise simultaneous interval. default 0.95 95% interval. unconditional logical; use smoothness selection-corrected Bayesian covariance matrix? frequentist logical; use frequentist covariance matrix? offset numeric; value use offset term ncores number cores generating random variables multivariate normal distribution. Passed mvnfast::rmvn(). Parallelization take place OpenMP supported (appears work Windows current R). partial_match logical; smooths selected partial matches term? TRUE, term can single string match . newdata Deprecated: use data instead.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/derivatives.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Derivatives of estimated smooths via finite differences — derivatives","text":"tibble, currently following variables: smooth: smooth row refers , var: name variable involved smooth, data: values var derivative evaluated, derivative: estimated derivative, se: standard error estimated derivative, crit: critical value derivative ± (crit * se) gives upper lower bounds requested confidence simultaneous interval (given level), lower: lower bound confidence simultaneous interval, upper: upper bound confidence simultaneous interval.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/derivatives.html","id":"note","dir":"Reference","previous_headings":"","what":"Note","title":"Derivatives of estimated smooths via finite differences — derivatives","text":"derivatives() ignore random effect smooths encounters object.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/derivatives.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Derivatives of estimated smooths via finite differences — derivatives","text":"Gavin L. Simpson","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/derivatives.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Derivatives of estimated smooths via finite differences — derivatives","text":"","code":"load_mgcv() dat <- data_sim(\"eg1\", n = 400, dist = \"normal\", scale = 2, seed = 42) mod <- gam(y ~ s(x0) + s(x1) + s(x2) + s(x3), data = dat, method = \"REML\")  ## first derivatives of all smooths using central finite differences derivatives(mod, type = \"central\") #> # A tibble: 400 x 12 #>    .smooth .by   .fs   .derivative   .se .crit .lower_ci .upper_ci      x0    x1 #>    <chr>   <chr> <chr>       <dbl> <dbl> <dbl>     <dbl>     <dbl>   <dbl> <dbl> #>  1 s(x0)   NA    NA           7.41  3.33  1.96     0.874      13.9 2.39e-4    NA #>  2 s(x0)   NA    NA           7.40  3.33  1.96     0.884      13.9 1.03e-2    NA #>  3 s(x0)   NA    NA           7.39  3.30  1.96     0.929      13.8 2.04e-2    NA #>  4 s(x0)   NA    NA           7.36  3.24  1.96     1.01       13.7 3.04e-2    NA #>  5 s(x0)   NA    NA           7.32  3.15  1.96     1.14       13.5 4.05e-2    NA #>  6 s(x0)   NA    NA           7.26  3.04  1.96     1.30       13.2 5.06e-2    NA #>  7 s(x0)   NA    NA           7.18  2.90  1.96     1.49       12.9 6.06e-2    NA #>  8 s(x0)   NA    NA           7.09  2.76  1.96     1.69       12.5 7.07e-2    NA #>  9 s(x0)   NA    NA           6.99  2.61  1.96     1.87       12.1 8.07e-2    NA #> 10 s(x0)   NA    NA           6.87  2.47  1.96     2.03       11.7 9.08e-2    NA #> # i 390 more rows #> # i 2 more variables: x2 <dbl>, x3 <dbl>  ## derivatives for a selected smooth derivatives(mod, type = \"central\", select = \"s(x1)\") #> # A tibble: 100 x 9 #>    .smooth .by   .fs   .derivative   .se .crit .lower_ci .upper_ci       x1 #>    <chr>   <chr> <chr>       <dbl> <dbl> <dbl>     <dbl>     <dbl>    <dbl> #>  1 s(x1)   NA    NA         -0.907  3.12  1.96     -7.02      5.20 0.000405 #>  2 s(x1)   NA    NA         -0.906  3.11  1.96     -7.01      5.20 0.0105   #>  3 s(x1)   NA    NA         -0.898  3.10  1.96     -6.97      5.17 0.0205   #>  4 s(x1)   NA    NA         -0.880  3.06  1.96     -6.88      5.12 0.0306   #>  5 s(x1)   NA    NA         -0.849  3.00  1.96     -6.73      5.03 0.0406   #>  6 s(x1)   NA    NA         -0.803  2.92  1.96     -6.52      4.92 0.0507   #>  7 s(x1)   NA    NA         -0.740  2.81  1.96     -6.25      4.77 0.0607   #>  8 s(x1)   NA    NA         -0.659  2.69  1.96     -5.93      4.61 0.0708   #>  9 s(x1)   NA    NA         -0.557  2.56  1.96     -5.57      4.46 0.0809   #> 10 s(x1)   NA    NA         -0.436  2.42  1.96     -5.19      4.32 0.0909   #> # i 90 more rows ## or via a partial match derivatives(mod, type = \"central\", select = \"x1\", partial_match = TRUE) #> # A tibble: 100 x 9 #>    .smooth .by   .fs   .derivative   .se .crit .lower_ci .upper_ci       x1 #>    <chr>   <chr> <chr>       <dbl> <dbl> <dbl>     <dbl>     <dbl>    <dbl> #>  1 s(x1)   NA    NA         -0.907  3.12  1.96     -7.02      5.20 0.000405 #>  2 s(x1)   NA    NA         -0.906  3.11  1.96     -7.01      5.20 0.0105   #>  3 s(x1)   NA    NA         -0.898  3.10  1.96     -6.97      5.17 0.0205   #>  4 s(x1)   NA    NA         -0.880  3.06  1.96     -6.88      5.12 0.0306   #>  5 s(x1)   NA    NA         -0.849  3.00  1.96     -6.73      5.03 0.0406   #>  6 s(x1)   NA    NA         -0.803  2.92  1.96     -6.52      4.92 0.0507   #>  7 s(x1)   NA    NA         -0.740  2.81  1.96     -6.25      4.77 0.0607   #>  8 s(x1)   NA    NA         -0.659  2.69  1.96     -5.93      4.61 0.0708   #>  9 s(x1)   NA    NA         -0.557  2.56  1.96     -5.57      4.46 0.0809   #> 10 s(x1)   NA    NA         -0.436  2.42  1.96     -5.19      4.32 0.0909   #> # i 90 more rows"},{"path":"https://gavinsimpson.github.io/gratia/reference/difference_smooths.html","id":null,"dir":"Reference","previous_headings":"","what":"Differences of factor smooth interactions — difference_smooths","title":"Differences of factor smooth interactions — difference_smooths","text":"Estimates pairwise differences (comparisons) factor smooth interactions (smooths factor argument) pairs groups defined factor. group means can optionally included difference.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/difference_smooths.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Differences of factor smooth interactions — difference_smooths","text":"","code":"difference_smooths(model, ...)  # S3 method for class 'gam' difference_smooths(   model,   select = NULL,   smooth = deprecated(),   n = 100,   ci_level = 0.95,   data = NULL,   group_means = FALSE,   partial_match = TRUE,   unconditional = FALSE,   frequentist = FALSE,   ... )"},{"path":"https://gavinsimpson.github.io/gratia/reference/difference_smooths.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Differences of factor smooth interactions — difference_smooths","text":"model fitted model. ... arguments passed methods. currently used. select character, logical, numeric; smooths plot. NULL, default, model smooths drawn. Numeric select indexes smooths order specified formula stored object. Character select matches labels smooths shown example output summary(object). Logical select operates per numeric select order smooths stored. smooth Use select instead. n numeric; number points evaluate difference pairs smooths. ci_level numeric 0 1; coverage credible interval. data data frame locations evaluate difference smooths. group_means logical; group means included difference? partial_match logical; smooth match partially smooths? partial_match = TRUE, smooth must single string, character vector length 1. Unlike similar functions, default TRUE intention users matching factor-smooth labels. unconditional logical; account smoothness selection model? frequentist logical; use frequentist covariance matrix?","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/difference_smooths.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Differences of factor smooth interactions — difference_smooths","text":"","code":"load_mgcv() df <- data_sim(\"eg4\", seed = 42) m <- gam(y ~ fac + s(x2, by = fac) + s(x0), data = df, method = \"REML\")  sm_dif <- difference_smooths(m, select = \"s(x2)\") sm_dif #> # A tibble: 300 x 9 #>    .smooth .by   .level_1 .level_2 .diff   .se .lower_ci .upper_ci      x2 #>    <chr>   <chr> <chr>    <chr>    <dbl> <dbl>     <dbl>     <dbl>   <dbl> #>  1 s(x2)   fac   1        2        0.386 0.618  -0.824        1.60 0.00359 #>  2 s(x2)   fac   1        2        0.479 0.574  -0.646        1.60 0.0136  #>  3 s(x2)   fac   1        2        0.572 0.534  -0.474        1.62 0.0237  #>  4 s(x2)   fac   1        2        0.665 0.497  -0.308        1.64 0.0338  #>  5 s(x2)   fac   1        2        0.758 0.464  -0.151        1.67 0.0438  #>  6 s(x2)   fac   1        2        0.850 0.435  -0.00342      1.70 0.0539  #>  7 s(x2)   fac   1        2        0.941 0.412   0.134        1.75 0.0639  #>  8 s(x2)   fac   1        2        1.03  0.393   0.262        1.80 0.0740  #>  9 s(x2)   fac   1        2        1.12  0.378   0.380        1.86 0.0841  #> 10 s(x2)   fac   1        2        1.21  0.367   0.489        1.93 0.0941  #> # i 290 more rows  draw(sm_dif)   # include the groups means for `fac` in the difference sm_dif2 <- difference_smooths(m, select = \"s(x2)\", group_means = TRUE) draw(sm_dif2)"},{"path":[]},{"path":"https://gavinsimpson.github.io/gratia/reference/dispersion.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Dispersion parameter for fitted model — dispersion","text":"","code":"dispersion(model, ...)  # S3 method for class 'gam' dispersion(model, ...)  # S3 method for class 'glm' dispersion(model, ...)"},{"path":"https://gavinsimpson.github.io/gratia/reference/dispersion.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Dispersion parameter for fitted model — dispersion","text":"model fitted model. ... arguments passed methods.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.basis.html","id":null,"dir":"Reference","previous_headings":"","what":"Plot basis functions — draw.basis","title":"Plot basis functions — draw.basis","text":"Plots basis functions using ggplot2","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.basis.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Plot basis functions — draw.basis","text":"","code":"# S3 method for class 'basis' draw(   object,   legend = FALSE,   labeller = NULL,   ylab = NULL,   title = NULL,   subtitle = NULL,   caption = NULL,   ncol = NULL,   nrow = NULL,   angle = NULL,   guides = \"keep\",   contour = FALSE,   n_contour = 10,   contour_col = \"black\",   ... )"},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.basis.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Plot basis functions — draw.basis","text":"object object, result call basis(). legend logical; legend drawn indicate basis functions? labeller labeller function label facets. default use ggplot2::label_both(). ylab character expression; label y axis. supplied, suitable label generated object. title character expression; title plot. See ggplot2::labs(). subtitle character expression; subtitle plot. See ggplot2::labs(). caption character expression; plot caption. See ggplot2::labs(). ncol, nrow numeric; numbers rows columns spread plots angle numeric; angle x axis tick labels drawn passed angle argument ggplot2::guide_axis(). guides character; one \"keep\" (default), \"collect\", \"auto\". Passed patchwork::plot_layout() contour logical; contours draw plot using ggplot2::geom_contour(). n_contour numeric; number contour bins. result n_contour - 1 contour lines drawn. See ggplot2::geom_contour(). contour_col colour specification contour lines. ... arguments passed methods. used method.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.basis.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Plot basis functions — draw.basis","text":"patchwork object.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.basis.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Plot basis functions — draw.basis","text":"Gavin L. Simpson","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.basis.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Plot basis functions — draw.basis","text":"","code":"load_mgcv() df <- data_sim(\"eg1\", n = 400, seed = 42) m <- gam(y ~ s(x0) + s(x1) + s(x2) + s(x3), data = df, method = \"REML\")  bf <- basis(m) draw(bf)   bf <- basis(m, \"s(x2)\") draw(bf)"},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.compare_smooths.html","id":null,"dir":"Reference","previous_headings":"","what":"Plot comparisons of smooths — draw.compare_smooths","title":"Plot comparisons of smooths — draw.compare_smooths","text":"Plot comparisons smooths","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.compare_smooths.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Plot comparisons of smooths — draw.compare_smooths","text":"","code":"# S3 method for class 'compare_smooths' draw(object, ncol = NULL, nrow = NULL, guides = \"collect\", ...)"},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.compare_smooths.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Plot comparisons of smooths — draw.compare_smooths","text":"object class \"compare_smooths\", result call compare_smooths(). ncol, nrow numeric; numbers rows columns spread plots guides character; one \"keep\" (default), \"collect\", \"auto\". Passed patchwork::plot_layout() ... additional arguments passed patchwork::wrap_plots().","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.conditional_values.html","id":null,"dir":"Reference","previous_headings":"","what":"Plot conditional predictions — draw.conditional_values","title":"Plot conditional predictions — draw.conditional_values","text":"Plot conditional predictions","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.conditional_values.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Plot conditional predictions — draw.conditional_values","text":"","code":"# S3 method for class 'conditional_values' draw(   object,   facet_scales = \"fixed\",   discrete_colour = NULL,   discrete_fill = NULL,   ... )"},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.conditional_values.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Plot conditional predictions — draw.conditional_values","text":"object object class \"conditional_values\", result call conditional_values(). facet_scales character; facets axis scales across facets? See ggplot2::facet_wrap() details. Options : \"fixed\" (default), \"free_x\", \"free_y\", \"free\". discrete_colour suitable colour scale used plotting discrete variables. discrete_fill suitable fill scale used plotting discrete variables. ... additional arguments passed patchwork::wrap_plots().","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.derivatives.html","id":null,"dir":"Reference","previous_headings":"","what":"Plot derivatives of smooths — draw.derivatives","title":"Plot derivatives of smooths — draw.derivatives","text":"Plot derivatives smooths","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.derivatives.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Plot derivatives of smooths — draw.derivatives","text":"","code":"# S3 method for class 'derivatives' draw(   object,   select = NULL,   scales = c(\"free\", \"fixed\"),   add_change = FALSE,   change_type = c(\"change\", \"sizer\"),   alpha = 0.2,   change_col = \"black\",   decrease_col = \"#56B4E9\",   increase_col = \"#E69F00\",   lwd_change = 1.5,   ncol = NULL,   nrow = NULL,   guides = \"keep\",   angle = NULL,   ... )  # S3 method for class 'partial_derivatives' draw(   object,   select = NULL,   scales = c(\"free\", \"fixed\"),   alpha = 0.2,   ncol = NULL,   nrow = NULL,   guides = \"keep\",   angle = NULL,   ... )"},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.derivatives.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Plot derivatives of smooths — draw.derivatives","text":"object fitted GAM, result call mgcv::gam(). select character, logical, numeric; smooths plot. NULL, default, model smooths drawn. Numeric select indexes smooths order specified formula stored object. Character select matches labels smooths shown example output summary(object). Logical select operates per numeric select order smooths stored. scales character; univariate smooths plotted y-axis scale? scales = \"free\", default, univariate smooth y-axis scale. scales = \"fixed\", common y axis scale used univariate smooths. Currently affect y-axis scale plots parametric terms. add_change logical; periods significant change highlighted plot? change_type character; type change indicate. \"change\", differentiation made periods significant increase decrease. \"sizer\", periods increase decrease differentiated resulting plot. alpha numeric; alpha transparency confidence simultaneous interval. change_col, decrease_col, increase_col colour specifications use indicating periods change. col_change used change_type = \"change\", col_decrease col_increase used `change_type = \"sizer\"“. lwd_change numeric; linewidth use change indicators. ncol, nrow numeric; numbers rows columns spread plots guides character; one \"keep\" (default), \"collect\", \"auto\". Passed patchwork::plot_layout() angle numeric; angle x axis tick labels drawn passed angle argument ggplot2::guide_axis(). ... additional arguments passed patchwork::wrap_plots().","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.derivatives.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Plot derivatives of smooths — draw.derivatives","text":"","code":"load_mgcv() dat <- data_sim(\"eg1\", n = 800, dist = \"normal\", scale = 2, seed = 42) mod <- gam(y ~ s(x0) + s(x1) + s(x2) + s(x3), data = dat, method = \"REML\")  ## first derivative of all smooths df <- derivatives(mod, type = \"central\") draw(df)  ## fixed axis scales draw(df, scales = \"fixed\")"},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.difference_smooth.html","id":null,"dir":"Reference","previous_headings":"","what":"Plot differences of smooths — draw.difference_smooth","title":"Plot differences of smooths — draw.difference_smooth","text":"Plot differences smooths","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.difference_smooth.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Plot differences of smooths — draw.difference_smooth","text":"","code":"# S3 method for class 'difference_smooth' draw(   object,   select = NULL,   rug = FALSE,   ref_line = FALSE,   contour = FALSE,   contour_col = \"black\",   n_contour = NULL,   ci_alpha = 0.2,   ci_col = \"black\",   smooth_col = \"black\",   line_col = \"red\",   scales = c(\"free\", \"fixed\"),   ncol = NULL,   nrow = NULL,   guides = \"keep\",   xlab = NULL,   ylab = NULL,   title = NULL,   subtitle = NULL,   caption = NULL,   angle = NULL,   ... )"},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.difference_smooth.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Plot differences of smooths — draw.difference_smooth","text":"object fitted GAM, result call mgcv::gam(). select character, logical, numeric; smooths plot. NULL, default, model smooths drawn. Numeric select indexes smooths order specified formula stored object. Character select matches labels smooths shown example output summary(object). Logical select operates per numeric select order smooths stored. rug logical; ref_line logical; contour logical; contour lines added smooth surfaces? contour_col colour specification contour lines. n_contour numeric; number contour bins. result n_contour - 1 contour lines drawn. See ggplot2::geom_contour(). ci_alpha numeric; alpha transparency confidence simultaneous interval. ci_col colour specification confidence/credible intervals band. Affects fill interval. smooth_col colour specification smooth difference line. line_col colour specification drawing reference lines scales character; univariate smooths plotted y-axis scale? scales = \"free\", default, univariate smooth y-axis scale. scales = \"fixed\", common y axis scale used univariate smooths. Currently affect y-axis scale plots parametric terms. ncol, nrow numeric; numbers rows columns spread plots guides character; one \"keep\" (default), \"collect\", \"auto\". Passed patchwork::plot_layout() xlab, ylab, title, subtitle, caption character; labels annotate plots angle numeric; angle x axis tick labels drawn passed angle argument ggplot2::guide_axis(). ... additional arguments passed patchwork::wrap_plots().","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.difference_smooth.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Plot differences of smooths — draw.difference_smooth","text":"","code":"load_mgcv() # simulate some data; a factor smooth example df <- data_sim(\"eg4\", seed = 42) # fit GAM m <- gam(y ~ fac + s(x2, by = fac) + s(x0), data = df, method = \"REML\")  # calculate the differences between pairs of smooths the f_j(x2) term diffs <- difference_smooths(m, select = \"s(x2)\") draw(diffs)"},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.evaluated_parametric_term.html","id":null,"dir":"Reference","previous_headings":"","what":"Plot estimated parametric effects — draw.evaluated_parametric_term","title":"Plot estimated parametric effects — draw.evaluated_parametric_term","text":"Plots estimated univariate bivariate smooths using ggplot2.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.evaluated_parametric_term.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Plot estimated parametric effects — draw.evaluated_parametric_term","text":"","code":"# S3 method for class 'evaluated_parametric_term' draw(   object,   ci_level = 0.95,   constant = NULL,   fun = NULL,   xlab,   ylab,   title = NULL,   subtitle = NULL,   caption = NULL,   rug = TRUE,   position = \"identity\",   response_range = NULL,   ... )"},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.evaluated_parametric_term.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Plot estimated parametric effects — draw.evaluated_parametric_term","text":"object object, result call evaluate_parametric_term(). ci_level numeric 0 1; coverage credible interval. constant numeric; constant add estimated values smooth. constant, supplied, added estimated value confidence band computed. fun function; function applied estimated values confidence interval plotting. Can function name function. Function fun applied adding constant, provided. xlab character expression; label x axis. supplied, suitable label generated object. ylab character expression; label y axis. supplied, suitable label generated object. title character expression; title plot. See ggplot2::labs(). subtitle character expression; subtitle plot. See ggplot2::labs(). caption character expression; plot caption. See ggplot2::labs(). rug evaluate_parametric_terms(), logical indicate rug plot drawn. position Position adjustment, either string, result call position adjustment function. response_range numeric; vector two values giving range response data guide. Used fix plots common scale/range. Ignored show set \"se\". ... arguments passed methods.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.evaluated_parametric_term.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Plot estimated parametric effects — draw.evaluated_parametric_term","text":"ggplot2::ggplot() object.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.evaluated_parametric_term.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Plot estimated parametric effects — draw.evaluated_parametric_term","text":"Gavin L. Simpson","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.gam.html","id":null,"dir":"Reference","previous_headings":"","what":"Plot estimated smooths from a fitted GAM — draw.gam","title":"Plot estimated smooths from a fitted GAM — draw.gam","text":"Plots estimated smooths fitted GAM model similar way mgcv::plot.gam() instead using base graphics, ggplot2::ggplot() used instead.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.gam.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Plot estimated smooths from a fitted GAM — draw.gam","text":"","code":"# S3 method for class 'gam' draw(   object,   data = NULL,   select = NULL,   parametric = FALSE,   terms = NULL,   residuals = FALSE,   scales = c(\"free\", \"fixed\"),   ci_level = 0.95,   n = 100,   n_3d = 16,   n_4d = 4,   unconditional = FALSE,   overall_uncertainty = TRUE,   constant = NULL,   fun = NULL,   dist = 0.1,   rug = TRUE,   contour = TRUE,   grouped_by = FALSE,   ci_alpha = 0.2,   ci_col = \"black\",   smooth_col = \"black\",   resid_col = \"steelblue3\",   contour_col = \"black\",   n_contour = NULL,   partial_match = FALSE,   discrete_colour = NULL,   discrete_fill = NULL,   continuous_colour = NULL,   continuous_fill = NULL,   position = \"identity\",   angle = NULL,   ncol = NULL,   nrow = NULL,   guides = \"keep\",   widths = NULL,   heights = NULL,   crs = NULL,   default_crs = NULL,   lims_method = \"cross\",   wrap = TRUE,   caption = TRUE,   envir = environment(formula(object)),   ... )"},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.gam.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Plot estimated smooths from a fitted GAM — draw.gam","text":"object fitted GAM, result call mgcv::gam(). data optional data frame used supply data smooths evaluated plotted. usually needed, option need fine control exactly data used plotting. select character, logical, numeric; smooths plot. NULL, default, model smooths drawn. Numeric select indexes smooths order specified formula stored object. Character select matches labels smooths shown example output summary(object). Logical select operates per numeric select order smooths stored. parametric logical; plot parametric terms also? Note select used selecting smooths plot. terms argument used select parametric effects plotted. default, mgcv::plot.gam(), draw parametric effects. terms character; model parametric terms drawn? Default NULL plot parametric terms can drawn. residuals logical; partial residuals smooth drawn? Ignored anything simple univariate smooth. scales character; univariate smooths plotted y-axis scale? scales = \"free\", default, univariate smooth y-axis scale. scales = \"fixed\", common y axis scale used univariate smooths. Currently affect y-axis scale plots parametric terms. ci_level numeric 0 1; coverage credible interval. n numeric; number points range covariate evaluate smooth. n_3d numeric; number new observations generate third dimension 3D smooth. n_4d numeric; number new observations generate dimensions higher 2 (!) kD smooth (k >= 4). example, smooth 4D smooth, dimensions 3 4 get n_4d new observations. unconditional logical; confidence intervals include uncertainty due smoothness selection? TRUE, corrected Bayesian covariance matrix used. overall_uncertainty logical; uncertainty model constant term included standard error evaluate values smooth? constant numeric; constant add estimated values smooth. constant, supplied, added estimated value confidence band computed. fun function; function applied estimated values confidence interval plotting. Can function name function. Function fun applied adding constant, provided. dist numeric; greater 0, used determine location far data plotted plotting 2-D smooths. data scaled unit square deciding exclude, dist distance within unit square. See mgcv::exclude..far() details. rug logical; draw rug plot bottom plot 1-D smooths plot locations data higher dimensions. contour logical; contours draw plot using ggplot2::geom_contour(). grouped_by logical; factor smooths drawn one panel per level factor (FALSE, default), individual smooths combined single panel containing levels (TRUE)? ci_alpha numeric; alpha transparency confidence simultaneous interval. ci_col colour specification confidence/credible intervals band. Affects fill interval. smooth_col colour specification smooth line. resid_col colour specification partial residuals. contour_col colour specification contour lines. n_contour numeric; number contour bins. result n_contour - 1 contour lines drawn. See ggplot2::geom_contour(). partial_match logical; smooths selected partial matches select? TRUE, select can single string match . discrete_colour suitable colour scale used plotting discrete variables. discrete_fill suitable fill scale used plotting discrete variables. continuous_colour suitable colour scale used plotting continuous variables. continuous_fill suitable fill scale used plotting continuous variables. position Position adjustment, either string, result call position adjustment function. angle numeric; angle x axis tick labels drawn passed angle argument ggplot2::guide_axis(). ncol, nrow numeric; numbers rows columns spread plots guides character; one \"keep\" (default), \"collect\", \"auto\". Passed patchwork::plot_layout() widths, heights relative widths heights column row grid. get repeated match dimensions grid. 1 plot widths = NULL, value widths set internally widths = 1 accommodate plots smooths use fixed aspect ratio. crs coordinate reference system (CRS) use plot. data projected CRS. See ggplot2::coord_sf() details. default_crs coordinate reference system (CRS) use non-sf layers plot. left default NULL, CRS used 4326 (WGS84), appropriate spline---sphere smooths, parameterized terms latitude longitude coordinates. See ggplot2::coord_sf() details. lims_method character; affects axis limits determined. See ggplot2::coord_sf(). careful; testing examples, changing \"orthogonal\" example chlorophyll-example Simon Wood's GAM book quickly used RAM test system OS killed R. incorrect usage part; right now grid points SOS smooths evaluated (supplied user) can produce invalid coordinates corners tiles grid generated tile centres without respect spacing tiles. wrap logical; wrap plots patchwork? FALSE, list ggplot objects returned, 1 per term plotted. caption logical; show smooth type caption plot? envir environment look data within. ... additional arguments passed patchwork::wrap_plots().","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.gam.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Plot estimated smooths from a fitted GAM — draw.gam","text":"object returned created patchwork::wrap_plots().","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.gam.html","id":"note","dir":"Reference","previous_headings":"","what":"Note","title":"Plot estimated smooths from a fitted GAM — draw.gam","text":"Internally, plots smooth created using ggplot2::ggplot() composed single plot using patchwork::wrap_plots(). result, possible use + add plots way one might typically work ggplot() plots. Instead, use & operator; see examples.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.gam.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Plot estimated smooths from a fitted GAM — draw.gam","text":"Gavin L. Simpson","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.gam.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Plot estimated smooths from a fitted GAM — draw.gam","text":"","code":"load_mgcv()  # simulate some data df1 <- data_sim(\"eg1\", n = 400, dist = \"normal\", scale = 2, seed = 2) # fit GAM m1 <- gam(y ~ s(x0) + s(x1) + s(x2) + s(x3), data = df1, method = \"REML\")  # plot all smooths draw(m1)   # can add partial residuals draw(m1, residuals = TRUE)   df2 <- data_sim(\"eg2\", n = 1000, dist = \"normal\", scale = 1, seed = 2) m2 <- gam(y ~ s(x, z, k = 40), data = df2, method = \"REML\") draw(m2, contour = FALSE, n = 50)   # See https://gavinsimpson.github.io/gratia/articles/custom-plotting.html # for more examples and for details on how to modify the theme of all the # plots produced by draw(). To modify all panels, for example to change the # theme, use the & operator"},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.gamlss.html","id":null,"dir":"Reference","previous_headings":"","what":"Plot smooths of a GAMLSS model estimated by GJRM::gamlss — draw.gamlss","title":"Plot smooths of a GAMLSS model estimated by GJRM::gamlss — draw.gamlss","text":"Provides draw() method GAMLSS (distributional GAMs) fitted GJRM::gamlss().","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.gamlss.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Plot smooths of a GAMLSS model estimated by GJRM::gamlss — draw.gamlss","text":"","code":"# S3 method for class 'gamlss' draw(   object,   scales = c(\"free\", \"fixed\"),   ncol = NULL,   nrow = NULL,   guides = \"keep\",   widths = NULL,   heights = NULL,   ... )"},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.gamlss.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Plot smooths of a GAMLSS model estimated by GJRM::gamlss — draw.gamlss","text":"object model, fitted GJRM::gamlss() scales character; univariate smooths plotted y-axis scale? scales = \"free\", default, univariate smooth y-axis scale. scales = \"fixed\", common y axis scale used univariate smooths. Currently affect y-axis scale plots parametric terms. ncol, nrow numeric; numbers rows columns spread plots guides character; one \"keep\" (default), \"collect\", \"auto\". Passed patchwork::plot_layout() widths, heights relative widths heights column row grid. get repeated match dimensions grid. 1 plot widths = NULL, value widths set internally widths = 1 accommodate plots smooths use fixed aspect ratio. ... arguments passed draw.gam()","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.gamlss.html","id":"note","dir":"Reference","previous_headings":"","what":"Note","title":"Plot smooths of a GAMLSS model estimated by GJRM::gamlss — draw.gamlss","text":"Plots smooths labelled linear predictor belong.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.gamlss.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Plot smooths of a GAMLSS model estimated by GJRM::gamlss — draw.gamlss","text":"","code":"if (require(\"GJRM\", quietly = TRUE)) {   # follow example from ?GJRM::gamlss   load_mgcv()   suppressPackageStartupMessages(library(\"GJRM\"))   set.seed(0)   n <- 100   x1 <- round(runif(n))   x2 <- runif(n)   x3 <- runif(n)   f1 <- function(x) cos(pi * 2 * x) + sin(pi * x)   y1 <- -1.55 + 2 * x1 + f1(x2) + rnorm(n)   dataSim <- data.frame(y1, x1, x2, x3)    eq_mu <- y1 ~ x1 + s(x2)   eq_s <- ~ s(x3, k = 6)   fl <- list(eq_mu, eq_s)   m <- gamlss(fl, data = dataSim)    draw(m) }"},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.html","id":null,"dir":"Reference","previous_headings":"","what":"Generic plotting via ggplot2 — draw","title":"Generic plotting via ggplot2 — draw","text":"Generic plotting via ggplot2","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Generic plotting via ggplot2 — draw","text":"","code":"draw(object, ...)"},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Generic plotting via ggplot2 — draw","text":"object R object plot. ... arguments passed methods.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Generic plotting via ggplot2 — draw","text":"ggplot2::ggplot() object.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Generic plotting via ggplot2 — draw","text":"Generic function plotting R objects uses ggplot2 package.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Generic plotting via ggplot2 — draw","text":"Gavin L. Simpson","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.mgcv_smooth.html","id":null,"dir":"Reference","previous_headings":"","what":"Plot basis functions — draw.mgcv_smooth","title":"Plot basis functions — draw.mgcv_smooth","text":"Plots basis functions using ggplot2","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.mgcv_smooth.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Plot basis functions — draw.mgcv_smooth","text":"","code":"# S3 method for class 'mgcv_smooth' draw(   object,   legend = FALSE,   use_facets = TRUE,   labeller = NULL,   xlab,   ylab,   title = NULL,   subtitle = NULL,   caption = NULL,   angle = NULL,   ... )"},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.mgcv_smooth.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Plot basis functions — draw.mgcv_smooth","text":"object object, result call basis(). legend logical; legend drawn indicate basis functions? use_facets logical; factor smooths, use facets show basis functions level factor? FALSE, separate ggplot object created level combined using patchwork::wrap_plots(). Currently ignored. labeller labeller function label facets. default use ggplot2::label_both(). xlab character expression; label x axis. supplied, suitable label generated object. ylab character expression; label y axis. supplied, suitable label generated object. title character expression; title plot. See ggplot2::labs(). subtitle character expression; subtitle plot. See ggplot2::labs(). caption character expression; plot caption. See ggplot2::labs(). angle numeric; angle x axis tick labels drawn passed angle argument ggplot2::guide_axis(). ... arguments passed methods. used method.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.mgcv_smooth.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Plot basis functions — draw.mgcv_smooth","text":"ggplot2::ggplot() object.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.mgcv_smooth.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Plot basis functions — draw.mgcv_smooth","text":"Gavin L. Simpson","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.mgcv_smooth.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Plot basis functions — draw.mgcv_smooth","text":"","code":"load_mgcv() df <- data_sim(\"eg4\", n = 400, seed = 42)  bf <- basis(s(x0), data = df) draw(bf)   bf <- basis(s(x2, by = fac, bs = \"bs\"), data = df) draw(bf)"},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.pairwise_concurvity.html","id":null,"dir":"Reference","previous_headings":"","what":"Plot concurvity measures — draw.pairwise_concurvity","title":"Plot concurvity measures — draw.pairwise_concurvity","text":"Plot concurvity measures","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.pairwise_concurvity.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Plot concurvity measures — draw.pairwise_concurvity","text":"","code":"# S3 method for class 'pairwise_concurvity' draw(   object,   title = \"Smooth-wise concurvity\",   subtitle = NULL,   caption = NULL,   x_lab = \"Term\",   y_lab = \"With\",   fill_lab = \"Concurvity\",   continuous_colour = NULL,   ... )  # S3 method for class 'overall_concurvity' draw(   object,   title = \"Overall concurvity\",   subtitle = NULL,   caption = NULL,   y_lab = \"Concurvity\",   x_lab = NULL,   bar_col = \"steelblue\",   bar_fill = \"steelblue\",   ... )"},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.pairwise_concurvity.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Plot concurvity measures — draw.pairwise_concurvity","text":"object object inheriting class \"concurvity\", usually result call model_concurvity() abbreviated form concrvity(). title character; plot title. subtitle character; plot subtitle. caption character; plot caption x_lab character; label x axis. y_lab character; label y axis. fill_lab character; label use fill guide. continuous_colour function; continuous colour (fill) scale use. ... arguments passed methods. bar_col colour specification bar colour. bar_fill colour specification bar fill","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.parametric_effects.html","id":null,"dir":"Reference","previous_headings":"","what":"Plot estimated effects for model parametric terms — draw.parametric_effects","title":"Plot estimated effects for model parametric terms — draw.parametric_effects","text":"Plot estimated effects model parametric terms","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.parametric_effects.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Plot estimated effects for model parametric terms — draw.parametric_effects","text":"","code":"# S3 method for class 'parametric_effects' draw(   object,   scales = c(\"free\", \"fixed\"),   ci_level = 0.95,   ci_col = \"black\",   ci_alpha = 0.2,   line_col = \"black\",   constant = NULL,   fun = NULL,   rug = TRUE,   position = \"identity\",   angle = NULL,   ...,   ncol = NULL,   nrow = NULL,   guides = \"keep\" )"},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.parametric_effects.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Plot estimated effects for model parametric terms — draw.parametric_effects","text":"object fitted GAM, result call mgcv::gam(). scales character; univariate smooths plotted y-axis scale? scales = \"free\", default, univariate smooth y-axis scale. scales = \"fixed\", common y axis scale used univariate smooths. Currently affect y-axis scale plots parametric terms. ci_level numeric 0 1; coverage credible interval. ci_col colour specification confidence/credible intervals band. Affects fill interval. ci_alpha numeric; alpha transparency confidence simultaneous interval. line_col colour specification used regression lines linear continuous terms. constant numeric; constant add estimated values smooth. constant, supplied, added estimated value confidence band computed. fun function; function applied estimated values confidence interval plotting. Can function name function. Function fun applied adding constant, provided. rug logical; draw rug plot bottom plot 1-D smooths plot locations data higher dimensions. position Position adjustment, either string, result call position adjustment function. angle numeric; angle x axis tick labels drawn passed angle argument ggplot2::guide_axis(). ... additional arguments passed patchwork::wrap_plots(). ncol, nrow numeric; numbers rows columns spread plots guides character; one \"keep\" (default), \"collect\", \"auto\". Passed patchwork::plot_layout()","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.penalty_df.html","id":null,"dir":"Reference","previous_headings":"","what":"Display penalty matrices of smooths using ggplot — draw.penalty_df","title":"Display penalty matrices of smooths using ggplot — draw.penalty_df","text":"Displays penalty matrices smooths heatmap using ggplot","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.penalty_df.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Display penalty matrices of smooths using ggplot — draw.penalty_df","text":"","code":"# S3 method for class 'penalty_df' draw(   object,   normalize = FALSE,   as_matrix = TRUE,   continuous_fill = NULL,   xlab = NULL,   ylab = NULL,   title = NULL,   subtitle = NULL,   caption = NULL,   ncol = NULL,   nrow = NULL,   guides = \"keep\",   ... )"},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.penalty_df.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Display penalty matrices of smooths using ggplot — draw.penalty_df","text":"object fitted GAM, result call mgcv::gam(). normalize logical; normalize penalty range -1, 1? as_matrix logical; plotted penalty matrix oriented? TRUE row 1, column 1 penalty matrix draw upper left, whereas, FALSE drawn lower left plot. continuous_fill suitable fill scale used plotting continuous variables. xlab character expression; label x axis. supplied, axis label drawn. May vector, one per penalty. ylab character expression; label y axis.  supplied, axis label drawn. May vector, one per penalty. title character expression; title plot. See ggplot2::labs(). May vector, one per penalty. subtitle character expression; subtitle plot. See ggplot2::labs(). May vector, one per penalty. caption character expression; plot caption. See ggplot2::labs(). May vector, one per penalty. ncol, nrow numeric; numbers rows columns spread plots. guides character; one \"keep\" (default), \"collect\", \"auto\". Passed patchwork::plot_layout() ... additional arguments passed patchwork::wrap_plots().","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.penalty_df.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Display penalty matrices of smooths using ggplot — draw.penalty_df","text":"","code":"load_mgcv() dat <- data_sim(\"eg4\", n = 400, seed = 42) m <- gam(y ~ s(x0) + s(x1, bs = \"cr\") + s(x2, bs = \"bs\", by = fac),   data = dat, method = \"REML\" )  ## produce a multi-panel plot of all penalties draw(penalty(m))   # for a specific smooth draw(penalty(m, select = \"s(x2):fac1\"))"},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.rootogram.html","id":null,"dir":"Reference","previous_headings":"","what":"Draw a rootogram — draw.rootogram","title":"Draw a rootogram — draw.rootogram","text":"rootogram model diagnostic tool assesses goodness fit statistical model. observed values response compared expected fitted model. discrete, count responses, frequency count (0, 1, 2, etc) observed data expected conditional distribution response implied model compared. continuous variables, observed expected frequencies obtained grouping data bins. rootogram drawn using ggplot2::ggplot() graphics. design closely follows Kleiber & Zeileis (2016).","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.rootogram.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Draw a rootogram — draw.rootogram","text":"","code":"# S3 method for class 'rootogram' draw(   object,   type = c(\"hanging\", \"standing\", \"suspended\"),   sqrt = TRUE,   ref_line = TRUE,   warn_limits = TRUE,   fitted_colour = \"steelblue\",   bar_colour = NA,   bar_fill = \"grey\",   ref_line_colour = \"black\",   warn_line_colour = \"black\",   ylab = NULL,   xlab = NULL,   ... )"},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.rootogram.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Draw a rootogram — draw.rootogram","text":"object R object plot. type character; type rootogram draw. sqrt logical; show observed fitted frequencies ref_line logical; draw reference line zero? warn_limits logical; draw Tukey's warning limit lines +/- 1? fitted_colour, bar_colour, bar_fill, ref_line_colour, warn_line_colour colours used draw respective element rootogram. xlab, ylab character; labels x y axis rootogram. May missing (NULL), case suitable labels used. ' ... arguments passed methods.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.rootogram.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Draw a rootogram — draw.rootogram","text":"'ggplot' object.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.rootogram.html","id":"references","dir":"Reference","previous_headings":"","what":"References","title":"Draw a rootogram — draw.rootogram","text":"Kleiber, C., Zeileis, ., (2016) Visualizing Count Data Regressions Using Rootograms. . Stat. 70, 296–303. doi:10.1080/00031305.2016.1173590","code":""},{"path":[]},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.rootogram.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Draw a rootogram — draw.rootogram","text":"","code":"load_mgcv() df <- data_sim(\"eg1\", n = 1000, dist = \"poisson\", scale = 0.1, seed = 6)  # A poisson example m <- gam(y ~ s(x0, bs = \"cr\") + s(x1, bs = \"cr\") + s(x2, bs = \"cr\") +   s(x3, bs = \"cr\"), family = poisson(), data = df, method = \"REML\") rg <- rootogram(m)  # plot the rootogram draw(rg)   # change the type of rootogram draw(rg, type = \"suspended\")"},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.smooth_estimates.html","id":null,"dir":"Reference","previous_headings":"","what":"Plot the result of a call to smooth_estimates() — draw.smooth_estimates","title":"Plot the result of a call to smooth_estimates() — draw.smooth_estimates","text":"Plot result call smooth_estimates()","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.smooth_estimates.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Plot the result of a call to smooth_estimates() — draw.smooth_estimates","text":"","code":"# S3 method for class 'smooth_estimates' draw(   object,   constant = NULL,   fun = NULL,   contour = TRUE,   grouped_by = FALSE,   contour_col = \"black\",   n_contour = NULL,   ci_alpha = 0.2,   ci_col = \"black\",   smooth_col = \"black\",   resid_col = \"steelblue3\",   decrease_col = \"#56B4E9\",   increase_col = \"#E69F00\",   change_lwd = 1.75,   partial_match = FALSE,   discrete_colour = NULL,   discrete_fill = NULL,   continuous_colour = NULL,   continuous_fill = NULL,   angle = NULL,   ylim = NULL,   crs = NULL,   default_crs = NULL,   lims_method = \"cross\",   caption = TRUE,   ... )"},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.smooth_estimates.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Plot the result of a call to smooth_estimates() — draw.smooth_estimates","text":"object fitted GAM, result call mgcv::gam(). constant numeric; constant add estimated values smooth. constant, supplied, added estimated value confidence band computed. fun function; function applied estimated values confidence interval plotting. Can function name function. Function fun applied adding constant, provided. contour logical; contours draw plot using ggplot2::geom_contour(). grouped_by logical; factor smooths drawn one panel per level factor (FALSE, default), individual smooths combined single panel containing levels (TRUE)? contour_col colour specification contour lines. n_contour numeric; number contour bins. result n_contour - 1 contour lines drawn. See ggplot2::geom_contour(). ci_alpha numeric; alpha transparency confidence simultaneous interval. ci_col colour specification confidence/credible intervals band. Affects fill interval. smooth_col colour specification smooth line. resid_col colour specification partial residuals. decrease_col, increase_col colour specifications use indicating periods change. col_change used change_type = \"change\", col_decrease col_increase used `change_type = \"sizer\"“. change_lwd numeric; value set linewidth ggplot2::geom_line(), used represent periods change. partial_match logical; smooths selected partial matches select? TRUE, select can single string match . discrete_colour suitable colour scale used plotting discrete variables. discrete_fill suitable fill scale used plotting discrete variables. continuous_colour suitable colour scale used plotting continuous variables. continuous_fill suitable fill scale used plotting continuous variables. angle numeric; angle x axis tick labels drawn passed angle argument ggplot2::guide_axis(). ylim numeric; vector y axis limits use panels drawn. crs coordinate reference system (CRS) use plot. data projected CRS. See ggplot2::coord_sf() details. default_crs coordinate reference system (CRS) use non-sf layers plot. left default NULL, CRS used 4326 (WGS84), appropriate spline---sphere smooths, parameterized terms latitude longitude coordinates. See ggplot2::coord_sf() details. lims_method character; affects axis limits determined. See ggplot2::coord_sf(). careful; testing examples, changing \"orthogonal\" example chlorophyll-example Simon Wood's GAM book quickly used RAM test system OS killed R. incorrect usage part; right now grid points SOS smooths evaluated (supplied user) can produce invalid coordinates corners tiles grid generated tile centres without respect spacing tiles. caption logical; show smooth type caption plot? ... additional arguments passed patchwork::wrap_plots().","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.smooth_estimates.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Plot the result of a call to smooth_estimates() — draw.smooth_estimates","text":"","code":"load_mgcv() # example data df <- data_sim(\"eg1\", seed = 21) # fit GAM m <- gam(y ~ s(x0) + s(x1) + s(x2) + s(x3), data = df, method = \"REML\") # plot all of the estimated smooths sm <- smooth_estimates(m) draw(sm)  # evaluate smooth of `x2` sm <- smooth_estimates(m, select = \"s(x2)\") # plot it draw(sm)   # customising some plot elements draw(sm, ci_col = \"steelblue\", smooth_col = \"forestgreen\", ci_alpha = 0.3)   # Add a constant to the plotted smooth draw(sm, constant = coef(m)[1])   # Adding change indicators to smooths based on derivatives of the smooth d <- derivatives(m, n = 100) # n to match smooth_estimates()  smooth_estimates(m) |>   add_sizer(derivatives = d, type = \"sizer\") |>   draw()"},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.smooth_samples.html","id":null,"dir":"Reference","previous_headings":"","what":"Plot posterior smooths — draw.smooth_samples","title":"Plot posterior smooths — draw.smooth_samples","text":"Plot posterior smooths","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.smooth_samples.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Plot posterior smooths — draw.smooth_samples","text":"","code":"# S3 method for class 'smooth_samples' draw(   object,   select = NULL,   n_samples = NULL,   seed = NULL,   xlab = NULL,   ylab = NULL,   title = NULL,   subtitle = NULL,   caption = NULL,   alpha = 1,   colour = \"black\",   contour = FALSE,   contour_col = \"black\",   n_contour = NULL,   scales = c(\"free\", \"fixed\"),   rug = TRUE,   partial_match = FALSE,   angle = NULL,   ncol = NULL,   nrow = NULL,   guides = \"keep\",   ... )"},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.smooth_samples.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Plot posterior smooths — draw.smooth_samples","text":"object fitted GAM, result call mgcv::gam(). select character, logical, numeric; smooths plot. NULL, default, model smooths drawn. Numeric select indexes smooths order specified formula stored object. Character select matches labels smooths shown example output summary(object). Logical select operates per numeric select order smooths stored. n_samples numeric; NULL, sample n_samples posterior draws plotting. seed numeric; random seed used sampling draws. xlab character expression; label x axis. supplied, suitable label generated object. ylab character expression; label y axis. supplied, suitable label generated object. title character expression; title plot. See ggplot2::labs(). subtitle character expression; subtitle plot. See ggplot2::labs(). caption character expression; plot caption. See ggplot2::labs(). alpha numeric; alpha transparency confidence simultaneous interval. colour colour use draw posterior smooths. Passed ggplot2::geom_line() argument colour. contour logical; contour lines added smooth surfaces? contour_col colour specification contour lines. n_contour numeric; number contour bins. result n_contour - 1 contour lines drawn. See ggplot2::geom_contour(). scales character; univariate smooths plotted y-axis scale? scales = \"free\", default, univariate smooth y-axis scale. scales = \"fixed\", common y axis scale used univariate smooths. Currently affect y-axis scale plots parametric terms. rug logical; draw rug plot bottom plot 1-D smooths plot locations data higher dimensions. partial_match logical; smooths selected partial matches select? TRUE, select can single string match . angle numeric; angle x axis tick labels drawn passed angle argument ggplot2::guide_axis(). ncol, nrow numeric; numbers rows columns spread plots guides character; one \"keep\" (default), \"collect\", \"auto\". Passed patchwork::plot_layout() ... arguments passed patchwork::wrap_plots().","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.smooth_samples.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Plot posterior smooths — draw.smooth_samples","text":"Gavin L. Simpson","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/draw.smooth_samples.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Plot posterior smooths — draw.smooth_samples","text":"","code":"load_mgcv() dat1 <- data_sim(\"eg1\", n = 400, dist = \"normal\", scale = 1, seed = 1) ## a single smooth GAM m1 <- gam(y ~ s(x0) + s(x1) + s(x2) + s(x3), data = dat1, method = \"REML\") ## posterior smooths from m1 sm1 <- smooth_samples(m1, n = 15, seed = 23478) ## plot draw(sm1, alpha = 0.7)  ## plot only 5 randomly smapled draws draw(sm1, n_samples = 5, alpha = 0.7)   ## A factor-by smooth example dat2 <- data_sim(\"eg4\", n = 400, dist = \"normal\", scale = 1, seed = 1) ## a multi-smooth GAM with a factor-by smooth m2 <- gam(y ~ fac + s(x2, by = fac) + s(x0), data = dat2, method = \"REML\") ## posterior smooths from m1 sm2 <- smooth_samples(m2, n = 15, seed = 23478) ## plot, this time selecting only the factor-by smooth draw(sm2, select = \"s(x2)\", partial_match = TRUE, alpha = 0.7)   # \\donttest{ ## A 2D smooth example dat3 <- data_sim(\"eg2\", n = 400, dist = \"normal\", scale = 1, seed = 1) ## fit a 2D smooth m3 <- gam(y ~ te(x, z), data = dat3, method = \"REML\") ## get samples sm3 <- smooth_samples(m3, n = 10) ## plot just 6 of the draws, with contour line overlays draw(sm3, n_samples = 6, contour = TRUE, seed = 42)  # }"},{"path":"https://gavinsimpson.github.io/gratia/reference/draw_parametric_effect.html","id":null,"dir":"Reference","previous_headings":"","what":"Internal function to draw an individual parametric effect — draw_parametric_effect","title":"Internal function to draw an individual parametric effect — draw_parametric_effect","text":"Internal function draw individual parametric effect","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/draw_parametric_effect.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Internal function to draw an individual parametric effect — draw_parametric_effect","text":"","code":"draw_parametric_effect(   object,   ci_level = 0.95,   ci_col = \"black\",   ci_alpha = 0.2,   line_col = \"black\",   constant = NULL,   fun = NULL,   xlab = NULL,   ylab = NULL,   title = NULL,   subtitle = NULL,   caption = NULL,   rug = TRUE,   position = \"identity\",   ylim = NULL,   angle = NULL,   factor_levels = NULL,   ... )"},{"path":"https://gavinsimpson.github.io/gratia/reference/draw_parametric_effect.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Internal function to draw an individual parametric effect — draw_parametric_effect","text":"object fitted GAM, result call mgcv::gam(). ci_level numeric 0 1; coverage credible interval. ci_col colour specification confidence/credible intervals band. Affects fill interval. ci_alpha numeric; alpha transparency confidence simultaneous interval. constant numeric; constant add estimated values smooth. constant, supplied, added estimated value confidence band computed. fun function; function applied estimated values confidence interval plotting. Can function name function. Function fun applied adding constant, provided. xlab character expression; label x axis. supplied, suitable label generated object. ylab character expression; label y axis. supplied, suitable label generated object. title character expression; title plot. See ggplot2::labs(). subtitle character expression; subtitle plot. See ggplot2::labs(). caption character expression; plot caption. See ggplot2::labs(). rug logical; draw rug plot bottom plot 1-D smooths plot locations data higher dimensions. position Position adjustment, either string, result call position adjustment function. angle numeric; angle x axis tick labels drawn passed angle argument ggplot2::guide_axis(). factor_levels list; named list factor levels ... additional arguments passed patchwork::wrap_plots().","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/edf.html","id":null,"dir":"Reference","previous_headings":"","what":"Effective degrees of freedom for smooths and GAMs — edf","title":"Effective degrees of freedom for smooths and GAMs — edf","text":"Extracts effective degrees freedom (EDF) model smooth terms overall EDF fitted GAMs","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/edf.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Effective degrees of freedom for smooths and GAMs — edf","text":"","code":"edf(object, ...)  # S3 method for class 'gam' edf(   object,   select = NULL,   smooth = deprecated(),   type = c(\"default\", \"unconditional\", \"alternative\"),   partial_match = FALSE,   ... )  model_edf(object, ..., type = c(\"default\", \"unconditional\", \"alternative\"))"},{"path":"https://gavinsimpson.github.io/gratia/reference/edf.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Effective degrees of freedom for smooths and GAMs — edf","text":"object fitted model extract smooth-specific EDFs. ... arguments passed methods. select character, logical, numeric; smooths plot. NULL, default, model smooths drawn. Numeric select indexes smooths order specified formula stored object. Character select matches labels smooths shown example output summary(object). Logical select operates per numeric select order smooths stored. smooth Use select instead. extracted. NULL, default, EDFs smooths returned. type character: type EDF return. \"default\" returns standard EDF; \"unconditional\" selects EDF corrected smoothness parameter selection, available; \"alternative\" returns alternative formulation EDF Wood (2017, pp. 252) partial_match logical; smooths selected partial matches select? TRUE, select can single string match .","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/edf.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Effective degrees of freedom for smooths and GAMs — edf","text":"Multiple formulations effective degrees freedom available. additional uncertainty due selection smoothness parameters can taken account computing EDF smooths. form EDF available type = \"unconditional\". Wood (2017; pp. 252) describes alternative EDF  model $$\\mathrm{EDF} = 2\\mathrm{tr}(\\mathbf{F}) -   \\mathrm{tr}(\\mathbf{FF}),$$ \\(\\mathrm{tr}\\) matrix trace \\(\\mathbf{F}\\) matrix mapping un-penalized coefficient estimates penalized coefficient estimates.  trace \\(\\mathbf{F}\\) effectively average shrinkage coefficients multipled number coefficients (Wood, 2017). Smooth-specific EDFs obtained summing relevent elements \\(\\mathrm{diag}(2\\mathbf{F} - \\mathbf{FF})\\).","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/edf.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Effective degrees of freedom for smooths and GAMs — edf","text":"","code":"load_mgcv() df <- data_sim(\"eg1\", n = 400, seed = 42) m <- gam(y ~ s(x0) + s(x1) + s(x2) + s(x3), data = df, method = \"REML\")  # extract the EDFs for all smooths edf(m) #> # A tibble: 4 x 2 #>   .smooth   .edf #>   <chr>    <dbl> #> 1 s(x0)   3.4248 #> 2 s(x1)   3.2213 #> 3 s(x2)   7.9049 #> 4 s(x3)   1.8847  # or selected smooths edf(m, select = c(\"s(x0)\", \"s(x2)\")) #> # A tibble: 2 x 2 #>   .smooth   .edf #>   <chr>    <dbl> #> 1 s(x0)   3.4248 #> 2 s(x2)   7.9049  # accounting for smoothness parameter uncertainty edf(m, type = \"unconditional\") #> # A tibble: 4 x 2 #>   .smooth   .edf #>   <chr>    <dbl> #> 1 s(x0)   3.7697 #> 2 s(x1)   3.8728 #> 3 s(x2)   8.0670 #> 4 s(x3)   2.8834  # over EDF of the model, including the intercept model_edf(m) #> # A tibble: 1 x 2 #>   .model   .edf #>   <chr>   <dbl> #> 1 m      17.436  # can get model EDF for multiple models m2 <- gam(y ~ s(x0) + s(x1) + s(x3), data = df, method = \"REML\") model_edf(m, m2) #> # A tibble: 2 x 2 #>   .model    .edf #>   <chr>    <dbl> #> 1 m      17.436  #> 2 m2      7.5777"},{"path":"https://gavinsimpson.github.io/gratia/reference/eval_smooth.html","id":null,"dir":"Reference","previous_headings":"","what":"S3 methods to evaluate individual smooths — eval_smooth","title":"S3 methods to evaluate individual smooths — eval_smooth","text":"S3 methods evaluate individual smooths","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/eval_smooth.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"S3 methods to evaluate individual smooths — eval_smooth","text":"","code":"eval_smooth(smooth, ...)  # S3 method for class 'mgcv.smooth' eval_smooth(   smooth,   model,   n = 100,   n_3d = NULL,   n_4d = NULL,   data = NULL,   unconditional = FALSE,   overall_uncertainty = TRUE,   dist = NULL,   ... )  # S3 method for class 'soap.film' eval_smooth(   smooth,   model,   n = 100,   n_3d = NULL,   n_4d = NULL,   data = NULL,   unconditional = FALSE,   overall_uncertainty = TRUE,   ... )  # S3 method for class 'scam_smooth' eval_smooth(   smooth,   model,   n = 100,   n_3d = NULL,   n_4d = NULL,   data = NULL,   unconditional = FALSE,   overall_uncertainty = TRUE,   dist = NULL,   ... )  # S3 method for class 'fs.interaction' eval_smooth(   smooth,   model,   n = 100,   data = NULL,   unconditional = FALSE,   overall_uncertainty = TRUE,   ... )  # S3 method for class 'sz.interaction' eval_smooth(   smooth,   model,   n = 100,   data = NULL,   unconditional = FALSE,   overall_uncertainty = TRUE,   ... )  # S3 method for class 'random.effect' eval_smooth(   smooth,   model,   n = 100,   data = NULL,   unconditional = FALSE,   overall_uncertainty = TRUE,   ... )  # S3 method for class 'mrf.smooth' eval_smooth(   smooth,   model,   n = 100,   data = NULL,   unconditional = FALSE,   overall_uncertainty = TRUE,   ... )  # S3 method for class 't2.smooth' eval_smooth(   smooth,   model,   n = 100,   n_3d = NULL,   n_4d = NULL,   data = NULL,   unconditional = FALSE,   overall_uncertainty = TRUE,   dist = NULL,   ... )  # S3 method for class 'tensor.smooth' eval_smooth(   smooth,   model,   n = 100,   n_3d = NULL,   n_4d = NULL,   data = NULL,   unconditional = FALSE,   overall_uncertainty = TRUE,   dist = NULL,   ... )"},{"path":"https://gavinsimpson.github.io/gratia/reference/eval_smooth.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"S3 methods to evaluate individual smooths — eval_smooth","text":"smooth currently object inherits class mgcv.smooth. ... arguments passed methods model fitted model; currently mgcv::gam() mgcv::bam() models suported. n numeric; number points range covariate evaluate smooth. n_3d, n_4d numeric; number points range last covariate 3D 4D smooth. default NULL achieves standard behaviour using n points range covariate, resulting n^d evaluation points, d dimension smooth. d > 2 can result many evaluation points slow performance. smooths d > 4, value n_4d used dimensions > 4, unless NULL, case default behaviour (using n dimensions) observed. data optional data frame values evaluate smooth . unconditional logical; confidence intervals include uncertainty due smoothness selection? TRUE, corrected Bayesian covariance matrix used. overall_uncertainty logical; uncertainty model constant term included standard error evaluate values smooth? dist numeric; greater 0, used determine location far data plotted plotting 2-D smooths. data scaled unit square deciding exclude, dist distance within unit square. See mgcv::exclude..far() details.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/evaluate_parametric_term.html","id":null,"dir":"Reference","previous_headings":"","what":"Evaluate parametric model terms — evaluate_parametric_term","title":"Evaluate parametric model terms — evaluate_parametric_term","text":"Returns values parametric model terms values factor terms grid covariate values linear parametric terms. function now deprecated favour parametric_effects().","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/evaluate_parametric_term.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Evaluate parametric model terms — evaluate_parametric_term","text":"","code":"evaluate_parametric_term(object, ...)  # S3 method for class 'gam' evaluate_parametric_term(object, term, unconditional = FALSE, ...)"},{"path":"https://gavinsimpson.github.io/gratia/reference/evaluate_parametric_term.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Evaluate parametric model terms — evaluate_parametric_term","text":"object object class \"gam\" \"gamm\". ... arguments passed methods. term character; parametric term whose effects evaluated unconditional logical; confidence intervals include uncertainty due smoothness selection? TRUE, corrected Bayesian covariance matrix used.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/evaluate_smooth.html","id":null,"dir":"Reference","previous_headings":"","what":"Evaluate a smooth — evaluate_smooth","title":"Evaluate a smooth — evaluate_smooth","text":"Evaluate smooth grid evenly spaced value range covariate associated smooth. Alternatively, set points smooth evaluated can supplied.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/evaluate_smooth.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Evaluate a smooth — evaluate_smooth","text":"","code":"evaluate_smooth(object, ...)"},{"path":"https://gavinsimpson.github.io/gratia/reference/evaluate_smooth.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Evaluate a smooth — evaluate_smooth","text":"object object class \"gam\" \"gamm\". ... arguments passed methods.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/evaluate_smooth.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Evaluate a smooth — evaluate_smooth","text":"data frame, class \"evaluated_1d_smooth\" evaluated_2d_smooth, inherit classes \"evaluated_smooth\" \"data.frame\".","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/evaluate_smooth.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Evaluate a smooth — evaluate_smooth","text":"evaluate_smooth() deprecated favour smooth_estimates(), provides cleaner way evaluate smooth range covariate values. smooth_estimates() can handle much wider range models evaluate_smooth() capable smooth_estimates() much easier extend handle new smooth types. code uses evaluate_smooth() work simply changing function call smooth_estimates(). However, differences: newdata argument becomes data","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/evenly.html","id":null,"dir":"Reference","previous_headings":"","what":"Create a sequence of evenly-spaced values — evenly","title":"Create a sequence of evenly-spaced values — evenly","text":"continuous vector x, evenly seq_min_max() create sequence n evenly-spaced values range lower – upper. default, lower defined min(x) upper max(x), excluding NAs. factor x, function returns levels(x).","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/evenly.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Create a sequence of evenly-spaced values — evenly","text":"","code":"evenly(x, n = 100, by = NULL, lower = NULL, upper = NULL)  seq_min_max(x, n, by = NULL, lower = NULL, upper = NULL)"},{"path":"https://gavinsimpson.github.io/gratia/reference/evenly.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Create a sequence of evenly-spaced values — evenly","text":"x numeric; vector evenly-spaced values returned n numeric; number evenly-spaced values return. default 100 used convenience typically used evaluating smooth. numeric; increment sequence. specified, argument n ignored sequence returned min(x) max(x) increments . lower numeric; lower bound interval. upper numeric; upper bound interval.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/evenly.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Create a sequence of evenly-spaced values — evenly","text":"numeric vector length n.","code":""},{"path":[]},{"path":"https://gavinsimpson.github.io/gratia/reference/evenly.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Create a sequence of evenly-spaced values — evenly","text":"","code":"x <- rnorm(10) n <- 10L  # 10 values evenly over the range of `x` evenly(x, n = n) #>  [1] -0.83562861 -0.56552757 -0.29542652 -0.02532547  0.24477557  0.51487662 #>  [7]  0.78497766  1.05507871  1.32517976  1.59528080  # evenly spaced values, incrementing by 0.2 evenly(x, by = 0.2) #>  [1] -0.83562861 -0.63562861 -0.43562861 -0.23562861 -0.03562861  0.16437139 #>  [7]  0.36437139  0.56437139  0.76437139  0.96437139  1.16437139  1.36437139 #> [13]  1.56437139  # evenly spaced values, incrementing by 0.2, starting at -2 evenly(x, by = 0.2, lower = -2) #>  [1] -2.0 -1.8 -1.6 -1.4 -1.2 -1.0 -0.8 -0.6 -0.4 -0.2  0.0  0.2  0.4  0.6  0.8 #> [16]  1.0  1.2  1.4"},{"path":"https://gavinsimpson.github.io/gratia/reference/factor_combos.html","id":null,"dir":"Reference","previous_headings":"","what":"All combinations of factor levels — factor_combos","title":"All combinations of factor levels — factor_combos","text":"combinations factor levels","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/factor_combos.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"All combinations of factor levels — factor_combos","text":"","code":"factor_combos(object, ...)  # S3 method for class 'gam' factor_combos(object, vars = everything(), complete = TRUE, ...)"},{"path":"https://gavinsimpson.github.io/gratia/reference/factor_combos.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"All combinations of factor levels — factor_combos","text":"object fitted model object. ... arguments passed methods. vars terms include exclude returned object. Uses tidyselect principles. complete logical; combinations factor levels returned? FALSE, combinations levels observed model retained.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/family.gam.html","id":null,"dir":"Reference","previous_headings":"","what":"Extract family objects from models — family.gam","title":"Extract family objects from models — family.gam","text":"Provides stats::family() method range GAM objects.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/family.gam.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Extract family objects from models — family.gam","text":"","code":"# S3 method for class 'gam' family(object, ...)  # S3 method for class 'gamm' family(object, ...)  # S3 method for class 'bam' family(object, ...)  # S3 method for class 'list' family(object, ...)"},{"path":"https://gavinsimpson.github.io/gratia/reference/family.gam.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Extract family objects from models — family.gam","text":"object fitted model. Models fitted mgcv::gam(), mgcv::bam(), mgcv::gamm(), gamm4::gamm4() currently supported. ... arguments passed methods.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/family_name.html","id":null,"dir":"Reference","previous_headings":"","what":"Name of family used to fit model — family_name","title":"Name of family used to fit model — family_name","text":"Extracts name family used fit supplied model.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/family_name.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Name of family used to fit model — family_name","text":"","code":"family_name(object, ...)"},{"path":"https://gavinsimpson.github.io/gratia/reference/family_name.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Name of family used to fit model — family_name","text":"object R object. ... arguments passed methods.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/family_name.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Name of family used to fit model — family_name","text":"character vector containing family name.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/family_type.html","id":null,"dir":"Reference","previous_headings":"","what":"Extracts the type of family in a consistent way — family_type","title":"Extracts the type of family in a consistent way — family_type","text":"Extracts type family consistent way","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/family_type.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Extracts the type of family in a consistent way — family_type","text":"","code":"family_type(object, ...)  # S3 method for class 'family' family_type(object, ...)  # Default S3 method family_type(object, ...)"},{"path":"https://gavinsimpson.github.io/gratia/reference/family_type.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Extracts the type of family in a consistent way — family_type","text":"object R object. Currently family() objects anything family() method. ... arguments passed methods.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/fderiv.html","id":null,"dir":"Reference","previous_headings":"","what":"First derivatives of fitted GAM functions — fderiv","title":"First derivatives of fitted GAM functions — fderiv","text":"function deprecated limited first order forward finite differences derivatives , improved offer needed functionality without breaking backwards compatability papers blog posts already used fderiv(). replacement, derivatives(), now available recommended new analyses.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/fderiv.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"First derivatives of fitted GAM functions — fderiv","text":"","code":"fderiv(model, ...)  # S3 method for class 'gam' fderiv(   model,   newdata,   term,   n = 200,   eps = 1e-07,   unconditional = FALSE,   offset = NULL,   ... )  # S3 method for class 'gamm' fderiv(model, ...)"},{"path":"https://gavinsimpson.github.io/gratia/reference/fderiv.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"First derivatives of fitted GAM functions — fderiv","text":"model fitted GAM. Currently models fitted mgcv::gam() mgcv::gamm() supported. ... Arguments passed methods. newdata data frame containing values model covariates evaluate first derivatives smooths. term character; vector one terms derivatives required. missing, derivatives smooth terms returned. n integer; newdata missing original data can reconstructed model n controls number values range covariate populate newdata. eps numeric; value finite difference used approximate first derivative. unconditional logical; TRUE, smoothing parameter uncertainty corrected covariance matrix used, available, otherwise uncorrected Bayesian posterior covariance matrix used. offset numeric; value offset use generating predictions.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/fderiv.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"First derivatives of fitted GAM functions — fderiv","text":"object class \"fderiv\" returned.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/fderiv.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"First derivatives of fitted GAM functions — fderiv","text":"Gavin L. Simpson","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/fderiv.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"First derivatives of fitted GAM functions — fderiv","text":"","code":"load_mgcv() dat <- data_sim(\"eg1\", seed = 2) mod <- gam(y ~ s(x0) + s(x1) + s(x2) + s(x3), data = dat, method = \"REML\")  ## first derivatives of all smooths... fd <- fderiv(mod) ## now use --> fd <- derivatives(mod)  ## ...and a selected smooth fd2 <- fderiv(mod, term = \"x1\") ## now use --> fd2 <- derivatives(mod, select = \"s(x1)\")  ## Models with factors dat <- data_sim(\"eg4\", n = 400, dist = \"normal\", scale = 2, seed = 2) mod <- gam(y ~ s(x0) + s(x1) + fac, data = dat, method = \"REML\")  ## first derivatives of all smooths... fd <- fderiv(mod) ## now use --> fd <- derivatives(mod)  ## ...and a selected smooth fd2 <- fderiv(mod, term = \"x1\") ## now use --> fd2 <- derivatives(mod, select = \"s(x1)\")"},{"path":"https://gavinsimpson.github.io/gratia/reference/fitted_samples.html","id":null,"dir":"Reference","previous_headings":"","what":"Draw fitted values from the posterior distribution — fitted_samples","title":"Draw fitted values from the posterior distribution — fitted_samples","text":"Expectations (fitted values) response drawn posterior distribution fitted model using Gaussian approximation posterior simple Metropolis Hastings sampler.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/fitted_samples.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Draw fitted values from the posterior distribution — fitted_samples","text":"","code":"fitted_samples(model, ...)  # S3 method for class 'gam' fitted_samples(   model,   n = 1,   data = newdata,   seed = NULL,   scale = c(\"response\", \"linear_predictor\"),   method = c(\"gaussian\", \"mh\", \"inla\", \"user\"),   n_cores = 1,   burnin = 1000,   thin = 1,   t_df = 40,   rw_scale = 0.25,   freq = FALSE,   unconditional = FALSE,   draws = NULL,   mvn_method = c(\"mvnfast\", \"mgcv\"),   ...,   newdata = NULL,   ncores = NULL )"},{"path":"https://gavinsimpson.github.io/gratia/reference/fitted_samples.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Draw fitted values from the posterior distribution — fitted_samples","text":"model fitted model supported types ... arguments passed methods. fitted_samples(), passed mgcv::predict.gam(). posterior_samples() passed fitted_samples(). predicted_samples() passed relevant simulate() method. n numeric; number posterior samples return. data data frame; new observations posterior draws model evaluated. supplied, data used fit model used data, available model. seed numeric; random seed simulations. scale character; scale fitted values returned ? \"linear predictor\" synonym \"link\" prefer terminology. method character; method used draw samples posterior distribution. \"gaussian\" uses Gaussian (Laplace) approximation posterior. \"mh\" uses Metropolis Hastings sampler alternates t proposals proposals based shrunken version posterior covariance matrix. \"inla\" uses variant Integrated Nested Laplace Approximation due Wood (2019), (currently implemented). \"user\" allows user-supplied posterior draws (currently implemented). n_cores number cores generating random variables multivariate normal distribution. Passed mvnfast::rmvn(). Parallelization take place OpenMP supported (appears work Windows current R). burnin numeric; number samples discard burnin draws. used method = \"mh\". thin numeric; number samples skip taking n draws. Results thin * n draws posterior taken. used method = \"mh\". t_df numeric; degrees freedom t distribution proposals. used method = \"mh\". rw_scale numeric; Factor scale posterior covariance matrix generating random walk proposals. Negative non finite skip random walk step. used method = \"mh\". freq logical; TRUE use frequentist covariance matrix parameter estimators, FALSE use Bayesian posterior covariance matrix parameters. unconditional logical; TRUE (freq == FALSE) Bayesian smoothing parameter uncertainty corrected covariance matrix used, available. draws matrix; user supplied posterior draws used method = \"user\". mvn_method character; one \"mvnfast\" \"mgcv\". default uses mvnfast::rmvn(), can considerably faster generate large numbers MVN random values mgcv::rmvn(), might work marginal fits, covariance matrix close singular. newdata Deprecated: use data instead. ncores Deprecated; use n_cores instead. number cores generating random variables multivariate normal distribution. Passed mvnfast::rmvn(). Parallelization take place OpenMP supported (appears work Windows current R).","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/fitted_samples.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Draw fitted values from the posterior distribution — fitted_samples","text":"tibble (data frame) 3 columns containing posterior predicted values long format. columns row (integer) row data posterior draw relates , draw (integer) index, range 1:n, indicating draw row relates , response (numeric) predicted response indicated row data.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/fitted_samples.html","id":"note","dir":"Reference","previous_headings":"","what":"Note","title":"Draw fitted values from the posterior distribution — fitted_samples","text":"Models offset terms supplied via offset argument mgcv::gam() etc. ignored mgcv::predict.gam(). , kind offset term also ignored posterior_samples(). Offset terms included model formula supplied mgcv::gam() etc ignored posterior samples produced reflect offset term values. side effect requiring new data values provided posterior_samples() via data argument must include offset variable.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/fitted_samples.html","id":"references","dir":"Reference","previous_headings":"","what":"References","title":"Draw fitted values from the posterior distribution — fitted_samples","text":"Wood, S.N., (2020). Simplified integrated nested Laplace approximation. Biometrika 107, 223–230. doi:10.1093/biomet/asz044","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/fitted_samples.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Draw fitted values from the posterior distribution — fitted_samples","text":"Gavin L. Simpson","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/fitted_samples.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Draw fitted values from the posterior distribution — fitted_samples","text":"","code":"load_mgcv() dat <- data_sim(\"eg1\", n = 1000, dist = \"normal\", scale = 2, seed = 2) m1 <- gam(y ~ s(x0) + s(x1) + s(x2) + s(x3), data = dat, method = \"REML\")  fs <- fitted_samples(m1, n = 5, seed = 42) # \\donttest{ fs #> # A tibble: 5,000 x 4 #>     .row .draw .parameter .fitted #>    <int> <int> <chr>        <dbl> #>  1     1     1 location      6.34 #>  2     2     1 location      5.08 #>  3     3     1 location      6.84 #>  4     4     1 location      7.71 #>  5     5     1 location      9.23 #>  6     6     1 location      8.03 #>  7     7     1 location      6.19 #>  8     8     1 location      7.28 #>  9     9     1 location     14.0  #> 10    10     1 location     12.7  #> # i 4,990 more rows # }  # can generate own set of draws and use them drws <- generate_draws(m1, n = 2, seed = 24) fs2 <- fitted_samples(m1, method = \"user\", draws = drws) # \\donttest{ fs2 #> # A tibble: 2,000 x 4 #>     .row .draw .parameter .fitted #>    <int> <int> <chr>        <dbl> #>  1     1     1 location      6.30 #>  2     2     1 location      5.12 #>  3     3     1 location      7.40 #>  4     4     1 location      7.42 #>  5     5     1 location      9.40 #>  6     6     1 location      8.04 #>  7     7     1 location      5.83 #>  8     8     1 location      7.30 #>  9     9     1 location     14.3  #> 10    10     1 location     13.1  #> # i 1,990 more rows # }"},{"path":"https://gavinsimpson.github.io/gratia/reference/fitted_values.html","id":null,"dir":"Reference","previous_headings":"","what":"Generate fitted values from a estimated GAM — fitted_values","title":"Generate fitted values from a estimated GAM — fitted_values","text":"Generate fitted values estimated GAM","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/fitted_values.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Generate fitted values from a estimated GAM — fitted_values","text":"","code":"fitted_values(object, ...)  # S3 method for class 'gam' fitted_values(   object,   data = NULL,   scale = c(\"response\", \"link\", \"linear predictor\"),   ci_level = 0.95,   ... )  # S3 method for class 'gamm' fitted_values(object, ...)  # S3 method for class 'scam' fitted_values(object, ...)"},{"path":"https://gavinsimpson.github.io/gratia/reference/fitted_values.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Generate fitted values from a estimated GAM — fitted_values","text":"object fitted model. Currently models fitted mgcv::gam() mgcv::bam() supported. ... arguments passed mgcv::predict.gam(). Note type, newdata, se.fit already used passed mgcv::predict.gam(). data optional data frame covariate values fitted values returned. scale character; scale fitted values returned ? \"linear predictor\" synonym \"link\" prefer terminology. ci_level numeric; value 0 1 indicating coverage credible interval.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/fitted_values.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Generate fitted values from a estimated GAM — fitted_values","text":"tibble (data frame) whose first m columns contain either data used fit model (data NULL), variables supplied data. Four columns added: fitted: fitted values specified scale, se: standard error fitted values (always link scale), lower, upper: limits credible interval fitted values, specified scale. Models fitted certain families include additional variables mgcv::ocat() models: scale = \"repsonse\", returned object contain row column category column, indicate row data row returned object belongs. Additionally, nrow(data) * n_categories rows returned object; row predicted probability single category response.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/fitted_values.html","id":"note","dir":"Reference","previous_headings":"","what":"Note","title":"Generate fitted values from a estimated GAM — fitted_values","text":"families, regardless scale fitted values returned, se component returned object link (linear predictor) scale, response scale. exception mgcv::ocat() family, se response scale scale = \"response\".","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/fitted_values.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Generate fitted values from a estimated GAM — fitted_values","text":"","code":"load_mgcv() sim_df <- data_sim(\"eg1\", n = 400, dist = \"normal\", scale = 2, seed = 2) m <- gam(y ~ s(x0) + s(x1) + s(x2) + s(x3), data = sim_df, method = \"REML\") fv <- fitted_values(m) fv #> # A tibble: 400 x 9 #>     .row       x0        x1         x2       x3  .fitted      .se .lower_ci #>    <int>    <dbl>     <dbl>      <dbl>    <dbl>    <dbl>    <dbl>     <dbl> #>  1     1 0.184882 0.617142  0.415244   0.132410  8.73875 0.354677   8.04360 #>  2     2 0.702374 0.569064  0.531439   0.365331  7.62581 0.337779   6.96378 #>  3     3 0.573326 0.153970  0.00324621 0.454532  3.12106 0.591862   1.96103 #>  4     4 0.168052 0.0348332 0.252100   0.537114 11.1124  0.402378  10.3237  #>  5     5 0.943839 0.997953  0.155229   0.185495 14.0533  0.452947  13.1655  #>  6     6 0.943475 0.835574  0.878840   0.449276  6.13080 0.364521   5.41635 #>  7     7 0.129159 0.586562  0.203511   0.256527 12.4838  0.355808  11.7864  #>  8     8 0.833449 0.339117  0.583528   0.618458  6.25215 0.344700   5.57655 #>  9     9 0.468019 0.166883  0.804473   0.880744  4.21463 0.372003   3.48552 #> 10    10 0.549984 0.807410  0.264717   0.317747 15.5283  0.369999  14.8031  #> # i 390 more rows #> # i 1 more variable: .upper_ci <dbl>"},{"path":"https://gavinsimpson.github.io/gratia/reference/fix_offset.html","id":null,"dir":"Reference","previous_headings":"","what":"Fix the names of a data frame containing an offset variable. — fix_offset","title":"Fix the names of a data frame containing an offset variable. — fix_offset","text":"Identifies variable, , model offset, fixed name offset(foo(var)) converted var, possibly sets values variable offset_val.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/fix_offset.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Fix the names of a data frame containing an offset variable. — fix_offset","text":"","code":"fix_offset(model, newdata, offset_val = NULL)"},{"path":"https://gavinsimpson.github.io/gratia/reference/fix_offset.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Fix the names of a data frame containing an offset variable. — fix_offset","text":"model fitted GAM. newdata data frame; new values predict . offset_val numeric, optional; provided, offset variable newdata set constant value returning newdata","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/fix_offset.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Fix the names of a data frame containing an offset variable. — fix_offset","text":"original newdata returned fixed names possibly modified offset variable.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/fix_offset.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Fix the names of a data frame containing an offset variable. — fix_offset","text":"Gavin L. Simpson","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/fix_offset.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Fix the names of a data frame containing an offset variable. — fix_offset","text":"","code":"load_mgcv() df <- data_sim(\"eg1\", n = 400, dist = \"normal\", seed = 2) m <- gam(y ~ s(x0) + s(x1) + offset(x2), data = df, method = \"REML\") names(model.frame(m)) #> [1] \"y\"          \"offset(x2)\" \"x0\"         \"x1\"         names(fix_offset(m, model.frame(m), offset_val = 1L)) #> [1] \"y\"  \"x2\" \"x0\" \"x1\""},{"path":"https://gavinsimpson.github.io/gratia/reference/fixef.gam.html","id":null,"dir":"Reference","previous_headings":"","what":"Extract fixed effects estimates from a fitted GAM — fixef.gam","title":"Extract fixed effects estimates from a fitted GAM — fixef.gam","text":"Extract fixed effects estimates fitted GAM","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/fixef.gam.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Extract fixed effects estimates from a fitted GAM — fixef.gam","text":"","code":"# S3 method for class 'gam' fixef(object, ...)  # S3 method for class 'gamm' fixef(object, ...)  # S3 method for class 'lm' fixef(object, ...)  # S3 method for class 'glm' fixef(object, ...)  fixed_effects(object, ...)  # Default S3 method fixed_effects(object, ...)"},{"path":"https://gavinsimpson.github.io/gratia/reference/fixef.gam.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Extract fixed effects estimates from a fitted GAM — fixef.gam","text":"object fitted GAM ... arguments passed methods","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/fixef.gam.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Extract fixed effects estimates from a fitted GAM — fixef.gam","text":"","code":"load_mgcv()  # run example if lme4 is available if (require(\"lme4\")) {   data(sleepstudy, package = \"lme4\")   m <- gam(     Reaction ~ Days + s(Subject, bs = \"re\") +       s(Days, Subject, bs = \"re\"),     data = sleepstudy, method = \"REML\"   )   fixef(m) } #> Loading required package: lme4 #> Loading required package: Matrix #> (Intercept)        Days  #>   251.40510    10.46729"},{"path":"https://gavinsimpson.github.io/gratia/reference/fixef.html","id":null,"dir":"Reference","previous_headings":"","what":"Extract fixed effects estimates — fixef","title":"Extract fixed effects estimates — fixef","text":"Extract fixed effects estimates","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/fixef.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Extract fixed effects estimates — fixef","text":"object fitted GAM ... arguments passed methods","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/gaussian_draws.html","id":null,"dir":"Reference","previous_headings":"","what":"Posterior samples using a simple Metropolis Hastings sampler — gaussian_draws","title":"Posterior samples using a simple Metropolis Hastings sampler — gaussian_draws","text":"Posterior samples using simple Metropolis Hastings sampler","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/gaussian_draws.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Posterior samples using a simple Metropolis Hastings sampler — gaussian_draws","text":"","code":"gaussian_draws(model, ...)  # S3 method for class 'gam' gaussian_draws(   model,   n,   n_cores = 1L,   index = NULL,   frequentist = FALSE,   unconditional = FALSE,   mvn_method = \"mvnfast\",   ... )  # S3 method for class 'scam' gaussian_draws(   model,   n,   n_cores = 1L,   index = NULL,   frequentist = FALSE,   parametrized = TRUE,   mvn_method = \"mvnfast\",   ... )"},{"path":"https://gavinsimpson.github.io/gratia/reference/gaussian_draws.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Posterior samples using a simple Metropolis Hastings sampler — gaussian_draws","text":"model fitted R model. Currently models fitted mgcv::gam() mgcv::bam(), return object inherits objects supported. , \"inherits\" used loose fashion; models fitted scam::scam() support even though models strictly inherit class \"gam\" far inherits() concerned. ... arguments passed methods. n numeric; number posterior draws take. n_cores integer; number CPU cores use generating multivariate normal distributed random values. used mvn_method = \"mvnfast\" method = \"gaussian\". index numeric; vector indices coefficients use. Can used subset mean vector covariance matrix extracted model. frequentist logical; TRUE, frequentist covariance matrix parameter estimates used. FALSE, Bayesian posterior covariance matrix parameters used. See mgcv::vcov.gam(). unconditional logical; TRUE  Bayesian smoothing parameter uncertainty corrected covariance matrix used, available model. See mgcv::vcov.gam(). mvn_method character; one \"mvnfast\" \"mgcv\". default uses mvnfast::rmvn(), can considerably faster generate large numbers MVN random values mgcv::rmvn(), might work marginal fits, covariance matrix close singular. parametrized logical; use parametrized coefficients covariance matrix, respect linear inequality constraints model. scam::scam() model fits.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/get_by_smooth.html","id":null,"dir":"Reference","previous_headings":"","what":"Extract an factor-by smooth by name — get_by_smooth","title":"Extract an factor-by smooth by name — get_by_smooth","text":"Extract factor-smooth name","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/get_by_smooth.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Extract an factor-by smooth by name — get_by_smooth","text":"","code":"get_by_smooth(object, term, level)"},{"path":"https://gavinsimpson.github.io/gratia/reference/get_by_smooth.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Extract an factor-by smooth by name — get_by_smooth","text":"object fitted GAM model object. term character; name smooth term extract. level character; level factor exrtact smooth .","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/get_by_smooth.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Extract an factor-by smooth by name — get_by_smooth","text":"single smooth object, list smooths several match named term.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/get_smooth.html","id":null,"dir":"Reference","previous_headings":"","what":"Extract an mgcv smooth by name — get_smooth","title":"Extract an mgcv smooth by name — get_smooth","text":"Extract mgcv smooth name","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/get_smooth.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Extract an mgcv smooth by name — get_smooth","text":"","code":"get_smooth(object, term)"},{"path":"https://gavinsimpson.github.io/gratia/reference/get_smooth.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Extract an mgcv smooth by name — get_smooth","text":"object fitted GAM model object. term character; name smooth term extract","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/get_smooth.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Extract an mgcv smooth by name — get_smooth","text":"single smooth object, list smooths several match named term.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/get_smooths_by_id.html","id":null,"dir":"Reference","previous_headings":"","what":"Extract an mgcv smooth given its position in the model object — get_smooths_by_id","title":"Extract an mgcv smooth given its position in the model object — get_smooths_by_id","text":"Extract mgcv smooth given position model object","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/get_smooths_by_id.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Extract an mgcv smooth given its position in the model object — get_smooths_by_id","text":"","code":"get_smooths_by_id(object, id)  # S3 method for class 'gam' get_smooths_by_id(object, id)  # S3 method for class 'scam' get_smooths_by_id(object, id)  # S3 method for class 'gamm' get_smooths_by_id(object, id)  # S3 method for class 'gamm4' get_smooths_by_id(object, id)  # S3 method for class 'list' get_smooths_by_id(object, id)"},{"path":"https://gavinsimpson.github.io/gratia/reference/get_smooths_by_id.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Extract an mgcv smooth given its position in the model object — get_smooths_by_id","text":"object fitted GAM model object. id numeric; position smooth model object.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/gratia-package.html","id":null,"dir":"Reference","previous_headings":"","what":"gratia: Graceful 'ggplot'-Based Graphics and Other Functions for GAMs Fitted Using 'mgcv' — gratia-package","title":"gratia: Graceful 'ggplot'-Based Graphics and Other Functions for GAMs Fitted Using 'mgcv' — gratia-package","text":"Graceful 'ggplot'-based graphics utility functions working generalized additive models (GAMs) fitted using 'mgcv' package. Provides reimplementation plot() method GAMs 'mgcv' provides, well 'tidyverse' compatible representations estimated smooths.","code":""},{"path":[]},{"path":"https://gavinsimpson.github.io/gratia/reference/gratia-package.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"gratia: Graceful 'ggplot'-Based Graphics and Other Functions for GAMs Fitted Using 'mgcv' — gratia-package","text":"Maintainer: Gavin L. Simpson ucfagls@gmail.com (ORCID) [copyright holder] contributors: Henrik Singmann (ORCID) [contributor]","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/gss_vocab.html","id":null,"dir":"Reference","previous_headings":"","what":"Data from the General Social Survey (GSS) from the National Opinion Research Center of the University of Chicago — gss_vocab","title":"Data from the General Social Survey (GSS) from the National Opinion Research Center of the University of Chicago — gss_vocab","text":"subset data carData::GSSvocab dataset carData package, containing observations 2016 .","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/gss_vocab.html","id":"format","dir":"Reference","previous_headings":"","what":"Format","title":"Data from the General Social Survey (GSS) from the National Opinion Research Center of the University of Chicago — gss_vocab","text":"data frame 1858 rows 3 variables: vocab: numeric; number words 10 correct vocabulary test. nativeBorn: factor; respondent born US? factor levels yes. ageGroup: factor; grouped age respondent levels 18-29 30-39, 40-49, 50-59, 60+.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/gw_functions.html","id":null,"dir":"Reference","previous_headings":"","what":"Gu and Wabha test functions — gw_f0","title":"Gu and Wabha test functions — gw_f0","text":"Gu Wabha test functions","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/gw_functions.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Gu and Wabha test functions — gw_f0","text":"","code":"gw_f0(x, ...)  gw_f1(x, ...)  gw_f2(x, ...)  gw_f3(x, ...)"},{"path":"https://gavinsimpson.github.io/gratia/reference/gw_functions.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Gu and Wabha test functions — gw_f0","text":"x numeric; vector points evaluate function , interval (0,1) ... arguments passed methods, ignored.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/gw_functions.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Gu and Wabha test functions — gw_f0","text":"","code":"x <- seq(0, 1, length = 6) gw_f0(x) #> [1] 0.000e+00 1.176e+00 1.902e+00 1.902e+00 1.176e+00 2.449e-16 gw_f1(x) #> [1] 1.000 1.492 2.226 3.320 4.953 7.389 gw_f2(x) #> [1] 0.000 8.591 4.261 3.199 1.100 0.000 gw_f3(x) # should be constant 0 #> [1] 0 0 0 0 0 0"},{"path":"https://gavinsimpson.github.io/gratia/reference/has_theta.html","id":null,"dir":"Reference","previous_headings":"","what":"Are additional parameters available for a GAM? — has_theta","title":"Are additional parameters available for a GAM? — has_theta","text":"additional parameters available GAM?","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/has_theta.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Are additional parameters available for a GAM? — has_theta","text":"","code":"has_theta(object)"},{"path":"https://gavinsimpson.github.io/gratia/reference/has_theta.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Are additional parameters available for a GAM? — has_theta","text":"object R object, either family() object object whose class family() method.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/has_theta.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Are additional parameters available for a GAM? — has_theta","text":"logical; TRUE additional parameters available, FALSE otherwise.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/has_theta.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Are additional parameters available for a GAM? — has_theta","text":"","code":"load_mgcv() df <- data_sim(\"eg1\", dist = \"poisson\", seed = 42, scale = 1 / 5) m <- gam(y ~ s(x0) + s(x1) + s(x2) + s(x3),   data = df, method = \"REML\",   family = nb() ) has_theta(m) #> [1] TRUE p <- theta(m)"},{"path":"https://gavinsimpson.github.io/gratia/reference/is_by_smooth.html","id":null,"dir":"Reference","previous_headings":"","what":"Tests for by variable smooths — is_by_smooth","title":"Tests for by variable smooths — is_by_smooth","text":"Functions check smooth -variable one test type -variable smooth factor-smooth continous-smooth interaction.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/is_by_smooth.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Tests for by variable smooths — is_by_smooth","text":"","code":"is_by_smooth(smooth)  is_factor_by_smooth(smooth)  is_continuous_by_smooth(smooth)  by_variable(smooth)  by_level(smooth)"},{"path":"https://gavinsimpson.github.io/gratia/reference/is_by_smooth.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Tests for by variable smooths — is_by_smooth","text":"smooth object class \"mgcv.smooth\"","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/is_by_smooth.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Tests for by variable smooths — is_by_smooth","text":"logical vector.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/is_by_smooth.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Tests for by variable smooths — is_by_smooth","text":"Gavin L. Simpson","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/is_factor_term.html","id":null,"dir":"Reference","previous_headings":"","what":"Is a model term a factor (categorical)? — is_factor_term","title":"Is a model term a factor (categorical)? — is_factor_term","text":"Given name (term label) term model, identify term factor term numeric. useful considering interactions, terms like fac1:fac2 num1:fac1 may requested user. terms type fac1:fac2 function return TRUE.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/is_factor_term.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Is a model term a factor (categorical)? — is_factor_term","text":"","code":"is_factor_term(object, term, ...)  # S3 method for class 'terms' is_factor_term(object, term, ...)  # S3 method for class 'gam' is_factor_term(object, term, ...)  # S3 method for class 'bam' is_factor_term(object, term, ...)  # S3 method for class 'gamm' is_factor_term(object, term, ...)  # S3 method for class 'list' is_factor_term(object, term, ...)"},{"path":"https://gavinsimpson.github.io/gratia/reference/is_factor_term.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Is a model term a factor (categorical)? — is_factor_term","text":"object R object method dispatch performed term character; name model term, sense attr(terms(object), \"term.labels\"). Currently checked see term exists model. ... arguments passed methods.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/is_factor_term.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Is a model term a factor (categorical)? — is_factor_term","text":"logical: TRUE variables involved term factors, otherwise FALSE.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/is_mgcv_smooth.html","id":null,"dir":"Reference","previous_headings":"","what":"Check if objects are smooths or are a particular type of smooth — is_mgcv_smooth","title":"Check if objects are smooths or are a particular type of smooth — is_mgcv_smooth","text":"Check objects smooths particular type smooth","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/is_mgcv_smooth.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Check if objects are smooths or are a particular type of smooth — is_mgcv_smooth","text":"","code":"is_mgcv_smooth(smooth)  stop_if_not_mgcv_smooth(smooth)  check_is_mgcv_smooth(smooth)  is_mrf_smooth(smooth)"},{"path":"https://gavinsimpson.github.io/gratia/reference/is_mgcv_smooth.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Check if objects are smooths or are a particular type of smooth — is_mgcv_smooth","text":"smooth R object, typically list","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/is_mgcv_smooth.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Check if objects are smooths or are a particular type of smooth — is_mgcv_smooth","text":"Check smooth inherits class \"mgcv.smooth\". stop_if_not_mgcv_smooth() wrapper around is_mgcv_smooth(), useful programming checking supplied object one mgcv's smooths, throwing consistent error . check_is_mgcv_smooth() similar stop_if_not_mgcv_smooth() returns result is_mgcv_smooth() invisibly.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/is_offset.html","id":null,"dir":"Reference","previous_headings":"","what":"Is a model term an offset? — is_offset","title":"Is a model term an offset? — is_offset","text":"Given character vector model terms, checks see , , model offset.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/is_offset.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Is a model term an offset? — is_offset","text":"","code":"is_offset(terms)"},{"path":"https://gavinsimpson.github.io/gratia/reference/is_offset.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Is a model term an offset? — is_offset","text":"terms character vector model terms.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/is_offset.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Is a model term an offset? — is_offset","text":"logical vector length terms.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/is_offset.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Is a model term an offset? — is_offset","text":"Gavin L. Simpson","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/is_offset.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Is a model term an offset? — is_offset","text":"","code":"load_mgcv() df <- data_sim(\"eg1\", n = 400, dist = \"normal\") m <- gam(y ~ s(x0) + s(x1) + offset(x0), data = df, method = \"REML\") nm <- names(model.frame(m)) nm #> [1] \"y\"          \"offset(x0)\" \"x0\"         \"x1\"         is_offset(nm) #> [1] FALSE  TRUE FALSE FALSE"},{"path":"https://gavinsimpson.github.io/gratia/reference/link.html","id":null,"dir":"Reference","previous_headings":"","what":"Extract link and inverse link functions from models — link","title":"Extract link and inverse link functions from models — link","text":"Returns link inverse estimated model, provides simple way extract functions complex models multiple links, location scale models.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/link.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Extract link and inverse link functions from models — link","text":"","code":"link(object, ...)  # S3 method for class 'family' link(object, parameter = NULL, which_eta = NULL, ...)  # S3 method for class 'gam' link(object, parameter = NULL, which_eta = NULL, ...)  # S3 method for class 'bam' link(object, parameter = NULL, which_eta = NULL, ...)  # S3 method for class 'gamm' link(object, ...)  # S3 method for class 'glm' link(object, ...)  # S3 method for class 'list' link(object, ...)  inv_link(object, ...)  # S3 method for class 'family' inv_link(object, parameter = NULL, which_eta = NULL, ...)  # S3 method for class 'gam' inv_link(object, parameter = NULL, which_eta = NULL, ...)  # S3 method for class 'bam' inv_link(object, parameter = NULL, which_eta = NULL, ...)  # S3 method for class 'gamm' inv_link(object, ...)  # S3 method for class 'list' inv_link(object, ...)  # S3 method for class 'glm' inv_link(object, ...)  extract_link(family, ...)  # S3 method for class 'family' extract_link(family, inverse = FALSE, ...)  # S3 method for class 'general.family' extract_link(family, parameter, inverse = FALSE, which_eta = NULL, ...)"},{"path":"https://gavinsimpson.github.io/gratia/reference/link.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Extract link and inverse link functions from models — link","text":"object family object fitted model extract family object.  Models fitted stats::glm(), mgcv::gam(), mgcv::bam(), mgcv::gamm(), gamm4::gamm4() currently supported. ... arguments passed methods. parameter character; parameter distribution. Usually \"location\" \"scale\" \"shape\" may provided location scale models. options include \"mu\" synonym \"location\", \"sigma\" scale parameter mgcv::gaulss(), \"pi\" zero-inflation term mgcv::ziplss(), \"power\" mgcv::twlss() power parameter, \"xi\", shape parameter mgcv::gevlss(), \"epsilon\" \"skewness\" skewness \"delta\" \"kurtosis\" kurtosis parameter mgcv::shash(), \"phi\" scale parameter mgcv::gammals() & mgcv::twlss(). which_eta numeric; linear predictor extract families mgcv::mvn() mgcv::multinom(). family family object, result call family(). inverse logical; return inverse link function?","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/link.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Extract link and inverse link functions from models — link","text":"Gavin L. Simpson","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/link.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Extract link and inverse link functions from models — link","text":"","code":"load_mgcv()  link(gaussian()) #> function (mu)  #> mu #> <environment: namespace:stats> link(nb()) #> function (mu)  #> log(mu) #> <environment: namespace:stats>  inv_link(nb()) #> function (eta)  #> pmax(exp(eta), .Machine$double.eps) #> <environment: namespace:stats>  dat <- data_sim(\"eg1\", seed = 4234) mod <- gam(list(y ~ s(x0) + s(x1) + s(x2) + s(x3), ~1),   data = dat,   family = gaulss )  link(mod, parameter = \"scale\") #> function (mu)  #> log(1/mu - 0.01) #> <environment: 0x563ac474a800> inv_link(mod, parameter = \"scale\") #> function (eta)  #> 1/(exp(eta) + 0.01) #> <environment: 0x563ac474a800>  ## Works with `family` objects too link(shash(), parameter = \"skewness\") #> function (mu)  #> mu #> <environment: namespace:stats>"},{"path":"https://gavinsimpson.github.io/gratia/reference/load_mgcv.html","id":null,"dir":"Reference","previous_headings":"","what":"Load mgcv quietly — load_mgcv","title":"Load mgcv quietly — load_mgcv","text":"Simple function loads mgcv package whilst suppressing startup messages prints console.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/load_mgcv.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Load mgcv quietly — load_mgcv","text":"","code":"load_mgcv()"},{"path":"https://gavinsimpson.github.io/gratia/reference/load_mgcv.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Load mgcv quietly — load_mgcv","text":"Returns logical vectors invisibly, indicating whether package loaded .","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/lp_matrix.html","id":null,"dir":"Reference","previous_headings":"","what":"Return the linear prediction matrix of a fitted GAM — lp_matrix","title":"Return the linear prediction matrix of a fitted GAM — lp_matrix","text":"lp_matrix() wrapper predict(..., type = \"lpmatrix\") returning linear predictor matrix model training data (data = NULL), user-specified data values supplied via data.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/lp_matrix.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Return the linear prediction matrix of a fitted GAM — lp_matrix","text":"","code":"lp_matrix(model, ...)  # S3 method for class 'gam' lp_matrix(model, data = NULL, ...)"},{"path":"https://gavinsimpson.github.io/gratia/reference/lp_matrix.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Return the linear prediction matrix of a fitted GAM — lp_matrix","text":"model fitted model ... arguments passed methods predict methods including mgcv::predict.gam() mgcv::predict.bam() data data frame values return linear prediction matrix.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/lp_matrix.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Return the linear prediction matrix of a fitted GAM — lp_matrix","text":"linear prediction matrix returned matrix. object returned class \"lp_matrix\", inherits classes \"matrix\" \"array\". special class allows printing matrix controlled, printing matrix tibble.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/lp_matrix.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Return the linear prediction matrix of a fitted GAM — lp_matrix","text":"linear prediction matrix \\(\\mathbf{X}_p\\) matrix maps values parameters \\(\\hat{\\mathbf{\\beta}}_p\\) values linear predictor model \\(\\hat{\\eta}_p = \\mathbf{X}_p \\hat{\\mathbf{\\beta}}_p\\). \\(\\mathbf{X}_p\\) model matrix spline covariates replaced values basis functions evaluated respective covariates. Parametric covariates also included.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/lp_matrix.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Return the linear prediction matrix of a fitted GAM — lp_matrix","text":"","code":"load_mgcv() df <- data_sim(\"eg1\", seed = 1) m <- gam(y ~ s(x0) + s(x1) + s(x2) + s(x3), data = df)  # linear prediction matrix for observed data xp <- lp_matrix(m) ## IGNORE_RDIFF_BEGIN xp #> Linear prediction matrix (400 x 37) #>   `(Intercept)` `s(x0).1` `s(x0).2` `s(x0).3` `s(x0).4` `s(x0).5` `s(x0).6` #>           <dbl>     <dbl>     <dbl>     <dbl>     <dbl>     <dbl>     <dbl> #> 1             1     0.961     0.227    0.706     -0.135     0.457  -0.146   #> 2             1     0.651    -0.241    0.0684    -0.308     0.394  -0.00994 #> 3             1    -0.385    -0.549    0.0660    -0.204    -0.416  -0.247   #> 4             1    -1.27      0.156   -1.53       0.222    -1.60    0.198   #> 5             1     1.05      0.420    1.11       0.214     0.893   0.0351  #> # i 395 more rows ## IGNORE_RDIFF_END  # the object `xp` *is* a matrix class(xp) #> [1] \"lp_matrix\" \"matrix\"    \"array\"     # but we print like a tibble to avoid spamming the R console  # linear predictor matrix for new data set ds <- data_slice(m, x2 = evenly(x2)) xp <- lp_matrix(m, data = ds) ## IGNORE_RDIFF_BEGIN xp #> Linear prediction matrix (100 x 37) #>   `(Intercept)` `s(x0).1` `s(x0).2` `s(x0).3` `s(x0).4` `s(x0).5` `s(x0).6` #>           <dbl>     <dbl>     <dbl>     <dbl>     <dbl>     <dbl>     <dbl> #> 1             1     0.170    -0.542   -0.0371   -0.0534     0.144    -0.353 #> 2             1     0.170    -0.542   -0.0371   -0.0534     0.144    -0.353 #> 3             1     0.170    -0.542   -0.0371   -0.0534     0.144    -0.353 #> 4             1     0.170    -0.542   -0.0371   -0.0534     0.144    -0.353 #> 5             1     0.170    -0.542   -0.0371   -0.0534     0.144    -0.353 #> # i 95 more rows ## IGNORE_RDIFF_END"},{"path":"https://gavinsimpson.github.io/gratia/reference/lss_parameters.html","id":null,"dir":"Reference","previous_headings":"","what":"General names of LSS parameters for each GAM family — lss_parameters","title":"General names of LSS parameters for each GAM family — lss_parameters","text":"General names LSS parameters GAM family","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/lss_parameters.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"General names of LSS parameters for each GAM family — lss_parameters","text":"","code":"lss_parameters(object)"},{"path":"https://gavinsimpson.github.io/gratia/reference/mh_draws.html","id":null,"dir":"Reference","previous_headings":"","what":"Posterior samples using a Gaussian approximation to the posterior distribution — mh_draws","title":"Posterior samples using a Gaussian approximation to the posterior distribution — mh_draws","text":"Posterior samples using Gaussian approximation posterior distribution","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/mh_draws.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Posterior samples using a Gaussian approximation to the posterior distribution — mh_draws","text":"","code":"mh_draws(model, ...)  # S3 method for class 'gam' mh_draws(   model,   n,   burnin = 1000,   thin = 1,   t_df = 40,   rw_scale = 0.25,   index = NULL,   ... )"},{"path":"https://gavinsimpson.github.io/gratia/reference/mh_draws.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Posterior samples using a Gaussian approximation to the posterior distribution — mh_draws","text":"model fitted R model. Currently models fitted mgcv::gam() mgcv::bam(), return object inherits objects supported. , \"inherits\" used loose fashion; models fitted scam::scam() support even though models strictly inherit class \"gam\" far inherits() concerned. ... arguments passed methods. n numeric; number posterior draws take. burnin numeric; length initial burn period discard. See mgcv::gam.mh(). thin numeric; retain thin samples. See mgcv::gam.mh(). t_df numeric; degrees freedom static multivariate t proposal. See mgcv::gam.mh(). rw_scale numeric; factor scale posterior covariance matrix generating random walk proposals. See mgcv::gam.mh(). index numeric; vector indices coefficients use. Can used subset mean vector covariance matrix extracted model.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/model_concurvity.html","id":null,"dir":"Reference","previous_headings":"","what":"Concurvity of an estimated GAM — model_concurvity","title":"Concurvity of an estimated GAM — model_concurvity","text":"Concurvity estimated GAM","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/model_concurvity.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Concurvity of an estimated GAM — model_concurvity","text":"","code":"model_concurvity(model, ...)  # S3 method for class 'gam' model_concurvity(   model,   terms = everything(),   type = c(\"all\", \"estimate\", \"observed\", \"worst\"),   pairwise = FALSE,   ... )  concrvity(   model,   terms = everything(),   type = c(\"all\", \"estimate\", \"observed\", \"worst\"),   pairwise = FALSE,   ... )"},{"path":"https://gavinsimpson.github.io/gratia/reference/model_concurvity.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Concurvity of an estimated GAM — model_concurvity","text":"model fitted GAM. Currently objects class \"gam\" supported ... arguents passed methods. terms currently ignored type character; pairwise logical; extract pairwise concurvity model terms?","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/model_concurvity.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Concurvity of an estimated GAM — model_concurvity","text":"","code":"## simulate data with concurvity... library(\"tibble\") load_mgcv() set.seed(8) n <- 200 df <- tibble(   t = sort(runif(n)),   x = gw_f2(t) + rnorm(n) * 3,   y = sin(4 * pi * t) + exp(x / 20) + rnorm(n) * 0.3 )  ## fit model m <- gam(y ~ s(t, k = 15) + s(x, k = 15), data = df, method = \"REML\")  ## overall concurvity o_conc <- concrvity(m) draw(o_conc)   ## pairwise concurvity p_conc <- concrvity(m, pairwise = TRUE) draw(p_conc)"},{"path":"https://gavinsimpson.github.io/gratia/reference/model_constant.html","id":null,"dir":"Reference","previous_headings":"","what":"Extract the model constant term — model_constant","title":"Extract the model constant term — model_constant","text":"Extracts model constant term(s), model intercept, fitted model object.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/model_constant.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Extract the model constant term — model_constant","text":"","code":"model_constant(model, ...)  # S3 method for class 'gam' model_constant(model, lp = NULL, ...)  # S3 method for class 'gamlss' model_constant(model, ...)"},{"path":"https://gavinsimpson.github.io/gratia/reference/model_constant.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Extract the model constant term — model_constant","text":"model fitted model coef() method exists. ... arguments passed methods. lp numeric; linear predictors extract constant terms .","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/model_constant.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Extract the model constant term — model_constant","text":"","code":"load_mgcv()  # simulate a small example df <- data_sim(\"eg1\", seed = 42)  # fit the GAM m <- gam(y ~ s(x0) + s(x1) + s(x2) + s(x3), data = df, method = \"REML\")  # extract the estimate of the constant term model_constant(m) #> [1] 7.495 #> attr(,\"par_names\") #> [1] \"location\" # same as coef(m)[1L] coef(m)[1L] #> (Intercept)  #>       7.495"},{"path":"https://gavinsimpson.github.io/gratia/reference/model_vars.html","id":null,"dir":"Reference","previous_headings":"","what":"List the variables involved in a model fitted with a formula — model_vars","title":"List the variables involved in a model fitted with a formula — model_vars","text":"List variables involved model fitted formula","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/model_vars.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"List the variables involved in a model fitted with a formula — model_vars","text":"","code":"model_vars(model, ...)  # S3 method for class 'gam' model_vars(model, ...)  # Default S3 method model_vars(model, ...)  # S3 method for class 'bam' model_vars(model, ...)  # S3 method for class 'gamm' model_vars(model, ...)  # S3 method for class 'gamm4' model_vars(model, ...)  # S3 method for class 'list' model_vars(model, ...)"},{"path":"https://gavinsimpson.github.io/gratia/reference/model_vars.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"List the variables involved in a model fitted with a formula — model_vars","text":"model fitted model object $pred.formula, $terms component \"terms\" attribute ... Arguments passed methods. Currently ignored.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/model_vars.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"List the variables involved in a model fitted with a formula — model_vars","text":"","code":"load_mgcv()  # simulate some Gaussian data df <- data_sim(\"eg1\", n = 50, seed = 2)  # fit a GAM with 1 smooth and 1 linear term m1 <- gam(y ~ s(x2, k = 7) + x1, data = df, method = \"REML\") model_vars(m1) #> [1] \"x1\" \"x2\"  # fit a lm with two linear terms m2 <- lm(y ~ x2 + x1, data = df) model_vars(m2) #> [1] \"x2\" \"x1\""},{"path":"https://gavinsimpson.github.io/gratia/reference/n_eta.html","id":null,"dir":"Reference","previous_headings":"","what":"The Number of linear predictors in model — n_eta","title":"The Number of linear predictors in model — n_eta","text":"Extracts number linear predictors fitted model.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/n_eta.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"The Number of linear predictors in model — n_eta","text":"","code":"n_eta(model, ...)  # S3 method for class 'gam' n_eta(model, ...)"},{"path":"https://gavinsimpson.github.io/gratia/reference/n_eta.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"The Number of linear predictors in model — n_eta","text":"model fitted model. Currently, models inheriting class \"gam\" supported. ... arguments passed methods.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/n_eta.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"The Number of linear predictors in model — n_eta","text":"integer vector length 1 containing number linear predictors model.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/n_smooths.html","id":null,"dir":"Reference","previous_headings":"","what":"How many smooths in a fitted model — n_smooths","title":"How many smooths in a fitted model — n_smooths","text":"many smooths fitted model","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/n_smooths.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"How many smooths in a fitted model — n_smooths","text":"","code":"n_smooths(object)  # Default S3 method n_smooths(object)  # S3 method for class 'gam' n_smooths(object)  # S3 method for class 'gamm' n_smooths(object)  # S3 method for class 'bam' n_smooths(object)"},{"path":"https://gavinsimpson.github.io/gratia/reference/n_smooths.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"How many smooths in a fitted model — n_smooths","text":"object fitted GAM related model. Typically result call mgcv::gam(), mgcv::bam(), mgcv::gamm().","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/nb_theta.html","id":null,"dir":"Reference","previous_headings":"","what":"Negative binomial parameter theta — nb_theta","title":"Negative binomial parameter theta — nb_theta","text":"Negative binomial parameter theta","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/nb_theta.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Negative binomial parameter theta — nb_theta","text":"","code":"nb_theta(model)  # S3 method for class 'gam' nb_theta(model)"},{"path":"https://gavinsimpson.github.io/gratia/reference/nb_theta.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Negative binomial parameter theta — nb_theta","text":"model fitted model.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/nb_theta.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Negative binomial parameter theta — nb_theta","text":"numeric vector length 1 containing estimated value theta.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/nb_theta.html","id":"methods-by-class-","dir":"Reference","previous_headings":"","what":"Methods (by class)","title":"Negative binomial parameter theta — nb_theta","text":"nb_theta(gam): Method class \"gam\"","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/nb_theta.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Negative binomial parameter theta — nb_theta","text":"","code":"load_mgcv() df <- data_sim(\"eg1\", n = 500, dist = \"poisson\", scale = 0.1, seed = 6)  m <- gam(y ~ s(x0, bs = \"cr\") + s(x1, bs = \"cr\") + s(x2, bs = \"cr\") +   s(x3, bs = \"cr\"), family = nb, data = df, method = \"REML\") ## IGNORE_RDIFF_BEGIN nb_theta(m) #> [1] 239333.8 ## IGNORE_RDIFF_END"},{"path":"https://gavinsimpson.github.io/gratia/reference/nested_partial_residuals.html","id":null,"dir":"Reference","previous_headings":"","what":"Partial residuals in nested form — nested_partial_residuals","title":"Partial residuals in nested form — nested_partial_residuals","text":"Computes partial residuals smooth terms, formats long/tidy format, nests partial_residual column result nested data frame one row per smooth.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/nested_partial_residuals.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Partial residuals in nested form — nested_partial_residuals","text":"","code":"nested_partial_residuals(object, terms = NULL, data = NULL)"},{"path":"https://gavinsimpson.github.io/gratia/reference/nested_partial_residuals.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Partial residuals in nested form — nested_partial_residuals","text":"object fitted GAM model terms vector terms include partial residuals . Passed argument terms mgcv::predict.gam()]. data optional data frame","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/nested_partial_residuals.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Partial residuals in nested form — nested_partial_residuals","text":"nested tibble (data frame) one row per smooth term. Contains two columns: smooth - label indicating smooth term partial_residual - list column containing tibble (data frame) 1 column partial_residual containing requested partial residuals indicated smooth.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/nested_rug_values.html","id":null,"dir":"Reference","previous_headings":"","what":"Values for rug plot in nested form — nested_rug_values","title":"Values for rug plot in nested form — nested_rug_values","text":"Extracts original data smooth terms, formats long/tidy format, nests data column(s) result nested data frame one row per smooth.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/nested_rug_values.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Values for rug plot in nested form — nested_rug_values","text":"","code":"nested_rug_values(object, terms = NULL, data = NULL)"},{"path":"https://gavinsimpson.github.io/gratia/reference/nested_rug_values.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Values for rug plot in nested form — nested_rug_values","text":"object fitted GAM model terms vector terms include original data . Passed argument terms mgcv::predict.gam()]. data optional data frame","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/nested_rug_values.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Values for rug plot in nested form — nested_rug_values","text":"nested tibble (data frame) one row per smooth term.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/null_deviance.html","id":null,"dir":"Reference","previous_headings":"","what":"Extract the null deviance of a fitted model — null_deviance","title":"Extract the null deviance of a fitted model — null_deviance","text":"Extract null deviance fitted model","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/null_deviance.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Extract the null deviance of a fitted model — null_deviance","text":"","code":"null_deviance(model, ...)  # Default S3 method null_deviance(model, ...)"},{"path":"https://gavinsimpson.github.io/gratia/reference/null_deviance.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Extract the null deviance of a fitted model — null_deviance","text":"model fitted model ... arguments passed methods","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/observed_fitted_plot.html","id":null,"dir":"Reference","previous_headings":"","what":"Plot of fitted against observed response values — observed_fitted_plot","title":"Plot of fitted against observed response values — observed_fitted_plot","text":"Plot fitted observed response values","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/observed_fitted_plot.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Plot of fitted against observed response values — observed_fitted_plot","text":"","code":"observed_fitted_plot(   model,   ylab = NULL,   xlab = NULL,   title = NULL,   subtitle = NULL,   caption = NULL,   point_col = \"black\",   point_alpha = 1 )"},{"path":"https://gavinsimpson.github.io/gratia/reference/observed_fitted_plot.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Plot of fitted against observed response values — observed_fitted_plot","text":"model fitted model. Currently class \"gam\". ylab character expression; label y axis. supplied, suitable label generated. xlab character expression; label y axis. supplied, suitable label generated. title character expression; title plot. See ggplot2::labs(). subtitle character expression; subtitle plot. See ggplot2::labs(). caption character expression; plot caption. See ggplot2::labs(). point_col colour used draw points plots. See graphics::par() section Color Specification. passed individual plotting functions, therefore affects points plots. point_alpha numeric; alpha transparency points plots.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/overview.html","id":null,"dir":"Reference","previous_headings":"","what":"Provides an overview of a model and the terms in that model — overview","title":"Provides an overview of a model and the terms in that model — overview","text":"Provides overview model terms model","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/overview.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Provides an overview of a model and the terms in that model — overview","text":"","code":"overview(model, ...)  # S3 method for class 'gam' overview(   model,   parametric = TRUE,   random_effects = TRUE,   dispersion = NULL,   frequentist = FALSE,   accuracy = 0.001,   stars = FALSE,   ... )"},{"path":"https://gavinsimpson.github.io/gratia/reference/overview.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Provides an overview of a model and the terms in that model — overview","text":"model fitted model object overview. ... arguments passed methods. parametric logical; include model parametric terms overview? random_effects tests fully penalized smooth terms (zero-dimensional null space, e.g. random effects) computationally expensive large data sets producing p values can take long time. random_effects = FALSE, tests expensive terms skipped. dispersion numeric; known value dispersion parameter. default NULL implies estimated value default value (1 Poisson distribution example) specified used instead. frequentist logical; default Bayesian estimated covariance matrix parameter estimates used calculate p values parametric terms. frequentist = FALSE, frequentist covariance matrix parameter estimates used. accuracy numeric; accuracy report p values, p values value displayed \"< accuracy\". stars logical; significance stars added output?","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/overview.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Provides an overview of a model and the terms in that model — overview","text":"","code":"load_mgcv() df <- data_sim(n = 400, seed = 2) m <- gam(y ~ x3 + s(x0) + s(x1, bs = \"bs\") + s(x2, bs = \"ts\"),   data = df, method = \"REML\" ) overview(m) #>  #> Generalized Additive Model with 4 terms #>  #>   term  type              k   edf statistic p.value #>   <chr> <chr>         <dbl> <dbl>     <dbl> <chr>   #> 1 x3    parametric       NA  1         4.28 0.03926 #> 2 s(x0) TPRS              9  3.02      6.25 < 0.001 #> 3 s(x1) B spline          9  2.81     71.0  < 0.001 #> 4 s(x2) TPRS (shrink)     9  7.91     83.8  < 0.001"},{"path":"https://gavinsimpson.github.io/gratia/reference/parametric_effects.html","id":null,"dir":"Reference","previous_headings":"","what":"Estimated values for parametric model terms — parametric_effects","title":"Estimated values for parametric model terms — parametric_effects","text":"Estimated values parametric model terms","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/parametric_effects.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Estimated values for parametric model terms — parametric_effects","text":"","code":"parametric_effects(object, ...)  # S3 method for class 'gam' parametric_effects(   object,   terms = NULL,   data = NULL,   unconditional = FALSE,   unnest = TRUE,   ci_level = 0.95,   envir = environment(formula(object)),   transform = FALSE,   ... )"},{"path":"https://gavinsimpson.github.io/gratia/reference/parametric_effects.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Estimated values for parametric model terms — parametric_effects","text":"object fitted model object. ... arguments passed methods. terms character; model parametric terms drawn? Default NULL plot parametric terms can drawn. data optional data frame may may used? FIXME! unconditional logical; confidence intervals include uncertainty due smoothness selection? TRUE, corrected Bayesian covariance matrix used. unnest logical; unnest parametric effect objects? ci_level numeric; coverage required confidence interval. Currently ignored. envir environment look data within. transform logical; TRUE, parametric effect plotted transformed scale result effect straight line. FALSE, effect plotted raw data (.e. log10(x), poly(z), x-axis plot x z respectively.)","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/parametric_terms.html","id":null,"dir":"Reference","previous_headings":"","what":"Names of any parametric terms in a GAM — parametric_terms","title":"Names of any parametric terms in a GAM — parametric_terms","text":"Names parametric terms GAM","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/parametric_terms.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Names of any parametric terms in a GAM — parametric_terms","text":"","code":"parametric_terms(model, ...)  # Default S3 method parametric_terms(model, ...)  # S3 method for class 'gam' parametric_terms(model, ...)"},{"path":"https://gavinsimpson.github.io/gratia/reference/parametric_terms.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Names of any parametric terms in a GAM — parametric_terms","text":"model fitted model. ... arguments passed methods.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/partial_derivatives.html","id":null,"dir":"Reference","previous_headings":"","what":"Partial derivatives of estimated multivariate smooths via finite differences — partial_derivatives","title":"Partial derivatives of estimated multivariate smooths via finite differences — partial_derivatives","text":"Partial derivatives estimated multivariate smooths via finite differences","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/partial_derivatives.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Partial derivatives of estimated multivariate smooths via finite differences — partial_derivatives","text":"","code":"partial_derivatives(object, ...)  # Default S3 method partial_derivatives(object, ...)  # S3 method for class 'gamm' partial_derivatives(object, ...)  # S3 method for class 'gam' partial_derivatives(   object,   select = NULL,   term = deprecated(),   focal = NULL,   data = newdata,   order = 1L,   type = c(\"forward\", \"backward\", \"central\"),   n = 100,   eps = 1e-07,   interval = c(\"confidence\", \"simultaneous\"),   n_sim = 10000,   level = 0.95,   unconditional = FALSE,   frequentist = FALSE,   offset = NULL,   ncores = 1,   partial_match = FALSE,   seed = NULL,   ...,   newdata = NULL )"},{"path":"https://gavinsimpson.github.io/gratia/reference/partial_derivatives.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Partial derivatives of estimated multivariate smooths via finite differences — partial_derivatives","text":"object R object compute derivatives . ... arguments passed methods. select character; vector one smooth terms derivatives required. missing, derivatives smooth terms returned. Can partial match smooth term; see argument partial_match . term Use select instead. focal character; name focal variable. partial derivative estimated smooth respect variable returned. variables involved smooth held constant. can missing supplying data, case, focal variable identified one variable constant. data data frame containing values model covariates evaluate first derivatives smooths. supplied, one variable must held constant value. order numeric; order derivative. type character; type finite difference used. One \"forward\", \"backward\", \"central\". n numeric; number points evaluate derivative . eps numeric; finite difference. interval character; type interval compute. One \"confidence\" point-wise intervals, \"simultaneous\" simultaneous intervals. n_sim integer; number simulations used computing simultaneous intervals. level numeric; 0 < level < 1; confidence level point-wise simultaneous interval. default 0.95 95% interval. unconditional logical; use smoothness selection-corrected Bayesian covariance matrix? frequentist logical; use frequentist covariance matrix? offset numeric; value use offset term ncores number cores generating random variables multivariate normal distribution. Passed mvnfast::rmvn(). Parallelization take place OpenMP supported (appears work Windows current R). partial_match logical; smooths selected partial matches term? TRUE, term can single string match . seed numeric; RNG seed use. newdata Deprecated: use data instead.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/partial_derivatives.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Partial derivatives of estimated multivariate smooths via finite differences — partial_derivatives","text":"tibble, currently following variables: .smooth: smooth row refers , .partial_deriv: estimated partial derivative, .se: standard error estimated partial derivative, .crit: critical value derivative ± (crit * se) gives upper lower bounds requested confidence simultaneous interval (given level), .lower_ci: lower bound confidence simultaneous interval, .upper_ci: upper bound confidence simultaneous interval.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/partial_derivatives.html","id":"note","dir":"Reference","previous_headings":"","what":"Note","title":"Partial derivatives of estimated multivariate smooths via finite differences — partial_derivatives","text":"partial_derivatives() ignore random effect smooths encounters object.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/partial_derivatives.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Partial derivatives of estimated multivariate smooths via finite differences — partial_derivatives","text":"Gavin L. Simpson","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/partial_derivatives.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Partial derivatives of estimated multivariate smooths via finite differences — partial_derivatives","text":"","code":"library(\"ggplot2\") library(\"patchwork\") load_mgcv() df <- data_sim(\"eg2\", n = 2000, dist = \"normal\", scale = 0.5, seed = 42)  # fit the GAM (note: for execution time reasons, k is set articifially low) m <- gam(y ~ te(x, z, k = c(5, 5)), data = df, method = \"REML\")  # data slice through te(x,z) holding z == 0.4 ds <- data_slice(m, x = evenly(x, n = 100), z = 0.4)  # evaluate te(x,z) at values of x & z sm <- smooth_estimates(m, select = \"te(x,z)\", data = ds) |>   add_confint()  # partial derivatives pd_x <- partial_derivatives(m, data = ds, type = \"central\", focal = \"x\")  # draw te(x,z) p1 <- draw(m, rug = FALSE) &   geom_hline(yintercept = 0.4, linewidth = 1) p1   # draw te(x,z) along slice cap <- expression(z == 0.4) p2 <- sm |>   ggplot(aes(x = x, y = .estimate)) +   geom_ribbon(aes(ymin = .lower_ci, ymax = .upper_ci), alpha = 0.2) +   geom_line() +   labs(     x = \"x\", y = \"Partial effect\", title = \"te(x,z)\",     caption = cap   ) p2   # draw partial derivs p3 <- pd_x |>   draw() +   labs(caption = cap) p3   # draw all three panels p1 + p2 + p3 + plot_layout(ncol = 3)"},{"path":"https://gavinsimpson.github.io/gratia/reference/partial_residuals.html","id":null,"dir":"Reference","previous_headings":"","what":"Partial residuals — partial_residuals","title":"Partial residuals — partial_residuals","text":"Partial residuals","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/partial_residuals.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Partial residuals — partial_residuals","text":"","code":"partial_residuals(object, ...)  # S3 method for class 'gam' partial_residuals(object, select = NULL, partial_match = FALSE, ...)"},{"path":"https://gavinsimpson.github.io/gratia/reference/partial_residuals.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Partial residuals — partial_residuals","text":"object R object, typically model. Currently objects class \"gam\" (inherit class) supported. ... arguments passed methods. select character, logical, numeric; smooths plot. NULL, default, model smooths drawn. Numeric select indexes smooths order specified formula stored object. Character select matches labels smooths shown example output summary(object). Logical select operates per numeric select order smooths stored. partial_match logical; smooths selected partial matches select? TRUE, select can single string match .","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/partial_residuals.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Partial residuals — partial_residuals","text":"","code":"## load mgcv load_mgcv()  ## example data - Gu & Wabha four term model df <- data_sim(\"eg1\", n = 400, seed = 42) ## fit the model m <- gam(y ~ s(x0) + s(x1) + s(x2) + s(x3), data = df, method = \"REML\")  ## extract partial residuals partial_residuals(m) #> # A tibble: 400 x 4 #>    `s(x0)`  `s(x1)` `s(x2)`  `s(x3)` #>      <dbl>    <dbl>   <dbl>    <dbl> #>  1 -0.3527 -1.321   -2.180   0.6077  #>  2 -0.1233  0.5013  -1.775   0.9613  #>  3  1.429   1.515    5.609   0.9910  #>  4 -1.110  -1.700   -0.8882 -0.6593  #>  5 -2.120  -0.01378 -2.733  -3.012   #>  6  1.254  -1.224    3.915   0.07275 #>  7 -0.5220  3.023   -0.8197 -1.019   #>  8  1.398   0.2184   7.055   1.897   #>  9  2.797   0.4969   7.329   2.498   #> 10  1.151  -0.2267   0.7202  0.7437  #> # i 390 more rows  ## and for a select term partial_residuals(m, select = \"s(x2)\") #> # A tibble: 400 x 1 #>    `s(x2)` #>      <dbl> #>  1 -2.180  #>  2 -1.775  #>  3  5.609  #>  4 -0.8882 #>  5 -2.733  #>  6  3.915  #>  7 -0.8197 #>  8  7.055  #>  9  7.329  #> 10  0.7202 #> # i 390 more rows  ## or with partial matching partial_residuals(m, select = \"x\", partial_match = TRUE) # returns all #> # A tibble: 400 x 4 #>    `s(x0)`  `s(x1)` `s(x2)`  `s(x3)` #>      <dbl>    <dbl>   <dbl>    <dbl> #>  1 -0.3527 -1.321   -2.180   0.6077  #>  2 -0.1233  0.5013  -1.775   0.9613  #>  3  1.429   1.515    5.609   0.9910  #>  4 -1.110  -1.700   -0.8882 -0.6593  #>  5 -2.120  -0.01378 -2.733  -3.012   #>  6  1.254  -1.224    3.915   0.07275 #>  7 -0.5220  3.023   -0.8197 -1.019   #>  8  1.398   0.2184   7.055   1.897   #>  9  2.797   0.4969   7.329   2.498   #> 10  1.151  -0.2267   0.7202  0.7437  #> # i 390 more rows"},{"path":"https://gavinsimpson.github.io/gratia/reference/penalty.html","id":null,"dir":"Reference","previous_headings":"","what":"Extract and tidy penalty matrices — penalty","title":"Extract and tidy penalty matrices — penalty","text":"Extract tidy penalty matrices","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/penalty.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Extract and tidy penalty matrices — penalty","text":"","code":"penalty(object, ...)  # Default S3 method penalty(   object,   rescale = FALSE,   data,   knots = NULL,   constraints = FALSE,   diagonalize = FALSE,   ... )  # S3 method for class 'gam' penalty(   object,   select = NULL,   smooth = deprecated(),   rescale = FALSE,   partial_match = FALSE,   ... )  # S3 method for class 'mgcv.smooth' penalty(object, rescale = FALSE, ...)  # S3 method for class 'tensor.smooth' penalty(object, margins = FALSE, ...)  # S3 method for class 't2.smooth' penalty(object, margins = FALSE, ...)  # S3 method for class 're.smooth.spec' penalty(object, data, ...)"},{"path":"https://gavinsimpson.github.io/gratia/reference/penalty.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Extract and tidy penalty matrices — penalty","text":"object fitted GAM smooth. ... additional arguments passed methods. rescale logical; default, mgcv scale penalty matrix better performance mgcv::gamm(). rescale TRUE, scaling undone put penalty matrix back original scale. data data frame; data frame values terms mentioned smooth specification. knots list data frame named components containing knots locations. Names must match covariates basis required. See mgcv::smoothCon(). constraints logical; identifiability constraints applied smooth basis. See argument absorb.cons mgcv::smoothCon(). diagonalize logical; TRUE, reparameterises smooth associated penalty identity matrix. effect turning last diagonal elements penalty zero, highlights penalty null space. select character, logical, numeric; smooths extract penalties . NULL, default, penalties model smooths drawn. Numeric select indexes smooths order specified formula stored object. Character select matches labels smooths shown example output summary(object). Logical select operates per numeric select order smooths stored. smooth Use select instead. partial_match logical; smooths selected partial matches select? TRUE, select can single string match . margins logical; extract penalty matrices tensor product marginal smooths tensor product?","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/penalty.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Extract and tidy penalty matrices — penalty","text":"'tibble' (data frame) class penalty_df inheriting tbl_df, following components: .smooth - character; label mgcv uses refer smooth, .type - character; type smooth, .penalty - character; label specific penalty. smooths multiple penalty matrices, penalty component identifies particular penalty matrix uses labelling mgcv uses internally, .row - character; label form fn n integer nth basis function, referencing columns penalty matrix, .col - character; label form fn n integer nth basis function, referencing columns penalty matrix, .value - double; value penalty matrix combination row col,","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/penalty.html","id":"note","dir":"Reference","previous_headings":"","what":"Note","title":"Extract and tidy penalty matrices — penalty","text":"print() method uses base::zapsmall() turn small numbers 0s display purposes ; underlying values penalty matrix matrices changed. smooths subject eigendecomposition (e.g. default thin plate regression splines, bs = \"tp\"), signs eigenvectors defined can expect differences across systems penalties smooths system-, OS-, CPU architecture- specific.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/penalty.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Extract and tidy penalty matrices — penalty","text":"Gavin L. Simpson","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/penalty.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Extract and tidy penalty matrices — penalty","text":"","code":"load_mgcv() dat <- data_sim(\"eg4\", n = 400, seed = 42) m <- gam(   y ~ s(x0, bs = \"cr\") + s(x1, bs = \"cr\") +     s(x2, by = fac, bs = \"cr\"),   data = dat, method = \"REML\" )  # penalties for all smooths penalty(m) #> # A tibble: 405 x 6 #>    .smooth .type .penalty .row  .col   .value #>    <chr>   <chr> <chr>    <chr> <chr>   <dbl> #>  1 s(x0)   CRS   s(x0)    F1    F1     0.783  #>  2 s(x0)   CRS   s(x0)    F1    F2    -0.635  #>  3 s(x0)   CRS   s(x0)    F1    F3     0.265  #>  4 s(x0)   CRS   s(x0)    F1    F4    -0.0203 #>  5 s(x0)   CRS   s(x0)    F1    F5     0.0441 #>  6 s(x0)   CRS   s(x0)    F1    F6     0.0378 #>  7 s(x0)   CRS   s(x0)    F1    F7     0.0482 #>  8 s(x0)   CRS   s(x0)    F1    F8     0.0216 #>  9 s(x0)   CRS   s(x0)    F1    F9     0.0247 #> 10 s(x0)   CRS   s(x0)    F2    F1    -0.635  #> # i 395 more rows  # for a specific smooth penalty(m, select = \"s(x2):fac1\") #> # A tibble: 81 x 6 #>    .smooth    .type .penalty   .row  .col   .value #>    <chr>      <chr> <chr>      <chr> <chr>   <dbl> #>  1 s(x2):fac1 CRS   s(x2):fac1 F1    F1     1.66   #>  2 s(x2):fac1 CRS   s(x2):fac1 F1    F2    -0.755  #>  3 s(x2):fac1 CRS   s(x2):fac1 F1    F3     0.430  #>  4 s(x2):fac1 CRS   s(x2):fac1 F1    F4     0.0846 #>  5 s(x2):fac1 CRS   s(x2):fac1 F1    F5     0.192  #>  6 s(x2):fac1 CRS   s(x2):fac1 F1    F6     0.152  #>  7 s(x2):fac1 CRS   s(x2):fac1 F1    F7     0.188  #>  8 s(x2):fac1 CRS   s(x2):fac1 F1    F8     0.164  #>  9 s(x2):fac1 CRS   s(x2):fac1 F1    F9     0.0597 #> 10 s(x2):fac1 CRS   s(x2):fac1 F2    F1    -0.755  #> # i 71 more rows"},{"path":"https://gavinsimpson.github.io/gratia/reference/post_draws.html","id":null,"dir":"Reference","previous_headings":"","what":"Low-level Functions to generate draws from the posterior distribution of model coefficients — post_draws","title":"Low-level Functions to generate draws from the posterior distribution of model coefficients — post_draws","text":"Low-level Functions generate draws posterior distribution model coefficients Generate posterior draws fitted model","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/post_draws.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Low-level Functions to generate draws from the posterior distribution of model coefficients — post_draws","text":"","code":"post_draws(model, ...)  # Default S3 method post_draws(   model,   n,   method = c(\"gaussian\", \"mh\", \"inla\", \"user\"),   mu = NULL,   sigma = NULL,   n_cores = 1L,   burnin = 1000,   thin = 1,   t_df = 40,   rw_scale = 0.25,   index = NULL,   frequentist = FALSE,   unconditional = FALSE,   parametrized = TRUE,   mvn_method = c(\"mvnfast\", \"mgcv\"),   draws = NULL,   seed = NULL,   ... )  generate_draws(model, ...)  # S3 method for class 'gam' generate_draws(   model,   n,   method = c(\"gaussian\", \"mh\", \"inla\"),   mu = NULL,   sigma = NULL,   n_cores = 1L,   burnin = 1000,   thin = 1,   t_df = 40,   rw_scale = 0.25,   index = NULL,   frequentist = FALSE,   unconditional = FALSE,   mvn_method = c(\"mvnfast\", \"mgcv\"),   seed = NULL,   ... )"},{"path":"https://gavinsimpson.github.io/gratia/reference/post_draws.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Low-level Functions to generate draws from the posterior distribution of model coefficients — post_draws","text":"model fitted R model. Currently models fitted mgcv::gam() mgcv::bam(), return object inherits objects supported. , \"inherits\" used loose fashion; models fitted scam::scam() support even though models strictly inherit class \"gam\" far inherits() concerned. ... arguments passed methods. n numeric; number posterior draws take. method character; algorithm use sample posterior. Currently implemented methods : \"gaussian\" \"mh\". \"gaussian\" calls gaussian_draws() uses Gaussian approximation posterior distribution. \"mh\" uses simple Metropolis Hasting sampler alternates static proposals based Gaussian approximation posterior, random walk proposals. Note, setting t_df low value result heavier-tailed statistic proposals. See mgcv::gam.mh() details. mu numeric; user-supplied mean vector (vector model coefficients). Currently ignored. sigma matrix; user-supplied covariance matrix mu. Currently ignored. n_cores integer; number CPU cores use generating multivariate normal distributed random values. used mvn_method = \"mvnfast\" method = \"gaussian\". burnin numeric; length initial burn period discard. See mgcv::gam.mh(). thin numeric; retain thin samples. See mgcv::gam.mh(). t_df numeric; degrees freedom static multivariate t proposal. See mgcv::gam.mh(). rw_scale numeric; factor scale posterior covariance matrix generating random walk proposals. See mgcv::gam.mh(). index numeric; vector indices coefficients use. Can used subset mean vector covariance matrix extracted model. frequentist logical; TRUE, frequentist covariance matrix parameter estimates used. FALSE, Bayesian posterior covariance matrix parameters used. See mgcv::vcov.gam(). unconditional logical; TRUE  Bayesian smoothing parameter uncertainty corrected covariance matrix used, available model. See mgcv::vcov.gam(). parametrized logical; use parametrized coefficients covariance matrix, respect linear inequality constraints model. scam::scam() model fits. mvn_method character; one \"mvnfast\" \"mgcv\". default uses mvnfast::rmvn(), can considerably faster generate large numbers MVN random values mgcv::rmvn(), might work marginal fits, covariance matrix close singular. draws matrix; user supplied posterior draws used method = \"user\". seed numeric; random seed use. NULL, random seed generated without affecting current state R's RNG.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/post_link_funs.html","id":null,"dir":"Reference","previous_headings":"","what":"A list of transformation functions named for LSS parameters in a GAMLSS — post_link_funs","title":"A list of transformation functions named for LSS parameters in a GAMLSS — post_link_funs","text":"list transformation functions named LSS parameters GAMLSS","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/post_link_funs.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"A list of transformation functions named for LSS parameters in a GAMLSS — post_link_funs","text":"","code":"post_link_funs(   location = identity_fun,   scale = identity_fun,   shape = identity_fun,   skewness = identity_fun,   kurtosis = identity_fun,   power = identity_fun,   pi = identity_fun )"},{"path":"https://gavinsimpson.github.io/gratia/reference/posterior_samples.html","id":null,"dir":"Reference","previous_headings":"","what":"Draw samples from the posterior distribution of an estimated model — posterior_samples","title":"Draw samples from the posterior distribution of an estimated model — posterior_samples","text":"Draw samples posterior distribution estimated model","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/posterior_samples.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Draw samples from the posterior distribution of an estimated model — posterior_samples","text":"","code":"posterior_samples(model, ...)  # S3 method for class 'gam' posterior_samples(   model,   n = 1,   data = newdata,   seed = NULL,   method = c(\"gaussian\", \"mh\", \"inla\", \"user\"),   n_cores = 1,   burnin = 1000,   thin = 1,   t_df = 40,   rw_scale = 0.25,   freq = FALSE,   unconditional = FALSE,   weights = NULL,   draws = NULL,   mvn_method = c(\"mvnfast\", \"mgcv\"),   ...,   newdata = NULL,   ncores = NULL )"},{"path":"https://gavinsimpson.github.io/gratia/reference/posterior_samples.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Draw samples from the posterior distribution of an estimated model — posterior_samples","text":"model fitted model supported types ... arguments passed methods. fitted_samples(), passed mgcv::predict.gam(). posterior_samples() passed fitted_samples(). predicted_samples() passed relevant simulate() method. n numeric; number posterior samples return. data data frame; new observations posterior draws model evaluated. supplied, data used fit model used data, available model. seed numeric; random seed simulations. method character; method used draw samples posterior distribution. \"gaussian\" uses Gaussian (Laplace) approximation posterior. \"mh\" uses Metropolis Hastings sampler alternates t proposals proposals based shrunken version posterior covariance matrix. \"inla\" uses variant Integrated Nested Laplace Approximation due Wood (2019), (currently implemented). \"user\" allows user-supplied posterior draws (currently implemented). n_cores number cores generating random variables multivariate normal distribution. Passed mvnfast::rmvn(). Parallelization take place OpenMP supported (appears work Windows current R). burnin numeric; number samples discard burnin draws. used method = \"mh\". thin numeric; number samples skip taking n draws. Results thin * n draws posterior taken. used method = \"mh\". t_df numeric; degrees freedom t distribution proposals. used method = \"mh\". rw_scale numeric; Factor scale posterior covariance matrix generating random walk proposals. Negative non finite skip random walk step. used method = \"mh\". freq logical; TRUE use frequentist covariance matrix parameter estimators, FALSE use Bayesian posterior covariance matrix parameters. unconditional logical; TRUE (freq == FALSE) Bayesian smoothing parameter uncertainty corrected covariance matrix used, available. weights numeric; vector prior weights. data null defaults object[[\"prior.weights\"]], otherwise vector ones. draws matrix; user supplied posterior draws used method = \"user\". mvn_method character; one \"mvnfast\" \"mgcv\". default uses mvnfast::rmvn(), can considerably faster generate large numbers MVN random values mgcv::rmvn(), might work marginal fits, covariance matrix close singular. newdata Deprecated: use data instead. ncores Deprecated; use n_cores instead. number cores generating random variables multivariate normal distribution. Passed mvnfast::rmvn(). Parallelization take place OpenMP supported (appears work Windows current R).","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/posterior_samples.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Draw samples from the posterior distribution of an estimated model — posterior_samples","text":"tibble (data frame) 3 columns containing posterior predicted values long format. columns row (integer) row data posterior draw relates , draw (integer) index, range 1:n, indicating draw row relates , response (numeric) predicted response indicated row data.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/posterior_samples.html","id":"note","dir":"Reference","previous_headings":"","what":"Note","title":"Draw samples from the posterior distribution of an estimated model — posterior_samples","text":"Models offset terms supplied via offset argument mgcv::gam() etc. ignored mgcv::predict.gam(). , kind offset term also ignored posterior_samples(). Offset terms included model formula supplied mgcv::gam() etc ignored posterior samples produced reflect offset term values. side effect requiring new data values provided posterior_samples() via data argument must include offset variable.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/posterior_samples.html","id":"references","dir":"Reference","previous_headings":"","what":"References","title":"Draw samples from the posterior distribution of an estimated model — posterior_samples","text":"Wood, S.N., (2020). Simplified integrated nested Laplace approximation. Biometrika 107, 223–230. doi:10.1093/biomet/asz044","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/posterior_samples.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Draw samples from the posterior distribution of an estimated model — posterior_samples","text":"Gavin L. Simpson","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/predicted_samples.html","id":null,"dir":"Reference","previous_headings":"","what":"Draw new response values from the conditional distribution of the response — predicted_samples","title":"Draw new response values from the conditional distribution of the response — predicted_samples","text":"Predicted values response (new response data) drawn fitted model, created via simulate() (e.g. simulate.gam()) returned tidy, long, format. predicted values include uncertainty estimated model; simply draws conditional distribution response.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/predicted_samples.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Draw new response values from the conditional distribution of the response — predicted_samples","text":"","code":"predicted_samples(model, ...)  # S3 method for class 'gam' predicted_samples(   model,   n = 1,   data = newdata,   seed = NULL,   weights = NULL,   ...,   newdata = NULL )"},{"path":"https://gavinsimpson.github.io/gratia/reference/predicted_samples.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Draw new response values from the conditional distribution of the response — predicted_samples","text":"model fitted model supported types ... arguments passed methods. fitted_samples(), passed mgcv::predict.gam(). posterior_samples() passed fitted_samples(). predicted_samples() passed relevant simulate() method. n numeric; number posterior samples return. data data frame; new observations posterior draws model evaluated. supplied, data used fit model used data, available model. seed numeric; random seed simulations. weights numeric; vector prior weights. data null defaults object[[\"prior.weights\"]], otherwise vector ones. newdata Deprecated: use data instead.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/predicted_samples.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Draw new response values from the conditional distribution of the response — predicted_samples","text":"tibble (data frame) 3 columns containing posterior predicted values long format. columns row (integer) row data posterior draw relates , draw (integer) index, range 1:n, indicating draw row relates , response (numeric) predicted response indicated row data.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/predicted_samples.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Draw new response values from the conditional distribution of the response — predicted_samples","text":"Gavin L. Simpson","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/predicted_samples.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Draw new response values from the conditional distribution of the response — predicted_samples","text":"","code":"load_mgcv() dat <- data_sim(\"eg1\", n = 1000, dist = \"normal\", scale = 2, seed = 2) m <- gam(y ~ s(x0) + s(x1) + s(x2) + s(x3), data = dat, method = \"REML\")  predicted_samples(m, n = 5, seed = 42) #> # A tibble: 5,000 x 3 #>     .row .draw .response #>    <int> <int>     <dbl> #>  1     1     1      8.93 #>  2     2     1      4.23 #>  3     3     1      7.71 #>  4     4     1      8.51 #>  5     5     1     10.1  #>  6     6     1      8.20 #>  7     7     1      8.95 #>  8     8     1      7.20 #>  9     9     1     18.1  #> 10    10     1     12.7  #> # i 4,990 more rows  ## Can pass arguments to predict.gam() newd <- data.frame(   x0 = runif(10), x1 = runif(10), x2 = runif(10),   x3 = runif(10) )  ## Exclude s(x2) predicted_samples(m, n = 5, newd, exclude = \"s(x2)\", seed = 25) #> # A tibble: 50 x 3 #>     .row .draw .response #>    <int> <int>     <dbl> #>  1     1     1      9.42 #>  2     2     1      6.97 #>  3     3     1      8.10 #>  4     4     1      9.95 #>  5     5     1      6.75 #>  6     6     1     10.3  #>  7     7     1     10.8  #>  8     8     1     10.5  #>  9     9     1      8.43 #> 10    10     1     12.2  #> # i 40 more rows  ## Exclude s(x1) predicted_samples(m, n = 5, newd, exclude = \"s(x1)\", seed = 25) #> # A tibble: 50 x 3 #>     .row .draw .response #>    <int> <int>     <dbl> #>  1     1     1      6.05 #>  2     2     1      5.28 #>  3     3     1      5.96 #>  4     4     1     13.7  #>  5     5     1      4.36 #>  6     6     1      5.11 #>  7     7     1     12.5  #>  8     8     1      5.66 #>  9     9     1     12.6  #> 10    10     1      8.38 #> # i 40 more rows  ## Select which terms --- result should be the same as previous ## but note that we have to include any parametric terms, including the ## constant term predicted_samples(m,   n = 5, newd, seed = 25,   terms = c(\"Intercept\", \"s(x0)\", \"s(x2)\", \"s(x3)\") ) #> # A tibble: 50 x 3 #>     .row .draw .response #>    <int> <int>     <dbl> #>  1     1     1    -1.94  #>  2     2     1    -2.71  #>  3     3     1    -2.03  #>  4     4     1     5.73  #>  5     5     1    -3.63  #>  6     6     1    -2.87  #>  7     7     1     4.48  #>  8     8     1    -2.33  #>  9     9     1     4.65  #> 10    10     1     0.395 #> # i 40 more rows"},{"path":"https://gavinsimpson.github.io/gratia/reference/qq_plot.html","id":null,"dir":"Reference","previous_headings":"","what":"Quantile-quantile plot of model residuals — qq_plot","title":"Quantile-quantile plot of model residuals — qq_plot","text":"Quantile-quantile plots (QQ-plots) GAMs using reference quantiles Augustin et al (2012).","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/qq_plot.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Quantile-quantile plot of model residuals — qq_plot","text":"","code":"qq_plot(model, ...)  # Default S3 method qq_plot(model, ...)  # S3 method for class 'gam' qq_plot(   model,   method = c(\"uniform\", \"simulate\", \"normal\", \"direct\"),   type = c(\"deviance\", \"response\", \"pearson\"),   n_uniform = 10,   n_simulate = 50,   seed = NULL,   level = 0.9,   ylab = NULL,   xlab = NULL,   title = NULL,   subtitle = NULL,   caption = NULL,   ci_col = \"black\",   ci_alpha = 0.2,   point_col = \"black\",   point_alpha = 1,   line_col = \"red\",   ... )  # S3 method for class 'glm' qq_plot(model, ...)  # S3 method for class 'lm' qq_plot(model, ...)"},{"path":"https://gavinsimpson.github.io/gratia/reference/qq_plot.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Quantile-quantile plot of model residuals — qq_plot","text":"model fitted model. Currently models inheriting class \"gam\", well classes \"glm\" \"lm\" calls stats::glm stats::lm supported. ... arguments passed ot methods. method character; method used generate theoretical quantiles. default \"uniform\", generates reference quantiles using random draws uniform distribution inverse cummulative distribution function (CDF) fitted values. reference quantiles averaged n_uniform draws. \"simulate\" generates reference quantiles simulating new response data model observed values covariates, residualised generate reference quantiles, using n_simulate simulated data sets. \"normal\" generates reference quantiles using standard normal distribution. \"uniform\" computationally efficient, \"simulate\" allows reference bands drawn QQ-plot. \"normal\" avoided used fall back random number generator (\"simulate\") inverse CDF available family used model fitting (`\"uniform\"“). Note method = \"direct\" deprecated favour method = \"uniform\". type character; type residuals use. \"deviance\", \"response\", \"pearson\" residuals allowed. n_uniform numeric; number times randomize uniform quantiles direct computation method (method = \"uniform\"). n_simulate numeric; number data sets simulate estimated model using simulation method (method = \"simulate\"). seed numeric; random number seed use method = \"simulate\" method = \"uniform\". level numeric; coverage level reference intervals. Must strictly 0 < level < 1. used method = \"simulate\". ylab character expression; label y axis. supplied, suitable label generated. xlab character expression; label y axis. supplied, suitable label generated. title character expression; title plot. See ggplot2::labs(). May vector, one per penalty. subtitle character expression; subtitle plot. See ggplot2::labs(). May vector, one per penalty. caption character expression; plot caption. See ggplot2::labs(). May vector, one per penalty. ci_col fill colour reference interval method = \"simulate\". ci_alpha alpha transparency reference interval method = \"simulate\". point_col colour points QQ plot. point_alpha alpha transparency points QQ plot. line_col colour used draw reference line.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/qq_plot.html","id":"note","dir":"Reference","previous_headings":"","what":"Note","title":"Quantile-quantile plot of model residuals — qq_plot","text":"wording used mgcv::qq.gam() uses direct reference simulated residuals method (method = \"simulated\"). avoid confusion, method = \"direct\" deprecated favour method = \"uniform\".","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/qq_plot.html","id":"references","dir":"Reference","previous_headings":"","what":"References","title":"Quantile-quantile plot of model residuals — qq_plot","text":"underlying methodology used method \"simulate\" \"uniform\" described Augustin et al (2012): Augustin, N.H., Sauleau, E.-., Wood, S.N., (2012) quantile quantile plots generalized linear models. Computational Statatistics Data Analysis 56, 2404-2409 doi:10.1016/j.csda.2012.01.026 .","code":""},{"path":[]},{"path":"https://gavinsimpson.github.io/gratia/reference/qq_plot.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Quantile-quantile plot of model residuals — qq_plot","text":"","code":"load_mgcv() ## simulate binomial data... dat <- data_sim(\"eg1\", n = 200, dist = \"binary\", scale = .33, seed = 0) p <- binomial()$linkinv(dat$f) # binomial p n <- sample(c(1, 3), 200, replace = TRUE) # binomial n dat <- transform(dat, y = rbinom(n, n, p), n = n) m <- gam(y / n ~ s(x0) + s(x1) + s(x2) + s(x3),   family = binomial, data = dat, weights = n,   method = \"REML\" )  ## Q-Q plot; default using direct randomization of uniform quantiles qq_plot(m)   ## Alternatively use simulate new data from the model, which ## allows construction of reference intervals for the Q-Q plot qq_plot(m,   method = \"simulate\",   seed = 42,   point_col = \"steelblue\",   point_alpha = 0.4 )   ## ... or use the usual normality assumption qq_plot(m, method = \"normal\")"},{"path":"https://gavinsimpson.github.io/gratia/reference/ref_level.html","id":null,"dir":"Reference","previous_headings":"","what":"Return the reference or specific level of a factor — ref_level","title":"Return the reference or specific level of a factor — ref_level","text":"Extracts reference specific level supplied factor, returning factor levels one supplied.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/ref_level.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Return the reference or specific level of a factor — ref_level","text":"","code":"ref_level(fct)  level(fct, level)"},{"path":"https://gavinsimpson.github.io/gratia/reference/ref_level.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Return the reference or specific level of a factor — ref_level","text":"fct factor; factor reference specific level extracted. level character; specific level extract case level().","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/ref_level.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Return the reference or specific level of a factor — ref_level","text":"length 1 factor levels supplied factor fct.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/ref_level.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Return the reference or specific level of a factor — ref_level","text":"","code":"f <- factor(sample(letters[1:5], 100, replace = TRUE))  # the reference level ref_level(f) #> [1] a #> Levels: a b c d e  # a specific level level(f, level = \"b\") #> [1] b #> Levels: a b c d e  # note that the levels will always match the input factor identical(levels(f), levels(ref_level(f))) #> [1] TRUE identical(levels(f), levels(level(f, \"c\"))) #> [1] TRUE"},{"path":"https://gavinsimpson.github.io/gratia/reference/ref_sims.html","id":null,"dir":"Reference","previous_headings":"","what":"Reference simulation data — ref_sims","title":"Reference simulation data — ref_sims","text":"set reference objects testing data_sim().","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/ref_sims.html","id":"format","dir":"Reference","previous_headings":"","what":"Format","title":"Reference simulation data — ref_sims","text":"named list simulated data sets created data_sim().","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/reorder_fs_smooth_terms.html","id":null,"dir":"Reference","previous_headings":"","what":"Reorder random factor smooth terms to place factor last — reorder_fs_smooth_terms","title":"Reorder random factor smooth terms to place factor last — reorder_fs_smooth_terms","text":"Reorder random factor smooth terms place factor last","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/reorder_fs_smooth_terms.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Reorder random factor smooth terms to place factor last — reorder_fs_smooth_terms","text":"","code":"reorder_fs_smooth_terms(smooth)"},{"path":"https://gavinsimpson.github.io/gratia/reference/reorder_fs_smooth_terms.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Reorder random factor smooth terms to place factor last — reorder_fs_smooth_terms","text":"smooth mgcv smooth object","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/reorder_tensor_smooth_terms.html","id":null,"dir":"Reference","previous_headings":"","what":"Reorder tensor product terms for nicer plotting — reorder_tensor_smooth_terms","title":"Reorder tensor product terms for nicer plotting — reorder_tensor_smooth_terms","text":"tensor product smooth 3 terms contains 2d marginal smooth, get nicer output smooth_estimates() hence nicer plot draw.smooth_estimates() method reorder terms smooth vary terms 2d marginal first, terms vary slowly generate data evaluate smooth . results automatically generated data focuses (first one) 2d marginal smooth, end result smooth_estimates() shows 2d smooth changes terms involved smooth.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/reorder_tensor_smooth_terms.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Reorder tensor product terms for nicer plotting — reorder_tensor_smooth_terms","text":"","code":"reorder_tensor_smooth_terms(smooth)"},{"path":"https://gavinsimpson.github.io/gratia/reference/reorder_tensor_smooth_terms.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Reorder tensor product terms for nicer plotting — reorder_tensor_smooth_terms","text":"smooth mgcv smooth object","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/rep_first_factor_value.html","id":null,"dir":"Reference","previous_headings":"","what":"Repeat the first level of a factor n times — rep_first_factor_value","title":"Repeat the first level of a factor n times — rep_first_factor_value","text":"Function repeat first level factor n times return vector factor original levels intact","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/rep_first_factor_value.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Repeat the first level of a factor n times — rep_first_factor_value","text":"","code":"rep_first_factor_value(f, n)"},{"path":"https://gavinsimpson.github.io/gratia/reference/rep_first_factor_value.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Repeat the first level of a factor n times — rep_first_factor_value","text":"f factor n numeric; number times repeat first level f","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/rep_first_factor_value.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Repeat the first level of a factor n times — rep_first_factor_value","text":"factor length n levels f, whose elements first level f.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/residuals_hist_plot.html","id":null,"dir":"Reference","previous_headings":"","what":"Histogram of model residuals — residuals_hist_plot","title":"Histogram of model residuals — residuals_hist_plot","text":"Histogram model residuals","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/residuals_hist_plot.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Histogram of model residuals — residuals_hist_plot","text":"","code":"residuals_hist_plot(   model,   type = c(\"deviance\", \"pearson\", \"response\"),   n_bins = c(\"sturges\", \"scott\", \"fd\"),   ylab = NULL,   xlab = NULL,   title = NULL,   subtitle = NULL,   caption = NULL )"},{"path":"https://gavinsimpson.github.io/gratia/reference/residuals_hist_plot.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Histogram of model residuals — residuals_hist_plot","text":"model fitted model. Currently class \"gam\". type character; type residuals use. \"deviance\", \"response\", \"pearson\" residuals allowed. n_bins character numeric; either number bins string indicating calculate number bins. ylab character expression; label y axis. supplied, suitable label generated. xlab character expression; label y axis. supplied, suitable label generated. title character expression; title plot. See ggplot2::labs(). subtitle character expression; subtitle plot. See ggplot2::labs(). caption character expression; plot caption. See ggplot2::labs().","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/residuals_linpred_plot.html","id":null,"dir":"Reference","previous_headings":"","what":"Plot of residuals versus linear predictor values — residuals_linpred_plot","title":"Plot of residuals versus linear predictor values — residuals_linpred_plot","text":"Plot residuals versus linear predictor values","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/residuals_linpred_plot.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Plot of residuals versus linear predictor values — residuals_linpred_plot","text":"","code":"residuals_linpred_plot(   model,   type = c(\"deviance\", \"pearson\", \"response\"),   ylab = NULL,   xlab = NULL,   title = NULL,   subtitle = NULL,   caption = NULL,   point_col = \"black\",   point_alpha = 1,   line_col = \"red\" )"},{"path":"https://gavinsimpson.github.io/gratia/reference/residuals_linpred_plot.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Plot of residuals versus linear predictor values — residuals_linpred_plot","text":"model fitted model. Currently class \"gam\". type character; type residuals use. \"deviance\", \"response\", \"pearson\" residuals allowed. ylab character expression; label y axis. supplied, suitable label generated. xlab character expression; label y axis. supplied, suitable label generated. title character expression; title plot. See ggplot2::labs(). subtitle character expression; subtitle plot. See ggplot2::labs(). caption character expression; plot caption. See ggplot2::labs(). point_col colour used draw points plots. See graphics::par() section Color Specification. passed individual plotting functions, therefore affects points plots. point_alpha numeric; alpha transparency points plots. line_col colour specification 1:1 line.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/response_derivatives.html","id":null,"dir":"Reference","previous_headings":"","what":"Derivatives on the response scale from an estimated GAM — response_derivatives","title":"Derivatives on the response scale from an estimated GAM — response_derivatives","text":"Derivatives response scale estimated GAM","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/response_derivatives.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Derivatives on the response scale from an estimated GAM — response_derivatives","text":"","code":"response_derivatives(object, ...)  # Default S3 method response_derivatives(object, ...)  # S3 method for class 'gamm' response_derivatives(object, ...)  # S3 method for class 'gam' response_derivatives(   object,   focal = NULL,   data = NULL,   order = 1L,   type = c(\"forward\", \"backward\", \"central\"),   scale = c(\"response\", \"linear_predictor\"),   method = c(\"gaussian\", \"mh\", \"inla\", \"user\"),   n = 100,   eps = 1e-07,   n_sim = 10000,   level = 0.95,   seed = NULL,   mvn_method = c(\"mvnfast\", \"mgcv\"),   ... )"},{"path":"https://gavinsimpson.github.io/gratia/reference/response_derivatives.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Derivatives on the response scale from an estimated GAM — response_derivatives","text":"object R object compute derivatives . ... arguments passed methods fitted_samples() focal character; name focal variable. response derivative response respect variable returned. variables involved model held constant values. can missing supplying data, case, focal variable identified one variable constant. data data frame containing values model covariates evaluate first derivatives smooths. supplied, one variable must held constant value. order numeric; order derivative. type character; type finite difference used. One \"forward\", \"backward\", \"central\". scale character; derivative estimated response linear predictor (link) scale? One \"response\" (default), \"linear predictor\". method character; method used draw samples posterior distribution. \"gaussian\" uses Gaussian (Laplace) approximation posterior. \"mh\" uses Metropolis Hastings sample alternates t proposals proposals based shrunken version posterior covariance matrix. \"inla\" uses variant Integrated Nested Laplace Approximation due Wood (2019), (currently implemented). \"user\" allows user-supplied posterior draws (currently implemented). n numeric; number points evaluate derivative (data supplied). eps numeric; finite difference. n_sim integer; number simulations used computing simultaneous intervals. level numeric; 0 < level < 1; coverage level credible interval. default 0.95 95% interval. seed numeric; random seed simulations. mvn_method character; one \"mvnfast\" \"mgcv\". default uses mvnfast::rmvn(), can considerably faster generate large numbers MVN random values mgcv::rmvn(), might work marginal fits, covariance matrix close singular.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/response_derivatives.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Derivatives on the response scale from an estimated GAM — response_derivatives","text":"tibble, currently following variables: .row: integer, indexing row data row output represents .focal: name variable partial derivative evaluated, .derivative: estimated partial derivative, .lower_ci: lower bound confidence simultaneous interval, .upper_ci: upper bound confidence simultaneous interval, additional columns containing covariate values derivative evaluated.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/response_derivatives.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Derivatives on the response scale from an estimated GAM — response_derivatives","text":"Gavin L. Simpson","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/response_derivatives.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Derivatives on the response scale from an estimated GAM — response_derivatives","text":"","code":"library(\"ggplot2\") library(\"patchwork\") load_mgcv() df <- data_sim(\"eg1\", dist = \"negbin\", scale = 0.25, seed = 42)  # fit the GAM (note: for execution time reasons using bam()) m <- bam(y ~ s(x0) + s(x1) + s(x2) + s(x3),   data = df, family = nb(), method = \"fREML\" )  # data slice through data along x2 - all other covariates will be set to # typical values (value closest to median) ds <- data_slice(m, x2 = evenly(x2, n = 100))  # fitted values along x2 fv <- fitted_values(m, data = ds)  # response derivatives - ideally n_sim = >10000 y_d <- response_derivatives(m,   data = ds, type = \"central\", focal = \"x2\",   eps = 0.01, seed = 21, n_sim = 1000 )  # draw fitted values along x2 p1 <- fv |>   ggplot(aes(x = x2, y = .fitted)) +   geom_ribbon(aes(ymin = .lower_ci, ymax = .upper_ci, y = NULL),     alpha = 0.2   ) +   geom_line() +   labs(     title = \"Estimated count as a function of x2\",     y = \"Estimated count\"   )  # draw response derivatives p2 <- y_d |>   ggplot(aes(x = x2, y = .derivative)) +   geom_ribbon(aes(ymin = .lower_ci, ymax = .upper_ci), alpha = 0.2) +   geom_line() +   labs(     title = \"Estimated 1st derivative of estimated count\",     y = \"First derivative\"   )  # draw both panels p1 + p2 + plot_layout(nrow = 2)"},{"path":"https://gavinsimpson.github.io/gratia/reference/rootogram.html","id":null,"dir":"Reference","previous_headings":"","what":"Rootograms to assess goodness of model fit — rootogram","title":"Rootograms to assess goodness of model fit — rootogram","text":"rootogram model diagnostic tool assesses goodness fit statistical model. observed values response compared expected fitted model. discrete, count responses, frequency count (0, 1, 2, etc) observed data expected conditional distribution response implied model compared. continuous variables, observed expected frequencies obtained grouping data bins. rootogram drawn using ggplot2::ggplot() graphics. design closely follows Kleiber & Zeileis (2016).","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/rootogram.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Rootograms to assess goodness of model fit — rootogram","text":"","code":"rootogram(object, ...)  # S3 method for class 'gam' rootogram(object, max_count = NULL, breaks = \"Sturges\", ...)"},{"path":"https://gavinsimpson.github.io/gratia/reference/rootogram.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Rootograms to assess goodness of model fit — rootogram","text":"object R object ... arguments passed methods max_count integer; largest count consider breaks continuous responses, group response. Can anything acceptable breaks argument graphics::hist.default()","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/rootogram.html","id":"references","dir":"Reference","previous_headings":"","what":"References","title":"Rootograms to assess goodness of model fit — rootogram","text":"Kleiber, C., Zeileis, ., (2016) Visualizing Count Data Regressions Using Rootograms. . Stat. 70, 296–303. doi:10.1080/00031305.2016.1173590","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/rootogram.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Rootograms to assess goodness of model fit — rootogram","text":"","code":"load_mgcv() df <- data_sim(\"eg1\", n = 1000, dist = \"poisson\", scale = 0.1, seed = 6)  # A poisson example m <- gam(y ~ s(x0, bs = \"cr\") + s(x1, bs = \"cr\") + s(x2, bs = \"cr\") +   s(x3, bs = \"cr\"), family = poisson(), data = df, method = \"REML\") rg <- rootogram(m) rg #> # A tibble: 21 x 3 #>     .bin .observed   .fitted #>    <dbl>     <int>     <dbl> #>  1     0       113 116.640   #>  2     1       236 227.869   #>  3     2       230 239.168   #>  4     3       200 181.679   #>  5     4        94 113.432   #>  6     5        68  62.4881  #>  7     6        27  31.6795  #>  8     7        22  15.1323  #>  9     8         4   6.88637 #> 10     9         3   2.99628 #> # i 11 more rows draw(rg) # plot the rootogram   # A Gaussian example df <- data_sim(\"eg1\", dist = \"normal\", seed = 2) m <- gam(y ~ s(x0) + s(x1) + s(x2) + s(x3), data = df, method = \"REML\") draw(rootogram(m, breaks = \"FD\"), type = \"suspended\")"},{"path":"https://gavinsimpson.github.io/gratia/reference/seq_min_max_eps.html","id":null,"dir":"Reference","previous_headings":"","what":"Create a sequence of evenly-spaced values adjusted to accommodate a small adjustment — seq_min_max_eps","title":"Create a sequence of evenly-spaced values adjusted to accommodate a small adjustment — seq_min_max_eps","text":"Creates sequence n evenly-spaced values range min(x) – max(x), minimum maximum adjusted always contained within range x x may shifted forwards backwards amount related eps. particularly useful computing derivatives via finite differences without adjustment may predicting values outside range data hence conmstraints penalty.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/seq_min_max_eps.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Create a sequence of evenly-spaced values adjusted to accommodate a small adjustment — seq_min_max_eps","text":"","code":"seq_min_max_eps(x, n, order, type = c(\"forward\", \"backward\", \"central\"), eps)"},{"path":"https://gavinsimpson.github.io/gratia/reference/seq_min_max_eps.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Create a sequence of evenly-spaced values adjusted to accommodate a small adjustment — seq_min_max_eps","text":"x numeric; vector evenly-spaced values returned n numeric; number evenly-spaced values return order integer; order derivative. Either 1 2 first second order derivatives type character; type finite difference used. One \"forward\", \"backward\", \"central\" eps numeric; finite difference","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/seq_min_max_eps.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Create a sequence of evenly-spaced values adjusted to accommodate a small adjustment — seq_min_max_eps","text":"numeric vector length n.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/shift_values.html","id":null,"dir":"Reference","previous_headings":"","what":"Shift numeric values in a data frame by an amount eps — shift_values","title":"Shift numeric values in a data frame by an amount eps — shift_values","text":"Shift numeric values data frame amount eps","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/shift_values.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Shift numeric values in a data frame by an amount eps — shift_values","text":"","code":"shift_values(df, h, i, FUN = `+`, focal = NULL)"},{"path":"https://gavinsimpson.github.io/gratia/reference/shift_values.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Shift numeric values in a data frame by an amount eps — shift_values","text":"df data frame tibble. h numeric; amount shift values df . logical; vector indexing columns df included shift. FUN function; function applut shift. Typically + -. focal character; focal variable computing partial derivatives. allows shifting focal variable eps.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/simulate.html","id":null,"dir":"Reference","previous_headings":"","what":"Simulate from the posterior distribution of a GAM — simulate.gam","title":"Simulate from the posterior distribution of a GAM — simulate.gam","text":"Simulations posterior distribution fitted GAM model involve computing predicted values observation data simulated data required, generating random draws probability distribution used fitting model.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/simulate.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Simulate from the posterior distribution of a GAM — simulate.gam","text":"","code":"# S3 method for class 'gam' simulate(   object,   nsim = 1,   seed = NULL,   data = newdata,   weights = NULL,   ...,   newdata = NULL )  # S3 method for class 'gamm' simulate(   object,   nsim = 1,   seed = NULL,   data = newdata,   weights = NULL,   ...,   newdata = NULL )  # S3 method for class 'scam' simulate(   object,   nsim = 1,   seed = NULL,   data = newdata,   weights = NULL,   ...,   newdata = NULL )"},{"path":"https://gavinsimpson.github.io/gratia/reference/simulate.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Simulate from the posterior distribution of a GAM — simulate.gam","text":"object fitted GAM, typically result call mgcv::gam` mgcv::gamm(). nsim numeric; number posterior simulations return. seed numeric; random seed simulations. data data frame; new observations posterior draws model evaluated. supplied, data used fit model used newdata, available object. weights numeric; vector prior weights. newdata null defaults object[[\"prior.weights\"]], otherwise vector ones. ... arguments passed methods. simulate.gam() simulate.scam() pass ... predict.gam(). can pass additional arguments terms, exclude, select model terms included predictions. may useful, example, excluding effects random effect terms. newdata Deprecated. Use data instead.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/simulate.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Simulate from the posterior distribution of a GAM — simulate.gam","text":"(Currently) matrix nsim columns.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/simulate.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Simulate from the posterior distribution of a GAM — simulate.gam","text":"simulate.gam() function, family component fitted model must contain, updateable contain, required random number generator. See mgcv::fix.family.rd().","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/simulate.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Simulate from the posterior distribution of a GAM — simulate.gam","text":"Gavin L. Simpson","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/simulate.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Simulate from the posterior distribution of a GAM — simulate.gam","text":"","code":"load_mgcv() dat <- data_sim(\"eg1\", n = 400, dist = \"normal\", scale = 2, seed = 2) m1 <- gam(y ~ s(x0) + s(x1) + s(x2) + s(x3), data = dat, method = \"REML\")  sims <- simulate(m1, nsim = 5, seed = 42) head(sims) #>           [,1]      [,2]      [,3]      [,4]      [,5] #> [1,] 11.445470 11.374304 10.098681  7.264881  8.796630 #> [2,]  6.510912  5.909584  9.057362  7.698084 11.444781 #> [3,]  3.837995  3.230610  3.550240  3.759380  4.774581 #> [4,] 12.361830 11.209226 10.714215 11.861957 10.746417 #> [5,] 14.851461 12.911440 11.356984 15.783913 15.106270 #> [6,]  5.921276  4.158963  5.520856  7.973614  9.654888"},{"path":"https://gavinsimpson.github.io/gratia/reference/smallAges.html","id":null,"dir":"Reference","previous_headings":"","what":"Lead-210 age-depth measurements for Small Water — smallAges","title":"Lead-210 age-depth measurements for Small Water — smallAges","text":"dataset containing lead-210 based age depth measurements SMALL1 core Small Water.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/smallAges.html","id":"format","dir":"Reference","previous_headings":"","what":"Format","title":"Lead-210 age-depth measurements for Small Water — smallAges","text":"data frame 12 rows 7 variables.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/smallAges.html","id":"source","dir":"Reference","previous_headings":"","what":"Source","title":"Lead-210 age-depth measurements for Small Water — smallAges","text":"Simpson, G.L. (Unpublished data).","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/smallAges.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Lead-210 age-depth measurements for Small Water — smallAges","text":"variables follows: Depth Drymass Date Age Error SedAccRate SedPerCentChange","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/smooth_coef_indices.html","id":null,"dir":"Reference","previous_headings":"","what":"Indices of the parametric terms for a particular smooth — smooth_coef_indices","title":"Indices of the parametric terms for a particular smooth — smooth_coef_indices","text":"Returns vector indices parametric terms represent supplied smooth. Useful extracting model coefficients columns covariance matrix.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/smooth_coef_indices.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Indices of the parametric terms for a particular smooth — smooth_coef_indices","text":"","code":"smooth_coef_indices(smooth)"},{"path":"https://gavinsimpson.github.io/gratia/reference/smooth_coef_indices.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Indices of the parametric terms for a particular smooth — smooth_coef_indices","text":"smooth object inherits class mgcv.smooth","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/smooth_coef_indices.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Indices of the parametric terms for a particular smooth — smooth_coef_indices","text":"numeric vector indices.","code":""},{"path":[]},{"path":"https://gavinsimpson.github.io/gratia/reference/smooth_coef_indices.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Indices of the parametric terms for a particular smooth — smooth_coef_indices","text":"Gavin L. Simpson","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/smooth_coefs.html","id":null,"dir":"Reference","previous_headings":"","what":"Coefficients for a particular smooth — smooth_coefs","title":"Coefficients for a particular smooth — smooth_coefs","text":"Returns vector model coefficients parametric terms represent supplied smooth.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/smooth_coefs.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Coefficients for a particular smooth — smooth_coefs","text":"","code":"smooth_coefs(object, ...)  # S3 method for class 'gam' smooth_coefs(object, select, term = deprecated(), ...)  # S3 method for class 'bam' smooth_coefs(object, select, term = deprecated(), ...)  # S3 method for class 'gamm' smooth_coefs(object, select, term = deprecated(), ...)  # S3 method for class 'gamm4' smooth_coefs(object, select, term = deprecated(), ...)  # S3 method for class 'list' smooth_coefs(object, select, term = deprecated(), ...)  # S3 method for class 'mgcv.smooth' smooth_coefs(object, model, ...)  # S3 method for class 'scam' smooth_coefs(object, select, term = deprecated(), ...)"},{"path":"https://gavinsimpson.github.io/gratia/reference/smooth_coefs.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Coefficients for a particular smooth — smooth_coefs","text":"object fitted GAM(M) object, , \"mgcv.smooth\" method, object inherits class mgcv.smooth. ... arguments passed methods. select character; label smooth whose coefficients returned. term Use select instead. model fitted GAM(M) object.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/smooth_coefs.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Coefficients for a particular smooth — smooth_coefs","text":"numeric vector model coefficients.","code":""},{"path":[]},{"path":"https://gavinsimpson.github.io/gratia/reference/smooth_coefs.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Coefficients for a particular smooth — smooth_coefs","text":"Gavin L. Simpson","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/smooth_coefs.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Coefficients for a particular smooth — smooth_coefs","text":"","code":"load_mgcv() df <- data_sim(\"eg1\", seed = 2) m <- gam(y ~ s(x0) + s(x1) + s(x2) + s(x3), data = df, method = \"REML\")  ## IGNORE_RDIFF_BEGIN smooth_coefs(m, select = \"s(x2)\") #>   s(x2).1   s(x2).2   s(x2).3   s(x2).4   s(x2).5   s(x2).6   s(x2).7   s(x2).8  #> -6.533373  9.694277  2.194078 -1.967280 -2.374874  1.207638 -1.572586  9.269744  #>   s(x2).9  #>  5.622738  ## IGNORE_RDIFF_END"},{"path":"https://gavinsimpson.github.io/gratia/reference/smooth_data.html","id":null,"dir":"Reference","previous_headings":"","what":"Generate regular data over the covariates of a smooth — smooth_data","title":"Generate regular data over the covariates of a smooth — smooth_data","text":"Generate regular data covariates smooth","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/smooth_data.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Generate regular data over the covariates of a smooth — smooth_data","text":"","code":"smooth_data(   model,   id,   n = 100,   n_2d = NULL,   n_3d = NULL,   n_4d = NULL,   offset = NULL,   include_all = FALSE,   var_order = NULL )"},{"path":"https://gavinsimpson.github.io/gratia/reference/smooth_data.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Generate regular data over the covariates of a smooth — smooth_data","text":"model fitted model id number ID smooth within model process. n numeric; number new observations generate. n_2d numeric; number new observations generate second dimension 2D smooth. Currently ignored. n_3d numeric; number new observations generate third dimension 3D smooth. n_4d numeric; number new observations generate dimensions higher 2 (!) kD smooth (k >= 4). example, smooth 4D smooth, dimensions 3 4 get n_4d new observations. offset numeric; value model offset use. include_all logical; include covariates involved smooth? FALSE, covariates involved smooth included returned data frame. TRUE, representative value included covariates model actually used smooth. can useful want pass returned data frame mgcv::PredictMat(). var_order character; order terms smooth processed. useful tensor products least one 2d marginal smooth.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/smooth_data.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Generate regular data over the covariates of a smooth — smooth_data","text":"","code":"load_mgcv() df <- data_sim(\"eg1\", seed = 42) m <- bam(y ~ s(x0) + s(x1) + s(x2) + s(x3), data = df)  # generate data over range of x1 for smooth s(x1) smooth_data(m, id = 2) #> # A tibble: 100 x 1 #>           x1 #>        <dbl> #>  1 0.0004050 #>  2 0.01046   #>  3 0.02052   #>  4 0.03057   #>  5 0.04063   #>  6 0.05069   #>  7 0.06074   #>  8 0.07080   #>  9 0.08086   #> 10 0.09091   #> # i 90 more rows  # generate data over range of x1 for smooth s(x1), with typical value for # other covariates in the model smooth_data(m, id = 2, include_all = TRUE) #> # A tibble: 100 x 4 #>           x1     x0     x2     x3 #>        <dbl>  <dbl>  <dbl>  <dbl> #>  1 0.0004050 0.4883 0.4708 0.4879 #>  2 0.01046   0.4883 0.4708 0.4879 #>  3 0.02052   0.4883 0.4708 0.4879 #>  4 0.03057   0.4883 0.4708 0.4879 #>  5 0.04063   0.4883 0.4708 0.4879 #>  6 0.05069   0.4883 0.4708 0.4879 #>  7 0.06074   0.4883 0.4708 0.4879 #>  8 0.07080   0.4883 0.4708 0.4879 #>  9 0.08086   0.4883 0.4708 0.4879 #> 10 0.09091   0.4883 0.4708 0.4879 #> # i 90 more rows"},{"path":"https://gavinsimpson.github.io/gratia/reference/smooth_dim.html","id":null,"dir":"Reference","previous_headings":"","what":"Dimension of a smooth — smooth_dim","title":"Dimension of a smooth — smooth_dim","text":"Extracts dimension estimated smooth.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/smooth_dim.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Dimension of a smooth — smooth_dim","text":"","code":"smooth_dim(object)  # S3 method for class 'gam' smooth_dim(object)  # S3 method for class 'gamm' smooth_dim(object)  # S3 method for class 'mgcv.smooth' smooth_dim(object)"},{"path":"https://gavinsimpson.github.io/gratia/reference/smooth_dim.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Dimension of a smooth — smooth_dim","text":"object R object. See Details list supported objects.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/smooth_dim.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Dimension of a smooth — smooth_dim","text":"numeric vector dimensions smooth.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/smooth_dim.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Dimension of a smooth — smooth_dim","text":"generic function methods objects class \"gam\", \"gamm\", \"mgcv.smooth\".","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/smooth_dim.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Dimension of a smooth — smooth_dim","text":"Gavin L. Simpson","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/smooth_estimates.html","id":null,"dir":"Reference","previous_headings":"","what":"Evaluate smooths at covariate values — smooth_estimates","title":"Evaluate smooths at covariate values — smooth_estimates","text":"Evaluate smooth grid evenly spaced value range covariate associated smooth. Alternatively, set points smooth evaluated can supplied. smooth_estimates() new implementation evaluate_smooth(), replaces function, removed package.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/smooth_estimates.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Evaluate smooths at covariate values — smooth_estimates","text":"","code":"smooth_estimates(object, ...)  # S3 method for class 'gam' smooth_estimates(   object,   select = NULL,   smooth = deprecated(),   n = 100,   n_3d = 16,   n_4d = 4,   data = NULL,   unconditional = FALSE,   overall_uncertainty = TRUE,   dist = NULL,   unnest = TRUE,   partial_match = FALSE,   ... )"},{"path":"https://gavinsimpson.github.io/gratia/reference/smooth_estimates.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Evaluate smooths at covariate values — smooth_estimates","text":"object object class \"gam\" \"gamm\". ... arguments passed methods. select character; select smooth's posterior draw . default (NULL) means posteriors smooths model wil sampled . supplied, character vector requested terms. smooth Use select instead. n numeric; number points range covariate evaluate smooth. n_3d, n_4d numeric; number points range last covariate 3D 4D smooth. default NULL achieves standard behaviour using n points range covariate, resulting n^d evaluation points, d dimension smooth. d > 2 can result many evaluation points slow performance. smooths d > 4, value n_4d used dimensions > 4, unless NULL, case default behaviour (using n dimensions) observed. data data frame covariate values evaluate smooth. unconditional logical; confidence intervals include uncertainty due smoothness selection? TRUE, corrected Bayesian covariance matrix used. overall_uncertainty logical; uncertainty model constant term included standard error evaluate values smooth? dist numeric; greater 0, used determine location far data plotted plotting 2-D smooths. data scaled unit square deciding exclude, dist distance within unit square. See mgcv::exclude..far() details. unnest logical; unnest smooth objects? partial_match logical; case character select, select match partially smooths? partial_match = TRUE, select must single string, character vector length 1.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/smooth_estimates.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Evaluate smooths at covariate values — smooth_estimates","text":"data frame (tibble), class \"smooth_estimates\".","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/smooth_estimates.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Evaluate smooths at covariate values — smooth_estimates","text":"","code":"load_mgcv() dat <- data_sim(\"eg1\", n = 400, dist = \"normal\", scale = 2, seed = 2) m1 <- gam(y ~ s(x0) + s(x1) + s(x2) + s(x3), data = dat, method = \"REML\")  ## evaluate all smooths smooth_estimates(m1) #> # A tibble: 400 x 9 #>    .smooth .type .by   .estimate      .se         x0    x1    x2    x3 #>    <chr>   <chr> <chr>     <dbl>    <dbl>      <dbl> <dbl> <dbl> <dbl> #>  1 s(x0)   TPRS  NA    -0.966542 0.316118 0.00710904    NA    NA    NA #>  2 s(x0)   TPRS  NA    -0.925391 0.297170 0.0171157     NA    NA    NA #>  3 s(x0)   TPRS  NA    -0.884233 0.279256 0.0271224     NA    NA    NA #>  4 s(x0)   TPRS  NA    -0.843050 0.262594 0.0371291     NA    NA    NA #>  5 s(x0)   TPRS  NA    -0.801824 0.247376 0.0471358     NA    NA    NA #>  6 s(x0)   TPRS  NA    -0.760536 0.233728 0.0571425     NA    NA    NA #>  7 s(x0)   TPRS  NA    -0.719175 0.221701 0.0671492     NA    NA    NA #>  8 s(x0)   TPRS  NA    -0.677736 0.211261 0.0771559     NA    NA    NA #>  9 s(x0)   TPRS  NA    -0.636220 0.202303 0.0871626     NA    NA    NA #> 10 s(x0)   TPRS  NA    -0.594641 0.194685 0.0971693     NA    NA    NA #> # i 390 more rows  ## or selected smooths smooth_estimates(m1, select = c(\"s(x0)\", \"s(x1)\")) #> # A tibble: 200 x 7 #>    .smooth .type .by   .estimate      .se         x0    x1 #>    <chr>   <chr> <chr>     <dbl>    <dbl>      <dbl> <dbl> #>  1 s(x0)   TPRS  NA    -0.966542 0.316118 0.00710904    NA #>  2 s(x0)   TPRS  NA    -0.925391 0.297170 0.0171157     NA #>  3 s(x0)   TPRS  NA    -0.884233 0.279256 0.0271224     NA #>  4 s(x0)   TPRS  NA    -0.843050 0.262594 0.0371291     NA #>  5 s(x0)   TPRS  NA    -0.801824 0.247376 0.0471358     NA #>  6 s(x0)   TPRS  NA    -0.760536 0.233728 0.0571425     NA #>  7 s(x0)   TPRS  NA    -0.719175 0.221701 0.0671492     NA #>  8 s(x0)   TPRS  NA    -0.677736 0.211261 0.0771559     NA #>  9 s(x0)   TPRS  NA    -0.636220 0.202303 0.0871626     NA #> 10 s(x0)   TPRS  NA    -0.594641 0.194685 0.0971693     NA #> # i 190 more rows"},{"path":"https://gavinsimpson.github.io/gratia/reference/smooth_label.html","id":null,"dir":"Reference","previous_headings":"","what":"Extract the label for a smooth used by 'mgcv' — smooth_label","title":"Extract the label for a smooth used by 'mgcv' — smooth_label","text":"label 'mgcv' uses smooths useful many contexts, including selecting smooths labelling plots. smooth_label() extracts label 'mgcv' smooth object, .e. object inherits class \"mgcv.smooth\". typically found $smooth component GAM fitted mgcv::gam() mgcv::bam(), related functions.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/smooth_label.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Extract the label for a smooth used by 'mgcv' — smooth_label","text":"","code":"smooth_label(object, ...)  # S3 method for class 'gam' smooth_label(object, id, ...)  # S3 method for class 'mgcv.smooth' smooth_label(object, ...)"},{"path":"https://gavinsimpson.github.io/gratia/reference/smooth_label.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Extract the label for a smooth used by 'mgcv' — smooth_label","text":"object R object. Currently, methods class \"gam\" mgcv smooth objects inheriting class \"mgcv.smooth\" supported. ... arguments passed methods. id numeric; indices smooths whose labels extracted. missing, labels smooths model returned.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/smooth_label.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Extract the label for a smooth used by 'mgcv' — smooth_label","text":"character vector.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/smooth_label.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Extract the label for a smooth used by 'mgcv' — smooth_label","text":"","code":"load_mgcv() df <- data_sim(\"gwf2\", n = 100) m <- gam(y ~ s(x), data = df, method = \"REML\")  # extract the smooth sm <- get_smooths_by_id(m, id = 1)[[1]]  # extract the label smooth_label(sm) #> [1] \"s(x)\"  # or directly on the fitted GAM smooth_label(m$smooth[[1]]) #> [1] \"s(x)\"  # or extract labels by idex/position smooth_label(m, id = 1) #> [1] \"s(x)\""},{"path":"https://gavinsimpson.github.io/gratia/reference/smooth_samples.html","id":null,"dir":"Reference","previous_headings":"","what":"Posterior draws for individual smooths — smooth_samples","title":"Posterior draws for individual smooths — smooth_samples","text":"Returns draws posterior distributions smooth functions GAM. Useful, example, visualising uncertainty individual estimated functions.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/smooth_samples.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Posterior draws for individual smooths — smooth_samples","text":"","code":"smooth_samples(model, ...)  # S3 method for class 'gam' smooth_samples(   model,   select = NULL,   term = deprecated(),   n = 1,   data = newdata,   method = c(\"gaussian\", \"mh\", \"inla\", \"user\"),   seed = NULL,   freq = FALSE,   unconditional = FALSE,   n_cores = 1L,   n_vals = 200,   burnin = 1000,   thin = 1,   t_df = 40,   rw_scale = 0.25,   rng_per_smooth = FALSE,   draws = NULL,   partial_match = NULL,   mvn_method = c(\"mvnfast\", \"mgcv\"),   ...,   newdata = NULL,   ncores = NULL )"},{"path":"https://gavinsimpson.github.io/gratia/reference/smooth_samples.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Posterior draws for individual smooths — smooth_samples","text":"model fitted model supported types ... arguments passed methods. fitted_samples(), passed mgcv::predict.gam(). posterior_samples() passed fitted_samples(). predicted_samples() passed relevant simulate() method. select character; select smooth's posterior draw . default (NULL) means posteriors smooths model wil sampled . supplied, character vector requested terms. term Use select instead. n numeric; number posterior samples return. data data frame; new observations posterior draws model evaluated. supplied, data used fit model used data, available model. method character; method used draw samples posterior distribution. \"gaussian\" uses Gaussian (Laplace) approximation posterior. \"mh\" uses Metropolis Hastings sampler alternates t proposals proposals based shrunken version posterior covariance matrix. \"inla\" uses variant Integrated Nested Laplace Approximation due Wood (2019), (currently implemented). \"user\" allows user-supplied posterior draws (currently implemented). seed numeric; random seed simulations. freq logical; TRUE use frequentist covariance matrix parameter estimators, FALSE use Bayesian posterior covariance matrix parameters. unconditional logical; TRUE (freq == FALSE) Bayesian smoothing parameter uncertainty corrected covariance matrix used, available. n_cores number cores generating random variables multivariate normal distribution. Passed mvnfast::rmvn(). Parallelization take place OpenMP supported (appears work Windows current R). n_vals numeric; many locations evaluate smooth data supplied burnin numeric; number samples discard burnin draws. used method = \"mh\". thin numeric; number samples skip taking n draws. Results thin * n draws posterior taken. used method = \"mh\". t_df numeric; degrees freedom t distribution proposals. used method = \"mh\". rw_scale numeric; Factor scale posterior covariance matrix generating random walk proposals. Negative non finite skip random walk step. used method = \"mh\". rng_per_smooth logical; TRUE, behaviour gratia version 0.8.1 earlier used, whereby separate call random number generator (RNG) performed smooth. FALSE, single call RNG performed model parameters draws matrix; user supplied posterior draws used method = \"user\". partial_match logical; smooths selected partial matches select? TRUE, select can single string match . mvn_method character; one \"mvnfast\" \"mgcv\". default uses mvnfast::rmvn(), can considerably faster generate large numbers MVN random values mgcv::rmvn(), might work marginal fits, covariance matrix close singular. newdata Deprecated: use data instead. ncores Deprecated; use n_cores instead. number cores generating random variables multivariate normal distribution. Passed mvnfast::rmvn(). Parallelization take place OpenMP supported (appears work Windows current R).","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/smooth_samples.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Posterior draws for individual smooths — smooth_samples","text":"tibble additional classes \"smooth_samples\" `\"posterior_samples\". \"gam\" method, columns currently returned (order) : .smooth; character vector. Indicates smooth function particular draw, .term; character vector. Similar smooth, contain full label smooth, differentiate factor-smooths example. .; character vector. smooth involves term, variable named , NA_character_ otherwise. .row; integer. vector values seq_len(n_vals), repeated n > 1L. Indexes row data particular draw. .draw; integer. vector integer values indexing particular posterior draw row belongs . .value; numeric. value smooth function posterior draw covariate combination. xxx; numeric. series one columns containing data required smooth, named per variables involved respective smooth. Additional columns present case factor smooths, contain level factor named by_variable particular posterior draw.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/smooth_samples.html","id":"warning","dir":"Reference","previous_headings":"","what":"Warning","title":"Posterior draws for individual smooths — smooth_samples","text":"set variables returned order tibble subject change future versions. rely position.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/smooth_samples.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Posterior draws for individual smooths — smooth_samples","text":"Gavin L. Simpson","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/smooth_samples.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Posterior draws for individual smooths — smooth_samples","text":"","code":"load_mgcv() dat <- data_sim(\"eg1\", n = 400, seed = 2) m1 <- gam(y ~ s(x0) + s(x1) + s(x2) + s(x3), data = dat, method = \"REML\")  sms <- smooth_samples(m1, select = \"s(x0)\", n = 5, seed = 42) # \\donttest{ sms #> # A tibble: 1,000 x 8 #>    .smooth .term .type .by    .row .draw .value      x0 #>    <chr>   <chr> <chr> <chr> <int> <int>  <dbl>   <dbl> #>  1 s(x0)   s(x0) TPRS  NA        1     1 -0.357 0.00711 #>  2 s(x0)   s(x0) TPRS  NA        1     2 -0.465 0.00711 #>  3 s(x0)   s(x0) TPRS  NA        1     3 -0.720 0.00711 #>  4 s(x0)   s(x0) TPRS  NA        1     4 -1.27  0.00711 #>  5 s(x0)   s(x0) TPRS  NA        1     5 -1.18  0.00711 #>  6 s(x0)   s(x0) TPRS  NA        2     1 -0.365 0.0121  #>  7 s(x0)   s(x0) TPRS  NA        2     2 -0.464 0.0121  #>  8 s(x0)   s(x0) TPRS  NA        2     3 -0.708 0.0121  #>  9 s(x0)   s(x0) TPRS  NA        2     4 -1.24  0.0121  #> 10 s(x0)   s(x0) TPRS  NA        2     5 -1.16  0.0121  #> # i 990 more rows # }  ## A factor by example (with a spurious covariate x0) dat <- data_sim(\"eg4\", n = 1000, seed = 2)  ## fit model... m2 <- gam(y ~ fac + s(x2, by = fac) + s(x0), data = dat) sms <- smooth_samples(m2, n = 5, seed = 42) draw(sms)"},{"path":[]},{"path":"https://gavinsimpson.github.io/gratia/reference/smooth_terms.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"List the variables involved in smooths — smooth_terms","text":"","code":"smooth_terms(object, ...)"},{"path":"https://gavinsimpson.github.io/gratia/reference/smooth_terms.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"List the variables involved in smooths — smooth_terms","text":"object R object result call mgcv::gam(), mgcv::bam(), mgcv::gamm(), inherits classes \"gam\" \"mgcv.smooth\", \"fs.interaction\". ... arguments passed methods. Currently unused.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/smooth_type.html","id":null,"dir":"Reference","previous_headings":"","what":"Determine the type of smooth and return it n a human readable form — smooth_type","title":"Determine the type of smooth and return it n a human readable form — smooth_type","text":"Determine type smooth return n human readable form","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/smooth_type.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Determine the type of smooth and return it n a human readable form — smooth_type","text":"","code":"smooth_type(smooth)  # Default S3 method smooth_type(smooth)  # S3 method for class 'tprs.smooth' smooth_type(smooth)  # S3 method for class 'ts.smooth' smooth_type(smooth)  # S3 method for class 'cr.smooth' smooth_type(smooth)  # S3 method for class 'cs.smooth' smooth_type(smooth)  # S3 method for class 'cyclic.smooth' smooth_type(smooth)  # S3 method for class 'pspline.smooth' smooth_type(smooth)  # S3 method for class 'cpspline.smooth' smooth_type(smooth)  # S3 method for class 'Bspline.smooth' smooth_type(smooth)  # S3 method for class 'duchon.spline' smooth_type(smooth)  # S3 method for class 'fs.interaction' smooth_type(smooth)  # S3 method for class 'sz.interaction' smooth_type(smooth)  # S3 method for class 'gp.smooth' smooth_type(smooth)  # S3 method for class 'mrf.smooth' smooth_type(smooth)  # S3 method for class 'random.effect' smooth_type(smooth)  # S3 method for class 'sw' smooth_type(smooth)  # S3 method for class 'sf' smooth_type(smooth)  # S3 method for class 'soap.film' smooth_type(smooth)  # S3 method for class 't2.smooth' smooth_type(smooth)  # S3 method for class 'sos.smooth' smooth_type(smooth)  # S3 method for class 'tensor.smooth' smooth_type(smooth)  # S3 method for class 'mpi.smooth' smooth_type(smooth)  # S3 method for class 'mpd.smooth' smooth_type(smooth)  # S3 method for class 'cx.smooth' smooth_type(smooth)  # S3 method for class 'cv.smooth' smooth_type(smooth)  # S3 method for class 'micx.smooth' smooth_type(smooth)  # S3 method for class 'micv.smooth' smooth_type(smooth)  # S3 method for class 'mdcx.smooth' smooth_type(smooth)  # S3 method for class 'mdcv.smooth' smooth_type(smooth)  # S3 method for class 'miso.smooth' smooth_type(smooth)  # S3 method for class 'mifo.smooth' smooth_type(smooth)"},{"path":"https://gavinsimpson.github.io/gratia/reference/smooth_type.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Determine the type of smooth and return it n a human readable form — smooth_type","text":"smooth object inheriting class mgcv.smooth.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/smooths.html","id":null,"dir":"Reference","previous_headings":"","what":"Names of smooths in a GAM — smooths","title":"Names of smooths in a GAM — smooths","text":"Names smooths GAM","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/smooths.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Names of smooths in a GAM — smooths","text":"","code":"smooths(object)  # Default S3 method smooths(object)  # S3 method for class 'gamm' smooths(object)"},{"path":"https://gavinsimpson.github.io/gratia/reference/smooths.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Names of smooths in a GAM — smooths","text":"object fitted GAM related model. Typically result call mgcv::gam(), mgcv::bam(), mgcv::gamm().","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/spline_values.html","id":null,"dir":"Reference","previous_headings":"","what":"Evaluate a spline at provided covariate values — spline_values","title":"Evaluate a spline at provided covariate values — spline_values","text":"Evaluate spline provided covariate values","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/spline_values.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Evaluate a spline at provided covariate values — spline_values","text":"","code":"spline_values(   smooth,   data,   model,   unconditional,   overall_uncertainty = TRUE,   frequentist = FALSE )"},{"path":"https://gavinsimpson.github.io/gratia/reference/spline_values.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Evaluate a spline at provided covariate values — spline_values","text":"smooth currently object inherits class mgcv.smooth. data data frame values evaluate smooth . model fitted model; currently mgcv::gam() mgcv::bam() models suported. unconditional logical; confidence intervals include uncertainty due smoothness selection? TRUE, corrected Bayesian covariance matrix used. overall_uncertainty logical; uncertainty model constant term included standard error evaluate values smooth? frequentist logical; use frequentist covariance matrix?","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/spline_values2.html","id":null,"dir":"Reference","previous_headings":"","what":"Evaluate a spline at provided covariate values — spline_values2","title":"Evaluate a spline at provided covariate values — spline_values2","text":"function spline_values2() renamed spline_values() version 0.9.0. allowed following removal evaluate_smooth(), function using spline_values(). spline_values2() renamed spline_values().","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/spline_values2.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Evaluate a spline at provided covariate values — spline_values2","text":"","code":"spline_values2(   smooth,   data,   model,   unconditional,   overall_uncertainty = TRUE,   frequentist = FALSE )"},{"path":"https://gavinsimpson.github.io/gratia/reference/spline_values2.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Evaluate a spline at provided covariate values — spline_values2","text":"smooth currently object inherits class mgcv.smooth. data optional data frame values evaluate smooth . model fitted model; currently mgcv::gam() mgcv::bam() models suported. unconditional logical; confidence intervals include uncertainty due smoothness selection? TRUE, corrected Bayesian covariance matrix used. overall_uncertainty logical; uncertainty model constant term included standard error evaluate values smooth? frequentist logical; use frequentist covariance matrix?","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/term_names.html","id":null,"dir":"Reference","previous_headings":"","what":"Extract names of all variables needed to fit a GAM or a smooth — term_names","title":"Extract names of all variables needed to fit a GAM or a smooth — term_names","text":"Extract names variables needed fit GAM smooth","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/term_names.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Extract names of all variables needed to fit a GAM or a smooth — term_names","text":"","code":"term_names(object, ...)  # S3 method for class 'gam' term_names(object, ...)  # S3 method for class 'mgcv.smooth' term_names(object, ...)  # S3 method for class 'gamm' term_names(object, ...)"},{"path":"https://gavinsimpson.github.io/gratia/reference/term_names.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Extract names of all variables needed to fit a GAM or a smooth — term_names","text":"object fitted GAM object (inheriting class \"gam\" mgcv::smooth.construct smooth object, inheriting class \"mgcv.smooth\". ... arguments passed methods. currently used.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/term_names.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Extract names of all variables needed to fit a GAM or a smooth — term_names","text":"vector variable names required terms model","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/term_variables.html","id":null,"dir":"Reference","previous_headings":"","what":"Names of variables involved in a specified model term — term_variables","title":"Names of variables involved in a specified model term — term_variables","text":"Given name (term label) term model, returns names variables involved term.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/term_variables.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Names of variables involved in a specified model term — term_variables","text":"","code":"term_variables(object, term, ...)  # S3 method for class 'terms' term_variables(object, term, ...)  # S3 method for class 'gam' term_variables(object, term, ...)  # S3 method for class 'bam' term_variables(object, term, ...)"},{"path":"https://gavinsimpson.github.io/gratia/reference/term_variables.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Names of variables involved in a specified model term — term_variables","text":"object R object method dispatch performed term character; name model term, sense attr(terms(object), \"term.labels\"). Currently checked see term exists model. ... arguments passed methods.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/term_variables.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Names of variables involved in a specified model term — term_variables","text":"character vector variable names.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/theta.html","id":null,"dir":"Reference","previous_headings":"","what":"General extractor for additional parameters in mgcv models — theta","title":"General extractor for additional parameters in mgcv models — theta","text":"General extractor additional parameters mgcv models","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/theta.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"General extractor for additional parameters in mgcv models — theta","text":"","code":"theta(object, ...)  # S3 method for class 'gam' theta(object, transform = TRUE, ...)"},{"path":"https://gavinsimpson.github.io/gratia/reference/theta.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"General extractor for additional parameters in mgcv models — theta","text":"object fitted model ... arguments passed methods. transform logical; transform natural scale parameter","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/theta.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"General extractor for additional parameters in mgcv models — theta","text":"Returns numeric vector additional parameters","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/theta.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"General extractor for additional parameters in mgcv models — theta","text":"","code":"load_mgcv() df <- data_sim(\"eg1\", dist = \"poisson\", seed = 42, scale = 1 / 5) m <- gam(y ~ s(x0) + s(x1) + s(x2) + s(x3),   data = df, method = \"REML\",   family = nb() ) p <- theta(m)"},{"path":"https://gavinsimpson.github.io/gratia/reference/tidy_basis.html","id":null,"dir":"Reference","previous_headings":"","what":"A tidy basis representation of a smooth object — tidy_basis","title":"A tidy basis representation of a smooth object — tidy_basis","text":"Takes object class mgcv.smooth returns tidy representation basis.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/tidy_basis.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"A tidy basis representation of a smooth object — tidy_basis","text":"","code":"tidy_basis(smooth, data = NULL, at = NULL, coefs = NULL, p_ident = NULL)"},{"path":"https://gavinsimpson.github.io/gratia/reference/tidy_basis.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"A tidy basis representation of a smooth object — tidy_basis","text":"smooth smooth object inheriting class \"mgcv.smooth\". Typically, objects returned part fitted GAM GAMM $smooth component model object $gam$smooth component model fitted mgcv::gamm() gamm4::gamm4(). data data frame containing variables used smooth. data frame containing values smooth covariate(s) basis evaluated. coefs numeric; optional vector coefficients smooth p_ident logical vector; used handling scam::scam() smooths.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/tidy_basis.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"A tidy basis representation of a smooth object — tidy_basis","text":"tibble.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/tidy_basis.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"A tidy basis representation of a smooth object — tidy_basis","text":"Gavin L. Simpson","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/tidy_basis.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"A tidy basis representation of a smooth object — tidy_basis","text":"","code":"load_mgcv() df <- data_sim(\"eg1\", n = 400, seed = 42)  # fit model m <- gam(y ~ s(x0) + s(x1) + s(x2) + s(x3), data = df, method = \"REML\")  # tidy representaition of a basis for a smooth definition # extract the smooth sm <- get_smooth(m, \"s(x2)\") # get the tidy basis - need to pass where we want it to be evaluated bf <- tidy_basis(sm, at = df)  # can weight the basis by the model coefficients for this smooth bf <- tidy_basis(sm, at = df, coefs = smooth_coefs(sm, model = m))"},{"path":"https://gavinsimpson.github.io/gratia/reference/to_na.html","id":null,"dir":"Reference","previous_headings":"","what":"Sets the elements of vector to NA — to_na","title":"Sets the elements of vector to NA — to_na","text":"Given vector indexing elements x, sets selected elements x NA.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/to_na.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Sets the elements of vector to NA — to_na","text":"","code":"to_na(x, i)"},{"path":"https://gavinsimpson.github.io/gratia/reference/to_na.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Sets the elements of vector to NA — to_na","text":"x vector values vector values used subset x","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/to_na.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Sets the elements of vector to NA — to_na","text":"Returns x possibly elements set NA","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/too_far.html","id":null,"dir":"Reference","previous_headings":"","what":"Exclude values that lie too far from the support of data — too_far","title":"Exclude values that lie too far from the support of data — too_far","text":"Identifies pairs covariate values lie far original data. function currently basic wrapper around mgcv::exclude..far().","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/too_far.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Exclude values that lie too far from the support of data — too_far","text":"","code":"too_far(x, y, ref_1, ref_2, dist = NULL)"},{"path":"https://gavinsimpson.github.io/gratia/reference/too_far.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Exclude values that lie too far from the support of data — too_far","text":"x, y numeric; vector values covariates compare observed data ref_1, ref_2 numeric; vectors covariate values represent reference x1 x2` compared dist supplied, numeric vector length 1 representing distance data beyond observation excluded. example, want exclude values lie observation 10% range observed data, use 0.1.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/too_far.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Exclude values that lie too far from the support of data — too_far","text":"Returns logical vector length x1.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/too_far_to_na.html","id":null,"dir":"Reference","previous_headings":"","what":"Set rows of data to NA if the lie too far from a reference set of values — too_far_to_na","title":"Set rows of data to NA if the lie too far from a reference set of values — too_far_to_na","text":"Set rows data NA lie far reference set values","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/too_far_to_na.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Set rows of data to NA if the lie too far from a reference set of values — too_far_to_na","text":"","code":"too_far_to_na(smooth, input, reference, cols, dist = NULL)"},{"path":"https://gavinsimpson.github.io/gratia/reference/too_far_to_na.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Set rows of data to NA if the lie too far from a reference set of values — too_far_to_na","text":"smooth mgcv smooth object input data frame containing input observations columns set NA reference data frame containing reference values cols character vector columns whose elements set NA data lies far reference set dist numeric, distance reference set beyond elements input set NA","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/transform_fun.html","id":null,"dir":"Reference","previous_headings":"","what":"Transform estimated values and confidence intervals by applying a function — transform_fun","title":"Transform estimated values and confidence intervals by applying a function — transform_fun","text":"Transform estimated values confidence intervals applying function","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/transform_fun.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Transform estimated values and confidence intervals by applying a function — transform_fun","text":"","code":"transform_fun(object, fun = NULL, ...)  # S3 method for class 'smooth_estimates' transform_fun(object, fun = NULL, constant = NULL, ...)  # S3 method for class 'smooth_samples' transform_fun(object, fun = NULL, constant = NULL, ...)  # S3 method for class 'mgcv_smooth' transform_fun(object, fun = NULL, constant = NULL, ...)  # S3 method for class 'evaluated_parametric_term' transform_fun(object, fun = NULL, constant = NULL, ...)  # S3 method for class 'parametric_effects' transform_fun(object, fun = NULL, constant = NULL, ...)  # S3 method for class 'tbl_df' transform_fun(object, fun = NULL, column = NULL, constant = NULL, ...)"},{"path":"https://gavinsimpson.github.io/gratia/reference/transform_fun.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Transform estimated values and confidence intervals by applying a function — transform_fun","text":"object object apply transform function . fun function apply. ... additional arguments passed methods. constant numeric; constant apply transformation. column character; \"tbl_df\" method, column transform.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/transform_fun.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Transform estimated values and confidence intervals by applying a function — transform_fun","text":"Returns object estimate upper lower values confidence interval transformed via function.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/transform_fun.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"Transform estimated values and confidence intervals by applying a function — transform_fun","text":"Gavin L. Simpson","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/typical_values.html","id":null,"dir":"Reference","previous_headings":"","what":"Typical values of model covariates — typical_values","title":"Typical values of model covariates — typical_values","text":"Typical values model covariates","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/typical_values.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Typical values of model covariates — typical_values","text":"","code":"typical_values(object, ...)  # S3 method for class 'gam' typical_values(   object,   vars = everything(),   envir = environment(formula(object)),   data = NULL,   ... )  # S3 method for class 'data.frame' typical_values(object, vars = everything(), ...)"},{"path":"https://gavinsimpson.github.io/gratia/reference/typical_values.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Typical values of model covariates — typical_values","text":"object fitted GAM(M) model. ... arguments passed methods. vars terms include exclude returned object. Uses tidyselect principles. envir environment within recreate data used fit object. data optional data frame data used fit model reconstruction data model work.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/user_draws.html","id":null,"dir":"Reference","previous_headings":"","what":"Handle user-supplied posterior draws — user_draws","title":"Handle user-supplied posterior draws — user_draws","text":"Handle user-supplied posterior draws","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/user_draws.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Handle user-supplied posterior draws — user_draws","text":"","code":"user_draws(model, draws, ...)  # S3 method for class 'gam' user_draws(model, draws, index = NULL, ...)"},{"path":"https://gavinsimpson.github.io/gratia/reference/user_draws.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Handle user-supplied posterior draws — user_draws","text":"model fitted R model. Currently models fitted mgcv::gam() mgcv::bam(), return object inherits objects supported. , \"inherits\" used loose fashion; models fitted scam::scam() support even though models strictly inherit class \"gam\" far inherits() concerned. draws matrix; user supplied posterior draws used method = \"user\". ... arguments passed methods. index vector index (subset) columns draws.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/user_draws.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Handle user-supplied posterior draws — user_draws","text":"supplied draws must matrix (currently), 1 column per model coefficient, 1 row per posterior draw. \"gam\" method argument index, can used subset (select) coefficients (columns) draws. index can valid way selecting (indexing) columns matrix. index useful set posterior draws entire model (say mgcv::gam.mh()) wish use draws individual smooth, via smooth_samples().","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/variance_comp.html","id":null,"dir":"Reference","previous_headings":"","what":"Variance components of smooths from smoothness estimates — variance_comp","title":"Variance components of smooths from smoothness estimates — variance_comp","text":"wrapper mgcv::gam.vcomp() returns smoothing parameters expressed variance components.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/variance_comp.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Variance components of smooths from smoothness estimates — variance_comp","text":"","code":"variance_comp(object, ...)  # S3 method for class 'gam' variance_comp(object, rescale = TRUE, coverage = 0.95, ...)"},{"path":"https://gavinsimpson.github.io/gratia/reference/variance_comp.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Variance components of smooths from smoothness estimates — variance_comp","text":"object R object. Currently models fitted mgcv::gam() mgcv::bam() supported. ... arguments passed methods rescale logical; numerical stability reasons penalty matrices smooths rescaled fitting. rescale = TRUE, rescaling undone, resulting variance components original scale. needed comparing mixed model software, lmer(). coverage numeric; value 0 1 indicating (approximate) coverage confidence interval returned.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/variance_comp.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Variance components of smooths from smoothness estimates — variance_comp","text":"function wrapper mgcv::gam.vcomp() performs three additional services suppresses annoying text output mgcv::gam.vcomp() prints terminal, returns variance smooth well standard deviation, returns variance components tibble.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/vars_from_label.html","id":null,"dir":"Reference","previous_headings":"","what":"Returns names of variables from a smooth label — vars_from_label","title":"Returns names of variables from a smooth label — vars_from_label","text":"Returns names variables smooth label","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/vars_from_label.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Returns names of variables from a smooth label — vars_from_label","text":"","code":"vars_from_label(label)"},{"path":"https://gavinsimpson.github.io/gratia/reference/vars_from_label.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Returns names of variables from a smooth label — vars_from_label","text":"label character; length 1 character vector containing label smooth.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/vars_from_label.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Returns names of variables from a smooth label — vars_from_label","text":"","code":"vars_from_label(\"s(x1)\") #> [1] \"x1\" vars_from_label(\"t2(x1,x2,x3)\") #> [1] \"x1\" \"x2\" \"x3\""},{"path":"https://gavinsimpson.github.io/gratia/reference/which_smooths.html","id":null,"dir":"Reference","previous_headings":"","what":"Identify a smooth term by its label — which_smooths","title":"Identify a smooth term by its label — which_smooths","text":"Identify smooth term label","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/which_smooths.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Identify a smooth term by its label — which_smooths","text":"","code":"which_smooths(object, ...)  # Default S3 method which_smooths(object, ...)  # S3 method for class 'gam' which_smooths(object, terms, ...)  # S3 method for class 'bam' which_smooths(object, terms, ...)  # S3 method for class 'gamm' which_smooths(object, terms, ...)"},{"path":"https://gavinsimpson.github.io/gratia/reference/which_smooths.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Identify a smooth term by its label — which_smooths","text":"object fitted GAM. ... arguments passed methods. terms character; one (partial) term labels identify required smooths.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/worm_plot.html","id":null,"dir":"Reference","previous_headings":"","what":"Worm plot of model residuals — worm_plot","title":"Worm plot of model residuals — worm_plot","text":"Worm plot model residuals","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/worm_plot.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Worm plot of model residuals — worm_plot","text":"","code":"worm_plot(model, ...)  # S3 method for class 'gam' worm_plot(   model,   method = c(\"uniform\", \"simulate\", \"normal\", \"direct\"),   type = c(\"deviance\", \"response\", \"pearson\"),   n_uniform = 10,   n_simulate = 50,   level = 0.9,   ylab = NULL,   xlab = NULL,   title = NULL,   subtitle = NULL,   caption = NULL,   ci_col = \"black\",   ci_alpha = 0.2,   point_col = \"black\",   point_alpha = 1,   line_col = \"red\",   ... )  # S3 method for class 'glm' worm_plot(model, ...)  # S3 method for class 'lm' worm_plot(model, ...)"},{"path":"https://gavinsimpson.github.io/gratia/reference/worm_plot.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Worm plot of model residuals — worm_plot","text":"model fitted model. Currently models inheriting class \"gam\", well classes \"glm\" \"lm\" calls stats::glm stats::lm supported. ... arguments passed ot methods. method character; method used generate theoretical quantiles. default \"uniform\", generates reference quantiles using random draws uniform distribution inverse cummulative distribution function (CDF) fitted values. reference quantiles averaged n_uniform draws. \"simulate\" generates reference quantiles simulating new response data model observed values covariates, residualised generate reference quantiles, using n_simulate simulated data sets. \"normal\" generates reference quantiles using standard normal distribution. \"uniform\" computationally efficient, \"simulate\" allows reference bands drawn QQ-plot. \"normal\" avoided used fall back random number generator (\"simulate\") inverse CDF available family used model fitting (`\"uniform\"“). Note method = \"direct\" deprecated favour method = \"uniform\". type character; type residuals use. \"deviance\", \"response\", \"pearson\" residuals allowed. n_uniform numeric; number times randomize uniform quantiles direct computation method (method = \"uniform\"). n_simulate numeric; number data sets simulate estimated model using simulation method (method = \"simulate\"). level numeric; coverage level reference intervals. Must strictly 0 < level < 1. used method = \"simulate\". ylab character expression; label y axis. supplied, suitable label generated. xlab character expression; label y axis. supplied, suitable label generated. title character expression; title plot. See ggplot2::labs(). May vector, one per penalty. subtitle character expression; subtitle plot. See ggplot2::labs(). May vector, one per penalty. caption character expression; plot caption. See ggplot2::labs(). May vector, one per penalty. ci_col fill colour reference interval method = \"simulate\". ci_alpha alpha transparency reference interval method = \"simulate\". point_col colour points QQ plot. point_alpha alpha transparency points QQ plot. line_col colour used draw reference line.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/worm_plot.html","id":"note","dir":"Reference","previous_headings":"","what":"Note","title":"Worm plot of model residuals — worm_plot","text":"wording used mgcv::qq.gam() uses direct reference simulated residuals method (method = \"simulated\"). avoid confusion, method = \"direct\" deprecated favour method = \"uniform\".","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/worm_plot.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Worm plot of model residuals — worm_plot","text":"","code":"load_mgcv() ## simulate binomial data... dat <- data_sim(\"eg1\", n = 200, dist = \"binary\", scale = .33, seed = 0) p <- binomial()$linkinv(dat$f) # binomial p n <- sample(c(1, 3), 200, replace = TRUE) # binomial n dat <- transform(dat, y = rbinom(n, n, p), n = n) m <- gam(y / n ~ s(x0) + s(x1) + s(x2) + s(x3),   family = binomial, data = dat, weights = n,   method = \"REML\" )  ## Worm plot; default using direct randomization of uniform quantiles ## Note no reference bands are drawn with this method. worm_plot(m)   ## Alternatively use simulate new data from the model, which ## allows construction of reference intervals for the Q-Q plot worm_plot(m,   method = \"simulate\", point_col = \"steelblue\",   point_alpha = 0.4 )   ## ... or use the usual normality assumption worm_plot(m, method = \"normal\")"},{"path":"https://gavinsimpson.github.io/gratia/reference/zooplankton.html","id":null,"dir":"Reference","previous_headings":"","what":"Madison lakes zooplankton data — zooplankton","title":"Madison lakes zooplankton data — zooplankton","text":"Madison lake zooplankton data long-term study seasonal dynamics zooplankton, collected Richard Lathrop. data collected chain lakes Wisconsin (Mendota, Monona, Kegnonsa, Waubesa) approximately bi-weekly 1976 1994. consist samples zooplankton communities, taken deepest point lake via vertical tow. data provided Wisconsin Department Natural Resources collection processing fully described Lathrop (2000).","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/zooplankton.html","id":"format","dir":"Reference","previous_headings":"","what":"Format","title":"Madison lakes zooplankton data — zooplankton","text":"data frame","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/zooplankton.html","id":"source","dir":"Reference","previous_headings":"","what":"Source","title":"Madison lakes zooplankton data — zooplankton","text":"Pedersen EJ, Miller DL, Simpson GL, Ross N. 2018. Hierarchical generalized additive models: introduction mgcv. PeerJ Preprints 6:e27320v1 doi:10.7287/peerj.preprints.27320v1 .","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/zooplankton.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Madison lakes zooplankton data — zooplankton","text":"record consists counts given zooplankton taxon taken subsample single vertical net tow, scaled account relative volume subsample versus whole net sample area net tow rounded nearest 1000 give estimated population density per m2 taxon point time sampled lake.","code":""},{"path":"https://gavinsimpson.github.io/gratia/reference/zooplankton.html","id":"references","dir":"Reference","previous_headings":"","what":"References","title":"Madison lakes zooplankton data — zooplankton","text":"Lathrop RC. (2000). Madison Wisonsin Lakes Zooplankton 1976–1994. Environmental Data Initiative.","code":""},{"path":[]},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"new-features-development-version","dir":"Changelog","previous_headings":"","what":"New features","title":"gratia (development version)","text":"conditional_values() draw() method compute plot predictions fitted GAM conditional one covariates. function wrapper around fitted_values() allows user simple ways specify covariates condition values covariates take. provides similar functionality marginaleffects::plot_predictions(), simpler. See #300. penalty() basis() can now allow smooth reparameterized resulting basis identity matrix. clearly highlights penalty null space, functions penalty effect . draw.gam() draw.smooth_estimates() gain argument caption, , set FALSE plot smooth basis type caption plot. #307 appraise() qq_plot.gam() now allow user set random seed used generating reference quantiles method = \"uniform\" method = \"simulate\".","code":""},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"bug-fixes-development-version","dir":"Changelog","previous_headings":"","what":"Bug fixes","title":"gratia (development version)","text":"derivative_samples() ignoring scale argument. #293 Reported @jonathonmellor Argument level derivative_samples() included accidentally. v0.9.2.9002 argument deprecated using now generate warning. #291 draw() plotting cyclic P spline smooths. Reported @Zuckerbrot #297 derivatives() fail \"fs\" smooths parametric effects model. Reported @mahonmb #301 Partial residuals partial_residuals() draw.gam() wrong GAMs fitted family = binomial() weights argument contained binomial sample sizes prior weights used form weighted working residuals. Now working weights used instead. Reported @emchuron #273 Internal function gammals_link() expecting \"theta\" synonym scale parameter master table \"phi\" coded synonym. Now work expected. level() assumed level single value even though handle multiple levels. #321","code":""},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"gratia-092","dir":"Changelog","previous_headings":"","what":"gratia 0.9.2","title":"gratia 0.9.2","text":"CRAN release: 2024-06-25","code":""},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"breaking-changes-0-9-2","dir":"Changelog","previous_headings":"","what":"Breaking changes","title":"gratia 0.9.2","text":"parametric_effects() slightly escaped great renaming happened 0.9.0. Columns type term gain prefix .. now rectified two columns now .type .term.","code":""},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"user-visible-changes-0-9-2","dir":"Changelog","previous_headings":"","what":"User visible changes","title":"gratia 0.9.2","text":"Plots random effects now labelled smooth label. Previously, title taken fro variable involved smooth, doesn’t work terms like s(subject, continuous_var, bs = \"re\") random slopes, previsouly title \"subject\". Now terms title \"s(subject,continuous_var)\". Simple random intercept terms, s(subject, bs = \"re\"), now titled \"s(subject)\". #287 vignettes custom-plotting.Rmd, posterior-simulation.Rmd moved vignettes/articles thus longer available package vignettes. Instead, accessible Articles package website: https://gavinsimpson.github.io/gratia/","code":""},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"new-features-0-9-2","dir":"Changelog","previous_headings":"","what":"New features","title":"gratia 0.9.2","text":"fitted_samples() now works gam() models multiple linear predictors, currently location parameter supported. parameter indicated new variable .parameter returned object.","code":""},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"bug-fixes-0-9-2","dir":"Changelog","previous_headings":"","what":"Bug fixes","title":"gratia 0.9.2","text":"partial_residuals() computing partial residuals deviance residuals. compatibility mgcv::plot.gam(), partial residuals now computed working residuals. Reported @wStockhausen #273 appraise() passing ci_col argument qq_plot() worm_plot(). Reported Sate Ahmed. Couldn’t pass mvn_method posterior sampling functions user facing functions fitted_samples(), posterior_samples(), smooth_samples(), derivative_samples(), repsonse_derivatives(). Reported @stefgehrig #279 fitted_values() works quantile GAMs fitted qgam(). confint.gam() applying shift estimate upper lower interval. #280 reported @TIMAVID & @rbentham parametric_effects() draw.parametric_effects() forget levels factors (intentionally), lead problems ordered factors ordering levels preserved. Now, parametric_effects() returns named list factor levels attribute \"factor_levels\" containing required information order levels preserved plotting. #284 Reported @mhpob parametric_effects() fail parametric terms model interaction terms (don’t currently handle). #282","code":""},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"gratia-090","dir":"Changelog","previous_headings":"","what":"gratia 0.9.0","title":"gratia 0.9.0","text":"CRAN release: 2024-03-27","code":""},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"breaking-changes-0-9-0","dir":"Changelog","previous_headings":"","what":"Breaking changes","title":"gratia 0.9.0","text":"Many functions now return objects different named variables. order avoid clashes variable names used user’s models data, period (.) now used prefix generated variable names. functions whose names changed : smooth_estimates(), fitted_values(), fitted_samples(), posterior_samples(), derivatives(), partial_derivatives(), derivative_samples(). addition, add_confint() also adds newly-named variables. derivatives() partial_derivatives() now work like smooth_estimates(); place var data columns, gratia now stores data variables derivatives evaluated columns object actual variable names. way spline---sphere (SOS) smooths (bs = \"sos\") plotted changed use ggplot2::coord_sf() instead previously-used ggplot2::coord_map(). changed made result coord_map() soft-deprecated (“superseded”) minor versions ggplot2 now already, changes guides system version 3.5.0 ggplot2. axes plots created coord_map() never really worked correctly changing angle tick labels never worked. coord_map() superseded, didn’t receive updates guides system side effect changes, code plotted SOS smooths producing warning release ggplot2 version 3.5.0. projection settings used draw SOS smooths previously controlled via arguments projection orientation. arguments affect ggplot2::coord_sf(), Instead projection used controlled new argument crs, takes PROJ string detailing projection use integer refers known coordinate reference system (CRS). default projection used +proj=ortho +lat_0=20 +lon_0=XX XX mean longitude coordinates data points.","code":"1. `est` is now `.estimate`, 2. `lower` and `upper` are now `.lower_ci` and `.upper_ci`, 3. `draw` and `row` and now `.draw` and `.row` respectively, 4. `fitted`, `se`, `crit` are now `.fitted`, `.se`, `.crit`, respectively 5. `smooth`, `by`, and `type` in `smooth_estimates()` are now `.smooth`,    `.by`, `.type`, respectively."},{"path":[]},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"defunct-0-9-0","dir":"Changelog","previous_headings":"Breaking changes > Defunct and deprecated functions and arguments","what":"Defunct","title":"gratia 0.9.0","text":"evaluate_smooth() deprecated gratia version 0.7.0. function ’s methods removed package. Use smooth_estimates() instead.","code":""},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"deprecated-functions-0-9-0","dir":"Changelog","previous_headings":"Breaking changes > Defunct and deprecated functions and arguments","what":"Deprecated functions","title":"gratia 0.9.0","text":"following functions deprecated version 0.9.0 gratia. eventually removed package part clean ahead eventual 1.0.0 release. functions become defunct version 0.11.0 1.0.0, whichever released soonest. evaluate_parametric_term() deprecated. Use parametric_effects() instead. datagen() deprecated. never really originally designed , replaced data_slice().","code":""},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"deprecated-arguments-0-9-0","dir":"Changelog","previous_headings":"Breaking changes > Defunct and deprecated functions and arguments","what":"Deprecated arguments","title":"gratia 0.9.0","text":"make functions package consistent, arguments select, term, smooth used thing hence latter two deprecated favour select. deprecated argument used, warning issued value assigned argument assigned select function continue.","code":""},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"user-visible-changes-0-9-0","dir":"Changelog","previous_headings":"","what":"User visible changes","title":"gratia 0.9.0","text":"smooth_samples() now uses single call RNG generate draws posterior smooths. Previous version 0.9.0, smooth_samples() separate call mvnfast::rmvn() smooth. result, result call smooth_samples() model multiple smooths now produce different results generated previously. regain old behaviour, add rng_per_smooth = TRUE smooth_samples() call. Note, however, using per-smooth RNG calls method = \"mh\" inefficient , method, posterior draws coefficients model sampled . , use rng_per_smooth = TRUE method = \"gaussian\". output smooth_estimates() draw() method changed tensor product smooths involve one 2D marginal smooths. Now, covariate values supplied via data argument, smooth_estimates() identifies one marginals 2d surface allows covariates involved surface vary fastest, ahead terms marginals. change made provides better default nothing provided data. also affects draw.gam(). fitted_values() now level support location, scale, shape families. Supported families mgcv::gaulss(), mgcv::gammals(), mgcv::gumbls(), mgcv::gevlss(), mgcv::shash(), mgcv::twlss(), mgcv::ziplss(). gratia now requires dplyr versions >= 1.1.0 tidyselect >= 1.2.0. new vignette Posterior Simulation available, describes posterior simulation fitted GAMs using {gratia}.","code":""},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"new-features-0-9-0","dir":"Changelog","previous_headings":"","what":"New features","title":"gratia 0.9.0","text":"Soap film smooths using basis bs = \"\" now handled draw(), smooth_estimates() etc. #8 response_derivatives() new function computing derivatives response respect (continuous) focal variable. First second order derivatives can computed using forward, backward, central finite differences. uncertainty estimated derivative determined using posterior sampling via fitted_samples(), hence can derived Gaussian approximation posterior using Metropolis Hastings sampler (see .) derivative_samples() work horse function behind response_derivatives(), computes returns posterior draws derivatives additive combination model terms. Requested @jonathanmellor #237 data_sim() can now simulate response data gamma, Tweedie ordered categorical distributions. data_sim() gains two new example models \"gwf2\", simulating data Gu & Wabha’s f2 function, \"lwf6\", example function 6 Luo & Wabha (1997 JASA 92(437), 107-116). data_sim() can also simulate data use GAMs fitted using family = gfam() grouped families different types data response handled. #266 part #265 fitted_samples() smooth_samples() can now use Metropolis Hastings sampler mgcv::gam.mh(), instead Gaussian approximation, sample posterior distribution model specific smooths respectively. posterior_samples() new function family fitted_samples() smooth_samples(). posterior_samples() returns draws posterior distribution response, combining uncertainty estimated expected value response dispersion response distribution. difference posterior_samples() predicted_samples() latter includes variation due drawing samples conditional distribution response (uncertainty expected values ignored), former includes sources uncertainty. fitted_samples() can new use matrix user-supplied posterior draws. Related #120 add_fitted_samples(), add_predicted_samples(), add_posterior_samples(), add_smooth_samples() new utility functions add respective draws posterior distribution existing data object covariate values object: obj |> add_posterior_draws(model). #50 basis_size() new function extract basis dimension (number basis functions) smooths. Methods available objects inherit classes \"gam\", \"gamm\", \"mgcv.smooth\" (individual smooths). data_slice() gains method data frames tibbles. typical_values() gains method data frames tibbles. fitted_values() now works models fitted using mgcv::ocat() family. predicted probability category returned, alongside Wald interval created using standard error (SE) estimated probability. SE estimated probabilities transformed logit (linear predictor) scale, Wald credible interval formed, back-transformed response (probability) scale. fitted_values() now works GAMMs fitted using mgcv::gamm(). Fitted (predicted) values use GAM part model, thus exclude random effects. link() inv_link() work models fitted using cnorm() family. worm plot can now drawn place QQ plot appraise() via new argument use_worm = TRUE. #62 smooths() now works models fitted mgcv::gamm(). overview() now returns basis dimension smooth gains argument stars TRUE add significance stars output plus legend printed tibble footer. Part wish @noamross #214 New add_constant() transform_fun() methods smooth_samples(). evenly() gains arguments lower upper modify lower / upper bound interval evenly spaced values generated. add_sizer() new function add information whether derivative smooth significantly changing (credible interval excludes 0). Currently, methods derivatives() smooth_estimates() objects implemented. Part request @asanders11 #117 draw.derivatives() gains arguments add_change change_type allow derivatives smooths plotted indicators credible interval derivative excludes 0. Options allow periods decrease increase differentiated via change_type = \"sizer\" instead default change_type = \"change\", emphasises either type change way. Part wish @asanders11 #117 draw.gam() can now group factor smooths given factor single panel, rather plotting smooths level separate panels. achieved via new argument grouped_by. Requested @RPanczak #89 draw.smooth_estimates() can now also group factor smooths given factor single panel. underlying plotting code used draw_smooth_estimates() univariate smooths can now add change indicators plots smooths change indicators added object created smooth_estimates() using add_sizer(). See example ?draw.smooth_estimates. smooth_estimates() can, evaluating 3D 4D tensor product smooth, identify one 2D smooths marginal tensor product. users provide covariate values evaluate smooths, smooth_estimates() focus 2D marginal smooth (first one involved tensor product), instead following ordering terms definition tensor product. #191 example, te(z, x, y, bs = c(cr, ds), d = c(1, 2)), second marginal smooth 2D Duchon spline covariates x y. Previously, smooth_estimates() generated n values z x n_3d values y, evaluated tensor product combinations generated values. ignore structure implicit tensor product, likely want know surface estimated Duchon spline x y smoothly varies z. Previously smooth_estimates() generate surfaces z x, varying y. Now, smooth_estimates() correctly identifies one marginal smooths tensor product 2D surface focus surface varying terms tensor product. improved behaviour needed bam() models always possible obvious thing reorder smooths defining tensor product te(x, y, z, bs = c(ds, cr), d = c(2, 1)). discrete = TRUE used bam() terms tensor product may get rearranged model setup maximum efficiency (See Details ?mgcv::bam). Additionally, draw.gam() now also works way. New function null_deviance() extracts null deviance fitted model. draw(), smooth_estimates(), fitted_values(), data_slice(), smooth_samples() now work models fitted scam::scam(). matters, current support extends univariate smooths. generate_draws() new low-level function generating posterior draws fitted model coefficients. generate_daws() S3 generic function extensible users. Currently provides simple interface simple Gaussian approximation sampler (gaussian_draws()) simple Metropolis Hasting sample (mh_draws()) available via mgcv::gam.mh(). #211 smooth_label() new function extracting labels ‘mgcv’ creates smooths smooth object . penalty() default method works s(), te(), t2(), ti(), create smooth specification. transform_fun() gains argument constant allow addition constant value objects (e.g. estimate confidence interval). enables single obj |> transform_fun(fun = exp, constant = 5) instead separate calls add_constant() transform_fun(). Part discussion #79 model_constant() new function simply extracts first coefficient estimated model.","code":""},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"bug-fixes-0-9-0","dir":"Changelog","previous_headings":"","what":"Bug fixes","title":"gratia 0.9.0","text":"link(), inv_link(), related family functions ocat() weren’t correctly identifying family name result throw error even passed object correct family. link() inv_link() now work correctly betar() family fitted GAM. print() method lp_matrix() now converts matrix data frame conversion tibble. makes sense results typical behaviour columns printed object doubles. Constrained factor smooths (bs = \"sz\") factor first variable mentioned smooth (.e. s(x, f, bs = \"sz\") continuous x factor f) now plotable draw(). #208 parametric_effects() unable handle special parametric terms like poly(x) log(x) formulas. Reported @fhui28 #212 parametric_effects() now works better location, scale, shape models. Reported @pboesu #45 parametric_effects now works missing values one variables used fitted GAM. #219 response_derivatives() incorrectly using .data tidyselect selectors. typical_values() handle logical variables GAM fit mgcv stores numerics var.summary. affected evenly() data_slice(). #222 parametric_effects() fail two ordered factors model. Reported @dsmi31 #221 Continuous smooths evaluated median value variable instead value 1. #224 fitted_samples() (hence posterior_samples()) now handles models offset terms formula. Offset terms supplied via offset argument ignored mgcv:::predict.gam() hence ignored also gratia. Reported @jonathonmellor #231 #233 smooth_estimates() fail \"fs\" smooth multivariate base smoother used factor last variable specified definition smooth: s(x1, x2, f, bs = \"fs\", xt = list(bs = \"ds\")) work, s(f, x1, x2, bs = \"fs\", xt = list(bs = \"ds\")) (ordering variables places factor last) emit obscure error. ordering terms involved smooth now doesn’t matter. Reported @chrisaak #249. draw.gam() fail plotting multivariate base smoother used \"sz\" smooth. Now, use case identified message printed indicating (currently) gratia doesn’t know plot smooth. Reported @chrisaak #249. draw.gam() fail plotting multivariate base smoother used \"fs\" smooth. Now, use case identified message printed indicating (currently) gratia doesn’t know plot smooth. Reported @chrisaak #249. derivative_samples() fail order = 2 computing forward finite differences, regardless type order = 1. Partly reported @samlipworth #251. draw() method penalty() normalizing penalty range 0–1, claimed documented -1–1 argument normalize = TRUE. now fixed. smooth_samples() failing data supplied contained variables used smooth sampled. Hence generally fail unless single smooth sampled model contained single smooth. function never intended retain variables data written way fail relocating data columns end posterior sampling object. #255 draw.gam() draw.smooth_estimates() fail plotting univariate tensor product smooth (e.g. te(x), ti(x), t2()). Reported @wStockhausen #260 plot.smooth() printing factor level subtitles ordered factor smooths.","code":""},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"gratia-082","dir":"Changelog","previous_headings":"","what":"gratia 0.8.2","title":"gratia 0.8.2","text":"CRAN release: 2024-01-09 Small fixes CRAN.","code":""},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"gratia-081","dir":"Changelog","previous_headings":"","what":"gratia 0.8.1","title":"gratia 0.8.1","text":"CRAN release: 2023-02-02","code":""},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"user-visible-changes-0-8-1","dir":"Changelog","previous_headings":"","what":"User visible changes","title":"gratia 0.8.1","text":"smooth_samples() now returns objects variables involved smooths correct name. Previously variables named .x1, .x2, etc. Fixing #126 improving compatibility compare_smooths() smooth_estimates() allowed variables named correctly. gratia now depends version 1.8-41 later mgcv package.","code":""},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"new-features-0-8-1","dir":"Changelog","previous_headings":"","what":"New features","title":"gratia 0.8.1","text":"draw.gam() can now handle tensor products include marginal random effect smooth. Beware plotting smooths many levels, however, separate surface plot produced level.","code":""},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"bug-fixes-0-8-1","dir":"Changelog","previous_headings":"","what":"Bug fixes","title":"gratia 0.8.1","text":"Additional fixes changes dplyr 1.1.0. smooth_samples() now works sampling posteriors multiple smooths different dimension. #126 reported @Aariq","code":""},{"path":[]},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"user-visible-changes-0-8-0","dir":"Changelog","previous_headings":"","what":"User visible changes","title":"gratia 0.8.0","text":"{gratia} now depends R version 4.1 later. new vignette “Data slices” supplied {gratia}. Functions {gratia} harmonised use argument named data instead newdata passing new data evaluate features smooths. message printed newdata used now . Existing code need changed data takes value newdata. Note due way ... handled R, R script uses data argument, run versions gratia prior 8.0 (released; 0.7.3.8 using development version) user-supplied data silently ignored. , scripts using data check installed version gratia >= 0.8 package developers update depend versions >= 0.8 using gratia (>= 0.8) DESCRIPTION. order plots smooths changed draw.gam() match order smooths specified model formula. See Bug Fixes detail #154.","code":""},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"new-features-0-8-0","dir":"Changelog","previous_headings":"","what":"New features","title":"gratia 0.8.0","text":"Added basic support GAMLSS (distributional GAMs) fitted gamlss() function package GJRM. Support currently restricted draw() method. difference_smooths() can now include group means difference, many users expected. include group means use group_means = TRUE function call, e.g. difference_smooths(model, smooth = \"s(x)\", group_means = TRUE). Note: function still differs plot_diff() package itsadug, essentially computes differences model predictions. main practical difference effects beyond factor smooth, including random effects, may included plot_diff(). implements main wish #108 (@dinga92) #143 (@mbolyanatz) despite protestations complicated cases (isn’t; complexity just cancels .) data_slice() totally revised. Now, user provides values variables want slice variables model specified held typical values (.e. value observation closest median numeric variables, modal factor level.) Data slices now produced passing name = value pairs variables values want appear slice. example value pair can expression looked (evaluated) data argument model frame fitted model (default). example, resulting slice data frame 100 observations, comprising x1, vector 100 values spread evenly range x1, constant value mean x2 x2 variable, constant factor level, model class fac, fac variable model. partial_derivatives() new function computing partial derivatives multivariate smooths (e.g. s(x,z), te(x,z)) respect one margins smooth. Multivariate smooths dimension handled, one dimensions allowed vary. Partial derivatives estimated using method finite differences, forward, backward, central finite differences available. Requested @noamross #101 overview() provides simple overview model terms fitted GAMs. new bs = \"sz\" basis released mgcv version 1.18-41 now supported smooth_estimates(), draw.gam(), draw.smooth_estimates() basis unique plotting method. #202 basis() now method fitted GAM(M)s can extract estimated basis model plot , using estimated coefficients smooth weight basis. #137 also new draw.basis() method plotting results call basis(). method can now also handle bivariate bases. tidy_basis() lower level function heavy lifting basis(), now exported. tidy_basis() returns tidy representation basis supplied object inheriting class \"mgcv.smooth\". objects returned $smooth component fitted GAM(M) model. lp_matrix() new utility function quickly return linear predictor matrix estimated model. wrapper predict(..., type = \"lpmatrix\") evenly() synonym seq_min_max() preferred going forward. Gains argument produce sequences covariate increment units . ref_level() level() new utility functions extracting reference specific level factor respectively. useful specifying covariate values condition data slice. model_vars() new, public facing way returning vector variables used model. difference_smooths() now use user-supplied data points evaluate pair smooths. Also note argument newdata renamed data. #175 draw() method difference_smooths() now uses better labels plot titles avoid long labels even modest factor levels. derivatives() now works factor-smooth interaction (\"fs\") smooths. draw() methods now allow angle tick labels x axis plots rotated using argument angle. Requested @tamas-ferenci #87 draw.gam() related functions (draw.parametric_effects(), draw.smooth_estimates()) now add basis plot using caption. #155 smooth_coefs() new utility function extracting coefficients particular smooth fitted model. smooth_coef_indices() associated function returns indices (positions) vector model coefficients (returned coef(gam_model)) coefficients pertain stated smooth. draw.gam() now better handles patchworks plots one plots fixed aspect ratios. #190","code":"m <- gam(y ~ s(x1) + x2 + fac) data_slice(model, x1 = evenly(x1, n = 100), x2 = mean(x2))"},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"bug-fixes-0-8-0","dir":"Changelog","previous_headings":"","what":"Bug fixes","title":"gratia 0.8.0","text":"draw.posterior_smooths now plots posterior samples fixed aspect ratio smooth isotropic. #148 derivatives() now ignores random effect smooths (derivatives don’t make sense anyway). #168 confint.gam(...., method = \"simultaneous\") now works factor smooths parm passed full name specific smooth s(x)faclevel. order plots produced gratia::draw.gam() matches order smooths entered model formula. Recent changes internals gratia::draw.gam() switch smooth_estimates() undertaken lead change behaviour resulting use dplyr::group_split(), ’s coercion internally character vector factor. factor now created explicitly, levels set correct order. #154 Setting dist argument set response smooth values NA lay far support data multivariate smooths, lead incorrect scale response guide. now fixed. #193 Argument fun draw.gam() applied parametric terms. Reported @grasshoppermouse #195 draw.gam() adding uncertainty linear predictors smooths overall_uncertainty = TRUE used. Now draw.gam() includes uncertainty linear predictors smooth takes part. #158 partial_derivatives() works provided single data point evaluate derivative. #199 transform_fun.smooth_estimates() addressing wrong variable names trying transform confidence interval. #201 data_slice() doesn’t fail error used model contains offset term. #198 confint.gam() longer uses evaluate_smooth(), soft deprecated. #167 qq_plot() worm_plot() compute wrong deviance residuals used generate theoretical quantiles exotic families (distributions) available mgcv. also affected appraise() QQ plot; residuals shown plots deviance residuals shown y-axis QQ plot correct. generation reference intervals/quantiles affected.","code":""},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"gratia-073","dir":"Changelog","previous_headings":"","what":"gratia 0.7.3","title":"gratia 0.7.3","text":"CRAN release: 2022-05-09","code":""},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"user-visible-changes-0-7-3","dir":"Changelog","previous_headings":"","what":"User visible changes","title":"gratia 0.7.3","text":"Plots smooths now use “Partial effect” y-axis label place “Effect”, better indicate displayed.","code":""},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"new-features-0-7-3","dir":"Changelog","previous_headings":"","what":"New features","title":"gratia 0.7.3","text":"confint.fderiv() confint.gam() now return results tibble instead common--garden data frame. latter mostly already . Examples confint.fderiv() confint.gam() reworked, part remove inconsistent output examples run M1 macs.","code":""},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"bug-fixes-0-7-3","dir":"Changelog","previous_headings":"","what":"Bug fixes","title":"gratia 0.7.3","text":"compare_smooths() failed passed non-standard model “names” like compare_smooths(m_gam, m_gamm$gam) compare_smooths(l[[1]], l[[2]]) even evaluated objects valid GAM(M) models. Reported Andrew Irwin #150","code":""},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"gratia-072","dir":"Changelog","previous_headings":"","what":"gratia 0.7.2","title":"gratia 0.7.2","text":"CRAN release: 2022-03-17","code":""},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"new-features-0-7-2","dir":"Changelog","previous_headings":"","what":"New features","title":"gratia 0.7.2","text":"draw.gam() draw.smooth_estimates() can now handle splines sphere (s(lat, long, bs = \"sos\")) special plotting methods using ggplot2::coord_map() handle projection spherical coordinates. orthographic projection used default, essentially arbitrary (northern hemisphere-centric) default orientation view. fitted_values() insures data (hence returned object) tibble rather common garden data frame.","code":""},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"bug-fixes-0-7-2","dir":"Changelog","previous_headings":"","what":"Bug fixes","title":"gratia 0.7.2","text":"draw.posterior_smooths() redundantly plotting duplicate data rug plot. Now unique set covariate values used drawing rug. data_sim() passing scale argument bivariate example setting (\"eg2\"). draw() methods gamm() gamm4::gamm4() fits passing arguments draw.gam(). draw.smooth_estimates() produce subtitle data continuous smooth factor smooth. Now subtitle contains name continuous variable.","code":""},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"gratia-071","dir":"Changelog","previous_headings":"","what":"gratia 0.7.1","title":"gratia 0.7.1","text":"Due issue size package source tarball, wasn’t discovered submission CRAN, 0.7.1 never released.","code":""},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"new-features-0-7-1","dir":"Changelog","previous_headings":"","what":"New features","title":"gratia 0.7.1","text":"draw.gam() draw.smooth_estimates(): {gratia} can now handle smooths 3 4 covariates plotting. smooths 3 covariates, third covariate handled ggplot2::facet_wrap() set (default n = 16) small multiples drawn, 2d surface evaluated specified value third covariate. smooths 4 covariates, ggplot2::facet_grid() used draw small multiples, default producing 4 rows 4 columns plots specific values third fourth covariates. number small multiples produced controlled new arguments n_3d (default = n_3d = 16) n_4d (default n_4d = 4, yielding n_4d * n_4d = 16 facets) respectively. affects plotting; smooth_estimates() able handle smooths number covariates . handling higher-dimensional smooths, actually drawing plots default device can slow, especially default value n = 100 (3D 4D smooths result 160,000 data points plotted). recommended reduce n smaller value: n = 50 reasonable compromise resolution speed. model_concurvity() returns concurvity measures mgcv::concurvity() estimated GAMs tidy format. synonym concrvity() also provided. draw() method provided produces bar plot heatmap concurvity values depending whether overall concurvity smooth pairwise concurvity smooth model requested. draw.gam() gains argument resid_col = \"steelblue3\" allows colour partial residuals (plotted) changed.","code":""},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"bug-fixes-0-7-1","dir":"Changelog","previous_headings":"","what":"Bug fixes","title":"gratia 0.7.1","text":"model_edf() using type argument. result ever returned default EDF type. add_constant() methods weren’t applying constant required variables. draw.gam(), draw.parametric_effects() now actually work model parametric effects. #142 Reported @Nelson-Gon parametric_effects() fail model parametric terms predict.gam() returns empty arrays passed exclude = character(0).","code":""},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"gratia-070","dir":"Changelog","previous_headings":"","what":"gratia 0.7.0","title":"gratia 0.7.0","text":"CRAN release: 2022-02-07","code":""},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"major-changes-0-7-0","dir":"Changelog","previous_headings":"","what":"Major changes","title":"gratia 0.7.0","text":"draw.gam() now uses smooth_estimates() internally consequently uses draw() method underlying plotting code. simplified code compared evaluate_smooth() methods, allow future development addition features easily evaluate_smooth() retained. Similarly, evaluate_parametric_terms() now deprecated favour parametric_effects(), also used internally draw.gam() parametric terms present model (parametric = TRUE). lot code reused differences plots result change minimal, corner cases may missed. File Issue notice something changed think shouldn’t. draw.gam() now plots 2D isotropic smooths (TPRS Duchon splines) equally-scaled x y coordinates using coord_equal(ratio = 1). Alignment plots little different now plotting models multiple smooths. See Issue #81.","code":""},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"deprecated-functions-0-7-0","dir":"Changelog","previous_headings":"Major changes","what":"Deprecated functions","title":"gratia 0.7.0","text":"version 0.7.0, following functions considered deprecated use discouraged: fderiv() soft-deprecated favour derivatives(), evaluate_smooth() soft-deprecated favour smooth_estimates(), evaluate_parametric_term() soft-deprecated favour parametric_effects(). first call one functions generate warning, pointing newer, alternative, function. safe ignore warnings, deprecated functions longer receive updates thus risk removed package future date. newer alternatives can handle types models smooths, especially case smooth_estimates().","code":""},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"new-features-0-7-0","dir":"Changelog","previous_headings":"","what":"New features","title":"gratia 0.7.0","text":"fitted_values() provides tidy wrapper around predict.gam() generating fitted values model. New covariate values can provided via argument data. credible interval fitted values returned, values can link (linear predictor) response scale. Note function returns expected values response. Hence, “fitted values” used instead “predictions” case new covariate values differentiate values case generating new response values fitted model. rootogram() draw() method produce rootograms diagnostic plots fitted models. Currently models fitted poisson(), nb(), negbin(), gaussian() families. New helper functions typical_values(), factor_combos() data_combos() quickly creating data sets producing predictions fitted models covariatess fixed come typical representative values. typical_values() new helper function return typical values covariates fitted model. returns value observation closest median numerical covariates modal level factor preserving levels factor. typical_values() useful preparing data slices scenarios fitted values estimated model required. factor_combos() extracts returns combinations levels factors found data used fit model. Unlike typical_values(), factor_combos() returns combinations factor levels observed data, just modal level. Optionally, combinations factor levels can returned, just observed data. data_combos() combines returns factor data factor_combos() plus typical values numerical covariates. useful want generate predictions model combination factor terms holding continuous covariates median values. nb_theta() new extractor function returns theta parameter fitted negative binomial GAM (families nb() negbin()). Additionally, theta() has_theta() provide additional functionality. theta() experimental function extracting additional parameters model family. has_theta() useful checking additional parameters available family model. edf() extracts effective degrees freedom (EDF) fitted model specific smooth model. Various forms EDF can extracted. model_edf() returns EDF overall model. supplied multiple models, EDFs model returned comparison. draw.gam() can now show “rug” plot bivariate smooth drawing small points high transparency smooth surface data coordinates. addition, rugs plots factor smooths now show locations covariate values specific level factor levels. better reflects data used estimate smooth, even though basis smooth set using covariate locations. draw.gam() draw.smooth_estimates() now allow aspects plot changed: fill (colour) alpha attributes credible interval, line colour smooth can now specified using arguments ci_col, ci_alpha, smooth_col respectively. Partial residuals can now plotted factor smooths. allow , partial residuals filtered residuals associated particular level’s smooth drawn plot smooth. smooth_estimates() uses check_user_select_smooths() handle user-specified selection smooth terms. flexible previously, allows easier selection smooths evaluate. fixef() now imported (re-exported) nlme package, methods models fitted gam() gamm(), extract fixed effects estimates fitted models. fixed_effects() alias fixef(). draw() method smooth_samples() can now handle 2D smooths. Additionally, number posterior draws plot can now specified plotting using new argument n_samples, result n_samples draws selected random set draws plotting. New argument seed allows selection draws repeatable.","code":""},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"bug-fixes-0-7-0","dir":"Changelog","previous_headings":"","what":"Bug fixes","title":"gratia 0.7.0","text":"smooth_estimates() filtering user-supplied data level specific smooth used factor smooths. result smooth evaluated rows user-supplied data, therefore result nrow(user_data) * nlevels(by_variable) rows returned object instead nrow(user_data) rows. add_confint() method smooth_estimates() upper lower intervals reversed. #107 Reported @Aariq draw.gam() smooth_estimates() ignoring dist argument allows covariate values lie far support data excluded returning estimated values smooth plotting . #111 Reported @Aariq smooth_samples() factor GAM return samples first factor level . Reported @rroyaute discussion #121 smooth_samples() fail model contained random effect “smooths”. now ignored message running smooth_samples(). Reported @isabellaghement #121 link(), inv_link() failing models fitted family = scat(). Reported @Aariq #130","code":""},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"gratia-060","dir":"Changelog","previous_headings":"","what":"gratia 0.6.0","title":"gratia 0.6.0","text":"CRAN release: 2021-04-18","code":""},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"major-changes-0-6-0","dir":"Changelog","previous_headings":"","what":"Major changes","title":"gratia 0.6.0","text":"{cowplot} package replaced {patchwork} package producing multi-panel figures draw() appraise(). shouldn’t affect code used {gratia} , passed additional arguments cowplot::plot_grid() used align axis arguments draw() appraise(), ’ll need adapt code accordingly. Typically, can simply delete align axis arguments {patchwork} just work align plots nicely. arguments passed via ... cowplot::plot_grid() just ignored patchwork::wrap_plots() unless passed arguments match arguments patchwork::wrap_plots().","code":""},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"new-features-0-6-0","dir":"Changelog","previous_headings":"","what":"New features","title":"gratia 0.6.0","text":"{patchwork} package now used multi-panel figures. , {gratia} longer Imports {cowplot} package. Worm plot diagnostic plots available via new function worm_plot(). Worm plots detrended Q-Q plots, deviation Q-Q reference line emphasized deviations around line occupy full height plot. worm_plot() methods available models classes \"gam\", \"glm\", \"lm\". (#62) Smooths can now compared across models using compare_smooths(), comparisons visualised associated draw() method. (#85 @dill) feature bit experimental; returned object uses nested lists may change future users find confusing. reference line qq_plot() method = \"normal\" previously drawn line intercept 0 slope 1, match methods. inconsistent stats::qqplot() drew line 1st 3rd quartiles. qq_plot() method = \"normal\" now uses robust reference line. Reference lines methods remain drawn slope 1 intercept 0. qq_plot() method = \"normal\" now draws point-wise reference band using standard error order statistic. draw() method penalty() now plots penalty matrix heatmaps -logical orientation, match matrices might written printed R console. link(), inv_link() now work models fitted gumbls() shash() families. (#84) extract_link() lower level utility function related link() inv_link(), now exported.","code":""},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"user-visible-changes-0-6-0","dir":"Changelog","previous_headings":"","what":"User visible changes","title":"gratia 0.6.0","text":"default method name generating reference quantiles qq_plot() changed \"direct\" \"uniform\", avoid confusion mgcv::qq.gam() help page description methods. Accordingly using method = \"direct\" deprecated message effect displayed used. way smooths/terms selected derivatives() switched use mechanism draw.gam()’s select argument. get partial match term, now need also specify partial_match = TRUE call derivatives().","code":""},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"bug-fixes-0-6-0","dir":"Changelog","previous_headings":"","what":"Bug fixes","title":"gratia 0.6.0","text":"transform_fun() copy paste bug definition generic. (#96 @Aariq) derivatives() user-supplied newdata fail factor smooths interval = \"simultaneous\" introduce rows derivative == 0 interval = \"confidence\" didn’t subset rows newdata specific level factor computing derivatives. (#102 @sambweber) evaluate_smooth() can now handle random effect smooths defined using ordered factor. (#99 @StefanoMezzini)","code":""},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"gratia-051","dir":"Changelog","previous_headings":"","what":"gratia 0.5.1","title":"gratia 0.5.1","text":"CRAN release: 2021-01-23","code":""},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"new-features-0-5-1","dir":"Changelog","previous_headings":"","what":"New features","title":"gratia 0.5.1","text":"smooth_estimates() can now handle bivariate multivariate thinplate regression spline smooths, e.g.  s(x, z, ), tensor product smooths (te(), t2(), & ti()), e.g. te(x, z, ) factor smooth interactions, e.g. s(x, f, bs = \"fs\") random effect smooths, e.g. s(f, bs = \"re\") penalty() provides tidy representation penalty matrices smooths. tidy representation suitable plotting ggplot(). draw() method provided, represents penalty matrix heatmap.","code":""},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"user-visible-changes-0-5-1","dir":"Changelog","previous_headings":"","what":"User visible changes","title":"gratia 0.5.1","text":"newdata argument smooth_estimates() changed data originally intended.","code":""},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"gratia-050","dir":"Changelog","previous_headings":"","what":"gratia 0.5.0","title":"gratia 0.5.0","text":"CRAN release: 2021-01-10","code":""},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"new-features-0-5-0","dir":"Changelog","previous_headings":"","what":"New features","title":"gratia 0.5.0","text":"Partial residuals models can computed partial_residuals(). partial residuals weighted residuals model added contribution smooth term (returned predict(model,   type = \"terms\"). Wish #76 (@noamross) Also, new function add_partial_residuals() can used add partial residuals data frames. Users can now control extent colour fill scales used plotting smooths draw() methods use . useful change fill scale plotting 2D smooths, change discrete colour scale used plotting random factor smooths (bs = \"fs\"). user can pass scales via arguments discrete_colour continuous_fill. effects certain smooths can excluded data simulated model using simulate.gam() predicted_samples() passing exclude terms predict.gam(). allows excluding random effects, example, model predicted values used simulate new data conditional distribution. See example predicted_samples(). Wish #74 (@hgoldspiel) draw.gam() related functions gain arguments constant fun allow user-defined constants transformations smooth estimates confidence intervals applied. Part wish Wish #79. confint.gam() now works 2D smooths also. smooth_estimates() early version code replace (likely supersede) evaluate_smooth(). smooth_estimates() can currently handle 1D smooths standard types.","code":""},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"user-visible-changes-0-5-0","dir":"Changelog","previous_headings":"","what":"User visible changes","title":"gratia 0.5.0","text":"meaning parm confint.gam changed. argument now requires smooth label match smooth. vector labels can provided, partial matching smooth label works single parm value. default behaviour remains unchanged however; parm NULL smooths evaluated returned confidence intervals. data_class() longer exported; ever intended internal function.","code":""},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"bug-fixes-0-5-0","dir":"Changelog","previous_headings":"","what":"Bug Fixes","title":"gratia 0.5.0","text":"confint.gam() failing tensor product smooth due matching issues. Reported @tamas-ferenci #88 also fixes #80 related issue selecting specific smooth. vdiffr package now used conditionally package tests. Reported Brian Ripley #93","code":""},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"gratia-041","dir":"Changelog","previous_headings":"","what":"gratia 0.4.1","title":"gratia 0.4.1","text":"CRAN release: 2020-05-30","code":""},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"user-visible-changes-0-4-1","dir":"Changelog","previous_headings":"","what":"User visible changes","title":"gratia 0.4.1","text":"draw.gam() scales = \"fixed\" now applies terms can plotted, including 2d smooths. Reported @StefanoMezzini #73","code":""},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"bug-fixes-0-4-1","dir":"Changelog","previous_headings":"","what":"Bug fixes","title":"gratia 0.4.1","text":"dplyr::combine() deprecated. Switch vctrs::vec_c(). draw.gam() scales = \"fixed\" wasn’t using fixed scales 2d smooths model. Reported @StefanoMezzini #73","code":""},{"path":[]},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"new-features-0-4-0","dir":"Changelog","previous_headings":"","what":"New features","title":"gratia 0.4.0","text":"draw.gam() can include partial residuals drawing univariate smooths. Use residuals = TRUE add partial residuals univariate smooth drawn. feature available smooths one variable, smooths, factor-smooth interactions (bs = \"fs\"). coverage credible confidence intervals drawn draw.gam() can specified via argument ci_level. default arbitrarily 0.95 reason (rough) compatibility plot.gam(). change effect making intervals slightly narrower previous versions gratia; intervals drawn ± 2 × standard error. default intervals now drawn ± ~1.96 × standard error. New function difference_smooths() computing differences factor smooth interactions. Methods available gam(), bam(), gamm() gamm4::gamm4(). Also draw() method, can handle differences 1D 2D smooths currently (handling 3D 4D smooths planned). New functions add_fitted() add_residuals() add fitted values (expectations) model residuals existing data frame. Currently methods available objects fitted gam() bam(). data_sim() tidy reimplementation mgcv::gamSim() added ability use sampling distributions Gaussian models implemented. Currently Gaussian, Poisson, Bernoulli sampling distributions available. smooth_samples() can handle continuous variable smooths varying coefficient models. link() inv_link() now work families available mgcv, including location, scale, shape families, specialised families described ?mgcv::family.mgcv. evaluate_smooth(), data_slice(), family(), link(), inv_link() methods models fitted using gamm4() gamm4 package. data_slice() can generate data 1-d slice (single variable varying). colour points, reference lines, simulation band appraise() can now specified via arguments point_col, point_alpha, ci_col ci_alpha line_col passed qq_plot(), observed_fitted_plot(), residuals_linpred_plot(), residuals_hist_plot(), also now take new arguments applicable. Added utility functions is_factor_term() term_variables() working models. is_factor_term() identifies named term factor using information terms() object fitted model. term_variables() returns character vector variable names involved model term. strictly working parametric terms models. appraise() now works models fitted glm() lm(), underlying functions calls, especially qq_plot. appraise() also works models fitted family gaulss(). location scale models models fitted extended family functions supported upcoming releases.","code":""},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"user-visible-changes-0-4-0","dir":"Changelog","previous_headings":"","what":"User visible changes","title":"gratia 0.4.0","text":"datagen() now internal function longer exported. Use data_slice() instead. evaluate_parametric_term() now much stricter can evaluate main effect terms, .e. whose order, stored terms object model 1.","code":""},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"bug-fixes-0-4-0","dir":"Changelog","previous_headings":"","what":"Bug fixes","title":"gratia 0.4.0","text":"draw() method derivatives() getting x-axis label factor smooths correctly, instead using NA second subsequent levels factor. datagen() method class \"gam\" couldn’t possibly worked anything simplest models fail even simple factor smooths. issues fixed, behaviour datagen() changed, function now intended use users. Fixed issue models terms form factor1:factor2 incorrectly identified numeric parametric terms. #68","code":""},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"gratia-031","dir":"Changelog","previous_headings":"","what":"gratia 0.3.1","title":"gratia 0.3.1","text":"CRAN release: 2020-03-29","code":""},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"new-features-0-3-1","dir":"Changelog","previous_headings":"","what":"New features","title":"gratia 0.3.1","text":"New functions link() inv_link() access link function inverse fitted models family functions. Methods classes: \"glm\", \"gam\", \"bam\", \"gamm\" currently. #58 Adds explicit family() methods objects classes \"gam\", \"bam\", \"gamm\". derivatives() now handles non-numeric creating shifted data finite differences. Fixes problem stringsAsFactors = FALSE default R-devel. #64","code":""},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"bug-fixes-0-3-1","dir":"Changelog","previous_headings":"","what":"Bug fixes","title":"gratia 0.3.1","text":"Updated gratia work tibble versions >= 3.0","code":""},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"gratia-030","dir":"Changelog","previous_headings":"","what":"gratia 0.3.0","title":"gratia 0.3.0","text":"CRAN release: 2020-01-19","code":""},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"new-features-0-3-0","dir":"Changelog","previous_headings":"","what":"New features","title":"gratia 0.3.0","text":"gratia now uses mvnfast package random draws multivariate normal distribution (mvnfast::rmvn()). Contributed Henrik Singmann #28 New function basis() generating tidy representations basis expansions mgcv-like definition smooth, e.g. s(), te(), ti(), t2(). basic smooth types also simple draw() method plotting basis. basis() simple wrapper around mgcv::smoothCon() post processing basis model matrix tidy format. #42 New function smooth_samples() draw samples entire smooth functions posterior distribution. Also draw() method plotting posterior samples.","code":""},{"path":"https://gavinsimpson.github.io/gratia/news/index.html","id":"bug-fixes-0-3-0","dir":"Changelog","previous_headings":"","what":"Bug fixes","title":"gratia 0.3.0","text":"draw.gam() produce empty plots panels parametric terms 2 parametric terms model. Reported @sklayn #39. derivatives() now works factor smooths, including ordered factor smooths. function also now works correctly complex models multiple covariates/smooths. #47 derivatives() also now handles 'fs' smooths. Reported @tomand-uio #57. evaluate_parametric_term() hence draw.gam() fail ziplss() model ) gratia didn’t handle parametric terms models multiple linear predictors correctly, ii) gratia didn’t convert naming convention mgcv terms higher linear predictors. Reported @pboesu #45","code":""}]