Scaling
-This is a dog.
+library(ramp.work) +library(deSolve) +An important question in malaria is the relationship between various +metrics, especially the relationship between the annual entomological +inoculation rate (aEIR) and the parasite rate (PR). To facilitate +analysis of malaria data, we have developed some functions to compute +scaling relationships.
xde_scaling
The function xde_scaling() defines the relationship
-between the EIR and the PR, and it outputs stable orbits for each value
-of aEIR in a mesh from \(10^{-1}\) up
-to \(10^{3}\) The code is in
+between the EIR and the PR. It analyzes stable orbits and outputs the
+average annual EIR and average annual PR for an even mesh of
+log(aEIR) values running from \(10^{-1}\) up to \(10^{3}\) The code is in
mob_library/Work
The cohort trace functions in ramp.xds take the form of
-F(a, bday=0, scale=1).
To illustrate, we pick a function describing a seasonal pattern using
+ramp.xds::make_F_sin
tt <- seq(0, 730, by=5)
p1 <- makepar_F_sin(floor=0.1)
Fsin <- make_function(p1)
plot(tt, Fsin(tt), type="l")
Next, we set up a cohort model:
xds_setup_cohort(Xname = "SIS", F_season=Fsin) -> sisxds_solve_cohort(sis) -> sisThe function xde_scaling_eir runs the model over a mesh
+of N=25 values:
xde_scaling_eir(sis, 25) -> sisThe results are attached as sis$outputs$eirpr
plot_eirpr(sis)
require(viridis)## Loading required package: viridis## Loading required package: viridisLitexde_scaling
lines(scaling[[15]]$aeir, scaling[[15]]$pr, col = clrs[15]) lines(scaling[[20]]$aeir, scaling[[20]]$pr, col = clrs[20]) }) -
xde_pr2eir()
@@ -436,65 +440,13 @@ xde_pr2eir()
other models can be bounded, so we don’t want to return nonsense
values.
preir_i$errors## pr1 pr2 pr3 pr4
-## 0.001000000 0.004284405 0.012302189 0.999000000## pr1 pr2
+## 0.001 0.999We can plot the others:
plot_eirpr(sis)
with(sis$outputs$eirpr, points(aeir, pr, pch = 15))
with(preir_i, points(365*eir, pr, pch = 19, col = "red"))
Scaling
-Seasonality
-sis0 <- xds_setup_cohort(Xname = "SIS", F_season = Fsin)
-xde_scaling_eir(sis0, 25) -> sis0clrs = turbo(25)
-with(sis$outputs$eirpr, plot(aeir, pr, type = "l", log = "x", xaxt= "n", xlab = "aEIR", ylab = "PR"))
-axis(1, 10^(-1:3), c(0.1, 1, 10, 100, 1000))
-lines(sis$outputs$eirpr$aeir, sis0$outputs$eirpr$pr, col = "tomato", lwd=2)
-
-with(sis0$outputs$eirpr, points(aeir, pr, col = clrs))
-with(sis0$outputs$eirpr, lines(scaling[[5]]$aeir, scaling[[5]]$pr, col = clrs[5]))
-with(sis0$outputs$eirpr, lines(scaling[[10]]$aeir, scaling[[10]]$pr, col = clrs[10]))
-with(sis0$outputs$eirpr, lines(scaling[[15]]$aeir, scaling[[15]]$pr, col = clrs[15]))
-with(sis0$outputs$eirpr, lines(scaling[[20]]$aeir, scaling[[20]]$pr, col = clrs[20]))
Drug taking
-sip = xds_setup_cohort(Xname = "SIP", F_season=Fsin)
-sip$Xpar[[1]]$eta = 1/40
-xde_scaling_eir(sip, 25) -> sipsip1 = setup_exposure_nb(sip, 1/50)
-xde_scaling_eir(sip1, 25) -> sip1with(sis$outputs$eirpr, plot(aeir, pr, type = "l", log = "x", xaxt= "n", xlab = "aEIR", ylab = "PR"))
-axis(1, 10^(-1:3), c(0.1, 1, 10, 100, 1000))
-with(sip$outputs$eirpr, lines(aeir, pr, col = "darkorange"))
-with(sip1$outputs$eirpr, lines(aeir, pr, col = "brown"))Environmental Heterogeneity
-sis4 <- setup_exposure_nb(sis, 1/50)
-xde_scaling_eir(sis4, 25) -> sis4with(sis$outputs$eirpr, plot(aeir, pr, type = "l", log = "x", xaxt= "n", xlab = "aEIR", ylab = "PR"))
-axis(1, 10^(-1:3), c(0.1, 1, 10, 100, 1000))
-#with(sis2$outputs$eir, lines(aeir, pr, col = "blue"))
-#with(sis3$outputs$eir, lines(aeir, pr, col = "purple"))
-with(sis4$outputs$eir, lines(aeir, pr, col = "darkblue"))
Travel
-sis5 <- setup_travel_static(sis, delta = 1/5/365)
-xde_scaling_eir(sis5, 25) -> sis5with(sis$outputs$eirpr, plot(aeir, pr, type = "l", log = "x", xaxt= "n", xlab = "aEIR", ylab = "PR"))
-axis(1, 10^(-1:3), c(0.1, 1, 10, 100, 1000))
-with(sis5$outputs$eir, lines(aeir, pr, col = "darkgreen")) 
