'rXX' notes

[drdhaval2785]

Earlier I thought that there would be uniform duplication or non-duplication dictionarywise.

A cursory reading of https://github.com/sanskrit-lexicon/hwnorm1/blob/master/proberrors/21violation.txt and https://github.com/sanskrit-lexicon/hwnorm1/blob/master/proberrors/rxx.txt showed me that it is not the case.

See the image.
akarkaSaH
but akarttA

S is not duplicated, t is.

[drdhaval2785]

I am trying to document some peculiarities if any.

[gasyoun]

21violation.txt is what kind of violation?

[drdhaval2785]

21violation.txt

It stores cases where there are NO duplication after 'r' even if the dictionary is in SKD,VCP,SHS,WIL,YAT,PD

Our hypothesis states that there is duplication after 'r' in these dictionaries.
Whenever our hypothesis doesn't hold true - it is stored in 21violation.txt.

If it fits in our hypothesis - it is stored in rxx.txt

For cursory statistics
rxstats.txt - Cases where particular pattern occurs in a given dictionary (no duplication - even though duplication was supposed to happen)

rxxstats.txt - Cases where particular pattern occurs in a given dictionary (with duplication).

[drdhaval2785]

Analysis of statistics

Observation 1 - PD dictionary doesn't do duplication as a rule. It does duplication as abberration.
See representative statistics

rkk pattern in PD dictionary is 2 / 12712
rKK pattern in PD dictionary is 0 / 12712
rgg pattern in PD dictionary is 0 / 12712
rGG pattern in PD dictionary is 0 / 12712
rcc pattern in PD dictionary is 0 / 12712
rjj pattern in PD dictionary is 0 / 12712
rtt pattern in PD dictionary is 20 / 12712
rdd pattern in PD dictionary is 0 / 12712
rpp pattern in PD dictionary is 0 / 12712
rmm pattern in PD dictionary is 2 / 12712
ryy pattern in PD dictionary is 0 / 12712
rvv pattern in PD dictionary is 2 / 12712

Therefore, PD deserves to be removed from list of duplicated dictionaries.

[drdhaval2785]

Observation 2 -
None of the dictionaries do duplication in case of 'K','G','N','C','J','Y','w','W','q','Q','T','D','n','P','B','r','l','S','z',s','h'

Exception - there is a sole entry 'rww' in YAT. dArwwura:YAT
Exception - there is a sole entry durnnirIkzya:WIL
Exception - there are two entries durllaBa:WIL and sudurllaBaH:SKD

Grammar note - the nonduplication in 'K','G','C','J','W','Q','T','D','P','B' are due to the fact that the duplicated entries converted to 'rkK','rgG','rcC','rjJ','rwW','rqQ','rtT','rdD','rpP','rbB' because of the rule झलां जश् झशि॥ ८।४।५२. This means that I would have to capture these patterns also.

[drdhaval2785]

Observation 3 -
Duplication in case of 'R' is predominantly seen in VCP
See

rRR pattern in SKD dictionary is 0 / 12712
rRR pattern in VCP dictionary is 212 / 12712
rRR pattern in SHS dictionary is 0 / 12712
rRR pattern in WIL dictionary is 3 / 12712
rRR pattern in YAT dictionary is 2 / 12712
rRR pattern in PD dictionary is 0 / 12712

[gasyoun]

The absence of uniform duplication rules makes me sad as well.
But I guess we do not need all the possible details, seems there
are too many. The general pattern stats gives the picture that is
enough, to filter out repetitive words if needed.

[funderburkjim]

addendum to dArwwura:YAT

Based on the sense,
WIL spells this dArddura.

MW spells this as dArdura.

dArwwura:YAT should be considered a print error, and be changed to dArddura:

• Since YAT closely follows WIL and
• dArwwura is out of alphabetical order in YAT: dArGasatra, dArQya, dArwwura, dArvvawa
  But dArddura would be restore alphabetical order.

Agree with correcting dArwwura to dArddura ?

[drdhaval2785]

Agree