Dtheta error

[Lisa239217]

Hi, an error for Dtheta is now popping up for me which I suspect is related to to the recent Dtheta changes.
Error in FUN(X[[i]], ...) : object 'Dtheta' not found

The help example also now throws the same error

data(beardata)
tdf<- cbind(1:38, beardata$trapmat, matrix(1, nrow=38, ncol=8))
edf<- beardata$edf
edf[,1]<- 1
data<- data2oscr(edf, sess.col = 1, id.col = 2, occ.col = 3, trap.col = 4, 
                 sex.col = NULL, tdf = list(tdf), K = c(8), ntraps = c(38),
                 remove.zeros = TRUE, remove.extracaps = TRUE, sex.nacode = NULL, 
                 tdf.sep = "/")
scrFrame  <- make.scrFrame(caphist=data$y3d, traps=data$traplocs, trapCovs=NULL ,
                           trapOperation=data$trapopp)
sf <- data$scrFrame
ssDF <- make.ssDF(sf, buffer=3, res = 0.5)

#fit the NULL model: 
m0 <- oSCR.fit(model=list(D~1,p0~1,sig~1), scrFrame=sf, ssDF=ssDF)`

Error in FUN(X[[i]], ...) : object 'Dtheta' not found

[jaroyle]

hi Lisa, Thanks for bringing this to my attention! I don't know what happened here to be honest. I was messing around with Claude a couple weeks ago and didn't realize I modified the repo by asking a few questions about improving the run time efficiency of oSCR.fit. Apparently Claude took some liberties with the repo..... This is even more unfortunate because I don't know exactly what the fix is right now! I will dig into it... regards andy

…

On Sat, Mar 28, 2026 at 7:33 PM Lisa239217 ***@***.***> wrote: *Lisa239217* created an issue (jaroyle/oSCR#11) <#11> Hi, an error for Dtheta is now popping up for me which I suspect is related to to the recent Dtheta changes. Error in FUN(X[[i]], ...) : object 'Dtheta' not found The help example also now throws the same error data(beardata) tdf<- cbind(1:38, beardata$trapmat, matrix(1, nrow=38, ncol=8)) edf<- beardata$edf edf[,1]<- 1 data<- data2oscr(edf, sess.col = 1, id.col = 2, occ.col = 3, trap.col = 4, sex.col = NULL, tdf = list(tdf), K = c(8), ntraps = c(38), remove.zeros = TRUE, remove.extracaps = TRUE, sex.nacode = NULL, tdf.sep = "/") scrFrame <- make.scrFrame(caphist=data$y3d, traps=data$traplocs, trapCovs=NULL , trapOperation=data$trapopp) sf <- data$scrFrame ssDF <- make.ssDF(sf, buffer=3, res = 0.5) #fit the NULL model: m0 <- oSCR.fit(model=list(D~1,p0~1,sig~1), scrFrame=sf, ssDF=ssDF)` Error in FUN(X[[i]], ...) : object 'Dtheta' not found — Reply to this email directly, view it on GitHub <#11?email_source=notifications&email_token=AAHGZC7KMGXSPNTZOJ33CNT4TBOLHA5CNFSL4Z3JMQ5C6L3HNF2C22DVMIXUS43TOVSS6NBRGYYTSNRRGU4TBJTSMVQXG33OVJZXKYTTMNZGSYTFMSSWK5TFNZ2KYZTPN52GK4S7MNWGSY3L>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AAHGZCZBQJYJLL6N3ICJE5D4TBOLHAVCNFSM6AAAAACXDSCTU2VHI2DSMVQWIX3LMV43ASLTON2WKOZUGE3DCOJWGE2TSMA> . You are receiving this because you are subscribed to this thread.Message ID: ***@***.***>

[jaroyle]

Claude identified the fix here (presumably). I'm just documenting that here and will see if I can get it fixed.

Wait — actually looking again at the error message object 'Dtheta' not found, the more likely cause is simpler: the beardata example uses the default distmet = "euc" and smallslow = FALSE, so D should be populated. The issue is more likely that Dtheta is precomputed in the body of oSCR.fit but the likelihood functions msLL.nosex and msLL.sex receive it as an explicit argument — and the Dtheta = Dtheta argument in the nlm() call is resolving Dtheta from the calling environment correctly, but when nlm calls the likelihood internally it passes Dtheta as a named argument. The real bug is that Dtheta needs to be inside the if (!smallslow) { if (distmet == "euc") block, and for the ecol and smallslow cases we need a fallback. The fix is to guard the precomputation and handle all cases:
r# After the D precomputation block, replace our insertion with:

if (!smallslow && distmet == "euc") {
# D is already fully populated - precompute D^theta once
Dtheta <- lapply(seq_len(ns), function(s) D[[s]]^theta)
} else {
# For ecol or smallslow, D is built inside the likelihood per evaluation.
# Dtheta must also be computed there - pass NULL as placeholder and
# fall back to D[[s]]^theta inside the likelihood.
Dtheta <- vector("list", ns)
}
And then inside msLL.nosex change the Kern line back to handle the NULL case:
rKern <- if (is.null(Dtheta[[s]])) {
exp(-alphsig[s] * D[[s]]^theta)
} else {
exp(-alphsig[s] * Dtheta[[s]])
}
And similarly in msLL.sex for the inline kernel:
rDth <- if (is.null(Dtheta[[s]])) D[[s]]^theta else Dtheta[[s]]

then use Dth[trimR..., trimC...] in place of Dtheta[[s]][...]

This way the optimisation gets the caching benefit for the common euc/!smallslow case, and gracefully falls back to computing on the fly for ecol and smallslow — which is the correct behaviour since D isn't fixed for those cases anyway. Want me to update the patch file with this fix?

[Lisa239217]

thanks so much Andy, that would be amazing if you could.

[dnorename]

Hi, I was trying to use the oSCR.fit function and got the same error: Error in FUN(X[[i]], ...) : object 'Dtheta' not found. I was curious if you already updated the file with the fix

[jaroyle]

sorry about that, I messed up an update to oSCR.fit. I just pushed the old version back so it should work please let me know. If you email me aroyle AT usgs.gov then I can also send you the function.

…

On Wed, Apr 1, 2026 at 7:32 AM dnorename ***@***.***> wrote: *dnorename* left a comment (jaroyle/oSCR#11) <#11 (comment)> Hi, I was trying to use the oSCR.fit function and got the same error: Error in FUN(X[[i]], ...) : object 'Dtheta' not found. I was curious if you already updated the file with the fix — Reply to this email directly, view it on GitHub <#11?email_source=notifications&email_token=AAHGZC5AYJMYDGV5TIEEJ3T4TT45JA5CNFSNUABFM5UWIORPF5TWS5BNNB2WEL2JONZXKZKDN5WW2ZLOOQXTIMJWHE2DEMZRGI2KM4TFMFZW63VHMNXW23LFNZ2KKZLWMVXHJLDGN5XXIZLSL5RWY2LDNM#issuecomment-4169423124>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AAHGZC32EC6Z4YIK7POX6U34TT45JAVCNFSM6AAAAACXDSCTU2VHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHM2DCNRZGQZDGMJSGQ> . You are receiving this because you commented.Message ID: ***@***.***>