ctmm.select() warnings

[Devin Fitzpatrick]

Hello Chris,

I have been analyzing red squirrel home ranges, following your vignette for guidance. When I try to do "Maximum Likelihood Fitting the Easy Way" using ctmm.select() function, I get many warnings such as these pasted below. In most cases I can continue with the workflow and estimate the IID and OUF AKDE using akde() for individuals despite these warnings, however, for some individuals the OUF AKDE fails with the error "error: cannot allocate vector of size…" (even on machines with sufficient RAM). Does this failure have to do with the earlier warnings? And are the estimates that do not fail inaccurate because of these warnings?

Any guidance would be greatly appreciated.

Thank you,

Devin 

Warning messages:

1: In ctmm.fit(data, GUESS, trace = trace2, ...) :

  pREML failure: indefinite ML Hessian or divergent REML gradient.

2: In min(x) : no non-missing arguments to min; returning Inf

3: In max(x) : no non-missing arguments to max; returning -Inf

4: In ctmm.fit(data, GUESS, trace = trace2, ...) :

  pREML failure: indefinite ML Hessian or divergent REML gradient.

5: In ctmm.fit(data, GUESS, trace = trace2, ...) :

  pREML failure: indefinite ML Hessian or divergent REML gradient.

6: In ctmm.fit(data, GUESS, trace = trace2, ...) :

  pREML failure: indefinite ML Hessian or divergent REML gradient.

[Christen Fleming]

Hi Devin,

Do these issues go away if you are running the development version of the package?

  devtools::install_github("ctmm-initiative/ctmm")

If the warnings about min(x) and max(x) and especially the memory error don't go away, then I'd really appreciate a minimal working example (data + script) to fix those issues.

I think in the development version of the package, I also fixed the pREML warnings to only issue if ctmm.fit is being run directly, because as long as you are running ctmm.select, that really doesn't matter.

Best,

Chris

[Devin Fitzpatrick]

Hi Chris,

Thank you for your reply. Running the development version has fixed all the warnings, and in some cases the error, but I am still getting the "error: cannot allocate vector of size..." for some individuals. I will send you over an example. 

Best wishes,

Devin