[R-sig-ME] How to check the iterative history of convergence and change convergence criterion in glmmadmb

[Ben Bolker]

If you specify verbose=TRUE when running the model you will see (quite
a lot of) interim output describing the optimization progress.  See
?admbControl for more control options; you can also use the
"extra.args" argument to pass additional arguments to ADMB, but you'll
need to look at the ADMB documentation
<http://www.admb-project.org/documentation> to figure out what to do:
from chapter 12 of the manual
<http://ftp.admb-project.org/admb-11.4/manuals/admb-11.4.1.pdf>, it
looks like  extra.args="-crit xxx" (where xxx is a numerical
convergence criterion -- not sure what makes sense).

See also <https://www.google.ca/url?sa=t&rct=j&q=&esrc=s&source=web&cd=9&cad=rja&uact=8&ved=0CEgQFjAIahUKEwju072Lq8rIAhVMGx4KHW5yAPU&url=ftp%3A%2F%2Fglpd.fw.msu.edu%2FQFC%2FADMB-Videos%2FCurrent_ADMB_Video_Release%2FDefault%2520Convergence%2520Criterion%2FADMB%2520-%2520Default%2520Convergence%2520Criterion%2520Video%2520Handout.pdf&usg=AFQjCNGcJXjNmcgBgu--8tMkxX__d6RMgQ&sig2=smswu71Be-z7X-PVMK-kGA>
, and consider asking about convergence criteria on the ADMB list if
you can't figure it out.


On Thu, Oct 15, 2015 at 10:58 PM,  <qiujing at udel.edu> wrote:
> I am using glmmadmb to fit negative binomial mixed effects model to my data. However it fails to converge sometimes. I want to understand what is causing the failure of convergence by first of all check whether it really fails to converge or just converges very slowly。 when I fit the data using was glmmmix， the iteration history reply shows that my model does converges but fail to meet the default convergency criteria. When I relax the convergence criterion, it converges well.
>
> Is there anyway I can find the iteration history in glmmadmb as well? And can I change the default convergence criterion?
>
> Thanks!
>
> Jing
>
> Jing qiu
> _______________________________________________
> R-sig-mixed-models at r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-sig-mixed-models