[R-sig-ME] A nAGQ=0 question for glmer()

Ben Bolker bbo|ker @end|ng |rom gm@||@com
Wed Sep 8 19:13:15 CEST 2021

On 9/8/21 12:24 PM, Hedyeh Ahmadi wrote:
> Hello all,
> I have the following two questions:
>    1.  If one is using a glmer() model with the nAGQ=0 option, I was wondering if there is a way to see if the model has actually converged?
>    2.  If the mentioned model runs with no error/warning , does that mean the model has converged?

   The convergence diagnostics and messages don't differ between nAGQ=0 
and nAGQ>0 (so, the answer to #2 is 'yes').

   For a fitted model gm1, any problems with convergence will be 
recorded in gm1 using modelinfo; specifically, see the $conv, $warnings, and 
$message components ($conv has two elements, $opt (optimizer convergence 
code) and $lme4 (warnings/non-convergence warnings from lme4). 
$conv$opt should be 0, $conv$lme4 and $warnings should be length-zero.

   For what it's worth, the best description of the 
assumption/approximation that nAGQ=0 makes is in the TMB (!) 
documentation: https://kaskr.github.io/adcomp/_book/Appendix.html, 
section 16.2

> Any help would be appreciated.
> Best,
> Hedyeh Ahmadi, Ph.D.
> Statistician
> Keck School of Medicine
> Department of Preventive Medicine
> University of Southern California
> LinkedIn
> www.linkedin.com/in/hedyeh-ahmadi<http://www.linkedin.com/in/hedyeh-ahmadi>
> <http://www.linkedin.com/in/hedyeh-ahmadi><http://www.linkedin.com/in/hedyeh-ahmadi>
> 	[[alternative HTML version deleted]]
> _______________________________________________
> R-sig-mixed-models using r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-sig-mixed-models

Dr. Benjamin Bolker
Professor, Mathematics & Statistics and Biology, McMaster University
Director, School of Computational Science and Engineering
Graduate chair, Mathematics & Statistics

More information about the R-sig-mixed-models mailing list