[R-sig-ME] pwrssUpdate Error with new version of lme4

Steve Walker steve.walker at utoronto.ca
Mon Sep 23 13:56:41 CEST 2013


On 2013-09-23 4:34 AM, Johannes Radinger wrote:
> Hi,
>
> I am building binomial (logit, binary response) models with the glmer
> function of lme4.
> Today I updated to the new version 1.0-4 and now I get following errors:
> "pwrssUpdate
> did not converge in 30 iterations". In the old version the model fit was
> generally working.
> So is what can I do to make the models working again?

It would definitely help if you could provide more information about 
your model. Better yet, it would be good if you could provide a minimal 
example, including (possibly fake, reduced, or permuted) data, that 
reproduces your problem.

http://stackoverflow.com/questions/5963269/how-to-make-a-great-r-reproducible-example


> Increase the number
> of iterations?

Currently the number of pwrss iterations is hard-coded at 30. However, 
you can try altering the tolerance by setting the control argument to 
something like glmerControl(tolPwrss=1e-3) say. In general, you can find 
more information about control parameters at ?glmerControl

Steve

> Change some other settings in my model?
>
> Thank you for your suggestions!
>
> Best regards,
> Johannes
>
> 	[[alternative HTML version deleted]]
>
> _______________________________________________
> R-sig-mixed-models at r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-sig-mixed-models
>



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