[R-sig-ME] lme4 to nlme translation

[Douglas Bates]

Unless group2 is nested within group1 it will be difficult to translate
such a model into nlme syntax.  We developed lme4 and its computational
methods specifically to be able to handle multiple non-nested grouping
factors, for which nlme was not well suited.  There is an example in
Pinheiro and Bates (2000) of fitting a model with crossed grouping factors
using the nlme package but the technique doesn't scale well.

Depending on the nature of the heteroscedasticity you may be better off
trying to transform the response, say with the Box-Cox family of
transformations.


On Wed, Mar 29, 2023 at 8:41 AM kalman.toth via R-sig-mixed-models <
r-sig-mixed-models using r-project.org> wrote:

> Dear All,
>
> I am struggling to 'translate' this lme4 syntax to nlme:
>
> lmer(response ~ (1|group1) + (1|group2) + (1|group1:group3), data)
>
> Let's say that the general assumptions hold but actually the main reason
> of the 'translation' would be to study the effect of heteroscedasticity on
> the variance estimates.
>
> Could someone help me?
>
> Thanks,
> Kalman
>         [[alternative HTML version deleted]]
>
> _______________________________________________
> R-sig-mixed-models using r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-sig-mixed-models
>

	[[alternative HTML version deleted]]