[R-sig-ME] fixed effects

[Mitchell Maltenfort]

What is the difference between re.form = ~0 and re.form = NA ?

On Saturday, October 3, 2015, Ben Bolker <bbolker at gmail.com> wrote:

>     Perhaps you want
>
>   predict(fm,re.form=~0)
>
> ?
>
>
> On Sat, Oct 3, 2015 at 2:38 PM, Chen, Gang (NIH/NIMH) [C]
> <gangchen at mail.nih.gov <javascript:;>> wrote:
> > With the following generalized linear mixed-effects model:
> >
> > fm <- glmer( Y ~ Treatment + ( Treatment  | Subject) + (1|Trial),
> family=poisson))
> >
> > How can I obtain the estimated fixed effects? fitted() seems to provide
> the sum of fixed and random effects, while predict() does not seem to work
> because I'm modeling the overdispersion with (1|Trial). I could manually
> compute the fixed effects based on the coefficients, but is there a
> function available for this purpose?
> >
> > Thanks,
> > Gang
> >
> >         [[alternative HTML version deleted]]
> >
> > _______________________________________________
> > R-sig-mixed-models at r-project.org <javascript:;> mailing list
> > https://stat.ethz.ch/mailman/listinfo/r-sig-mixed-models
>
> _______________________________________________
> R-sig-mixed-models at r-project.org <javascript:;> mailing list
> https://stat.ethz.ch/mailman/listinfo/r-sig-mixed-models
>


-- 
Sent from Gmail Mobile

	[[alternative HTML version deleted]]