Message-ID: <BANLkTinoXrC=8YugKgAY6=kJfTS-GL0Qpg@mail.gmail.com>
Date: 2011-06-14T16:47:07Z
From: Douglas Bates
Subject: Question regarding glmer-function of R-package lme4
In-Reply-To: <DC6A62F93770C44B8D2856DA3744EF7C06FD717D@VS1.exc.top.gwdg.de>
As far as I know the REML criterion is not defined for generalized
linear mixed models. I have taken the liberty of copying the
R-SIG-Mixed-Models at R-project.org list on this response in case others
who read the list can offer suggestions.
On Tue, Jun 14, 2011 at 11:31 AM, Greb, Friederike
<fgreb at uni-goettingen.de> wrote:
> Dear Professor Bates,
>
>
>
> I am currently working with generalized linear mixed models using your
> R-package lme4, in particular the glmer function. I need the value of the
> REML-function, but don?t really understand how to get it for a non-gaussian
> family. Is it possible to extract it from the output of a call to glmer (or
> lmer)? Thank you very much for your help!
>
>
>
> Best regards,
>
> Friederike Greb
>
>
>
>
>
> -----------------------------------------------------------------------
>
> Friederike Greb
>
> Research Associate / PhD student
>
> Georg-August-Universit?t G?ttingen
>
>
>
> CRC "Poverty, Equity and Growth in Developing Countries"
>
> Wilhelm-Weber-Str. 2
>
> 37073 G?ttingen, Germany
>
> Tel: +49 (0)551 39 200 25
>
>
>
> Institute for Mathematical Stochastics
>
> Goldschmidstr. 7
>
> 37077 G?ttingen, Germany
>
> Tel: +49 (0)551 39 13 504
>
>
>
> http://www.uni-goettingen.de/en/102480.html
>
> Email: fgreb at uni-goettingen.de
>
>
>
>