[R] Longitudinal multivariate data analysis

[Marc Schwartz]

On Oct 7, 2010, at 6:48 AM, abderrahim youssef wrote:

> Dear all,
> 
> I am looking for an R package that fits multivariate gaussian or
> non-gaussian longitudinal outcomes.
> I am especially interested to non-gaussian outcomes  since the outcomes I've
> got are discrete (some are binomial and some are count data).
> Many thanks in advance,
> Abderrahim


You will want to start by looking at the lme4 package for R, within which are the lmer() and glmer() functions to fit such models.

More information here:

  http://cran.us.r-project.org/web/packages/lme4/index.html

and:

  http://lme4.r-forge.r-project.org/

There is also a R-SIG-MIXED-MODELS e-mail list focused on this subject matter:

  https://stat.ethz.ch/mailman/listinfo/r-sig-mixed-models

Finally, there is also a working draft FAQ, which I believe Ben Bolker is maintaining:

  http://glmm.wikidot.com/faq

HTH,

Marc Schwartz