[R-sig-ME] robust linear mixed model
koller at stat.math.ethz.ch
Tue Oct 9 17:01:35 CEST 2012
(forgot to cc the mailinglist)
I am currently working on a package - robustlmm - that does exactly
this. It is currently available via GitHub only
(https://github.com/kollerma/robustlmm). (It hasn't been released to
CRAN since it depends on the new lme4.)
It is a combination of Fellner's (1986) approach to estimate the fixed
and random effects and the Design Adaptive Scale estimate developed
for MM-estimates (available in the package robustbase). The
documentation is not great, but there are some slides on my website
(http://stat.ethz.ch/people/kollerma) - a vignette is due soon. The
package provides a function "rlmer" that can be used just like "lmer".
On Tue, Oct 9, 2012 at 5:32 AM, Ahmad Rabiee <AhmadR at sbscibus.com.au> wrote:
> I have a dataset with some outliers or influential observations, and intend to do a mixed model regression. I am looking for procedures or a package that can do "robust linear mixed model", to account for the outliers or influential observations or groups.
> There are similar functions for fixed models (rlm) in MASS package, but I haven't been able to find anything on mixed models.
> R-sig-mixed-models at r-project.org mailing list
Manuel Koller <koller at stat.math.ethz.ch>
Seminar für Statistik, HG G 18, Rämistrasse 101
ETH Zürich 8092 Zürich SWITZERLAND
phone: +41 44 632-4673 fax: ...-1228
More information about the R-sig-mixed-models