unbalanced mixed models (was Re: R-beta: R and S: Differences to a person who has used neither)

[Douglas Bates]

Peter Dalgaard BSA <p.dalgaard at kubism.ku.dk> writes:

> The things that you *will* be missing in R are more likely to be
> specific statistical analyses, like unbalanced mixed models (we really
> don't have these, do we?). One might hope that code for such things
> get contributed over time. Actually, a lot of the code in S-plus stems
> from contributed functions on Statlib, and can be ported without too
> much trouble.

One way of fitting unbalanced mixed models in S-plus is with the lme
function in the nlme library.  That function uses the ms() function
for optimizing the likelihood.  It may be possible to modify it to use
the optimization functions available in R but it would not be trivial
to do so.  I could promise to offer advice and help if someone wanted
to try to do that (the nlme library itself is GPL'd) but I don't think
either Jose Pinheiro or I could commit to doing an implementation in R
right now.  We are busy redesigning the whole library itself.

-- 
Douglas Bates                            bates at stat.wisc.edu
Statistics Department                    608/262-2598
University of Wisconsin - Madison        http://www.stat.wisc.edu/~bates/
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
r-help mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html
Send "info", "help", or "[un]subscribe"
(in the "body", not the subject !)  To: r-help-request at stat.math.ethz.ch
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=