This may sound like an April Fool's joke but it isn't really.
I just uploaded version 0.95-1 of the lme4 package for R. This is the
package with the new versions of the lme (linear mixed-effects) function
for R and the GLMM (generalized linear mixed models) function.
One distinguishing feature of the 0.95 series of releases of this
package is the absence of the lme and GLMM functions. They have been
replaced by a single function called lmer (pronounced "Elmer"). The
name could stand for "lme revised" or for "lme for R".
The method of specifying models for lmer is slightly different from that
for lme, which is the reason for using a new name for this function. In
lmer the random effects terms and their grouping factors are specified
in the formula. Furthermore, if a family argument is given then a
generalized linear mixed model is fit - otherwise a linear mixed model
is fit.
I am writing an article for the next edition of R News describing the
changes from lme to lmer. In the meantime you can look at the examples
and the tests directory of the mlmRev package to see examples of
specifying lmer models.
My apologies for springing this on the community without warning. I had
planned to do a bit more development and create more documentation
before releasing this new version but yesterday I accidently triggered a
cascade of inconsistencies by uploading a new version of the Matrix package.
If you are using the current lme4 and prefer not to make changes at this
time then stay with the 0.9 series of releases and don't update to 0.95-1.
_______________________________________________
R-packages mailing list
R-packages at stat.math.ethz.ch
https://stat.ethz.ch/mailman/listinfo/r-packages
