Skip to content

lme4.0

6 messages · Andrzej Galecki, Ben Bolker, Viechtbauer Wolfgang (STAT)

Original
Viechtbauer Wolfgang (STAT)
# 
Dear lme4 Authors,

I am wondering if a lme4.0 version (i.e., a backwards compatible, bug fixes only version) is still going to be released (via CRAN). I noticed that lme4 version 1.0-4 is (currently) available for R-devel, but there is no lme4.0. Does this indicate that lme4.0 is off the table?

Best,
Wolfgang

--   
Wolfgang Viechtbauer, Ph.D., Statistician   
Department of Psychiatry and Psychology   
School for Mental Health and Neuroscience   
Faculty of Health, Medicine, and Life Sciences   
Maastricht University, P.O. Box 616 (VIJV1)   
6200 MD Maastricht, The Netherlands   
+31 (43) 388-4170 | http://www.wvbauer.com
Ben Bolker
# 
Viechtbauer Wolfgang (STAT <wolfgang.viechtbauer at ...> writes:
It's not off the table, but is very much up in the air (to mix
metaphors).  As of right now we haven't convinced the CRAN maintainers
to accept lme4.0, but we haven't given up either. Our
best chance of convincing them seems to be to get almost all of the
downstream packages go smoothly to lme4, so that only a few are left
depending on lme4.0 (thus hopefully convincing the CRAN maintainers 
that the scope for lme4.0/lme4 trouble is small).
  I don't expect this to be settled in the very near future; we will
focus our attention on helping package maintainers with the upgrade
process for the next couple of weeks, then assess the situation/renew
our lobbying efforts.
  In case anyone was thinking of contacting the CRAN maintainers
independently to plead for lme4.0, please *don't* -- if you have
thoughts on the subject or just want to make strong statements about
how important it is, please bring it up with the lme4 authors
instead.

  cheers
    Ben Bolker

PS this response is my own personal take on the situation, but I
think it is fairly representative of the view of all of the lme4
authors ...
Viechtbauer Wolfgang (STAT)
# 
Thank you for the feedback. It's good to have a better idea where things stand.

One argument in favor of lme4.0:

In the metafor package, I use lme4 to fit a particular Poisson/logistic mixed-effects model that involves a nonscalar random effects term. In lme4.0, this could be fitted with nAGQ > 1. In lme4 (1.0.x), this is (currently) not possible (only nAGQ=1, i.e., Laplace approximation). See also:

https://mailman.stat.ethz.ch/pipermail/r-sig-mixed-models/2013q3/020556.html

Best,
Wolfgang

--   
Wolfgang Viechtbauer, Ph.D., Statistician   
Department of Psychiatry and Psychology   
School for Mental Health and Neuroscience   
Faculty of Health, Medicine, and Life Sciences   
Maastricht University, P.O. Box 616 (VIJV1)   
6200 MD Maastricht, The Netherlands   
+31 (43) 388-4170 | http://www.wvbauer.com
Ben Bolker
# 
Viechtbauer Wolfgang (STAT <wolfgang.viechtbauer at ...> writes:
I agree that this is one of the best arguments for sticking with
lme4.0: I really hope we can implement this in the relatively near future
(unfortunately no guarantees on what "relatively near" means!), but we
need to find someone who has the time and energy to devote to this,
as well as being technically capable of understanding and implementing
multi-dimensional adaptive Gauss-Hermite quadrature within the current
R/reference class/RcppEigen framework ... (For anyone who's interested,
start from https://github.com/lme4/lme4/blob/master/src/external.cpp --
search for "glmerAGQ" ... and compare it with
https://r-forge.r-project.org/scm/viewvc.php/pkg/lme4.0
  /src/lmer.c?view=markup&revision=1725&root=lme4
[broken URL], starting at line 1355, to see how AGQ was implemented
in lme4.0 ...)

  Ben Bolker


  Ben Bolker