[R-sig-ME] weights in lmer

Parsons, Van L. (CDC/CCHIS/NCHS) vlp1 at CDC.GOV
Fri Apr 4 16:39:48 CEST 2008

Prof Bates,
  There is a problem with the "weights" option in lmer.
 I checked a run with lme4_0.999375-12 for a model

lmer(    Reaction ~ (1|Subject) , data=sleepstudy     )
Vs a weight=1 model 

   wt1 = rep(1, 180 )    
 lmer(    Reaction ~ (1|Subject) , data=sleepstudy ,weight= wt1    )  
and observed different random and fixed effects.
The old version lme4_0.99875-9 was OK.

This issue has come up in a previous thread, and I 
wondering on the status of a correction. 




More information about the R-sig-mixed-models mailing list