[R-sig-ME] random regression with MCMCglmm

Celine Teplitsky teplitsky at mnhn.fr
Wed Jun 23 17:07:29 CEST 2010

Dear all,

I would like to fit a random regression using MCMCglmm. The problem is I 
don't manage to include heterogeneous residual variance (when I use just 
one residual variance the model is running). I have created a factor 
(Cat) grouping the environments into 4 food qualities from good to bad, 
so the model looks like:
priorA<-list(G=list(G1=list(V=diag(2), n=2)),R=list(V=diag(4), n= 2))

MCMCglmm(LD ~ Age+stFood , random=~us(1+st2Food):ID, rcov = 
~us(Cat):units, data=Data,prior=priorA, family = 
c("gaussian"),nitt=120000, burnin=20000, thin=100)

but I get this error message, and I really don't know how to fix the 

Error in `[.data.frame`(data, , components[[1]]) :
   undefined columns selected

Many thanks in advance for any help



Celine Teplitsky
Département Ecologie et Gestion de la Biodiversité UMR 7204
Unité Conservation des Espèces, Restauration et Suivi des Populations
Case Postale 51
55 rue Buffon 75005 Paris

Webpage : http://www2.mnhn.fr/cersp/spip.php?rubrique96
Fax : (33-1)-4079-3835
Phone: (33-1)-4079-3443

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