[R-sig-ME] two random statements with a correlation structures

Nina Bhola nina.bhola at gmail.com
Fri Mar 12 13:49:49 CET 2010


Dear list,

I am modelling trends in counts of trees across years and plots in
different height classes across a savanna ecosystem.
I am using a generalized linear mixed model with a negative binomial
distribution
(many zeros).
My fixed effects are height class + year + plot

I have two random statements which are not about random height class,
plot or year. These variables are inducing correlations between all
observations made on the same  plot in a given year (first statement)
and correlations between observations for a given age in a given, plot
and year (second random statement).

Therefore, I would like model these random effects in R like we could
do in SAS using a random term with a correlation structure.

Random year /sub=plot*rep type=sp(pow)(year2);
Random year /sub=plot*rep*age;



Is there any way of doing this? I prefer to use R so i am trying very
hard to code this in R.


Thanks in advance




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