Displaying 1 result from an estimated 1 matches for "blockxvar1".
2003 May 20
0
Problem on model simplification with glmmPQL
..."h" "i" "j" "k" "l" "m"
> levels(Block)
[1] "B1" "B2" "B3"
My general model in aov is:
aov(Yvar~Xvar1*Xvar2+Error(Block/Xvar1/Xvar2))
In glmmPQL I make this:
create the nested high level variable:
> BlockXvar1 <- as.factor(paste(Block,Xvar1))
> levels(BlockXvar1)
[1] "B1 A" "B1 B" "B2 A" "B2 B" "B3 A" "B3 B"
Make the model
> m1 <-
glmmPQL(Yvar~Xvar1*Xvar2+Block,random=~1|BlockXvar1/Xvar2,family=poisson)
iteration 1
iteration 2...