Binomial glm in asreml-r

In asreml-r, how do I specify the model for a binomial glm?

I think SAS uses something like "model x/n = fixedeffects" and I think
lme4 uses "cbind(x,n-x)~fixedeffects".

I tried variations on those syntaxes, but the only thing I could get
to run in asreml is to define a new variable "proportion = x/n" and
then
asreml(proportion ~ fixedeffects, random, data,
family=asreml.binomial(link="logit"))

That doesn't use the information about "n" anywhere that I can see, so
I'm wondering if I have the right syntax.

Any pointers would be appreciated.

Kevin Wright
Received on Tue Dec 21 2007 - 16:15:08 EST