[R] R not responding for nested glm model

[Sundar Dorai-Raj]

Yuval Sapir said the following on 10/12/2006 1:08 PM:
> Hi,
> I'm trying to perform a glm model on count data (poisson distribution of 
> the errors) where data are nested.
> glmmodel<-glm(y~x/z,poisson)
> x and z are factors, z nested within x, y is count data.
> In that point the R just "stuck" and not respond anymore. I tried
> glmmodel<-glm(y~x,poisson)
> and there were no problems. Trying
> glmmodel<-glm(y~x/z)
> gave the same no response. Similar problem with continuous data (normal 
> distribution of the errors).
> I am using R 2.3.1 on Windows XP
> Thanks
> Yuval
> 

Please see the signature "and provide commented, minimal, 
self-contained, reproducible code." Namely, what is 'x' and 'z'? My 
guess is ~x/z produces so large a model matrix you are approaching your 
limits of memory. But, there's no way to tell without more information.

HTH,

--sundar