Dear all,
I'm using R function "glmmPQL" in "MASS" package for generalized linear mixed
model considering the temporal correlations in random effect. There are 1825
observations in my data, in which the random effect is called "Date", and there
are five levels in "Date", each repeats 365 times.
When I tried
fit.model1=glmmPQL(y~f1+f2+f3+f4+f5+f6+f7+f8+f9+f10+f11+f12+f13+f14+f15+f16+f17+f18+f19+f20+f21+f22+f23+f24,
family=poisson,random=~1|Date,data=mydata,correlation=corCompSymm(value=0.2,form=~1|Date)),
the model was fitted well. But because of my particular interest, I need to
specify the correlation structure by myself, so I tried the following code,
fit.model2=glmmPQL(y~f1+f2+f3+f4+f5+f6+f7+f8+f9+f10+f11+f12+f13+f14+f15+f16+f17+f18+f19+f20+f21+f22+f23+f24,family=poisson,random=~1|Date,data=mydata,correlation=corSymm(value=B[lower.tri(B)],form=~1|Date)),
where B is a 365*365 correlation matrix that's specified by me. Then there's
an error message "Error in vector("double", length) : vector size specified is
too large". Even I wrote B exactly the same as the one used in model1, i.e.
diagonal elements 1, off-diagonal elements 0.2, the same error message shows.
Is this error something inherited from the "glmmPQL" function that I couldn't
change, or something wrong I made so that I could make certain modifications?
Thanks so much in advance for any kind help!
--
Best,
Vicky
[[alternative HTML version deleted]]
______________________________________________
R-help@r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.
