Dear R users, I've been trying to find an R package which does the PL estimation of spatial GLMMs especially with the negative binomial model. so it would be something similar to the "proc GLIMMIX" with the PL method in SAS. I've looked up some possible packages related to GLMMs, but it doesn't seem to be anyone using the PL estimation. Thanks for your help! Fei He UCR Statistics -- View this message in context: http://r.789695.n4.nabble.com/Pseudolikelihood-Estimation-of-spatial-GLMM-using-R-tp4633896.html Sent from the R help mailing list archive at Nabble.com.
Prof Brian Ripley
2012-Jun-20 11:09 UTC
[R] Pseudolikelihood Estimation of spatial GLMM using R
On 19/06/2012 22:29, Herrfay wrote:> Dear R users, > > I've been trying to find an R package which does the PL estimation of > spatial GLMMs especially with the negative binomial model. so it would be > something similar to the "proc GLIMMIX" with the PL method in SAS. I've > looked up some possible packages related to GLMMs, but it doesn't seem to be > anyone using the PL estimation.Well, have you looked at MASS? You didn't tell us what 'PL' is, but from memory GLIMMIX does something close to PQL as done by MASS::glmmPQL. And AFAIR that does support negative-binomial GLMs (which of necessity have fixed theta and as parametrized as in MASS, the book).> > Thanks for your help! > > Fei He > UCR Statistics >-- Brian D. Ripley, ripley at stats.ox.ac.uk Professor of Applied Statistics, http://www.stats.ox.ac.uk/~ripley/ University of Oxford, Tel: +44 1865 272861 (self) 1 South Parks Road, +44 1865 272866 (PA) Oxford OX1 3TG, UK Fax: +44 1865 272595