I´m looking for a package in R that can deal with glmm´s and spatial and temporal autocorrelation, the only one I´ve found is the glmmPQL function in MASS, but it has been heavily criticized and does not provide AIC values. Also, in this package, the model has to be mixed (i.e., with random factors), and I´d like to find one where you don´t necessarily have to include random factors. Thanks!
Alexis,
You could use R, R2Winbugs, and geobugs to perform this kind of analysis. Also, check out Roger Bivand's book on Applied Spatial analysis using R.http://www.asdar-book.org/
Cheers,
Robert
Hi Alexis,
'spdep' package might be of your interest. I sometimes use spautolm function to fit autoregressive models. I also suggest to consult Bivand's book.
moreover, GAM (general additive models) can also account for spatial and temporal autocorrelation, see package 'mgcv'.
cheers,
Adam
Hi Alexis,
look at :Mixed Effects Models and Extensions in Ecology with R (200)
by Alain Zuur,Elena N. Ieno,Neil Walker,Anatoly A. Saveliev,Graham M. Smith
It is very helpful and easy to apply.
Good luck,
Carolina
Alexis,
Check the lme4 package:
lmer(y~x+(1|random.effect))
returns AIC,BIC,log likelihood and deviance
Hi Alexis,
Check out MCMCglmm package by Jarrod Hadfield. Specifically the ginverse argument.
- Badges
- Science topic
- Similar topics
- Mathematics
- Statistics
- Statistical Software
- R Statistical Package