[R-sig-ME] Modeling Residuals vs. Modeling Random-effects Covariances

Simon Harmel @|m@h@rme| @end|ng |rom gm@||@com
Mon Jul 27 05:48:25 CEST 2020

Dear all,

I had a quick question. The variance-covariance structure from the
"glmmTMB" package and the "nlme" package are not the same thing?

Is there any other package capable of modeling the Level-1 residual
variance-covariance structure like in "nlme"? *-- Many Thanks, Simon *

*glmmTMB::glmmTMB(y ~ times + ar1(times | subjects), data = data) ## DON'T
RUN   nlme::lme    (y ~ times, random = ~ times | subjects, correlation =
corAR1(), data = data) ## DON'T RUN *

	[[alternative HTML version deleted]]

More information about the R-sig-mixed-models mailing list