A quirk of nlme::varIdent – James E. Pustejovsky

Wherein I document how to control the reference level in models that use nlme::varIdent() for specifying a variance structure.

James E. Pustejovsky
@AbuMirchi @damiacer Perhaps they’d like to fund development of an open source #NLME #Pharmacometrics model development tool

@statstas @GaborBekes

Genuine question: Aren't you able to use weights with #nlme or #lme4 in R? Or are they doing something different with the weights compared with Stata?

Or is this one of those things where clustering and weights cause R to explode?

I'm on fosstodon.org because I do a lot of #rstats development, most notably as a member of the #nlmixr2 development team. nlmixr2 is a set of packages - let's call it the #mixrverse - for R that provides an #OpenSource alternative for nonlinear mixed-effects (#NLME) model development, which are the core of most #Pharmacometrics workflows (amongst others).