The penalized least squares (PLS) and penalized iteratively reweighted least squares (PIRLS) algorithms from lme4, but written entirely in pure R. The purposes of this package are:
-
clarify how PLS and PIRLS work without having to read through C++ code,
-
as a sandbox for trying out modified versions of PLS and PIRLS,
-
as a companion to manuscripts describing the lmer and glmer functions.
Note: Currently, this branch of lme4pureR depends on the flexLambda branch of lme4.