Partially Linear Mixed Effects Model
Description
This package fits the partially linear mixed effects model
(semiparametric random intercept model) using kernel
regression, without distributional assumptions for the random
terms. Estimation procedure is an iterative generalized least
squares type. A nonparametric heteroskedastic variance function
is allowed for the regression error. Bootstrap resampling is
provided for inference. The package implements bandwidth
selection by an alternative cross validation for correlated
data.