glmLC {RPMM}

Weighted GLM for latent class covariates

Description

Wrapper for glm function to incorporate weights corresponding to latent classes

Usage

glmLC(y,W,family=quasibinomial(),eps=1E-8,Z=NULL)

Arguments

y

outcome

W

weight matrix (rows=cases, # rows = length of y)

family

glm family (default = quasibinomial for logistic regression)

eps

threshold below which to delete pseudo-subject corresponding to a specific weight

Z

matrix of additional covariates

Details

This function is a wrapper for glm to incorporate weights corresponding to latent classes (e.g. from an RPMM prediction)

Value

a glm object