update_beta

A character formula with target variable and predictor variables

formula

The numeric value of the specific cluster to be updated

A character value corresponding to id column name

col_id

A matrix of dimension : rows number is the number of cluster K, columns number is the number of distinct levels id

An inverse matrix representing the W matrix in the beta equation for the M step

w_inv

A list of design matrices filter by id levels

matrix_id

M-step : update of beta parameters

Provides a clustering method of non-gaussian longitudinal data with a mixture of generalized linear models. The longitudinal data should be defined as repeated observations for each individual. The number of observations for each individual can be different. In runEM(), an expectation-maximization algorithm is developed for both binomial and longitudinal data mixture model.

update_beta: M-step : update of beta parameters

Description

Usage

Arguments

Value