loglikcs: The loglikelihood function for two independent overdispersed...

In IvonneMartin/CombinedMultinomial:

Description

The loglikelihood function for two independent overdispersed multivariate count outcome

Usage

loglikcs(pars, dataset, Des, model)

Arguments

pars	The vector of fixed effect parameters followed by the overdispersion parameter theta.
dataset	A list of two elements, where the first element is the multivariate outcome and the second element is the matrix of categorical covariates.
Des	A list of two elements, where the first element is the left hand side of the matrix equation of the loglinear model and the second element is the design matrix for the regression covariates.
model	A model for the multivariate count outcome. Either DMM, CNBM or UNBM.

Value

The estimated regression parameters of regression model, the loglikelihood values, and the Hessian matrix of the estimated parameters. This loglikelihood function is used to estimate parameters of the regression of multivariate count outcome on categorical covariate in the repeated measurement setting in which the correlation between observations at two different time point is ignored ($\sigma^2_u = 0$).

IvonneMartin/CombinedMultinomial documentation built on Dec. 17, 2021, 11:32 p.m.