cold: An R Package for the Analysis of Count Longitudinal Data

This paper describes the R package cold for analysis of count longitudinal data. In this marginal and random effects models are considered. both cases estimation is via maximization exact likelihood serial dependence among observations assumed to be Markovian type referred as integer-valued autoregressive order one process. For adaptive Gaussian quadrature Monte Carlo methods used compute integrals whose dimension depends on structure effects. written partly in language, Fortran 77, interfaced through built following S4 formulation methods.