The computer generation of bivariate binomial variables with given marginals and correlation

Abstract

In this paper simple mixture models are proposed for generating bivariate binomial variables when the marginal distributions and correlation coefficient are specified. The methods of Loukas and Kemp (1986, Commun. Statist. B) are also considered and comparative timings are given. Analogous models for the bivariate negative binomial and gamma distributions are also examined.

Keywords:

• Mixture
• conditional distribution
• common elements
• simulation
• varying dependence
• bivariate negative binomial
• bivariate gamma