BayesGESM: Bayesian Analysis of Generalized Elliptical Semiparametric Models and Flexible Measurement Error Models

This package allows to perform the statistical inference based on the Bayesian approach for regression models under the assumption that independent additive errors follow normal, Student-t, slash, contaminated normal, Laplace or symmetric hyperbolic distributions, i.e., additive errors follow a scale mixtures of normal distributions. The regression models considered in this package are: (i) Generalized elliptical semiparametric models, where both location and dispersion parameters of the response variable distribution include nonparametric additive components described by using B-splines; and (ii) Flexible measurement error models, which admit explanatory variables with and without measurement additive errors as well as the presence of a nonparametric component approximated by using B-splines.

Downloads: