Alejandro Jara

Associate Professor

Department of Statistics

Pontificia Universidad Católica de Chile

 

R package for Bayesian Non- and Semi-parametric Analysis (DPpackage 1.1-6) (Windows Binary) (MacOS X Binary) (Source) (Manual)

- Author: Alejandro Jara.

- Description: This package contains functions to perform inference via simulation from the posterior distributions for Bayesian nonparametric and semiparametric models. Although the name of the package was motivated by the Dirichlet Process prior, the package considers and will consider other priors on functional spaces. So far, DPpackage includes models considering Dirichlet Processes, Dependent Dirichlet Processes, Dependent Poisson- Dirichlet Processes, Hierarchical Dirichlet Processes, Polya Trees, Linear Dependent Tailfree Processes, Mixtures of Triangular distributions, Random Bernstein polynomials priors and Dependent Bernstein Polynomials. The package also includes models considering Penalized B-Splines. Currently the package includes semiparametric models for marginal and conditional density estimation, ROC curve analysis, interval censored data, binary regression models, generalized linear mixed models, IRT type models, and generalized additive models. The package also contains functions to compute Pseudo-Bayes factors for model comparison, and to elicitate the precision parameter of the Dirichlet Process. To maximize computational efficiency, the actual sampling for each model is done in compiled FORTRAN. The functions return objects which can be subsequently analyzed with functions provided in the coda package.

- Contributors: Timothy Hanson, Fernando Quintana, Peter Mueller and Gary L. Rosner.

- Awards: A previous version of this package won a John M. Chambers Statistical Software Award (2008) from the Statistical Computing Section of the American Statistical Association.


R package for to fit Conditionally Specified Logistic Regression Models (cslogistic 0.1-2) (Windows Binary) (MacOS X Binary) (Source) (Manual)

- Authors: Alejandro Jara and Maria Jose Garcia-Zattera.

- Description: This package contains functions for Bayesian (BayesCslogistic) and Likelihood (MleCslogistic) analysis of conditionally specified logistic regression models. All the computations are done in compiled FORTRAN. 'BayesCslogistic' return mcmc objects which can be subsequently analyzed with functions provided in the coda package.