Sample menu:



This is an R package based on C code that implements a Bayesian selection or switching model for a multinomial probit.  This package borrows heavily from the MNP package of Kosuke Imai and David A. van Dyk.  Source and compiled versions along with documentation are available here.


This is an R package that implements a Bayesian quantile regression model that assumes a factor structure for part of the design matrix.  The package is available here.  See also the accompanying paper under the Research tab.

Imputation via sequential regression trees

Here is a basic R implementation of multiple imputation via sequential trees.  This code relies on the tree package of Brian Ripley to do the heavy lifting.  This code does not detect dependent variables automatically, so users should make sure that no algebraic equalities exist between columns of the data set.  See Burgette and Reiter (2010) "Multiple imputation for missing data via sequential regression trees" for details.

See here for a toy example that uses the code.