r-cran-party binary package in Ubuntu Mantic ppc64el

 A computational toolbox for recursive partitioning.
 The core of the package is ctree(), an implementation of
 conditional inference trees which embed tree-structured
 regression models into a well defined theory of conditional
 inference procedures. This non-parametric class of regression
 trees is applicable to all kinds of regression problems, including
 nominal, ordinal, numeric, censored as well as multivariate response
 variables and arbitrary measurement scales of the covariates.
 Based on conditional inference trees, cforest() provides an
 implementation of Breiman's random forests. The function mob()
 implements an algorithm for recursive partitioning based on
 parametric models (e.g. linear models, GLMs or survival
 regression) employing parameter instability tests for split
 selection. Extensible functionality for visualizing tree-structured
 regression models is available. The methods are described in
 Hothorn et al. (2006) <doi:10.1198/106186006X133933>,
 Zeileis et al. (2008) <doi:10.1198/106186008X319331> and
 Strobl et al. (2007) <doi:10.1186/1471-2105-8-25>.

Source package