r-cran-rms binary package in Ubuntu Xenial ppc64el

 Regression modeling, testing, estimation, validation, graphics,
 prediction, and typesetting by storing enhanced model design
 attributes in the fit. rms is a collection of 229 functions that
 assist with and streamline modeling. It also contains functions for
 binary and ordinal logistic regression models and the Buckley-James
 multiple regression model for right-censored responses, and implements
 penalized maximum likelihood estimation for logistic and ordinary
 linear models. rms works with almost any regression model, but it
 was especially written to work with binary or ordinal logistic
 regression, Cox regression, accelerated failure time models,
 ordinary linear models, the Buckley-James model, generalized least
 squares for serially or spatially correlated observations, generalized
 linear models, and quantile regression.
 .
 See Frank Harrell (2001), Regression Modeling Strategies, Springer
 Series in Statistics, as well as http://biostat.mc.vanderbilt.edu/Rrms.

Source package