multiPIM (1.0-1)

Variable Importance Analysis with Population Intervention Models.

http://www.stat.berkeley.edu/users/sritter/multiPIM/
http://cran.r-project.org/web/packages/multiPIM

Performs variable importance analysis for possibly many exposures of interest and possibly many outcomes of interest. This is done by fitting Population Intervention Models, using either Inverse Probability of Censoring Weighted (IPCW) or Double-Robust IPCW (DR-IPCW) estimators. Inference can be obtained from the influence curve or by bootstrapping.

Maintainer: Stephan Ritter
Author(s): Stephan Ritter <sritter@berkeley.edu>, Alan Hubbard <hubbard@berkeley.edu>, Nicholas Jewell <jewell@berkeley.edu>

License: GPL (>= 3)

Uses: lars, penalized, polspline, randomForest, rpart, rlecuyer, multicore

Released over 8 years ago.