multiPIM: Variable Importance Analysis with Population Intervention Models

Performs variable importance analysis using a causal inference approach. This is done by fitting Population Intervention Models. The default is to use a Targeted Maximum Likelihood Estimator (TMLE). The other available estimators are Inverse Probability of Censoring Weighted (IPCW), Double-Robust IPCW (DR-IPCW), and Graphical Computation (G-COMP) estimators. Inference can be obtained from the influence curve (plug-in) or by bootstrapping.

Version: 1.4-1
Depends: lars (≥ 0.9-8), penalized, polspline, rpart
Suggests: parallel
Published: 2014-04-14
Author: Stephan Ritter, Alan Hubbard, Nicholas Jewell
Maintainer: Stephan Ritter <stephanritterRpacks at gmail.com>
License: GPL-2 | GPL-3 [expanded from: GPL (≥ 2)]
URL: http://www.jstatsoft.org/v57/i08/
NeedsCompilation: no
Citation: multiPIM citation info
Materials: ChangeLog
CRAN checks: multiPIM results

Downloads:

Reference manual: multiPIM.pdf
Package source: multiPIM_1.4-1.tar.gz
OS X binary: multiPIM_1.4-1.tgz
Windows binary: multiPIM_1.4-1.zip
Old sources: multiPIM archive