fastpolicytree: Constructs Policy Trees from Covariate and Reward Data

Constructs optimal policy trees which provide a rule-based treatment prescription policy. Input is covariate and reward data, where, typically, the rewards will be doubly robust reward estimates. This package aims to construct optimal policy trees more quickly than the existing 'policytree' package and is intended to be used alongside that package. For more details see Cussens, Hatamyar, Shah and Kreif (2025) <doi:10.48550/arXiv.2506.15435>.

Version:	1.0
Imports:	Rcpp (≥ 1.0.7)
LinkingTo:	Rcpp
Suggests:	policytree
Published:	2025-06-24
DOI:	10.32614/CRAN.package.fastpolicytree
Author:	James Cussens [aut, cre], Julia Hatamyar [ctb], Vishalie Shah [ctb], University of Bristol [cph], MRC [fnd]
Maintainer:	James Cussens <james.cussens at bristol.ac.uk>
License:	GPL (≥ 3)
URL:	https://github.com/jcussens/tailoring
NeedsCompilation:	yes
CRAN checks:	fastpolicytree results

Documentation:

Reference manual:

fastpolicytree.html , fastpolicytree.pdf

Downloads:

Package source:	fastpolicytree_1.0.tar.gz
Windows binaries:	r-devel: fastpolicytree_1.0.zip, r-release: fastpolicytree_1.0.zip, r-oldrel: fastpolicytree_1.0.zip
macOS binaries:	r-release (arm64): fastpolicytree_1.0.tgz, r-oldrel (arm64): fastpolicytree_1.0.tgz, r-release (x86_64): fastpolicytree_1.0.tgz, r-oldrel (x86_64): fastpolicytree_1.0.tgz

Linking:

Please use the canonical form https://CRAN.R-project.org/package=fastpolicytree to link to this page.