Package: policytree 1.2.3

Erik Sverdrup

policytree: Policy Learning via Doubly Robust Empirical Welfare Maximization over Trees

Learn optimal policies via doubly robust empirical welfare maximization over trees. Given doubly robust reward estimates, this package finds a rule-based treatment prescription policy, where the policy takes the form of a shallow decision tree that is globally (or close to) optimal.

Authors:Erik Sverdrup [aut, cre], Ayush Kanodia [aut], Zhengyuan Zhou [aut], Susan Athey [aut], Stefan Wager [aut]

policytree_1.2.3.tar.gz
policytree_1.2.3.tar.gz(r-4.5-noble)policytree_1.2.3.tar.gz(r-4.4-noble)
policytree_1.2.3.tgz(r-4.4-emscripten)policytree_1.2.3.tgz(r-4.3-emscripten)
policytree.pdf |policytree.html✨
policytree/json (API)

# Install 'policytree' in R:
install.packages('policytree', repos = 'https://cloud.r-project.org')

Bug tracker:https://github.com/grf-labs/policytree/issues13 issues

Uses libs:
  • c++– GNU Standard C++ Library v3

On CRAN:

Conda:r-policytree-1.2.3(2025-03-25)

cpp

1.95 score 3 packages 700 downloads 7 exports 10 dependencies

Last updated 9 months agofrom:f355509d8a. Checks:3 OK. Indexed: yes.

TargetResultLatest binary
Doc / VignettesOKMar 11 2025
R-4.5-linux-x86_64OKMar 11 2025
R-4.4-linux-x86_64OKMar 11 2025

Exports:conditional_meansdouble_robust_scoresgen_data_eplgen_data_maplhybrid_policy_treemulti_causal_forestpolicy_tree

Dependencies:BHDiceKriginggrflatticelmtestMatrixRcppRcppEigensandwichzoo

Citation

To cite package ‘policytree’ in publications use:

Sverdrup E, Kanodia A, Zhou Z, Athey S, Wager S (2024). policytree: Policy Learning via Doubly Robust Empirical Welfare Maximization over Trees. R package version 1.2.3, https://CRAN.R-project.org/package=policytree.

Corresponding BibTeX entry: 

  @Manual{,
    title = {policytree: Policy Learning via Doubly Robust Empirical
      Welfare Maximization over Trees},
    author = {Erik Sverdrup and Ayush Kanodia and Zhengyuan Zhou and
      Susan Athey and Stefan Wager},
    year = {2024},
    note = {R package version 1.2.3},
    url = {https://CRAN.R-project.org/package=policytree},
  }

Readme and manuals

Help Manual

Help pageTopics
Estimate mean rewards mu for each treatment aconditional_means conditional_means.causal_forest conditional_means.causal_survival_forest conditional_means.instrumental_forest conditional_means.multi_arm_causal_forest
Matrix Gamma of scores for each treatment adouble_robust_scores double_robust_scores.causal_forest double_robust_scores.causal_survival_forest double_robust_scores.instrumental_forest double_robust_scores.multi_arm_causal_forest
Example data generating process from Policy Learning With Observational Datagen_data_epl
Example data generating process from Offline Multi-Action Policy Learning: Generalization and Optimizationgen_data_mapl
Hybrid tree searchhybrid_policy_tree
(deprecated) One vs. all causal forest for multiple treatment effect estimationmulti_causal_forest
Plot a policy_tree tree object.plot.policy_tree
Fit a policy with exact tree searchpolicy_tree
Predict method for policy_treepredict.policy_tree
Print a policy_tree object.print.policy_tree