Learn R Programming

policytree (version 1.2.5)

Policy Learning via Doubly Robust Empirical Welfare Maximization over Trees

Description

Learn optimal policies via doubly robust empirical welfare maximization over trees. Given reward estimates, the algorithm finds a rule-based treatment allocation, where the policy takes the form of a shallow decision tree that is globally optimal (or nearly so). Methods are described in Sverdrup, Kanodia, Zhou, Athey, and Wager (2020) , Athey and Wager (2021) , and Zhou, Athey, and Wager (2023) .

Copy Link

Version

Install

install.packages('policytree')

Monthly Downloads

2,511

Version

1.2.5

License

MIT + file LICENSE

Issues

13

Pull Requests

0

Stars

89

Forks

16

Maintainer

Erik Sverdrup

Last Published

August 4th, 2026

Functions in policytree (1.2.5)

double_robust_scores.causal_forest

Matrix \(\Gamma\) of scores for each treatment \(a\)
gen_data_epl

Example data generating process from Policy Learning With Observational Data
conditional_means.causal_forest

Estimate mean rewards \(\mu\) for each treatment \(a\)
make_tree

A utility function for generating random trees for test purposes.
hybrid_policy_tree

Hybrid tree search
multi_causal_forest

(deprecated) One vs. all causal forest for multiple treatment effect estimation
plot.policy_tree

Plot a policy_tree tree object.
gen_data_mapl

Example data generating process from Offline Multi-Action Policy Learning: Generalization and Optimization
predict.policy_tree

Predict method for policy_tree
print.policy_tree

Print a policy_tree object.
policy_tree

Fit a policy with exact tree search
export_graphviz

Export a tree in DOT format. This function generates a GraphViz representation of the tree, which is then written into dot_string.
policytree-package

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

Predict with the above test tree.
create_dot_body

Writes each node information If it is a leaf node: show it in different color, show number of samples, show leaf id If it is a non-leaf node: show its splitting variable and splitting value