This package accompanies the paper "Challenges in Obtaining Valid Causal Effect Estimates with Machine Learning Algorithms," by Ashley Naimi, Alan Mishler, and Edward Kennedy (https://arxiv.org/abs/1711.07137).

It implements simulations and visualizations to illustrate the estimation of the Average Treatment Effect (aka Average Causal Effect) via singly and doubly robust estimators, using parametric and nonparametric estimators of the nuisance propensity scores and outcome regressions.

This package can be installed via devtools::install_github("amishler/nonparametricDoublyRobust").

This package was inspired by ainaimi/NPDR.