This package provides a variety of tools for nonparametric estimation of causal effects across a wide range of settings. The methods are based on the theory of influence functions, and can incorporate flexible machine learning and high-dimensional regression tools, while still yielding inference in the form of confidence intervals and hypothesis tests. Many of the methods are doubly robust.
To install and load this package in R from GitHub, run the following commands:
install.packages("devtools")
library(devtools)
install_github("ehkennedy/npcausal")
library(npcausal)