This package implements the Accelerated Bayesian Causal Forests approach for conditional average treatment effect estimation; the manuscript is available here. This approach builds on the methodology behind Bayesian Causal Forests outlined in Hahn et al. (2020) and incorporates several improvements to Bayesian Additive Regression Trees implemented by He et al. (2019).
This package is based on the source code of the XBART package and was originally developed as a branch of that repository.
To install the package, run from the R console:
library(devtools)
install_github("socket778/XBCF")
To install XBCF from PyPI use pip install xbcausalforest