The CausalExpAnalysis R package contains functions to perform causal and experiment analysis, i.e., analysis for the causal inference of both experiments and observational studies, within a single framework, with a unifying theme of evaluating covariate balance. Three kinds of samples can be created—unadjusted, weighted, and matched—and two kinds of estimators of treatment effects—an analysis of variance (ANOVA) estimator, i.e., a simple difference-in-means estimator, and an analysis of covariance (ANCOVA) estimator, i.e., a regression adjustment estimator, with either equal or unequal slopes assumption for the covariates. Contains helper functions for statistical inference and modeling, visualization of results, and data manipulation.
install.packages("devtools")
library(devtools)
devtools::install_github("google/CausalExpAnalysis")
library(CausalExpAnalysis)
This is not an officially supported Google product.