Tidy Methods for Bayesian Treatment Effect Models

Functions for extracting tidy data from Bayesian treatment effect models, in particular BART, but extensions are possible. Functionality includes extracting tidy posterior summaries as in 'tidybayes' < https://github.com/mjskay/tidybayes>, estimating (average) treatment effects, common support calculations, and plotting useful summaries of these.

Tidy methods for Bayesian treatment effect models

tidytreatment is an R package that provides functions for extracting tidy data from Bayesian treatment effect models, estimating treatment effects, and plotting useful summaries of these. This package closely follows the output style from the tidybayes R package in order to use some functions provided by tidybayes.

The package currently supports the following models:

• BART: see CRAN
• bartMachine: see CRAN.
• bcf: see CRAN (in development, see branch bcf-hold on github)

in addition to models supported by tidybayes.

See this HTML vignette or vignette("use-tidytreatment-BART") for examples of usage.

How to install

CRAN

Install the release version from CRAN with install.packages("tidytreament").

Github

To install the latest development version:

1. Make sure at least one of the above model fitting packages is installed.
2. In R make sure remotes is installed. Install with install.packages("remotes").
3. Run remotes::install_github("bonStats/tidytreatment")