The paper associated with this package is on arXiv: https://arxiv.org/abs/2310.12358.
The R package causalBETA is an MCMC-based implementation of piecewise exponential model for survival data, and it uses G-computation to conduct causal inference for binary treatment on survival data. We provide generic plot functions to visualize estimates easily and allow users to run MCMC diagnostics by coda package.
install using devtools package
## install.packages(devtools) ## make sure to have devtools installed
devtools::install_github("RuBBiT-hj/causalBETA")
library(causalBETA)
The following packages are required for causalBETA:
cmdstanr≥ 0.5.3codametssurvivalLaplacesDemonstatsrlanggrDevicesgraphics
This information is also listed in DESCRIPTION.
The paper associated with this package contains the statistical details of the model as well as a detailed walk-through demonstration.
Help documentation in R is also available, and it has example code for each function. After installing the package and loading it with library(causalBETA), use ? to access help documentation for specific functions. For example, for the two main functions:
?causalBETA::bayeshaz # Construct the MCMC-based Bayesian piece-wise exponential model
?causalBETA::bayesgcomp # Apply G-computation to obtain posterior draws of the average difference in survial probabilities between two treatments
The code for demostration in the paper is available in the folder demo_code.
If you encounter any bugs or have feature requests, please open an issue on GitHub.
Please use the following LaTex cite as follows:
@misc{ji2023causalbeta,
title={causalBETA: An R Package for Bayesian Semiparametric Causal Inference with Event-Time Outcomes},
author={Han Ji and Arman Oganisian},
year={2023},
eprint={2310.12358},
archivePrefix={arXiv},
primaryClass={stat.ME}
}
The corresponding package author are Han Ji (email: han_ji@brown.edu) and Arman Oganisian (email: arman_oganisian@brown.edu).