Overview

The binomialDP package makes it easy to perform differentially private inference for a population proportion in R. The package offers a quick solution to find differentially private uniformly most powerful (UMP) tests, private p-values, and private confidence intervals. Also included in the package are the CDF function and the random deviate generator of the Truncated-Uniform-Laplace (Tulap) distribution.

Installation

To install the package, run one of the following:

## The latest official version from CRAN
install.packages("binomialDP")
library(binomialDP)

## The latest development version from Github
install.packages("devtools")
devtools::install_github("tranntran/binomialDP")

Examples

Please see Get Started for the package demonstration and working examples.

Citation

If you use this package for academic purposes, please cite this package along with the paper on differentially private inference for binomial data:

Jordan Awan, Aleksandra Slavkovic and Tran Tran (2020). binomialDP: Differentially Private Inference for Binomial Data. R package version 1.0.0.

Awan, Jordan Alexander, and Aleksandra Slavkovic. “Differentially Private Inference for Binomial Data.” Journal of Privacy and Confidentiality 10, no. 1 (2020).