PyDP

A Python wrapper for Google's C++ Differential Privacy library that enables epsilon-private data analysis.

PyDP provides a high-level Python interface to the core algorithms used in Google's Differential Privacy project. It allows data scientists to calculate aggregate statistics (like mean, variance, and count) while maintaining mathematical privacy guarantees for individuals in the dataset. By leveraging the OpenMined ecosystem, PyDP handles the complex noise injection and sensitivity calculations required to prevent data leakage in sensitive environments.

https://github.com/OpenMined/PyDP