The equiboots library is a fairness-aware model evaluation toolkit designed to audit performance disparities across demographic groups. It provides robust, bootstrapped metrics for binary, multi-class, and multi-label classification, as well as regression models. The library supports group-wise performance slicing, fairness diagnostics, and customizable visualizations to support equitable AI/ML development.
equiboots is particularly useful in clinical, social, and policy domains where transparency, bias mitigation, and outcome fairness are critical for responsible deployment.
Before installing equiboots, ensure your system meets the following requirements:
equiboots requires Python 3.7.4 or higher. Specific dependency versions vary depending on your Python version.
The following dependencies will be automatically installed with equiboots:
matplotlib>=3.5.3, <=3.10.1
numpy>=1.21.6, <=2.2.4
pandas>=1.3.5, <=2.2.3
scikit-learn>=1.0.2, <=1.5.2
scipy>=1.8.0, <=1.15.2
seaborn>=0.11.2, <=0.13.2
tqdm>=4.66.4, <=4.67.1
You can install equiboots directly from PyPI:
pip install equibootsComing soon.
equiboots is distributed under the Apache License. See LICENSE for more information.
If you use model_tuner in your research or projects, please consider citing it.
@software{shpaner_2025_15086941,
author = {Shpaner, Leonid and
Funnell, Arthur and
Rahrooh, Al and
Beam, Colin and
Petousis, Panayiotis},
title = {EquiBoots},
month = mar,
year = 2025,
publisher = {Zenodo},
version = {0.0.1a1},
doi = {10.5281/zenodo.15086941},
url = {https://doi.org/10.5281/zenodo.15086941}
}If you have any questions or issues with equiboots, please open an issue on this GitHub repository.
This work was supported by the UCLA Medical Informatics Institute (MII) and the Clinical and Translational Science Institute (CTSI). Special thanks to Dr. Alex Bui for his invaluable guidance and support, and to Panayiotis Petousis, PhD, for his contributions to this codebase.