This repository contains the code for the corresponding paper "SoK: Descriptive Statistics Under Local Differential Privacy" accepted at PETS 2025.
An eprint of the paper is available at https://eprint.iacr.org/2024/1464.
To work with the project you need to install poetry and have a working python environment. The project was tested with python 3.9 and 3.10.
See https://github.com/mad-lab-fau/mad-cookiecutter/blob/main/python-setup-tips.md#global-tooling for a convenient way of installing poetry globally without polluting your global python environment based on pipx.
Afterwards run:
poetry install
All dependencies are manged using poetry.
Poetry will automatically create a new venv for the project, when you run poetry install.
See ARTIFACT-EVALUATION.md for instructions on how to reproduce the results of the paper.
We have uploaded the simulation results used for producing the figures in the paper to Zenodo. You can find the data here: https://doi.org/10.5281/zenodo.13683977
Unzip the results into the experiments folder (experiments/results and experiments/results_grouped) to reproduce
the figures using the scripts in the plotting folder.