Finite-time convergence to an ϵ-efficient Nash equilibrium in potential games

This repository accompanies the paper:

Maddux, A., Ouhamma, R., Catic, H., and Kamgarpour, M., "Finite-time convergence to an ϵ-efficient Nash equilibrium in potential games", IEEE Transactions on Control of Network Systems (2026).

Dependencies

The project was tested on Windows 11 and is written in Python 3.11. Dependencies are managed via Poetry (see pyproject.toml).

Setup

1. Install Git LFS. Preprocessed experiment data are tracked using Git LFS.

   git lfs install
   

2. Clone the repository:

   git clone https://github.com/hanacatic/potentialgames.git
   cd potentialgames
   

3. Install dependencies:

   pipx install poetry
   poetry install
   

Usage

Scripts and modules for running experiments are provided in the scripts/ and functions to visualise results are provided in the potentialgames/utils/ directories. See the code and docstrings for details.

Quick example

```
poetry run python main.py
```