PASCO

This repository contains the code to run PASCO (PArallel Structured COarsening): an overlay to speed up graph clustering algorithms. It is associated to the paper: soon to come.

Installation

To run PASCO, follow the instruction below.
In a terminal, go into the PASCO folder, and run the following commands :

• conda create -n pasco -c conda-forge graph-tool
• conda activate pasco
• pip install matplotlib networkx scikit-learn POT pygsp scikit-network infomap leidenalg pandas
• pip install -e .

The libraries graph-tool and infomap are sometimes problematic to install. If you do not intend to use neither MDL nor infomap as clustering algorithm, you can set a simpler conda environment. To do so, run the following command from the PASCO folder.

• conda create -n pasco python
• conda activate pasco
• pip install matplotlib networkx scikit-learn POT pygsp scikit-network leidenalg pandas
• pip install -e .

Tutorial

Once the conda environment is installed, you can check our tutorial notebook tutorial.ipynb to see how to simply use PASCO (with clustering algorithm already implemented, or your own clustering algorithm).