SPLab-aviyente / GRASCale

Simultaneous graph signal clustering and multiview graph learning

GRASCale

Python implementation of simultaneous GRAph Signal Clustering And graph LEarning (GRASCale) algorithm presented in [1].

Installation

Once you download the repo, go to the repo directory and start a terminal. First, create an environment and then install the required packages listed in requirements.txt. This can be done as follows for conda:

conda create -n grascale python=3.10
conda activate grascale

This will create a conda environment. To install the requirements:

pip install -r requirements.txt

Usage

Please see experiment1.py under scripts folder for an illustration about how to use the code. The script includes data generation process of experiment 1 from the paper and shows how to run GRASCale on the generated simulated data. To run the script first start a terminal on the repo directory, then:

conda activate grascale
python scripts/experiment1.py

It will print F1 score, NMI and density of the learned graphs associated with clusters. The script for the experiment 2 of the paper will also be published soon.

References

• [1] Karaaslanli, Abdullah, and Selin Aviyente. "Simultaneous Graph Signal Clustering and Graph Learning." International Conference on Machine Learning. PMLR, 2022.