Graph Wavelets for Multiscale Community Mining

Python implementation of the method proposed in the paper:

Tremblay, Nicolas, and Pierre Borgnat. "Graph wavelets for multiscale community mining." IEEE Transactions on Signal Processing 62.20 (2014): 5227-5239.

Structure

• Final report: ./report.pdf
• All the necessary code is provided in the directory ./src
• Examples of the method application are provided in jupyter-notebooks:
  • experiment on a small Sales-Pardo graph with visualization: ./example_sp_graph_toy.ipynb
  • experiment on a bigger Sales-Pardo graph reproducing results provided in the original paper: ./example_sp_graph.ipynb
  • experiment on a non-uniform swiss roll manifold: ./example_suiss_roll.ipynb
• All plots and graphs' visualizations are saved to the directory ./plots