https://github.com/snap-stanford/graphwave

https://github.com/snap-stanford/graphwave

Last synced: 3 months ago
JSON representation

• Host: GitHub
• URL: https://github.com/snap-stanford/graphwave
• Owner: snap-stanford
• Created: 2017-11-03T03:45:27.000Z (about 7 years ago)
• Default Branch: master
• Last Pushed: 2018-12-03T17:11:11.000Z (almost 6 years ago)
• Last Synced: 2024-05-21T17:23:34.733Z (6 months ago)
• Language: Jupyter Notebook
• Size: 189 MB
• Stars: 167
• Watchers: 54
• Forks: 53
• Open Issues: 13
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

        
# GraphWave

## Spectral Wavelets for learning structural signatures in complex networks

This folder contains the code for GraphWave, an algorithm for computing structural signatures for nodes in a network using heat spectral wavelets. 

This code folder is organized as follows:

+ __shapes/__: contains the functions for generating (more or less) regular graphs and shapes

+ __performance_evaluation/__: functions computing different metrics for assessing the quality of the embeddings generated by GraphWave

+ __test\_perturbation\_synthetic/__: set of ipython notebooks for running the synthetic

experiments described in the paper.

+ __utils/__: set of  helper functions

+ files:

    + __characteristic\_functions.py__: functions for computing the characteristic functions.

    + __graphwave.py__: wrapper function for computing the embeddings.

 

 

### Prerequisites

GraphWave was written for __Python 2.7__ and requires the installation of the following Python libraries:

+ __networkx__: allows easy manipulation and plotting of graph objects (more information in the [Networkx website](https://networkx.github.io)).

+ __pyemd__: module for computing Earth Mover distances (for trying out other distances between diffusion distributions. More information in the [pyemd website](https://github.com/wmayner/pyemd))

Also, need standard packages: __scipy, sklearn, seaborn__ for analyzing and plotting results.

Note: heat diffusion scaling wavelets can also be computed with the Graph Signal Processing toolbox  __pygsp__ (accessible through the  [EPFL website](https://pygsp.readthedocs.io/en/stable/)  ), which, beyond structural similarities, has plenty of extremely useful features for handling signals on graphs.

 

 

## Running Graphwave

A full example on how to use GraphWave is provided in the ''Using GraphWave.ipynb" of this directory.

In a nutshell:

+ input: nx (or pygsp) Graph structure

+ compute the heat wavelets

+ embed the distributions in Euclidean space using the characteristic function

+ output: signatures, which can be used in one's favorite Machine Learning framework.

For a given graph __G__ (of type pygsp or nx),GraphWave structural signatures can be simply 

computing by calling:

```

>from graphwave import graphwave

>chi,heat_print, taus=graphwave_alg(G, 'automatic', verbose=False)

```

## Authors

* Anonymous at this time

## Acknowledgements

We would like to thank the authors of [struc2vec](https://arxiv.org/abs/1704.03165) for the [open access of the implementation of their method](https://github.com/leoribeiro/struc2vec), as well as [Lab41](http://lab41.github.io/blog/2014/12/18/rolx-discovering-individuals-roles-in-a-social-network/) for its open-access implementation of [RolX](https://dl.acm.org/citation.cfm?id=2339723).

## License

This project is licensed under the MIT License - see the [LICENSE.md](LICENSE.md) file for details

=======

# graphwave

>>>>>>> 5d0c25e5cc6baee37b445b1c0d232dc0d7e43e52