Variational Inference for Random Graph Models (VIRGMo)

Dependencies

• python>=3.6
• pytorch>=1.4.0: https://pytorch.org
• scipy: https://scipy.org
• numpy: https://www.numpy.org

Installation

To install, run

$ python setup.py install

Source code:

• vi_rg: VI for random graphs (abstract parent class)
• vi_sbm: VI for stochastic block models
• vi_graphon - VI for graphons
• vi_hrg: - VI for hyperbolic random graphs
• utils: util functions
• graph_models: random graph models for network data generation and EdgesDataset for effective training
• distributions: several Pytorch implementations of non-build-in distributions (including the von Mises-Fisher and hyperspherical Uniform distributions from https://github.com/nicola-decao/s-vae-pytorch)

Examples:

1. SBMs on simulated data
2. WSBMs on simulated data
3. SBM on simulated interbank network
4. Karate club network
5. NFL 2009 network
6. World trade network (unweighted)
7. World trade network (weighted)

Feedback

For questions and comments, feel free to contact Iurii Mozzhorin.

License

MIT