Flow Stability, a python package for dynamic community detection in temporal networks.

Speaker: Alexandre Bovet

Website: https://github.com/alexbovet/flow_stability

Abstract: Many networks of interest are changing with time and extracting a simplified view of their evolution in interesting for many network science practitioners. The flow stability method (Bovet et al.) offers a dynamic community detection method based on the modeling of a dynamical process evolving on the temporal network. The Python package “flowstab” implements this method using sparse matrix computations, high parallelization, and Cython optimization to offer fast computations of diffusion flows in temporal networks and clustering to find optimal partitions. We are currently working on the code to offer a more user-friendly interface, add more examples and tutorials, and complement the documentation. We plan to finish this work and submit it to the Journal of Open Source Software before NetSci.

Reference: