Giuseppe Torrisi et al J. Stat. Mech. (2020) 083501 https://doi.org/10.1088/1742-5468/aba7b0.
See this twitter thread for a short summary of what is in the paper.
This repository contins all the code to reproduce the figures of the paper. This repository requires:
- Python version 3.6 or higher
- Mathematica and wolframscript installed
- jupyter notebooks
Each folder performs a task. Inside each folder, use the following order to execute files:
- *.wls
- *.py
- *ipynb
Also Section_4/macroscopic-cavity should be run first.
-
lib: a folder with interal routines for python scripts
-
Section 3 contains a Mathematica notebook
-
Section 4 contains:
- macroscopic-cavity: this folder compute the macroscopic cavity theory for type 1 networks.
- pruning-aOC; percolation on synthetic network:
- single-instance: microscopic cavity dynamics
- knockout-cascade: single gene removal
-
Section 5 contains: strongly connected component evaluation.
This repository relies heavily on networkx. Some method for direct bipartite graph generation have been extendend. In particular:
- directed biparite random graph generation
- configurational model for directed bipartite graph
Moreover in- and out-component are computed for directed graph.
Go in the lib
folder to know more