epydemic

epydemic is a library for simulating epidemic (and other) processes over complex networks (and other combinatorial structures).

Network science uses epidemic spreading processes a lot, both for the obvious application of modelling diseases, but also for things that are mathematically similar such as computer viruses and rumour-spreading. Some simulations use discrete-time approaches where lots of events happen in a single timestep; others — and this is more common for research-grade software — use continuous-time stochastic or Gillespie simulation, which can be significantly faster and is statistically exact, but is also a lot harder to code up. There didn’t seem to be a reusable simulation library that would play well with NetworkX, so I wrote one in the course of writing my complex networks blog/book. epydemic releases this code from the book and allows it to be used stand-alone. For good measure I integrated it with epyc to allow for reproducible simulations at scale.