mesa

Why we included this project

Mesa is a Python library for agent-based modeling, aimed at people who want to simulate complex systems by defining rules for individual agents and then watching what happens at the level of the whole population. It comes with the standard machinery, spatial grids and scheduling loops, so you don't have to build that plumbing yourself, but you can still drop in your own logic for how each agent behaves. Models render in a browser interface, and results feed straight into the usual Python data-analysis tools, so writing a model and inspecting its output happen in one place. Researchers in fields like social science, ecology, economics, or epidemiology often pick it up as a scriptable alternative to NetLogo and similar tools, and anyone who prefers to keep simulation work inside Python will find it a comfortable fit.