Multi-Level Mesa

TL;DR

Multi-Level Mesa extends the Mesa ABM library to support complex networks with hierarchical agent groups, enabling multi-layered simulations of adaptive systems in Python.

Contribution

It introduces a new multi-level extension to the Mesa library, allowing for hierarchical and networked agent-based modeling in Python.

Findings

Supports complex networks with modules and hierarchies

Enables simulation of multi-layered adaptive networks

Demonstrated with Sugarscape model examples

Abstract

Multi-level Mesa is an extension to support the Python based Agents Based Model (ABM) library Mesa. Multi-level Mesa provides ABM infrastructure to allow for the inclusion of complex networks, which have modules (groups) and hierarchies (layers) of agents. This approach allows for users to define and simulate multi-layered adaptions of complex networks. This study reviews other multi-level libraries currently in the field, describes the main functions and classes of the Multi-level Mesa, and describes its implementation and impact in numerous varieties using the seminal ABM - Sugarscape. Multi-level Mesa and Sugarscape examples are available on GitHub at https://github.com/tpike3/multilevel_mesa and https://github.com/tpike3/SugarScape.