PanaXea

Introduction

Welcome to PanaXea!

A minimalist framework to easily and quickly develop agent-based models in Python. PanaXea aims to provide a set of tools and utilities to rapidly achieve common tasks in agent-based model development such as schedule and environment management, agent behaviour setup, etc.

It provides:

• Easy configuration model properties and keep track of parameter values and outputs;
• A scheduling system that implements model progression as a set of epochs, and makes sure all agents are executed within a single epoch;
• Classes for 2D and 3D environments, and helper methods for these;
• Generic classes for agents and other steppables providing commonly used functionalities.

Table of Contents

• Contributing - PanaXea welcomes contributors, see the guidelines for an overview of how to participate in the project;
• Getting started - A quick description of how to get started designing your own models;
• Installation - A description of how to install the project;
• License - PanaXea is distributed under MIT license;

Getting started

See examples in ./examples for a few simple models to get you started.

The documentation is available on readthedocs.

PanaXea will work with both Python 2 and Python 3.

The core and toolkit modules do not require any additional packages to be used.

It might be useful to have Nox installed to run the end-to-end linting/testing/documenting pipeline. This will also require you to have pytest, sphinx and flake8 installed.

Running some of the examples requires NumPy and Matplotlib.

Overall, you might be more comfortable using an Anaconda distribution of Python.

Installation

As a first possibility you may clone/download the project and make sure the core module (and the toolkit module if you plan on using it) are added to your PYTHONPATH.

You may also install using pip:

pip install PanaXea

Contact

The project is maintained by Dario Panada, PhD student at The University of Manchester (UK). - dario.panada at manchester dot ac dot uk