Many existing agent-based modeling frameworks, like Mesa, focus their interactive and visualization capabilities on spatial modeling. In addition, agent-based modeling in Python – while much more powerful than an environment like NetLogo – is much more difficult to get started with.
Helipad makes it dead-simple to get started writing and interacting with your agent-based model.
Live dataSee your model’s data being generated in real time.
InteractiveParameters can be set beforehand from the control panel GUI, or adjusted live while the model is running.
ShocksInteract with your model in real time, or shock parameters automatically with timer functions.
Easy to useA hook-based API makes it easy to manipulate the functionality you need, without worrying about what you don’t.
Jupyter IntegrationRun models as a standalone GUI application, package them as a Jupyter notebook, or – if you’re brave – run them without a GUI at all.
Network FunctionalityBuild networks among agents and analyze them with NetworkX integration.
Powerful Agent ClassesAgents come with the ability to barter, buy and sell with money, reproduce both haploid and polyploid, and more.
Flexible Model-BuildingBuild sequential or random-activation models, matching models, multi-level models, and exhaust all the possibilities with parameter sweeps.
Jump Right In
Helipad is cross-platform and runs anywhere that Python 3.6 runs.
pip install helipad, or download from Github.
Write a ModelScan the Getting Started Guide or peruse the Function and Hook Reference.