- Streamlit
- Agent Based Modeling via the python library Mesa
- Using python classes
- Docstrings
In this toy model we use Mesa to simulate the process of a positive behavior (like composting) moving through a neighborhood. Via a streamlit app, we allow the user to explore how such propagation depends on the level of compatibility between neighbors and a the propensity of neighbors to talk about that behavior with others / adopt the behavior if encouraged to do so.
The behavioral model used here is a toy model with no scientific ground, but it could be improved to include a more complex one.
- Install the required packages from pipenv via
pipenv install. - Run the script via
pipenv run streamlit run sim_app.py - Interact with the simulation by clicking on the local URL that appears in the terminal at runtime. You can also share this simulation externally via the network URL.
- https://towardsdatascience.com/intro-to-agent-based-modeling-3eea6a070b72 Supermarket cashier model
- https://dmnfarrell.github.io/bioinformatics/abm-mesa-python COVID infection model
- One of the most famous ABMs: the Schelling segregation model https://ncase.me/polygons/
- Agent-based modeling: Methods and techniques for simulating human systems https://www.pnas.org/doi/10.1073/pnas.082080899
- Simple or complicated agent-based models? A complicated issue (doi: 10.1016/j.envsoft.2016.09.006) http://manuscript.elsevier.com/S1364815216306041/pdf/S1364815216306041.pdf
Streamlit will let you develop locally with the latest python (3.9.x) but can only host up to python 3.7.12. The latest pandas has moved on from python 3.7. All this is to say, please don't change the python and pandas versions specified in the pipenv files.