How northseawatch/honors-abm works
Project type and purpose
• This is a Python-based Agent-Based Model (ABM) built on the Mesa framework.
• Purpose: simulate ship movements, port interactions, and scrubber-water discharge policies in the North Sea.
• It is best classified as an open-source simulation tool / research prototype.
File structure
• Top-level:
• README.md, requirements.txt
• mesa/ directory containing core code, data, experiments, and output folders.
• mesa/data/: input datasets (CSV, Parquet).
• mesa/graphs/: generated visualization outputs.
• mesa/mesa_model.py: the main simulation engine and Mesa ModularServer setup.
• mesa/ship.py, port.py, filtered_ports_with_x_y.csv: agent definitions and positioning data.
• Experiment scripts (sweden_denmark_ban_exp.py, all_countries_ban_exp.py, nl_ban.py) configure different policy scenarios.
• plot_comparison.py: post-processing and chart generation.
README for additional insights
• Architecture overview is implicit:
• Agent types: Port, Ship, ScrubberTrail, Terrain.
• Core simulation loop in mesa_model.py with scheduling, grid, data collectors, and server/visualization.
• Data I/O: CSV input for ports, Parquet output for experiment results, PNGs in graphs/.
• Experiments orchestrate parameter sets and invoke the model.
• Visualization via Mesa’s CanvasGrid in a browser.