mazsim

Installation

1. install UV package manager

2. install Microsoft Visual Studio Community and make sure the C++ MSVC build tools option is selected during install

3. Setup pandana:

   • manually clone the jkolberg pandana fork from github.com/jkolberg/pandana and switch to the pandas_23 branch
   • change the tool.uv.sources path in pyproject.toml to your cloned pandana directory location

4. Place base year data in projects/baseline_summer2026/data

5. create the uv venv:

   uv sync

6. run estimation with the following command:

   uv run mazsim estimate -c projects\baseline_summer2026\configs

7. run calibration with the following command:

   uv run mazsim calibrate -c projects\baseline_summer2026\configs

8. run validation with the following command:

   uv run mazsim validate -c projects\baseline_summer2026\configs

   validation runs the simulation out the most recent year that's included in the observed_data table. The simulation can then be compared to the observerd data before running the full simulation.