MAY is a high-performance, configuration-driven framework for building synthetic populations and distributing them across geography, residences, schools, workplaces and other venues. Nothing in it assumes a particular country or era, so it works with any administrative hierarchy.

The shipped configuration targets the modern-day UK (England 2021). Out of the box it builds County Durham and Darlington, two local authorities sized to run on a laptop. Since the data covers all four nations, widening the build is a config change. The USER_GUIDE explains how.

What it does

You supply census-style inputs: a geographical hierarchy, age × sex demographics for the smallest unit, household composition counts, and venue inventories. create_world.py turns those into one HDF5 file, world_state.h5, holding the whole synthetic world. That means every person, where they live, where they go to school or work or receive care, and the friendship and romantic-partnership networks between them.

YAML configuration files drive the pipeline, so building a new world means editing YAMLs and CSVs rather than Python.

Documentation

• Docs page holds the full documentation.
• USER_GUIDE.md walks through every YAML and CSV: configuring geography filters, editing household allocation, swapping census years, turning debug outputs on. Read it before changing any config.

Install

Requires Python 3.13+. Any environment manager will do, and Conda is the one we use:

conda create -n MayEnv python=3.13 -y
conda activate MayEnv
pip install -r requirements.txt

Or with venv:

python3.13 -m venv .venv
source .venv/bin/activate          # macOS / Linux
.venv\Scripts\activate             # Windows
pip install -r requirements.txt

Get the data

The repo ships without the bulky census and venue CSVs. Fetch them once:

python scripts/get_data.py

This downloads the archive and unpacks it into data/. It uses only the standard library, so it runs the same way on Windows, macOS and Linux. Pass --force to replace an existing data/ directory.

On macOS and Linux, bash scripts/get_data.sh does the same thing.

Run

# Default config (configs/2021/config.yaml)
python create_world.py

# Custom config / output file
python create_world.py --config configs/2021/config_uk_test.yaml --filename uk.h5

The run writes output/2021/world_state.h5. Its directory comes from serialization.output_dir in the config and its name from serialization.filename, which --filename overrides.

Project layout

MAY/
├── create_world.py     # Main entry point
├── configs/            # All user-facing configuration (2021, 1911)
├── data/               # Input CSVs (census-style), fetched separately
├── may/                # Core engine (generic, world-agnostic)
├── docs/               # Source for the documentation site
├── scripts/            # get_data.py and get_data.sh
├── tests/              # test_unit/, test_integration/
└── output/             # Written by a run; world_state.h5 lands here

Testing

pytest                                          # all tests
pytest tests/test_unit/may/population/          # specific module

pytest is not in requirements.txt, so install it separately to run the suite.

Viewing the world

Two separate tools can display a finished world:

• MAY-viewer
• MAY-world-visualiser

Requirements

Python 3.13+ and the packages pinned in requirements.txt (numpy, pandas, scipy, numba, h5py, PyYAML).

License

GNU General Public License v3.0. See LICENSE.

Copyright (C) 2026 Martha Correa. This program is free software: you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation, either version 3 of the License, or (at your option) any later version. This program is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details.