Python Automata Library (PAL)

PAL is a Python modeling library for high-performance agent-based, lattice population, and PDE models. Model code uses a compact object-oriented API and can be compiled with @pal.njit; PAL keeps performance-sensitive state and algorithms in a precompiled native core.

import PythonAutomataLibrary as pal

PAL provides:

• Grid: typed 1D/2D/3D lattice data.
• AgentGrid: lattice or continuous-position agents with identity and properties.
• PopGrid: integer lattice populations without individual agent identity.
• PDEgrid: diffusion/advection fields.
• IList: reusable integer query results.
• Multinomial: binomial/multinomial sampling.
• PixWindow and OpenGLWindow: 2D/3D visualization.
• @pal.njit: Numba compilation with PAL-specific source transformation and diagnostics.

PAL runs in safe mode by default. Call pal.FastMode() before constructing PAL objects to select the unchecked high-performance implementations after a model has been validated.

Documentation

• PAL Guide (PDF) — first model, worked modeling workflows, and detailed API behavior.
• PAL Architecture and Design (PDF) — implementation structure, design rationale, subsystem choices, testing, and performance.
• API Reference (PDF) — source-derived public signatures.
• Cheatsheet (PDF) — compact one-page lookup for common operations.
• Examples/ — complete working models.
• Testing — correctness, API, integration, and regression tests, with local and CI run instructions.

Installation

Install PAL from PyPI (including visualization dependencies):

python -m pip install PythonAutomataLibrary

For a minimal headless installation, install without dependencies and add only the core requirements:

python -m pip install --no-deps PythonAutomataLibrary
python -m pip install numpy numba

Note that --no-deps skips all dependencies, not just visualization; this minimal installation is for advanced users and may be reported as missing dependencies by pip check.

Release wheels bundle the native core for Windows, Linux, and macOS, without manual compilation. PAL v1.0.0 is published on PyPI. See distribution details.

Examples

Install PAL, then copy its bundled examples into a new, editable folder in your current directory:

python -m pip install PythonAutomataLibrary
python -m PythonAutomataLibrary.ExampleInstaller

To choose a different parent directory, pass its path, for example python -m PythonAutomataLibrary.ExampleInstaller ~/Documents (or python -m PythonAutomataLibrary.ExampleInstaller "C:\\Users\\YourName\\Documents" on Windows). This creates PAL-Examples/ inside the chosen directory, with runnable examples and a Documentation/ folder containing the PAL PDF manuals, without downloading from GitHub. The command refuses to overwrite an existing PAL-Examples/ folder. Run individual Python scripts from that folder; examples requiring a graphical display should be run on a desktop computer.

Build from source

PAL currently loads its native library from the package directory. Build it before importing PAL.

Windows, from an x64 Visual Studio Developer Command Prompt:

python build_native.py

Linux/macOS:

python build_native.py

Then run the tests:

python -m pytest -q Testing

Architecture

NativeCore.py is the Python/Numba API and compilation layer. pal_native.c owns stable mutable simulation state and the workhorse native algorithms. Safe and fast PAL objects are selected at construction time, so compiled hot paths do not branch on safety mode.

The public API has dedicated regression coverage in Testing/API/; the full Testing/ tree is the integration gate.

Citation

If you use PAL in your research, please cite:

Bravo, R. R. (2026). Python Automata Library (PAL) [Computer software]. Zenodo. https://doi.org/10.5281/zenodo.23439298

For reproducibility, cite the specific release you used: PAL v1.0.0.

GitHub's Cite this repository button provides citation formats from CITATION.cff.

PAL is based on the Hybrid Automata Library (HAL). Please also acknowledge or cite HAL when its underlying ideas or algorithms are relevant to your work.