CellularAutomata

pyCA is a Python library for building cellular automata and measuring the information their histories carry: the elementary CAs (Wolfram rules 0–255), an Ising variant where a deterministic rule competes with a heat bath, noisy and asynchronous rule corruptions, and outer-totalistic 2d automata, Conway's Game of Life included. Block entropy, entropy rate, mutual information, and Lempel-Ziv complexity read the resulting spacetime diagrams.

The original MATLAB classes (2016) live on in matlab/.

Installation

git clone https://github.com/EternalTime/CellularAutomata.git
cd CellularAutomata
python3 -m venv .venv
source .venv/bin/activate
python -m pip install --upgrade pip
pip install -e .

Requires Python 3.8+; numpy and matplotlib come along. The Getting Started guide is the authority on installation.

Quick start

from pyCA import ECA, CA2D, measures

# rule 110 on a random lattice of 256 cells
ca = ECA(110, N=256)
ca.run(500)
print(measures.entropy_rate(ca.spacetime(), k=4))

# watch it live (close the window to stop)
ca.play()

# Conway's Game of Life
life = CA2D.life((128, 128))
life.run(200)
print(life.population)

The Ising and stochastic families follow the same pattern:

from pyCA import ICA, NoisyECA, AsyncECA

ica = ICA(110, temperature=1.5, stochfrac=0.3, N=256)
noisy = NoisyECA(90, noise=0.01, N=256)
lazy = AsyncECA(30, update_fraction=0.7, N=256)

Documentation

The docs are hosted at damiansowinski.com/pyCA, and import pyCA; pyCA.docs() opens them. To build locally:

source .venv/bin/activate
pip install -e ".[docs]"
make -C docs html

Tests

source .venv/bin/activate
pip install -e ".[test]"
pytest

License

MIT