TFG-CellularAutomata

Cellular Automata Properties Prediction is a Final Degree Project developed by Ezequiel González for his 22-23 last year in ETSII college of Seville, Spain.

It lets you generate different types of datasets based on 3 rule types. With those you can:

• Generate interesting plots based mainly in their density evolution over time.
• Plot individual estates.
• Train, evaluate and test Machine Learning models to predict new data.
• Compare those models performance with plots.

Main Workflow

The main idea is to provide an easy to use library to make this process as automated as possible.

First you define the CA based on certain attributes like initial state, size, rules etc. Those 3 rules can be also easy extended. This is made in the package CA.

Then you can build the dataset with this new CA and the free parameters. Those datasets are stored in data/datasets. This is made in package datasets.

Now, with the new built dataset, you can start analyzing it via the multiple plots that are available in the package statistics.

Finally, you can train, evaluate and test Machine Learning models to predict new data. This is made in the package learning.

Getting started

Requirements

• Python 3.10.7
• pipenv (2022.10.11) for virtual environment management

Setup

First clone the repository from Github:

git clone https://github.com/ezegonmac/TFG-CellularAutomata

Create a virtual environment:

cd TFG-CellularAutomata
pipenv shell

Install requirements:

pip install -r requirements.txt

Usage

The main package is src. It contains the following principal subpackages:

• CA: Contains the CA class and the 3 rule types.
• datasets: Contains the definition of datasets and the functions to build them.
• statistics: Contains the functions to generate statistical plots of the datasets.
• learning: Contains the functions to train, evaluate and test Machine Learning models.

Their functions must be called from their specific main files. Those files are located in the src folder.

Those main files are:

• main_datasets.py: Main functions for generating the default datasets.
• main_density_datasets.py: Main functions for generating the density datasets.
• main_density_plots.py: Main functions for generating the density plots.
• main_doc: Main functions for generating the documentation figures.
• main_learning.py: Main functions for managing the learning process.
• main_states.py: Main functions for generating states plots.

Now you can comment/uncomment the functions you want to run in those files, or create your own functions. You can run those files from the src folder with the following command:

python src/main_file.py

If you dont have your virtual environment activated, you should run the following command first:

pipenv shell