RL Tic-Tac-Toe

This project is a command-line implementation of Tic-Tac-Toe where a user can play against a Reinforcement Learning (RL) agent. The agent is trained using the Q-learning algorithm. The project also allows for training the AI agents and watching them play against each other.

Features

• Play Tic-Tac-Toe against an AI.
• Train the AI through self-play.
• Watch two AI agents play against each other.
• Command-line interface.

Installation

1. Ensure you have Python 3.12+ and poetry installed.
2. Clone the repository:

   git clone https://github.com/your-username/rl-tic-tac-toe.git

3. Install dependencies:

   poetry install

Usage

To start the interactive game menu, run:

poetry run python src/rl_tic_tac_toe/main.py

From the menu, you can choose to:

1. Train the AI agents.
2. Play against a trained AI.
3. Watch two AI agents play against each other.

Running Tests

To execute the test suite, run:

poetry run pytest

Development

Code Style and Linting

This project uses Ruff for code formatting and linting. To check the code for style issues, run:

poetry run ruff check .

Static Type Checking

This project uses MyPy for static type checking. To check for type errors, run:

poetry run mypy .