Installation

This guide covers how to install the UNO Card Game RL project on your system.

Requirements

• Python 3.8 or higher

• pip package manager

• Git (optional, for cloning)

System Dependencies

Windows

No additional system dependencies required.

Linux

You may need SDL libraries for Pygame:

sudo apt-get install python3-dev libsdl2-dev libsdl2-image-dev libsdl2-mixer-dev libsdl2-ttf-dev

macOS

Install SDL via Homebrew:

brew install sdl2 sdl2_image sdl2_mixer sdl2_ttf

Standard Installation

1. Clone the Repository

git clone https://github.com/user/uno-card-game-rl.git
cd uno-card-game-rl

2. Create Virtual Environment (Recommended)

# Create virtual environment
python -m venv venv

# Activate (Windows)
venv\Scripts\activate

# Activate (Linux/macOS)
source venv/bin/activate

3. Install Dependencies

pip install -r requirements.txt

Development Installation

For development, install with editable mode:

pip install -e .
pip install -r requirements-dev.txt

Required Packages

Core Dependencies

Package	Version	Purpose
gymnasium	≥0.29.0	RL environment framework
stable-baselines3	≥2.0.0	RL algorithms (PPO, DQN, A2C)
sb3-contrib	≥2.0.0	RecurrentPPO (LSTM policies)
pygame	≥2.5.0	GUI rendering
numpy	≥1.24.0	Numerical operations
pandas	≥2.0.0	Data analysis
torch	≥2.0.0	Deep learning backend

Optional Dependencies

Package	Version	Purpose
tensorboard	≥2.14.0	Training visualization
matplotlib	≥3.7.0	Plotting results
seaborn	≥0.12.0	Statistical plots

Verifying Installation

Run the following to verify everything is installed correctly:

# Test basic imports
python -c "from src.game import UnoGame; print('✓ Game engine OK')"

# Test GUI
python -c "import pygame; pygame.init(); print('✓ Pygame OK')"

# Test RL libraries
python -c "from sb3_contrib import RecurrentPPO; print('✓ RecurrentPPO OK')"

Troubleshooting

Common Issues

ImportError: No module named ‘pygame’

Run pip install pygame and ensure your virtual environment is activated.

CUDA not available

Training will work on CPU. For GPU acceleration, install PyTorch with CUDA support:

pip install torch --index-url https://download.pytorch.org/whl/cu118

Model file not found

Ensure you’re running from the project root directory where the models/ folder exists.

Getting Help

If you encounter issues:

1. Check the Quickstart guide

2. Search existing GitHub issues

3. Open a new issue with your error message and system info