Tutorials

Learn to use Coopetition-Gym through hands-on examples, from basic environment interaction to advanced multi-agent training.

Getting Started

Before diving into tutorials, ensure you have: 1. Installed Coopetition-Gym

  1. Verified your installation works
  2. Basic familiarity with Python and NumPy

Tutorial Overview

Beginner

Tutorial Description Time
Quick Start First steps with Coopetition-Gym 15 min

Intermediate

Tutorial Description Time
Environment Deep Dive Understanding observations, actions, and rewards 30 min
Training with Stable-Baselines3 PPO training on TrustDilemma-v0 45 min
PettingZoo APIs Using Parallel and AEC interfaces 30 min

Advanced

Tutorial Description Time
Multi-Agent Training Independent and centralized training 60 min
Custom Strategies Implementing game-theoretic policies 45 min
Experiment Design Reproducible research workflows 45 min

Quick Start Guide

The Quick Start Tutorial covers: 1. Creating environments - Using the factory functions

  1. Basic interaction - Reset, step, observe
  2. Understanding observations - What agents see
  3. Understanding rewards - How payoffs work
  4. Running episodes - Complete interaction loops

Learning Path

For MARL Researchers

  1. Quick Start → Basic API familiarity
  2. Environment Deep Dive → Understanding state spaces
  3. Multi-Agent Training → Algorithm implementation
  4. Experiment Design → Reproducible benchmarks

For Game Theorists

  1. Quick Start → Basic API familiarity
  2. Custom Strategies → Implementing equilibrium strategies
  3. Case Studies → SLCD and Renault-Nissan analysis

For Engineers

  1. Quick Start → Basic API familiarity
  2. Training with Stable-Baselines3 → Practical training
  3. PettingZoo APIs → Integration patterns

Prerequisites by Tutorial

Tutorial Prerequisites
Quick Start Python basics, NumPy
Environment Deep Dive Quick Start completed
Training with SB3 Quick Start, familiarity with RL concepts
PettingZoo APIs Quick Start
Multi-Agent Training SB3 tutorial, MARL concepts
Custom Strategies Game theory basics
Experiment Design All previous tutorials

Code Repository

All tutorial code is available in the examples/ directory:

examples/
├── quickstart.py           # Quick start examples
├── sb3_training.py         # Stable-Baselines3 training
├── pettingzoo_demo.py      # PettingZoo API examples
├── multi_agent/            # Multi-agent training scripts
└── strategies/             # Game-theoretic strategies

Next Steps

Start with the Quick Start Tutorial to begin working with Coopetition-Gym.

After completing tutorials, explore: