Multi-Agent Deep Reinforcement Learning in 13 Lines of Code Using PettingZoo