It's fun to work with Openai-gym and reinforcement learning. In this repository you are going to find a notebook and several python scripts to introduce to you only the basics.

It is expected that you follow the order below inspecting the codebase:

• openai-gym-intro.ipynb: a jupyter notebook describing openai-gym. It includes description of the following python scripts. It also includes Q-learning based reinforcement learning for the Cartpole-v0 gym environment.

• ALE-Breakout-v5_code1.py : The atari Breakout powered by ALE. The agent acts randomly per step.
• Blackjack-v1-code1.py : Blackjack gym. The agent acts randomly per step.
• CarRacing-v2-code1.py : CarRacing gym. The agent acts randomly per step.
• CartPole-v0-code1.py : CartPole gym. The agent acts randomly per step.
• CartPole-v0-code2.py : Basics / introduction to the CartPole-v0 gym.
• CartPole-v0-code3.py : Random agent for Cartpole-v0 gym.
• CartPole-v0-code4.py : Training an agent to play CartPole-v0 gym with Q-learning.
• Riverraid-v0_code1.py : Riverraid-v0 gym. The agent acts randomly per step.
• lunarlanderv2.py : LunarLanderv2 gym. The agent acts randomly per step.
• pongv0_code1.py : pongv0 gym. The agent acts randomly per step.

• list_all_envs_registry.py :

• Python 3.10.x
• (MacOS specific dependency!) Install package swig with brew install swig in case your gymnasium install throws error like Failed building wheel for box2d-py.
• Here are few pip commands you might be interested to run if got hiccups multiple times with the requirements.txt
  • Installing the package gymnasium with all its dependencies, plus all ROM license to create Atari game environments: pip install "gymnasium[all,accept-rom-license]"
  • pip install tqdm
  • pip install joblib
  • pip install termcolor
• pip install -r requirements.txt (the file is given)
• Make sure you have gymnasium==0.29.1 installed. And verify print(gymnasium.__version__) returns 0.29.1.