README.md

Gym Games

This is a collection of Gymnasium compatible games for reinforcement learning.

Note

For Gym compatible version, please check v1.0.4.

For PyGame Learning Environment, the default observation is a non-visual state representation of the game.

For MinAtar, the default observation is a visual input of the game.

Environments

• PyGame learning environment:

  • Catcher-PLE-v0
  • FlappyBird-PLE-v0
  • Pixelcopter-PLE-v0
  • PuckWorld-PLE-v0
  • Pong-PLE-v0

• MinAtar:

  • Asterix-MinAtar-v1
  • Breakout-MinAtar-v1
  • Freeway-MinAtar-v1
  • Seaquest-MinAtar-v1
  • SpaceInvaders-MinAtar-v1

• Exploration games:

  • NChain-v1
  • LockBernoulli-v0
  • LockGaussian-v0
  • SparseMountainCar-v0
  • DiabolicalCombLock-v0

Installation

Gymnasium

Please read the instruction here.

Pygame

• On OSX:

  brew install sdl sdl_ttf sdl_image sdl_mixer portmidi
  pip install pygame==2.5.2
  

• On Ubuntu:

  sudo apt-get -y install python-pygame
  pip install pygame==2.5.2
  

• Others: Please read the instruction here.

PyGame Learning Environment

git clone https://github.com/ntasfi/PyGame-Learning-Environment.git
cd PyGame-Learning-Environment/
pip install -e .

MinAtar

pip install minatar==1.0.15

Gym-games

pip install git+https://github.com/qlan3/gym-games.git

Example

Run python test.py.

Cite

Please use this bibtex to cite this repo:

@misc{gym-games,
  author = {Lan, Qingfeng},
  title = {Gym Compatible Games for Reinforcement Learning},
  year = {2019},
  publisher = {GitHub},
  journal = {GitHub Repository},
  howpublished = {\url{https://github.com/qlan3/gym-games}}
}

References

• gym
• gym-ple
• SRNN
• MinAtar
• Latent State Decoding
• Gymnasium