The repository contains environments used in LCS literature that are compliant with OpenAI Gym interface.
- Maze (different variants)
- Binary Multiplexer
- Real Multiplexer
- Hand Eye
- Checkerboard
- Real-valued toy problems
- 1D Corridor
- 2D Grid
- Mountain Car with energy reward
- Finite State World (FSW)
For some usage examples look at examples/ directory.
conda create --name openai-envs python=3.7
conda activate openai-envs
pip install -e ".[testing]"
python setup.py test