bwosh / papersoccer_env

Environment for Reinforcement Learning with PAPER SOCCER

Paper Soccer RL Environment

This code is python code of Environment for Reinforcement Learning for game:

Paper Soccer https://en.wikipedia.org/wiki/Paper_soccer

Dependencies

Image preview is implemented and it depends on:

• opencv
• numpy (basic code only so any version should work fine)

Interface

Board class (inside papersoccer_env.py file) has couple interesting functions/variables including:

• possible_moves() - gets dictionary o fpossibel moves
• move(...) - executes one of the move from possible moves returning outcome information
• draw() - prepares RGB uint8 image with current state visualization
• data variable containing board fields data

Notebook/Quick Start

For quick start you can check this notebook. (some more dependencies here)

License

Code author:

• Błażej Matuszewski

The project is licensed upon:
MIT License