A reinforcement learning environment for the game Pong

In the repos, i provide An Environement for the game Pong. The Environement is based on OpenAI gym and can be trained with any RL approach.

Developer Guide

Then Environement is designed with python 3.7 in mind. How an environment can be created, how to train and evaluate models and how to modify the environment is discussed below.

How to set up the Environment.

python3 -m venv game-env
source game-env/bin/activate
pip3 install -r requirements.txt

The game is already implemented, nor yet perfect but work.

How to play malually

The game can be play using the keyboard. The player at the top can move using a and d. The at the bottom moves using KEY right and Key left. The game can be lauch using the following command.

python3 play.py

Todo

• implement the environment
• optimize the collision check
• write docstring for methods and classes
• write the tests for the game
• write the tests for the environement

How to contribute