Fixes to make this project compatible with modern Python versions (3.10+) and packages updates

[Leirbag-gabrieL]

Description

Hello, I wanted to use this repo to learn Q-learning and globally deep reinforcement learning. Unfortunately, right now when you pip install everything, some packages used (gym, and numpy) have received updates that broke the project. So I fixed them. It may break the current CI as I am maybe using some new Python features since I use Python 3.12.11.

For now, I've tested my code manually, it passes the tests and will soon add tests which targets changes I made.
This update is targeted in order to make changes for the Super Mario gym work properly, Kautenja/gym-super-mario-bros#137.

Here is a non-exhaustive list of changes I made :

• Switched from gym (outdated and not maintained) to gymnasium. Now envs should return (observation, reward, terminated, truncated, info)
• Added truncation options with a new truncate_function argument and max_episode_steps arguments. Basically early stopping mechanics.
• Fixed some type issues which made numpy crash the program
• Fixed the joypad wrapper to stay compatible with gymnasium
• Overall cleanup and refacto
• Added the same kind of pre-commit gymnasium is using to maintain a proper code base (that's why 23 files changed)

Type of change

Please select all relevant options:

• Bug fix (non-breaking change which fixes an issue)
• New feature (non-breaking change which adds functionality)

How Has This Been Tested?

Manually with super mario gym and using the test pipeline already made.

• ran python3 -m unittest discover . a lot

Test Configuration

• Operating System: Windows 11
• Python version: 3.12.11
• C++ compiler version: Microsoft Visual Studio (MSVC) 14.43.34808

Checklist

• My code follows the style guidelines of this project
• I have performed a self-review of my own code
• I have commented my code, particularly in hard-to-understand areas
• I have made corresponding changes to the documentation
• I have added tests that prove my fix is effective or that my feature works