Skip to content

Latest commit

 

History

History
25 lines (17 loc) · 1.47 KB

README.md

File metadata and controls

25 lines (17 loc) · 1.47 KB

TabulaRL

Simple tabular reinforcement learning in Python.

This is a very simple project for tabular reinforcement learning (TabulaRL) that I wrote during my PhD studies. There are already several great open source packages for reinforcement learning, including:

Out of these, it looks like OpenAI Gym is currently the most promising... if I was starting this all again maybe I would have focused my attention here!

That said, I found that for studying simple tabular RL problems these frameworks were all a little bit of overkill. If you are experimenting with small, simple, tabular MDPs or just learning about RL for the first time it can be nice to have a really simple bit of code to get started. The nice thing about this code instead of something like Gym is:

  • All written in simple Python+Numpy, no complicated dependencies beyond standard Anaconda
  • Very few lines of code, potentially more accessible to novice programmers.
  • Focus on very simple problems.

But the main reason I'm putting this code up is so that people can more easily implement/verify some of the ideas in some of my papers!