rlutils: A utility package for implementing reinforcement learning simulations.

rlutils is a python package containing utility functions to implement reinforcement learning simulations. The goal of rlutils is to provide a useful base library to help implement reinforcement learning (RL) research projects. The examples directory contains jupyter notebooks demonstrating how to use rlutils.

rlutils provides an API to implement RL agents, policies, and simulate them in a control task (also called environment or Markov Decision Process (MDP)). rlutils also provides implementations to some basic policies, value iteration, and Q-learning. Other utility functions are also included to collect trajectories, store them, and replay them. rlutils' goal is to provide a framework and utility functions for implementing RL algorithms.

If you have any questions about rlutils, please send email to lucas_lehnert@brown.edu.

Installation

rlutils is tested on python 3.5, 3.6, 3.7, and 3.8.

This package can be installed using pip:

pip install rlutils

Alternatively, the repository can also be cloned and the package can be installed without pulling from PyPI:

git clone https://github.com/lucaslehnert/rlutils.git
cd rlutils
pip install .