This code performs SSX algorithm on Four Rooms. 

****Note that current setup only runs in Python 2.****

This code is based on a Four Rooms environment from the github repository available at: https://github.com/david-abel/rl_info_theory which goes along with the paper "State Abstraction as Compression in Apprenticeship Learning", by David Abel, Dilip Arumugam, Kavosh Asadi, Yuu Jinnai, Michael L. Littman, and Lawson L.S. Wong, AAAI (2019). The Four Rooms environment is also dependent on the simple_rl package (https://github.com/david-abel/simple_rl) which can be installed via pip and is detailed in the paper "simple_rl: Reproducible Reinforcement Learning in Python" by David Abel, ICLR Workshop on Reproducibility in Machine Learning (2019). 

Some extra steps are necessary in order to run SSX on Four Rooms:
1. Install simple_rl via pip install simple_rl
2. Download rl_info_theory repository from https://github.com/david-abel/rl_info_theory
3. Copy fourrooms_savefigs.py, fourrooms_helper_functions.py, and figures from this directory to rl_info_theory directory 
4. Move to rl_info_theory directory

Command to run SSX on Four Rooms:

python fourrooms_savefigs.py

Result will be created and saved in figures directory.