Sokobandits, A Hybrid Q-Learning and BFS Approach to the Sokoban Game

Contributors: Dan Nguyen, Jillian Ness, Brooke Ryan

Final project for CS 271, Introduction to Artificial Intelligence, UC Irvine Fall Quarter 2020.

Instructor: Professor Kalev Kask

README

To Run the Project

To run trials, there are a few options:

Run Trials on Remote Server via GitHub Actions

This option should be followed if you would prefer to run trials on a remote server instead of your own machine.

In our project's GitHub repository, trials are configured to run every time a pull request is posted to the branch.

1. Follow these steps to create a pull request for the project Sokoban-AI.

2. Once you have a pull request posted to the project, navigate to the project's Actions. On this page, you can see the history of all trial runs for the project. Click on the workflow that corresponds to your pull request.

3. There, you can view a build job for each benchmark problem:

Run Trials on Local Machine

If desired, trials can also be run on your local machine. Please note, the project can currently find a solution only for benchmark problems 00 through 07a.

1. Ensure you have Python >=3.8 set up on your machine.

2. Navigate to the root directory of the project folder. (This is so the dependencies step can utilize the setup.py script that specifies library dependencies.)

3. Install dependencies.

   python -m pip install --upgrade pip
   pip install .
   

4. Navigate to the testing directory.

   cd test/trials_no_logging/
   

5. Run trials.

   0. python -m unittest sokoban00.py
      

   1. python -m unittest sokoban01.py
      

   2. python -m unittest sokoban02.py
      

   3. python -m unittest sokoban03.py
      

   4. python -m unittest sokoban04.py
      

   5. python -m unittest sokoban05a.py
      

      python -m unittest sokoban05b.py
      

   6. python -m unittest sokoban06a.py
      

      python -m unittest sokoban06b.py
      

      python -m unittest sokoban06c.py
      

   7. python -m unittest sokoban07a.py