Tests for CS540 Spring 2022 HW10: Reinforcement Learning
- More tests needed, please feel free to contribute. *
- Updated to support Spring 2022 Implementation
- Checks EPSILON values for 50,000 , 20,000 , 10,000 and 5,000 episodes
- Checks whether implementation was able to make progress before 6 different episodes. If not, implementation is likely inaccurate.
IN_DEVELOPMENT
Download tests.py, Q_learning.csv, and move them into the same directory. In order for tests.py to successfully run, it requires (Q_TABLE.pkl) which is created by running Q_learning.py.
The contents of your directory should look like this:
$ tree
.
├── tests.py
├── Q_learning.py
└── Q_TABLE.pklTo run the tests, reference hw10.pdf where you'll find the required dependencies needed.
Once all dependencies are install, you may run the tests using:
$ python3 tests.pyThese tests are not endorsed or created by anyone working in an official capacity with UW Madison or any staff for CS540. These tests take portions from the official tests with modification to assist in grading the robustness of student implementations. Tests created/modified are not guaranteed to be correct. Use tests at your own risk.
By running tests.py, you are executing code you downloaded from the internet. Back up your files and take a look at what you are running first.
For any comments, questions, and contribution -- create a pull request.