Monte Carlo Tree Search agent in a game of Isolation - Debug Suggestions Monte Carlo Tree Search agent in a game of Isolation - Debug Suggestions python-3.x python-3.x

Monte Carlo Tree Search agent in a game of Isolation - Debug Suggestions


I would recommend to write unit tests for each of the functions you have. Verify your assumptions about the behavior of your code. Try to be comprehensive about testing it, including corner cases. Just the need to input abstract test data usually reveals a lot about the architecture and details of a solution.

Furthermore, you could try to let your agent play any other suitable game. These steps will give you a good chance to discover whatever the bug is in your code, and make it more suitable for reuse in future.