search

udacity / deep-reinforcement-learning

Repo for the Deep Reinforcement Learning Nanodegree program
https://www.udacity.com/course/deep-reinforcement-learning-nanodegree--nd893
MIT License
4.9k stars 2.34k forks source link

Fix discretization Q-Learning agent's act() #44

Closed NiklasZ closed 5 years ago

NiklasZ commented 5 years ago

The self.last_state and self.last_action are not defined on the first call to act(), resulting in an TypeError: unsupported operand type(s) for +: 'NoneType' and 'tuple' when trying to update the Q table.

NiklasZ commented 5 years ago

Closing, as I think it might be a different problem.