rl.logger.Logger Class Reference

This class provides logging functionality to the reinforcement part of the code. More...

Public Member Functions
def	__init__ (self)
def	log_episode (self, int episode, np.ndarray final_state, np.ndarray goal, int time_steps, int total_finished, List[bool] episodes_finished, float reward, float eps)
	Log 1 episode. More...

Detailed Description

This class provides logging functionality to the reinforcement part of the code.

Member Function Documentation

log_episode()

def rl.logger.Logger.log_episode	(		self,
		int	episode,
		np.ndarray	final_state,
		np.ndarray	goal,
		int	time_steps,
		int	total_finished,
		List[bool]	episodes_finished,
		float	reward,
		float	eps
	)

Log 1 episode.

Parameters

episode	Episode number
final_state	State at the end of the episode
goal	Goal this episode
time_steps	Amount of steps taken in the episode
total_finished	Total amount of times during the reinforcement run the arm reached the goal.
episodes_finished	boolean array with length 50. It describes the last 50 episodes. True = the episode finished, False = the episode did not finish.
reward	The reward of the last step in the episode
eps	Epsilon parameter of the DQN at this moment.

---

The documentation for this class was generated from the following file:

• src/rl/logger.py