rlgt.environments.graph_environment.EpisodeStatus

class documentation

class EpisodeStatus(Enum):

View In Hierarchy

This enumeration represents the possible statuses of an episode in a reinforcement learning environment for graph theory applications.

Note
This enumeration also applies to batches of episodes. In episodic tasks, all parallel episodes are guaranteed to reach a terminal state after the same number of actions. As a result, regardless of whether the tasks are episodic or continuing, all episodes within a batch always share the same status. This common value is regarded as the status of the entire batch.
Constant IN_PROGRESS The episode is ongoing, meaning that the environment may continue to accept actions.
Constant TERMINATED The episode has ended because a terminal state has been reached, so the environment can no longer accept actions. This status arises only in environments with episodic tasks.
Constant TRUNCATED The episode has ended because a predefined step limit has been reached. The current state is not terminal, but no further actions should be taken. This status arises only in environments with continuing tasks.
IN_PROGRESS: int =

The episode is ongoing, meaning that the environment may continue to accept actions.

Value
0
TERMINATED: int =

The episode has ended because a terminal state has been reached, so the environment can no longer accept actions. This status arises only in environments with episodic tasks.

Value
1
TRUNCATED: int =

The episode has ended because a predefined step limit has been reached. The current state is not terminal, but no further actions should be taken. This status arises only in environments with continuing tasks.

Value
2