ray.rllib.env.single_agent_episode.SingleAgentEpisode.get_return

SingleAgentEpisode.get_return() → float

Calculates an episode’s return, excluding the lookback buffer’s rewards.

The return is computed by a simple sum, neglecting the discount factor. Note that if self is a continuation chunk (resulting from a call to self.cut()), the previous chunk’s rewards are NOT counted and thus NOT part of the returned reward sum.

Returns:

The sum of rewards collected during this episode, excluding possible data inside the lookback buffer and excluding possible data in a predecessor chunk.