ray.rllib.env.multi_agent_episode.MultiAgentEpisode.get_agents_that_stepped

MultiAgentEpisode.get_agents_that_stepped() → Set[Any]

Returns a set of agent IDs of those agents that just finished stepping.

These are all the agents that have an observation logged at the last env timestep, which may include agents, whose single agent episode just terminated or truncated.

Returns:

A set of AgentIDs of those agents that just finished stepping (that have a most recent observation on the env timestep scale), regardless of whether their single agent episodes are done or not.