Multi agent env

Jittable abstract base class for all jaxmarl Environments.

agent_classes: dict property

Returns a dictionary with agent classes, used in environments with hetrogenous agents.

Format

agent_base_name: [agent_base_name_1, agent_base_name_2, ...]

name: str property

Environment name.

__init__(num_agents)

num_agents (int): maximum number of agents within the environment, used to set array dimensions

action_space(agent)

Action space for a given agent.

get_avail_actions(state)

Returns the available actions for each agent.

get_obs(state)

Applies observation function to state.

observation_space(agent)

Observation space for a given agent.

reset(key)

Performs resetting of the environment.

step(key, state, actions, reset_state=None)

Performs step transitions in the environment. Resets the environment if done. To control the reset state, pass reset_state. Otherwise, the environment will reset randomly.

step_env(key, state, actions)

Environment-specific step transition.