smarts.zoo.agent_spec module

class smarts.zoo.agent_spec.AgentSpec(interface: ~smarts.core.agent_interface.AgentInterface | None = None, agent_builder: ~typing.Callable[[...], ~smarts.core.agent.Agent] | None = None, agent_params: ~typing.Any | None = None, observation_adapter: ~typing.Callable = <function AgentSpec.<lambda>>, action_adapter: ~typing.Callable = <function AgentSpec.<lambda>>, reward_adapter: ~typing.Callable = <function AgentSpec.<lambda>>, info_adapter: ~typing.Callable = <function AgentSpec.<lambda>>)[source]

A configuration that is used by SMARTS environments.

agent_spec = AgentSpec(
    interface=AgentInterface.from_type(AgentType.Laner),
    agent_params={"agent_function": lambda _: "keep_lane"},
    agent_builder=Agent.from_function,
)

env = gym.make(
    "smarts.env:hiway-v1",
    scenarios=["scenarios/sumo/loop"],
    agent_specs={agent_id: agent_spec},
)

agent = agent_spec.build_agent()

Refer to the Agent documentation.

action_adapter(): Deprecated. Do not use. An adaptor that allows shaping of the action. Defaults to lambda act: act.

agent_builder: Callable[[...], Agent] | None = None: A callable to build an smarts.core.agent.Agent given AgentSpec.agent_params (default None)

agent_params: Any | None = None: Parameters to be given to AgentSpec.agent_builder (default None)

build_agent() → Agent[source]: Construct an Agent from the AgentSpec configuration.

info_adapter(reward, info): Deprecated. Do not use. An adaptor that allows shaping of info. Defaults to lambda obs, reward, info: info.

interface: AgentInterface | None = None: the adaptor used to wrap agent observation and action flow (default None)

observation_adapter(): Deprecated. Do not use. An adaptor that allows shaping of the observations. Defaults to lambda obs: obs.

reward_adapter(reward): Deprecated. Do not use. An adaptor that allows shaping of the reward. Defaults to lambda obs, reward: reward.