flatland.core.env module

The env module defines the base Environment class. The base Environment class is adapted from rllib.env.MultiAgentEnv (https://github.com/ray-project/ray).

class flatland.core.env.Environment

Bases: object

Base interface for multi-agent environments in Flatland.

Derived environments should implement the following attributes:

action_space: tuple with the dimensions of the actions to be passed to the step method

Agents are identified by agent ids (handles). Examples:

>>> obs, info = env.reset()
>>> print(obs)
{
    "train_0": [2.4, 1.6],
    "train_1": [3.4, -3.2],
}
>>> obs, rewards, dones, infos = env.step(
    action_dict={
        "train_0": 1, "train_1": 0})
>>> print(rewards)
{
    "train_0": 3,
    "train_1": -1,
}
>>> print(dones)
{
    "train_0": False,    # train_0 is still running
    "train_1": True,     # train_1 is done
    "__all__": False,    # the env is not done
}h
>>> print(infos)
{
    "train_0": {},  # info for train_0
    "train_1": {},  # info for train_1
}

get_agent_handles(self)

Returns a list of agents’ handles to be used as keys in the step() function.

reset(self)

Resets the env and returns observations from agents in the environment.

Returns:

obs : dict

New observations for each agent.

step(self, action_dict)

Environment step.

Performs an environment step with simultaneous execution of actions for agents in action_dict. Returns observations for the agents. The returns are dicts mapping from agent_id strings to values.

Parameters:

action_dict : dict

Dictionary of actions to execute, indexed by agent id.

Returns:

obs : dict

New observations for each ready agent.

rewards: dict

Reward values for each ready agent.

dones : dict

Done values for each ready agent. The special key “__all__” (required) is used to indicate env termination.

infos : dict

Optional info values for each agent id.