Module ai.environments

Module containing the abstract definition of an environment, as well as implementations of it.

"""Module containing the abstract definition of an environment, as well as
implementations of it."""

from . import action_spaces
from ._base import Base
from ._factory import Factory
from ._simulator_wrapper import SimulatorWrapper
from ._gym_wrapper import GymWrapper


__all__ = ["Base", "SimulatorWrapper", "action_spaces", "Factory", "GymWrapper"]

Sub-modules

ai.environments.action_spaces

Action spaces for environments.

Classes

class Base

Environment base class.

An environment is a stateful environment upon which action may be executed. It has an internal state that is modified by the action and (potentially only partially) observable from the outside.

Ancestors

• abc.ABC

Subclasses

• GymWrapper
• SimulatorWrapper

Static methods

def get_factory(*args, **kwargs) ‑> Factory

Creates and returns a factory object that spawns simulators when called.

Args and kwargs are passed along to the class constructor. However, if other behavior is required, feel free to override this method and return a factory class of your choice.

Expand source code

@classmethod
def get_factory(cls, *args, **kwargs) -> "environments.Factory":
    """Creates and returns a factory object that spawns simulators when called.

    Args and kwargs are passed along to the class constructor. However, if other
    behavior is required, feel free to override this method and return a factory
    class of your choice."""
    return environments.Factory(cls, *args, **kwargs)

Instance variables

var action_space : Base

The action space instance used by the environment instance.

Expand source code

@property
@abstractmethod
def action_space(self) -> environments.action_spaces.Base:
    """The action space instance used by the environment instance."""
    raise NotImplementedError

Methods

def close(self)

Disposes resources used by the environment.

Expand source code

@abstractmethod
def close(self):
    """Disposes resources used by the environment."""
    raise NotImplementedError

def render(self)

Renders the environment in its current state.

Expand source code

@abstractmethod
def render(self):
    """Renders the environment in its current state."""
    raise NotImplementedError

def reset(self) ‑> numpy.ndarray

Resets the environment to a new initial state.

Returns

np.ndarray

Initial state.

Expand source code

@abstractmethod
def reset(self) -> np.ndarray:
    """Resets the environment to a new initial state.

    Returns:
        np.ndarray: Initial state.
    """
    raise NotImplementedError

def step(self, action: int) ‑> Tuple[numpy.ndarray, float, bool, Dict]

Executes an action in the environment.

Args

action : int

Action index

Returns

Tuple[np.ndarray, float, bool, Dict]

Tuple of next state, reward, terminal

flag, and debugging dictionary.

Expand source code

@abstractmethod
def step(self, action: int) -> Tuple[np.ndarray, float, bool, Dict]:
    """Executes an action in the environment.

    Args:
        action (int): Action index

    Returns:
        Tuple[np.ndarray, float, bool, Dict]: Tuple of next state, reward, terminal
        flag, and debugging dictionary.
    """
    raise NotImplementedError

class Factory (cls: Type[Base], *args, **kwargs)

Factories are callable objects that spawn environment instances.

Args

cls : Type[environments.Base]

Environment class.

*args, **kwargs: arguments and key-word arguments passed to the environment __init__ method.

Subclasses

• ai.environments._simulator_wrapper.SimulatorWrapper.Factory

class GymWrapper (env: Union[gym.core.Env, str])

Environment wrapper for openAI gym environments.

Args

env : Union[gym.Env, str]

Environment instance to wrap, or the gym environment identifier..

Ancestors

• Base
• abc.ABC

Class variables

var ActionSpace

Discrete action space that wraps a discrete openAI Gym action space.

Inherited members

• Base:
  • action_space
  • close
  • get_factory
  • render
  • reset
  • step

class SimulatorWrapper (simulator: Base)

Environment that wraps a simulator, exposing it as an environment.

Args

simulator : simulators.Base

Simulator class to wrap into an environment.

Ancestors

• Base
• abc.ABC

Class variables

var Factory

Factories are callable objects that spawn environment instances.

Static methods

def get_factory(factory: Factory) ‑> ai.environments._simulator_wrapper.SimulatorWrapper.Factory

Wraps a simulator factory into an environment factory.

Expand source code

@classmethod
def get_factory(cls, factory: simulators.Factory) -> "SimulatorWrapper.Factory":
    """Wraps a simulator factory into an environment factory."""
    return SimulatorWrapper.Factory(factory)

Inherited members

• Base:
  • action_space
  • close
  • render
  • reset
  • step