evolvepy.integrations.gym package

Submodules

evolvepy.integrations.gym.gym module

class evolvepy.integrations.gym.gym.GymFitnessFunction(env_name: str, show: bool = False, save: bool = False)

Bases: ProcessFitnessFunction, ABC

Evaluates individuals using a Gym. Must be inherited by the user.

Can be used with ProcessEvaluator. As the sum of the rewards obtained by the individual during the evaluation.

__init__(env_name: str, show: bool = False, save: bool = False) → None

GymFitnessFunction constructor.

Parameters:

• env_name (str) – Name of the environment that will be used for the evaluation.

• show (bool, optional) – Whether to show the graphical output of the environment. Defaults to False.

• save (bool, optional) – Whether to save the graphical output in a file. Defaults to False.

abstract behaviour(obs: object, individual: ndarray) → object

Individual behavior. Receives the observation and returns the action.

Must be implemented by the user.

Parameters:

• obs (object) – Environment observation.

• individual (np.ndarray) – Individual being evaluated.

Returns:

Individual action.

Return type:

object

evaluate(individuals: ndarray) → ndarray

Evaluates the individual through the environment.

Parameters:

individuals (np.ndarray) – Individuals to be evaluated. Evaluates only the first individual.

Returns:

Individuals scores.

Return type:

np.ndarray

setup() → None

Initializes the environment.

Module contents

EvolvePy’s integration with OpenAi’s Gym (https://gym.openai.com/).