rlberry.envs.PipelineEnv

rlberry.envs.PipelineEnv(env_ctor, env_kwargs, wrappers)

Environment defined as a pipeline of wrappers and an environment to wrap.

Parameters:

env_ctor: environment class

env_kwargs: dictionary

kwargs fed to the environment

wrappers: list of tuple (wrapper, wrapper_kwargs)

list of tuple (wrapper, wrapper_kwargs) to be applied to the environment. The list [wrapper1, wrapper2] will be applied in the order wrapper1(wrapper2(env))

Examples

>>> from rlberry.envs import PipelineEnv
>>> from rlberry.envs import gym_make
>>> from rlberry.wrappers import RescaleRewardWrapper
>>>
>>> env_ctor, env_kwargs = PipelineEnv, {
>>>     "env_ctor": gym_make,
>>>     "env_kwargs": {"id": "Acrobot-v1"},
>>>     "wrappers": [(RescaleRewardWrapper, {"reward_range": (0, 1)})],
>>> }
>>> eval_env = (gym_make, {"id":"Acrobot-v1"}) # unscaled env for evaluation