d3rlpy.datasets.get_dataset

d3rlpy.datasets.get_dataset(env_name, transition_picker=None, trajectory_slicer=None, render_mode=None)

Returns dataset and envrironment by guessing from name.

This function returns dataset by matching name with the following datasets.

• cartpole-replay

• cartpole-random

• pendulum-replay

• pendulum-random

• d4rl-pybullet

• d4rl-atari

• d4rl

import d3rlpy

# cartpole dataset
dataset, env = d3rlpy.datasets.get_dataset('cartpole')

# pendulum dataset
dataset, env = d3rlpy.datasets.get_dataset('pendulum')

# d4rl-atari dataset
dataset, env = d3rlpy.datasets.get_dataset('breakout-mixed-v0')

# d4rl dataset
dataset, env = d3rlpy.datasets.get_dataset('hopper-medium-v0')

Parameters

• env_name (str) – environment id of the dataset.

• transition_picker (Optional[d3rlpy.dataset.transition_pickers.TransitionPickerProtocol]) – TransitionPickerProtocol object.

• trajectory_slicer (Optional[d3rlpy.dataset.trajectory_slicers.TrajectorySlicerProtocol]) – TrajectorySlicerProtocol object.

• render_mode (Optional[str]) – Mode of rendering (human, rgb_array).

Returns

tuple of d3rlpy.dataset.ReplayBuffer and gym environment.

Return type

Tuple[d3rlpy.dataset.replay_buffer.ReplayBuffer, gym.core.Env[Any, Any]]