Source code for minari.data_collector.callbacks.episode_metadata
from typing import Dict
import numpy as np
[docs]
class EpisodeMetadataCallback:
"""Callback to full episode after saving to hdf5 file as a group.
This callback can be overridden to add extra metadata attributes or statistics to
each episode in the Minari dataset. The custom callback can then be
passed to the DataCollector wrapper to the `episode_metadata_callback` argument.
TODO: add more default statistics to episode datasets
"""
def __call__(self, episode: Dict):
"""Callback method.
Override this method to add custom attribute metadata to the episode group.
Args:
episode (dict): the dict that contains an episode's data
"""
return {
"rewards_sum": np.sum(episode["rewards"]),
"rewards_mean": np.mean(episode["rewards"]),
"rewards_std": np.std(episode["rewards"]),
"rewards_max": np.max(episode["rewards"]),
"rewards_min": np.min(episode["rewards"]),
}