BIDS Dataset Example

In this example, we show how to fetch and prepare a BIDS dataset for usage with Braindecode.

# Authors: Pierre Guetschel <pierre.guetschel@gmail.com>
#
# License: BSD (3-clause)

from pathlib import Path

import openneuro

from braindecode.datasets import BIDSDataset

First, we download a collection of (fake/empty) BIDS datasets.

# import tempfile
# data_dir = tempfile.mkdtemp()
data_dir = Path("~/mne_data/openneuro/").expanduser()
dataset_name = "ds004745"  # 200Mb dataset
dataset_root = data_dir / dataset_name

if not dataset_root.exists():
    openneuro.download(dataset=dataset_name, target_dir=dataset_root)

Now, loading the dataset is simply a one-line command:

bids_ds = BIDSDataset(dataset_root)

And we can see that the events of this dataset are set in the .annotations attribute of the raw data:

print(bids_ds.datasets[0].raw.annotations)

<Annotations | 54 segments: 1 (6), 10 (6), 2 (6), 3 (6), 4 (6), 5 (6), 6 ...>

Estimated memory usage: 854 MB