meeg_utils.Dataset

class meeg_utils.Dataset(root, *, datatype=None, subjects=None, sessions=None, tasks=None, runs=None, exclude=())

The MEG/EEG recordings of a BIDS dataset, selected by entity.

Parameters:

• root (str | Path) – Root of the BIDS dataset.

• datatype ({"meg", "eeg"} | None) – Keep only this datatype.

• subjects (list of str | None) – Keep only these labels (without the sub- etc. prefixes). Empty-room recordings (sub-emptyroom) are left out unless "emptyroom" is listed.

• sessions (list of str | None) – Keep only these labels (without the sub- etc. prefixes). Empty-room recordings (sub-emptyroom) are left out unless "emptyroom" is listed.

• tasks (list of str | None) – Keep only these labels (without the sub- etc. prefixes). Empty-room recordings (sub-emptyroom) are left out unless "emptyroom" is listed.

• runs (list of str | None) – Keep only these labels (without the sub- etc. prefixes). Empty-room recordings (sub-emptyroom) are left out unless "emptyroom" is listed.

• exclude (list of str) – Recordings to leave out: any whose file name contains one of these strings (e.g. ["sub-07", "task-rest_run-02"]).

recordings

The selected recordings, sorted.

Type:

list of BIDSPath

Examples

>>> ds = meu.Dataset("/data/bids", datatype="meg", tasks=["faces"])
>>> ds.groups()
{'sub-01_ses-01_task-faces_meg': [BIDSPath(...run-01...), BIDSPath(...run-02...)], ...}

__init__(root, *, datatype=None, subjects=None, sessions=None, tasks=None, runs=None, exclude=())

Methods

__init__(root, *[, datatype, subjects, ...])
groups()	Group the recordings that belong together: the runs of one session and task.
table()	One row per recording with its BIDS entities and file.

groups()

Group the recordings that belong together: the runs of one session and task.

Returns:

Group name (the file name without the run) mapped to its recordings, in run order.

Return type:

dict

table()

One row per recording with its BIDS entities and file.

Returns:

subject, session, task, acquisition, run, datatype, group and path.

Return type:

list of dict

__len__()

Return the number of recordings.

__iter__()

Iterate over the recordings.

__repr__()

Summarize the selection.