neuralfetch.utils.bids.BidsExporter¶

pydantic model neuralfetch.utils.bids.BidsExporter[source][source]¶

Export a Neuralset Study to BIDS format, optionally in parallel via SLURM.

Parameters:

path – Root directory for the BIDS output.
device – Neurophysiology recording type. Must be one of "Eeg", "Meg", "Ieeg", "Emg", or "Fnirs".
task – BIDS task label. If None, the "task" column in the events DataFrame is used.
anonymize – Passed to mne_bids.write_raw_bids. Requires a daysback key. If None, no anonymization is performed.
overwrite – If True, overwrite existing BIDS files.
infra_bids – Caching/compute backend for per-timeline BIDS writes. Uses a local process pool by default; set cluster="slurm" (with a folder and SLURM parameters) to parallelise across cluster nodes.

Examples

Sequential (default):

BidsExporter(path="/data/bids", device="Meg", task="mytask").export(study)

Parallel SLURM:

BidsExporter(
    path="/data/bids",
    device="Meg",
    task="mytask",
    infra_bids=exca.MapInfra(
        cluster="slurm",
        folder="/tmp/bids_jobs",
        slurm_partition="learnfair",
        mem_gb=64,
        timeout_min=60,
    ),
).export(study)

Fields:

anonymize (dict[str, Any] | None)
device (str)
infra_bids (exca.map.MapInfra)
overwrite (bool)
path (pathlib.Path)
task (str | None)

field path: Path [Required][source]¶

field device: str [Required][source]¶

field task: str | None = None[source]¶

field anonymize: dict[str, Any] | None = None[source]¶

field overwrite: bool = False[source]¶

field infra_bids: MapInfra = MapInfra(cluster='processpool')[source]¶

export(study: Study) → Path[source][source]¶

Run the full BIDS export for study.

Returns the BIDS root path.

← Back to API reference