API Reference

API Reference#

Below you’ll find the public Python API, auto‑generated from docstrings. The summaries link to detailed pages for each class or module.

`pnpl.datasets.libribrain2025.dataset.LibriBrain`	LibriBrain MEG dataset for speech and language tasks.
`pnpl.datasets.libribrain2025.compat.LibriBrainSpeech`	Speech detection dataset wrapper.
`pnpl.datasets.libribrain2025.compat.LibriBrainPhoneme`	Phoneme classification dataset wrapper.
`pnpl.datasets.libribrain2025.compat.LibriBrainWord`	Word classification dataset wrapper.
`pnpl.datasets.libribrain2025.sentence_dataset.LibriBrainSentence`
`pnpl.datasets.libribrain100.dataset.LibriBrain100`	Task-driven dataset for the full LibriBrain100 release.
`pnpl.datasets.libribrain100.compat.LibriBrain100Speech`	Speech-vs-silence binary classification on LibriBrain100.
`pnpl.datasets.libribrain100.compat.LibriBrain100Phoneme`	Multi-class phoneme classification on LibriBrain100.
`pnpl.datasets.libribrain100.compat.LibriBrain100Word`	Word classification / keyword detection on LibriBrain100.
`pnpl.datasets.gwilliams2022.dataset.Gwilliams2022`	MEG-MASC continuous MEG dataset.
`pnpl.datasets.armeni2022.dataset.Armeni2022`	Armeni 2022 continuous MEG dataset (CTF, audiobook listening).
`pnpl.datasets.schoffelen2019.dataset.Schoffelen2019`	Schöffelen 2019 (MOUS) continuous MEG dataset (CTF).
`pnpl.datasets.pallier2025.dataset.Pallier2025`	LittlePrince audiobook-listening continuous MEG dataset.
`pnpl.datasets.grouped_dataset.GroupedDataset`
`pnpl.datasets.hdf5.dataset.HDF5Dataset`	Torch Dataset for Serialized hdf5 files.

`pnpl.tasks.base.TaskProtocol`	Protocol defining what a Task must implement.
`pnpl.tasks.libribrain.SpeechDetection`	Binary speech vs silence classification task.
`pnpl.tasks.libribrain.PhonemeClassification`	Multi-class phoneme classification task.
`pnpl.tasks.libribrain.WordClassification`	Word classification task.
`pnpl.tasks.gwilliams2022.PhonemeClassification`	Multi-class phoneme classification on MEG-MASC.
`pnpl.tasks.gwilliams2022.WordClassification`	Word-onset classification on MEG-MASC.
`pnpl.tasks.armeni2022.PhonemeClassification`	Multi-class phoneme classification on Armeni 2022.
`pnpl.tasks.schoffelen2019.TrialEpoching`	Trial-onset epoching with a UPPT001 trigger-code label.
`pnpl.tasks.pallier2025.WordClassification`	Word-onset classification on Pallier 2025 (LittlePrince Listen).

`pnpl.preprocessing.Pipeline`	MEG preprocessing pipeline.
`pnpl.preprocessing.BadChannels`	Detect and mark bad channels.
`pnpl.preprocessing.HeadPosition`	Load cached head position data.
`pnpl.preprocessing.MaxwellFilter`	Apply Maxwell filter (Signal Space Separation).
`pnpl.preprocessing.NotchFilter`	Apply notch filter to remove line noise.
`pnpl.preprocessing.BandpassFilter`	Apply bandpass filter.
`pnpl.preprocessing.Downsample`	Downsample the data.
`pnpl.preprocessing.Epoch`	Create epochs from continuous data.
`pnpl.preprocessing.fif_to_h5`	Convert MNE Raw data to H5 format.
`pnpl.preprocessing.epochs_to_h5`	Convert MNE Epochs to H5 format.

`pnpl.datasets.mixins.HFDownloadMixin`	Mixin providing HuggingFace download functionality.
`pnpl.datasets.mixins.OSFDownloadMixin`	Mixin providing OSF download functionality.
`pnpl.datasets.mixins.RadboudDownloadMixin`	Mixin providing WebDAV-backed download from the Radboud Data Repository.
`pnpl.datasets.mixins.OpenNeuroDownloadMixin`	Mixin providing OpenNeuro download functionality.
`pnpl.datasets.mixins.BIDSMixin`	Mixin for loading raw BIDS MEG data.
`pnpl.datasets.mixins.ContinuousH5Mixin`	Mixin for loading continuous H5 MEG data.
`pnpl.datasets.mixins.EpochedH5Mixin`	Mixin for loading epoched H5 MEG data.
`pnpl.datasets.mixins.StandardizationMixin`	Mixin providing standardization functionality for MEG data.

Constants for LibriBrain dataset.