Bundled Plugins and Extensions

EEGPrep ships Python ports of several EEGLAB plugin workflows. They are bundled with the package, exposed in the GUI, available from eegprep-console, and documented in the API reference. External extensions use a separate Python package registry described in External Extensions.

Bundled Plugin Inventory

Plugin	Main entry points	Implemented workflow
clean_rawdata	pop_clean_rawdata, clean_artifacts, clean_asr, vis_artifacts	Drift, flatline, noisy-channel, ASR, window cleaning, and visual artifact review for continuous data.
FIRFilt	pop_eegfiltnew, pop_firws, pop_firpm, pop_firma, pop_xfirws	FIR filtering, filter-order helpers, response reports, and boundary-aware filtering.
ICLabel	pop_iclabel, pop_icflag, eeg_icalabelstat, pop_viewprops, pop_prop_extended	Default ICLabel network classification, threshold flagging, component dashboards, and label statistics.
DIPFIT	pop_dipfit_settings, pop_dipfit_gridsearch, pop_dipfit_nonlinear, pop_multifit, pop_dipplot, pop_leadfield	Standalone spherical source-localization workflows for ICA components and explicit limits for unavailable external backends.
EEG-BIDS	pop_importbids, pop_exportbids, pop_load_frombids, bids_preproc, bids_list_eeg_files	BIDS import/export, metadata loading, and batch preprocessing helpers.

Inspect Installed Plugin Metadata

From normal Python or eegprep-console:

import eegprep

for plugin in eegprep.bundled_plugins():
    print(plugin["plugin"], plugin["funcname"], plugin["menu"])

status, names, records = eegprep.plugin_status("ICLabel", exactmatch=True)
print(status, names)

inventory = eegprep.plugin_menu(show=False)
print(eegprep.format_plugin_menu())

The Extension Manager dialog uses the same metadata and adds curated catalog records when available.

clean_rawdata and ASR

Use the pop_* wrapper when you want GUI behavior or command history:

from eegprep import pop_clean_rawdata

EEG, com = pop_clean_rawdata(
    EEG,
    FlatlineCriterion=5,
    ChannelCriterion=0.8,
    LineNoiseCriterion=4,
    Highpass=(0.25, 0.75),
    BurstCriterion=20,
    WindowCriterion=0.25,
    return_com=True,
)

Use clean_artifacts when you need lower-level tuple outputs:

from eegprep import clean_artifacts

clean_eeg, highpass_state, burst_state, removed_channels = clean_artifacts(EEG)

Channel correlation cleaning uses the spatial reconstruction method when the dataset has usable X, Y, and Z channel coordinates. If channel locations are empty, pop_clean_rawdata automatically uses the bundled location-free correlation method instead. The RANSAC subset size is round(subset_size * nbchan) channels, and a larger count raises ValueError. The line-noise filter is a short FIR, so a NaN stays in a neighborhood of that channel instead of removing the channel from every correlation.

clean_windows keeps the removal duration from eeg_eegrej and sets xmax to xmin + (pnts - 1) / srate only when that duration is off the sample grid. A recording whose fit widths round to 1 or less raises ValueError, and clean_asr calibrates on the whole recording instead. Eleven to twenty-two windows still fit.

FIRFilt

pop_eegfiltnew is the usual filtering entry point. When the filter order is left at its default, the transition width is recomputed from that even order (3.3 / order * srate) so the reported cutoff matches the coefficients:

EEG, com = pop_eegfiltnew(EEG, locutoff=1, hicutoff=40, plotfreqz=False, return_com=True)

Use the lower-level FIRFilt helpers for custom order/window work:

beta = pop_kaiserbeta(0.001)
order = pop_firwsord("kaiser", EEG["srate"], 2, 0.001)
EEG, com = pop_firws(
    EEG,
    fcutoff=[1, 40],
    ftype="bandpass",
    wtype="kaiser",
    warg=beta,
    forder=order,
    return_com=True,
)

ICLabel

The standalone Python engine ships the default ICLabel network. lite and beta network requests require MATLAB or Octave with an EEGLAB ICLabel checkout that provides those artifacts.

EEG, com = pop_iclabel(EEG, "default", return_com=True)
stats = eeg_icalabelstat(EEG, threshold=0.9, verbose=False)
EEG, com = pop_icflag(EEG, return_com=True)

Review components visually with pop_viewprops before removing them.

DIPFIT

DIPFIT helpers support standalone spherical workflows. They store model fields under EEG["dipfit"]["model"] and keep command strings replayable:

EEG, com = pop_dipfit_settings(EEG, model="standardBESA", return_com=True)
EEG, com = pop_multifit(EEG, [1, 2, 3], threshold=40, return_com=True)

Unavailable MRI/BEM, AFNI, LORETA, and STUDY-level source-statistics workflows raise explicit errors.

EEG-BIDS

Use pop_load_frombids for one EEG recording inside a BIDS dataset and bids_preproc for batch workflows:

EEG, report = pop_load_frombids(
    "sub-01/eeg/sub-01_task-rest_eeg.set",
    return_report=True,
)

See BIDS Workflow for BIDS dataset layout and batch notes.

External Extensions

External EEGPrep extensions are Python packages, not MATLAB plugin folders. They declare an eegprep.extensions entry point and return an ExtensionSpec. The Extension Manager shows installed and curated extension metadata but does not install, update, remove, unzip, or execute package-manager commands for the user.

Continue with External Extensions for package authoring and Extension Curation and Catalog Policy for catalog/trust policy.