braindecode 1.3.0.dev180329405__py3-none-any.whl → 1.3.0.dev182330353__py3-none-any.whl

braindecode/modules/layers.py

@@ -70,26 +70,27 @@ class TimeDistributed(nn.Module):
 class DropPath(nn.Module):
     """Drop paths, also known as Stochastic Depth, per sample.
 
-        When applied in main path of residual blocks.
+    When applied in main path of residual blocks.
 
-        Parameters:
-        -----------
-        drop_prob: float (default=None)
-            Drop path probability (should be in range 0-1).
+    Parameters
+    ----------
+    drop_prob: float (default=None)
+        Drop path probability (should be in range 0-1).
 
-        Notes
-        -----
-        Code copied and modified from VISSL facebookresearch:
+    Notes
+    -----
+    Code copied and modified from VISSL facebookresearch:
     https://github.com/facebookresearch/vissl/blob/0b5d6a94437bc00baed112ca90c9d78c6ccfbafb/vissl/models/model_helpers.py#L676
-        All rights reserved.
-
-        THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
-        IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
-        FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
-        AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
-        LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
-        OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
-        SOFTWARE.
+
+    All rights reserved.
+
+    THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+    IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+    FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+    AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+    LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+    OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+    SOFTWARE.
     """
 
     def __init__(self, drop_prob=None):

braindecode/preprocessing/__init__.py

@@ -1,10 +1,185 @@
+"""Preprocessing module for EEG data.
+
+This module provides preprocessing functionality for EEG data through:
+
+1. **MNE-based preprocessing classes**: Automatically generated classes that wrap
+   MNE-Python functions for easy integration with Braindecode workflows.
+2. **Custom preprocessing functions**: Specialized functions for EEG preprocessing.
+3. **Windowing functions**: Tools for creating epochs/windows from continuous data.
+
+MNE Preprocessing Classes
+--------------------------
+The following preprocessing classes are automatically generated from MNE functions.
+Each class inherits from :class:`~braindecode.preprocessing.Preprocessor` and can be
+used with :func:`~braindecode.preprocessing.preprocess`.
+
+**Signal Processing**
+
+- :class:`~braindecode.preprocessing.Resample` : Resample data to different sampling frequency
+- :class:`~braindecode.preprocessing.Filter` : Apply bandpass, highpass, or lowpass filter
+- :class:`~braindecode.preprocessing.FilterData` : Low-level filter function for data arrays
+- :class:`~braindecode.preprocessing.NotchFilter` : Remove specific frequencies (e.g., 50/60 Hz power line noise)
+- :class:`~braindecode.preprocessing.SavgolFilter` : Apply Savitzky-Golay polynomial filter
+- :class:`~braindecode.preprocessing.ApplyHilbert` : Compute analytic signal or envelope
+- :class:`~braindecode.preprocessing.Rescale` : Rescale channel amplitudes
+- :class:`~braindecode.preprocessing.OversampledTemporalProjection` : Apply oversampled temporal projection
+
+**Channel Management**
+
+- :class:`~braindecode.preprocessing.Pick` : Select specific channels or channel types
+- :class:`~braindecode.preprocessing.PickChannels` : Pick channels by name
+- :class:`~braindecode.preprocessing.PickTypes` : Pick channels by type (EEG, MEG, etc.)
+- :class:`~braindecode.preprocessing.DropChannels` : Remove specific channels
+- :class:`~braindecode.preprocessing.AddChannels` : Append new channels from other MNE objects
+- :class:`~braindecode.preprocessing.CombineChannels` : Combine data from multiple channels
+- :class:`~braindecode.preprocessing.RenameChannels` : Rename channels
+- :class:`~braindecode.preprocessing.ReorderChannels` : Reorder channels
+- :class:`~braindecode.preprocessing.SetChannelTypes` : Specify sensor types of channels
+- :class:`~braindecode.preprocessing.InterpolateBads` : Interpolate bad channels
+- :class:`~braindecode.preprocessing.InterpolateTo` : Interpolate EEG data onto new montage
+- :class:`~braindecode.preprocessing.InterpolateBridgedElectrodes` : Interpolate bridged electrodes
+- :class:`~braindecode.preprocessing.ComputeBridgedElectrodes` : Identify bridged electrodes
+- :class:`~braindecode.preprocessing.EqualizeChannels` : Make channel sets identical across datasets
+- :class:`~braindecode.preprocessing.EqualizeBads` : Equalize bad channels across instances
+- :class:`~braindecode.preprocessing.FindBadChannelsLof` : Find bad channels using LOF algorithm
+
+**Reference & Montage**
+
+- :class:`~braindecode.preprocessing.SetEEGReference` : Specify EEG reference (Raw method)
+- :class:`~braindecode.preprocessing.SetBipolarReference` : Set bipolar reference
+- :class:`~braindecode.preprocessing.AddReferenceChannels` : Add zero-filled reference channels
+- :class:`~braindecode.preprocessing.SetMontage` : Set channel positions/montage
+
+**SSP Projections**
+
+- :class:`~braindecode.preprocessing.AddProj` : Add SSP projection vectors
+- :class:`~braindecode.preprocessing.ApplyProj` : Apply SSP operators
+- :class:`~braindecode.preprocessing.DelProj` : Remove SSP projection vector
+
+**Data Transformation**
+
+- :class:`~braindecode.preprocessing.Crop` : Crop data to specific time range
+- :class:`~braindecode.preprocessing.CropByAnnotations` : Crop data by annotations
+- :class:`~braindecode.preprocessing.ComputeCurrentSourceDensity` : Apply CSD transformation
+- :class:`~braindecode.preprocessing.FixStimArtifact` : Remove stimulation artifacts
+- :class:`~braindecode.preprocessing.MaxwellFilter` : Apply Maxwell filtering (for MEG data)
+- :class:`~braindecode.preprocessing.RealignRaw` : Realign raw data
+- :class:`~braindecode.preprocessing.RegressArtifact` : Regress out artifacts
+
+**Artifact Detection & Annotation**
+
+- :class:`~braindecode.preprocessing.AnnotateAmplitude` : Annotate periods based on amplitude
+- :class:`~braindecode.preprocessing.AnnotateBreak` : Annotate breaks in the data
+- :class:`~braindecode.preprocessing.AnnotateMovement` : Annotate movement artifacts
+- :class:`~braindecode.preprocessing.AnnotateMuscleZscore` : Annotate muscle artifacts using z-score
+- :class:`~braindecode.preprocessing.AnnotateNan` : Annotate NaN values in data
+
+**Metadata & Configuration**
+
+- :class:`~braindecode.preprocessing.Anonymize` : Anonymize measurement information
+- :class:`~braindecode.preprocessing.SetAnnotations` : Set annotations
+- :class:`~braindecode.preprocessing.SetMeasDate` : Set measurement start date
+- :class:`~braindecode.preprocessing.AddEvents` : Add events to stim channel
+- :class:`~braindecode.preprocessing.FixMagCoilTypes` : Fix Elekta magnetometer coil types
+- :class:`~braindecode.preprocessing.ApplyGradientCompensation` : Apply CTF gradient compensation
+
+Usage Examples
+--------------
+Using the new preprocessing classes::
+
+    from braindecode.preprocessing import (
+        Resample, Filter, NotchFilter, SetEEGReference, preprocess
+    )
+
+    preprocessors = [
+        Resample(sfreq=100),
+        NotchFilter(freqs=[50]),  # Remove 50 Hz power line noise
+        Filter(l_freq=4, h_freq=30),  # Bandpass filter
+        SetEEGReference(ref_channels='average'),
+    ]
+    preprocess(dataset, preprocessors)
+
+Using the generic Preprocessor class (legacy approach)::
+
+    from braindecode.preprocessing import Preprocessor, preprocess
+
+    preprocessors = [
+        Preprocessor('resample', sfreq=100),
+        Preprocessor('filter', l_freq=4, h_freq=30),
+    ]
+    preprocess(dataset, preprocessors)
+
+See Also
+--------
+:class:`braindecode.preprocessing.Preprocessor` : Base class for all preprocessors
+:func:`braindecode.preprocessing.preprocess` : Apply preprocessors to datasets
+:func:`braindecode.preprocessing.create_windows_from_events` : Create epochs from events
+:func:`braindecode.preprocessing.create_fixed_length_windows` : Create fixed-length epochs
+"""
+
+from .eegprep_preprocess import (
+    EEGPrep,
+    ReinterpolateRemovedChannels,
+    RemoveBadChannels,
+    RemoveBadChannelsNoLocs,
+    RemoveBadWindows,
+    RemoveBursts,
+    RemoveCommonAverageReference,
+    RemoveDCOffset,
+    RemoveDrifts,
+    RemoveFlatChannels,
+    Resampling,
+)
 from .mne_preprocess import (  # type: ignore[attr-defined]
+    AddChannels,
+    AddEvents,
+    AddProj,
+    AddReferenceChannels,
+    AnnotateAmplitude,
+    AnnotateBreak,
+    AnnotateMovement,
+    AnnotateMuscleZscore,
+    AnnotateNan,
+    Anonymize,
+    ApplyGradientCompensation,
+    ApplyHilbert,
+    ApplyProj,
+    CombineChannels,
+    ComputeBridgedElectrodes,
+    ComputeCurrentSourceDensity,
     Crop,
+    CropByAnnotations,
+    DelProj,
     DropChannels,
+    EqualizeBads,
+    EqualizeChannels,
     Filter,
+    FilterData,
+    FindBadChannelsLof,
+    FixMagCoilTypes,
+    FixStimArtifact,
+    InterpolateBads,
+    InterpolateBridgedElectrodes,
+    InterpolateTo,
+    MaxwellFilter,
+    NotchFilter,
+    OversampledTemporalProjection,
     Pick,
+    PickChannels,
+    PickTypes,
+    RealignRaw,
+    RegressArtifact,
+    RenameChannels,
+    ReorderChannels,
     Resample,
+    Rescale,
+    SavgolFilter,
+    SetAnnotations,
+    SetBipolarReference,
+    SetChannelTypes,
     SetEEGReference,
+    SetMeasDate,
+    SetMontage,
 )
 from .preprocess import (
     Preprocessor,
@@ -25,12 +200,66 @@ __all__ = [
     "filterbank",
     "preprocess",
     "Preprocessor",
-    "Resample",
+    "AddChannels",
+    "AddEvents",
+    "AddProj",
+    "AddReferenceChannels",
+    "Anonymize",
+    "AnnotateAmplitude",
+    "AnnotateBreak",
+    "AnnotateMovement",
+    "AnnotateMuscleZscore",
+    "AnnotateNan",
+    "ApplyGradientCompensation",
+    "ApplyHilbert",
+    "ApplyProj",
+    "CombineChannels",
+    "ComputeBridgedElectrodes",
+    "ComputeCurrentSourceDensity",
+    "Crop",
+    "CropByAnnotations",
+    "DelProj",
     "DropChannels",
-    "SetEEGReference",
+    "EqualizeBads",
+    "EqualizeChannels",
     "Filter",
+    "FilterData",
+    "FindBadChannelsLof",
+    "FixMagCoilTypes",
+    "FixStimArtifact",
+    "InterpolateBads",
+    "InterpolateBridgedElectrodes",
+    "InterpolateTo",
+    "MaxwellFilter",
+    "NotchFilter",
+    "OversampledTemporalProjection",
     "Pick",
-    "Crop",
+    "PickChannels",
+    "PickTypes",
+    "RealignRaw",
+    "RegressArtifact",
+    "RenameChannels",
+    "ReorderChannels",
+    "Resample",
+    "Rescale",
+    "SavgolFilter",
+    "SetAnnotations",
+    "SetBipolarReference",
+    "SetChannelTypes",
+    "SetEEGReference",
+    "SetMeasDate",
+    "SetMontage",
+    "EEGPrep",
+    "RemoveDCOffset",
+    "Resampling",
+    "RemoveFlatChannels",
+    "RemoveDrifts",
+    "RemoveBadChannels",
+    "RemoveBadChannelsNoLocs",
+    "RemoveBursts",
+    "RemoveBadWindows",
+    "ReinterpolateRemovedChannels",
+    "RemoveCommonAverageReference",
     "create_windows_from_events",
     "create_fixed_length_windows",
     "create_windows_from_target_channels",