biosip_tools.eeg package

Submodules

biosip_tools.eeg.constants module

biosip_tools.eeg.timeseries module

class biosip_tools.eeg.timeseries.EEGSeries(data: Optional[numpy.ndarray] = None, path: Optional[str] = None, sample_rate: int = 500)

Bases: object

Class for EEG time series.

Parameters

  • data (np.ndarray) – EEG data.

  • path (str) – Path to .npy array. Expected shape is (n_subjects, n_channels, n_samples)

  • sample_rate (int, optional) – sample rate, defaults to 500

append(new_data) → None

Append new data to the EEG data.

Parameters

new_data (EEGSeries) – Data to append.

apply_cheby_filter(lowcut: float, highcut: float, order: int = 6, rs: float = 40, plot_response=False)

Apply a Chebyshev II filter to the EEG data.

Parameters

  • lowcut (float) – Lower pass-band edge.

  • highcut (float) – Upper pass-band edge.

  • order (int, optional) – [description], defaults to 6

  • rs (float, optional) – [description], defaults to 40

  • plot_response (bool, optional) – [description], defaults to False

Returns

Filtered data

Return type

EEGSeries

cheby_filter_bands(**kwargs) → dict

Return a dictionary of filtered EEG data.

Returns

Dictionary of filtered data. {band_name: data}

Return type

dict

fir_filter(l_freq: float, h_freq: float, verbose=False, **kwargs) → numpy.ndarray

Apply a FIR filter to the EEG data. Accepts arguments for mne.filter.filter_data.

Parameters

  • l_freq (float) – Lower pass-band edge.

  • h_freq (float) – Upper pass-band edge.

Returns

Filtered data

Return type

EEGSeries

fir_filter_bands(**kwargs) → dict

Return a dictionary of filtered EEG data.

Returns

Dictionary of filtered data. {band_name: data}

Return type

dict

biosip_tools.eeg.utils module

biosip_tools.eeg.utils.window_data_loader(eegs: biosip_tools.eeg.timeseries.EEGSeries, batch_size: int = 32, window_size: float = 1, infinity: bool = False, labels: Optional[numpy.ndarray] = None, epochs=1, shuffle: bool = False, return_subjects=False, stride: Optional[float] = None) → tuple

[summary]

Parameters

  • eegs (EEGSeries) – [description]

  • batch_size (int, optional) – batch size, defaults to 32

  • window_size (float, optional) – eeg window in seconds, defaults to 1

  • infinity (bool, optional) – whether the loop should be infinity, defaults to False

  • labels (np.ndarray, optional) – Labels per subject, defaults to None

  • epochs (int, optional) – number of epochs, defaults to 1

  • shuffle (bool, optional) – whether to shuffle the data, defaults to False

  • return_subjects (bool, optional) – batch size correspond to number of windows per subject, defaults to False

  • stride (float, optional) – stride in seconds, defaults to None

Yield

batch and labels if labels are provided

Return type

tuple

Module contents