mne_denoise.qa.peak_attenuation_db#

mne_denoise.qa.peak_attenuation_db(freqs: ndarray, psd_before: ndarray, psd_after: ndarray, target_freq: float, bandwidth: float = 2.0) → ndarray[source]#

Attenuation (dB) of the dominant peak around a target frequency.

Parameters:

  • freqs (array of shape (n_freqs,)) – Frequency vector.

  • psd_before (array of shape (n_channels, n_freqs) or (n_freqs,)) – PSD before cleaning.

  • psd_after (array of shape (n_channels, n_freqs) or (n_freqs,)) – PSD after cleaning.

  • target_freq (float) – Centre frequency of the peak (Hz).

  • bandwidth (float) – Half-bandwidth (Hz) around target_freq to search for the peak.

Returns:

attenuation – Per-channel attenuation in dB for 2D PSD input, or a scalar value for 1D PSD input.

Return type:

ndarray | float

Notes

This metric compares the maximum PSD value in a narrow band around target_freq:

10 * log10(max_before / max_after)

Positive values indicate suppression of the target peak.

Examples

>>> import numpy as np
>>> from mne_denoise.qa import peak_attenuation_db
>>> freqs = np.arange(0, 100, 0.5)
>>> before = np.ones_like(freqs) * 0.01
>>> after = before.copy()
>>> band = (freqs >= 49) & (freqs <= 51)
>>> before[band] = 1.0
>>> after[band] = 0.5
>>> float(peak_attenuation_db(freqs, before, after, 50.0)) > 0
True