mne_denoise.dss.CombFilterBias

class mne_denoise.dss.CombFilterBias(fundamental_freq: float, sfreq: float, *, n_harmonics: int = 3, q_factor: float = 30.0, q_mode: str = 'fixed', weights: ndarray | None = None)

Comb-filter bias for harmonic frequencies.

Parameters:

fundamental_freqfloat

Fundamental frequency in Hz.

sfreqfloat

Sampling frequency in Hz.

n_harmonicsint, default=3

Number of harmonics to consider.

q_factorfloat, default=30.0

Quality factor passed to each peak filter.

q_mode{“fixed”, “proportional”}, default=”fixed”

With "fixed", Q is constant and bandwidth = frequency / Q therefore increases in absolute width at higher harmonics. With "proportional", Q is multiplied by the harmonic number, giving approximately constant absolute bandwidth.

weightsarray-like or None, default=None

Harmonic weights. None uses inverse harmonic-number weights.

Notes

The implementation sums peak-filter outputs and excludes harmonics at or above 95 percent of Nyquist.

apply(data: ndarray) → ndarray

Apply the summed harmonic peak-filter bias.

Parameters:

datandarray, shape (n_channels, n_times) or (n_channels, n_times, n_epochs)

Channel-first data.

Returns:

ndarray

Weighted harmonic content with the input shape.

property harmonic_frequencies: list[float]

Return list of harmonic frequencies being filtered.