mne_denoise.spectrum_interpolation.SpectrumInterpolation#

Line-noise remover based on amplitude spectrum interpolation.

Parameters:

sfreqfloat or None, default=None: Sampling frequency in Hz; inferred from MNE metadata when available.
line_freqfloat or array-like, default=50.0: Fundamental frequency or explicit target frequencies in Hz.
n_harmonicsint or None, default=None: Number of harmonics when line_freq is scalar; None uses all harmonics below Nyquist.
bandwidthfloat, default=1.0: Half-width of each replaced band in Hz.
neighbour_widthfloat, default=2.0: Width of each neighboring reference band in Hz.
verbosebool, str, int, or None, default=None: Logging level.

Attributes:

sfreq_float: Sampling frequency used during fit.
freqs_ndarray: Resolved target frequencies.

See also

interpolate_spectrum: One-shot array interface.
mne_denoise.zapline.ZapLine: DSS-based spatial line-noise removal.

Notes

NumPy input uses 2-D or 3-D channel-first layouts; 3-D records are processed independently. MNE data channels are processed while non-data channels and container metadata are preserved. Short segments may have limited spectral resolution [1].

References

Examples

>>> import numpy as np
>>> from mne_denoise.spectrum_interpolation import SpectrumInterpolation
>>> rng = np.random.default_rng(0)
>>> data = rng.standard_normal((8, 2000))
>>> model = SpectrumInterpolation(sfreq=250.0, line_freq=60.0, n_harmonics=2)
>>> clean = model.fit_transform(data)

fit(X: Any, y: Any = None, *, verbose: bool | str | int | None = None) → SpectrumInterpolation[source]#

Resolve the sampling frequency and target frequencies.

Parameters:

XRaw, Epochs, Evoked, or ndarray: Input whose metadata or shape is inspected.
yNone, default=None: Ignored for scikit-learn compatibility.
verbosebool, str, int, or None, default=None: Logging level.

Returns:

SpectrumInterpolation: The fitted estimator.

fit_transform(X: Any, y: Any = None, *, verbose: bool | str | int | None = None, **fit_params: Any) → Any[source]#

Fit spectrum interpolation and transform X.

Parameters:

XRaw, Epochs, Evoked, or ndarray: Data to clean.
yNone, default=None: Ignored for scikit-learn compatibility.
verbosebool, str, int, or None, default=None: Logging level.
**fit_paramsdict: Ignored for scikit-learn compatibility.

Returns:

same type as X: Cleaned data.

transform(X: Any, *, verbose: bool | str | int | None = None) → Any[source]#

Apply spectrum interpolation.

Parameters:

XRaw, Epochs, Evoked, or ndarray: Data to clean.
verbosebool, str, int, or None, default=None: Logging level.

Returns:

same type as X: Cleaned data with the same shape.