Skip to main content

MultitaperPsd

Namespace: RTPROCESSINGLIB  ·  Library: DSP Library

Python equivalent

mne.time_frequency.psd_multitaper in MNE-Python.

#include <dsp/multitaper_psd.h>

class UTILSLIB::MultitaperPsd

Multitaper power spectral density estimator using DPSS (Slepian) tapers.

Applies multiple orthogonal DPSS tapers to the data, computes the FFT of each tapered segment, and averages the resulting periodograms weighted by the taper eigenvalues. This provides a PSD estimate with reduced variance compared to a single-taper (periodogram) approach.

// 600 Hz data, half-bandwidth 4, default tapers
MultitaperPsdResult r = MultitaperPsd::compute(matData, 600.0);
// r.matPsd   → n_channels × (n_times/2+1)
// r.vecFreqs → frequency axis in Hz

Static Methods​

compute(matData, sfreq, halfBandwidth, nTapers, nFft)​

Compute multitaper PSD for every channel of a data matrix.

Parameters:

  • matData : const Eigen::MatrixXd & Data matrix (n_channels × n_times).

  • sfreq : double Sampling frequency in Hz.

  • halfBandwidth : double Half-bandwidth parameter (NW); default 4.0.

  • nTapers : int Number of DPSS tapers; -1 → floor(2*halfBandwidth - 1).

  • nFft : int FFT length; -1 → n_times.

Returns:

  • MultitaperPsdResult — MultitaperPsdResult with matPsd and vecFreqs.

Authors of this file​