cogpy.spectral.multitaper

Multitaper spectral estimation utilities.

This module provides core DPSS/multitaper primitives and a multitaper spectrogram wrapper (Ghostipy backend) with Dask-friendly outer parallelism.

Status

STATUS: ACTIVE Reason: Core multitaper primitives in active use; legacy wrappers deprecated in favour of sliding_core. Superseded by: n/a Safe to remove: no

Functions

dpss_tapers(N[, NW, K_max])	Generate Discrete Prolate Spheroidal Sequences (DPSS) tapers.
mtm_kwarg_from_gsp(bandwidth, fs, nperseg, ...)	convert ghostipy mtm_spectrogram kwargs to NW, window_size, window_step
mtm_kwarg_to_gsp(NW, fs, window_size, ...)	convert NW, window_size, window_step to ghostipy mtm_spectrogram kwargs
mtm_spectrogram(x, bandwidth[, axis])	Compute a multitaper spectrogram (Thomson method) on a NumPy or Dask array by applying ghostipy.mtm_spectrogram independently along a chosen time axis.
mtm_spectrogramx(xsig[, dim])	multitaper spectrogram using ghostipy backend, wrapped as xarray DataArray
multitaper_fft(y[, axis, NW, nfft, K_max, ...])	multitaper density estimate
multitaper_psd(y[, axis, NW, nfft, K_max, ...])	multitaper power spectral density estimate
nperseg_from_ncycle(fm[, fs, ncycle, ...])	rel_nperseg: number of cycles per segment
take_first_fiber_along_axis(x[, axis])	Return the fiber of x along the given axis, fixing all other dimensions at index 0.