mtspec.mt_deconvolve¶
-
mtspec.multitaper.mt_deconvolve(data_a, data_b, delta, nfft=None, time_bandwidth=None, number_of_tapers=None, weights='adaptive', demean=True, fmax=0.0)[source]¶ Deconvolve two time series using multitapers.
This uses the eigencoefficients and the weights from the multitaper spectral estimations and more or less follows this paper:
Receiver Functions from Multiple-Taper Spectral Correlation Estimates Jeffrey Park, Vadim Levin
Bulletin of the Seismological Society of America Dec 2000, 90 (6) 1507-1520 http://dx.doi.org/10.1785/0119990122Parameters: - data_a (
numpy.ndarray) – Data for first time series. - data_b (
numpy.ndarray) – Data for second time series. - delta (float) – Sample spacing of the data.
- nfft (int) – Number of points for the FFT. If
nfft == None, no zero padding will be applied before the FFT. - time_bandwidth (float) – Time-bandwidth product. Common values are 2, 3, 4, and numbers in between.
- number_of_tapers (int) – Number of tapers to use. Defaults to
int(2*time_bandwidth) - 1. This is maximum senseful amount. More tapers will have no great influence on the final spectrum but increase the calculation time. Use fewer tapers for a faster calculation. - weights (str) –
"adaptive"or"constant"weights. - demean – Force the complex TF to be demeaned.
- fmax (float) – Maximum frequency for lowpass cosine filter. Set this to zero to not have a filter.
Returns: Returns a dictionary with 5
numpy.ndarray’s. See the note below.Note
Returns a dictionary with five arrays:
"deconvolved": Deconvolved time series."spectrum_a": Spectrum of the first time series."spectrum_b": Spectrum of the second time series."spectral_ratio": The ratio of both spectra."frequencies": The used frequency bins for the spectra.
- data_a (
