
| HOME | DOWNLOAD | DOCUMENTATION | FAQ |
> Home > Documentation > Python > Global and Localized Spectral Analysis
SHMultiTaperMaskCSE
Perform a localized multitaper cross-spectral analysis using arbitrary windows derived from a mask.
Usage
mtse, sd = SHMultiTaperMaskCSE (sh1, sh2, tapers, [lmax1, lmax2, lmaxt, k, taper_wt, norm, csphase])
Returns
mtse: float, dimension (lmax-lmaxt+1)- The localized multitaper cross-power spectrum estimate.
lmaxis the smaller oflmax1andlmax2. sd: float, dimension (lmax-lmaxt+1)- The standard error of the localized multitaper cross-power spectral estimates.
lmaxis the smaller oflmax1andlmax2.
Parameters
sh1: float, dimension (2,lmax1in+1,lmax1in+1)- The spherical harmonic coefficients of the first function.
sh2: float, dimension (2,lmax2in+1,lmax2in+1)- The spherical harmonic coefficients of the second function.
tapers: float, dimension ((lmaxtin+1)**2,kin)- An array of the
kwindowing functions, arranged in columns, obtained from a call toSHReturnTapersMap. The spherical harmonic coefficients are packed according to the conventions inSHCilmToVector. lmax1: optional, integer, default =lmax1in- The spherical harmonic bandwidth of
sh1. This must be less than or equal tolmax1in. lmax2: optional, integer, default =lmax2in- The spherical harmonic bandwidth of
sh2. This must be less than or equal tolmax2in. lmaxt: optional, integer, default =lmaxtin- The spherical harmonic bandwidth of the windowing functions in the array
tapers. k: optional, integer, default =kin- The number of tapers to be utilized in performing the multitaper spectral analysis.
taper_wt: optional, float, dimension (kin), default = -1- The weights used in calculating the multitaper spectral estimates and standard error. Optimal values of the weights (for a known global power spectrum) can be obtained from the routine
SHMTVarOpt. The default value specifies not to usetaper_wt. norm: optional, intger, default = 1- 1 (default) = 4-pi (geodesy) normalized harmonics; 2 = Schmidt semi-normalized harmonics; 3 = unnormalized harmonics; 4 = orthonormal harmonics.
csphase: optional, integer, default = 1- 1 (default) = do not apply the Condon-Shortley phase factor to the associated Legendre functions; -1 = append the Condon-Shortley phase factor of (-1)^m to the associated Legendre functions.
Description
SHMultiTaperMaskCSE will perform a localized multitaper cross-spectral analysis of two input functions expressed in spherical harmonics, SH1 and SH2, using an arbitrary set of windows derived from a mask. The maximum degree of the localized multitaper power spectrum estimate is lmax-lmaxt, where lmax is the smaller of lmax1 and lmax2. The matrix tapers contains the spherical harmonic coefficients of the windows and can be obtained by a call to SHReturnTapersMap. The coefficients of each window are stored in a single column, ordered according to the conventions used in SHCilmToVector.
If the optional array taper_wt is specified, then these weights will be used in calculating a weighted average of the individual k tapered estimates (mtse) and the corresponding standard error of the estimates (sd). If not present, the weights will all be assumed to be equal. When taper_wt is not specified, the mutltitaper spectral estimate for a given degree is calculated as the average obtained from the k individual tapered estimates. The standard error of the multitaper estimate at degree l is simply the population standard deviation, S = sqrt(sum (Si - mtse)^2 / (k-1)), divided by sqrt(k). See Wieczorek and Simons (2007) for the relevant expressions when weighted estimates are used.
The employed spherical harmonic normalization and Condon-Shortley phase convention can be set by the optional arguments norm and csphase; if not set, the default is to use geodesy 4-pi normalized harmonics that exclude the Condon-Shortley phase of (-1)^m.
References
Wieczorek, M. A. and F. J. Simons, Minimum-variance multitaper spectral estimation on the sphere, J. Fourier Anal. Appl., 13, doi:10.1007/s00041-006-6904-1, 665-692, 2007.
See also
shmultitapermaskse, shreturntapersmap, shcilmtovector
> Home > Documentation > Python > Global and Localized Spectral Analysis
| Laboratoire Lagrange | Observatoire de la Côte d'Azur | © 2017 SHTOOLS |