Mercurial > hg > ltpda
comparison m-toolbox/classes/+utils/@math/overlapCorr.m @ 0:f0afece42f48
Import.
| author | Daniele Nicolodi <nicolodi@science.unitn.it> |
|---|---|
| date | Wed, 23 Nov 2011 19:22:13 +0100 |
| parents | |
| children |
comparison
equal
deleted
inserted
replaced
| -1:000000000000 | 0:f0afece42f48 |
|---|---|
| 1 % OVERLAPCORR Compute correlation introduced by segment overlapping | |
| 2 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% | |
| 3 % OVERLAPCORR Compute correlation introduced by segment overlapping in WOSA | |
| 4 % spectral estimations. It estimates the term contributing to the variance | |
| 5 % because of segment overlapping. | |
| 6 % | |
| 7 % CALL | |
| 8 % | |
| 9 % R = utils.math.overlapCorr(w,N,T) | |
| 10 % [R,n] = utils.math.overlapCorr(w,N,T) | |
| 11 % | |
| 12 % INPUT | |
| 13 % | |
| 14 % - w, window samples, Nsx1 double, Ns must be the effective length of the | |
| 15 % segments used for spectral estimation. E.g. For a periodogram Ns is equal | |
| 16 % to the length of the data series. For a WOSA estimation Ns is the length | |
| 17 % of each averaging segment. | |
| 18 % - N total length of data series, 1x1 double | |
| 19 % - navs, number of averages, 1x1 double | |
| 20 % | |
| 21 % REFERENCES | |
| 22 % | |
| 23 % D. B. Percival and A. T. Walden, Spectral Analysis for Physical | |
| 24 % Applications (Cambridge University Press, Cambridge, 1993) p 292. | |
| 25 % | |
| 26 % L Ferraioli 09-03-2011 | |
| 27 % | |
| 28 % $Id: overlapCorr.m,v 1.1 2011/03/28 16:37:23 luigi Exp $ | |
| 29 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% | |
| 30 function varargout = overlapCorr(w,N,navs) | |
| 31 | |
| 32 Ns = numel(w); | |
| 33 % willing to work with columns | |
| 34 [nn,mm] = size(w); | |
| 35 if nn<mm | |
| 36 w = w.'; | |
| 37 end | |
| 38 | |
| 39 % make suaqre integrable | |
| 40 a = sqrt(sum(w.^2)); | |
| 41 w = w./a; | |
| 42 | |
| 43 n = floor((N-Ns)/(navs-1)); | |
| 44 win = [w; zeros((navs-1)*n,1)]; | |
| 45 | |
| 46 R = 0; | |
| 47 for kk=1:navs-1 | |
| 48 ew = zeros(size(win)); | |
| 49 ew(kk*n+1:kk*n+Ns) = win(1:Ns); | |
| 50 | |
| 51 R = R + (navs-kk)*abs(win.'*ew)^2; | |
| 52 end | |
| 53 R = R./navs^2; | |
| 54 | |
| 55 if nargout == 1 | |
| 56 varargout{1} = R; | |
| 57 else | |
| 58 varargout{1} = R; | |
| 59 varargout{2} = n; | |
| 60 end | |
| 61 | |
| 62 | |
| 63 end |