EDDYLEVELS Eddy ridge significance levels using the survival function. RHO=EDDYLEVELS(DX,XMID,YBIN,X,Y,XNOISE,YNOISE,N) returns the eddy ridge significance levels RHO for a Lagrangian dataset, based on comparision of the survival function of the data with that of an ensemble of noise. For details and theory on the use of EDDYLEVELS, see Lilly, J. M. and P. Perez-Brunius (2021). Extracting statistically significant eddy signals from large Lagrangian datasets using wavelet ridge analysis, with application to the Gulf of Mexico. Submitted to Nonlinear Processes in Geohysics. This paper will be referred to as LPB21 afterwards. The eddy ridges for the data and the noise are assumed to have been calculated using EDDYRIDGES, while the noise itself is generated using NOISEDRIFTERS. See MAKE_GOMED for an example. The significance levels are computed in terms of X and Y, two ridge- averaged quantities. XNOISE and YNOISE are the comparable quantities in the noise ensemble, which is N times the size of the data. In LPB21, the ridge-averaged nondimensional frequency OMEGA_AST_BAR is used for X, and the ridge length L times the square of the circularity XI is used for Y. See therein for a discussion and other possibilties. A survival function is created for Y as a function of X, with the x-bins centered on midpoints XMID and spanning XMID-DX/2 to XMID+DX/2, and with YBIN giving the bin edges for the Y variable. RHO is then the significance level of the ridges based on interpolating within the ratio of the two survival functions. Smaller values of RHO correspond to higher significance levels. RHO has the size of X and Y. [RHO,YMID,S,SNOISE,SRAT]=EDDYLEVELS(...) also returns the y-bin midpoints YMID, the data and noise survival functions S and SNOISE, and their ratio SRAT=SNOISE./S. The survival functions are all matrices with LENGTH(YMID) rows and LENGTH(XMID) columns. EDDYLEVELS(...,'symmetric') optionally forces the noise survival function to be symmetric about its x-axis, as expected if the x-axis spans a symmetric band of frequencies centered about zero as in LPB21. EDDYLEVELS(...,'sort') optionally sorts the survival function ratio SRAT along its y-axis in descending order, in order to remove small-scale jaggedness that can arise in sparsely sampled bins. This is the recommended setting, for reasons discussed in LPB21. Usage: rho=eddylevels(dx,xmid,ybin,x,y,xnoise,ynoise,N); [rho,ymid,S,Snoise,Srat] =eddylevels(dx,xmid,ybin,x,y,xnoise,ynoise,N); __________________________________________________________________ This is part of JLAB --- type 'help jlab' for more information (C) 2021 J.M. Lilly --- type 'help jlab_license' for details