function sl = cohere_signif_level( df, p ) % COHERE_SIGNIF_LEVEL - estimate significance level coherence using % theoretical formula. % % Usage: sig_lev = cohere_bootstrap_signif_level( df, p ) % % where p is the significance level and df is the number of degrees of % freedom as described in Shumway and Stoffer (see below). p defaults to 0.95. % % sig_lev is the significance level. % % This function uses the finv function which is in the Matlab Statistical % Toolbox. % % This function uses an equation taken from Shumway and Stoffer % (2000), "Time series analysis and its applications", pg. 250, % equ. 3.82. % % The degrees of freedom is given by: % % df = 2Ln/n' when a series of length n has been zero padded to length % n'. L is the number of frequencies in the averaging used to compute % the spectra. % %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% % % $Id: cohere_signif_level.m,v 1.2 2005-01-08 02:49:29 dmk Exp $ % % Copyright (C) 2004 David M. Kaplan % Licence: GPL (Gnu Public License) % %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% if nargin < 2 p = 0.95; end sl = finv( p, 2, df - 2 ); sl = sl / ( df/2 - 1 + sl );