1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
|
contwtspec(2) Scilab Function contwtspec(2)
NAME
contwtspec - Continuous L2 wavelet based Legendre spectrum
Author: Paulo Goncalves
Estimates the multifractal Legendre spectrum of a 1-D signal from the
wavelet coefficients of a L2 continuous decomposition
Usage
[alpha,f_alpha,logpart,tau] = contwtspec(wt,scale,Q[,FindMax,ChooseReg])
Input parameters
o wt : Real or complex matrix [N_scale,N] Wavelet coefficients of a
continuous wavelet transform (output of contwt or contwtmir))
o scale : real vector [1,N_scale] Analyzed scale vector
o Q : real vector [1,N_Q] Exponents of the partition function
o FindMax : 0/1 flag. FindMax = 0 : estimates the Legendre spectrum
from all coefficients FindMax = 1 : estimates the Legendre spectrum
from the local Maxima coefficients of the wavelet transform Default
value is FindMax = 1
o ChooseReg : 0/1 flag or integer vector [1,N_reg], (N_reg <= N_scale)
ChooseReg = 0 : full scale range regression ChooseReg = 1 : asks
online the scale indices setting the range for the linear regression
of the partition function. ChooseReg = [n1 ... nN_reg] : scale
indices for the linear regression of the partition function.
Output parameters
o alpha : Real vector [1,N_alpha], N_alpha <= N_Q Singularity support
of the multifractal Legendre spectrum
o f_alpha : real vector [1,N_alpha] Multifractal Legendre spectrum
o logpart : real matrix [N_scale,N_Q] Log-partition function
o tau : real vector [1,N_Q] Regression function
See also:
contwt, cwtspec, cwt, dwtspec, FWT
Example:
N = 2048 ; H = 0.7 ; Q = linspace(-4,4,11) ;
[x] = fbmlevinson(N,H) ;
[wt,scale] = contwtmir(x,2^(-8),2^(-1),16,8) ;
[alpha,f_alpha,logpart,tau] = contwtspec(wt,scale,Q,1,1) ;
plot(alpha,f_alpha),
|
