How to calculate Yaghoobi Fractal dimension

7 views (last 30 days)
How to calculate the Yaghoobi fractal dimension for a high dimension trajectory in the high dimensional phase space.
Article title: A new approach to measure the fractal dimension of a trajectory in the high-dimensional phase space
  1 Comment
reza yaghoobi
reza yaghoobi on 2 Jul 2021
function main()
clc;clear all;close all
Kmax = 8;
Y = randn(55,4);
F = YaghoobiFractalDimension(Y,Kmax);
end
function YFD = YaghoobiFractalDimension(Y,Kmax)
%% Estimate Yaghoobi fractal dimension
% Y: a high dimension trajectory
% Kmax: The length of actions
% YFD: Fractal Dimension of the input signal
dim = size(Y,2);
N = size(Y,1);
for k = 1:Kmax
clear xx L
for i = 1:k
a = Y(i:k:(i+fix((N-i)/k)*k),:);
d = diff(a);
r = sqrt(sum(d.^2,2));
alpha = [N-1]/(k*floor([N-i]/k));
L(i) = alpha*sum(r)/(k^(dim-1));
end
LM(k) = mean(L);
end
LM(LM==0)=1e-200;
lLM = log10(LM);
lk = log10(1:Kmax);
lk(LM==0) = [];
lLM(LM==0) =[];
%% LSE
[a,b] = LSE(lk,lLM);
%% Plot
plot(lk,a*lk+b,':k');hold on
plot(lk,lLM,'.','MarkerSize',15);
xlabel('Log_1_0(k)')
ylabel('Log_1_0(L_k)')
xlim([lk(1) lk(end)])
YFD = abs(a);
title(['FD = ',num2str(YFD)])
end
%% fitting
function [a,b] = LSE(x,y)
N = size(x,2);
a = [N*x*y'-sum(x)*sum(y)]/[N*sum(x.^2)-sum(x)^2];
b = [sum(x.^2)*sum(y)-x*y'*sum(x)]/[N*sum(x.^2)-sum(x)^2];
end

Sign in to comment.

Answers (0)

Categories

Find more on Fractals in Help Center and File Exchange

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!