TSTOOL home page | TSTOOL documentation page | TSTOOL link page

next up previous contents
Next: 6.11 henon - Generate Up: 6. Mex-Function Reference Previous: 6.9 fnearneigh - Fast

6.10 gendimest - Estimate generalized dimension spectrum

The Renyi dimension spectrum of a points set can be estimated using information about the distribution of the interpoint distances. Since we are interested in the scaling behaviour of the Renyi information for small distances, we don't need to compute all interpoint distances, the distances to k nearest neighbors for each reference point are sufficient [150].

Robust estimation is used instead of mean square error fitting.

Syntax:

Input arguments:

Output arguments:

Example:

 
x = chaosys(25000, 0.025, [0.1 -0.1 0.02], 0);  % generate data from Lorenz system
x = x(5001:end,:);      % discard first 5000 samples due to transient 
[nn, dist] = fnearneigh(x, randref(1, 20000, 1000), 128, 0);
gammas = -5:0.5:5;
gedims = gendimest(dist, gammas, 8, 8, 128);
plot(1-gammas./gedims', gedims)
xlabel('q');ylabel('D_q');title('Renyi dimension')

next up previous contents
Next: 6.11 henon - Generate Up: 6. Mex-Function Reference Previous: 6.9 fnearneigh - Fast   Contents
TSTOOL

Copyright © 1997-2009 DPI Göttingen