multifractal-spectrum

所属分类:matlab编程
开发工具:matlab
文件大小:79KB
下载次数:66
上传日期:2012-04-03 15:11:03
上 传 者zy-nwu
说明:  国外用MATLAB编写的多维分形分析断裂结构的程序。在地质建模中有潜在用途。
(A MATLAB® program based on the Hou algorithm for estimation of fractal dimension and multifractal spectrum of fractures is presented. The program performance was tested with many synthetical fractals and field data. )

文件列表:
2s_cantor.dat (278528, 2011-06-22)
box_count.m (2566, 2011-06-22)
falpha.dat (693, 2011-06-22)
fit_frac.m (1721, 2011-06-22)
leg_transf.m (721, 2011-06-22)
n_box.dat (1056, 2011-06-22)

A) INTRODUCTION Thank you for your interest in our program. This file contain instructions for using box-count program and auxiliary programs leg_transf and fit_frac. Box-count program estimates the fractal dimension and the multifractal spectrum of multidimensional ASCII data. Auxiliary programs leg_transf and fit_frac are used to obtain the fractal dimension and the multifractal spectrum, respectively. B) INSTRUCTION 0) Unzip the files. 1) Copy the files (scripts, test files and output files) to your directory. For example, ``\home\"your_account"\programs'' (under linux) or ``C:/PROGRAMS'' (under windows). 2) Open the matlab (or Octave) prompt as usual. 3) Change to the directory in which the scripts and files are located, by typing cd [path]. 4) In MATLAB/Octave prompt type [np,s,q]=box_count; 5) Provide the filename ([path/filename]) of the set to be processed; 6) Enter the precision from 4 to ***; 7) Enter the kind of analysis: 1 for monofractal or 2 or multifractal; 8-a) If it was chosen monofractal analysis type in the MATLAB/Octave prompt [x np]=fit_frac(np,s); 8-b) If it was chosen multifractal fractal analysis type the MATLAB/Octave [a f_a]=leg_transf(np,s,q); 9) Enter the range (m_low and m_up) to fit the curves, if the fitting is not good repeat procedures (8-a) or (8-b). 10) Use the fractal dimension item 8-a to control the support dimension of the spectrum f(alpha) 8-b. 11) You can change the q values (moments) in box_count code by choosing other values for min_q, step and max_q, in the vector qv in line 15: qv=single([min_q]:[step]:[max_q]); The plots are presented in the relative scale log(e)=m (m=1,2,3,...,k) and the real scale, obtained by s*(2^m/2^k), where $s$ is the maximal linear length of the set. C) FORMAT OF INPUT DATA. ASCII NxD or DxN arrays, where N is the number of points and D the embedding dimension of the set (D may be lesser than N). D) TEST DATA. The test data (2s_cantor.dat) is a Two-scale Cantor Set (Halsey et al., 1***6) with 2^14 points, generate by IFS. The output files are ``n_box.dat'' (generated using box_count program with fractal option and fit_frac program) and ``falpha.dat'' (generated box_count program with multifractal option and leg_transf program). The output was generated using k=32 and (m_low,m_up)=(18,30). These matlab scripts are distributed as part of the article: Jeferson de Souza and Sidnei Pires Rostirolla. A fast MATLAB program to estimate the multifractal spectrum of multidimensional data: application to fractures, Submitted to Computers and Geosciences, 2007 Contact: jdesouza@ufpr.br

近期下载者

相关文件


收藏者