Convergence of Numerical Box-Counting and Correlation Integral Multifractal Analysis Techniques.
Abstract
A systematic study of the rate of convergence for a numerical box-counting and a numerical correlation integral algorithm applied to Euclidean point sets, Koch constructions, and a symmetric chaotic mapping is described. The number of points N(5) required for 5 percent convergence of the box-counting (for 0 < or = q < or = 25) and correlation integral (for q between -25 and 25) algorithms for the fractal sets studied is determined by the generalized dimension D(q) and is given by log10(N5) approx. equals to 2.54 D(q)-O.11. Approximately 25 times as many points are required for 1 percent convergence. The box-based correlation integral(BBCI) algorithm employed in the present studies, which is well suited to the analysis of large data sets, is also described.
Document Details
- Document Type
- Technical Report
- Publication Date
- Apr 01, 1995
- Accession Number
- ADA300068
Entities
People
- Lawrence V. Meisel
- Marc A. Johnson
Organizations
- United States Army Armament Research, Development and Engineering Center