mfdfa: Multifractal Detrended Fluctuation Analysis

In wol-fi/multifractal: Multifractal and Surrogate Analysis

Multifractal Detrended Fluctuation Analysis

Description

Performs the multifractal detrended fluctuation analysis upon a time-series x. Allows the improvement of overlapping windows.

Usage

mfdfa(x, scale=NA, q=-5:5, m=1, overlap=FALSE)

Arguments

x	a numeric vector containing the values of the time-series.
scale	an integer vector containing the time scales. It is recommended to use a log-equally spaced sequence for scale, for example via the function logseq(). When set to NA then it is calculated as scale = round(logseq(from=32, to=length(x)/10, n=20)).
q	a numeric vector containing values of moment orders.
m	an integer representing the polynomial order of detrending, m=1 defines linear detrending.
overlap	a boolean indicating if using overlapping windows or not. If set to TRUE then the estimation is more robust and reliable, if set to FALSE then the computational time is faster.

Details

The algorithm allows a basic but also improved (overlapping windows) estimation of a time-series' multi-scaling characteristica. It comes with the methods plot() and print() to easily visualize/access the multifractal spectra and statistics. The function significance() allows to easily test the significance of multifractality.

Value

Returns an object of class multifractal containg multifractal statistics and more.

Author(s)

Wolfgang Schadner

References

Kantelhardt, J. W., Zschiegner, S. A., Koscielny-Bunde, E., Havlin, S., Bunde, A., & Stanley, H. E. (2002). Multifractal detrended fluctuation analysis of nonstationary time series. Physica A: Statistical Mechanics and its Applications, 316(1-4), 87-114.

See Also

significance

Examples

  n <- 1024

  # random t-distributed variable:
  x <- rt(n, 3)

  mdl <- mfdfa(x, overlap=T)
  plot(mdl)
  significance(mdl)

wol-fi/multifractal documentation built on May 31, 2022, 1:18 a.m.