localH: Local Hurst Exponent

In wol-fi/multifractal: Multifractal and Surrogate Analysis

Local Hurst Exponent

Description

Estimates the local Hurst exponent based on Ihlen and Vereijken (2014). This measures the current, temporal degree of persistence. A value larger 0.5 means positive auto-correlation and below 0.5 describes negative auto-correlation.

Usage

localH(x, mdl, scale=5:21, m=1, align="center")

Arguments

x	a numeric vector containing the values of the time-series.
mdl	an object of class multifractal estimated from the function mfdfa.
scale	an integer vector containing the time scales. These should be small to capture local correlation behavior, i.e. the current degree of persistence.
m	an integer representing the polynomial order of detrending, m=1 defines linear detrending.
align	specifies the alignment within the estimation windows. Possible values are 'center', 'left' or 'right'.

Details

The intertemporal persistence is estimated from the interpolating routine of Ihlen and Vereijken (2014). The multifractal model should be estimated in a first step using mfdfa.

Value

Returns a two-column matrix of local Hurst exponent and related interquartile range.

Author(s)

Wolfgang Schadner

References

Ihlen E. A. F., & Vereijken B. (2014). Detection of co-regulation of local structure and magnitude of stride time variability using a new local detrended fluctuation analysis. Gait Posture, 39(1), p.466-471.

See Also

mfdfa

Examples

  n <- 1024

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

  mdl <- mfdfa(x, overlap=T)

  Ht <- localH(mdl)
  
  plot(Ht)

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