RoverS: Estimate the Hurst coefficient by rescaled range (R/S) method

In fractal: A Fractal Time Series Modeling and Analysis Package

Description

The series is partitioned into m groups. The R/S statistic is computed as described in the references, the number of groups is increased, and the calculation is repeated. A log-log plot of R/S versus number of groups is, ideally, linear, with a slope related to H, so H can be determined by linear regression.

Usage

RoverS(x, n.block.min=2, scale.ratio=2, scale.min=8)

Arguments

x	a vector containing a uniformly-sampled real-valued time series.
n.block.min	minimum number of blocks in partitioning the data. Must be at least 2. Default: 2.
scale.min	minimum number of data values allowed in a block This may be restricted so the statistic evaluated within each group is from a reasonable sample. Default: 8.
scale.ratio	ratio of successive scales to use in partitioning the data. For example, if scale.min=8 and scale.ratio=2, the first scale will be 8, the second scale 16, the third scale 32, and so on. Default: 2.

Value

estimated Hurst parameter H of the time series.

References

B.B. Mandelbrot and J.R. Wallis (1969), Water Resources Research, 5, 228–267.

See summary in M.S. Taqqu and V. Teverovsky (1998), On Estimating the Intensity of Long-Range Dependence in Finite and Infinite Variance Time Series, in A practical Guide to Heavy Tails: Statistical Techniques and Applications, 177–217, Birkhauser, Boston.

See Also

hurstBlock, hurstACVF, dispersion.

Examples

RoverS(wmtsa::ocean)

Example output

Loading required package: splus2R
Loading required package: ifultools
        X 
0.9255632 

fractal documentation built on May 1, 2019, 8:04 p.m.