nonlinearTseries (version 0.3.1)

FFTsurrogate: Generate surrogate data using the Fourier transform

Description

Generates surrogate samples from the original time series.

Usage

FFTsurrogate(time.series, n.samples = 1)

Value

A matrix containing the generated surrogate data (one time series per row).

Arguments

time.series

The original time.series from which the surrogate data is generated.

n.samples

The number of surrogate data sets to generate,

Author

Constantino A. Garcia

Details

This function uses the phase randomization procedure for generating the surrogated data. This algorithm generates surrogate data with the same mean and autocorrelation function (and thus, the same power spectrum because of the Wiener-Khinchin theorem) as the original time series.

The phase randomization algorithm is often used when the null hypothesis being tested consist on the assumption that the time.series data comes from a stationary linear stochastic process with Gaussian inputs. The phase randomization preserves the Gaussian distribution.

References

H. Kantz and T. Schreiber: Nonlinear Time series Analysis (Cambridge university press)

Examples

if (FALSE) {
# generate 20 surrogate sets using as original time series
# an arma(1,1) simulation
time.series = arima.sim(list(order = c(1,0,1), ar = 0.6, ma = 0.5), n = 200)
surrogate = FFTsurrogate(time.series, 20)
}