make_surrogate_data: Generate surrogate data for permutation/randomization tests

In ha0ye/rEDM: Applications of Empirical Dynamic Modeling from Time Series

Description

This is a wrapper function for generating surrogate time series using several different null models.

See test_nonlinearity for an example context for usage of this function.

make_surrogate_shuffle() creates surrogates by randomly permuting the values of the original time series.

make_surrogate_ebisuzaki() creates surrogates by randomizing the phases of a Fourier transform, preserving the power spectra of the null surrogates

make_surrogate_seasonal() creates surrogates by computing a mean seasonal trend of the specified period and shuffling the residuals.

make_surrogate_twin() creates surrogates using the twin- surrogate method, with the option to preserve the phase for seasonal/ periodic data

Usage

make_surrogate_data(ts, method = c("random_shuffle", "ebisuzaki",
  "seasonal", "twin"), num_surr = 100, ...)

make_surrogate_shuffle(ts, num_surr = 100)

make_surrogate_ebisuzaki(ts, num_surr = 100)

make_surrogate_seasonal(ts, num_surr = 100, T_period = 12)

make_surrogate_twin(ts, num_surr = 1, dim = 1, tau = 1,
  phase_lock = TRUE, T_period = 24, initial_point = "same_season",
  ...)

Arguments

ts	the original time series
method	which algorithm to use to generate surrogate data
num_surr	the number of null surrogates to generate
...	remaining arguments are passed on to the specific function to make surrogate time series of that type
T_period	the period of seasonality for seasonal surrogates (ignored for other methods)
dim	the embedding dimension for the state-space reconstruction, in which twins are identified
tau	the lag for the state-space reconstruction
phase_lock	whether twins have to occur at the same phase
initial_point	how to sample the initial point. If '"same_season"', then the initial point is chosen from the same phase in a different cycle, and the surrogate is not allowed to line up in both phase and cycle with the original time series.

Value

A matrix where each column is a separate surrogate with the same length as 'ts'.

Examples

data("two_species_model")
ts <- two_species_model$x[1:200]
make_surrogate_data(ts, method = "ebisuzaki")

make_surrogate_shuffle(rnorm(100), 10)

make_surrogate_ebisuzaki(rnorm(100), 10)

make_surrogate_seasonal(rnorm(100) + sin(1:100 * pi / 6), 10)

make_surrogate_twin(rnorm(100, sd = 0.1) + sin(1:100 * pi / 6), 10)

ha0ye/rEDM documentation built on March 30, 2021, 11:21 p.m.