Create a synthetic timeseries that emulates the characteristics of a variable

create synthetic timeseries based on a timeseries. Useful for null hypothesis testing

ar1Surrogates(time, vals, detrend = TRUE, method = "redfit", n.ens = 1)

Arguments

time

LiPD "variable list" or vector of year/age values

vals

LiPD "variable list" or vector of values

detrend

boolean variable, indicating whether one should remove a linear trend (default) or not. Note: does not affect the AR(1) fit, but does affect the generated timeseries

method

Method used for estimation. Possible values are "even" or "redfit"

n.ens

Number of ensemble members to simulate

Value

a vector or matrix of AR(1) surrogates for series X

See also

Other spectra: computeSpectraEns(), createSyntheticTimeseries(), periodAnnotate(), plotSpectrum(), reverselog10_trans()

Other pca: createSyntheticTimeseries(), pcaEns(), plotPcaEns(), plotScreeEns()

Other correlation: ar1(), corMatrix, effectiveN(), plotCorEns(), pvalMonteCarlo(), pvalPearsonSerialCorrected()