hadron (version 3.1.2)

bootstrap.cf: bootstrap a set of correlation functions

Description

bootstrap a set of correlation functions

Usage

bootstrap.cf(cf, boot.R = 400, boot.l = 2, seed = 1234, sim = "geom",
  endcorr = TRUE)

Arguments

cf

correlation matrix of class cf e.g. obtained with a call to extrac.obs.

boot.R

number of bootstrap samples.

boot.l

block size for autocorrelation analysis

seed

seed for the random number generation used for boostrapping.

sim

The type of simulation required to generate the replicate time series. The possible input values are <U+2018>"fixed"<U+2019> (block resampling with fixed block lengths of <U+2018>boot.l<U+2019>) and <U+2018>"geom"<U+2019> (block resampling with block lengths having a geometric distribution with mean <U+2018>boot.l<U+2019>). Default is <U+2018>"geom"<U+2019>. See tsboot for details.

endcorr

A logical variable indicating whether end corrections are to be applied when <U+2018>sim<U+2019> is <U+2018>"fixed"<U+2019>. When <U+2018>sim<U+2019> is <U+2018>"geom"<U+2019>, <U+2018>endcorr<U+2019> is automatically set to <U+2018>TRUE<U+2019>; <U+2018>endcorr<U+2019> is not used when <U+2018>sim<U+2019> is <U+2018>"model"<U+2019> or <U+2018>"scramble"<U+2019>. See tsboot for details.

Value

returns an object of class cf with bootstrap samples added for th correlation function called cf.tsboot. Moreover, the original average of cf is returned as cf0 and the bootstrap errors as tsboot.se. We also copy the input parameters over and set bootstrap.samples to TRUE.

See Also

tsboot, jackknife.cf

Examples

# NOT RUN {
data(samplecf)
samplecf <- bootstrap.cf(cf=samplecf, boot.R=99, boot.l=2, seed=1442556)
plot(samplecf, log=c("y"))

# }