ARsieve_bootstrap: AR-sieve bootstrap

In rwigren/TSAutils: A set of utilities for time series analysis

Description

Bootstrapping method for linear stationary univariate time series. Returns a bootstrap distribution of the chosen statistic.

Usage

ARsieve_bootstrap(series, statistic, B, sim.length = NROW(series),
  max.p = 10 * log10(NROW(series)), ar.method = "yw", ...)

Arguments

series	A univariate linear stationary time series from which the AR-sieve sampling will be done.
statistic	Function: the statistic for which a bootstrap-distribution will be created.
B	Amount of bootstrap samples.
sim.length	Length of the sampled series. Will usually be the same length as the original series, but can in some cases (such as residual bootstrapping) be useful to be set to another value.
max.p	Maximum order the fitted AR model can take.
ar.method	Method used to fit the auto-regressive process. ()
...	Extra values passed on to the statistic function.

Details

The implementation follows the AR-sieve bootstrap method as presented by Bühlmann (1997). Fits an AR process based on AIC minimization and creates samples based on fitted coefficients and empirical residuals. The statistic is then evaluated on these new samples. Can for example be used to estimate the variance or confidence intervals of a statistic (e.g. auto-regressive coefficients of the series).

References

Bühlmann, Peter (1997) "Sieve bootstrap for time series". Bernoulli, 3(2), 123–148.

See Also

blockwise_bootstrap, stats::ar

Examples

series <- arima.sim(model=list(ar=c(0.6, -0.4)), n=100)
ars.res <- ARsieve_bootstrap(series, mean, B=500)
boxplot(ars.res)

rwigren/TSAutils documentation built on Nov. 5, 2019, 5:10 a.m.