xbooter: Additional Bootstrap Functions for Univariate EVA

Description

Additonal bootstrap capabilities for extreme-value analysis for fevd objects.

Usage

xbooter(x, B, rsize, block.length = 1,
	return.period = c(10, 20, 50, 100, 200, 500),
	qcov = NULL, qcov.base = NULL, shuffle = NULL,
	replace = TRUE, verbose = FALSE, ...)

Value

Object of class “booted” is returned. See the help file for booter for more information.

Arguments

x: list object of class “fevd”
B, rsize, block.length, shuffle, replace: See the help file for booter from the distillery package.
return.period: numeric value for the desired return period for which CIs are desired.
qcov: numeric matrix with rows the same length as q and columns equal to the number of parameters (+ 1 for the threshold, if a POT model). This gives any covariate values for a nonstationary model. If NULL, and model is non-stationary, only the intercept terms for modeled parameters are used, and if a non-constant threshold, only the first threshold value is used. Not used if model is stationary.
qcov.base: numeric matrix analogous to qcov. When provided, the function returns the difference in return levels between the level for the covariates in qcov and the level for covariates in qcov.base.
verbose: logical if TRUE progress information is printed to the screen.
...: Additonal optional arguments to the booter function.

Author

Eric Gilleland

Details

The ci method function will perform parametric bootstrapping for “fevd” objects, but this function is a wrapper to booter, which allows for greater flexibility with “fevd” objects. Gives CIs for the EVD parameters and return levels.

References

Gilleland, E. (2020) Bootstrap methods for statistical inference. Part I: Comparative forecast verification for continuous variables. Journal of Atmospheric and Oceanic Technology, 37 (11), 2117 - 2134, doi: 10.1175/JTECH-D-20-0069.1.

Gilleland, E. (2020) Bootstrap methods for statistical inference. Part II: Extreme-value analysis. Journal of Atmospheric and Oceanic Technology, 37 (11), 2135 - 2144, doi: 10.1175/JTECH-D-20-0070.1.

Examples

Run this code

set.seed( 409 )
z <- apply( matrix( rnorm( 100 * 1000 ), 1000, 100 ), 2, max )
fit <- fevd( z )

# In order to keep the code fast for CRAN compiling, 
# a low value for B is used here, but should use a larger
# value in general.
bfit <- xbooter( fit, B = 50, verbose = TRUE )
ci( bfit, type = "perc" )

Run the code above in your browser using DataLab

Last chance! 50% off unlimited learning