xtibber: Test-Inversion Bootstrap for Extreme-Value Analysis

In extRemes: Extreme Value Analysis

Test-Inversion Bootstrap for Extreme-Value Analysis

Description

Test-inversion bootstrap (TIB) for fevd class objects.

Usage

xtibber(x, type = c("return.level", "parameter"), which.one,
	tib.method = c("interp", "rm"), nuisance = "shape", B,
	test.pars, rsize, block.length = 1, shuffle = NULL,
	replace = TRUE, alpha = 0.05, qcov = NULL,
	qcov.base = NULL, stud = FALSE, step.size, tol = 1e-04,
	max.iter = 1000, keep.iters = TRUE, verbose = FALSE, ...)

Arguments

x	List object of class “fevd”.
type	character string stating whether to calculate TIB intervals for a return level or a parameter as this funciton will only calculate an interval for a single parameter/return level at a time.
which.one	number or character stating which return level or which parameter to find CIs for.
tib.method	character stating whether to estimate the TIB interval by interpolating from a series of pre-determined values of the nuisance parameter or to use the Robbins-Monroe (RM) method. See the help file for tibber from the distillery package for more information.
nuisance	character naming the nuisance parameter.
B, rsize, block.length, shuffle, replace	See the help file for booter from the distillery package for more information on these arguments.
test.pars	numeric vector giving the sequence of nuisance parameter values for the interpolation method, or a numeric vector of length two giving the starting values for the RM method.
alpha	numeric between zero and one giving the desired confidence level.
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.
stud	logical if TRUE will calculate Studentized intervals (generally not profitable with the TIB method).
step.size	Used with the RM method only. Numeric giving the size of increments to use in the root-finding algorithm.
tol	Used with the RM method only. Numeric stating how close to the desired level of confidence is satisfactory.
max.iter	numeric giving the maximum number of iterations for the root-finding algorithm before giving up.
keep.iters	logical, should all of the values in the root-finding search be kept? Needed if a plot will be made.
verbose	logical, if TRUE will print progress information to the screen.
...	optional arguments to nlminb.

Details

This function provides a wrapper to the tibber function from distillery for “fevd” objects.

Value

See the help file for tibber for more information on the value

Author(s)

Eric Gilleland

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.

See Also

fevd, tibber, booter

Examples

## Not run: 
data("ftcanmax")
fit <- fevd( Prec, data = ftcanmax )

tbfit <- xtibber( fit, which.one = 100, B = 500,
		 test.pars = seq(-0.01,0.2,,100), verbose = TRUE )

tbfit

plot( tbfit )


## End(Not run)

extRemes documentation built on Nov. 19, 2022, 1:07 a.m.