conf.int: Confidence intervals for GEV and MEV distributions

In SpatialModelsZAMG: Spatial Modeling of Extreme Events

Description

Computes confidence intervals for GEV and MEV distributions using a parametric bootstrap method.

Usage

conf.int(x, alpha = 0.05, return.periods = c(2,10,30,50,100),
         dist = c("GEV","MEV"), method = "boot", R = 502, verbose = FALSE)

Arguments

x	data.frame with either columns "W", "C", "n" for the MEV distribution, or "loc","scale","shape" for the GEV distribution (see details).
alpha	Number between zero and one giving the 1 - alpha confidence level.
return.periods	Numeric vector giving the desired return periods.
dist	Character string naming which EVD to calculate confidence intervals from.
method	character naming which method for obtaining confidence intervals should be used. Option "boot" employs a parametric bootstrap that simulates data from the fitted model, and then fits the EVD to each simulated data set to obtain a sample of parameters or return levels. Currently only "boot" is available.
R	the number of bootstrap iterations.
verbose	Should additional information be printed during rundtime? Can be TRUE or FALSE, which is the default.

Details

x must be a data.frame, whose content depends on the distribution dist:

MEV: If dist='MEV', then the data.frame x must contain columns "W", "C", "n". They represent shape and scale parameter of the MEV, and mean number of wet days.

GEV: If dist='GEV', columns "loc","scale","shape" must be location, scale and shape parameter of the GEV .

Value

Either a numeric vector of length 3 (if only one return level is used) or a matrix.

Author(s)

Harald Schellander

References

Schellander, H., Lieb, A. and Hell, T. (2019) 'Error Structure of Metastatistical and Generalized Extreme Value Distributions for Modeling Extreme Rainfall in Austria', Earth and Space Science, 6, pp. 1616-1632. doi: 10.1029/2019ea000557.

SpatialModelsZAMG documentation built on Nov. 11, 2019, 3 p.m.