mev (version 2.1)

thselect.cbm: Threshold selection by shape mean square error minimization

Description

Use a semiparametric bootstrap to calculate the mean squared error of the shape parameter using maximum likelihood for different thresholds, and return the one that minimize the mean squared error.

Usage

thselect.cbm(xdat, thresh, B = 100)

Value

an object of class mev_thselect_cbm containing

• thresh: ordered vector of candidate thresholds

• thresh0: selected threshold

• shape: shape parameter coefficient estimates at each threshold

• nexc: number of exceedances at each threshold

• bias: vector of bootstrap bias estimates

• var: vector of bootstrap variance estimates

• mse: vector of mean squared error bootstrap estimates

Arguments

xdat

vector of observations

thresh

vector of thresholds

B

number of bootstrap replications

References

Caers, J., Beirlant, J. and Maes, M.A. (1999). Statistics for Modeling Heavy Tailed Distributions in Geology: Part I. Methodology. Mathematical Geology, 31, 391–410. <doi:10.1023/A:1007538624271>

Examples

set.seed(2025)
xdat <- rnorm(1000)
thresh <- qnorm(c(0.8, 0.9, 0.95))
thselect.cbm(xdat, thresh, B = 50)