Threshold selection via SAMSEE

Smooth asymptotic mean squared error estimator of Schneider et al. (2021) for threshold selection. The implementation uses a second-order regular variation index of -1

Usage

thselect.samsee(xdat)

Arguments

xdat

vector of positive exceedances

Value

a list with elements

k0

optimal number of exceedances

shape

Hill estimator of the tail index

thresh0

selected threshold

References

Schneider, L.F., Krajina, A. & Krivobokova, T. (2021). Threshold selection in univariate extreme value analysis, Extremes, 24, 881–913 doi:10.1007/s10687-021-00405-7