Random block maxima shape estimator of Wager

Computes the shape estimator for varying k up to sample size of maximum kmax largest observations

Usage

shape.rbm(xdat, k = 10:floor(length(xdat)/2), ...)

Arguments

xdat: [vector] sample exceedances
k: [int] vector of integers for which to compute the estimator
...: additional parameters, currently ignored

Value

a list with elements

k: number of exceedances
shape: tail index estimate, strictly positive
risk: empirical Bayes estimate of risk
thresh: threshold given by the smallest order statistic considered in the sample

References

Wager, S. (2014). Subsampling extremes: From block maxima to smooth tail estimation, Journal of Multivariate Analysis, 130, 335-353, doi:10.1016/j.jmva.2014.06.010