mev (version 2.1)

tstab.cv: Coefficient of variation threshold stability plot

Description

This function calculates parametric estimates of the coefficient of variation with pointwise Wald confidence intervals along with empirical estimates and returns a threshold stability plot.

Usage

tstab.cv(
  xdat,
  thresh,
  method = c("empirical", "mle", "wcv", "cv"),
  nthresh = 10L,
  nsim = 99L,
  plot = TRUE,
  level = 0.95,
  ...
)

Arguments

xdat

[vector] vector of observations

thresh

[vector] vector of threshold. If missing, set to \(p^k\) for \(k=0\) to \(k=\)nthresh

method

[string], either moment estimator for the (weighted) coefficient of variation (wcv and cv) or maximum likelihood (mle)

nthresh

[integer] number of thresholds, if thresh is not supplied by the user

nsim

[integer] number of bootstrap replications

plot

[logical] if TRUE, returns a plot of the p-value path

level

[numeric] probability level for sequential testing procedure

...

additional parameters, notably for package boot, for the type of confidence intervals.

Examples

tstab.cv(
   xdat = rgp(1000),
   thresh = qgp(seq(0,0.9, by = 0.1)),
   method = "cv")
tstab.cv(
   xdat = rgp(1000),
   thresh = qgp(seq(0,0.9, by = 0.1)),
   method = "empirical")