依你的数据,beta是0.4070101
qgpd()
function (p, xi, mu = 0, beta = 1)
{
mu + (beta/xi) * ((1 - p)^(-xi) - 1)
}
beta就改采用0.4070101
tailplot()的scale是用来计算location
threshold <- x$threshold
xi <- x$par.ests["xi"]
beta <- x$par.ests["beta"]
prob <- x$p.less.thresh
scale <- beta * (1 - prob)^xi
location <- threshold - (scale * ((1 - prob)^(-xi) - 1))/xi
依此公式,你也可以自己算出location