extRemes (version 1.65)

exi.intervals: Extremal Index Estimator

Description

Evaluates the intervals estimator for the extremal index.

Usage

exi.intervals(z, blocks = NULL)

Arguments

z

Logical vector indicating which positions correspond to extreme values.

blocks

vector describing physical blocks across which exceedance times should not be calculated. For example, if data are taken over only summer across years, then it might not make sense to consider the time between exceedances that cross from one summer to an

Value

• Estimate of the extremal index.

Warning

The estimator is not constrained to lie in [0,1] and a default value of 1 is returned if there are fewer than two extreme values.

References

Ferro, C. A. T. and Segers, J. (2003) Inference for clusters of extreme values. Journal of the Royal Statistical Society B, 65, 545--556.

See Also

decluster.intervals

Examples

x <- rnorm(1000)
exi.intervals(x > quantile(x, 0.9))

data(FCwx)
x <- FCwx[,"MxT"]
y <- FCwx[,"Year"]
# Extract just the July and August time points.
id <- is.element(FCwx[,"Mn"],7:8)
x <- x[id]
y <- y[id]
# Does not account for crossing from one year to the next.
exi.intervals(x)

# Takes account of crossing from one year to the next.
exi.intervals(x>93.5, blocks=y)