decluster: Decluster a time series.

In theGreatWhiteShark/climex: Estimating Extreme Events in Climatic Time Series (using Data of the DWD)

Decluster a time series.

Description

Decluster point over threshold data as a preprocessing step for the fitting of the generalized Pareto (GP) distribution.

Usage

decluster(x, threshold, cluster.distance = NULL, silent = FALSE,
  mc.cores = NULL)

Arguments

x	Either an object of class xts or numeric or a list of those. The full time series has to be provided. Not just the exceedances over the threshold!
threshold	Numerical value. Has to be set sufficient high to fulfill the asymptotic condition for the GP distribution.
cluster.distance	Numerical value specifying how many points have to be below the threshold for the next point to be considered the starting point of a new cluster. Only supply a value when you really know what you are doing! Default = NULL
silent	Whether or not to display warnings.
mc.cores	A numerical input specifying the number of cores to use for the multi core application of the function (see detectCores). This functionality is only available if the input is a list of different objects. If NULL, the function will be calculated classically with only one core. Default = NULL.

Details

This function determines clusters in an object of class xts and extracts just their maximal values in order to remove short-range correlations. All position of exceedances will be considered belonging to a cluster until at least cluster.distance consecutive points fall below the threshold.

The argument cluster.distance will be determined using the extremal index as suggested in Ferro & Segers (2003) when set to NULL. It, thus, provides a non-parametric way of declustering.

It also features a special treatment of missing values. All of them will be kept and the returned time series will have the same length as the input. Separate missing values and small sequences will be omitted in the cluster determination. But if more than 15 missing values appear in a row, they will be replaced with the minimal value of the time series for the cluster detection. This way exceedances separated over a big temporal distance will not be considered to belong to the same cluster.

This function can be applied to lists of xts or numeric class elements too.

Value

Same class as the input. The time series will be similar to their original versions with all the elements within a cluster having smaller values than the clusters maximum being replaced by NA.

Author(s)

Philipp Mueller

See Also

Other extremes: block.list, block.xts, block, decluster.list, decluster.xts, extremal.index, gev.density, gpd.density, qevd, return.level.climex.fit.gev, return.level.climex.fit.gpd, return.level.list, return.level.numeric, return.level, revd, rlevd, threshold.list, threshold.xts, threshold, upper.limit.climex.fit.gev, upper.limit.climex.fit.gpd, upper.limit.list, upper.limit.numeric, upper.limit

theGreatWhiteShark/climex documentation built on July 13, 2022, 9:11 a.m.