follow_ups: Postprocessing for changepoint detection

In ChrisKust/rexpar: Simplicial Depth for Explosive Autoregressive Processes

Description

The function collects subsequent marked candidate points in the changepoints_lin2 function to calculate final estimates of the change points.

Usage

follow_ups(dat, mincper, steps, mincp)

Arguments

dat	Marked candidate points which could be changepoints based on the method selected in the changepoints_lin2 function.
mincper	A value for a minimal percentage of the size of the largest marked cluster to limit the size of clusters used for the final changepoints.
steps	Steps in which the candidates are evaluated.
mincp	This value defines which cluster, sorted by length, is relevant for the mincper-treshold. I.e. Just clusters with a lenght of at least 'mincper' of the 'mincp' longest cluster are considered as final estimates.

Details

This function is just used in the changepoint detection algorihms internally.

Value

clusters	Sizes of the jump clusters.
jumps	Detected changpoint locations based on the median values in each relevant cluster.

Author(s)

Kustosz, Christoph

References

Kustosz, C. (2016). Depth based estimators and tests for autoregressive processes with application. Ph. D. thesis. TU Dortmund.

See Also

changepoints_lin2, changepoints_lin2_cl

Examples

x <- c(1, 2, 3, 4, 7, 8, 11, 35, 40, 50, 60)
follow_ups(x, 0.75, 1)

ChrisKust/rexpar documentation built on May 6, 2019, 11:48 a.m.