FindBestSplit: FindBestSplit

Description Usage Arguments

View source: R/binary_segmentation.R

Description

Takes a segment of the data and dispatches the choosen method to the different optimizers.

Usage

1
2
FindBestSplit(x, start, end, delta, n_obs, control, SegmentLossFUN,
  optimizer = c("line_search", "section_search"))

Arguments

x

A n times p matrix or data frame.

start

The start index of the given segment x.

end

The end index of the given segment x.

delta

Numeric value between 0 and 0.5. This tuning parameter determines the minimal segment size proportional to the size of the dataset and hence an upper bound for the number of changepoints (roughly 1/δ).

n_obs

The number of observations in the data set.

control

A list with parameters that is accessed by the selected optimizer:

  • stepsize: Numeric value between 0 and 0.5. Used by section search.

  • min_points: Integer value larger than 3. Used by section search.

SegmentLossFUN

A loss function as created by closure SegmentLoss.

optimizer

Which search technique should be used for performing individual splits in the binary segmentation alogrithm? Possible choices are

  • line_search: Exhaustive linear search. All possivle split candidates are evaluated and the index with maximal loss reduction is returned.

  • section_search: Iteratively cuts the search space according by a flexible ratio as determined by parameter stepsize in control parameter list and approximately finds an index at a local maximum. See Haubner (2018) for details.


lorenzha/hdcd documentation built on Sept. 2, 2018, 8:20 p.m.