backtrace: Back-tracing step of the 'segmenTier' algorithm.

In segmenTier: Similarity-Based Segmentation of Multidimensional Signals

Description

back-tracing step: collect clustered segments from the scoring matrix S(i,c) by back-tracing the position j=k which delivered the maximal score at position i.

Usage

backtrace(S, K, multib, nextmax = FALSE, verb = TRUE)

Arguments

S	matrix S, containing the local scores
K	matrix K, containing the position k used for score maximization
multib	if multiple k produce the maximal score, take either the shortest k ("max") or the longest k ("min"); if multib is set to "skip" the next unique k will be searched
nextmax	proceed backwards while score is increasing before opening a new segment
verb	print messages

segmenTier documentation built on May 2, 2019, 2:49 p.m.