DTW: Dynamic Time Warping for Quaternion Time Series

In astamm/squad: Statistics for Quaternion Temporal Data

Dynamic Time Warping for Quaternion Time Series

Description

This function evaluates the Dynamic Time Warping (DTW) distance between two quaternion time series (QTS).

Usage

DTW(
  qts1,
  qts2,
  resample = TRUE,
  disable_normalization = FALSE,
  distance_only = FALSE,
  step_pattern = dtw::symmetric2
)

Arguments

qts1	An object of class qts.
qts2	An object of class qts.
resample	A boolean specifying whether the QTS should be uniformly resampled on their domain before computing distances. Defaults to TRUE.
disable_normalization	A boolean specifying whether quaternion normalization should be disabled. Defaults to FALSE which ensures that we always deal with unit quaternions.
distance_only	A boolean specifying whether to only compute distance (no backtrack, faster). Defaults to FALSE.
step_pattern	A dtw::stepPattern specifying the local constraints on the warping path. Defaults to dtw::symmetric2 which uses symmetric and normalizable warping paths with no local slope constraints. See dtw::stepPattern for more information.

Details

If no evaluation grid is provided, the function assumes that the two input QTS are evaluated on the same grid.

Value

An object of class dtw::dtw storing the dynamic time warping results.

Examples

DTW(vespa64$igp[[1]], vespa64$igp[[2]])

astamm/squad documentation built on Jan. 26, 2024, 5:30 p.m.