Comprehensive implementation of Dynamic Time Warping (DTW) algorithms in R.
DTW finds the optimal (least cumulative distance) mapping between a given query into a given template time series.
Most variants of the algorithm are supported: symmetric, asymmetric and
custom step patterns, with weighting (see stepPattern).
Supports windowing: none, "Itakura" parallelogram, Sakoe-Chiba band,
custom (see dtwWindowingFunctions). Handles query and
template of arbitrary lengths. Multivariate matching and arbitrary
definition for a distance function are supported via user-supplied local
distance matrix.
Package provides minimum cumulative distance, warping function, plots, etc. A fast, compiled version of the algorithm is normally used. Should it not be available, a slower pure-R equivalent is automatically used as a fall-back.
Please see documentation for function dtw, which is the
main entry point to the package.
If you use this software, please cite it according to
citation("dtw"). The package home page is at
To get the latest stable version from CRAN, use
install.packages("dtw"). To get the development version
(possibly unstable), use
install.packages("dtw",repos="http://r-forge.r-project.org").
dtw for the main entry point to the package;
dtwWindowingFunctions for global constraints;
stepPattern for local constraints;
distance, outer for
building a local cost matrix with multivariate
timeseries and custom distance functions.