compute_envelope: Time series warping envelopes

In dtwclust: Time Series Clustering Along with Optimizations for the Dynamic Time Warping Distance

Time series warping envelopes

Description

This function computes the envelopes for DTW lower bound calculations with a Sakoe-Chiba band for a given univariate time series using the streaming algorithm proposed by Lemire (2009).

Usage

compute_envelope(x, window.size, error.check = TRUE)

Arguments

x	A univariate time series.
window.size	Window size for envelope calculation. See details.
error.check	Logical indicating whether the function should try to detect inconsistencies and give more informative errors messages. Also used internally to avoid repeating checks.

Details

The windowing constraint uses a centered window. The calculations expect a value in window.size that represents the distance between the point considered and one of the edges of the window. Therefore, if, for example, window.size = 10, the warping for an observation x_i considers the points between x_{i-10} and x_{i+10}, resulting in 10(2) + 1 = 21 observations falling within the window.

Value

A list with two elements (lower and upper envelopes respectively): lower and upper.

Note

This envelope is calculated assuming a Sakoe-Chiba constraint for DTW.

References

Lemire D (2009). “Faster retrieval with a two-pass dynamic-time-warping lower bound .” Pattern Recognition, 42(9), pp. 2169 - 2180. ISSN 0031-3203, \Sexpr[results=rd]{tools:::Rd_expr_doi("10.1016/j.patcog.2008.11.030")}, https://www.sciencedirect.com/science/article/pii/S0031320308004925.

Examples

data(uciCT)

H <- compute_envelope(CharTraj[[1L]], 18L)

matplot(do.call(cbind, H), type = "l", col = 2:3)
lines(CharTraj[[1L]])

dtwclust documentation built on Sept. 11, 2024, 9:07 p.m.