search

DynamicTimeWarping / dtw-python

Python port of R's Comprehensive Dynamic Time Warp algorithms package
https://dynamictimewarping.github.io
GNU General Public License v3.0
279 stars 35 forks source link

Efficient computation of a cross-distances for several time series #77

Closed vspinu closed 11 months ago

vspinu commented 11 months ago

Is my understanding correct that there current does not exist an eficient implementation of the cross distance matrix for several time series? Basically a parity for R's proxy:dist(..., method="DTW") described in section 3.7 of the vignete?

Would it be too difficult to add?

Thanks!

tonigi commented 11 months ago

Hi, I chose to leave it out because

. It's simple enough to implement . It was understandably confusing for some people . Dtw is not a distance in the mathematical sense . it couldn't handle TS of different lengths . Nor multivariate (vector) TS

The natural way to do it is to iterate a list of timeseries.

T

Il dom 5 nov 2023, 17:38 Vitalie Spinu @.***> ha scritto:

Is my understanding correct that there current does not exist an eficient implementation of the cross distance matrix for several time series? Basically a parity for R's proxy:dist(..., method="DTW") described in section 3.7 of the vignete https://cran.r-project.org/web/packages/dtw/vignettes/dtw.pdf?

Would it be too difficult to add?

Thanks!

— Reply to this email directly, view it on GitHub https://github.com/DynamicTimeWarping/dtw-python/issues/77, or unsubscribe https://github.com/notifications/unsubscribe-auth/AAU74CY53XMFNRZGQ5YADVLYC66IDAVCNFSM6AAAAAA66OJR26VHI2DSMVQWIX3LMV43ASLTON2WKOZRHE3TOOBWGUYDOMA . You are receiving this because you are subscribed to this thread.Message ID: @.***>