An implementation of the Fast DTW algorithm used in Chronix. This implementation is 40 times faster than the provided implementation at Google Code.
This implementation includes some performance optimizations. The following results show the benefits.
2 months of CPU load with 442.005 data points.
- Raw Dataset
Times are seconds.
Search Radius | Original | Chronix |
---|---|---|
1 | 35.448 | 0.694 |
5 | 37.629 | 0.719 |
10 | 43.572 | 0.843 |
15 | 41.557 | 1.028 |
20 | 40.303 | 1.139 |
25 | 39.797 | 1.338 |
30 | 40.597 | 1.567 |
This work is initially started by:
- Stan Salvador, [email protected]
- Philip Chan, [email protected]
- Florian Lautenschlager, @flolaut