Refactored Backend Scalable, Bugfixes

v0.6

[ENH, BUG] New Backend "scalable" Implementation, Multivariate Suppor…

Add Naive Support for Multivariate TS

This does not implement sub-dimensional motif discovery.

Please refer to our other publication if you are interested in sub-dim MD: https://github.com/patrickzib/leitmotifs

v0.5.0 - New Distance Measures and improved Plotting using Down-Sampling

• Adds support for four distance measures: ED, z-normed ED, Cosine, Complexity Invariant Distance (CID)
• Improves Plotting by using down-sampling for (too) large time series

v0.4.1 Bugfix Release

• Bumps version numbers of dependencies
• Fixes a bug with auto length detection
• Faster k-NN search O(n) average case instead ok O(kn)

v0.4 Large Time Series Support

Adds support for large time series using a sparse distance matrix representation storing only needed pairs.

v0.3 Adds an Object Oriented API for simple use.

Features:

• Adds an Object Oriented API for simple use.
• Refactored motiflets

v0.2.7

Fixes a bug in window length selection. Now uses the position of the k-motiflet for length l+1 to compute the extent for length l as initial bound on distance.

v0.2.6 - Speedup part 2

Numbarize get_approximate_k_motiflet

v0.2.5 Speedup Part 1

• Faster distance computations
• Introduce slack as a parameter

v0.2.3

Rerun notebooks