News for hackers. Stuff that matters.

seanlaw

Discovering Multidimensional Time Series Motifs

Happy 2021! Please enjoy our brand new STUMPY tutorial on “Discovering Multidimensional Time Series Motifs”.This is a sophisticated matrix profile approach that allows you to find conserved behavior across the relevant time series dimensions.https://towardsdatascience.com/part-10-discovering-multidimensional-time-series-motifs-45da53b594bb

( Read More... | no comments | 2 points | on HN )

Stumpy Release v1.6.1 – Modern Time Series Analysis

Happy New Year! To thank our community and to welcome in 2021, we just released STUMPY v1.6.1! This version adds a ton of new features, hands-on tutorials, 100% test coverage, and more. The future of modern time series analysis is looking brighthttps://stumpy.readthedocs.io/en/latest/

( Read More... | no comments | on HN )
Re:Tutorial Series: Modern Time Series Analysis with Stumpy on Friday December 25, 2020 @05:07PM (#25539406)
STUMPY faithfully reproduces the matrix profile work (https://www.cs.ucr.edu/~eamonn/MatrixProfile.html) in a user friendly and scalable Python package. At the end of the day, there is no silver bullet but, now, you can quickly and easily compute matrix profiles (using distributed CPUs and multi-GPUs) and compare the results with your favorite time series analysis approach. That way, you can choose the right tool for the right data/right job.

Tutorial Series: Modern Time Series Analysis with Stumpy

Happy Holidays! Sit back, relax, and please enjoy our multi-part tutorial series on "Modern Time Series Analysis with STUMPY". Whether you're a visual learner or love hands-on walkthroughs, there's something for everyone!https://seanmylaw.medium.com/stumpy-fdb9f8f1f261

( Read More... | 2 comments | 4 points | on HN )

Consensus Time Series Motifs with Stumpy

Finding motifs (i.e., conserved patterns) in only two time series is so 2020! Check out this new STUMPY tutorial and learn how to discover a “consensus motif” that exists across a set of time series using the latest “ostinato” algorithm:https://towardsdatascience.com/time-series-consensus-motifs-8345152bd6ac

( Read More... | no comments | 2 points | on HN )

Stumpy v1.6.0 (For Modern Time Series Analysis)

STUMPY v1.6.0 is now available on Conda-Forge and PyPI with new features for computing a “consensus motif” across multiple time series, a novel distance measure called “MPdist” that can be used for clustering time series, and improved support/stability for Dask distributed matrix profile computations! Let us know what you think:https://stumpy.readthedocs.io/en/latest/install.html

( Read More... | no comments | on HN )

AB-Joins with Stumpy (Time Series Analysis)

"AB-Joins" are like SQL Joins but for two time series data sets. In this new post, we use STUMPY to analyze and compare Queen/Bowie's song "Under Pressure" with Vanilla Ice's "Ice Ice Baby" and discover remarkable similarities in the bass line:https://towardsdatascience.com/part-8-ab-joins-with-stumpy-af985e12e391

( Read More... | no comments | on HN )
Re:Significant Pattern Mining for Time Series on Thursday June 13, 2019 @12:30PM (#20176423)
Hi all, I am the creator of STUMPY and wanted to thank you for your interest. Please feel free to post questions on our Github issues and we'll try to assist where we can.