book

Time Series Indexing

by Mihalis Tsoukalos

June 2023

Intermediate to advanced

248 pages

5h 41m

English

Packt Publishing

Read now

Unlock full access

ContributorsAbout the authorAbout the reviewer
Who this book is forWhat this book coversTo get the most out of this bookDownload the example code filesDownload the color images Conventions usedGet in touchShare your thoughtsDownload a free PDF copy of this book
Technical requirementsUnderstanding time seriesTime series are everywhereEssential definitionsTime series data miningComparing time seriesThe Euclidean distanceThe Chebyshev distanceWhat is an index and why do we need indexing?The Python knowledge that we are going to needTiming Python codeAn introduction to AnacondaThe required Python packagesSetting up our environmentPrinting package versionsCreating sample dataPublicly available time series dataHow time series are processedReading time series from diskIs all data numeric?Do all lines have the same amount of data?Creating subsequencesVisualizing time seriesWorking with the Matrix ProfileExploring the MPdist distanceSummaryResources and useful linksExercises
Technical requirementsThe required theoryWhy do we need SAX?NormalizationVisualizing normalized time seriesAn introduction to SAXThe cardinality parameterThe segments parameterHow to manually find the SAX representation of a subsequenceΗow can we divide 10 data points into 3 segments?Reducing the cardinality of a SAX representationDeveloping a Python packageThe basics of Python packagesThe SAX Python packageWorking with the SAX packageComputing the SAX representations of the subsequences of a time seriesCounting the SAX representations of a time seriesThe tsfresh Python packageCreating a histogram of a time seriesCalculating the percentiles of a time seriesSummaryUseful linksExercises
Technical requirementsBackground informationTrees and binary treesUnderstanding how iSAX worksThe cardinality parameterThe segments parameterThe threshold parameterComputing the normalized mean valuesHow big can an iSAX index get?What happens when there is no space left for adding more subsequences to an iSAX index?How iSAX is constructedHow iSAX is searchedPromotion strategySplitting nodesManually constructing an iSAX indexUpdating the counting.py utilitySummaryUseful linksExercises
Technical requirementsA quick look at the iSAX Python packageThe class for storing subsequencesThe class for iSAX nodesThe class for entire iSAX indexesExplaining the missing partsExploring the remaining filesThe tools.py fileThe variables.py fileThe sax.py fileUsing the iSAX Python packageReading the iSAX parametersHow to process subsequences to create an iSAX indexCreating our first iSAX indexCounting the subsequences of an iSAX indexHow long does it take to create an iSAX index?Dealing with iSAX overflowsSummaryUseful linksExercises
Technical requirementsHow the sliding window size affects the iSAX construction speedChecking the search speed of iSAX indexesJoining iSAX indexesImplementing the joining of iSAX indexesExplaining the Python codeUsing the Python codeWe have a long list of Euclidean distances, so what?Saving the outputFinding iSAX nodes without a matchWriting Python testsWhat are we going to test?Comparing the number of subsequencesChecking the number of node splitsAll Euclidean distances are 0Running the testsSummaryUseful linksExercises
Technical requirementsStoring an iSAX index in JSON formatDownloading the JavaScript code locallyRunning the code locallyVisualizing an iSAX indexVisualizing iSAX as a treeTrying something radicalMore iSAX index visualizationsUsing icicle plotsVisualizing iSAX as a Collapsible TreeSummaryUseful linksExercises
Technical requirementsUnderstanding the Matrix ProfileWhat does the Matrix Profile compute?Manually computing the exact Matrix ProfileComputing the Matrix Profile using iSAXWhat happens if there is not a valid match?Calculating the errorApproximate Matrix Profile implementationComparing the accuracy of two different parameter setsUnderstanding MPdistHow to compute MPdistManually computing MPdistCalculating MPdist using iSAXImplementing the MPdist calculation in PythonUsing the approximate Matrix Profile wayUsing the join of two iSAX indexesUsing the Python codeComparing the accuracy and the speed of the methodsSummaryUseful linksExercises
Concluding all that we have learned so farOther variations of iSAXInteresting research papers on time seriesInteresting research papers on databasesUseful booksUseful books on databasesBuilding a strong computer science backgroundBooks on UNIX and LinuxBooks on the Python programming languageSummaryUseful linksExercises

Other Books You May EnjoyPackt is searching for authors like youShare your thoughtsDownload a free PDF copy of this book

Overview

Time series data is at the heart of many applications, from finance and system monitoring to weather forecasting and medical data analysis. "Time Series Indexing" offers a hands-on guide to implementing and leveraging the iSAX indexing technique in Python to efficiently manage, search, and analyze time series data.

What this Book will help me do

Gain the know-how to implement algorithms like SAX and iSAX with illustrative Python examples.
Learn to construct robust time series indexes tailored to real-world data sets.
Understand the theoretical underpinnings of time series processing and indexing techniques.
Explore and employ visualization techniques to interpret time series structures and insights.
Gain the skills to adapt iSAX methodologies to other programming environments and practices.

Author(s)

Mihalis Tsoukalos is an accomplished developer and author specializing in Python programming and data processing techniques. With years of experience translating complex academic research into practical applications, Mihalis excels at bridging the gap between theory and practice. His writing approach ensures readers grasp both the foundational principles and the hands-on methods needed to succeed.

Who is it for?

This book best suits researchers, analysts, and developers who work with time series data and seek to elevate their proficiency in indexing and managing such data. It is perfect for professionals with a foundational knowledge of Python and programming concepts. This material also supports learners eager to derive actionable insights from theory-heavy academic research.