Icon

01_​Accessing_​Transforming_​and_​Modeling_​Time_​Series

Accessing, Transforming and Modeling Time Series

This workflow shows how to access time series data, make it equally-spaced, impute missing values, aggregate it at a greater granularity, and explore it visually. After these steps, the time series is decomposed into trend, seasonality, and residual. The residual is modeled with an ARIMA model, and deployment data are saved for testing the model's out-of-sample forecast accuracy.



Visual Exploration - Line Plot, Seasonal Plot, Conditional Box Plot, Lag Plot Inspecting and Transforming Time Series Modeling Time Series Accessing to preprocessing, transforming, and modelingtime series Deploying Time Series Ascending by timeOriginal daily datamonthYearly seasonalityIntroduce missing daysSales is 0 wheremissingOnly 2017One value perdayyearTurning point, TrendSeasonal plotMonthly distributions Max lag = 12 (yearly seasonality)2017 for testingNumber rowsMean(Sales)ResidualOutliers in residualDecomposed time seriesRead Sample - Superstore dataDownload the dataset from Kagglevia the link in the workflow descriptionYear, month, weekas seasonal cycleslags.tabletraining.tabledeployment.tabletrend.pmml Sorter Line Plot AggregationGranularity Line Plot Timestamp Alignment Missing Value Date&Time-basedRow Filter GroupBy AggregationGranularity Line Plot Line Plot Pivoting ConditionalBox Plot Number To String Decompose Signal Rule-basedRow Splitter Extract TableDimension Column Filter Column Filter ARIMA Predictor Line Plot Numeric Scorer Auto ARIMA Learner Numeric Outliers Line Plot Line color Line color Line color Line color Excel Reader Extract Date&TimeFields Joiner Variable toTable Row Table Writer Table Writer Table Writer PMML Writer Visual Exploration - Line Plot, Seasonal Plot, Conditional Box Plot, Lag Plot Inspecting and Transforming Time Series Modeling Time Series Accessing to preprocessing, transforming, and modelingtime series Deploying Time Series Ascending by timeOriginal daily datamonthYearly seasonalityIntroduce missing daysSales is 0 wheremissingOnly 2017One value perdayyearTurning point, TrendSeasonal plotMonthly distributions Max lag = 12 (yearly seasonality)2017 for testingNumber rowsMean(Sales)ResidualOutliers in residualDecomposed time seriesRead Sample - Superstore dataDownload the dataset from Kagglevia the link in the workflow descriptionYear, month, weekas seasonal cycleslags.tabletraining.tabledeployment.tabletrend.pmml Sorter Line Plot AggregationGranularity Line Plot Timestamp Alignment Missing Value Date&Time-basedRow Filter GroupBy AggregationGranularity Line Plot Line Plot Pivoting ConditionalBox Plot Number To String Decompose Signal Rule-basedRow Splitter Extract TableDimension Column Filter Column Filter ARIMA Predictor Line Plot Numeric Scorer Auto ARIMA Learner Numeric Outliers Line Plot Line color Line color Line color Line color Excel Reader Extract Date&TimeFields Joiner Variable toTable Row Table Writer Table Writer Table Writer PMML Writer

Nodes

Extensions

Links

Download

To use this workflow in KNIME, download it from the below URL and open it in KNIME:

Download Workflow
Created by: Maarit Laukkanen
Created at: 2020-06-09
On NodePit since: 2024-06-28
Last update: 2024-10-17
Created with KNIME version: v4.5.2
Tags: time seriesARIMAgranularitytrendseasonalityclassical decompositionseasonal plotlag plot

Deploy, schedule, execute, and monitor your KNIME workflows locally, in the cloud or on-premises – with our brand new NodePit Runner.

Try NodePit Runner!