Contents x
    Forecast
    • 06 Jan 2023
    • 1 Minute to read
    • Print
    • Dark
      Light
    Contents

    Forecast

    • Updated on 06 Jan 2023
    • 1 Minute to read
    • Print
    • Dark
      Light
    Article summary

    Description

    The Forecast node is used to apply a forecast model to a data set.


    Configuration Options

    Basic Configuration Options

    SettingDescription\Parameters
    Split ColumnsThe selected column(s) will be used to split the data set over to build additional models.
    Target ColumnThe column we are predicting.
    Sort By ColumnsSort data rows by these columns.
    Outliers ColumnColumn to set an outlier (0/1).
    PredictorsThe columns we are using to predict.

    Advanced Configuration Options

    Model Options

    SettingDescription
    Model TypeThe model type.

    Options
    • Auto
    • Seasonal Causal
    Consider Linear Seasonality
    Periodicity
    Force Non-Negative Forecasts
    Prune Weak Predictors
    Include Linear TermMust be included if no predictors are specified.
    Standardize PredictorsStandardize predictors between 1 and 0.
    Auto-lag Predictors
    Fill Target Zeroes With

    Advanced Options

    SettingDescription
    Automatically Detect/Exclude Outliers
    Minimum MAPE Improvement for Outlier
    Seasonality Extraction BandwidthBandwidth is the fraction of all data points for calculating the localized regression used in smoothing (loess). Smaller numbers = less smoothing.
    Trend Smoothing BandwidthBandwidth is the fraction of all data points for calculating the localized regression used in smoothing (loess). Smaller numbers = less smoothing.

    Actions

    ActionDescription
    PreviewAfter configuring the node, the combined result set can be previewed by clicking the Preview button.
    Was this article helpful?
    What's Next