Add a Forecast to a Visualization

Add forecasts to your workbooks based on Auto-Regressive Integrated Moving Average (ARIMA), Seasonal ARIMA, Exponential Triple Smoothing (ETS), or Prophet. For example, you might want to forecast summer temperatures based on data from previous summers.

Before you start, make sure that you have the required data in your visualization for the type of analytics you want to add. For a forecast, you need a time dimension and a measure. For example, you might have Week as a time dimension and Amount of rainfall (in millimetres) as the measure. If you don't have the required data columns in your visualization, when you click Forecast in the Analytics pane, you see this message: "This visualization does not support Forecast".
  1. On your home page, hover over a workbook, click Actions, then select Open.
  2. In the Data Panel, click AnalyticsUse the Analytics icon on the Data pane or Properties pane to view statistical analytics available..
    Description of stat_analytics-png.png follows
    Description of the illustration stat_analytics-png.png

  3. Drag and drop Forecast from the Analytics pane to a visualization.
  4. Use Analytics pane in Grammar Panel to change the default settings for the model.
    For example, you can change the default model type Seasonal ARIMA to Prophet, or use Periods to specify the number of periods to forecast.
    Description of stat-analytics3-png.png follows
    Description of the illustration stat-analytics3-png.png