You can use the Predictor data screening features to:
Fill in values that should exist in historical data but do not, such as data missing for one month in a five-year series (see Selecting Data Attributes—Seasonality, Events, Screening)
Screen (exclude) outliers, values that differ significantly from the normal range of historical data
Specify the statistical algorithms used to fill in or screen data (see Setting Screening Options)
To examine the effects of filling in or screening data, and to change screening settings:
Click View Screened Data in the Data Attributes panel.
The Historical Data - Data Screening dialog opens. Any screened data values are highlighted in the chart.
Optional: Select Show screened data only to gray out unscreened data in the chart.
Optional: Click Screening Options to specify data filling and screening options. For details, see Setting Screening Options.