After you run a Predictor forecast, you can adjust the forecasted data to customize it for your particular situation. For example, you could add 50 to each forecasted value, or you could round each value to the nearest hundreds. Adjustments are applied to all methods for that series.
To adjust forecasted data:
Run a Predictor forecast and display the Predictor Results window.
In the Adjust Forecast for Series dialog, change any or all settings in the listed order:
1. Percentile of forecast range—See the note below for an explanation; select Custom to enter a percentile into the Custom Percentile dialog (Default = Median)
2. Adjust values by—Increases or decreases each value by the specified amount (Default = 0.00)
3. Round values to—Rounds values to the specified numeric place; for example Integer rounds to the nearest number in units place (Default = No rounding; select Custom to specify a decimal place as described in Custom Rounding)
4. Restrict values to range—Limits adjusted values to the specified range (Default = –Infinity to Infinity)
Note:
When a time-series forecast is pasted into a model as a Crystal Ball assumption, each forecast value is assumed to be the median, or 50th percentile, of a normal distribution assumption. The percentile adjustment is ignored in this case.
Optional: Click Apply to All Series to apply the settings to all data series, except for dependent variables in a regression analysis.
Optional: Click Defaults to restore all the default settings.