If the METHOD=WINTERS or METHOD=ADDWINTERS option is specified, and the SEASONS= option is incorrectly specified based on the INTERVAL= frequency, the FORECAST procedure issues an error message such as the following:

ERROR: Read Access Violation FORECAST

ERROR: An exception has been encountered.
Please contact technical support and provide them with the following traceback information:

The SAS task name is [FORECAST]
Segmentation Violation

When running in SAS Studio, the procedure might instead result in the error below:

The issue occurs under either of the following conditions when the METHOD=WINTERS or METHOD=ADDWINTERS option is specified:

• The input data frequency is annual or lower and the SEASONS= option is specified. For example, INTERVAL=YEAR and SEASONS=MONTH.
• The interval specified in the SEASONS= option has a higher frequency than the interval specified in the INTERVAL= option. For example, INTERVAL=QTR and SEASONS=MONTH.

In the first specification above, month-of-year seasonality is requested for annual data, which cannot be computed because there is no information about the month-to-month variation in the annual data. In the second specification above, month-of-year seasonality is requested when the input data is quarterly. It also cannot be computed for the same reason.

To circumvent the problem, only specify a seasonal model if the input data has a frequency that supports a seasonal model. If the data has a frequency that is higher than annual and therefore supports a seasonal model, then specify a frequency in the SEASONS= option that is the same or lower than the frequency of the input data.

The FORECAST procedure is obsolete. Please consider using the ESM procedure to fit exponential smoothing models.