Seasonal and Non-Seasonal Exponential Smoothing
Click the Exponential smoothing & forecasting button in the Time Series Analysis Startup Panel to display the Seasonal and Non-Seasonal Exponential Smoothing dialog box, which contains six tabs: Quick, Advanced, Grid search, Automatic search, Autocorrelations, and Review Series.
Use the options in the Seasonal and Non-Seasonal Exponential Smoothing dialog box to perform exponential smoothing for time series with or without a seasonal component and trend (see also, Time Series Analysis Index). The implementation of these techniques in the STATISTICA Time Series module follows closely the comprehensive survey by Gardner, 1985; see exponential smoothing for details).
Click in the Variable or Long variable (series) name column to highlight the variable (series) in the active work area to be analyzed. Then, select the appropriate exponential smoothing model from among the options in the Model box (Quick tab or Advanced tab). For seasonal models, specify the seasonal lag (length of seasonal period) or accept the default lag of 12.
Element Name | Description |
---|---|
Summary: Exponential smoothing | Click the Summary: Exponential smoothing button to begin the exponential smoothing process. By default, 10 forecast values are computed (past the last observed value in the series). You can change the number of forecasts in the Forecast __ cases box on the Advanced tab. If the Add pred./errors to work area check box on the Advanced tab is selected, then when you click the Summary: Exponential smoothing button, the smoothed (forecast) values as well as residuals (observed minus smoothed values) are appended to the active work area. If you choose to append those series, make sure that there is sufficient space in the active work area for the respective variable, that is, that there are at least two (not locked) backups available. You can increase the Number of backups parameter (see below) if necessary. Also, be sure to Lock (see below) all important backups of the original variable (those that you want to retain); otherwise, those backups may be overwritten (refer to the section describing the memory management in Active Work Area). |
Active work area | The Active work area is composed of three columns: Lock, Variable, and Long variable (series) name. For detailed information on the Active work area, see Time Series Analysis - Active Work Area. |
Lock | Double-click on a transformed variable in the Lock column to lock or unlock the variable in the active work area. An L is listed in the column to indicate that the variable is locked. Locked variables will not be replaced as successive transformations exceed the current maximum number of backups. Note that original (untransformed) variables are always locked, and they cannot be unlocked. |
Variable | The Variable column contains the short name of the variable(s) currently available for analysis. Click on a variable name in the list to highlight that variable (series) in the Active work area that is to be analyzed For more information, see Active Work Area. |
Long variable (series) name | When a new (e.g., transformed) variable is appended to the active work area, the Long variables (series) name column will display a new long variable name that consists of the old long variable name (as much of it as will fit) and a brief description of the respective transformation that was performed. |
Number of backups per variable (series) | The parameter entered in the Number of backups per variable (series) box determines the maximum number of backups (i.e., previous versions of the series) that is kept in the active work area for each variable. After a transformation (or other analysis) is performed on a series, the resulting transformed series (or residuals, forecasts, etc.) will be appended to the active work area, and the values of the series prior to the transformation will be maintained as a backup. When a new transformation is appended and the maximum number of backups for the respective original variable is exceeded, the "oldest" previous backup will be replaced. You can lock a variable (prevent it from being overwritten) by double-clicking on the respective variable in the Lock column. For more details, refer to Active Work Area. |
Save variables | Click the Save variables button to save the variables (series) in the Active work area into a standard STATISTICA data file. You can save all variables or only selected variables. |
Delete | Click the Delete button to delete a highlighted transformed variable from the Active work area. Original (untransformed) variables cannot be deleted. |
Cancel | Click the Cancel button to return to the Time Series Analysis Startup Panel. |
Options | Click the Options button to display the Options menu. |