Modeling Scenarios (Desktop Application)

Scenario modeling is the modeling, which considers several alternative ways of model development. It is one of the most effective tools in strategic analysis.

Modeling scenarios are objects that are used to create a series in an output variable, to which the data should be loaded after calculation of the modeling problem is completed.

To manage all the scenarios that are in included in the modeling container, use the modeling scenarios panel.

All scenarios, by which modeling problem is calculated, are displayed on the Scenarios panel:

NOTE. This panel displays only the scenarios the user has read permissions for. If a parent object of the modeling container is a time series database, access permissions to the Fact scenario are not considered.
On copying the modeling container only the scenarios are copied, to which the user has permissions.

By default, the modeling container has only the Fact scenario, which is not displayed on this panel. The Fact scenario is used in all variables, models and modeling problems to get data about sample period. If other scenarios are absent, calculation results may be loaded according to the Fact scenario.

To get calculation results for the entire forecasting period by all output variables and a certain scenario, take into account features of calculated models. Depending on whether scenario data is necessary, models can be classified as follows:

Forecasting models which do not consider the use of factors (ARIMA, Exponential Smoothing, Grey Forecast, and so on). To identify the model, the input variable must contain data for the entire sample period by the Fact scenario. Results of model data calculation by different scenarios will not differ.
Forecasting models which consider the use of factors (Determinate Equation, Linear Regression (OLS Estimation), Non-Linear Regression (Non-Linear OLS Estimation), and so on). To identify the model, all input variables must contain data for the entire sample period by the Fact scenario (identification for determinate equation is not required). To calculate forecasting values by scenarios, input variables (factors) must contain data for the forecasting period by all used scenarios. If the output variable is used as a factor with lag, calculation of the first forecast point by the required scenario in the output variable requires data for the last point for the sample period (in accordance with the lag value) by the Fact scenario.

Operations with Scenarios

Drag a scenario from the Modeling Scenarios panel.
Click the Add button or select the same context menu item. The Calculation Scenarios Selection dialog box opens that contains the list of all available scenarios in the current modeling container:

Select checkboxes of the scenarios to be used in the problem. The context menu can be used to work with a selection:

- Select. It selects checkboxes for selected tree elements. If a folder is selected, its child elements are also selected.
- Deselect. It deselects checkboxes of selected tree elements. If there is a folder among them, it is selected together with its child elements.
- Select All. It selects all elements.
- Deselect All. It deselects all elements.

Click the Copy button.
Select the same context menu item.

The dialog box opens:

Determine the parameters:

- Source. Specify the scenario data, which should be copied.
- New Scenario. Set name and identifier of the new scenario, to which the data should be copied.
- Variables. The list includes all variables that are used to calculate a modeling problem. Check the variables, to which data should be copied.
- Select All. Deselect All. The buttons are used to work with elements selection in the variables' menu.