Select Dimension Values (Temporal Causal Modeling)
For multidimensional data, you can customize the analysis by specifying which dimension values apply to a particular metric field with a particular role. For example, if sales is a metric field and channel is a dimension with values 'retail' and 'web', you can specify that 'web' sales is an input and 'retail' sales is a target. You can also specify dimension subsets that apply to all metric fields used in the analysis. For example, if region is a dimension field that indicates geographical region, then you can limit the analysis to particular regions.
- All values
- Specifies that all values of the current dimension field are included. This option is the default.
- Select values to include or exclude
- Use this option to specify the set of values for the current dimension field. When
Include is selected for the Mode, only
values that are specified in the Selected values list are included.
When Exclude is selected for the Mode, all
values other than the values that are specified in the Selected
values list are included.You can filter the set of values from which to choose. Values that meet the filter condition appear in the Matched tab and values that do not meet the filter condition appear in the Unmatched tab of the Unselected values list. The All tab lists all unselected values, regardless of any filter condition.
- You can use asterisks (*) to indicate wildcard characters when you specify a filter.
- To clear the current filter, specify an empty value for the search term on the Filter Displayed Values dialog.
To specify a dimension subset
From the menus, choose:
- Specify the dimension fields that identify the time series.
- Select the dimension field for which you want to specify a dimension subset.
- Click Specify Subset...
To specify dimension values for a metric field
- Select the Time Series settings on the Fields tab.
- Click the ellipsis button in the row for the metric and dimension field for which you want to specify the dimension values.