Calculation views offer a lot of flexibility, in particular the possibility to have a large number of nodes. As calculation views become more complex, maintenance can be challenging. The SAP Business Application Studio provides functionality to manage nodes and semantics. The following features are available:
| Feature | Purpose |
|---|---|
| Switching Node Types | Switch one node type with another without losing the link to the data source or the upper and lower nodes. |
| Replacing a Data Source | Replace one node source with another without losing the output columns, calculated columns, and so on, throughout all of the upper nodes. |
| Extract Semantics | Apply the semantics from an underlying node or data source to the Semantics node of the calculation view. |
| Propagate to Semantics | Map columns to the output across all upper nodes, up to the top node of the calculation view. |
| Previewing the Output of Intermediate Nodes | To troubleshoot problems in view design, preview the data of an intermediate node rather than the whole calculation view. |
| Map Input Parameters between nodes | Map input parameters of a view to input parameters defined in underlying (source) views of other nodes. |
Switching Between Certain Types of Node
It's possible to convert a Projection node into an Aggregation node and the other way around without losing the reference to the data sources of the node.
For the top node, it's also possible to switch to a Star Join node.
Replacing the Data Sources of a Node
Depending on the type of node and view configuration, you can replace the data source of a node (for example, a table, a Calculation view, or another node) by another data source (table, Calculation View).
This functionality is useful, for example, when you want to have the columns still propagated to all the upper nodes. You can also keep the existing calculated columns defined in this node working despite the change of data source.
When you do the change, a column mapping dialog box appears. It helps you assign the columns of the new data source to the output columns previously defined in the node.
Extracting Semantics
It's possible to extract the semantics from the underlying data sources. This especially useful when these underlying data sources include rich semantics and you don't want to have to redefine them again. You can choose to extract different types of semantics such as column labels and aggregation functions. It's also possible to extract hierarchies and variables from the included calculation views. To allow this, the corresponding columns must be available in the current view (the "extracting" view). For example, it doesn't make sense to extract a hierarchy based on product if the column product doesn't exist in the extracting calculation view.
Propagate to Semantics
This feature is helpful in calculation views with multiple stacked nodes. The purpose is to map one or several columns, already in the output of a node, to the output of all upper nodes right up to the Semantics node.
This is an alternative approach, and less tedious than adding manually, the same column(s) to the output of each of the nodes in a calculation view.
To do this, select one or several columns in the output of a node. Then right-click the selection, and choose Propagate to Semantics.
Previewing the Output of Intermediate Nodes
To understand and fine-tune the different steps of a calculation view, each node can be previewed independently from the calculation view itself.
To achieve this preview, the calculation view must have already been successfully deployed. Be sure to select the node header and not the data source within the node, to preview the results of the node. If you select the data source and use Data Preview, you'll see the data in the original data source and not the results of the node output.
Map Input Parameters Between Nodes
You can map input parameters (and also variables) defined in a calculation view to the input parameters and variables defined in the source views. You can do this to pass the parameter value itself, but also to pass the relevant list of values to the input parameter dialog box at runtime (when the view is executed).
Note
In the Manage Mappings view, you can even copy and map an input parameter from a source view.
That is, you don't need to create a new input parameter in the calculation view (and define all its settings) before mapping it. This task can be done automatically in just one click.
Copy and Paste Part of a Scenario
It's possible to copy and paste a node within the Scenario pane of the same calculation view. If this node uses other nodes as its data sources, it's possible to decide whether you want to copy this node only, or the node and all nodes below, together with the connections between them.
Insert a Node between Two Other Nodes
In the Calculation View scenario, you can insert a new node between two existing nodes that are already linked. By doing this, you keep the columns consistent along the entire scenario.
Note that this is only available as a graphical feature: You must drag the new node from the palette and drop it exactly onto the existing link between the existing nodes.
For example, if you've started designing a CUBE calculation view where the aggregation node uses a join node as its data source. You've already mapped the required columns, defined the aggregate functions, semantics, and so on. Then you realize you need a projection to calculate a column before aggregating. In that case, by adding the new Projection between the Join node and the Aggregation node (instead of basically removing the Join node from the aggregation node and adding it to a new projection node), you keep most of the work you already made and save time.
Navigating to Source Calculation Views
It's possible to open a calculation view directly from calculation scenario of another view that consumes it. To do this, in the Scenario pane, right-click the data source (Calculation View) in the node that consumes it, and choose Open.