Modeling an Analysis ProcessTo set up a calculation process (for example, in the context of an analytical application), you can model an Analysis Process. There are different types of sources, data targets, and transformations available for you to use for the modeling. When you execute the modeled analysis process, the data is transferred to the data target and transformed where necessary.
You model an analysis process in a specific context, that is, for a specific application, such as Filling CRM Attributes or Creating Target Group for BW Survey. Each analysis process is only valid for the application for which it was created. You also assign authorizations specific to the application (authorization object RSANPR). Which sources, data targets, and transformations are available to you depends on the application.
Sources provide the input data for an analysis process. Depending on the application, a query or the master data table of a characteristic, for example, can be used as the source.
You can use a transformation to change the data provided by the source. There are basic transformations that just change the order of the data (Sort) or restrict the data selection (Filter). Some transformations, however, change the structure of the data. The transformation Data Record - List is one such example, since it transforms a data record with many fields into a list of data records with smaller numbers of fields.
After the transformation, if any, you can write the data to different data targets. A data target can be a transactional ODS object, for instance, or an attribute in an SAP CRM system (for the application Filling CRM Attributes).
Microsoft Internet Explorer 5.5
For the definition of a link: Installation of the BW AddOn for the SAP GUI
For the graphical overview display of the field assignments (data flow arrow between two nodes): Adobe SVG Viewer
An analysis process is modeled using the following graphical elements:
Nodes
Sources, transformations, and data targets are
depicted as nodes, and the nodes are linked by data flow arrows. The different
types of nodes are depicted by different symbols. By double-clicking a
selected node, you call up a dialog box in which you then make the detailed
settings. In each dialog box, you can call up the documentation by choosing
the symbol
(with the quick info text Help). Each node has
one or more connectors (small red triangles) that you can select with the
mouse to link up the nodes. Incoming connectors (not used with sources) are
placed to the left of the nodes and outgoing connectors (not used with data
targets) to the right.
Data flow arrow
You use data flow arrows to link the nodes. In the case of data flow arrows that have a symbol, you need to explicitly define a field assignment between input and output nodes. You make these assignments in a dialog box that you call up by double-clicking the arrow. In the case of the other data flow arrows, field assignments are made automatically.
All analysis processes of the current application are listed on the left-hand side of the screen. Under Templates, you find the sample analysis processes delivered by SAP. These templates serve as an example of how you can define an analysis process for a specific application. You cannot change a template directly; you first need to copy it and then you can edit it.
On the
right-hand side of the screen, you find the nodes valid for a particular
application. You can use Drag&Drop to move nodes to the work area and then
use the mouse to link them. To delete a node, select it simply by clicking it
and choose the symbol
(with the quick info text Delete Selected
Objects.
During
modeling, especially when you use transformations, it can be useful to display
the output data. To do this, select the node concerned and choose the symbol
(with the quick info text Display Data for Selected Node). The
entire analysis process up to this node is then run. Note that, when using
this function with high volumes of data, it can terminate due to a timeout. In
such cases, use a filter to reduce the data quantity.
When you execute the analysis process (online or in the background), the data that has been determined is posted to the data target.
You can export a model into an XML file and then import this file back into the system. You may need to do this when transferring, for example, a model into another system. Currently, analysis process models cannot be transported.
You can use XML export and import to transfer a model to another application. Note that, for this, all of the nodes used also have to be available in the application where they are being imported.
You
import the XML file first by creating a model in the application and then by
entering the properties of the model. You then choose the symbol
(with the quick info text Import Analysis Process).
To create and execute a simple analysis process that does not involve transformation, proceed as follows:
...
1. Create a model.
2. Enter the model properties.
3. Move a source into the work area and make the detailed settings in the dialog box that appears.
4. Move a data target into the work area.
5. Use the mouse to link up the two nodes.
6. Double-click the data target and make the detailed settings in the dialog box that appears.
7. To make an explicit field assignment, double-click the data flow arrow linking the nodes.
8.
Select the
data target and choose the symbol
(with the quick info text Display Data for Selected Node) to
display the data that would be posted to the data target during actual
execution.
9. Execute the analysis process.
The data is posted and a log appears.