Transforming Lists into Data Records

Transformation allows you to change the internal data structure. This creates a one-line data record at the node exit from a list of table rows at the node entrance.

Transferring a list of data records to a data record is done using a transformation field, usually a characteristic. You also select a transposition field - usually a key figure - whose content is placed in the new fields of the output structure depending on the characteristic value of the transformation field.

Every combination of a characteristic value of the transformation field and the selected transposition field can be assigned a new field in the output structure of the node. The new field has the same technical definition as the transposition field. The values of all non-selected fields are forwarded to the subsequent node without being changed.

Input and output structures differ in transformation of a list into a data record. The transformation field and the transposition field are not passed on to the subsequent node. Instead only the selected fields and the new fields that are defined in the details are passed on. The structure of the data is wider in the data record form, meaning that the internal tables for the analysis process contain fewer data records but more fields. This can cause problems to occur, especially if characteristics with lots of characteristic values are used as transformation fields. If new field names are now specified for all or a number of characteristic values, the maximum permitted length of the data structures can be exceeded. You can then expect error messages from the process.

1. On the Definition tab page, select a transformation field from all of the available fields which can be used to fill the newly generated fields in the output structure.

2. Select a transposition field.

3. On the Details tab page, assign a new field for the flat data record to every combination of a characteristic value of the transformation field and a transposition field. Enter a technical name for each new field. If you do not enter a field name, the value for this combination will not be passed on and will not be retained when you exit the application.

The questionnaire for a survey contains questions and answers in table form. Every record thus contains the number of the question as the characteristic and the value of the answer to this question as the key figure. For evaluation purposes, the data has to be in a transformed form where the answer to each question has its own field. To do this, the transformation into a data record can be used.

In this example, the questionnaire contains three questions. If you take over the questionnaire number directly and define the question number as the transformation field and the answer as the transposition field, this output structure for this node would have the following form: