Getting Started in APD
It is hoped that this basic step through of creating an APD will be helpful for first-time users
Analysis Process Designer allows the user to create new information out of existing data. It is possible to transform the information available so that the new data is relevant for the users requirements. The results can then be written back to the BW database or can be made available in other operational systems.
The user creates the APD within transaction RSANWB. Within the APD workbench, you can easily manipulate your data by combining the available steps into one single data process. Data sources can be any infoprovider or even a query which allows you to further combine data to be used in new scenarios.
Getting started ...
Transaction RSANWB :
Now, you have 4 options here ~ Refresh Create Find Find Again :
To CREATE a new APD, use the button :
Select the ‘application’ which is the folder (on the left) where the APD will be located.
For showing purposes, I’m using the ‘General’ folder and then click continue.
This brings you to the APD design screen:
When you create your APD, you will read data from data sources and run transformations on this data and the resulting data will be saved in data targets. Above you can see that for each area there are multiple types to choose from. Below is the options available in each area.
Data Sources are:
- Attributes of a Characteristic
- Data from Inforpovider
- Query to Read Data
- Data from File
- Data from Database Table
Transformations allowed are:
- Restrict Amount of Data (Filter)
- Aggregate Data
- Join from Two Data Sources
- Unify from Two Data Sources (UNION)
- Hide or Rename Columns
- Sort Data
- Transform List into Data Record
- Transform Data Record into List
- ABAP Routine
- ABC Classification
- Weighted Table Scoring
- Regression Analysis
- Prediction with Decision Trees
- Prediction with Cluster Model
- Prediction with 3rd Party Data Mining Model
Data Targets available:
- Change Attributes of Characteristic
- Write Data Directly to DSO
- CRM Attribute
- Train Decision Tree
- Train Cluster Model
- Create Association Analysis Model
- Train Scoring Model
- Train 3rd Party Data Mining Model