Hero Backgroud Elements 2
Reading Time: 4 Min

How to Start Modeling your Data Using the Graphical View

October 1, 2019

Learning
Author
Helena Nacinovic Helena Nacinovic
Helena is a Solution & Product Marketing Senior Specialist and Digital Learning Designer at SAP. An enthusiast about learning new things, she's passionate about digital media, technology, and digital learning in all its forms.

 

It’s easy to start using the business semantic layer in SAP Data Warehouse Cloud and add business attributes to your data. Check out what the business semantic layer is here.

Your first step is to go to the Data Builder and select the Space in which you want to model your data.


Then, click on New Graphical View.


The first step now is to use the left side panel to select your data sources. You can either use the data already in the Repository of this Space, or select from the Sources connected to this Space.

 

Data Modeling

To start modeling your data, select either one table or view, click on it, and drag it to the canvas in the middle of the screen. Once you drop it on the canvas, an Output node will be automatically created and linked to the table or view you dragged to the canvas.


You can now drag and drop other tables or views on top of the one already in the canvas. As you drag the first table or view, you will see a pop-up listing Union and Join.


This is your opportunity to choose to make a union of these tables, which would mean an output with the complete contents of both tables or views. If you don’t specifically choose a Union, all tables and views will be joined by default. You don’t have to specifically choose Join, just drop the table and it will automatically be a join. Once you have your first join, you will notice that the canvas rearranges itself to show a Join node and a Projection node. If you click on the Join node, you can see the right sidebar show details of this join, including which columns are mapped together. You can either change the type of join or change the mappings between the two tables.

If you want to delete an existing mapping, just click on the line between the two columns and then click on the trash icon in red.

After you add all of your data sources and determine the ways in which these tables or views are connected (or joined), you might notice some yellow alerts on top of some of the Projection nodes.

This usually means that your join includes some columns which have exactly the same technical name. You can click on one of the Projection nodes with a yellow alert, then click on the shield icon to see what the issue is.

To solve this issue, you can either rename or hide a duplicate column. If you are sure that the columns have exactly the same information, you should hide it. To do that, just click on the column name, then click on Set Hidden. If, however, the column contents are similar, but not exactly the same, you can rename the column to make sure the information is present on your output but does not create issues with duplication.

To rename a column, click on the Projection node, then click on the column you want to rename. This will highlight in the canvas from which table or view this table is coming from.

Now that you are sure this is the right column from the right table, just click on the three dots icon and click on Change Business Name.

Data Transformation and Preview

If you want to make sure your data output includes only the right data, you can drill down into your data by choosing to add filters to your join, projection or output nodes. Just click on the node you want to filter, then click on the Filter icon right next to the node.

Now you will see the Filter Properties on the right sidebar. You can filter based on columns or type in your expressions as needed. All of your changes here are automatically saved, so when you are done, just go to the next step.

Another transformation you can do with SAP Data Warehouse Cloud is to create a new calculated column and add it to your output. To start, just click on the fx icon next to the output, a projection or a join node.

Then, on the right sidebar, you can select an existing column to add a calculation to, and click on the pencil icon next to the column. Alternatively, you can click on the plus icon to add a new calculated column.

Again, you can use the functions, columns, and operators on the right sidebar to create your calculation.

Finally, you can see a sample of your data by clicking on the Preview Data button. This button can be found in multiple places. For example, you can click on it next to your output node. You can also click on it on the top menu, next to the save and deploy icons.

But you can also preview just one table or one join. To do that, just click on the Preview Data icon next to the table or node you want to preview.

You will see the preview on the center lower part of the screen. Notice that you can confirm which preview you are looking at by checking the title of the preview Data Preview: Output.

Business Semantics

You can check out our detailed articles on Business Semantics and how to add business semantics attributes to your data models in the following articles:

Ready to deploy

When you are done with your data modeling and transformation, you still need to make sure to either save or save and deploy your data model. You should choose Save if you are not yet done with your modeling and still will make changes to it before using the data model in a data story.

But you can directly click on the Deploy icon to deploy your data model, which also saves it on your Space’s Repository. If you haven’t yet added a business and technical name to this data model, you will be asked to do so when clicking either the Save or Deploy icons. Your data model will be available for use in Data Stories if you determine it is a Fact and after you deploy it successfully.