Now let us use SAP Datasphere. If you want to create a model, you first need to consider:
- What information is needed to build the model?
- Where (in which sources) is data currently available?
- Is virtual (remote) access fast enough to avoid data duplication, or is it necessary to replicate data to the SAP HANA database of the SAP Datasphere?
- What selections, joins, unions, projections, and aggregations are needed?
- How do I visualize the model?
- How should the business user interact with the data?
Then, you first need to log on to SAP Datasphere. Let's have a look at the main page and the navigation elements.
Now, you can create different objects as you like. In a very simple case, you have a view. But suppose you want to persist data. In this case, you could start by creating a new table. Tables created in SAP Datasphere can be populated with data from a CSV file or via a data flow. You can also import tables from a connection. The most important part when creating a new table is to define how your table should be used for data modeling purposes, which is called Semantic Usage. If you are looking for measures and attributes next to each other which is basically the principle type of object used by reporting tools, you have to select Fact. If you want to provide textual context in one or more languages, you have to select Text. Apart from that, SAP Datasphere offers further techniques such as the linkage of its data with another source object or even the option to store the table data directly in memory to improve performance.
Launch the video to learn how a table is created:
If you want to integrate different sources, you need to define connections and remote tables. In the remote table editor, you can start or schedule uploading an snapshot. You can check the status at the Data Integration Monitor
You can then create a view on top of the remote tables.
Launch the video to learn how to create a data model: