Creation of a Datamart consists of the following general steps:
- Create a Data Source.
- It must have a primary key.
- Once deployed, the system automatically creates a Data Feed which loads data into the Data Source.
- Create a Datamart.
- Select the primary Data Source.
- Pick columns from that Data Source and place them into the Datamart (by drag&drop).
- If you want to join other, “secondary” Data Sources, the rule is: the name of the primary key column of those other Data Sources must already exist in your Datamart. (Also, the Datamart column which is used for such join must be set as a dimension.)
- Populate all the Data Sources with data.
- The standard way is to have data in Data Feed and using the Data Load event 'Flush' upload the data from Data Feed to Data Source.
- Or if you have data in Excel, you can upload it directly into Data Source (or alternatively into Data Feed and then use the Data Load to copy from Data Feed into Data Source).
- Run the Data Load of Datamart (this is called “refresh”). It will load all rows from the primary Data Source and then also the columns from other Data Sources.
Datamart is a physical table but it does not store a copy of all attribute values. Only keys and calculated columns are stored, the rest of the data is joined dynamically when a query is run.
Datamart (DM) is a data table, with data coming generally from one or more Data Sources (it is possible to make a Datamart without a Data Source). This data set can then be enriched:
- with data from other joined data sources (through related matching columns); Datamart is then a combination of joined data sources
- or with data from 99584273 on the fly (not stored in DB)
- or with columns, whose values are calculated by a logic (via a Calculation Data Load; these are stored in DB).