Hi,
I would like to get a better understanding of the star-join calculation view modelling. Each star-join view consists of fact tables and dimensions. Could either the fact tables or the dimensions be considered as “leading” table for implementing filter conditions?
Example:
- Star-Join view consist of one dimension and one fact table (simplified view)
- Dimension retrieves Leads and implements filter like IF X = Y… to recognize table rows as Lead objects (table contains multiple object types)
- Fact table returns some measures (open, closed, won, … leads) and implements the same filter IF X = Y… again to recognize table rows as leads in the same table
- Dimension and fact are joined by the lead UUID
Could I remove the filter either from the dimension or from the fact tables? Does the dimension table restrict the amount of data while evaluating the filter condition or is the fact table the leading object for implementing the filter condition and the fact table “triggers” the retrieving of the key figures in each dimension view?
Best regards,
Ingo