If your source data for a graphical calculation view is of a star schema type, you can create a calculation view of type CUBE with Star Join.
A Star Join node enables you to join the fact data with the descriptive data. The input allowed to the Star Joinnode includes the lower node and all the relevant calculation views of type DIMENSION.
This way, you are logically creating a star schema where the join is created from the central entity to the other entities.
During deployment, the Star Joinnode is always deployed with an Aggregation node on top of it. The Star Joinnode is deployed first with a series of joins, and then the aggregation logic is deployed. This allows the view to aggregate the measures dynamically, based on the attribute columns you include in the result set.
The Star Joinnode in calculation views supports the Referential Join type, which can speed up the execution of the calculation view when you do not query any column from one or several of the dimension calculation views assigned to the Star Joinnode, because the join will not be executed (at least for the cardinalities 1..1 and n..1).
As always, using a Referential Join requires that the referential integrity of the joined tables is ensured.
As already said, in addition to the Facts Data Source ("Fact Table"), the only other possible sources for a Star Joinnode are calculation views of type DIMENSION. But DIMENSION views can also be used in classical CUBE calculation views (without star join) if relevant.