Desktop-Specialist Exam Dumps : Tableau Desktop Specialist Exam

When using a data source that has relationships, you can add a join between two or more tables by dragging a table directly on top of an existing logical table from the Tables pane. This will create a logical table that contains the joined tables. You can then edit the join type, join clauses, and join calculation as needed2 The other options are not valid ways to add a join between tables when using relationships. From the Tables pane, double-clicking a table will add it to the data source as a separate logical table, not joined with any other table. From the Connections pane, selecting Add will allow you to add another connection to the data source, not join tables within the same connection. Double-clicking a logical table in the canvas will open the Data Source page, where you can view and edit the fields in the logical table, not join it with another table2

In fact, the opposite of this is true.

Trick : Whenever you think of joins -> Think that after the join is created, we get 1 single flat combined (joined) table. This flat combined table is created prior to us creating our visualizations. This happens at the physical layer.

If you ever think about relationships, know that all tables will remain distinct and separate, and relationships sit at the logical layer. At run time, when you bring in the dimensions and measures to create your viz, Tableau very smartly creates the necessary joins, relates the tables and sends queries to these tables to get the resultant data back in the most meaningful way possible. This allows you to focus on using your data and revealing insights from it and focus less on the data preparation aspect!

Refer to logical layer vs physical layer from the official documentation: 

From the official documentation, the following are the major advantages of using Tableau Data Extracts:

From Tableau 2020.2 onwards, .hyper is the recommended way since it is faster than .tde!