Forum Discussion

Anonymous's avatar
Anonymous
Not applicable
4 years ago
Solved

Dimension Driving Dimension Selection through Fact

A single fact table attaches to two conformed dimensions. Dimension_1 is actively attached. Dimension_2 is inactively attached. For this fact table only, the goal is for Dimension_2 to filter the list of Dimension_1 as associated to the Fact table, but then to disappear from the calculation, allowing only related Dimension_1 records to affect the results. Another way to look at the goal is that Dimension_2 chooses the Dimension_1 values, but then all Fact records associated with Dimension_1 are calculated, ignoring the Dimension_2 choice. Note that other dimensions might be affecting the results, but these describe the interactions of these two dimensions.

 

For example, I have tried the following DAX that does not accomplish the goal. What does get returned is filtered by both Dim1 and Dim2.

 

VAR __dim1_list =
CALCULATETABLE(
        VALUES('Fact'[Dim1Key]),
        USERELATIONSHIP('Fact'[Dim2Key],'Dim2'[Dim2Key])
)
VAR __result =
    CALCULATE(
        COUNTROWS('Fact'),
        __dim1_list
    )
 
I am obviously not understanding something about the processing of the contexts, but not sure what.
  • Anonymous's avatar
    Anonymous
    4 years ago

    It was solved with a bridge table between Dim2 and the Fact table. The bridge contains the appropriate Dim1 and Dim2 pairings, with a relationship {Dim2.Dim2Key, Bridge.Dim2Key} one-to-many join and a {Bridge.Dim1Key, Fact.Dim1Key} many-to-many join. Both joins are uni-directional toward the fact table.

     

    Thanks for your suggestions amitchandak .

     

2 Replies

  • amitchandak's avatar
    amitchandak
    Super User
    4 years ago

    Anonymous , Try one of the two ways

     

    CALCULATE(
    COUNTROWS('Fact'),
    USERELATIONSHIP('Fact'[Dim2Key],'Dim2'[Dim2Key])
    )

    or


    CALCULATE(
    COUNTROWS(summarize('Fact', 'Fact'[Dim1Key])),
    USERELATIONSHIP('Fact'[Dim2Key],'Dim2'[Dim2Key])
    )

  • Anonymous's avatar
    Anonymous
    Not applicable
    4 years ago

    It was solved with a bridge table between Dim2 and the Fact table. The bridge contains the appropriate Dim1 and Dim2 pairings, with a relationship {Dim2.Dim2Key, Bridge.Dim2Key} one-to-many join and a {Bridge.Dim1Key, Fact.Dim1Key} many-to-many join. Both joins are uni-directional toward the fact table.

     

    Thanks for your suggestions amitchandak .