Forum Discussion

Sabarikumar7579's avatar
4 years ago
Solved

Summarize with multiple inactive relations and cross filtering visuals not working.

I am trying to summarize total items using inactive relations, which i am able to do separately using "userelationship" function. I have 3 inactive relations to different dimension tables. Everything looks great when I visualize them separately. The challenge is when cross-filtering the visuals, The filter doesn't flow to other Tables as they don't have active relations.

I tried using the below function which doesn't work. Even tried replacing Userelationship with "CrossFilter" which dint help.

 

 CALCULATE (
                [Total items],
                USERELATIONSHIP ( Date[Date], DataTable[CreatedDate] ),
                USERELATIONSHIP ( dimAging[OpenAge_Bucket], DataTable[Open_Age_Bucket]),
		USERELATIONSHIP (dimType[Open_type],DataTable[Open_type])
              )

 

 This is how my relationship looks:
Relationship

Below is my challenge on getting cross-filtering working

 

 

  • Sabarikumar7579 , The formula seems correct , Check createddate so not have any timestamp (Change datatype to datetime and use a format with date and check )

     

    Try like

     

     

    CALCULATE (CALCULATE (CALCULATE (
    [Total items],
    USERELATIONSHIP ( Date[Date], DataTable[CreatedDate] )),
    USERELATIONSHIP ( dimAging[OpenAge_Bucket], DataTable[Open_Age_Bucket])),
    USERELATIONSHIP (dimType[Open_type],DataTable[Open_type])
    )

     

     

    If this does not help
    Can you share sample data and sample output in table format? Or a sample pbix after removing sensitive data.

2 Replies

  • Sabarikumar7579 , The formula seems correct , Check createddate so not have any timestamp (Change datatype to datetime and use a format with date and check )

     

    Try like

     

     

    CALCULATE (CALCULATE (CALCULATE (
    [Total items],
    USERELATIONSHIP ( Date[Date], DataTable[CreatedDate] )),
    USERELATIONSHIP ( dimAging[OpenAge_Bucket], DataTable[Open_Age_Bucket])),
    USERELATIONSHIP (dimType[Open_type],DataTable[Open_type])
    )

     

     

    If this does not help
    Can you share sample data and sample output in table format? Or a sample pbix after removing sensitive data.