Forum Discussion

pvarley's avatar
pvarley
Helper I
9 years ago
Solved

Power BI Desktop hangs when adding a column

I'm adding a column to an existing table. The table contains about 20,000 rows, so it should not take long, but after I hit Enter, Power BI Desktop says "Working on it" and never finishes. I have upd...
Need Help
  • pvarley's avatar
    pvarley
    9 years ago
    Anonymous wrote:

    HI pvarley,

     

    I think the performanc issue is related to your masure, each measure will calculate throught the UsersTenants table. If this is a case, your calculate column will calculate about almost 1,368,000,000(19,000* 36,000*2) times.

     

    Regards,

    Xiaoxin Sheng

     

    That may be the problem. I worked around it by moving the string comparison into the query and generating two new columns in the UsersTenants table with True/False values. Now the measures just look at those two values, and the calculation is much faster. The new measures look like this:

     

    CALCULATE( DISTINCTCOUNT( UsersTenants[Email] ),
    FILTER( UsersTenants,
    UsersTenants[IsProvisioningUser] = FALSE && UsersTenants[IsInternalUser] = TRUE
    )
    )

pvarley's avatar
pvarley
Helper I
9 years ago

I left it alone for about an hour, and it had finished when I returned. I am creating a column because I want to use it as a filter in reports. I don't believe I can do that with a measure because a measure would not have a row context. 

Anonymous's avatar
Anonymous
Not applicable
9 years ago

Hi pvarley,

 

Can you share some detail content of your issue? It will be help for troubleshooting.(e.g. screenshots, formula, operate steps)

 

In addition, I'd like to suggest you use measure instead the calculate column and test again.

 

Regards,

Xiaoxin Sheng

  • pvarley's avatar
    pvarley
    Helper I
    9 years ago

    I created a measure with the same formula as the column, but Power BI will not allow me to use the measure as a filter in a report. That is why I'm using a calculated column.

     

    The measures behind the column look like this:

     

    Measure = CALCULATE( DISTINCTCOUNT( UsersTenants[Email] ),
    FILTER( UsersTenants,
    UsersTenants[Email] <> "[email protected]" && FIND("@example.com", UsersTenants[Email], ,0) > 0
    )
    )

     

    The UserTenants table has about 36,000 rows, and the Orgs table, which is related and has the column has about 19.000 rows, so it's not a lot of data. The column takes at least 20 minutes to calculate, which seems like an eternity for 19,000 rows. 

    • Anonymous's avatar
      Anonymous
      Not applicable
      9 years ago

      HI pvarley,

       

      I think the performanc issue is related to your masure, each measure will calculate throught the UsersTenants table. If this is a case, your calculate column will calculate about almost 1,368,000,000(19,000* 36,000*2) times.

       

      Regards,

      Xiaoxin Sheng

      • pvarley's avatar
        pvarley
        Helper I
        9 years ago
        Anonymous wrote:

        HI pvarley,

         

        I think the performanc issue is related to your masure, each measure will calculate throught the UsersTenants table. If this is a case, your calculate column will calculate about almost 1,368,000,000(19,000* 36,000*2) times.

         

        Regards,

        Xiaoxin Sheng

         

        That may be the problem. I worked around it by moving the string comparison into the query and generating two new columns in the UsersTenants table with True/False values. Now the measures just look at those two values, and the calculation is much faster. The new measures look like this:

         

        CALCULATE( DISTINCTCOUNT( UsersTenants[Email] ),
        FILTER( UsersTenants,
        UsersTenants[IsProvisioningUser] = FALSE && UsersTenants[IsInternalUser] = TRUE
        )
        )