Forum Discussion

Chris2016's avatar
Chris2016
Resolver I
3 years ago
Solved

Optimizing performance when using time comparison

Hello,   I have a large dataset where some of the data looks like this: Name Test Activity DateStarted DateCompleted PercComplete Key _TestCompleteDate _CompletedIndividual _TestCo...
  • tamerj1's avatar
    tamerj1
    3 years ago

    Chris2016 

    Please try

    TestCompleteIndividuals =
    COUNTROWS (
    DISTINCT (
    SELECTCOLUMNS (
    FILTER ( 'Table', 'Table'[_TestCompleteUpToLastMonth] = 1 ),
    "@Name", 'Table'[Name]
    )
    )
    )

Chris2016's avatar
Chris2016
Resolver I
3 years ago

Thanks a lot for your help, Tamer, this is great. The visuals no longer have issues loading.
Any idea why this measures has performance issues in the given context as opposed to your approach?

TestCompleteIndividuals = CALCULATE(DISTINCTCOUNT('Table'[Name]),FILTER('Table','Table'[_TestCompleteUpToLastMonth]=1))

 Thanks, and best regards!

    • Chris2016's avatar
      Chris2016
      Resolver I
      3 years ago

      "... counting unique values in complex reports can still create performance issues. The main reason for this is that DISTINCTCOUNT is a non-additive aggregation that must be computed for every cell in the report. Understanding the behavior of this aggregation can explain why other equivalent expressions that are slower in theory, can provide better performance in specific reports."
      This is great, thanks, tamerj1 !