Alternate approach to SUMMARIZE

[JBaker0585]

Stuck with a slow performing report, and I have a number of measures that are based on the below code only modifying the measure used in the CALCULATE call to get a different result.

I am putting this out to the group to see if I can find a better solution and move away from the use of SUMMARIZE which based on everything I have read is a performance nightmare.

__% Priority Orders Rank =

VAR MedianItems =

MEDIANX (

    SUMMARIZE (

        ALLSELECTED ( Sales ), Sales[bill_to_id], Sales[billto_customer_name] ),

            CALCULATE (

                [__% Priority Orders]

            )

    )

RETURN

IF ( [__% Priority Orders] > 0 && [__% Priority Orders] <= .5 * MedianItems, "A",

    IF ( [__% Priority Orders] > .5 * MedianItems && [__% Priority Orders] <= MedianItems, "B",

        IF ( [__% Priority Orders] > MedianItems && [__% Priority Orders] <= 1.5 * MedianItems, "C", "D"

        )

    )

)

[sturlaws]

Hi @JBaker0585 ,

If I should guess, it's not the SUMMARIZE-function which slows down your query, but the MEDIANX-function probably does. MEDIANX is an iterator function.

What is the purpose of the ALLSELECTED-function?

If you post a sample of your report, I can try to tune it.

Cheers,
Sturla

[JBaker0585]

Thanks, @sturlaws. The ALLSELECTED was used in another piece of DAX that I modified for my particular solution. I have not attempted the code without it.

For my sample data, what exactly would you need to help you set up a test?