Help optimizing a row by row calculated column that filters on the past 30 minutes

[PJStevens]

Hello All,

I have spent all day today and yesterday trying to solve an issue I have with a calculated column a former colleague made. It is taking extremely long to run over data sets of 300-400k rows, and I haven't been able to solve the issue.

  var Summed = CALCULATE(

                    (SUM('Leveringszekerheid MI'[Counter Onderbreking CV]))/4,

                    FILTER(

                        ALL('Leveringszekerheid MI'),

                        'Leveringszekerheid MI'[DateTimeKey] <= CurrenDateTime && 'Leveringszekerheid MI'[DateTimeKey] > 'Leveringszekerheid MI'[DateTimeKey] - Time(0, 30, 0)

                        )

                    )

return

    IF(Summed = 1,1,0)

Seems to be an issue summing the values over the filtered period (datetimekey)-time(0,30,0) combined with the logic if statement. The column Counter Onderbreking CV is a simply calculated column that is counting the amount of interferences their were recorded (based on certain conditions that aren't really relevant, but they result in 1's or 0's)

Any tips would be greatly appreciated as I cannot seem to solve this

[v-kongfanf-msft · Accepted solution]

Hi @PJStevens ,

Maybe you can try below formula to create calculated column:

Summed Calculation = 
VAR CurrentDateTime = 'Table'[DateTimeKey]
VAR StartTime = CurrentDateTime - TIME(0, 30, 0)
VAR Summed = 
    CALCULATE(
        SUM('Table'[Counter Onderbreking CV]) / 4,
        FILTER(
            'Table',
            'Table'[DateTimeKey] <= CurrentDateTime &&
            'Table'[DateTimeKey] > StartTime
        )
    )
RETURN
    IF(Summed = 1, 1, 0)

Best Regards,
Adamk Kong

If this post helps, then please consider Accept it as the solution to help the other members find it more quickly.

[PJStevens]

Hello,

Thanks for the reply. This is indeed the solution I went with, unfortunately, due to the nature of the calculation, I have decided to only apply it to the past 2 months as to not overload it.

I will mark it as solved, as this is essentially the same calculation I am using, but instead of using it on the entire data set, I only use it on the past 2 months.

Thanks anyways for the reply.

[lbendlin]

Install DAX Studio and then use it to examine the query plan and modify according to your findings.

No need to specify FILTER inside CALCULATE - that will be done for you automatically.

Consider using aggregation functions instead of CALCULATE.