Lack of computational power (memory)

[Anonymous]

Hi.

I have tried to make a calcauated collumn for a 200 hour rolling average. 

The code is running fine and return the expected results in my test set of 2,000 rows. 
However, when I implement it on the complete data set of +100,000 rows, i lact memory, (I have 16 gb).

I used the following code:

RollingAvg =

VAR _HourStep = 1 / 24

VAR _CutOffTime = Table1[Date/time] - ( _HourStep * 200 )

RETURN

    CALCULATE (

        AVERAGE ( Table1[Msteam] );

        Table1[Date/time] > _CutOffTime;

            Table1[Date/time] <= _CutOffTime + ( _HourStep * 200 ) ;

        ALL(Table1)

    )

[Anonymous]

Hi @Anonymous ,

Actually, I think 200 hours range are too large to calculate.
When you use calculated column, it will calculate on each row with matched rows(e.g. 10k rows matched), the total calculate amount on your calculated column will increase to 1M * 10k.(it calculated when data model updated)

I'd like to suggest you use measure formula to instead, it calculated when you use it on visuals.

Ten Techniques for Optimising Memory Usage in Microsoft Power BI

Power BI performance best practices

If measure still spend huge amount of memory, you can consider to add more conditions on filters on table to reduce rows which calculated in measure expression.

Regards,

Xiaoxin Sheng