Forum Discussion

Linger's avatar
Linger
Frequent Visitor
6 years ago
Solved

Crossfilter customer table to improve performance

Hi there ðŸ˜€   I am currently working with a data model that has about 40 million rows in the fact table and about 300k rows in the customer table. The measure that is keeping things interesting is...
  • Linger's avatar
    Linger
    Frequent Visitor
    6 years ago

    amitchandak Thank you very much for referring to your blog posts. These lead me to a solution I am happy with ðŸ˜Š

    I tried your version with SUMX which worked fine and gave me a new way to approach my measure.

    It is slightly slower than the measure I posted previously but as this one returns incorrect results on the month level I can't use it anyway.

    What also worked was slimply replacing CustomerTable by VALUES(CustomerTable[customerid]). Not sure why I didn't try that in the first place. This way the measure is as fast as the SUMX version.

     

    CNENFRNL Thank you very much for taking the time to explain your approach. With some tweaking I think this could also work fine for No of Active Customers. However I also need other versions of this measure considering Sales 12M > X so I prefer to reference the measure.