The ultimate Fabric, Power BI, SQL, and AI community-led learning event. Save €200 with code FABCOMM.
Get registeredEnhance your career with this limited time 50% discount on Fabric and Power BI exams. Ends August 31st. Request your voucher.
Good Morning all,
Ok my query is this, I have a General Ledger Table, where i need to Calculate the Monthy Revenue, for Each 'Territory' based on a specific Reporting Segment, to Gauge the % of Total Revenue for that Month. ( For that Reporting Segment and Supsequent Territory)
Using that % to calculate, from a Total COGS for the month, the proportional amount of that COGS total, using the % that has been calculated.
This is to then be part of the FACT TAble, not a DAX Calc..
DAX Calc's used are:
Solved! Go to Solution.
Hi @Anonymous ,
Power Query can do all of this if you push it hard enough, but the query would be disproportionally complicated and would take a week to refresh.
Power Query/M is just not designed for what you are wanting it to do here, DAX is.
Your original path appears to be the correct one from the limited info I have. The only comment I would make on what you have done so far would be to say that DAX calculated columns should only be used as a very last resort. They have to be materialised in-memory at runtime so can cause significant performance degradation.
The calculated columns you described should be recreated as measures.
Pete
Proud to be a Datanaut!
Hi @Anonymous ,
Power Query can do all of this if you push it hard enough, but the query would be disproportionally complicated and would take a week to refresh.
Power Query/M is just not designed for what you are wanting it to do here, DAX is.
Your original path appears to be the correct one from the limited info I have. The only comment I would make on what you have done so far would be to say that DAX calculated columns should only be used as a very last resort. They have to be materialised in-memory at runtime so can cause significant performance degradation.
The calculated columns you described should be recreated as measures.
Pete
Proud to be a Datanaut!
Hi @BA_Pete , many thanks for that, your anser was pretty much what i was expecting. Totally agree in creating measures rather than the calc columns, my viewpoint was to create a table of the proportioned data to show an end result. I think in the end calculated measures may be my answer.
Regards Mike.