Forum Discussion
Anonymous
5 years agoNot applicable
Sum based on column (different grannularities)
Doc number Date Customer Sales TOTAL SALES per customer 1 01/01/2021 Microsoft 100 600 2 02/01/2021 Apple 50 50 3 02/01/2021 Microsoft 200 600 4 03/01/2021 Microsoft ...
Anonymous
5 years agoNot applicable
Hi,
thanks for your reply.
I tried both but they did not work. Both return me the same values as in the sample data column "sales"
This is what I tried: (note: [(€) Turnover CP] is a measure that performs a formula for sales already (it calculates with discounts and quantities and several prices))
First option: calculate([(€) Turnover CP]; allexcept(Customer;Customer[Customer]))
Second option:
calculate([(€) Turnover CP]; filter(allselected(Customer); Customer[Customer] = max(Customer[Customer])))
amitchandak
Super User
5 years agoAnonymous , Please find the attached file after the signature. I created same formula
- Anonymous5 years agoNot applicable
Hi,
It works indeed as desired, but I have a memory issue if I perform it on my full dataset :(.
Would you have another idea?