firstsortedvalue with filters

[LG]

I read many other posts, but I can't apply it to my case...

I think "firstsortedvalue" might be the solution.

Here's my Data (I gave my best to get a good example):

#	Reporting Plant	Material Number	Material Description	Direct Savings Year	Invoice Amount CHF	Invoice Quantity	Diff. prev. Year	Calc. Direct Savings CHF	Savings / Cost	Vendor Category
1	Germany	12345	Apple	2019	480’250.86	163’681	-0.4802	-115’571.35	Savings	R
2	Germany	12345	Apple	2019	3’952.29	1’125	-0.4802	-115’571.35	Savings	M
3	Germany	12345	Apple	2018	694’002.40	240’158	-0.4802	-115’571.35	Savings	R
4	Germany	12345	Apple	2018	940.13	503	-0.4802	-115’571.35	Savings	M

What do I want to do:

There are multiple same values for "Calc. Direct Savings CHF", but I only want to see the "Vendor Category", based/ranked on the <highest> "Invoice Amount CHF" per "Direct Savings Year"

Example:

#1, "Vendor Category" <R> would be shown, because the "Invoice Amount CHF" for 2019 is higher than #2

#3, "Vendor Category" <R> would be shown, because the "Invoice Amount CHF" for 2018 is higher than #4

Finally, out of the 4 lines, only 2 shall remain --> #1 & #3

hope this makes sense!

Thanks already 🙂

[JGMDataAnalysis]

To help you solve your problem, I need information about the data model and the chart in which you are using the measure. In principle I think that if you are using the set identifier {1} you do not depend on the selections made by the user, then, if the calculation does not depend on the "intervention" of the user, it could be solved at load time (script). Ideally, you should attach a qvf file with reduced information to evaluate your case.