Forum Discussion
Rankx ignoring 1 column
- 1 year ago
You're right Arnault_ , I just updated the measure:
Product ranking (volumes moy. stockés) = VAR fx = [Volumes moy. stockés] RETURN IF ( fx = 0, BLANK (), RANKX ( ALL ( XN_ART_DEPOT[ART/VER], XN_ART_DEPOT[Dubai_stock] ), [Volumes moy. stockés], , DESC ) ) - 1 year ago
Hi All,
I finally figured this out. Here is the correct formula:
Product ranking (volumes moy. stockés) v 3 = VAR fx = [Volumes moy. stockés] RETURN IF ( fx = 0, BLANK (), RANKX ( ALLSELECTED ( XN_ART_DEPOT[ART/VER], XN_ART_DEPOT[Dubai_stock] ), [Volumes moy. stockés],, DESC ) )Thank you for your help.
Hi Arnault_ ,
The issue you're encountering occurs because the ALL function is removing all filters from the XN_ART_DEPOT[ART/VER] column, but when you add a new column (Dubai_stock), it starts affecting the results. This happens because the rank calculation interacts with all columns in the table.
Please try this measure:
Product ranking (volumes moy. stockés) =
VAR fx = [Volumes moy. stockés]
RETURN
IF (
fx = 0,
BLANK(),
CALCULATE(
RANKX(
ALL(XN_ART_DEPOT[ART/VER]), -- Rank by ART/VER only
[Volumes moy. stockés],,
DESC
),
ALLEXCEPT(XN_ART_DEPOT, XN_ART_DEPOT[Dubai_stock]) -- Keep all other filters except Dubai_stock
)
)
Let me know if you're facing any issues.
- Arnault_1 year agoResolver III
Hi Bibiano_Geraldo ,
Thanks you very much for you help, unfortunately, it does not work. The rank is now "1" for all the products.
- Bibiano_Geraldo1 year agoSuper User
Hi Arnault_ ,
Please try the bellow Updated DAX, if the problem persist, consider to share no sensitive data sample:
Product ranking (volumes moy. stockés) = VAR fx = [Volumes moy. stockés] RETURN IF ( fx = 0, BLANK(), RANKX ( FILTER ( ALL ( XN_ART_DEPOT[ART/VER] ), -- Ignore filter on ART/VER NOT ( ISBLANK ( [Volumes moy. stockés] ) ) -- Ensure non-blank values ), [Volumes moy. stockés],, DESC ) )- Arnault_1 year agoResolver III
Hi Bibiano_Geraldo ,
In the revised formula, you don't remove XN_ART_DEPOT[Dubai_stock], then it goes back to my initial result. I will try to figure this out. I am not sure the sample data would be usable because my measure calls a formula from another table. But anyway, I think you for your help.