Calculate rank over dimensions from different tables

[alexa_0028]

Hi PBI Experts,

I have a scenario where I have customer and product dimensions connected to my facts table "Sales" in a star schema.
The 2 dimension tables are not related.
I want to rank my table like below by descending measure and applying rank on each row of the table. 

Main goal : Rank measure based on customer and product combination.
My outcome:
Rank = RANKX(SUMMARIZE('Sales','Customer'[Customer],'Product'[Product]),[Sales],,DESC,Dense)

Expectation:

Can someone pls help me here?

I will highly appreciate your help.

[alexa_0028]

Hi @floriangaerner /All,

Meanwhile, I applied some slicers so that I don't get the error for threshold increase.
I have a doubt in the output. My ranking doesn't seem to get update after applying slicer. I sorted the measure DESC in visual but my ranking is not starting from 1,2,3,4 and so on. Its like 0,52 and weird numbers.

Can you pls help me to fix it?

Appreciate your time & help here pls.

[v-kelly-msft]

Hi  @alexa_0028 ,

Could you pls show us your .pbix file?

Remember to remove the confidential information.

Best Regards,
Kelly

Did I answer your question? Mark my reply as a solution!

[alexa_0028]

@floriangaerner 

thank you so much for your help !!
Indeed, it looks like very promising, and something that should work and suits my requirement. 

But when I am running this in my model I am getting error out , maybe due to my large data model and the mode is Direct Query.

So, I tried running it on import mode directly. But still it is so slow to return any result and getting error out too.

Is there any other quicker way to solve this issue? As in excel it is very quick to sort and apply rank, I am not sure why in PBI it is so different.

[floriangaerner]

Hello @alexa_0028 ,

may I propose you a solution:

I've created .pbix file to show the outcome. Based on your description I have following data model:

The tables are filled according to your description.

I assumed that you want to group sales for multiple Customer-Product combinations together.

My proposed measure:

Proposed Rank =

RANKX (
    ALLSELECTED ( Sales ),
    CALCULATE (
        SUM ( Sales[Sale] ),
        ALLEXCEPT ( Sales, 'Customer'[Customer_Key], 'Product'[Product_Key] )
    ),
    ,
    DESC,
    DENSE
)

Result:

Please find attached the .pbix file.

Hope I could help!

Calculate rank over dimensions from different tables.pbix 

[juleshorton]

This code doesn't work at all & returns error: Function 'RANK' only accepts the ORDERBY function for its OrderBy parameter.