Forum Discussion

2366's avatar
2366
Icon for Helper I rankHelper I
4 years ago
Solved

RANKX across multiple tables

Hi,

 

I have the below tables in my data model - With the current RANKX expression I am able to Rank product sales across all cities but my desired output is to calculate sales of my products across different countries which is my dimension table and Rank sales based on product and across countries. Since I cannot use ALL(ProductID,Country) I am not sure how to get around this. Please help!

 

Thanks 

 

1) Fact Sales 

ProductIDCitySales
1London100
2Paris300
1Paris300
3London250
2Milan350
1Milan150

 

2) Dim Country

CityCountry
LondonUK
ParisFrance
MilanItaly

 

Existing Measure - Calculate(RANKX(All('Fact Sales'[ProductID]), SUM(Sales),,DESC,),Removefilter('Dim Country'[Country]))

 

ProductIDCountrySalesRank
1UK1002
1France3002
1Italy1502
2France3001
2Italy3501
3UK2503

 

Expected Output

 

ProductIDCountrySalesRank
1UK1005
1France3002
1Italy1504
2France3002
2Italy3501
3UK2503

 

DAX
RANKX

2 Replies

  • Jos_Woolley's avatar
    Jos_Woolley
    Icon for Solution Sage rankSolution Sage
    4 years ago

    Hi,

    In this case, auto-exist will not apply and so you need to be careful to avoid ending up with a visual which contains additional entries due to the cross-join between the two tables. So I would try something like:

     

    Measure =
VAR T1 =
    ADDCOLUMNS(
        ADDCOLUMNS(
            'Fact Sales',
            "Country", RELATED( 'Dim Country'[Country] )
        ),
        "MyRank",
            RANKX(
                ALL( 'Fact Sales' ),
                'Fact Sales'[Sales],
                ,
                ,
                DENSE
            )
    )
RETURN
    MINX(
        T1,
        [MyRank]
    )

     

    Regards