rank based on grouping of multiple columns

[chayanupadhyay]

Hi All,

I have below data in my table and i need to rank the sales on the basis of country, area.brand and product, PFB the sample data:

Country	area	product	Brands	sales	rank
UK	area1	xyz	a1	500.0	7/7
UK	area1	xyz	a2	200.0	2/7
UK	area1	xyz	a3	399.0	6/7
UK	area1	xyz	a4	233.0	3/7
UK	area1	xyz	a5	134.0	1/7
UK	area1	xyz	a6	284.0	4/7
UK	area1	xyz	a7	374.0	5/7
US	area1	xyz	a8	279.0	2/7
US	area1	xyz	a9	376.0	5/7
US	area1	xyz	a10	373.0	4/7
US	area1	xyz	a11	253.0	1/7
US	area1	xyz	a12	379.0	6/7
US	area1	xyz	a13	743.0	7/7
US	area1	xyz	a14	333.0	3/7

i have seven different brands which is selling same product in same area in same country, i need to get the above mentioned rank, for an example: Brand a1 to a7 is selling their product in UK in same area and its sales is captured now i need to find the rank of the brand as per their sales amount.

I have multiple countries data in same table with same format as shown  in above sample.

I really appreciate your help on this, thank you in advance.

Please let me know if you need more details on above scenario.

[v-eachen-msft]

Hi @chayanupadhyay ,

You need to add a new calculated column.

Column =
VAR A =
    RANKX (
        CALCULATETABLE (
            test,
            ALLEXCEPT ( test, test[Country], test[area], test[product] )
        ),
        test[sales],
        ,
        ASC,
        DENSE
    )
VAR B =
    COUNTROWS (
        CALCULATETABLE (
            test,
            ALLEXCEPT ( test, test[Country], test[area], test[product] )
        )
    )
RETURN
    A & "/" & B

Here is the result.

Best Regards,

Eads

If this post helps, then please consider Accept it as the solution to help the other members find it more quickly.

[chayanupadhyay]

Thank you @v-eachen-msft  for your help. It worked .

I have to rank Quarter and year as well, PFB the sample data:

country	Product	Sales	Period	Rank
x	a	200	Q3 2019	1
x	b	300	Q4 2018	2
y	b	200	Q4 2018	2
z	b	200	Q4 2018	2
z	a	239	Q3 2018	3
f	a	134	Q3 2018	3
g	b	143	Q1 2018	4
h	c	234	Q1 2018	4

If you could help me on the DAX statement to get the ranking of Period column, since Q3 2019 is the latest which will rank as 1 and then following quarter with the year will be ranked, if the quarter and year are same then same rank needs to be applied.

Period column is text column, Period is coming from source as column and i am unpivoting it and making it as Rows but unable to change its data type to Date and sort it according to that.

Thank you in advance for your help.

Regards,

Chayan Upadhyay

[v-eachen-msft · Accepted solution]

Hi @chayanupadhyay ,

You can try this calculated column.

Column =
RANKX (
    'Table',
    RIGHT ( 'Table'[Period], 4 ) & LEFT ( 'Table'[Period], 2 ),
    ,
    DESC,
    DENSE
)

Best Regards,

Eads

If this post helps, then please consider Accept it as the solution to help the other members find it more quickly.

[chayanupadhyay]

@v-eachen-msftThank you so much for your help and support.

Apologies for replying late.

It helped me alot 🙂

[amitchandak]

Check, if this can help

https://community.powerbi.com/t5/Desktop/Summarize-and-Rank-by-multiple-columns-and-rows/td-p/330108

[chayanupadhyay]

Thank you @amitchandak  for sharing the link.

In that use case, they have ranked total sales amount for each combination, in my case i need to rank each brand on their individual sales amount comparing to the other brand sales amount for same country, same area and same product.

[amitchandak]

Please check these formulae. These have been created as measures, not columns. If one create them as column they will become aggregatable. There is diff, how you calculate these two

RANK Test = VAR x = SUM(Sales[Sales])
RETURN 
	RANKX(
			ALL(Customer[Age Group]),
			CALCULATE(SUM(Sales[Sales])
			
		))
		
RANK Test 1 = VAR x = SUM(Sales[Sales])
RETURN 
	RANKX(
			ALL(Customer[Income Group]),
			CALCULATE(SUM(Sales[Sales])
			
		))
RANK Test 2 = VAR x = SUM(Sales[Sales])
RETURN 
	RANKX(
			ALL(Customer[Age Group],Customer[Income Group]),
			CALCULATE(SUM(Sales[Sales])
			
		))