Forum Discussion

Anonymous's avatar
Anonymous
Not applicable
5 years ago
Solved

Earlier function on the fly

Hi all, I'm trying to create rank calculation on PBI. I'm able to do that with earlier function which only allows create on column. When I created on column It is not serve to purpose completely....
Need Help
Show and Tell
  • MFelix's avatar
    MFelix
    5 years ago

    Hi Anonymous ,

     

    You need to create the following measures, be aware that I'm divinding this in several measure so you can have the calculations checked.

     

    Amount = SUM(PCC_Analysis2[Amount])

Rank_Values = RANKX(ALLSELECTED(PCC_Analysis2[Customer]); CALCULATE([Amount]) + + INT ( CALCULATE ( MIN ( PCC_Analysis2[Customer]) ) ) / 100000)

Cumulative Total based on Rank = 
           CALCULATE([Amount];
                     TOPN([Rank_Values];ALLSELECTED(PCC_Analysis2[Customer]);[Amount] + INT ( CALCULATE ( MIN ( PCC_Analysis2[Customer]) ) ) / 100000 )
           )

% of Products Running Total = 
         DIVIDE([Cumulative Total based on Rank];CALCULATE([Amount];ALLSELECTED(PCC_Analysis2[Customer])))

     

    The idea here is to pick up the TOPN lines that are equal to the ranking.

     

    Adding the + INT ( CALCULATE ( MIN ( PCC_Analysis2[Customer]) ) ) / 100000 allows to make the equal values to be different.

     

    Be aware that I'm making a relationship between slicer and fact table.

     

    Anonymous  Using you PBIX file thanks in advance

     

    Check PBIX file attach.

Anonymous's avatar
Anonymous
Not applicable
5 years ago

Thank you for all AlB , MFelix and amitchandak 

It was absolutely working fine for this scenario but there is another point which I forget to mention on there.

All of you guys used RANKX function but actually I want to find reciprocity to EARLIER function. You are all right, RANKX looks like fit for this scenario but in I'm using EARLIER function with other calculations. Like cumulative sum.

Let me show you my main problem.

I want to find customers % amount for total amount.
Then I want to cumulative sum customer % amount from top to bottom rank.
Then make filter on Cumulative Amount % column. For example show the values less than %80

It is working fine for one month but not working well as I mentioned on my first message for more than one month.


I created Group Amount, Group Rank, Amount % and Cumulative Amount % on column and I used EARLIER function most of this calculations. 
Thats the point which I'm looking for a solution.
By the way If you have another solution to make this cumulative amount % calculation It is fine for me no matter.

MFelix's avatar
MFelix
Super User
5 years ago

Hi Anonymous ,

 

The best option for % in my opinion is to do it with measures, because when you make them has columns on your datatables you then loose the flexibility of having the filters calculation.

 

If you do the % as measures then you can do the rank based on those measures making a temporary table on your measure.

 

Can you please provide the way you are calculating the %s.

 

Please AlB  and amitchandak  if you have any other ideas please share them.

  • Anonymous's avatar
    Anonymous
    Not applicable
    5 years ago

    Hi MFelix ,

    First of all I find the total amount for customers for each month. 

     

    Group Amount = CALCULATE (
    SUM( PCC_Analysis2[Amount] ),
    FILTER (
        'PCC_Analysis2',
        'PCC_Analysis2'[Tarih] = EARLIER ( 'PCC_Analysis2'[Tarih] )))

     


    Then I create calculation rank for customers amount 

     

    Group Rank =
COUNTROWS (
    FILTER (
        'PCC_Analysis2',
        'PCC_Analysis2'[Tarih] = EARLIER ( 'PCC_Analysis2'[Tarih] )
            && PCC_Analysis2[Amount %] > EARLIER ( PCC_Analysis2[Amount %] )
    )
) + 1

     

     
    Then I find % Amount for each customer with this divide amount and total amount

     

    Amount % = PCC_Analysis2[Amount] / PCC_Analysis2[Group Amount]

     

     

    Finally I'm making this cumulative sum calculation.

     

    Cumulative Amount % = 
CALCULATE (
    SUM( PCC_Analysis2[Amount %] ),
    FILTER (
        'PCC_Analysis2',
        'PCC_Analysis2'[Tarih] = EARLIER ( 'PCC_Analysis2'[Tarih] ) && PCC_Analysis2[Group Rank] <= EARLIER(PCC_Analysis2[Group Rank])))

     


    This is the final result

     

     

    • Anonymous's avatar
      Anonymous
      Not applicable
      5 years ago

      Hi Anonymous 

      You may try to build a measure to achieve your goal.

      Firstly, build a slicer table.

      Slicer = VALUES('PCC_Analysis2'[Tarih])

      Measure:

      Amount% = 
VAR _sel =
    ALLSELECTED ( Slicer[Tarih] )
VAR _GroupAmount =
    IF (
        ISFILTERED ( Slicer[Tarih] ),
        CALCULATE (
            SUM ( PCC_Analysis2[Amount] ),
            FILTER ( ALL ( 'PCC_Analysis2' ), 'PCC_Analysis2'[Tarih] IN _sel )
        ),
        CALCULATE (
            SUM ( PCC_Analysis2[Amount] ),
            FILTER (
                ALL ( 'PCC_Analysis2' ),
                'PCC_Analysis2'[Tarih] = MAX ( PCC_Analysis2[Tarih] )
            )
        )
    )
VAR _Amount =
    IF (
        ISFILTERED ( Slicer[Tarih] ),
        CALCULATE (
            SUM ( PCC_Analysis2[Amount] ),
            FILTER ( 'PCC_Analysis2', 'PCC_Analysis2'[Tarih] IN _sel )
        ),
        SUM ( PCC_Analysis2[Amount] )
    )
RETURN
    _Amount / _GroupAmount
      Cumulative Amount % = 
IF (
    [Amount%] = BLANK (),
    BLANK (),
    IF (
        ISFILTERED ( Slicer[Tarih] ),
        SUMX (
            FILTER (
                ALLSELECTED ( PCC_Analysis2 ),
                PCC_Analysis2[Tarih] IN ALLSELECTED ( Slicer[Tarih] )
                    && PCC_Analysis2[Rank] <= MAX ( PCC_Analysis2[Rank] )
            ),
            [Amount%]
        )
    )
)

      Result:

      As Default:

      Select two Tarih in Slicer.

      You can download the pbix file from this link: Earlier function on the fly

       

      Best Regards,

      Rico Zhou

       

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

      • Anonymous's avatar
        Anonymous
        Not applicable
        5 years ago

        Hi Anonymous ,

        Thank you for your helps.

        The Amount measure is working fine which you created but the cumulative amount not working as it's supposed to be.

        You are making group the values by Rank. When we do this, different customers can be found in the same group.

        For ex, in your example customer 0003 and 0005 were grouped. This is different from our logic.

        My expected result should be contain Customer, amount and Cumulative amount. not rank. and shoul be like this.


        I tried to change Cumulative Amount % dax formula. It must not be contain Rank column. Because Rank column is static for each month. We need on the fly rank formula or something different.

        I tried to change

        && PCC_Analysis2[Rank] <= MAX ( PCC_Analysis2[Rank] )

        part of your dax with Amount % but it not accepting measures.

        Do you have any idea about this ?