Forum Discussion

ahood89's avatar
ahood89
Regular Visitor
4 years ago

Un-accumulate a set values

I have a set of finance data that I have already had to transform and combine in Power Query, including applying exchange rates. 

 

The values are accumulated with each Period. So P1 = P1, P2 = P1+P2. I now want to un-do that and just have P2 so P2-P1. I have looked at various online help articles but I am getting nowhere! I think I am confusing Measues and Columns as well which isn't helping and I don't understand the limitations of either/how they work. 

 

I can do it in Excel so easily so I feel like this should be simple in PBI.

 

Excel:

In PBI:

Column = IF('TAGETIK EXTRACTS'[Period] = 2,CALCULATE(SUMX('TAGETIK EXTRACTS','TAGETIK EXTRACTS'[Transaction Amount €]),('TAGETIK EXTRACTS'[Period]=2))-CALCULATE(SUMX('TAGETIK EXTRACTS','TAGETIK EXTRACTS'[Transaction Amount €]),('TAGETIK EXTRACTS'[Period]=1)),0)
 
If that worked I was then going to expand it e.g.
Column = IF('TAGETIK EXTRACTS'[Period] = 2,CALCULATE(SUMX('TAGETIK EXTRACTS','TAGETIK EXTRACTS'[Transaction Amount €]),('TAGETIK EXTRACTS'[Period]=2))-CALCULATE(SUMX('TAGETIK EXTRACTS','TAGETIK EXTRACTS'[Transaction Amount €]),('TAGETIK EXTRACTS'[Period]=1))
IF('TAGETIK EXTRACTS'[Period] = 3,CALCULATE(SUMX('TAGETIK EXTRACTS','TAGETIK EXTRACTS'[Transaction Amount €]),('TAGETIK EXTRACTS'[Period]=3))-CALCULATE(SUMX('TAGETIK EXTRACTS','TAGETIK EXTRACTS'[Transaction Amount €]),('TAGETIK EXTRACTS'[Period]=2)),0)
 
Any ideas please?
 
Thanks,

5 Replies

  • AntonioM's avatar
    AntonioM
    Solution Sage

    You could add a custom column

     

    Periodic Revenue = 
    'Table'[Measure] - MAXX(FILTER('Table', 'Table'[Period] < EARLIER('Table'[Period])), 'Table'[Measure])

     

    Where Table is your table name and Period and Measure are the columns from your screenshot

     

    This will work if you only have one year's data (so one set of periods 1 through 12)

     

     

    made up example:

     

  • ahood89's avatar
    ahood89
    Regular Visitor

    Thanks for replying so quickly.

     

    Apologies, I have multiple sets of periods. e.g. 2021 Actual, 2021 Budget, 2022 Actual, 2022 Budget etc. Even though on the current page I have a Filter so there is only 1 of each period. 

     

    Also, sorry I should have said before, the screenshots are not the raw data, they're already 'grouped'. The raw data is more like the below, so multiple entries for 1, 2 etc. 

     

    I did as you suggested above and only P1 is correct, I am assuming that's something to do with this?

     

    Raw data structure example (simplified):

    FilePeriodDepartmentTypeAmount
    Budget1ARevenue1
    Budget1BCost2
    Budget2ARevenue3
    Budget2BCost4
    Budget3ARevenue6
    Budget3BCost7
    Budget4ARevenue10
    Budget4BCost11
    Actual1ARevenue1
    Actual1BCost2
    Actual2ARevenue3
    Actual2BCost4
    Actual3ARevenue6
    Actual3BCost7
    Actual4ARevenue10
    Actual4BCost11

    Desired result (example of 1 sub group):

    FilePeriodDepartmentTypeAmountAmount per Period
    Budget1BCost22
    Budget2BCost42
    Budget3BCost73
    Budget4BCost114

     

    So Amount per Period is the value I am trying to create.

    • AntonioM's avatar
      AntonioM
      Solution Sage

      Sure, this works for the data you posted, is the real data much more complicated?

       

      Amount per Period = 
      var _period = 'Table'[Period]
      
      var _previous = 
      CALCULATE(
          MAX('Table'[Amount]),
          'Table'[Period] < _period,
          ALL('Table'[Amount])
          )
      
      return 'Table'[Amount] - _previous

       

       

      • ahood89's avatar
        ahood89
        Regular Visitor

        Thanks. I can also get this to work for my test data:

        But unfortunately it isn't working for my actual dataset. When adding the new column and Period to a table, it still shows the accumulated values i.e. the same as the SUM of (my equivalent of) Test[Amount]? There are many more columns and rows of data in the real data as opposed to my example, but unfortunately I can't share it. 

         

        My real data contains negative values, but I just did the above test data with negative values and it was OK so I don't think it can be that.