Forum Discussion

dirkkoch's avatar
dirkkoch
Helper III
3 years ago
Solved

Dynamically subtract two values in same column between two dates

Hi,   I have two columns (one with values, one with dates, see Screenshot marked red). I want to create a calculated column which subtracts each value with the one from the former date (see desired...
Need Help
Show and Tell
Tips and Tricks
  • MahyarTF's avatar
    MahyarTF
    3 years ago

    Hi,

    I did as below :

    1- Add the Index column to the table in Power Query.

    2- Then use below code for create the new column :

    DiffVal Col =
    Var __CurrIndex = Sheet57[Index]
    Var __CurVal = (Sheet57[Value])
    Var __BeforeVal = CALCULATE(sum(Sheet57[Value]), filter(sheet57, Sheet57[Index] = __CurrIndex - 1 ))
    return __CurVal - __BeforeVal

    Appreciate your Kudos and please mark it as solution if it helps you

MahyarTF's avatar
MahyarTF
Memorable Member
3 years ago

Hi,

I did as below :

1- Add the Index column to the table in Power Query.

2- Then use below code for create the new column :

DiffVal Col =
Var __CurrIndex = Sheet57[Index]
Var __CurVal = (Sheet57[Value])
Var __BeforeVal = CALCULATE(sum(Sheet57[Value]), filter(sheet57, Sheet57[Index] = __CurrIndex - 1 ))
return __CurVal - __BeforeVal

Appreciate your Kudos and please mark it as solution if it helps you

  • daXtreme's avatar
    daXtreme
    Solution Sage
    3 years ago

    MahyarTF 

     

    THe problem with your solution is that it makes the model unnecessarily bloated (as the index column is not really needed). Second, calculated columns are not compressed the way they are when created in Power Query - this results in even more bloat. In big models this may be unacceptable. Therefore, please use Power Query to do such calculations. Unless, of course, you don't care about how big the model is and how fast DAX could be. But I think you should since you pay money for storage. You also want to have fast DAX and columns optimized for access and calculations.

    • MahyarTF's avatar
      MahyarTF
      Memorable Member
      3 years ago

      Hi daXtreme ,

      Thanks for your useful advice.

      Just learning and trying to improve my knowledge.

      This is the solution it was at top of my head but absolutely there are better solutions.

  • dirkkoch's avatar
    dirkkoch
    Helper III
    3 years ago

    MahyarTF 
    Thanks! Yes, that would work if there is a value in each row of the column. Unfortunately there might be some blanks, meaning that if there is no value, it should substract the actual with the last availabe value in the column 

     

    • MahyarTF's avatar
      MahyarTF
      Memorable Member
      3 years ago

      Hi dirkkoch 

      Please try the below code in the new column I described earlier :

      DiffVal Col =
      Var __CurrIndex = Sheet57[Index]
      Var __PreviousIndex = CALCULATE(max(Sheet57[Index]), all(Sheet57), Sheet57[Index] < __CurrIndex, not(ISBLANK(Sheet57[Value])))
      Var __BeforePreviousIndex = CALCULATE(max(Sheet57[Index]), all(Sheet57), Sheet57[Index] < __PreviousIndex, not(ISBLANK(Sheet57[Value])))
      Var __PreviousVal = CALCULATE(sum(Sheet57[Value]), filter(sheet57, Sheet57[Index] = __PreviousIndex))
      Var __BeforePreviousVal = CALCULATE(sum(Sheet57[Value]), filter(sheet57, Sheet57[Index] = __BeforePreviousIndex))
      Var __NextVal = CALCULATE(sum(Sheet57[Value]), filter(sheet57, Sheet57[Index] = __CurrIndex+1))
      Var __CurVal = if(ISBLANK(Sheet57[Value]), __NextVal,Sheet57[Value])

      return if(ISBLANK(Sheet57[Value]), __PreviousVal - __BeforePreviousVal, __CurVal - __PreviousVal)

      Appreciate your Kudos and please mark it as a solution if it helps you

      • dirkkoch's avatar
        dirkkoch
        Helper III
        3 years ago

        MahyarTF  Thanks for your support! I managed to do it myself now by creating an adapted Index (with Rankx-Function) leaving out the blank spots. That worked for me!