Forum Discussion

AA622's avatar
AA622
Helper II
3 years ago
Solved

Rolling yearly Average

Hello

 

I am trying to calculate a rolling average. Its to calculate the yearly usage of a certain product over the last rolling 12 months and then from there I will calculate the daily average by working days (#260). 

 

I was using this formula but it doesnt match the actual yearly usage based on historical data.

UsageRollingYearly = CALCULATE(
    [Usage],
    DATESINPERIOD(
        'Calendar'[Date],
        MAX('Calendar'[Date]),
        -365,DAY))
 

I have another formula but I also cant seem to get it 100% right across all products

Usage12mo = CALCULATE (
[Usage],
FILTER ( ALL ( 'Calendar' ),
'Calendar'[Date] >= TODAY() - 365))
 
My question is which formula is more indicative of an accurate measure to calculate rolling 12 months usage?
 
Whatever number is calculated from either of these measures will be used to calculate daily average.
  • Anonymous's avatar
    Anonymous
    3 years ago

    Hi AA622 ,

    Please update the formula of your measure [UsageRolling6mo] as below and check if it works or not...

    UsageRolling6mo =
    VAR __MaxDate =
        TODAY ()
    VAR __EOM12 =
        EOMONTH ( __MaxDate, -12 )
    VAR __MinDate =
        DATE ( YEAR ( __EOM12 ), MONTH ( __EOM12 ), DAY ( __MaxDate ) )
    VAR __Table =
        FILTER (
            ALLSELECTED ( 'Calendar' ),
            'Calendar'[Date] <= __MaxDate
                && 'Calendar'[Date] >= __MinDate
        )
    RETURN
        AVERAGEX ( __Table, [Usage] )

    In addition, you can refer the following links to get the rolling average values:

    Rolling Averages In Power BI

    Rolling 12 Months Average in DAX

    Best Regards

3 Replies

  • Greg_Deckler's avatar
    Greg_Deckler
    Community Champion

    AA622 Try:

     

    Better RA =
        VAR __MaxDate = TODAY() 
        VAR __EOM12 = EOMONTH(__MaxDate,-12)
        VAR __MinDate = DATE(YEAR(__EOM12),MONTH(__EOM12),DAY(__MaxDate)
        VAR __Table = FILTER(ALLSELECTED('Table'),[Date] <= __MaxDate && [Date] >= __MinDate)
    RETURN
        AVERAGEX(__Table,[Usage])

     

    Both of the formulas you have currently are equally bad in my opinion because they do not account for leap years.

    • AA622's avatar
      AA622
      Helper II
      Better RA =
          VAR __MaxDate = TODAY() 
          VAR __EOM12 = EOMONTH(__MaxDate,-12)
          VAR __MinDate = DATE(YEAR(__EOM12),MONTH(__EOM12),DAY(__MaxDate)
          VAR __Table = FILTER(ALLSELECTED('Table'),[Date] <= __MaxDate && [Date] >= __MinDate)
      RETURN
          AVERAGEX(__Table,[Usage])

       

      HAving issues with the last VAR line. Line 5

       

      Am I supposed to reference a table?

       

      I tried referencing our Calendar Table. What would I put at the last line? I also seem to be having an issue with the mindate VAR in line 5

      • Anonymous's avatar
        Anonymous
        Not applicable

        Hi AA622 ,

        Please update the formula of your measure [UsageRolling6mo] as below and check if it works or not...

        UsageRolling6mo =
        VAR __MaxDate =
            TODAY ()
        VAR __EOM12 =
            EOMONTH ( __MaxDate, -12 )
        VAR __MinDate =
            DATE ( YEAR ( __EOM12 ), MONTH ( __EOM12 ), DAY ( __MaxDate ) )
        VAR __Table =
            FILTER (
                ALLSELECTED ( 'Calendar' ),
                'Calendar'[Date] <= __MaxDate
                    && 'Calendar'[Date] >= __MinDate
            )
        RETURN
            AVERAGEX ( __Table, [Usage] )

        In addition, you can refer the following links to get the rolling average values:

        Rolling Averages In Power BI

        Rolling 12 Months Average in DAX

        Best Regards