Rolling 3 month sum, divide by value across tables

[DCELL]

Hello,

My problem is a bit difficult to explain in words alone, so here's a spreadsheet to make it clearer:

I have 2 fact tables. Columns with black text are columns that I have, and columns in blue are ones that I don't have.
The only number I care about obtaining is the one on the far right (in blue). My dataset is small so creating calculated columns isn't a problem if necessary.

I have created a rolling 3 month sum measure as follows:

Rolling 3 month sum = 
    CALCULATE(
        SUM(Table1[COGS]),
        DATESINPERIOD(Table1[Date],
        LASTDATE(Table1[Date]), -3, MONTH)
    )

but when dividing 'Table2'[Balance] by the result in this measure, it needs to divide it by country, not as a whole.

Any help is appreciated as always 🙂

[MFelix · Accepted solution]

Hi @DCELL ,

I created a model with the following tables:

• COGS
• BALANCE
• Country (Dimension table)
• Date (Dimension table)

The dimension table are use to make a relationship between the COGS and Balance table then create the following two measures:

Rolling 3 month sum =
IF (
    DATEDIFF (
        MINX (
            DATESINPERIOD ( 'Calendar'[Date]; LASTDATE ( 'Calendar'[Date] ); -3; MONTH );
            'Calendar'[Date]
        );
        MAXX (
            DATESINPERIOD ( 'Calendar'[Date]; LASTDATE ( 'Calendar'[Date] ); -3; MONTH );
            'Calendar'[Date]
        );
        DAY
    ) < 61;
    BLANK ();
    CALCULATE (
        SUM ( COGS[COGS] );
        DATESINPERIOD ( 'Calendar'[Date]; LASTDATE ( 'Calendar'[Date] ); -3; MONTH )
    )
)

Balance 3 Months =
IF (
    [Rolling 3 month sum] = BLANK ();
    BLANK ();
    SUM ( Balance[Balance] ) / [Rolling 3 month sum] * 91
)

See result below and in attach file:

[DCELL]

Hi @MFelix ,

Thank you so much for your help. It works!