Filtering on SUMMARIZECOLUMNS possible ?

Hi v-ljerr-msft

Sounds good. Internally we are only running december version, so i'll try and make a version on my own. One thing i noticed is that that the SUMMARIZECOLUMNS is running in the current date context.

Ex. if i have the following:

[Total Sales] = SUM('Sales'[Sales])

[Total Sales SPLY] = CALCULATE([Total Sales];SAMEPERIODLASTYEAR ('Date'[Date])))

[Total Sales Diff] = [Total Sales]-[Total Sales SPLY]

[Total Sales Diff%] = DIVIDE([Total Sales Diff];[Total Sales SPLY])

Measure 4 =
SUMX (
    FILTER (
        SUMMARIZECOLUMNS (
            Locations[Country],
            Products[Category],
            "abc", [Total Sales Diff],
            "efg", SUMX (Sales, [Total Sales Diff%] )
        ),
        [efg] < 0.05
    ),
    [abc]
)

So when in PowerBI selects Month= February 2018 it summarizes the combinations of Country and Category for [Total Sales] in February 2018, meaning that it will elave out the potential combinations of Country and Category for February 2017 (the SPLY measure used for the [Total Sales Diff] calcuation. In other words i don't get the [Total Sales Diff] for those combinations of Country and Category that were in February 2017 that was not in February 2018 - they are just left out of the SUMMARIZECOLUMNS.

In the current calculation i can solve it by using VALUES instead of SUMMARIZE, but it heavy on the ressources. Any idea of how to include that in the SUMMARIZECOLUMNS ?