Sum values of Measure 1 when condition applies to measure 2

[athgeor]

Hi all, 

I'm trying to create a measure that sums the values of a measure (1) when the values of another measure (2) is above 20. 

I have tried different measures I found in the forum but none of them seems to work. (Note: Measure 1 and Measure 2 can change depending on the time bucket I apply)

e.g I tried  CALCULATE( [Customers], IF ([Spend]>20)), I tried variations of SUMX and didn't get a result

	Measure 1	Measure 2	Measure Neeeded
Time Bucket	Customers	Spend
00:00 - 00:15	10	45	10
00:15 - 00:30	25	40	25
00:30 - 00:45	30	40	30
00:45 - 01:00	5	10
01:00 - 01:15	80	20
01:15 - 01:30	90	10
01:30 - 01:45	100	15
01:45 - 02:00	120	58	120
02:00 - 02:15	90	50	90
…	…	…	…
…	…	…	…

Can anyone help?

Many thanks in advance

[AlexisOlson]

You can create a calculated table within a measure that you then apply SUMX to like this:

BigSpenders =
VAR _Summary_ =
    ADDCOLUMNS (
        SUMMARIZE ( 'Table', 'Table'[Time Bucket] ),
        "@Customers", CALCULATE ( SUM ( 'Table'[Customers] ) ),
        "@Spend", CALCULATE ( SUM ( 'Table'[Spend] ) )
    )
RETURN
    SUMX ( FILTER ( _Summary_, [@Spend] > 20 ), [@Customers] )

[athgeor]

Hi @AlexisOlson 

I appreciate the help.

I don't seem to be able to select the measures in the sum functions...

I believe this works with the native columns but does not seem to work with my measures (@Customers & @Spend)

[AlexisOlson · Accepted solution]

I created a _Summary_ table that has columns named [@Customers] and [@Spend] and the used SUMX to iterate over the _Summary_ table, filtering on the [@Spend] column and summing on the [@Customers] column.

This pattern works assuming [Customers] and [Spend] are columns of 'Table'. If those are actually measures instead of columns, then the DAX is slightly simpler:

BigSpenders =
VAR _Summary_ =
    ADDCOLUMNS (
        SUMMARIZE ( 'Table', 'Table'[Time Bucket] ),
        "@Customers", [Customers],
        "@Spend", [Spend]
    )
RETURN
    SUMX ( FILTER ( _Summary_, [@Spend] > 20 ), [@Customers] )

[athgeor]

Thanks @AlexisOlson, the last code worked like a charm! Many thanks!