Cumulative Total for timestamp

[azakir]

Hi Guys. 

I have the following situation. I need to get the cumulative based on every hour. My table has time stamp with seconds / minutes, however, my target is for a 12 hour mark. I can divide the total benchmark by 12 and apply a running sum formula, but that does a running sum based on every time stamp. I would like this running sum to be based on every hour: 

Ideally, I'd like this red line to start from 3635 (43626 / 12) at 6:00PM, incrementing by 3635 every hour and at 6:00AM it should give 43626

[v-yilong-msft · Accepted solution]

Hi @azakir ,

I create a table as you mentioned.

Then I create some calculated columns and measures.

Here are calculated columns DAX codes:

Hour = HOUR('Table'[Timestamp])

HourlyBin =
'Table'[Timestamp]
    - SECOND ( 'Table'[Timestamp] ) * ( 1 / 86400 )
    - MINUTE ( 'Table'[Timestamp] ) * ( 1 / 1440 )

IncrementPerHour = DIVIDE(43626,12)

Here are measures DAX codes:

TargetRunningSum = 
    VAR StartTime = TIME(18,0,0) 
    VAR CurrentTime = MAX('Table'[HourlyBin])
    VAR _Incremental = MAX('Table'[IncrementPerHour])
    RETURN 
        IF(CurrentTime >= StartTime,
            (DATEDIFF(StartTime, CurrentTime,HOUR)+1) *_Incremental ,
            BLANK()
        )

CumulativeSum = 
    CALCULATE(
        SUM('Table'[Value]),
        FILTER(
            ALLSELECTED('Table'),
            'Table'[HourlyBin] <= MAX('Table'[HourlyBin])
        )
    )

Finally when you add columns you want, you will see what you want.

Best Regards

Yilong Zhou

If this post helps, then please consider Accept it as the solution to help the other members find it more quickly.

[azakir]

Hi @v-yilong-msft This has resolved the issue. Thank you for explaining in such a detail.