Forum Discussion
Help Flagging Status Change
I want to have a measure that will display 1 when a status has changed, and 0 when the status is the same as yesterday's value. My data set is rather large, so I want to avoid using a calculated column if possible.
I feel like I'm so close to having the right DAX, but this measure is not giving me the result I am looking for.
Change Status Flag =
INT (
COUNTROWS (
FILTER (
'Table',
'Table'[Key]
= EARLIER ( 'Table'[Key] )
&& 'Table'[Status]
<> EARLIER ( 'Table'[Status] )
&& 'Table'[Snapshot Date]
= EARLIER ( 'Table'[Snapshot Date] )
)
) > 0
)
What I'm getting -
value set to 1 even though the status from the previous day hasn't changed
What I want:
Mark only the rows where the status is different than yesterday.
Could anyone help me out with the DAX to accomplish this? Thank you in advance for your help.
Attached is my sample data.
Snapshot DateKeyStatus
|
Use the following DAX to create the column:
Change Status Flag = VAR CurrentDate = 'Table'[Snapshot Date] VAR CurrentKey = 'Table'[Key] VAR CurrentStatus = 'Table'[Status] VAR PreviousStatus = CALCULATE( FIRSTNONBLANK('Table'[Status], 1), FILTER( 'Table', 'Table'[Snapshot Date] = CurrentDate - 1 && 'Table'[Key] = CurrentKey ) ) RETURN IF(ISBLANK(PreviousStatus), 0, IF(CurrentStatus <> PreviousStatus, 1, 0))
5 Replies
- foodd
Community Champion
Use the following DAX to create the column:
Change Status Flag = VAR CurrentDate = 'Table'[Snapshot Date] VAR CurrentKey = 'Table'[Key] VAR CurrentStatus = 'Table'[Status] VAR PreviousStatus = CALCULATE( FIRSTNONBLANK('Table'[Status], 1), FILTER( 'Table', 'Table'[Snapshot Date] = CurrentDate - 1 && 'Table'[Key] = CurrentKey ) ) RETURN IF(ISBLANK(PreviousStatus), 0, IF(CurrentStatus <> PreviousStatus, 1, 0)) - Ashish_Mathur
Super User
Hi,
Please find attached the PBI file.
Hope this helps.
- ThxAlot
So far, few users opt for window functions; perhaps few have profound comprehension of them.
I hope the window functions are more performant in such scenarios than "old school" FILTER().
