Skip to main content
cancel
Showing results for 
Search instead for 
Did you mean: 

Earn the coveted Fabric Analytics Engineer certification. 100% off your exam for a limited time only!

Reply
AGo
Post Patron
Post Patron

Rolling sum optimization

Hello,

 

I populated a line chart with a cumulative function:

Rolling count = SUMX(FILTER(ALLSELECTED(Tab[timestamp]),Tab[timestamp]<=MAX(Tab[timestamp])),CALCULATE(SUM(Tab[DeltaWithPreviousRowValue]),Tab[Item]="Console1", Tab[Sensor]="Counters/Cuts"))
 
For each record I calculated the column [DeltaWithPreviousRowValue], I've got also the [value] which is a event incremental counter.
I can't use time hiearchy on my axis because I need to prompt the line automatically in hours if few days are selected, and days if a long time range is selected in the date filter.
Maybe a function that subtracts always the [Value] at start time and not using [DeltaWithPreviousRowValue] can work fine? In this case I have no idea how to fix pre-calculated MIN(Tab[value]) as a variable/sort of parameter.
 
timestampItemSensorvalueRankDeltaWithPreviousRowValue
11/06/2020 17:04:12Console1Counters/Cuts241610031
11/06/2020 17:04:07Console1Counters/Cuts241510021
11/06/2020 17:04:01Console1Counters/Cuts241410011
11/06/2020 17:03:56Console1Counters/Cuts241310001
11/06/2020 17:03:50Console1Counters/Cuts24129991
11/06/2020 17:03:45Console1Counters/Cuts24119981
11/06/2020 17:03:35Console1Counters/Cuts24109971
11/06/2020 17:03:14Console1Counters/Cuts24099961
11/06/2020 17:02:53Console1Counters/Cuts24089951
11/06/2020 17:02:35Console1Counters/Cuts24079941
11/06/2020 17:02:27Console1Counters/Cuts24069931
11/06/2020 17:02:22Console1Counters/Cuts24059921
11/06/2020 17:02:18Console1Counters/Cuts24049911
11/06/2020 17:02:15Console1Counters/Cuts24039902
11/06/2020 17:02:09Console1Counters/Cuts24019891
11/06/2020 17:02:05Console1Counters/Cuts24009882
11/06/2020 17:01:59Console1Counters/Cuts23989871
11/06/2020 17:01:56Console1Counters/Cuts23979862
11/06/2020 17:01:50Console1Counters/Cuts23959852
11/06/2020 17:01:44Console1Counters/Cuts23939841
11/06/2020 17:01:40Console1Counters/Cuts23929832
11/06/2020 17:01:34Console1Counters/Cuts23909821
11/06/2020 17:01:31Console1Counters/Cuts23899812
11/06/2020 17:01:25Console1Counters/Cuts23879802
11/06/2020 17:01:19Console1Counters/Cuts23859791
11/06/2020 17:01:15Console1Counters/Cuts23849782
11/06/2020 17:01:09Console1Counters/Cuts23829771
11/06/2020 17:01:06Console1Counters/Cuts23819762
11/06/2020 17:01:00Console1Counters/Cuts23799752
11/06/2020 17:00:54Console1Counters/Cuts23779741

 

Many thanks in advance!

 
3 REPLIES 3
sturlaws
Resident Rockstar
Resident Rockstar

Hi @AGo 

 

I don't get what your issue is. Could you create a sample report and share it?

 

Cheers,
Sturla

amitchandak
Super User
Super User

@AGo , a New column like

maxx(filter(Table, table[timestamp] <earlier(table[timestamp] )),table[timestamp]) -table[timestamp]

No, I'm sorry, the resulting value I need in my measure is not from timestamp column, it has to be extracted from calculations based on [vaue] or [DeltaWithPreviousRowValue] depending on which one is easier to use.

Helpful resources

Announcements
April AMA free

Microsoft Fabric AMA Livestream

Join us Tuesday, April 09, 9:00 – 10:00 AM PST for a live, expert-led Q&A session on all things Microsoft Fabric!

March Fabric Community Update

Fabric Community Update - March 2024

Find out what's new and trending in the Fabric Community.

Top Solution Authors