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

Next up in the FabCon + SQLCon recap series: The roadmap for Microsoft SQL and Maximizing Developer experiences in Fabric. All sessions are available on-demand after the live show. Register now

Reply
GunnerJ
Post Patron
Post Patron

Help with Circular Dependency

First I want to say that I've looked through several of the other post related to cirular dependency but they haven't helped me correct this issue. The two calculated columns below create a circular dependency. "Last SO" was created first and I'd prefer to not change it if at all possible. "Last Count" is looking at a cumulative total and trying to show what the value was before the current row where account is the same and the date came before. Is there some kind of filter magic we can use to keep the same functionality but to get rid of the error?

 

 

Last SO =
CALCULATE(LASTNONBLANK('SO Attrition'[BI_SO_TYPE_CD],1),FILTER('SO Attrition', 'SO Attrition'[BI_ACCT] = EARLIER('SO Attrition'[BI_ACCT]) && 'SO Attrition'[NEEDED_DT] < EARLIER('SO Attrition'[NEEDED_DT]) && 'SO Attrition'[SRV_MAP_LOC] = EARLIER('SO Attrition'[SRV_MAP_LOC])&& 'SO Attrition'[BI_SO_TYPE_CD] <> EARLIER('SO Attrition'[BI_SO_TYPE_CD])))

 

Last Count =
CALCULATE(LASTNONBLANK('SO Attrition'[DISC running Total],1),FILTER('SO Attrition', 'SO Attrition'[BI_ACCT] = EARLIER('SO Attrition'[BI_ACCT]) && 'SO Attrition'[NEEDED_DT] < EARLIER('SO Attrition'[NEEDED_DT])))
1 ACCEPTED SOLUTION
AlexisOlson
Super User
Super User

I think you can avoid this by eliminating the context transition induced by CALCULATE.

 

See if these versions work any better:

Last SO =
VAR FilteredTable =
    FILTER (
        'SO Attrition',
        'SO Attrition'[BI_ACCT] = EARLIER ( 'SO Attrition'[BI_ACCT] )
            && 'SO Attrition'[NEEDED_DT] < EARLIER ( 'SO Attrition'[NEEDED_DT] )
            && 'SO Attrition'[SRV_MAP_LOC] = EARLIER ( 'SO Attrition'[SRV_MAP_LOC] )
            && 'SO Attrition'[BI_SO_TYPE_CD] <> EARLIER ( 'SO Attrition'[BI_SO_TYPE_CD] )
    )
RETURN
    MAXX ( FilteredTable, 'SO Attrition'[BI_SO_TYPE_CD] )
Last Count =
VAR FilteredTable =
    FILTER (
        'SO Attrition',
        'SO Attrition'[BI_ACCT] = EARLIER ( 'SO Attrition'[BI_ACCT] )
            && 'SO Attrition'[NEEDED_DT] < EARLIER ( 'SO Attrition'[NEEDED_DT] )
    )
RETURN
    MAXX ( FilteredTable, 'SO Attrition'[DISC running Total] )

 

It's possible to get it to work with CALCULATE but easier to mess up, especially if you're not comfortable with context transition concepts.

View solution in original post

5 REPLIES 5
AlexisOlson
Super User
Super User

I think you can avoid this by eliminating the context transition induced by CALCULATE.

 

See if these versions work any better:

Last SO =
VAR FilteredTable =
    FILTER (
        'SO Attrition',
        'SO Attrition'[BI_ACCT] = EARLIER ( 'SO Attrition'[BI_ACCT] )
            && 'SO Attrition'[NEEDED_DT] < EARLIER ( 'SO Attrition'[NEEDED_DT] )
            && 'SO Attrition'[SRV_MAP_LOC] = EARLIER ( 'SO Attrition'[SRV_MAP_LOC] )
            && 'SO Attrition'[BI_SO_TYPE_CD] <> EARLIER ( 'SO Attrition'[BI_SO_TYPE_CD] )
    )
RETURN
    MAXX ( FilteredTable, 'SO Attrition'[BI_SO_TYPE_CD] )
Last Count =
VAR FilteredTable =
    FILTER (
        'SO Attrition',
        'SO Attrition'[BI_ACCT] = EARLIER ( 'SO Attrition'[BI_ACCT] )
            && 'SO Attrition'[NEEDED_DT] < EARLIER ( 'SO Attrition'[NEEDED_DT] )
    )
RETURN
    MAXX ( FilteredTable, 'SO Attrition'[DISC running Total] )

 

It's possible to get it to work with CALCULATE but easier to mess up, especially if you're not comfortable with context transition concepts.

@AlexisOlson 

That worked in getting rid of the circular dependency for those columns! Sadly one bit of bad news is that it seems to have passed the dependency onto another column. "DISC Running Total" is a cumulative sum working on an account level. Its saying its dependency is related to "Last Count" which you've provided. I know "Last Count" literally references it but I'm unsure how to make a cumulative total without calculate. Any help?

DISC running Total =
VAR LastNeededDt = CALCULATE(MAX('SO Attrition'[NEEDED_DT]))
VAR countOfDISC =
CALCULATE(
SUM('SO Attrition'[add or disc]),
FILTER(
ALLSELECTED('SO Attrition'),
'SO Attrition'[NEEDED_DT] <= LastNeededDt),
'SO Attrition'[BI_ACCT] = EARLIER('SO Attrition'[BI_ACCT]))
Return countOfDISC



You could rewrite this one like I did with the others.

 

Here's a CALCULATE version too that I think should work:

 

DISC running Total =
CALCULATE (
    SUM ( 'SO Attrition'[add or disc] ),
    ALLEXCEPT ( 'SO Attrition', 'SO Attrition'[BI_ACCT] ),
    'SO Attrition'[NEEDED_DT] <= EARLIER ( 'SO Attrition'[NEEDED_DT] )
)

 

@AlexisOlson it worked! thank you! 

You referenced Context transitions. Just for future knowledge do you have a best source to get more familiar?

I included a link to an article upon my first mention.

 

Here's that link again along with another more recently written article:
https://www.sqlbi.com/articles/understanding-context-transition/

https://www.burningsuit.co.uk/context-transition-where-the-row-context-becomes-a-filter-context/

Helpful resources

Announcements
New to Fabric survey Carousel

New to Fabric Survey

If you have recently started exploring Fabric, we'd love to hear how it's going. Your feedback can help with product improvements.

Power BI DataViz World Championships carousel

Power BI DataViz World Championships - June 2026

A new Power BI DataViz World Championship is coming this June! Don't miss out on submitting your entry.

Join our Fabric User Panel

Join our Fabric User Panel

Share feedback directly with Fabric product managers, participate in targeted research studies and influence the Fabric roadmap.

March Power BI Update Carousel

Power BI Community Update - March 2026

Check out the March 2026 Power BI update to learn about new features.