This time we’re going bigger than ever. Fabric, Power BI, SQL, AI and more. We're covering it all. You won't want to miss it.
Learn moreLevel up your Power BI skills this month - build one visual each week and tell better stories with data! Get started
I have two tables called 'left' and 'right'
| left.colA | left.Val |
|-----------|----------|
| 1 | A |
| 2 | B |
| 4 | D |
| 5 | E |
| right.colA | right.colB |
|------------|------------|
| 1 | 3 |
| 2 | 4 |
| 3 | 5 |
They have one ACTIVE (left.colA-right.colA) and one INACTIVE (left.colA-right.colB) relationship.
I want to perform a DAX query that generates what a LEFT ANTI join (FIlter 'left' such as it does not contain any value in left.colA that exists in right.colB) would produce on tbl[left] where the join columns are left.colA to right.colB.
So I want to end up with the following
Now, if I perform a join on ACTIVE relationship columns, I can take advantage of in-built ISEMPTY. But is there any way to utilize ISEMPTY if the joins are on INACTIVE relationship columns? I want to avoid COUNTROWS as I suspect it will be a lot slower on a large dataset [Ref-SQLBI]
Table =
var leftAntiActiveRelation = FILTER('left',NOT ISEMPTY(RELATEDTABLE('right')))
var rightAntiInActiveRelation = FILTER('left',CALCULATE(COUNTROWS('right'),USERELATIONSHIP('left'[left.colA],'right'[right.colB]))=0)
return rightAntiInActiveRelation
@AlexisOlson @CNENFRNL @OwenAuger
Thank you in advance.
Sample pbix is attached
Solved! Go to Solution.
Hi, @smpa01 , you always come up with tricky and interesting questiongs.
In theory, the expected result can be achieved this way,
But from perspective of modeling, such an anti-join is a typical violation of reference integrity
| Thanks to the great efforts by MS engineers to simplify syntax of DAX! Most beginners are SUCCESSFULLY MISLED to think that they could easily master DAX; but it turns out that the intricacy of the most frequently used RANKX() is still way beyond their comprehension! |
DAX is simple, but NOT EASY! |
You can also skip relationships entirely and use right.colA or right.colB the same way like this:
Table =
VAR _Right = VALUES ( 'right'[right.colB] )
RETURN
FILTER ( 'left', NOT 'left'[left.colA] IN _Right )
I'm not sure about the efficiency of this though. This might be better since it doesn't have an iterator:
Table =
VAR _ColA =
EXCEPT ( VALUES ( 'left'[left.colA] ), VALUES ( 'right'[right.colB] ) )
RETURN
CALCULATETABLE ( 'left', TREATAS ( _ColA, 'left'[left.colA] ) )
Hi, @smpa01 , you always come up with tricky and interesting questiongs.
In theory, the expected result can be achieved this way,
But from perspective of modeling, such an anti-join is a typical violation of reference integrity
| Thanks to the great efforts by MS engineers to simplify syntax of DAX! Most beginners are SUCCESSFULLY MISLED to think that they could easily master DAX; but it turns out that the intricacy of the most frequently used RANKX() is still way beyond their comprehension! |
DAX is simple, but NOT EASY! |
Check out the April 2026 Power BI update to learn about new features.
Sign up to receive a private message when registration opens and key events begin.
If you have recently started exploring Fabric, we'd love to hear how it's going. Your feedback can help with product improvements.
| User | Count |
|---|---|
| 35 | |
| 32 | |
| 26 | |
| 21 | |
| 18 |
| User | Count |
|---|---|
| 68 | |
| 36 | |
| 32 | |
| 25 | |
| 23 |