Tutorial with example
A common example in a Power BI table is to have multiple dates, which are related to the same Date table. As you are limited to one active relationship, we need to use the DAX function USERELATIONSHIP to manage the inactive relationships.
Download example file:
To get the files:
1. Go to Curbal Download Center > Dax Fridays
2. Get File #21
No download file available for the sumif video. Please use the northwind dataset to practice.
- USERELATIONSHIP can only be used in functions that take a filter as an argument, for example: CALCULATE, CALCULATETABLE, CLOSINGBALANCEMONTH, CLOSINGBALANCEQUARTER, CLOSINGBALANCEYEAR, OPENINGBALANCEMONTH, OPENINGBALANCEQUARTER, OPENINGBALANCEYEAR, TOTALMTD, TOTALQTD and TOTALYTD functions.
- USERELATIONSHIP uses existing relationships in the model, identifying relationships by their ending point columns.
- In USERELATIONSHIP, the status of a relationship is not important; that is, whether the relationship is active or not does not affect the usage of the function. Even if the relationship is inactive, it will be used and overrides any other active relationships that might be present in the model but not mentioned in the function arguments.
- An error is returned if any of the columns named as an argument is not part of a relationship or the arguments belong to different relationships.
- If multiple relationships are needed to join table A to table B in a calculation, each relationship must be indicated in a different USERELATIONSHIP function.
- If CALCULATE expressions are nested, and more than one CALCULATE expression contains a USERELATIONSHIP function, then the innermost USERELATIONSHIP is the one that prevails in case of a conflict or ambiguity.
- Up to 10 USERELATIONSHIP functions can be nested; however, your expression might have a deeper level of nesting, ie. the following sample expression is nested 3 levels deep but only 2 for USEREALTIONSHIP: =CALCULATE(CALCULATE( CALCULATE( <anyExpression>, USERELATIONSHIP( t1[colA], t2[colB])), t99[colZ]=999), USERELATIONSHIP( t1[colA], t2[colA])).