I added the following column to a table named 'User Story'. It uses a relationship that is disabled because it is redundant and only useful for this particular set of calculations. The values calculated for each row are correct.
Count of Failing Tests =
COUNTROWS (
CALCULATETABLE (
'Test Case Most Recent Result',
USERELATIONSHIP ('User Story'[WorkItemId], '_Relation'[WorkItemId]),
'Test Case Most Recent Result'[Outcome] = "Failed"
)
)
The problem comes when I add a nearly identical column. The only difference is that instead of results with an outcome of "Failed", I want to calculate the ones with an outcome of "Passed."
Count of Passing Tests =
COUNTROWS (
CALCULATETABLE (
'Test Case Most Recent Result',
USERELATIONSHIP ('User Story'[WorkItemId], '_Relation'[WorkItemId]),
'Test Case Most Recent Result'[Outcome] = "Passed"
)
)
When I try to add this column, I get the following message about circular dependencies.
A circular dependency was detected: User Story[Count of Passing Tests], User Story[Count of Failing Tests], User Story[Count of Passing Tests].
How is this circular? Is there a general pattern for counting the rows in a related table, specifying both a relationship and a filter? Any help is greatly appreciated.
