Power BI is turning 10! Tune in for a special live episode on July 24 with behind-the-scenes stories, product evolution highlights, and a sneak peek at what’s in store for the future.
Save the dateEnhance your career with this limited time 50% discount on Fabric and Power BI exams. Ends August 31st. Request your voucher.
Hi Forum,
Inspired by this article, I'd like to create automated unit tests to test measures within my Power BI Semantic Model (Dataset).
I'm planning to:
- Use Powershell to connect to and run through the tests
- DAX scripts to query the Semantic Model itself
Since these are unit tests, I want to contain "fake" data within the test itself (not have to worry about the underlying database changing).
A few approaches I've attempted.
1. Overwrite Existing Table for Test (Doesn't work; PBI won't let me overwrite the actual_table in the semantic model)
DEFINE
TABLE actual_table = UNION(ROW("order","A1"),ROW("order","B1")) -- attempt to update as a "local variable"
MEASURE 'Measures Table'[COUNT_ORDERS] = DISTINCTCOUNT('actual_table'[order]))
EVALUATE
ROW("COUNT_ORDERS",'Measures Table'[COUNT_ORDERS])
Error: Table '<ccon>actual_table</ccon>' cannot be created because a table or variable with the same name already exists.
2. Update Table Reference (Pretty sure PBI was designed purposely not to do this)
DEFINE
TABLE test_table = UNION(ROW("order","A1"),ROW("order","B1"))
MEASURE 'Measures Table'[COUNT_ORDERS] = DISTINCTCOUNT('actual_table'[order])) -- somehow update table reference 'actual_table' to 'test_table'
EVALUATE
ROW("COUNT_ORDERS",'Measures Table'[COUNT_ORDERS])
3. SWITCH between actual table and test table (I don't like this approach because I'd have to duplicate the SWITCH across all measures)
DEFINE
TABLE test_table = UNION(ROW("order","A1"),ROW("order","B1"))
MEASURE 'Measures Table'[COUNT_ORDERS] =
SWITCH(2,
1,DISTINCTCOUNT('actual_table'[order]),
2,DISTINCTCOUNT('test_table'[order])
)
EVALUATE
ROW("COUNT_ORDERS",'Measures Table'[COUNT_ORDERS])
Do you have any alternative recommendations? Your help is greatly appreciated.
Have your test data be part of the regular data source refreshes. For production use a filter that excludes the test data. For unit testing flip the filter to exclude the production data.
User | Count |
---|---|
22 | |
11 | |
8 | |
6 | |
6 |
User | Count |
---|---|
25 | |
12 | |
11 | |
7 | |
6 |