The ultimate Fabric, Power BI, SQL, and AI community-led learning event. Save €200 with code FABCOMM.
Get registeredEnhance your career with this limited time 50% discount on Fabric and Power BI exams. Ends September 15. Request your voucher.
List.Average is not computed correctly in this code.
[
Source = #table({"Value"}, {{null}, {1}}) & #table({"Value"}, {{2}}),
Avg = List.Average(Source[Value])
]
It computes to 1 whereas it should be 1.5. Such incorrect List.Average calculation seems to occur anytime you are working with a non-buffered table built by combining tables. If you do an average on a column in the table like the code above or in a Table.Group, the average calculation erroneously includes nulls in the value count for the denominator. If you buffer the table or buffer the list before averaging, the error does not occur. I do not see this error with List.NonNullCount, List.Median, or List.StandardDeviation.
Has this been raised as a defect with the Power BI team? I've tested this behavior with multiple tables in both Power Query for Excel and Power BI and the defect shows up in all cases.
Solved! Go to Solution.
Thank you, Vijay, for directing me to the right place to submit such issues. I have done that now and here's a link to the submission. Please fix this defect in List.Average - Microsoft Fabric Community
This is a nice reporting. The issues (or bugs) can be reported to Power BI Dev team from here
https://community.fabric.microsoft.com/t5/Issues/idb-p/Issues
After searching there, I don't see any issue logged in this regard.
So, you will have to help out the community by logging an issue there.
Thank you, Vijay, for directing me to the right place to submit such issues. I have done that now and here's a link to the submission. Please fix this defect in List.Average - Microsoft Fabric Community