I noticed a different behaviour in pivot tables when row are collapsed vs expanded. I found a work-around but I think it is a bug!
I made a quik example in QS Desktop to illustrate the case.
- The pivot table has 3 lines and one very basic measure (count distinct(Email_Event_Id))
- In the data I can have null values in ay of the 3 lines / dimensions.
- In this particular pivot table I want to consider only non null data so I uncheked the "show null"
What happens:
- If I show the table expanded (right in the print screen), everything is ok. I see only the lines that I want, and the QTY measure is good
- If I show the table collapsed, the QTY measure includes records with null values in the hidden dimensions. It changes when expanding the line, and again when collapsing...
For me expanding a a line is not like doing a selection... Results should be the same at the top level expanded or collapsed. Is it a bug?
![pivot table example.jpg](/legacyfs/online/185455_pivot table example.jpg)