![]() | |
![]() |
| | Thread Tools | Display Modes |
#1
| |||
| |||
|
#2
| |||
| |||
|
#3
| |||
| |||
|
|
A couple of approaches that come to mind: 1) If there is a "Quote" dimension, with the leaf level consisting of unique QuoteID's, then a Count(Filter(.. can count the subset of unique QuoteID's with purchases (QPC). 2) An alternative with better performance: create a 2nd cube by copying the original. But add a SQL 'where' fact table filter condition to its partition: PC > 0. You need only the distinct count measure (QC2) in this 2nd cube. Now combine both cubes in a virtual cube, where QC is QC1 from the 1st cube, and QPC is QC2 from the 2nd cube (only includes fact table rows where Purchase Count > 0). - Deepak *** Sent via Developersdex http://www.developersdex.com *** Don't just participate in USENET...get rewarded for it! |
![]() |
| Thread Tools | |
| Display Modes | |
| |