Forum Discussion
Substitute value when filtered
I have a dataset where some Items are members of a Group. For Groups, one Item is designated Type of Parent and the others, if any, are designated Type of Child. Each Item not in a Group carries an Amount. For Groups, only the Parent carries the Amount, however there is a Display Amount field that shows the Amount against each Child Item members of Groups. A correct Total of Amount would be SUM(Amount), while a SUM(Display Amount) would overcount.
I need a measure that sums Amount but works even if the Parent item is filtered out via a slicer or a user selection. Effectively I need to Sum the Amount column, but if a specific Child is selected I need to subsitute the Display Amount. If a selection filters out the Parent but includes more than one Child I need to use the distinct Amount for the Group.
I've tried playing around with VALUES and DISTINCT but I haven't found a solution that works for both Group and non-Group items.
| Item | Group | Type | Amount | Display Amount |
| 1 | A | Parent | 3 | 3 |
| 2 | A | Child | 3 | |
| 3 | A | Child | 3 | |
| 4 | B | Parent | 5 | 5 |
| 5 | B | Child | 5 | |
| 6 | B | Child | 5 | |
| 7 | C | Parent | 3 | 3 |
| 8 | C | Child | 3 | |
| 9 | C | Child | 3 | |
| 10 | null | 7 | 7 | |
| 11 | null | 5 | 5 | |
| 12 | null | 2 | 2 |
