July 10, 2005 at 11:50 am
Hi all,
Suppose I have a dimension containing members which are some sort of group numbers, e.g. 111,123,155, ....,266,288, .......
Querying my cube will give this result:
111 6
123 8
155 9
...
266 16
288 18
..............
If I want to insert a member, call it 100, which contains the total value of all group numbers starting with "1", i.e. 1XX, and this member '100' is NOT belonging to the original dimension ... So it would look like:
100 23
111 6
123 8
155 9
...
200 34
266 16
288 18
..............
What should I use to achieve this? Custom rollup formula or calculated member? Thanks.
delpiero
July 11, 2005 at 1:08 am
Hi, you have two ways:
Aggregate
(Filter([Dim_X].[Level_Y].Members, Left([Dim_X].CurrentMember.Name, 1) = "1"))I prefer method 1, because it is automatic, and if you add a new member 539, parent 500 will be created next time you process your dimension.
JG
July 11, 2005 at 2:19 am
Thanks JG,
For Method 1, is it I can't use it if my original fact table does not have the members 100,200? They are just treated as rollup members and they dun actually exist in the fact table.
Actually what I want to do is just what the ROLLUP operator in T-SQL does. Is there a disadvantage to retrieve the cube data into a temp table in T-SQL and then use the rollup operator to produce my desired results? Will it have a poor performance?
Best regards,
delpiero
Viewing 3 posts - 1 through 2 (of 2 total)
You must be logged in to reply to this topic. Login to reply