nikipedia
Technical User
- May 27, 2009
- 2
Ok here is a nice tasty Transformer problem we can’t work out here.
We have a measure called ‘actives’ which is the count of distinct active user Ids for a given dimension. For instance against our locations visited we have the following
North 4255
South 2649
East 3311
West 2420
Central 4032
Scotland659
However when it is working out the aggregation in the cube build, it then totals this up to be 17,326 but this isn’t the count of distinct Ids for that level of granularity as a member could have visited 2 or more locations... Whatever subtotals it calculates it needs to be the count distinct of that level. The aggregation type of the data item in FM is calculated which transformer doesn’t support. Has anyone any ideas how this can be done?
Many Thanks
We have a measure called ‘actives’ which is the count of distinct active user Ids for a given dimension. For instance against our locations visited we have the following
North 4255
South 2649
East 3311
West 2420
Central 4032
Scotland659
However when it is working out the aggregation in the cube build, it then totals this up to be 17,326 but this isn’t the count of distinct Ids for that level of granularity as a member could have visited 2 or more locations... Whatever subtotals it calculates it needs to be the count distinct of that level. The aggregation type of the data item in FM is calculated which transformer doesn’t support. Has anyone any ideas how this can be done?
Many Thanks