You can pre-filter or post-filter data.
Here is an example of post-filtering, where the filter is applied to the destination buffer just before it is written back to the cube:
api.Data.Calculate("S#POV = RemoveZeros(S#SrcScenario)","A#Root.Base.Remove(A1,A2,A3))")
Here is an example of pre-filtering, where the filter can be applied to any one or all of the source buffers as desired:
api.Data.Calculate("S#POV = FilterMembers(RemoveZeros(S#SrcScenario),[A#Root.Base.Remove(A1,A2,A3)]))")
It usually more performance efficient to pre-filter the data, otherwise you are having OS calculate way more rows than are needed, only to toss them out after the fact.
Of course, with more advanced data buffer handlers, it is possible to filter or manipulate any data buffer way beyond the above two simple filtering methods.
