Hi

Is there any reason why I shouldn't mix the cubes referred to in the
ORDER and FILTER parts of an MDX statement.

This query is ignoring the FILTER, but if I remove the ORDER part the
FILTER works.

I am wondering if there is a problems caused by the fact that the
FILTER is based on one cube, whilst the ORDER is based on another

The middle section {XFILTERBYLEVEL( {XSUBSETALL( [Product] )}, 0)}uses
MDX functions specific to a 3rd party product to retrive all leaf
elements in a product dimension.

{ ORDER( {FILTER( {XFILTERBYLEVEL( {XSUBSETALL( [Product] )},
0)},[XProductScore].([Data].[SALES],[Variable Type].[% +/- Budget]) >
0.000000)}, [Category Plan].([Variable Type].[% +/- Budget],[Cost
Type].[Retail],[Version].[Actual
TY],[Data].[SALES],[Time].[05/06-W05]), BDESC) }

Regards

John

MDX queries are restricted to a single cube. The only way to include data
from other cubes is to use the LookupCube function.
As an alternative, you can build a virtual cube.
HTH,
Brian
www.geocities.com/brianaltmann/olap.html


"John Hobson" wrote:

Many thanks Brian

Regards

John

Brian Altmann wrote: