I'm trying to form an MDX query such that it returns only the combinations of two dimensions where a measure meets a certain criteria. I thought this would be pretty straight forward using the FILTER function, i.e.
SELECT
NON EMPTY FILTER({[Program].[ByRegion].[Program] * [Performance Metric].[Metric].CHILDREN }, [Measures].[Point Percentage] < .95) ON ROWS,
NON EMPTY ( HIERARCHIZE([Calendar Period].[Y-Q-M].[Month of Quarter].&[3]&[1]&[2009]) , [Measures].[Point Percentage] )ON COLUMNS
FROM [QEP Revenue]
However, after running the query, it is pretty easy to see that I have a mistake because the very first result has a Point Percentage of 1.5172 which is obviously more than .95.
If I completely remove the filter:
SELECT
--NON EMPTY FILTER({[Program].[ByRegion].[Program] * [Performance Metric].[Metric].CHILDREN }, [Measures].[Point Percentage] < .95) ON ROWS,
NON EMPTY ({[Program].[ByRegion].[Program] * [Performance Metric].[Metric].CHILDREN }) ON ROWS,
NON EMPTY ( HIERARCHIZE([Calendar Period].[Y-Q-M].[Month of Quarter].&[3]&[1]&[2009]) , [Measures].[Point Percentage] )ON COLUMNS
FROM [QEP Revenue]
I get a similar result set including values above .95. Am I completely missing the point of a filter, or is there an issue with attempting to filter two dimensions at once?