I am creating a report that uses a cube as its data source, and within the cube is a calculated measure that's a percentage. I need to display this percentage measure based on a geographical hiearchy I have set up in the cube; however when I do a real simple build of the dataset, add the measure to the matrix in the report, the percentages are off from what I would expect to see. (When validated in a pivot table in Excel or in BIDS browser) I used the query builder in SSRS which produced the following MDX:
SELECT NON EMPTY { [Measures].[%Measure] } ON COLUMNS, NON EMPTY {[Geography].[GeoHierarchy].[Level1].ALLMEMBERS, ([Geography].[GeoHierarchy].[Level2].ALLMEMBERS ) } DIMENSION PROPERTIES MEMBER_CAPTION, MEMBER_UNIQUE_NAME ON ROWS FROM [Cube] CELL PROPERTIES VALUE, BACK_COLOR, FORE_COLOR, FORMATTED_VALUE, FORMAT_STRING, FONT_NAME, FONT_SIZE, FONT_FLAGS
At the moment I only want to show the Level 1 results in the column grouping, but will need to add the drill down, so the matrix I have right now looks something like this:
Europe Africa North America
20% 60% 35%
NOTE - the percentages above do not need to sum to 100%, they are measured within each Level 1 geography.
I pulled the Level 1 into the column grouping area, and the % Measure into the data portion of the matrix. I removed the "sum" portion out of the field information and tried both Aggregate(Fields!%measure.value, "Level 1") and just Avg(Fields!%Measure.value). The aggregate function produces blank results and the Avg function produces results that are kind of close to what I expect, but not quite. (And in the interest of data quality, I'd like the SSRS report values to match what I see in Excel or BIDS)
What else can I try to get consistent results in SSRS? This is on SQL Server 2008 R2. Ultimately the report will need to have parameters/filters which in reading other related threads means aggregate is out.