Hi,
maybe I'm too tired, but I have been dealing with this for hours ... I get diferent result for the aggregate value of a level than the sum of its members.
Here is my MDX code for a time dimension Year-Week-Weekday
And here the results I get. All weeks of all years are correctly added up but weeks 25 & 26 2015 (in red).
Using another time dimension (Year-Month-Day), I also get a diferent value for June 2015 than the sum of its days. Only June 2015 sum is incorrect, all other months of other years are correct.
The diference in both cases is exactly the same amount, so I guess for eom reason some records are not considered on the sum. The detail data is the correct one checked with an SQL.
Any idea what can result in such behaviour?
maybe I'm too tired, but I have been dealing with this for hours ... I get diferent result for the aggregate value of a level than the sum of its members.
Here is my MDX code for a time dimension Year-Week-Weekday
Code:
SELECT
NON EMPTY {[Measures].[Vendes]} ON COLUMNS,
NON EMPTY {Hierarchize(
{{[TempsDimension.SetmanaHie].[AnyANSI].members},
{{[TempsDimension.SetmanaHie].[SetmanaLev].members},{[TempsDimension.SetmanaHie].[DiaSetmana].members}}}
)} ON ROWS
FROM [CubVendes]| 24 | 850.418,79 | ||
|---|---|---|---|
| Dilluns | 97.156,64 | ||
| Dimarts | 163.442,31 | ||
| Dimecres | 164.059,12 | ||
| Dijous | 167.497,27 | ||
| Divendres | 206.763,09 | ||
| Dissabte | 51.500,36 | ||
| 25 | 870.462,31 | ||
| Dilluns | 83.199,16 | ||
| Dimarts | 180.153,47 | ||
| Dimecres | 174.032,31 | ||
| Dijous | 172.740,95 | ||
| Divendres | 206.504,39 | ||
| Dissabte | 49.578,26 | ||
| Diumenge | 276,58 | ||
| 26 | 78.761,61 | ||
| Dilluns | 97.557,47 | ||
| Dimarts | 268.112,54 | ||
| Dijous | 137.315,38 | ||
| Divendres | 80,61 | ||
| 27 | 38,40 | ||
| Dilluns | 38,40 | ||
| 28 | 0,00 | ||
Using another time dimension (Year-Month-Day), I also get a diferent value for June 2015 than the sum of its days. Only June 2015 sum is incorrect, all other months of other years are correct.
The diference in both cases is exactly the same amount, so I guess for eom reason some records are not considered on the sum. The detail data is the correct one checked with an SQL.
Any idea what can result in such behaviour?