December 18, 2010 at 11:48 am
Suppose i have a table
Name|Roll_No|Test_Date|Marks
which contains the following row
Kevin| 1003 | 12/1/2010 |100
Now the cube created contains 2 dimensions (Name) and (Time) and a measure which is RollNo.
Now when the cube is viewed using 2 dimensions along with the measure everything is shown.
However when exploring this cube if we choose one dimension with one measure we get nothing..
The reason why i chose one dimension(Name) and one measure(RollNo) was to expect an output similar to a query like "list RollNo of all names" but i got an empty screen ... can we get such an ouput from the same cube or will have to make a new one??
December 18, 2010 at 12:07 pm
I think you've maybe got to change your approach. RollNo would never be (IMHO) a measure. It's an attribute of the student and so should be an attribute within that dimension. Then you *could* put both student name and roll# on (say) rows in a pivot table, even with *no* measure selected and see the roll# per student.
The reason why it's not a measure is that it provides no value when aggregated using any standard aggregation (ie sum of roll# == rubbish; avg of roll# == rubbish; *maybe* min roll# *might* mean something to someone, but i doubt it).
Using the data you've supplied, the only measure I see is Marks, and then, you'd likely want to set the aggregation to average, otherwise, when Kevin has two records in the fact table, and you're not filtered to a single subject, assuming he got 100 for both, then his Marks measure value would be 200 (using sum). Using avg, it would be 100. Of course, putting the courses dimension on the rows would let you see the avg mark per course (which, if there was only a single fact entry for each course, the avg would equal the actual mark attained).
HTH,
Steve.
December 18, 2010 at 12:11 pm
Hm... An excellent explanation will definitely try this now...thanks..
Viewing 3 posts - 1 through 2 (of 2 total)
You must be logged in to reply to this topic. Login to reply