no, you can't
you must use an Integer source.
but you can create a dynamic DCount calculation if you have a dimension
which contains the members you want to count.
like:
count(crossjoin(dimension.level.members, {measures.[anymeasure]},
exclude_empty)
"NimsPatel" <NimsPatel (AT) discussions (DOT) microsoft.com> wrote
Quote:
I am trying to get a "distinct count" aggregate function specificed for a
measure that is non-numeric. Is this possible? When I use the cube editor,
the Distinct Count appears in the drop down list only when the measure is
numeric or datetime. If the measure is varchar, the Distinct Count option
is
not there. The measure is CustomerID, which is char data type.
BTW, I am using SQL server/Analysis Services version 2000.
Any ideas ? |