Hi Randi,
analytic views are meant to serve as 'cube'-constructs, not as fancy SQL tables.
While you could perform a comparison between two strings (e.g. byte by byte) and thereby define a MAX or MIN, the data warehouse cube don't know how to compare string measures.
Therefore, only date and numeric values can be aggregated by more than CNT().
The question here is of course: what exactly is your requirement here?
If this request is about e.g. replacing NULL values with one value that occurs in the same group of data ('flattening e.g. the result of a a UNION'), then you can build this easily in a calculation view or via SQL.
- Lars