"No fact table exists at the requested level of detail": means "Did you set the content levels appropriately?"
another way to do this is in Answers you can UNION the query that joins rows and a query that has dim only columns and a zero (0) for your fact columns ... skip the dummy table and the 1=1 join
Are you talking about UNION-ing reports from different subject areas? I tried it and after i UNION-ed reports, dimension values were duplicated. I didn't find a way to group by dimension values and sum up. Selection doesn't work on UNIONed reports too.
You could create a matrix with your dimensions (use a full alter join or just "no join") in physical layer
Note: OBIEE works with diemensional model so you must have at least 2 tables
Thank you for the reply. I actually created Star schema with inner joins. In addition to "inner join"-ed, there are so many reports that use outer join with every dimensions (the fact table is aggregated and all dimension values are visible even no related fact exists).
If the solution above works, it's very good and general solution for all subjects areas. By the standard way, it seems i should create 2 joins (one outer and one inner) for every fact<->dimension pairs. Am i right?