Is it possible to create a joint data set that doesn't have all the fields of its input data set

mstaniloiu

I want to create a joint data set from two data sets each having 6 columns. 4 of them are the same, and they have to be equal to eachother as a condition for the join.
I only want to keep each of the common columns once in the joint data set (there's no use of having so much redundant information) but I see no way of doing it (the Edit Data Set -> Output Columns panel automatically has all the columns from both data sets, and doesn't have any way of removing them).

Also - is there any way of using an existing data set as a data source for creating a new data set? It would be really useful, as besides joining the two data sets I also have to do some computations on the data (and the Edit Data Set dialog doesn't allow me to write an sql query for it).

mwilliams

I don't think there's a way to disclude the duplicated columns by doing the join in BIRT. You may be able to do this in your SQL, but I'm not certain on this. As for the computations you need to do, do computed columns not work for you?