KoldCoffee (5/30/2013)
That worked. Thanks Phil. I was trying to accomplish before unpivot.In my dataset I sometimes have rows, where all 12 of the 12 rows that I am unpivoting (using unpivot transform) contain a NULL. I wanted to filter those out before the unpivot.
If I am understanding, then this could be done by passing the data through an Asynchronous Script Component Transformation in your Data Flow. This way, you would keep track of and construct the aggregated rows manually in .NET and if any of the 12 incoming rows failed to meet your criteria, you would not add a new aggregated row to the Output for that group of incoming rows. In effect you would be doing the work in you are currently doing with a Conditional Split plus an Unpivot inside of a single Script Component. More work, but it will probably perform better.
However, if I use conditional split after the unpivot, on the aggregated column, it excludes those rows I'm after. So for now I am fine:-)
I think what you have might be the simplest way to solve the problem case using the out of the box SSIS components.
There are no special teachers of virtue, because virtue is taught by the whole community.
--Plato