Transposing rows to columns

[chulett]

Excellent points. I tend to forget about the possibility of doing something after stage, the before/after job ones are so much more... visible. I should have mentioned the fact that you can land data and pick it back up again in the same job in the manner you illustrate. Heck, I just did something very much like this in a job today, as a matter of fact. Doh! [smacks forehead]

And thanks for the clarifications on the lookups. I didn't think too much about the inner/outer/left/right-ness of it all, just started typing. Still helps illustrate my point - they basically are (or can be) whatever you make them. You decide what gets 'rejected' by the lookup, not DataStage.

Annotations are something no job should be without. I heartily second Ray's suggestion to document the presence of anything not immediately obvious about a job.

[siddharthkaushik]

Craig,
I did get your point earlier and as Ray pointed out, it is the constraint that is making the join inner.

After some tweeking of the constraints I was able to get the row through and the transpose worked perfectly alright.

But the question, the curiosity and the interest still remains .... Is it possible to add a dummy row in a Transformer Stage or rather is it possible to synthesize a dummy row in a trasformer and pass it to the next stage in the end?

Thank you guys for all the help. There is a long weekend coming up and I plan to put my head into this problem then. If I am able to do it, I will definitely post it here.

Thanks,
Siddharth