DSXchange

Hi,

I am doing a simple lookup:

Database
v
v
Dataset>>>>Lookup>>>>>Dataset


In the Database i am having the data as

ID POS DEPT
1 QA FIN
2 QA FIN
3 QA FIN
1 BA DIN
3 BA DIN

and the source is having data like

ID Name1 Name2
1 A AA
2 B BB
3 C CC


The key is ID for lookup.I want only 3 rows in the output like

ID POS1 POS2 Name1
1 QA BA A
2 QA B
3 QA BA C

ANY HELP IS APPRECIATED

THANKS
SAM

There is no common key on which a lookup can be based. Please describe the algorithm you would use to generate the output if DataStage were not available.

ray.wurlod wrote:There is no common key on which a lookup can be based. Please describe the algorithm you would use to generate the output if DataStage were not available.

Hi Ray,

I can do this thing by doing the lookup based on ID and allowing "Multiple rows" through the lookup link.

ID POS Name1
1 QA A
1 BA A
2 QA B
3 QA C
3 BA C

After this i am getting stuck on how to convert this multiple rows into a single row.

Any Help is appreciated
Thanks
sam

This technique is called a vertical pivot - a search will reveal a few approaches that people have taken to it. It's tricky to do in parallel jobs and retain parallelism.

Hi

Can somebody tell me what this Error means.
Thanks
sam

After doing a search i found that by clearing the preserve partion flag this warning is handled.But can anyone help me in understanding why this happened?

Thanks
sam

Hi Sam,

By default the previous stages would have a setting of propogating the existing partition to the following stages. But you would have defined a new partitioning for the following stage.

So DataStage gives a warning stating that the existing partiton cannot be preserved and rather the new partition will be used for that particular stage.

HTH
--Rich