count the rows

arshi · Post by **arshi** » Thu Jun 12, 2008 9:36 am

Hi,

I have the requirement like,

INPUT

COL1

ABC
ABC
XYZ
ABC
XYZ

I want the OUTPUT as,

COL2 COL3
ABC 3
XYZ 2

Can you tell how to implement in datastage.

Regards,
Arshi

gateleys · Post by **gateleys** » Thu Jun 12, 2008 9:43 am

Aggregate based on Col1 and count.

Minhajuddin · Post by **Minhajuddin** » Thu Jun 12, 2008 11:38 am

If you have more than one column and you want other columns to be intact, you need to split it using a copy stage aggregate it on COL1 and then join it back with the original stream on COL1

DSguru2B · Post by **DSguru2B** » Thu Jun 12, 2008 11:42 am

If your source is a database you can select that column and a count(*) next to it, with the first column in the group by clause.

ag_ram · Post by **ag_ram** » Thu Jun 12, 2008 11:47 am

Minhajuddin wrote:If you have more than one column and you want other columns to be intact, you need to split it using a copy stage aggregate it on COL1 and then join it back with the original stream on COL1

Didn't see that Job type is Server, Minhajuddin.

pranay · Post by **pranay** » Thu Jun 12, 2008 11:57 am

use the aggregator stage with columns COL1 and count(COL1).

Minhajuddin · Post by **Minhajuddin** » Fri Jun 13, 2008 10:43 am

My apologies for giving the wrong solution

DSXchange

count the rows

count the rows

Re: count the rows