Variable field calculation

js_j · Post by **js_j** » Mon Aug 22, 2011 1:12 am

Hi,
The following is my file structure:
A B C D
2001;40S0;17.50;123456;
2001;500A;02.00;123456;
2001;50S0;16.50;123556; etc...

the required output is for every value of B in (40S0 and 500A) the corresponding C column values should be added for the same/different D column value.

In the Output i have an extra column YYYY to which the above calculation is mapped to. Input is dataset-->transformer-->dataset.

I am trying to use the transformer for the above. I am also open to using aggregator stage for the same.

Kindly let me know.

thanks,
JJ

ray.wurlod · Post by **ray.wurlod** » Mon Aug 22, 2011 2:27 am

This requires a classic fork-join design (though, once you're on version 8.5, you may be able to use looping constructs in the Transformer stage).
One side of the fork (Copy stage) leads to an Aggregator stage that performs the required calculation. The two streams are then united using a Join stage.

js_j · Post by **js_j** » Mon Aug 22, 2011 2:46 am

ray.wurlod wrote:This requires a classic fork-join design (though, once you're on version 8.5, you may be able to use looping constructs in the Transformer stage).
One side of the fork (Copy stage) leads to an Aggreg ...

Ok. But how do i implement the same in 7.5 version?

priyadarshikunal · Post by **priyadarshikunal** » Mon Aug 22, 2011 3:38 am

js_j wrote: Ok. But how do i implement the same in 7.5 version?

Although I can't see the post from Ray (waiting for membership director's action), I can see Ray explaining classic fork-join using copy aggregator and join... I think all you have to do is read the message completely.. Or search for fork join to get some examples to move forward.

js_j · Post by **js_j** » Mon Aug 22, 2011 4:11 am

will try and get back.

thanks,
JJ