Originally posted by: Ole Haavard Haug-Larsen

Hi.

I have a simple input-file with only two fields: Bankid (4 numeric) and Amount (17 numeric). Here is en example:

285000000000000017949
285000000000000166500
303000000000000012912
303000000000000006398
303000000000000004720
306000000000000001643
306000000000000006034
 

I want to create a new file where each record consist of only unique bankid's - and the sum of amount pr. bankid

How do I do that??

Originally posted by: Laurent(Satisco)

Hi,

you need to use Control Break Logic with component rules.

The attached sample show this on your data.

Attachment(s)

demo.zip   5 KB 1 version

Originally posted by: DennisRitche

Hey Laurent,

can you please let me know what exactly the mentioned component rule(Bank:$ = Bank:$[LAST]) do?

Thanks in Advance.

Originally posted by: Laurent(Satisco)

Please look at this: WTX documentation

Originally posted by: paul.brett

This is how I did it....

Input:

285000000000000017949
285000000000000166500
303000000000000012912
303000000000000006398
303000000000000004720
306000000000000001643
306000000000000006034

Output:

3 : 7

285000000000000184449
303000000000000024030
306000000000000007677

....using Control-Break logic on the input Typetree, and a simple SUM() function on output.

Thank-you.

Paul.

Follow me on Twitter

Attachment(s)

CtrlBreakSum.zip   3 KB 1 version

Originally posted by: Ole Haavard Haug-Larsen

Thanks a lot! It solved my problem.

Just one final question: is it possible to do this without changes in the Typetree? I have a similar case where I would prefer to not make changes in the Typetree (which are shared with other).

Originally posted by: paul.brett

Yes, it is possible with a little bit of messing about with EXTRACT().

Thank-you.

Paul.

Follow me on Twitter

Attachment(s)

simple.zip   4 KB 1 version

Originally posted by: Laurent(Satisco)

Agree with Paul, you can do with EXTRACT, but this can be a performances killer if the data is not very small.