IBM Mainframe Forum

I didn't get you.If possible, can u explain it.
There is no need of splitting. I split the records because at a time i can arrange either MAX-NMBR in descending order or MIN-NMBR in ascending order. Thats why i used two steps of sort.
I tried using OUTFIL option so that single step will achieve two things, but OUTFIL does not support DUPKEYS keyword.
In that old post, 3nadh suggested the merge solution, but with that at time only one field will be arranged in either max or min order.

Exactly its not a client requirement, I am doing this to reduce the CPU time used by a JOB.

3nadh suggested a SORT solution. I wondered about the possibility of MERGE, as you file was in order already.

Are you using SyncSort now or DFSORT?

If SyncSort, try specifying both MIN and MAX on different fields of the DUPKEYS.

If DFSORT, forget DUPKEYS, and use OUTFIL with SECTIONS which have their own MIN and MAX. You were given that code twice elsewhere.

I m using SYNCSORT. The file could be in any order.
I tried the below mentioned code and got the error message. Thats why i used two sort steps.

Can you try DUPKEYS MIN=(03,03,ZD),MAX=(06,03,ZD)

If you get an error message, post the sysout from the step.

Ya it worked.
Thanks Billy.
Don't need to put comma. Correct code is mentioned below.


Now one more thing i need to do.

There is one other file which has the layout mentioned below.



The file which we created from the above mentioned merge step has three things LIC-NMBR, MIN-NMBR and MAX-NMBR.
These two files has to be compared.


From the Merge step we got the output(mentioned below).
File-1
11001005
12000007
13000001

The second file has the data
File-2
11003
11002
11004
12001
12002
13001

it could be in ascending or descending order.
We have to compare these two files on the basis of Field NMBR, this field should be in between MIN-NMBR & MAX-NMBR of file 1 for the particular LIC.If it doennot belong to then it has to be written in output file.
For example
11001005
In this NMBR is 005.
In file-2
11003
11002
11004
For the LIC (11) Min-NMBR is 001 and MAX-NMBR IS 005. So above given number (003,002,004) are in between 001 & 005.They will not be written to the file. But if there would have been one more record like 11007 then it has to be written to the error file as 11007007.
How could it be achieved??

If you use JOINKEYS on the LIC, then for matched records:



You will have each of multiple key on File 2 matched to the single record on File 1.

You can then use OUTFIL OMIT/INCLUDE to output 9,5 from the REFORMAT record, if number 11,3 is less than 3,3 or greater than 6,3.

Getting some error in joinkey statement....

Hello,

And you intend to keep this all to yourself . . .

I was little bit busy.
There is the complete code.Its working now.


Now i need one include statement. So that i can compare the number with maximum and minimum number.
I will update soon.

I included omit condition instead of include.


I want to reduce the length of Output dataset to 8, now its 11.
How could it be achieved? Any Suggestions?