Hai

working in DB2-environment with Datamigrator GUI.
This is my field in input table:
FIELDNAME=ACTU_DAT, ALIAS=ACTU_DAT, USAGE=A8, ACTUAL=A8
An example is '20190105'

It has to be joined with this field in another table:
FIELDNAME=DAT, ALIAS=DAT, USAGE=YYMD, ACTUAL=DATE
Example is '2019-01-05'

In the source transforms calculator I putted this transformation
DATECVT(ACTU_DAT, 'A8', 'YYMD')

But the inner join gives no result.

Where does it go wrong?

Regards
Ron

This message has been edited. Last edited by: FP Mod Chuck,

Personally I would do the opposite : convert the date into alpha (A8) and then join the two alpha.
Or convert both into integer date (I8YYMD) or alpha date (A8YYMD) and then join them.

You need to be sure that both field have the same format and content.
You problem is probably there where one field doesn't have the exact same format and content then the other.

Have you print them individually to compare and see if they have the same representation ?

Hoi Martin

you give three solutions:
- convert the date into alpha (A8) and then join the two alpha
- convert both into integer date (I8YYMD) and then join them or
- convert both into alpha date (A8YYMD) and then join them.

Can you , as a learning option for me, show the three solutions?

Hi,
Why dont you change the Usage from A8 to YYMD in the master file rather than converting it.
Is there are reaons you want to keep format as A8 in the master file?

Master file is used in multiple dataflows. I don't know what the effect will be in the other flows by changing the master.

quote:

Master file is used in multiple dataflows

You can add a DEFINE in the synonym and use that without affecting other users...

quote:

Originally posted by Ronibi:
Hi Martin

you give three solutions:
- convert the date into alpha (A8) and then join the two alpha
- convert both into integer date (I8YYMD) and then join them or
- convert both into alpha date (A8YYMD) and then join them.

Can you , as a learning option for me, show the three solutions?

They are just format conversion and may or may not apply to your situation and/or input format.

You can use DEFINE and apply a DATECVT function to move from A8 to I8YYMD par example or to any other specified format
The important thing is that at the end you are comparing two fields with the same format and content representation.

You should first look at individual field content after transformation and ensure that they can be compared.

It can have plenty of solutions, you need to find the proper one for you

quote:

quote:
Master file is used in multiple dataflows

You can add a DEFINE in the synonym and use that without affecting other users...

FYI

With Relation DB's, you can have the same field multiple times.

I your case you could have:
FIELDNAME=ACTU_DAT, ALIAS=ACTU_DAT, USAGE=A8, ACTUAL=A8
FIELDNAME=ACTU_YYMD, ALIAS=ACTU_DAT, USAGE=YYMD, ACTUAL=A8


BUT
What is the format of the column in DB2 ?

Thnx. The problem in the last answer is that we always do a recreate of a master, not a refresh.
So when I add something in the master, it will be lost once it is recreated because of some other change.

As I said before I already tried DATECVT

In the source transforms calculator I putted this transformation
DATECVT(ACTU_DAT, 'A8', 'YYMD')

But the inner join gives no result

quote:

we always do a recreate of a master

Why do you always do a recreate of a master instead of just refreshing / updating the data?

Does anything change in the master, like fields being added or deleted or format changes?