Creating new character variable based on two existing variables

[Sean_OConnor]

Folks,

I wonder could I get some help on the following. I have two existing character variables which are both ids. One is a product of method a while the other is a product of method b. 

I would like to create a third variable which takes the accurate matches from both. A simple table illustrates my point.

Also where we have a contrasting match this  output is placed in a second variable. 

ID1	ID2	Final id	Second Id
xxxx	5b	5b	-
6i	xxxx	6i	-
8c	5j	8c	5j
9g	xxxx	9g	-
10b	10b	10b	-

Any help would be great! 

[Reeza]

You have the 'xxx' in your dataset as indicated?

[Sean_OConnor]

Hi Reeza,

Yes all failed matches have a 'xxxx' inserted.

[Reeza]

Look at the IFC function. 

Or in Query builder, created a new column. You can use a CASE statement in your computed column. 

[Astounding]

It sounds like some simple logic would do.

if ID1=ID2 then Final_id = ID1;

else if ID1='xxxx' then Final_id = ID2;

else if ID2 = 'xxxx' then Final_id = ID1;

else do;

   Final_id = ID1;

   Second_id = ID2;

end;

Optionally, if you really want to, you could add:

if Second_id = ' ' then Second_id = '-';