Assign the same value to identical ids

[Sean_OConnor]

Folks,

I've a bit of a query which maybe people can clear up. I have two string variables one being an id and the other being a location. What I would like to do is if two ids are the same then geographic location gets written to all identical ids.

So from something like this;

ID	County
41A55F2DD6000000
41A55F2DD6000000	KILKENNY
41A7508F18000000
41A7508F18000000
41CA171A0A000000
41CA171A0A000000	DUBLIN
41CD59FD4B000000	GALWAY
41CD59FD4B000000	GALWAY
41D0A741A4400000	KILDARE
41D0A741A4400000

To this;

ID	County
41A55F2DD6000000	KILKENNY
41A55F2DD6000000	KILKENNY
41A7508F18000000
41A7508F18000000
41CA171A0A000000	DUBLIN
41CA171A0A000000	DUBLIN
41CD59FD4B000000	GALWAY
41CD59FD4B000000	GALWAY
41D0A741A4400000	KILDARE
41D0A741A4400000	KILDARE

Any help is appreicated. 

[ballardw]

Do you ever have a case in your data where an ID has two different non-missing values for the county/ geography?

If not:

data have;
   infile datalines missover;
   informat id $17. county $10.;
   input ID County ;
datalines;
41A55F2DD6000000   
41A55F2DD6000000 KILKENNY 
41A7508F18000000   
41A7508F18000000   
41CA171A0A000000   
41CA171A0A000000 DUBLIN 
41CD59FD4B000000 GALWAY 
41CD59FD4B000000 GALWAY 
41D0A741A4400000 KILDARE 
41D0A741A4400000   
;
run;

proc sql;
   create table want as
   select a.id, b.county
   from have as a left join
        (select distinct id,county from have 
         where not missing(county)) as b
       on a.id=b.id;
quit;

The proc sql part is the important one, the data step is just to have something to test the code with.

[Jagadishkatam]

Please check the new_county variable with the expected output

data have;
infile cards missover;
input ID$20.	County$15.;
cards;
41A55F2DD6000000	 
41A55F2DD6000000	KILKENNY
41A7508F18000000	 
41A7508F18000000	 
41CA171A0A000000	 
41CA171A0A000000	DUBLIN
41CD59FD4B000000	GALWAY
41CD59FD4B000000	GALWAY
41D0A741A4400000	KILDARE
41D0A741A4400000	      
;

proc sort data=have;
by id descending County;
run;

data want;
set have;
by id descending County;
retain new_county;
if first.id then new_county=county;
run;

[Ksharp]

data have;
   infile datalines missover;
   informat id $17. county $10.;
   input ID County ;
datalines;
41A55F2DD6000000   
41A55F2DD6000000 KILKENNY 
41A7508F18000000   
41A7508F18000000   
41CA171A0A000000   
41CA171A0A000000 DUBLIN 
41CD59FD4B000000 GALWAY 
41CD59FD4B000000 GALWAY 
41D0A741A4400000 KILDARE 
41D0A741A4400000   
;
run;
data want;
 merge have(keep=id) have(where=(county is not missing));
 by id;
run;