Extract new ID from records with one hyphen

[lalohg]

Hi there

I have a data set with a variable "ID" shown below and would like to create a new ID variable as shown below in column "New ID" (need to get rid of the hyphen and numbers after the hyphen). Please notice that the ID changes ONLY when ONE hyphen is present in variable "ID".

ID	New ID
1170192-01	1170192
1171775-01	1171775
1175345-01	1175345
12-06-0219	12-06-0219
12-06-0223	12-06-0223
12-06-0235	12-06-0235
12-06-0236	12-06-0236
12-06-0263	12-06-0263
13-06-0291	13-06-0291
13-06-0294	13-06-0294
13-06-0321	13-06-0321
14-06-0347	14-06-0347
14-06-0351	14-06-0351
14-06-0597	14-06-0597
15-06-0116	15-06-0116
15-06-0353	15-06-0353
15-06-0365	15-06-0365
15-06-0367	15-06-0367
15-06-0369	15-06-0369
15-06-0371	15-06-0371
99-06-9921	99-06-9921
SP0100470-	SP0100470
SP080004-0	SP080004
SP080005-0	SP080005

All help will be appreciated it

Thank you

[Steelers_In_DC · Accepted solution]

Here you go:

DATA have;

infile cards dsd;

informat id $10.;

format id $10.;

input id;

cards;

1170192-01

1171775-01

1175345-01

12-06-0219

12-06-0223

12-06-0235

12-06-0236

12-06-0263

13-06-0291

13-06-0294

13-06-0321

14-06-0347

14-06-0351

14-06-0597

15-06-0116

15-06-0353

15-06-0365

15-06-0367

15-06-0369

15-06-0371

99-06-9921

SP0100470-

SP080004-0

SP080005-0

;

run;

data want;*(rename=(new_id=id));

set have;

count = countc(id,'-');

if count = 1 then do;

New_ID = scan(id,1,'-');

end;

if count ne 1 then do;

New_ID = ID;

end;

/*drop id count;*/

run;

After you run it and see what the dataset looks like uncomment the few things that are commented and see if that is the desired output.

[lalohg]

THANK YOU VERY MUCH MARK!!

[Astounding]

This is a good set of tools to use here, but you might want to simplify things a bit:

data want;

set have;

if countc(id, '-') = 1 then id = scan(id, 1, '-');

run;

[lalohg]

Thanks again Mark,

how should I modify the code if I want to keep in one separated column the hyphen and numbers?, this column for the fist three cases will have -01

Eduardo.

[Steelers_In_DC]

Here's what you would want:

data want;

set have;

count = countc(id,'-');

if count = 1 then do;

New_ID = scan(id,1,'-');

end;

if count ne 1 then do;

New_ID = ID;

end;

drop count;

run;

The changes I made, remove the rename statement from the first line, we won't be renaming anything, and I'm only dropping the count at the bottom, I used that to identify what has more than one hyphens.