topic Re: proc sort nodup in SAS Procedures

proc sort nodup

KevinC_ — Tue, 22 Sep 2009 14:10:36 GMT

Hello Everyone,

I am trying to get rid of records with duplicate key (the by value in proc sort). I tried different syntax and could never get it to work. The field ID1 is the key. I want to have unique ID1 in the outfile.

I have used:
PROC SORT; BY ID1; NODUP;
PROC SORT; BY ID1 NODUP;

none of these work. Does anyone have any suggestion?

Thank you so much!

Re: proc sort nodup

Flip — Tue, 22 Sep 2009 14:13:20 GMT

PROC SORT; BY ID1; NODUPkey;

Re: proc sort nodup

data_null__ — Tue, 22 Sep 2009 14:20:50 GMT

actually NODUPKEY is a PROC statement option.

[pre]
proc sort data=sashelp.class out=sex nodupkey;
by sex;
run;
[/pre]

Re: proc sort nodup

KevinC_ — Tue, 22 Sep 2009 14:22:10 GMT

Thank you Flip.

I tried it with your exact syntax and got the following error:
"Statement is not valid or it is used out of proper order."

Any idea why? I am using sas version 8.2.

Thanks!

Re: proc sort nodup

KevinC_ — Tue, 22 Sep 2009 14:26:11 GMT

DATA_NULL,

That worked. Thank you so much!!
:)

Re: proc sort nodup

Flip — Tue, 22 Sep 2009 14:28:19 GMT

Oops, yup

Re: proc sort nodup

DanielSantos — Tue, 22 Sep 2009 15:07:07 GMT

Very handy adresses...

Everything you wanted to know about SAS procedures:
http://support.sas.com/documentation/cdl/en/proc/61895/HTML/default/procwhatsnew902.htm#

And off course, the SAS Online documentation at:
http://support.sas.com/documentation/index.html

Cheers from Portugal.

Daniel Santos @ www.cgd.pt

Re: proc sort nodup

sbb — Tue, 22 Sep 2009 18:08:58 GMT

Also, remember, for NODUP to work, the duplicate observations must be adjacent, which implicitly calls for a reasonably granular BY variable list when using the function. And, PROC SORT yields no mercy if you forget and code a BY variable list that will not achieve the required condition in order for the sort package to do its job removing duplicates.

Scott Barry
SBBWorks, Inc.