Date:         Wed, 26 Feb 1997 09:42:18 -0600
Reply-To:     "Alan M. Sherkow" <alsherko@EXECPC.COM>
Sender:       "SAS(r) Discussion" <SAS-L@UGA.CC.UGA.EDU>
From:         "Alan M. Sherkow" <alsherko@EXECPC.COM>
Subject:      Re: Losing Duplicates Inconsistently (PROC SORT)
Content-Type: text/plain; charset="us-ascii"

Dan -- No dups works by verifying that the record about to be output is not
the same, field for field, as the last record output. Your 'by statements'
are not specific enough. To really make sure this works, the most complete
technique is to use all fields in the dataset. In your case this would be
"by copid offname offnum agency".
 
You know your data, and I don't. Your example suggests that 'by copid
agency' may be enough of a key.
 
Al
+---------------------------------+
| Alan M. Sherkow                 |
| I/S Management Strategies, Ltd. |
| 4942 N. Hollywood Avenue        |
| Whitefish Bay, WI   53217-5935  |
| Voice: (414)332-3062            |
| Fax:   (414)332-8771            |
+---------------------------------+