Date: Tue, 15 Nov 2005 10:37:49 -0500
Reply-To: "Richard A. DeVenezia" <radevenz@IX.NETCOM.COM>
Sender: "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From: "Richard A. DeVenezia" <radevenz@IX.NETCOM.COM>
Subject: Re: Base vs. SQL
data _null_; wrote:
> Perhaps a simple interleave would work.
> data work.dupids2;
> set _1(in=in1) _2(in=in2) _3(in=in3) _4(in=in4) _5(in=in5);
> by id;
> if not (first.id and last.id);
Too aggressive filtering, it excludes single occurring ids potentially
duplicated in a different table.
> source2 = in1*1e4 + in2*1e3 + in3*1e3 + in4*1e1 + in5*1e0;
Nice! in3*1e2 though. Also, if true integer types were supported, the
source state could be indicated by a bit instead of 'power of ten' bit.