Date: Thu, 1 Oct 1998 14:32:35 -0400
Reply-To: Steve Giblin <Steve_Giblin@VAPOWER.COM>
Sender: "SAS(r) Discussion" <SAS-L@UGA.CC.UGA.EDU>
From: Steve Giblin <Steve_Giblin@VAPOWER.COM>
Subject: Re: No Subject
Content-type: text/plain; charset=us-ascii
I'm making an assumption here, given lack of details, that the merge/purge
that you are looking for is in the are of name and addresses, and/or
these in some type of customer/prospect/householding/bundling or some type
routine update process.
If this assumption is correct, then there really are two parts to your
1, For names and addresses, you will need some type of standardization
2. Then, once things are standardized, you need a match/merge/purge
Indeed, the expensive packages like POSTALSOFT will perform both of these
If though, you have condition #1 satisfied ( everthing standardized ), then
you could really
perform the match/merge/purge that you are looking via a SAS job ( albeit
many lines of
code to such a SAS job ).
I have electronically generated lists for direct mail over the years (
several million pieces
worth ) and have used SAS exclusively to accomplish this. Let me know if
this is an option
worthy of further pursuit.
I have seen other merge/purge mechanisms using C++ that operates on the
where there is more direct control over the parallel processors and what
can be optimally
and simultaneously processed.