Date: Fri, 20 Apr 2001 10:24:27 -0600
Reply-To: Darran Zenger <DZENGER@KLUNE.COM>
Sender: "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From: Darran Zenger <DZENGER@KLUNE.COM>
Subject: Re: Finding Differences in 2 Datasets
Content-Type: text/plain; charset="iso-8859-1"
Maybe I'll clarify what I'm attempting to do. Some have suggested using PROC
COMPARE, which for me, I'm looking through the SAS Help to learn about it.
What I've got are 2 datasets pulled from a Vantage DB using Report Builder
(since we are still working on ODBC connecting SAS to the DB) so I'm pulling
data out. How can I put this...hmm...wish there was some teleconferencing
here! It's rather hard to explain, but it all boils down to this. I've got
two datasets that I need to find out why one is summing up dollar values
more than the other, given the same filtering requirements in Report Builder
from two different tables. Now having said this, one data set has more
variables than the other (non essential variables, like account numbers,
entry person). But when running the same code on both data sets, my HBAR
should all be the same regardless of which dataset I use, but yet, they are
off by several thousand dollars (roughly $16K). That's a big difference for
us so need to find out why this is occurring.
Make sense?
Darran Zenger
The SAS System
Programming
Klune Industries, INC
Spanish Fork Division
dzenger@klune.com
-----Original Message-----
From: Charles Patridge [mailto:Charles_S_Patridge@PRODIGY.NET]
Sent: Friday, April 20, 2001 10:22 AM
To: SAS-L@LISTSERV.UGA.EDU; Darran Zenger
Subject: Re: Finding Differences in 2 Datasets
Darren,
Not sure of your exact requirements but you might try using PROC COMPARE for
finding differences. I use it and it works very well when comparing two SAS
Datasets for minor and major differences. You will have to set your "BY"
values to meet your needs.
If record position is a problem then COMPARE will not work for you in this
situation unless you sort both files first. In addition, it is assumed both
files have the same number of records to do the comparison.
HTH,
Charles Patridge
Email: Charles_S_Patridge@prodigy.net
| 
