Date:         Mon, 5 Oct 2009 01:02:54 -0700
Reply-To:     Savian <savian.net@GMAIL.COM>
Sender:       "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From:         Savian <savian.net@GMAIL.COM>
Organization: http://groups.google.com
Subject:      Re: Merging Very Large Datasets
Comments: To: sas-l@uga.edu
Content-Type: text/plain; charset=ISO-8859-1

On Oct 4, 5:32 pm, Claus Yeh <phoebe.caulfiel...@gmail.com> wrote:
> Dear SAS guru's,
>
> I have two very large SAS datasets.  Each has 100000 variables and
> 5000 observations. Is there an efficient way to merge these two
> datasets without taking hours?
>
> thank you so much,
> claus

Why do you possibly have 100K variables? That is a major issue.
Systems are designed for efficiencies on number of records. You are
probably filling up your memory footprint and having to page out. You
would need to run performance tools to see what is happening but I
wouldn't be surprised to see thrashing happening.

Alan

Savian, LLC