Date:         Thu, 31 May 2001 16:59:48 -0400
Reply-To:     "Muhlbaier, Lawrence H." <lawrence.muhlbaier@DUKE.EDU>
Sender:       "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From:         "Muhlbaier, Lawrence H." <lawrence.muhlbaier@DUKE.EDU>
Organization: Duke Clinical Research Institute
Subject:      Re: merge replacement
Content-Type: text/plain; charset=us-ascii

Both Harry Droogendyk and James Brittain's suggestions have sorts implied in
them.  The SQL code does a sort in the work space as it goes and the INDEX
building process is in fact a tagsort with just the sorted variable and key
stored in the index.

I think that you are pretty much stuck.  You could write data step code to do a
binary search and link that way, but why (it passes the datasets multiple
times, probably more often than a sort would).  If you are worried about sort
disk space, using the tagsort option on the sort will minimize disk space.

Doc

wendy.watson@COVANCE.COM wrote:

> Hi SAS-L,
>
> I have two very large datasets and would like to attach variables from one
> dataset to the other without a sort and merge.
> Data 1 has the variables SSN, DRUG, etc.   Data 2 has the variables SSN and
> DATE OF BIRTH.  I would like to attach DATE OF BIRTH from Data 2 to Data 1
> (the merge variable would be SSN).  Data 1 contains one record for each SSN
> and  Data 2 has one record per SSN containing their Date of Birth.  Any
> ideas?
>
> Also, If Data 1 had multiple records for each SSN, how would I attach DOB
> from DATA 2 in this scenario?  Thanks.
>
> Data 1                                  Data 2
>
> SSN             DRUG                    SSN             DOB
>
> SSN02           X                       SSN03           12/04/1973
> SSN01           Y                       SSN02           09/18/1951
> SSN03           X                       SSN01           05/15/1987
> ...
>                         -OR-
>
> Data 1                                  Data 2
>
> SSN             DRUG                    SSN             DOB
>
> SSN02           X                       SSN03           12/04/1973
> SSN01           Y                       SSN02           09/18/1951
> SSN03           X                       SSN01           05/15/1987
> SSN02           Z
> SSN01           D
> SSN03           C
> ...
>
> -----------------------------------------------------
> Confidentiality Notice: This e-mail transmission
> may contain confidential or legally privileged
> information that is intended only for the individual
> or entity named in the e-mail address. If you are not
> the intended recipient, you are hereby notified that
> any disclosure, copying, distribution, or reliance
> upon the contents of this e-mail is strictly prohibited.
>
> If you have received this e-mail transmission in error,
> please reply to the sender, so that we can arrange
> for proper delivery, and then please delete the message
> from your inbox. Thank you.

--
Lawrence H. ('Doc') Muhlbaier           muhlb001@mc.duke.edu
Assistant Research Professor
Duke University Medical Center          919-668-8774 (office)
DUMC 3865                               919-383-0595 (home)
Durham, NC 27710-7510                   919-668-7057 (FAX)