|Date: ||Fri, 10 Dec 1999 13:50:46 -0500|
|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: SAS V5 dataset question|
|Content-Type: ||text/plain; charset=us-ascii|
Boy are you exercising old synapses....
SAS 5.18 sequential was called 'TAPE' format in the documentation. I don't
know that there was information on how to read with other software in the
documentation. My recollection is that the TAPE format actually stored the
data in a fixed-block orientation, so you might be able to unscramble some of
it. If your colleague has a CONTENTS printout of the dataset, you are in much
My old experience with MVS was that if the originator just copied (with MVS
utilities) a SAS (default) dataset to another place, all of the binary, direct
access, pointers were toast, as the pointers were to hardware sectors. CMS,
with it's mini-disks, may be different.
There are still some V5 systems on mainframes; perhaps a user there can shed
more definitive light.
"F.J. Kelley" wrote:
> I think I know the answer to this one, but one never knows what will turn
> up ...
> A faculty member was given a file which (supposedly) had some data he
> needed on it. The file was sent from another institution and their folks
> had been unable to do anything with it.
> It appears to be a SAS V5 file, probably in sequential format. I can't be
> sure, but I believe it was written on a CMS system. When the original
> institution shut down their mainframe, the data was copied (they _did_
> preserve the binary attributes). Of course there is no end-of-record, so
> a binary upload to a mainframe results in a single line. vi doesn't much
> care for it either, however the Cygnus hex editor on my NT box seems to
> have no problems. And the file appears to be EBCDIC hex. I've looked
> over the old "SAS Programmer's Guide (V5)" but can't seem to find info re
> the structure of a sequential file (I think the SAS 82 version had that,
> but it _has_ been a while). My feeling is that the data may well be
> there, but the file is trash. Any thoughts? --Joe
> (oh yes, I believe this was written with V5 SAS, based on the "85.18"
> string in the first 80 chars)
Lawrence H. ('Doc') Muhlbaier firstname.lastname@example.org
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)