Convert SAS v 6.12 VAX files to windows

[rlmacdonald]

I have a clinical trial data from the 1980s that is in SAS v6.12 and was created on the old DEC VAX system (DEC model 4000 MODEL 610) and I want to convert it to be able to be read in Windows. SAS technical support said they don't have any versions of SAS like this anymore. Any suggestions? Thanks.

RL Macdonald, MD

Department of Neurological Surgery

UCSF Fresno

[TomKari]

1. Is it actually a SAS dataset, or is it a dataset in some other format that you were using in SAS?

2. Do you now have the file physically on your PC?

3. Do you know anything about the data contained in the file?

Tom

[rlmacdonald]

Tom

Yes. I have all of the available files on my PC sent to me from my collaborator, including the data. The files do not open in the SAS free online viewer. I can open them in the text reader and can see some description of what the file is at the beginning. Below that info, the rest of most of the files is a series of various irrelevant characters. I am sure that one file called rec123.ssd is the full data from the study which is 3521 patients with 287 variables per patient because the top info lists that and then I can see the variable names all appearing throughout the file in the txt reader. There are a bunch of .sfn files that show the variable names and how they are coded.

Here is what one of the files called rec123.log says:

1 The SAS System 21:28 Friday, May 15, 1998

NOTE: Copyright (c) 1989-1996 by SAS Institute Inc., Cary, NC, USA.
NOTE: SAS (r) Proprietary Software Release 6.12 TS020
Licensed to UNIVERSITY OF IOWA, Site 0001386002.

NOTE: Running on DEC Model 4000 MODEL 610 Serial Number 80000000.

Welcome to the new SAS Information Delivery System, Release 6.12.

Preventive Medicine - State Health Registry
University of Iowa - Iowa City, Iowa

1 options nocenter nofmterr;
2 libname castos '[torner.castos.data]';
NOTE: Libref CASTOS was successfully assigned as follows:
Engine: V5
Physical Name: PMSYS$DKA100:[TORNER.CASTOS.DATA]
3
4 data rec123; set castos.rec123;

Then there is a list of Note: Format CENTER is unknown. and it says this for basically all 250+ variables. and ends with:

NOTE: Format FCDO is unknown.
NOTE: The data set WORK.REC123 has 3521 observations and 287 variables.

5 proc contents;
6 run;

NOTE: The PROCEDURE CONTENTS printed pages 1-6.

7

NOTE: SAS Institute Inc., SAS Campus Drive, Cary, NC USA 27513-2414

This is the header info of rec123.ssd:

d   
ˆ 
|™ ø Á DISK 
d tô{ SAS REC123 DATA 5.18 VMS A BOOK! WHISPER 09APR89:19:49:3309APR89:19:49:33DATA - followed by the variable names interspersed with improperly converted characters and symbols.

Thanks.

R. Loch Macdonald

[Tom]

So you have a SAS VMS version 5.18 dataset, not a version 6.12 dataset. Did they provide any listings of the data (.lst or .lis or .listing files)? If you have a complete listing of the data you might be able to re-create the dataset from the text in the listing.

[rlmacdonald]

The .lis file says:

Created: 21:28 Friday, May 15, 1998 Observation Length: 2296
Last Modified: 21:28 Friday, May 15, 1998 Deleted Observations: 0
Protection: Compressed: NO
Data Set Type: Sorted: NO
Label:

-----Engine/Host Dependent Information-----

Data Set Page Size: 32256
Number of Data Set Pages: 253
File Format: 607
First Data Page: 2
Max Obs per Page: 14
Obs in First Data Page: 12
Filename: PMSYS$DKA100:[TORNER.CASTOS.SAS$WORK000095A5]REC123.SASEB$DATA
Host Format: AXP
Disk Blocks Allocated: 15940

-----Alphabetic List of Variables and Attributes-----

# Variable Type Len Pos Format Label
------------------------------------------------------------------------------------------------
4 AGE Num 8 24
31 AMNHA Num 8 240 AMNHA7. MENINGEAL SIGNS - HEADACHE

then lists this info for every one of the variables.

I also have files that seem to define the variables eg.

AMNHA.sfn:

d 
X  UTILX

d <SAS AMNHA SFN 5.16 VMS T_GERMANPROC SASUTIL :
AMNHA TOùö ´ 5.16

   Œ,̼o ‰P 9 € €
9 €@ €@ À€ À€ <   ABSENT <   PRESENT

This says they are SAS 5.16 files, the database rec123.ssd looks like the ssd is short for SASEB$DATA

The variable AMNHA is coded as absent or present but one can't tell what the coding system is (0, 1 or something) and it can't be seen in the rec123.ssd

[JackHamilton]

Does the file have an extension of saseb$data?  

If so, there used to be a Windows program named DBMSCOPY.  It could read unencrypted files from the VAX.  You might be able to find someone with a working copy.  https://dss.princeton.edu/online_help/stats_packages/dbmscopy/dbms_stat.htm 

In fact, someone at SAS Institute should have a working copy.  DBMSCOPY was made by a very small company, and SAS made the owner an offer he couldn't refuse.  He went on to fulfill his thespian dreams, and SAS discontinued his program (it was competition for SAS/Access).  But SAS owned it, and someone there should have a copy.  

I assume that you've tried just reading the data sets from SAS, and it didn't work.  The documentation implies that OpenVMS files are readable by SAS (possibly requiring the outrep=vax_vms libname option), but I don't know how different the OpenVMS format is from the VAX format.

What's the urgency on this? 

[rlmacdonald]

Thanks. Not urgent. The database is called rec123.ssd. The .lis file that describes it says file extension is saseb$data so maybe it was abbreviated to ssd. I'll contact SAS tech support again and see if I can get a copy of dbmscopy

[JackHamilton]

Good luck!  If SAS can't/won't help, there might be someone on one of the user mailing lists that still has a copy.  I don't remember the software had an expiration date or was permanently licensed.