SAS
SAS is a software suite for statistical analysis, etc.[1]
Picotte
SAS 9.4 is installed on Picotte. To use, load this modulefile:
sas/9.4
The installation does not include SAS Studio or Simulation Studio.
Running SAS
SAS is GUI (graphical user interface) application by default. This means that SAS needs to be able to display its application windows on your PC. For information on the software you will need on your PC to do this, see:
It can also be run non-interactively, a.k.a. batch mode. This is how compute jobs would be submitted.
The command to run SAS (EN) is:
sas
Or one of these, based on language:
sas_en- Englishsas_u8- UTF-8 Unicode; NB the SAS Display Manager System does not fully support UTF-8sas_es- Spanish (Castillian)sas_zh- Chinese (Simplified)sas_zt- Chinese (Traditional)sas_zt.euc- Chinese (Trad. - EUC encoding)
Interactive, On the Login Node
Please do this only for brief sessions. When SAS is run on the login node, it will be sharing the resources (CPU cores, memory) of that node with others who may be doing work on it. Thus, performance may not be the best. And if SAS uses a lot of resources, it may affect the other users on the login node. See the next section for details on requesting a compute node for exclusive use.
To run SAS on picottelogin:
[juser@picotte001 ~]$ module load sas/9.4
[juser@picotte001 ~]$ sas_en
this results in the multiple-window SAS application:
To quit SAS, in the "SAS: Explorer" window, use the menus to select File → Exit…
Interactive, On a Compute Node
This is the preferred way to run SAS interactively. This guarantees exclusive access to all resources on a compute node to do your work, rather than sharing resources with other logged-in users. Since there are 74 (standard) compute nodes in Picotte, chances are high that the wait time for free resources is low.
See: Running GUI Applications on Compute Nodes
Non-Interactive (Batch)
Batch mode is the best way to exploit the resources that Picotte offers to perform large amounts of computation work.
- Use example code by P. MacDonald at McMaster University: https://ms.mcmaster.ca/peter/s4p03/s4p03_0304/sasnotes.htm
- This is the SAS program -- filename
lranova.sas:
data crack;
input id age load agef;
datalines;
1 20 11.45 20
2 20 10.42 20
3 20 11.14 20
4 25 10.84 25
5 25 11.17 25
6 25 10.54 25
7 31 9.47 31
8 31 9.19 31
9 31 9.54 31
;
proc glm data=crack;
class agef;
model load = age agef / p;
output out=crackreg p=pred r=resid;
run;
proc plot data=crackreg;
plot load*age="*" pred*age="+"/ overlay;
run;
- Job script -- filename
testsas.sh; submit job with "sbatch testsas.sh":
#!/bin/bash
#SBATCH --nodes=1
#SBATCH --ntasks=1
#SBATCH --cpus-per-task=12
#SBATCH --mem=44G
#SBATCH --account=myrsrchprj
#SBATCH --time=1:00:00
module load sas/9.4
sas lranova.sas
- Output files generated:
lranova.lstlranova.log
lranova.lstcontents (blank lines removed):
The SAS System 00:54 Thursday, February 25, 2021 1
The GLM Procedure
Class Level Information
Class Levels Values
agef 3 20 25 31
Number of Observations Read 9
Number of Observations Used 9
^L The SAS System 00:54 Thursday, February 25, 2021 2
The GLM Procedure
Dependent Variable: load
Sum of
Source DF Squares Mean Square F Value Pr > F
Model 2 4.69668889 2.34834444 17.07 0.0033
Error 6 0.82566667 0.13761111
Corrected Total 8 5.52235556
R-Square Coeff Var Root MSE load Mean
0.850487 3.560833 0.370960 10.41778
Source DF Type I SS Mean Square F Value Pr > F
age 1 4.03621252 4.03621252 29.33 0.0016
agef 1 0.66047637 0.66047637 4.80 0.0710
Source DF Type III SS Mean Square F Value Pr > F
age 0 0.00000000 . . .
agef 1 0.66047637 0.66047637 4.80 0.0710
^L The SAS System 00:54 Thursday, February 25, 2021 3
The GLM Procedure
Observation Observed Predicted Residual
1 11.45000000 11.00333333 0.44666667
2 10.42000000 11.00333333 -0.58333333
3 11.14000000 11.00333333 0.13666667
4 10.84000000 10.85000000 -0.01000000
5 11.17000000 10.85000000 0.32000000
6 10.54000000 10.85000000 -0.31000000
7 9.47000000 9.40000000 0.07000000
8 9.19000000 9.40000000 -0.21000000
9 9.54000000 9.40000000 0.14000000
Sum of Residuals -0.00000000
Sum of Squared Residuals 0.82566667
Sum of Squared Residuals - Error SS -0.00000000
First Order Autocorrelation -0.61749428
Durbin-Watson D 2.96961378
^L The SAS System 00:54 Thursday, February 25, 2021 4
Plot of load*age. Symbol used is '*'.
Plot of pred*age. Symbol used is '+'.
load |
|
11.50 +
| *
|
|
11.25 +
| *
| *
|
11.00 + +
|
| +
| *
10.75 +
|
|
| *
10.50 +
| *
|
|
10.25 +
|
|
|
10.00 +
|
|
|
9.75 +
|
|
| *
9.50 + *
|
| +
|
9.25 +
| *
|
|
9.00 +
|
--+----------+----------+----------+----------+----------+----------+----------+----------+----------+----------+----------+--
20 21 22 23 24 25 26 27 28 29 30 31
age
NOTE: 6 obs hidden.
- Contents of
lranova.log:
1 The SAS System 00:54 Thursday, February 25, 2021
NOTE: Copyright (c) 2016 by SAS Institute Inc., Cary, NC, USA.
NOTE: SAS (r) Proprietary Software 9.4 (TS1M7)
Licensed to DREXEL UNIVERSITY - T&R - SFA, Site 70131026.
NOTE: This session is executing on the Linux 4.18.0-147.el8.x86_64 (LIN X64) platform.
NOTE: Analytical products:
SAS/STAT 15.2
SAS/ETS 15.2
SAS/OR 15.2
SAS/IML 15.2
SAS/QC 15.2
NOTE: Additional host information:
Linux LIN X64 4.18.0-147.el8.x86_64 #1 SMP Thu Sep 26 15:52:44 UTC 2019 x86_64 Red Hat Enterprise Linux release 8.1 (Ootpa)
You are running SAS 9. Some SAS 8 files will be automatically converted
by the V9 engine; others are incompatible. Please see
http://support.sas.com/rnd/migration/planning/platform/64bit.html
PROC MIGRATE will preserve current SAS file attributes and is
recommended for converting all your SAS libraries from any
SAS 8 release to SAS 9. For details and examples, please see
http://support.sas.com/rnd/migration/index.html
This message is contained in the SAS news file, and is presented upon
initialization. Edit the file "news" in the "misc/base" directory to
display site-specific news and information in the program log.
The command line option "-nonews" will prevent this display.
NOTE: SAS initialization used:
real time 0.15 seconds
cpu time 0.02 seconds
1 data crack;
2 input id age load agef;
3 datalines;
NOTE: The data set WORK.CRACK has 9 observations and 4 variables.
NOTE: DATA statement used (Total process time):
real time 0.00 seconds
cpu time 0.01 seconds
13 ;
14
15 proc glm data=crack;
16 class agef;
17 model load = age agef / p;
18 output out=crackreg p=pred r=resid;
19 run;
^L2 The SAS System 00:54 Thursday, February 25, 2021
20
NOTE: The data set WORK.CRACKREG has 9 observations and 6 variables.
NOTE: The PROCEDURE GLM printed pages 1-3.
NOTE: PROCEDURE GLM used (Total process time):
real time 0.04 seconds
cpu time 0.01 seconds
21 proc plot data=crackreg;
22 plot load*age="*" pred*age="+"/ overlay;
23 run;
24
NOTE: There were 9 observations read from the data set WORK.CRACKREG.
NOTE: The PROCEDURE PLOT printed page 4.
NOTE: PROCEDURE PLOT used (Total process time):
real time 0.00 seconds
cpu time 0.00 seconds
NOTE: SAS Institute Inc., SAS Campus Drive, Cary, NC USA 27513-2414
NOTE: The SAS System used:
real time 0.22 seconds
cpu time 0.05 seconds
Moving from SAS to R
R is an open source statistical analysis language. The advantage over SAS is that it requires no expensive license, and also provides a terminal-based interface which does not require a graphical display.
Here are some tutorials for going from SAS to R:
There are also several case studies online by groups who converted from SAS to R: do a search for "SAS to R".
See Also
References
[1] SAS Website