SAS
| Site | Docs | Ref | Resources | Community | FeedBack |
| Mail | Web | Software | Classes | FAQs |
|ACCESS |DIALUP |DOS |EXCEL |FOX |HTML |MAPLE |POWERPT |SAS |
|SPSS |WINDOW |WORD |WP |WWW-HTML |WWW |XWIN |
|
You have reached an Archive Page on the CCTR web server! This web page is no longer being updated, and the information may be outdated or inaccurate. If you have reached this page from a link on another page, please inform the CCTR web maintainer. If you need help finding information regarding CCTR, please contact us at ops@cctr.umkc.edu. |
SAS
| Site
| Docs
| Ref
| Resources
| Community
| FeedBack
|
| Mail
| Web
| Software
| Classes
| FAQs
|
|ACCESS
|DIALUP
|DOS
|EXCEL
|FOX
|HTML
|MAPLE
|POWERPT
|SAS
|
|SPSS
|WINDOW
|WORD
|WP
|WWW-HTML
|WWW
|XWIN
|
Introduction to the SAS system:
-------------------------------
SAS is statistical software developed to perform data analysis. Its
capabilities range from language translation to data management and
statistical application. SAS is a set of many interrelated software products,
at the heart of those programs is Base SAS which contains SAS supervisor
which
communicate and work with the operating system, the DATA step language
compiler, input/output access method, a number of SAS procedures that take
care
of complex computing tasks, and other services that are required for the
SAS
job.
SAS Files:
----------
For SAS to be able to analyze data, the data have to be in a format that
is
readable by SAS. The data have to be stored as a SAS file ready to be
accessed
by SAS software using any of the methods. Furthermore, SAS files that
are created by SAS/Access software and
contain data that the interface engine may use to implement an access view
are
called Access descriptor files. Another kind of SAS file is Stored programs
which contain compiled code produced by a data step program. SAS Catalogs
are
a kind of SAS file that contain one of the many form catalog entries. All
these
files can be stored in SAS Libraries
The SAS Job:
------------
There are two types of SAS job step: DATA steps, where data are
read,
stored, rearranged, and written in a way the user may specify using the
DATA
step language, and PROCedure Steps which start a certain SAS service.
Spacing
is not important in SAS and the file could contain extra spacing within
it.
We can distinguish between two types of statements, executable, and
non-executable. Most DATA step statements are executable which means
that the
compiler is ordered to prepare code for an immediate action. Non-executable
commands such as TITLE, DROP, KEEP, and LABEL are not associated with a
specific demand for immediate computational action. Some commands are
only
meaningful when used under another command, Commands such as SET, MERGE,
INFILE, FORMAT, LABEL, TITLE, LIBNAME, and OPTIONS are examples of such
command. There is also what is called statement modifiers or sometimes
they are
referred to as OPTIONS, which if present in a certain context will tell
SAS to
make specific refinements to the way it carries out the instructions in
the
statement.
For Example:
TABLES COUNT;
TABLES COUNT / NOPERCENT;
or
INFILE FILE1;
INFILE FILE1 END=AGE;
In the example we modified the first statement by using the keyword
NOPERCENT, and similarly we modified the second statement by using END=VARIABLE
NAME. SAS names could not start with numbers, should be 8 characters long
or
less, should not contain any non-alphanumeric characters except the underscore
(_), and there should be no spaces included in the name. The following
Names
are invalid:
PREDICTABLE (more than 8 characters)
12AGE (cannot start with numbers)
THAT.ONE (contains a non-alphanumeric character".")
THIS ONE (contains space)
Statement Labels could be added by starting the statement with the
label
followed directly with a fixed post colon":", for Example,
THIS: INPUT X Y Z;
Operators (both numeric and character) could be used in SAS also, for example
on numeric operators we have, + is for adding, - is for subtracting, /
is for
dividing, and * is for multiplication. For character we have, LT(<)
for less
than, GT() for greater than, EQ(=) for equality, NQ(^=) for non-equality,
LE (=<) is less than or equal to. The following table should aid with
respect
to the different operators:
SAS DATA STEP LANGUAGE OPERATORS
SYM ALT Definition Example Result
--- --- ---------- ------- ------
+ (prefix) Makes Value Positive +2 2
- (prefix) Makes Value Negative -2 -2
** Exponential 3**3 27
* Multiplication 3*3 9
/ Division 4/2 2
+ Addition 4+4 8
- Substraction 4-4 0
< Maximum 5<10 10
< Minimum 5<10 5
Comparison Operators
--------------------
= EQ Equal 3+4=7 1(true)
^= NQ Not Equal 3+4^= 0(false)
< LT Less Than 3+4<7 0
GT Greater Than 3+47 0
<= LE Less Than or Equal to 3+4=<7 1
= GE Greater Than or Equal to 3+4= 1
none IN List Membership 3+ IN(6,7<8) 0
SAS also have DATA step language Functions that are specified with
basic operations, and are a SAS function name followed by one or more function
arguments in parentheses.
For example:
ROUND (387.26,.1) which equals 387.3.
The following table have a list of some of those Functions:
SOME SAS DATA STEP LANGUAGE FUNCTIONS
-------------------------------------
Numeric Functions
Purpose Example Result
------- ------- ------
Round to the nearest unit ROUND(27.3,.5) 27.5
Closest lower integer FLOOR(18.9) 18
Absolute Value ABS(-83) 83
Modular remainder MOD(16.3) 1
Summation SUM(1,2,3,4,5) 15
Average MEAN(1,2,3,4,5) 3
Maximum MAX(1,2,3,4,5) 5
Character Function
Locate Character INDEXC('HELLO','L') 3
Locate String INDEX('HELLO','LO') 4
Find nth word SCAN('HELLO THERE',2) 2
Find Substring SUBSTR('ABCDE',2,3) BCD
Uppercase UPPER('Pqrst) PQRST
In SAS the user can add any comment that he want to show on the output
files. The Comment statement must start with an astrict and end with the
SAS
semi-colon
* This is a COMMENT;
A delimited comma can also be added to enable the user to use characters
like
the semi-colon and it should start with (slash-astrict) and should end
with
(astrict-slash), for example,
/*This is a delimited Comment; yes, it ! is*/
The Following are the Statements used in the DATA step:
-------------------------------------------------------
ABORT: Used to stop the execution of SAS job Immediately,
format : IF_ERROR_THEN ABORT;
ARRAY: Used to enable the processing of many variable the same way,
format : ARRAY ITEM(|) $ 12 X1-X10;
BY: Used to control the operation of a SET, MERGE, or UPDATE Statements,
format : BY MONTH NONSORTED; or BY STATE CITY MONTH;
CALL: Used to call a routine,
format : CALL SYMPUT(macro variable, SAS variable);
CARDS: Used to note that the data is after it.
DATA: Used to begin the data set and assign a name for the data set,
format : DATA EXAMPLE;
DELETE: Used to stop the processing of a current observation,
format : IF GRADE <^ THEN DELETE;
DO: Used to specify that the statement following the DO are to be executed
as a unit until a matching END Statement appears,
format : IF X=Y THEN DO;
X=X+!;
PUT X;
END;
OUTPUT;
DROP: Used to specify variables that are not included.
END: Used to indicate the end of a DO statement.
ERROR: Used to print a message on the SAS log and sets the automatic SAS
variable_ERROR_to 1,
format : _ERROR_=1;
FILE: Used to define the current output file,
format : FILE EXAMPLE;
FORMAT: Used to associate output formats with variables in a DATA step,
format : PROC FORMAT;
VALUE SEXFMT 1=Male 2=Female;
DATA ALL;
INPUT NAME $ SEX@@;
FORMAT SEX SEXFMT.;
GO TO: Used to tell SAS to jump to a different Statement,
format : If 1<=X<=5 THEN GO TO OK;
IF: Used when you want to execute a SAS statement for some of the
observations.
INFILE: Used to identify a non-SAS file on a disk or
tape.
INFORMAT: Used to specify a default informat for variables
listed in an input statement.
INPUT: Used to read values into variables from input data lines.
KEEP: Used to specify the variable that are to be included in any SAS data
sets being created.
LABEL: Used to give labels to variables.
LENGTH: Used to specify the number of bytes for sorting values of variables in
each data set being created.
LINK: Used to tell SAS to jump to another statement in the same DATA step and
begin executing statements from that point on.
LIST: Used to list on the log file the input data lines for the observation
being processed.
LOSTCARD: used to resychronize the input data when SAS encounters a missing
record in data with multiple records per observation.
MERGE: Used to join corresponding observations from two or more SAS data sets
into single observation in a new SAS file.
MISSING: Used to declare that certain values in the data are missing.
NULL: It does not perform any task.
OUTPUT: Used to tell SAS to write the current observation to the data set being
created.
PUT: Used to write output lines.
RENAME: Used to give variables a new name.
RETAIN: Used to reatain a variable value from previous execution.
RETURN: Used to stop SAS from executing and to go back to the beginning.
SET: Used to tell SAS to read observations from different data files.
STOP: Used to tell SAS to stop processing a DATA step.
UPDATE: Used to combine observations from two SAS data sets in a manner similar
to the MERGE statement.
The following is an example of a SAS program file.
--------------------------------------------------
This file was created in CCTR by using the EVE editor as follows:
$EVE Example.SAS
type in the following program
********A SAS Program********
DATA Example;
THAT: Input Name $ Height Weight Age;
*This is a comment;
/*This is a delimited comment;yes it is !!!*/
Cards;
Joey 62 189 23
Frannie 54 121 20
Robby 59 172 19
Joy 51 108 17
Run;
Proc print data= example;
run;
proc means data= example;
Title1 'Averages of Data';
title2 'Example';
proc plot data= example;
plot height*weight;
title 'PLOTTTTTT1';
proc plot data= example;
plot height*age;
title 'plottt2';
Run;
********************
In the previous program we used the statement command to give our data
a name
“Example”, and the INPUT statement to identify the variablnames. CARDS
is the
statement that tells SAS that the data is next. After we entered the data,
we
are ready to start different procedures or “PROC”. In the example we ditwo
procedures: PRINT which give us a printout of the data, and MEANS which
give us
a printout of the averages for the data example. All statements ends with
a
semicolon”;” and this is mandatory iSAS.
If we run the previous program by using the batch command CCTR by typing
$ runsas Example.SAS, we get the following two files:
1. The Log file which should be under Example.log
The SAS System
NOTE: Copyright(c) 1989 by SAS Institute Inc., Cary, NC USA.
NOTE: SAS (r) Proprietary Software Release 6.09 TS027
Licensed to UNIVERSITY OF MISSOURI - KANSAS CITY, Site 0007049001.
NOTE: Running on DEC Model 3000 - M600 Serial Number 80000000.
Welcome to the SAS Information Delivery System, Version 6.09
===================================================================
Computing Services is now offering the full slate of SAS products. That
includes Base SAS, AF, Assist, Calc, Connect, Graph, English, ETS, FSP,
IML, Insight, Lab, OR, QC, and Stat. We also licensed the six Computer
Based Tutorials. Give the Tachyon command $ SASTUTOR to start those
tutorials.
For a short description of each product, give the VMS command
$ TYPE/PAGE PUBLIC:SAS-PRODUCTS.LIST
Our local SAS support representative is Deborah Wall. She can be
reached either at 235-1184 or username wall@vax1.umkc.edu.
1 DATA Example;
2 THAT: Input Name $ Height Weight Age;
3 *This is a comment;
4 /*This is a delimited comment;yes it is !!!*/
5 Cards;
NOTE: Unreferenced label defined: THAT.
NOTE: The data set WORK.EXAMPLE has 4 observations and 4 variables.
10 Run;
11 Proc print data= example;
12 run;
NOTE: The PROCEDURE PRINT printed page 1.
13 proc means data= example;
14 Title1 'Averages of Data';
15 title2 'Example';
NOTE: The PROCEDURE MEANS printed page 2.
16 proc plot data= example;
17 plot height*weight;
18 title 'PLOTTTTTT1';
NOTE: The PROCEDURE PLOT printed page 3.
19 proc plot data= example;
20 plot height*age;
21 title 'plottt2';
22
23 Run;
NOTE: The PROCEDURE PLOT printed page 4.
NOTE: SAS Institute Inc., SAS Campus Drive, Cary, NC USA 27513-2414.
The log file is basically a description of the interpretation of the
different data steps and proce
dures that SAS found in the file we ran. The
importance of the Log file is mainly that it tells the researcher about any
errors found or warnings that might be a reason for alarm.
2. The output file which is found under Example.Lis
1
The SAS System 10:21 Monday, May 22, 1995 1
OBS NAME HEIGHT WEIGHT AGE
--- ---- ------ ------ ---
1 Joey 62 189 23
2 Frannie 54 121 20
3 Robby 59 172 19
4 Joy 51 108 17
1
Averages of Data
Example
-------
Variable N Mean Std Dev Minimum Maximum
-------------------------------------------------------------------
HEIGHT 4 56.5000000 4.9328829 51.0000000 62.0000000
WEIGHT 4 147.5000000 39.0939040 108.0000000 189.0000000
AGE 4 19.7500000 2.5000000 17.0000000 23.0000000
-------------------------------------------------------------------
PLOTTTTTT1
----------
Plot of HEIGHT*WEIGHT. Legend: A = 1 obs, B = 2 obs,
HEIGHT |
| A
62 +
|
60 +
|
58 +
|
56 +
|
|
54 + A
|
|
52 +
| A
50 +
|
--------+------------+------------+------------+------------+---------
100 110 120 130 140 150 160 170 WEIGHT
plottt2
Plot of HEIGHT*AGE.
Legend: A = 1 obs, B = 2 obs, etc.
HEIGHT |
|
62 + A
|
60 +
| A
58 +
|
56 +
|
54 + A
|
52 +
| A
-----------+----------------+----------------+----------------+------
17 18 19 20 21 22 AGE
The output file is a list of all the results of the calculations
that
were requested. SAS as could be seen from the latter example is a very
powerful
programming language that is easy to understand and perform. It is true
that a
very simplistic example was used however, with a little extension on the
program we can do almost everything. The following are the different
statements used in the PROC step:
BY, CLASS, DROP, FORMAT, FREQ, ID, KEEP, LABEL, MODEL, OUTPUT, PROC,
VAR,
WEIGHT
The Following are Features for both, the DATA step and the PROC step:
CLEAR, CMS, COMMENT, ENDSAS, HELP, %INCLUDE, %LIST, MACRO, OPTIONS, PAGE,
%PUT,
RUN, %RUN, SKIP, TITLE, TSO, X.
As in the example SAS is a very easy to use software. Even thought, SAS
can
offer much more than that of the example SAS files are only an extension
of
our example. There is a SAS TUTOR available on UMKC's CCTR and you can
access
it by typing SASTUTOR on the dollar prompt.
Information Services, Central Computing
University
of Missouri Kansas City