First version: June 18, 2004

Revised: January 24, 2008

PATRICIA LEDESMA LIÉBANA

SAS programming skills

This document is not a self-paced SAS tutorial; rather, it focuses on data management skills
and features of SAS that are helpful in a research context. However, it does provide the basic
elements of the SAS Language. If you have some previous programming experience, the skills
listed here should be easy to understand. For a comprehensive SAS tutorial, please refer to “The
Little SAS Book” by Delwiche and Slaughter, available in the Research Computing library (room
4219).

Contents
1. SYNTAX RULES 2
2. BASIC PARTS OF A SAS PROGRAM 2
3. TYPES OF DATA FILES 3
4. PROVIDING DATA IN THE PROGRAM 3
5. CREATING VARIABLES 4
6. SAS OPERATORS AND FUNCTIONS 5
7. READING “RAW” DATA – EXTERNAL FILES 6
8. COMBINING DATASETS 6
8.1. Appending datasets 6
8.2. Merging datasets 7
8.3. PROC SQL 7
9. USING PROCEDURES TO GENERATE DATASETS 9
10. LIBNAMES AND SAS DATASETS 9
11. BY PROCESSING 10
12. CREATING COUNTERS 11
13. LAGS IN PANEL DATASETS 11
14. ARRAYS 12
15. WRITING ASCII OUTPUT DATASETS 12
15.1. Plain ASCII files: FILE and PUT statements 12
15.2. Creating or reading CSV files: EXPORT and IMPORT 13
16. USEFUL PROCEDURES 13
16.1. “Flipping” your data: PROC TRANSPOSE 13
16.2. Dealing with rolling windows: PROC EXPAND 13
17. ODS TO GENERATE CUSTOMIZED DATASETS FROM PROCS 14
18. IML – INTERACTIVE MATRIX LANGUAGE 15
19. MACROS – AN EXAMPLE 16
20. USING SAS MANUALS 18

©2004-2008 Patricia Ledesma Liébana

SAS PROGRAMMING SKILLS

1. Syntax rules
• All commands end in a semi-colon

• SAS statements are not case sensitive. You may use upper or lower case.

• Variables names can be upper or lower case.

• When you refer to "external" file name, it is case-sensitive (interaction with UNIX
operating system)

• Commands can extend over several lines as long as words are not split

• You may have more than one command per line

• SAS name rules (for datasets and variables): up to 32 characters long; must start with
a letter or an underscore (“_”). Avoid special characters.

2. Basic parts of a SAS program

• There are two basic building blocks in a SAS program: DATA ‘steps’ and PROC
(procedures). SAS procedures do not require the execution of a data step before them.

• In addition, OPTIONS to control appearance of output and log files.

• The DATA step is where you manipulate the data (creating variables, recoding,
subsetting, etc). The data step puts the data in a format SAS can understand.

• A SAS data file can have up to three parts: (a) the headers – descriptive information
of the dataset’s contents; (b) the matrix of data values; and (c) if created, indexes.
Indexes are saved to separate “physical” files, but SAS considers it part of the data
file. Thus, if an index exists, it should not be removed from the directory where the
data file resides. Version 8 data files have the extension “.sas7bdat”, while index files
have extension “.sas7bndx”. Most data files in WRDS have indexes.

• SAS reads and executes a data step statement by statement, observation by

observation. All the variables in the portion of memory that processes the current
each observation (input buffer or program data vector, depending on whether you are
reading “raw” data or a SAS data file) are reset to missing in each iteration of the
data step. The RETAIN statement prevents this from happening.

• Missing values: Generally, missing values are denoted by a period (“.”). Most
operators propagate missing values. For example, if you have three variables (v1, v2,
v3) and for observation 10, v2 is missing, creating total=v1+v2+v3 will result in a
missing value for observation 10 [if you would like the sum of the non-missing
values, use the SUM function: total=sum(v1,v2,v3)]. However, for comparison
operators (<, >, <= or >=), SAS treats missing values as infinitely negative numbers
(- ∞) [unlike Stata, which treats them as infinitely positive]

2 KELLOGG SCHOOL OF MANAGEMENT

SAS PROGRAMMING SKILLS

• PROCs are used to run statistics on existing datasets and, in turn, can generate
datasets as output. Output data sets are extremely useful and can simplify a great deal
of the data manipulation.

• DATA step and PROC boundary: Under Windows, all procedures must end with a
“RUN;” statement. In UNIX, this is not necessary since SAS reads the entire program
before execution. Does, it determines DATA step boundaries when it encounters a
PROC and knows the boundary of a PROC when it is followed by another PROC or
a DATA step. The one exception to this rule is in the context of SAS macros, where
you may need to add the “RUN;” statement.

3. Types of data files

• SAS distinguishes two types of files: “internal” (created by SAS) and “external”.
Referring to files in SAS maybe the most complicated point in learning SAS. Here,
the conventions date to mainframe systems. In brief, files external to SAS are ASCII
(or text) files and files generated by other applications (such as Excel, SPSS, etc).
Files internal to SAS are divided into “temporary” or “scratch” files, which exist only
for the duration of the SAS session, and “permanent” files, which are saved to disk.

 ASCII
External → 
 Files generated by other programs
Files in SAS → 
Internal → Temporary, " scratch" - one level names
 Permanent - two level names (libref.name)

• SAS scratch files are created and held until the end of the session in a special
directory created by SAS in a scratch area of the workstation. When SAS finishes
normally, all the files in this directory are deleted.

• SAS permanent files are written to a “SAS library”, which is simply a directory in the
system with a collection of SAS data files. In the program, the directory is assigned a
nickname, called a “libref” in the manuals, with a LIBNAME statement.
Subsequently, to refer to a permanent file, you type the libref, followed by a period,
followed by a name.

• The examples below build a data file. Until a complete version of the file is created,
all intermediate data files are “scratch” files. We ignore the LIBNAME statement
until the sample code in section 7.

4. Providing data in the program

Use a text editor to create the following program (call the file “ipos.sas”):

data ipo;
input ticker $ offer_prc offer_amount open_prc

KELLOGG SCHOOL OF MANAGEMENT 3

SAS PROGRAMMING SKILLS

date_public mmddyy8.;
format date_public yymmddn8.;
cards;
ACAD 7 37.5 7.44 05/27/04
CYTK 13 86.3 14.5 04/29/04
CRTX 7 48 7.26 05/27/04
EYET 21 136.5 30 01/30/04
;
proc sort data=ipo; by descending offer_amount;
proc print data=ipo;
title 'Recent IPOs';

Note that the name of the date variable is followed by “mmddyyn8.”. This is a date
“informat”, an instruction that tells SAS how to read the values. SAS will read the dates into SAS
dates (number of days elapsed since January 1, 1960), enabling you to do calculations with the
date. The format statement includes a SAS format, an instruction to SAS on how to display the
date (internally, it keep the number of days since 01/01/1960). The “n” in the format stands
indicates that we do not want any separators between the year, month and day (for a slash, the
format would be yymmdds10.).

To run the program, exit the text editor, type the following command at the UNIX prompt
and hit enter to execute:

sas ipos

SAS will create two files:

• ipos.log: The LOG file has all the messages, errors and warnings that the SAS
compiler has issued during execution of your program. Always check it.

• ipos.lst: The listing file contains any “printed” output generated by a procedure.

Use the UNIX “more” command to browse these files.

5. Creating variables
New variables can be defined in the data step. For example, in the previous program, type the
following statement after the INPUT statement (between INPUT and CARDS):

open_offer_ret=open_prc/offer_prc;

Run the program again to view the results. SAS has a wide range of functions that can also be
used to modify your data.

4 KELLOGG SCHOOL OF MANAGEMENT

SAS PROGRAMMING SKILLS

6. SAS operators and functions

For a complete listing of functions and their syntax, refer to the SAS/BASE “SAS Language
Reference: Dictionary” manual. For detailed explanations of the precedence each operator takes
and how SAS expressions are built, refer to the SAS/BASE “SAS Language Reference:
Concepts” manual. The list below only includes some of the functions I use most.

Arithmetic Logical Comparison

As in Excel: +, -, *, / And: & or AND Equal: =
Exponentiation: ** Or: | or OR Less than: < or LT
Not: ^ or NOT Less than or equal: <= or LE
Greater than: > or GT
Greater than or equal: >= or GE
Not equal to: ^= or NE

An additional important operator not included in the categories above, is the concatenation
operator, which creates one alphanumeric by concatenating two character strings: ||

Functions work across variables (in rows). Some important functions:

• ABS(x): absolute value

• DIFn(X): difference between argument and its n-th lag

• EXP(x): raises e to x power.

• FLOOR(x): integer that is less than or equal to x. Same results as INT(x)

• INDEX(argument, excerpt) searches an argument and returns the position of the first
occurrence.

• LAGn(x): lagged values from queue (LAG(.) is first lag)).

• LOG(x): natural log of x (LOG2(x) logarithm to the base 2, LOG10(x) logarithm to

the base 10)

• MAX(x), MIN(x): Maximum and minimum

• MEAN(X), SUM(X)

• N(x) number of non missing values

• NMISS(x) number of missing values

• SUBSTR(argument, position, length) retrieves a portion of a given string

KELLOGG SCHOOL OF MANAGEMENT 5

SAS PROGRAMMING SKILLS

7. Reading “raw” data – external files

To read data from an ASCII file, you must specify the location of the file using either the
FILENAME statement or the INFILE statement.

We will use the following sample data file:

~ple531/sastraining/sampledata.prn

In the next example, we will use the INFILE version, where the location of the external file is
referenced in quotes:

data t1;
infile '~ple531/sastraining/sampledata.prn' firstobs=2;
input year
ticker $
data6
data12
fiction;
proc print data=t1;

Note: The FILENAME would assign a “fileref” or handle to the external file, which would
then be used by the INFILE statement. The first lines of the code would be written as follows:

filename raw '~ple531/sastraining/sampledata.prn';

data t1;
infile raw firstobs=2;
etc.

Note that “raw” is an arbitrary nickname for the external file, which is then used in the
INFILE statement instead of the file location. The FILENAME command is structured like the
LIBNAME command in section 7 and, if your program is reading several external files, it might
be a more organized way of writing the program: declare all the external file names at the
beginning of the program (a sequence of FILENAME statements), making easier replication of
results if you must move the data files and programs to a different computer.

8. Combining datasets
There are two common types of dataset operations: appending datasets with the same
variables or combining (merging) datasets with different variables.

8.1. Appending datasets

In “~ple531/sastraining/additionaldata.txt” there are three additional years of sample data,

with four of the same variables as “sampledata.prn”. To append or concatenate these datasets,
first add the following statements to read the additional data into a SAS dataset:

6 KELLOGG SCHOOL OF MANAGEMENT

SAS PROGRAMMING SKILLS

data t2;
infile '~ple531/sastraining/additionaldata.txt' firstobs=2;
input year ticker $ data6 data12;

Next, concatenate both files with a new data step:

data t3;
set t1 t2;

Note: This can also be done with APPEND procedure, which is more resource efficient.
Instead of the additional data step, you can try:

proc append base=t1 data=t2;

In general, proc append will be faster than the data step: the data step will read both
datasets (t1 and t2), while proc append will only read the second (t2, the dataset being
appended to the end of t1, the base dataset).

8.2. Merging datasets

File “~ple531/sastraining/moredata.txt” has an additional variable, a fictional dummy for fair

trade coffee. It also includes the tickers. To merge these data with the rest of our coffee data:

data t4;
infile '~ple531/sastraining/moredata.txt';
input ticker $ fair;
proc sort data=t4; by ticker;
proc sort data=t3; by ticker year;
data t5;
merge t3 t4;
by ticker;

One thing to keep in mind to control the result of the merge, is that SAS keeps track of the
dataset that contributes to each observation with some internal dummies. Using the “IN=” dataset
option (for example, “merge t3 (in=a) t4 (in=b);”) it is possible to control what
observations are in the final dataset. By default, it would be all observations (from t3 and t4).
Using an IF statement such as the one below writes only those observations for which both
datasets contributed. Similarly, one could specify “in=a;”, “in=b;”, “if not b;”, etc.

data t5;
merge t3 (in=a) t4 (in=b);
by ticker;
if a and b;

8.3. PROC SQL

PROC SQL is the SAS implementation of the Standard Query Language (SQL) standard to
work with databases and is used in Oracle, Microsoft SQL Server, Microsoft Access and other
database systems. PROC SQL is a powerful language that allows sophisticated merges. When

KELLOGG SCHOOL OF MANAGEMENT 7

SAS PROGRAMMING SKILLS

you merge datasets with PROC SQL you do not need to sort them. Given that most files in
WRDS are indexed, PROC SQL also works fast when creating queries on CRSP and Compustat,
for example. Consider the example below, from the WRDS sample file “ina.sas” to retrieve data
from the Compustat Industrial file.

To end PROC SQL you must type use the QUIT statement. The SQL query is between PROC
SQL and QUIT (as a very long statement). Unlike the rest of SAS, datasets are called “tables”,
while variable names now have two levels (e.g., inanames.coname): the first level denotes the
table of origin (dataset “inanames”), while the second level is the variable name itself
(“coname”). There is no reference to SAS libnames except in the “FROM” part of the SQL query.

In the example below, the query itself has been broken into lines for convenience. The first
line defines the dataset that will be created (temp2) by selecting variables from dataset inanames
(coname and iname) and dataset “temp” (all variables). The second line of the query defines the
sources of the datasets: temp is a temporay dataset (defined in earlier SAS statements) and inames
is in the “comp” library (the libref assigned to the location of Compustat files in WRDS). Finally,
the WHERE portion of the query defines how the merge is done. In this example, observations
from temp are matched to observations of inanames on the basis of CNUM (CUSIP issuer
number) and CIC (the issue number and check digit). Note that the variables in temp are named
left of the equal sign; this implicitly defines a “left join” in which all the records of temp will be
preserved (whether there is matching Compustat data or not) and only matching records of the
Compustat inanames file.

PROC SQL;
CREATE TABLE temp2 AS SELECT
inanames.coname, inanames.iname, temp.*
FROM temp, comp.inanames
WHERE temp.cnum=inanames.cnum AND temp.cic=inanames.cic;
QUIT;

To achieve the same with a MERGE statement, assuming both datasets are sorted by CNUM
and CIC, the DATA step would read:

data temp2;
merge temp (in=a) comp.inanames;
if a;

Given that “inanames” is a large file and that the DATA step will copy each merged
observation before evaluating the IF statement, this option is less efficient than the PROC SQL
version.

The next example shows a more complex SQL statement in which observations from TAQ
consolidated trades (test) and consolidated quotes (test2) are matched based on the date and
symbol, as well as time stamp for each record. In this case, for each trade, we match all the quotes
in a 10 second window. Note that variable “secs” is a calculation based on variables from
different datasets. Also, notice that the time stamp variable was renamed on the fly.

proc sql;
create table combined as
select test.symbol, test.date, test.price, test.size,
test2.bid, test2.ofr, test2.bidsiz, test2.ofrsiz,

8 KELLOGG SCHOOL OF MANAGEMENT

SAS PROGRAMMING SKILLS

test.ct_time, test2.cq_time,
(test2.cq_time-test.ct_time) as secs
from test(rename=(time=ct_time)) left join
test2(rename=(time=cq_time))
on -10<=(ct_time-cq_time)<=10
where test.symbol=test2.symbol and test.date=test2.date;
quit;

Some limitations of PROC SQL are: (i) it can merge up to 16 tables, versus SAS normal limit
of 100; (ii) if you are merging many tables, PROC SQL can be slower relative to using a MERGE
statement.

9. Using procedures to generate datasets

Most (if not all) SAS procedures can generate output datasets that you can use for further
processing (merge with another datasets, etc), with either an OUT= option or an OUTEST=
option. Note that you can use DATA step options when you create these datasets.

proc sort data=t5; by fair;

proc means data=t5 noprint;
var data6 data12;
by fair;
output out=t6 (drop=_freq_ _type_) mean= / autoname;
proc print data=t6;

The resulting dataset can be merged with the original data to “propagate” the mean values,
allowing us to use it for further computations:

proc sort data=t5; by fair;

data spiff.t7;
merge t5 t6;
by fair;

For more options to generate output datasets from SAS procedures, see section 16 below, on
the “Output Delivery System” (ODS).

10. LIBNAMES and SAS datasets

SAS creates two types of SAS “system” data files: temporary (“scratch” or “work” files) o
“permanent”. Both types of files are SAS data files. Physical SAS files have the “sas7bdat”
extension for SAS versions 7 and up.

Temporary files are files created during a SAS session (either in interactive mode or during
the execution of a SAS program. Temporary files are named in the program with a one-level
name. For example in “DATA oranges”, the dataset “oranges” is temporary. If the program ends
successfully, the “oranges” file will be deleted by SAS at the end. If the execution fails,
temporary files will not be deleted.

KELLOGG SCHOOL OF MANAGEMENT 9

SAS PROGRAMMING SKILLS

In each SAS installation, there is a default location for the work directory, where temporary
files are created during a session. In Kellogg’s UNIX server, temporary files are directed to
“/scratch/sastemp”. This location can be changed by specifying the “-work” option before
execution:

sas –work /someotherpath progname

All default options for SAS are stored in a configuration file (sasv8.cfg) in the SAS program
directory.

“Permanent” files are SAS files named with a two-level name during the program. The first
level of this name is a handle or “libref” that associates the dataset with a specific directory where
it will be created and stored. The libref is created in a LIBNAME statement. The following
example associates the “~/training” directory to the libref “spiff”. The libref is then used to write
dataset “t7” to the directory named in the LIBNAME statement.

libname spiff '~/training';

data spiff.t7;
merge t5 t6 (drop=_freq_ _type_);
by fraud;

11. BY processing
There are many operations that require handling the data in groups of observations, as defined
by common values of one or more variables. Most procedures, if not all, allow BY and/or CLASS
statements. BY was also used above for merging data. The following example illustrates another
use of BY-processing in a data step.

To use BY-processing in a data step, first you data must be sorted according to the variables
needed in the BY statement. When the BY statement is invoked in the DATA step, SAS
automatically creates a set of internal dummy variables that identify the boundaries of each
group. In the following example, it will create four dummy variables: first.sic, last.sic,
first.permno, and last.permno. First.sic takes a value of one for the first observation in every SIC
code, and is zero otherwise. Last.sic will be one in the last observation of every SIC code and
zero otherwise.

These internal variables can be used for a variety of purposes. In the example below, we are
keeping the last observation of every “permno”, and in the example that follows in section 12, we
use the internal variables to create a counter.

PROC SORT DATA=t1; BY sic permno date;

DATA test;
SET t1;
BY sic permno;
IF LAST.permno THEN OUTPUT;

10 KELLOGG SCHOOL OF MANAGEMENT

SAS PROGRAMMING SKILLS

12. Creating counters

The following example creates a counter (called “qobs”) that counts that starts over for each
group. In this case, it is used to subset the first two observations in each group.

PROC SORT DATA=subset1;

BY qdate DESCENDING idtdtd;
DATA subset2;
SET subset1;
BY qdate DESCENDING idtdtd;
RETAIN qobs;
IF first.qdate THEN qobs=0;
qobs=qobs+1;
if qobs in(1,2);

The RETAIN statement is extremely powerful. SAS normally reads and processes
observation by observation in a data step. In each iteration, all the variables in the “program data
vector” (the area of memory where the observation processed is being read) are reset to missing
before reading the next observation. The RETAIN statement changes this behavior and keeps the
value of the previous iteration. This is what enables the counter to increase from iteration to
iteration.

13. Lags in panel datasets

The LAGn(argument) function can be used to create different lags of a variable. In the
context of a panel dataset, the problem is the SAS DATA step does not distinguish between
different cross-section units: if dealing with a monthly panel of securities, the lagged value of
variable for the first observation of every cross-section unit (starting with the second one) will be
set to the value of the variable for the last observation of the previous cross-section unit. A quick
solution is to use BY processing and “clean” the lags:

PROC SORT DATA=test; BY permno date;

DATA test; set test;
BY PERMNO;
lagprc=LAG(prc);
IF FIRST.PERMNO then lagprc=.;

Higher order lags can be created by specifying the order in the function: LAG2(prc),
LAG3(prc), etc.

There is no “lead” function, but the LAG function can be used to create leads by using it after
sorting the dataset in descending order (BY permno DESCENDING date;)

KELLOGG SCHOOL OF MANAGEMENT 11

SAS PROGRAMMING SKILLS

14. Arrays
SAS arrays are named groups of variables within a DATA step that allow you to repeat the
same tasks for related variables. Arrays may have one or more dimensions. In the following
example, we use only one dimension. Suppose you have received a data file where each
observation is a firm (identified by permno) and 552 columns, each of them for a daily return,
where missing values were coded as -99. Instead of writing 552 IF statements (“if ret1=-99
then ret1=.; if ret2=-99 then ret2=.; etc”), you could define an array that groups
all 552 columns and the cycle through each variable replacing the values:

data test;
set retdata;
array returns(552) ret1-ret552;
do i=1 to 552;
if returns(i)=-99 then returns(i)=.;
end;
drop i;

There is more flexibility in SAS arrays than this example shows. For example, if the firm
identifier was a character variable (and all variables that you want in the array were numeric),
instead of declaring the list of variables in the array as “array returns(552) ret1-
ret552”, you could declare it as “array returns(*) _numeric_”. In this case, SAS will
group all the numeric variables in the array and count how many there are (the array dimension).
The DO statement would then be rewritten as “do i=1 to dim(returns);”

15. Writing ASCII output datasets

15.1. Plain ASCII files: FILE and PUT statements

Within a data step, the combination of the FILE (naming the destination of the data) and PUT
(which writes lines to the destination named in FILE) statements provides great flexibility. You
may use it to generate anything from a plain ASCII data file to a LaTex table by including the
proper formatting. The code below provides a very simple output file:

data temp;
set inststat;
file 'spectruminst.dat';
put cusip $ year totsh noinst;

The PUT statement can be very refined. For example, if the data being written contains
missing values, then you may want to write a formatted PUT statement, in which each variable
will take a specific column location in the output file (e.g. cusip in columns 1-8 of the file, year in
columns 10-13, etc).

12 KELLOGG SCHOOL OF MANAGEMENT

SAS PROGRAMMING SKILLS

15.2. Creating or reading CSV files: EXPORT and IMPORT

PROC EXPORT and PROC IMPORT are procedures that allow SAS to create and read
comma-separated-values (CSV) files, which include headers. These files are easy to read in
Matlab (csvread function – start reading in row 2 to avoid the header) and Stata (insheet
command).

proc export data=temp

outfile="spectruminst.csv"
dbms=csv
replace;

When you run a SAS program that includes a PROC EXPORT or a PROC IMPORT, be sure
to specify the “-noterminal” option, as these procedures normally require X Windows:

sas –noterminal program_name

16. Useful procedures

16.1. “Flipping” your data: PROC TRANSPOSE

PROC TRANSPORT, part of the SAS/BASE procedures, allows to restructure a dataset

converting variables into observations and vice versa.

FILENAME spiff 'externaldata/multi.csv';

PROC IMPORT DATAFILE=spiff OUT=test DBMS=CSV;
GETNAMES=YES;
*PROC CONTENTS DATA=test; *PROC PRINT DATA=test (OBS=10);
PROC TRANSPOSE DATA=test OUT=test1 NAME=ticker PREFIX=date;
VAR msft ibm ba xrx;
ID date;
PROC PRINT DATA=test1;

16.2. Dealing with rolling windows: PROC EXPAND

The following program uses PROC EXPAND (part of SAS/ETS) to compute a moving sum
over a 3-month window. The example in question is for cumulative returns, calculated using
natural logs. The data must be sorted by the group identifier (permno) and time identifier (date);
the variable specified in the ID statement must be a SAS date or time. By default, PROC
EXPAND uses the spline method to interpolate missing values; by specifying METHOD=NONE,
this is avoided. Finally, in the CONVERT statement, the options NOMISS and TRIMLEFT
(which will set to missing the first two observations of every BY group) will ensure that all the
computed cumulative returns include three dates. Otherwise, the first return for each permno
would be in the second date, computed with only two observations.

DATA test; INFILE "rets2.txt";

KELLOGG SCHOOL OF MANAGEMENT 13

SAS PROGRAMMING SKILLS

INPUT permno date YYMMN6. ret;

FORMAT date YYMMN6.;
logret=LOG(ret+1);
PROC PRINT DATA=test (OBS=12);
PROC EXPAND DATA=test OUT=out METHOD=NONE;
BY permno;
ID date;
CONVERT logret=movsum / TRANSFORMOUT=(NOMISS MOVSUM 3 TRIMLEFT 2);
DATA out;
SET out;
cumret=EXP(movsum)-1;
PROC PRINT DATA=OUT;

This procedure can also compute statistics for centered windows, forward looking windows,
lags, leads within a period, etc, so it can be a welcome alternative to a more complex data step.

17. ODS to generate customized datasets from PROCs

Starting with version 7, the “Output Delivery System” (ODS) allows the user to control each
portion of a procedure’s output. In particular (among many features), ODS allows the user to
create datasets from specific portions of the output that could not be sent to a dataset with an
OUT= or OUTEST= option. The complete ODS documentation is available in the SAS/BASE
manual “The Complete Guide to the SAS Output Delivery System”.

To use ODS to create a dataset from a PROC statement:

1. Find out how SAS calls the piece of output you need. For this, type “ODS TRACE
ON” before the PROC you need to examine, and “ODS TRACE OFF” after the
PROC. Example:

ods trace on;

proc model data=uspop;
pop = a / ( 1 + exp( b - c * (year-1790) ) );
fit pop start=(a 1000 b 5.5 c .02);
run;
ods trace off;

2. Run the program and examine the log file – it will include a listing of each object of
output sent to the LST file. For example, the table with parameter estimates in PROC
MODEL is called “ParameterEstimates”.

3. Use the object name to create the output file. Furthermore, you can “turn off” the
listing with the “ODS LISTING OFF” statement. This is extremely useful if you need
to run the same procedure repeatedly for a large number of units – if you add the
NOPRINT option to the procedure, no output will be written at all. Example:

ods listing close;

ods output ParameterEstimates=d2 (drop=Esttype);
proc model data=uspop;

14 KELLOGG SCHOOL OF MANAGEMENT

SAS PROGRAMMING SKILLS

pop = a / ( 1 + exp( b - c * (year-1790) ) );

fit pop start=(a 1000 b 5.5 c .02);
quit;
ods listing; /* Allow output of PROC PRINT to be listed */
proc print data=d2;
title "ODS output dataset";

18. IML – Interactive Matrix Language

SAS has a matrix language in which expressions are close to Matlab. You can read a SAS
dataset into a matrix or produce a dataset from a matrix – you may go back and forth between
using SAS data steps and procedures and using IML. Like in PROC SQL, you end the procedure
with a QUIT statement.

The following example shows some functions in IML: (i) first, we read the “fake” dataset
into a matrix beta1; (ii) compute betax the inverse of beta1T*beta1 and print it (the matrix will be
printed in the LST file); (iii) define a column vector and a 2 by 2 matrix providing the data in
IML; (iv) define a row vector with elements: x1, x2, x3; (v) create a new dataset, fake2, with the
contents of betax. The CREATE statement defines the dataset with the characteristics of the betax
matrix, using newvars as the source of the column names. The APPEND statement “populates”
the dataset with the data from the matrix. Note that the transpose symbol is the backquote
character, generally on the top left of a keyboard; alternatively, you may write t(beta1).

data fake;
input cusip $ v1 v2 v3;
cards;
a 1 34 65
b 1 78 32
c 2 12 01
d 4 54 29
;
proc print data=fake;
proc iml;
use fake;
read all var _num_ into beta1 [rowname=cusip colname=names];
print beta1;
betax=inv(beta1`*beta1);
print betax;
columnvector={1,2,3,4};
twobytwomatrix={1 2, 3 4};
print columnvector, twobytwomatrix;
newvars='x1' : 'x3';
create fake2 from betax [colname=newvars];
append from betax;
close fake2;
quit;
proc print data=fake2;

KELLOGG SCHOOL OF MANAGEMENT 15

SAS PROGRAMMING SKILLS

19. Macros – an example

The next step in learning SAS is to learn macros. The SAS macro language allows you to run
a program for different parameters, such as tickers, dates, etc.

• Build your macros step by step based on a program that works. At each stage of the
program, use PROC PRINT to take a look at intermediate output.

• Add one macro variable at a time.

• Test extensively. For example, test for two tickers, then for 10. Make sure that the
macro produces the desired results before trying it on the entire universe of tickers!
At each stage of the program, use PROC PRINT to take a look at intermediate output.

• Complex macros are extremely hard to debug, which is why you should follow the
rules above.

Suppose you have some data in which you would like to replace the missing values for each
variable with the median for the variable. In this example, there are two variables, but the
macro was written for any number of variables. First, we compute the median with PROC
MEANS, and transpose the output dataset so that there will be only one variable (col1) with
the medians for each variable.

data fake;
input id $ v2 v3;
cards;
1 37483 .
2 . 83478
3 758 7848
4 584 4785
5 849 49549
6 . 73
7 . .
8 347 .
;
proc print data=fake;

proc means data=fake noprint;

var _numeric_;
output out=fake2 (drop=_type_ _freq_) median= / autoname;

proc transpose data=fake2 out=fake2;

proc print data=fake2;

Next, we use the SYMPUT function, one of the DATA step interfaces to the macro facility.
The SYMPUT function assigns a value produced in the data step to a macro variable. Since the
macro could be used for many variables (and we don’t want to count), the macro variable names
are automatically generated by appending the observation number to the “med” prefix: med1,
med2, ... , med_n_. The LEFT function trims leading blank spaces that would be added if there
were more than 9 observations – numbers are “right aligned”: e.g., . With the help of the “END=”

16 KELLOGG SCHOOL OF MANAGEMENT

SAS PROGRAMMING SKILLS

option in the SET statement (which defines a dummy variable that will be set to 1 when it reaches
the end of the file, in the last observation). We also define an additional macro variable “maxi”,
which will have the total number of observations (number of medians, in this case).

data _null_;
set fake2 end=last;
call symput('med'||left(_n_),col1);
if last then call symput('maxi',_n_);
run;

Now that we have the necessary ingredients, we can write the macro that will replace missing
values with the corresponding median. Macro statements are identified by the leading percentage
(%) sign, while macro variables are identified by the leading ampersand (&). You start a macro
by defining it: %macro macroname; (where macroname is a name assigned by the user) and
end the macro with a %mend; statement. To run the macro, you use assigned the macro name:
%macroname;.

The tricky part in this macro is how SAS resolves it: SAS will read the entire macro and
replace the macro variables with their assigned values. We want it to cycle from the first variable
and corresponding median to the last (nth) variable. For the first variable in the array, we want to
assign the first median for each missing value. Notice that in the IF statement, the item
corresponding to the median has two ampersands: “&&med&n”. This is to force SAS to resolve the
macro in two passes. In the first one, for n=1, the IF statement will resolve to: “if vars(1)=.
then vars(1)=&med1;”

In the second pass, &med1 will be replaced by the actual value for the median of the first
variable in the array. Without the double ampersand, SAS would try to evaluate the IF statement
in the first pass (“if vars(1)=. then vars(1)=med1;”) and produce an error since it
expected a numeric value. Finally, note that the n index in the %do statement is also a macro
variable.

%macro substit;
data fake3;
set fake;
array vars (*) _numeric_;
%do n=1 %to &maxi;
if vars(&n) = . then vars(&n)=&&med&n;
%end;
%mend;
%substit;

proc print data=fake3;

KELLOGG SCHOOL OF MANAGEMENT 17

 SAS PROGRAMMING SKILLS

20. Using SAS manuals

SAS has extensive documentation for each of its modules. The complete set of manuals is
available online:

http://support.sas.com/onlinedoc/913/docMainpage.jsp

All DATA step statements, functions and options are explained in detail in the BASE SAS
manual. The SAS macro language is also part of BASE. The following table provides a list of
important procedures and their location in the SAS manuals. Most procedures allow you to create
output datasets and avoid listing output with the “NOPRINT” option.

Manual Procedure Use

Base append Append one dataset to another
contents List variables in a dataset
corr Computes correlations (Pearson, Spearman, …)
Lists datasets available in a specified library; allows you to delete datasets
datasets
within a program
export Translate dataset to a different format, such as CSV
freq Computes frequencies and crosstabs
Import files in a different format, such as CSV. The Windows version allows
import
you to read Excel or Access files with this procedure.
means Computes summary statistics
“Prints” observations in the output (LST) file. Important options are OBS=n to
print control the number of observation and specifying the variables to be printed
with the VAR statement.
Computes ranks and can divide observations in groups (percentiles, deciles,
rank
etc)
transpose Restructure dataset
Computest detailed univariate statistics, including plots of a variable’s
univariate
distribution
STAT logistic Logistic regression (logit, probit, cloglog)
nlin Non linear estimation
probit Probit
reg Regression, OLS
ETS arima ARIMA models, including identification tests (ACF, PACF, Q tests, etc)
autoreg GARCH models
expand Time series data interpolation, data manipulation, etc
Multidimensional choice models (nested logit, multinomial probit, conditional
mdc
logit, etc). Experimental in version 8.
Estimate one or more nonlinear equations: OLS, 2SLS, SUR, 3SLS, GMM,
model
FIML
Qualitative and limited dependent variable models: logit, probit, bivariate
qlim
logit/probit, etc. Experimental in version 8
syslin Simultaneous linear equations
tscsreg Panel data – disadvantage: does not allow for unbalanced panels
x11 Seasonal adjustment
IML IML Matrix language
OR nlp Non Linear Programming

SAS also has a GRAPH module, which produces very good graphics. Useful additional
recommendations

18 KELLOGG SCHOOL OF MANAGEMENT

SAS PROGRAMMING SKILLS

An extremely useful reference book (not part of the on-line documentation) is the SAS Institute’s
“Combining and Modifying SAS Data Sets: Examples.” Also useful is the volume by Boehmer,
Broussard and Kallunki, “Using SAS in Financial Research.”

KELLOGG SCHOOL OF MANAGEMENT 19