LWPRG1

SAS Programming 1:
Essentials
Course Notes
SAS Programming 1: Essentials Course Notes was developed by Charlot Bennett and Kathy Passarella.
Additional contributions were made by Davetta Dunlap, Michele Ensor, Susan Farmer, Ted Meleky,
Linda Mitterling, Bill Powers, Jim Simon, and Roger Staum. Editing and production support was
provided by the Curriculum Development and Support Department.
SAS and all other SAS Institute Inc. product or service names are registered trademarks or trademarks of
SAS Institute Inc. in the USA and other countries. indicates USA registration. Other brand and product
names are trademarks of their respective companies.
SAS Programming 1: Essentials Course Notes
Copyright 2013 SAS Institute Inc. Cary, NC, USA. All rights reserved. Printed in the United States of
America. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in
any form or by any means, electronic, mechanical, photocopying, or otherwise, without the prior written
permission of the publisher, SAS Institute Inc.
Book code E2132, course code LWPRG1/PRG1, prepared date 14Feb2013.
LWPRG1_003
ISBN 978-1-61290-489-4
For Your Information
Table of Contents
Course Description ...................................................................................................................... ix
Prerequisites ................................................................................................................................. x
Chapter 1
Introduction .......................................................................................... 1-1
1.1
Overview of SAS Foundation .......................................................................................... 1-3
1.2
Course Logistics .............................................................................................................. 1-6

Demonstration: Investigating the Help Facility ...................................................... 1-12
1.3
Course Data Files ........................................................................................................... 1-14

Demonstration: Creating Course Data Files............................................................ 1-15
Exercises.................................................................................................................. 1-16
Chapter 2
SAS Programs ..................................................................................... 2-1
2.1
Introduction to SAS Programs ......................................................................................... 2-3
2.2
Submitting a SAS Program .............................................................................................. 2-7

Demonstration: Submitting a SAS Program: SAS Enterprise Guide ........................ 2-8
Demonstration: Submitting a SAS Program: SAS Windowing Environment ........ 2-12
Exercises.................................................................................................................. 2-18
2.3
SAS Program Syntax ..................................................................................................... 2-20

Demonstration: Automatic Formatting with SAS Enterprise Guide ....................... 2-23
Demonstration: Diagnosing and Correcting Syntax Errors..................................... 2-27
Demonstration: Correcting Unbalanced Quotation Marks: SAS Windowing
Environment .................................................................................. 2-30
Demonstration: Correcting Unbalanced Quotation Marks: SAS Enterprise
Guide ............................................................................................. 2-33
Exercises.................................................................................................................. 2-35
2.4
Solutions ........................................................................................................................ 2-36

Solutions to Exercises ............................................................................................. 2-36
iii
iv
Solutions to Student Activities (Polls/Quizzes) ....................................................... 2-38

Chapter 3
3.1
Accessing Data ..................................................................................... 3-1
Examining SAS Data Sets ............................................................................................... 3-3

Exercises.................................................................................................................. 3-12
3.2
Accessing SAS Libraries ............................................................................................... 3-13

Demonstration: Browsing SAS Libraries: SAS Enterprise Guide .......................... 3-25
Demonstration: Browsing SAS Libraries: SAS Windowing Environment ............. 3-29
Exercises.................................................................................................................. 3-33
3.3
Solutions ........................................................................................................................ 3-34

Chapter 4
4.1
Producing Detail Reports..................................................................... 4-1
Subsetting Report Data .................................................................................................... 4-3

Exercises.................................................................................................................. 4-24
4.2
Sorting and Grouping Report Data ................................................................................ 4-27

Exercises.................................................................................................................. 4-38
4.3
Enhancing Reports ......................................................................................................... 4-40

Exercises.................................................................................................................. 4-48
4.4
Solutions ........................................................................................................................ 4-50

Chapter 5
5.1
Formatting Data Values........................................................................ 5-1
Using SAS Formats.......................................................................................................... 5-3

Exercises.................................................................................................................. 5-10
5.2
User-Defined Formats .................................................................................................... 5-12

Demonstration: Defining and Using a Numeric Format ......................................... 5-18
Exercises.................................................................................................................. 5-23
5.3
Solutions ........................................................................................................................ 5-25

Chapter 6
6.1
Reading SAS Data Sets ...................................................................... 6-1
Reading a SAS Data Set .................................................................................................. 6-3

Exercises.................................................................................................................. 6-15
6.2
Customizing a SAS Data Set ......................................................................................... 6-17

Exercises.................................................................................................................. 6-39
6.3
Solutions ........................................................................................................................ 6-42

Chapter 7
7.1
Reading Spreadsheet and Database Data .......................................... 7-1
Reading Spreadsheet Data ............................................................................................... 7-3

Demonstration: Reading Excel Data in SAS Enterprise Guide .............................. 7-10
Demonstration: Reading Excel Data in the SAS Windowing Environment ........... 7-13
Demonstration: Creating a SAS Data Set ............................................................... 7-16
Exercises.................................................................................................................. 7-17
7.2
Reading Database Data .................................................................................................. 7-20
7.3
Solutions ........................................................................................................................ 7-24

Chapter 8
Reading Raw Data Files ....................................................................... 8-1
8.1
Introduction to Reading Raw Data Files .......................................................................... 8-3
8.2
Reading Standard Delimited Data.................................................................................... 8-7

Demonstration: Examining Data Errors .................................................................. 8-33
vi
Exercises.................................................................................................................. 8-34
8.3
Reading Nonstandard Delimited Data ........................................................................... 8-37

Demonstration: Using List Input: Importance of Colon Format Modifier .............. 8-48
Exercises.................................................................................................................. 8-51
8.4
Handling Missing Data .................................................................................................. 8-54

Exercises.................................................................................................................. 8-60
8.5
Solutions ........................................................................................................................ 8-63

Chapter 9
9.1
Manipulating Data................................................................................. 9-1
Using SAS Functions ....................................................................................................... 9-3

Exercises.................................................................................................................. 9-10
9.2
Conditional Processing .................................................................................................. 9-12

Exercises.................................................................................................................. 9-37
9.3
Solutions ........................................................................................................................ 9-40

Chapter 10
Combining Data Sets.......................................................................... 10-1
10.1 Concatenating Data Sets ................................................................................................ 10-3

Exercises................................................................................................................ 10-22
10.2 Merging Data Sets One-to-One.................................................................................... 10-25
10.3 Merging Data Sets One-to-Many ................................................................................. 10-30
Exercises................................................................................................................ 10-47
10.4 Merging Data Sets with Non-Matches ......................................................................... 10-49
Exercises................................................................................................................ 10-71
10.5 Solutions ...................................................................................................................... 10-73
Solutions to Exercises ........................................................................................... 10-73

Solutions to Student Activities (Polls/Quizzes) ..................................................... 10-77
Chapter 11
Creating Summary Reports ............................................................... 11-1
11.1 The FREQ Procedure ..................................................................................................... 11-3

Exercises................................................................................................................ 11-24
11.2 The MEANS and UNIVARIATE Procedures .............................................................. 11-28
Exercises................................................................................................................ 11-40
11.3 Using the Output Delivery System .............................................................................. 11-44
Demonstration: Using the Output Delivery System.............................................. 11-55
Exercises................................................................................................................ 11-62
11.4 Solutions ...................................................................................................................... 11-65
Solutions to Exercises ........................................................................................... 11-65
Solutions to Student Activities (Polls/Quizzes) ..................................................... 11-69
Chapter 12
Learning More ..................................................................................... 12-1
12.1 Introduction .................................................................................................................... 12-3
vii
viii
Course Description
This course is for users who want to learn how to write SAS programs. It is the entry point to learning
SAS programming and is a prerequisite to many other SAS courses. If you do not plan to write SAS
programs and you prefer a point-and-click interface, you should attend the SAS Enterprise Guide 1:
Querying and Reporting course.
To learn more
For information about other courses in the curriculum, contact the SAS
Education Division at 1-800-333-7660, or send e-mail to [email protected].
You can also find this information on the Web at support.sas.com/training/
as well as in the Training Course Catalog.
For a list of other SAS books that relate to the topics covered in this
Course Notes, USA customers can contact our SAS Publishing Department
at 1-800-727-3228 or send e-mail to [email protected]. Customers outside
the USA, please contact your local SAS office.
Also, see the Publications Catalog on the Web at support.sas.com/pubs for
a complete list of books and a convenient order form.
ix
Prerequisites
Before attending this course, you should have experience using computer software. Specifically, you
should be able to
understand file structures and system commands on your operating systems
access data files on your operating systems.
No prior SAS experience is needed. If you do not feel comfortable with the prerequisites or are new to
programming and think that the pace of this course might be too demanding, you can take the SAS
Programming Introduction: Basic Concepts course before attending this course. SAS Programming
Introduction: Basic Concepts is designed to introduce you to computer programming and presents a
portion of the SAS Programming 1: Essentials material at a slower pace.
Chapter 1 Introduction
1.1
Overview of SAS Foundation ........................................................................................ 1-3
1.2
Course Logistics ............................................................................................................ 1-6

Demonstration: Investigating the Help Facility ..................................................................... 1-12
1.3
Course Data Files ......................................................................................................... 1-14

Demonstration: Creating Course Data Files ........................................................................ 1-15
Exercises .............................................................................................................................. 1-16
1-2
Copyright 2013, SAS Institute Inc., Cary, North Carolina, USA. ALL RIGHTS RESERVED.
1.1 Overview of SAS Foundation

Objectives
Characterize SAS software.

Describe the functionality of Base SAS and
SAS Foundation tools.
3
3
What Is SAS?
SAS is a suite of business solutions and technologies
to help organizations solve business problems.
1-3
1-4
What Can You Do with SAS?

SAS software enables you to do the following:
access data across multiple sources
manage data
perform sophisticated analyses
deliver information across your organization
5
5
What Is Base SAS?

Base SAS is the centerpiece of all SAS software.
Flexible
Extensible
Integrated
Powerful
Base SAS is a product within the SAS Foundation set

of products that provides
a highly flexible, highly extensible fourth-generation
programming language
a rich library of encapsulated programming procedures
a graphic user interface for administering SAS tasks.
6
6
About This Class

This class focuses on writing SAS programs to do the
following:
access data in various forms
create SAS data sets
use prewritten procedures to analyze
data and write basic reports
combine data sets
generate detail and summary reports
in various formats, including HTML,
RTF, PDF
1.01 Multiple Choice Poll

Have you worked with Base SAS?
a. yes, just maintaining programs
b. yes, writing some programs
c. no, not at all
1-5
1-6
1.2 Course Logistics

Objectives
Describe the data used in the course.

Designate the editors and processing mode available
for workshops.
Specify the naming convention used for course files.
Define the three levels of exercises.
Navigate the Help facility.
11
11
Orion Star Sports & Outdoors

This course focuses on a fictitious global sports and
outdoors retailer that has traditional stores, an online
store, and a catalog business.
12
Orion Star Data

Large amounts of data are stored in transactional systems
in various formats.
1,000
employees
90,000
customers
150,000
orders
64 suppliers
13
13
Orion Star Business Scenarios

In this course, you write SAS programs that access
Orion Star data and create reports using an editor.
data newemp;
set orion.emp;
where Salary le 100000;
run;
proc means data=newemp;
class Job_Title;
var Salary;
run;
SAS Enterprise Guide

Program Editor
SAS Windowing
Environment Editor
14
1-7
1-8
What Is SAS Enterprise Guide?

SAS Enterprise Guide is a powerful Windows
client application that provides a GUI for
transparently accessing the power of SAS.
It provides the following:
a point-and-click interface with menus and wizards
that enable the user to define tasks
SAS code generation and execution based on user
selections
a full programming interface that can be used to write,
edit, and submit SAS code
This class uses the programming interface.
15
15
What Is the SAS Windowing Environment?

The SAS windowing environment consists of a
series of windows that you can use to edit and
submit programs, and view the results.
The SAS windowing environment editor contains the
following windows:
the Enhanced Editor and Program Editor windows
for preparing and submitting a program
the Log window for viewing notes, warning messages,
and error messages
the Output window, which contains the output
generated by most SAS procedures
16
16
SAS Windowing Environment
Editor
Program Editor
Enhanced Editor or Program Editor
Formatting
automatic
manual
Syntax Help
context-sensitive
menu- or function key-based
Output
SAS Report
HTML

Projects
yes
no
Autocomplete yes
no
Program
Flow
Analysis
no
yes

Which editor will you use outside of this class to write
SAS programs?
a. SAS Enterprise Guide Program Editor
b. the Program Editor in the SAS windowing
environment
c. a different editor
d. I do not know.
17
Running SAS Programs

In this course, you invoke SAS in interactive mode
(SAS Enterprise Guide or windowing environment)
to process programs.
18
18
1-9
1-10
Running SAS Programs

There are other modes for processing SAS programs.
Batch Mode for z/OS (OS/390)
Noninteractive Mode
Use any editor to create a file

with SAS statements plus job
control statements (JCL), and
then submit the file to the
operating system.
Use any editor to create a file

with SAS statements, and then
issue the SAS command
referencing the file.
Example file:
//jobname JOB
// EXEC SAS
//SYSIN DD *
proc freq data=x.pay;
tables ID;
run;
Directory-based example:
SAS filename
z/OS (OS/390) example:
SAS INPUT(filename)
19
19

How will you invoke SAS outside of this class?
a. interactive mode using SAS Enterprise Guide
b. interactive mode using the windowing environment
c. batch mode
d. noninteractive mode
e. I do not know.
20
Program Naming Conventions

In this course, you retrieve and save SAS programs using
the structure below.
course ID
p104d01x
chapter #
type
a=activity
d=demo
e=exercise
s=solution
item #
Programming 1, Chapter 4, Demo 1
placeholder
21
21
Filename and Library Name References

In this course, macro variable references are used
to give a more flexible approach for locating files.
Examples:
%let path=s:\workshop;
filename sales "&path\sales.dat";
infile "&path\payroll.dat";
22
22
1-11
1-12
Three Levels of Exercises

The course is designed to have you complete only one
set of exercises. Select the level most appropriate for
your skill set.
Level 1
Provides step-by-step instructions.
Level 2
Provides less information and guidance.
Challenge Provides minimal information and guidance.

You might need to use the Help facility.
23
23
Getting Help
In class, you can get product help in several ways,
depending on the editor being used.
Getting Started tutorials
Help facilities included in the software
web-based help, if web access is available
24
Investigating the Help Facility

The demonstration illustrates how to access and explore the Help facility in Enterprise Guide and in the
windowing environment.

1. Invoke a SAS Enterprise Guide session.
2. Select Help from the menu.
3. Investigate the Help facility.

1. Invoke a SAS session.
2. Select Help from the menu.
3. Investigate the Help facility.

Were you able to open the Help facility in your session?
a. I opened Help in Enterprise Guide.
b. I opened Help in SAS.
c. I was not able to open Help.
27
1-13
1-14
Extending Your Learning

After class, you will have access to an extended learning
page that was created for this course. The page includes
course data and program files
a PDF file of the course notes
other course-specific resources.
This page might also be available during class.
28
1.3 Course Data Files

Objectives
Execute a SAS program to create the course data

files.
Execute a SAS program to define the data location.
31
31
1-15
Business Scenario
Identify a location for the course data files and execute
programs to create the files and define the location.
Data
location?
cre8data.sas
libname.sas
32
32
Creating Course Data Files

cre8data
The cre8data program creates data files for this course. The program must be executed once, at the start
of the course.
1. Define target locations for your course data files. The default location for all course data is
s:\workshop. If your data files are to be created in a location other than s:\workshop, you must
identify a location for the SAS data files.
Create the SAS data files here: ________________________________________
2. Select File Open Program.
3. Navigate to the data folder, select cre8data, and click Open. The program is displayed in an editor.
4. Note the default values for the %LET statement. If your files are to be created at a location other than
s:\workshop, change the value assigned to PATH= to reflect the location of the SAS data files.
If your files are to be created in s:\workshop, then no change is needed.
5. Press F3 to submit the program.

6. Click the Results tab and verify that the output contains a list of data files.
1-16
Defining the Data Location

libname
The libname program tells SAS where to find the course data files. This program must be executed each
time you start a new session.
1. Open the libname program. The LIBNAME statement starts with an asterisk. This means it is
commented out and SAS will not execute this statement. At this point, you only want SAS to execute
the %LET statement to tell SAS where the data files are located. You learn about the LIBNAME
statement and the use of comments later in this course.
*libname orion "s:\workshop";
The data location might be different in your libname program. It was defined based on the
data location specified in cre8data.
2. Submit the program. Click the Log tab and verify that there are no errors or warnings.
Exercises
You must complete the exercises to create the course data files. If you do not create the data files,
all programs in this course will fail.
Required Exercise
1. Creating Course Data
a. The default location for all course data is s:\workshop. If your data files are to be created in a
location other than s:\workshop, you must identify a location for your SAS data files.
Create the SAS data files here: ________________________________________
b. Select File Open Program.
c. Navigate to the data folder, select cre8data, and click Open. The program is displayed in an
editor. Observe the default value in the %LET statement.
1-17
d. If your files are to be created at a location other than s:\workshop, change the value assigned to
PATH= to reflect the location of your SAS data files.
If your files are to be created in s:\workshop, then no change is needed.
e. Press F3 to submit the program.

f. Click the Results tab and verify that the output contains a list of data files, similar to the list
below:
2. Defining the Data Location

a. Open the libname program. Do not change anything in this program.
b. Submit the program.
c. Click the Log tab and verify that there are no errors or warnings.
1-18
Chapter 2 SAS Programs

2.1
Introduction to SAS Programs ...................................................................................... 2-3
2.2
Submitting a SAS Program ............................................................................................ 2-7

Demonstration: Submitting a SAS Program: SAS Enterprise Guide ..................................... 2-8
Demonstration: Submitting a SAS Program: SAS Windowing Environment ....................... 2-12
Exercises .............................................................................................................................. 2-18
2.3
SAS Program Syntax .................................................................................................... 2-20

Demonstration: Automatic Formatting with SAS Enterprise Guide ..................................... 2-23
Demonstration: Diagnosing and Correcting Syntax Errors .................................................. 2-27
Demonstration: Correcting Unbalanced Quotation Marks: SAS Windowing
Environment ............................................................................................... 2-30
Demonstration: Correcting Unbalanced Quotation Marks: SAS Enterprise Guide .............. 2-33
Exercises .............................................................................................................................. 2-35
2.4
Solutions ....................................................................................................................... 2-36

Solutions to Exercises .......................................................................................................... 2-36
Solutions to Student Activities (Polls/Quizzes) ..................................................................... 2-38
2-2
2.1 Introduction to SAS Programs

Objectives
List the components of a SAS program.
3
3
SAS Programs
A SAS program is a sequence of one or more steps.
DATA Step
PROC Step
DATA steps typically create SAS data sets.

PROC steps typically process SAS data sets to
generate reports and graphs, and to manage data.
4
4
2-3
2-4
SAS Program Steps

A step is a sequence of SAS statements. This program
has a DATA step and a PROC step.
data work.newemps;
infile "&path\newemps.csv" dlm=',';
input First $ Last $ Title $ Salary;
run;
proc print data=work.newemps;
run;
5
5
Step Boundaries
SAS steps begin with either of the following:
a DATA statement
a PROC statement
SAS detects the end of a step when it encounters
one of the following:
a RUN statement (for most steps)
a QUIT statement (for some procedures)
the beginning of another step (DATA statement
or PROC statement)
6
6
2.01 Quiz
How many steps are in this program?
data work.newsalesemps;
length First_Name $ 12
Last_Name $ 18 Job_Title $ 25;
input First_Name $ Last_Name $
Job_Title $ Salary;
run;
proc print data=work.newsalesemps;
run;
proc means data=work.newsalesemps;
var Salary;
run;
p102d01
7
7
SAS Program Example

This DATA step creates a temporary SAS data set named
work.newsalesemps by reading four fields from a file.
Job_Title $ Salary;
run;
run;
var Salary;
run;
p102d01
2-5
2-6
SAS Program Example

This PROC PRINT step lists the work.newsalesemps
data set.
Job_Title $ Salary;
run;
run;
var Salary;
run;
p102d01
10
10
SAS Program Example

This PROC MEANS step summarizes the Salary variable
in the work.newsalesemps data set.
Job_Title $ Salary;
run;
run;
var Salary;
run;
11
p102d01
11
2.2 Submitting a SAS Program
2.02 Quiz
How does SAS detect the end of each step in this program?
Job_Title $ Salary;
run;
var Salary;
12
12

Objectives
Use SAS Enterprise Guide to open and submit a

SAS program and browse the results.
Use the SAS windowing environment to open and
submit a SAS program and browse the results.
16
16
2-7
2-8
Business Scenario
Orion Star programmers will create and execute SAS
programs and view results in an interactive environment.
They must become familiar with both environments.
SAS Enterprise
Guide
SAS Windowing
Environment
17
17
Submitting a SAS Program: SAS Enterprise Guide

p102d01
Start SAS Enterprise Guide.

Open and submit a SAS program.
Examine the results.
Select output formats.
This demonstration is based on SAS Enterprise Guide 5.1.
Starting SAS Enterprise Guide

1. Start SAS Enterprise Guide. The method that you use varies by customizations in effect at your site.
2. Close the Welcome to SAS Enterprise Guide window by clicking
2-9
Opening and Submitting a SAS Program

1. Open p102d01. To open a SAS program in Enterprise Guide, select File Open Program or
and then select Program. Navigate to your data folder, select the file that you want to
click
include, and click Open. The program is displayed on the Program tab of the workspace.
You can use the Program tab to access and edit existing SAS programs, write new programs, submit
programs, and save programs to a file. The program is color-coded in the Program Editor.
2. To submit the program for execution, select Program Run program-name on <server> or click
Run on the Program tab. The F3 or F8 keys can also be used to submit a program.
2-10
Examining the Log

You should always check the log to make sure that the program ran successfully, even when output is
generated. The log echoes the programming statements that were submitted as well as statements added
by SAS Enterprise Guide. The log also contains notes about the program execution and results, as well as
error and warning messages when appropriate. In this example, there are no warning or error messages.
The statements added by Enterprise Guide are used during processing and are referred to as
wrapper code. To suppress the display of the wrapper code in the log, select Tools Options
Results General and clear Show generated wrapper code in SAS log.
To scroll horizontally in the log, use the horizontal scroll bar. To scroll vertically, use the vertical scroll
bar or use the PAGE UP or PAGE DOWN keys on the keyboard.
Examining the Results

The Results tab displays the program output. It becomes the active tab each time it receives output.
To scroll vertically in the Results window, use the vertical scroll bar or use the PAGE UP or PAGE
DOWN keys on the keyboard.
2-11
Selecting Output Formats

SAS Enterprise Guide 5.1 creates SAS Report output by default. Other formats, including HTML, PDF,
RTF, or Text output, can be requested instead of, or in addition to, the default output.
1. To change the result format, select Tools Options Results Results General. The changes
affect all further output, until another change is made.
2. Select the desired result formats: SAS Report and Text output. You can also set an alternate default
format. Click OK.
3. Click the Program tab to display the program and then run the program again.
4. Click Yes to replace the results from the previous run.
5. Click the Results tabs to view the different result formats.
6. Reset the output formats to generate SAS Report or text output (or both) as desired.
Text output is shown throughout this book.
2-12
Idea Exchange
Have you worked with SAS Enterprise Guide? If so, what
do you like about it?
19
19
Submitting a SAS Program: SAS Windowing Environment

p102d01
Start a SAS session.

Open and submit a SAS program.
Examine the results.
Select output formats.
This demonstration is based on SAS 9.3.
Starting a SAS Session

1. Start SAS. The method that you use varies by customizations in effect at your site.
2. Close the Change Notice/Getting Started with SAS window by clicking
or Close.
2-13
Opening and Submitting a SAS Program

1. Open p102d01. To open a SAS program, select File Open Program, or with the Editor window
active, click
. Use the Open window to navigate to the folder that contains your SAS programs.
Select the desired file in the Open Program window and click Open. The program is displayed in an
Editor window.
You use the editor to access and edit existing SAS programs, write new programs, submit programs,
and save programs to a file. Programs are color-coded in the Editor window.
2-14
Alternatively, you can issue the INCLUDE command to open or include a program in your session.
With the Editor window active, type include on the command bar followed by the name of the
program file enclosed in quotation marks. SAS looks for the file in the current or active folder.
To change the current folder, double-click the path indicator

on the
SAS taskbar. A Change Folder dialog box is displayed. Navigate to the appropriate folder
and click OK.
Type include 'p102d01.sas' and press ENTER. The program is displayed in the Editor window.
, or select Run
2. To submit the program for execution, with the Editor window active, click
Submit. Alternatively, you can type submit on the command bar or press F3 or F8.
Examining the Log

The Log window is one of the three primary windows and is open by default. The Log window provides
an audit trail of your SAS session. It echoes the programming statements that were submitted and
includes error or warning messages, as well as any notes about the program execution.
You should always check the log for errors or warnings, even if the program generates output.
1. To access the Log window, click the Log tab near the bottom of the SAS window. You can also select
View Log or submit the LOG command using the command bar.
With the Log window active, use the vertical scroll bar or the Page Up/Page Down keys to scroll up
to the first line of the program that you just submitted. The contents of the Log window are
cumulative, with the most recent information added to the bottom, so be sure you are looking at the
most recent information.
In this example, the Log window contains no warning or error messages.
2. To clear the contents of the Log window, make sure that the window is active and then click
can also issue the CLEAR command or select Edit Clear All.
2-15
. You
Examining the Results

Results Viewer
SAS HTML output, the default output in SAS 9.3, is displayed automatically in the Results Viewer.
Use the vertical scroll bar or the PAGE UP and PAGE DOWN keys on the keyboard to scroll from top to
bottom. You can also enter the TOP and BOTTOM commands on the command bar to scroll vertically.
Use the horizontal scroll bar, if displayed, to scroll side to side.
Selecting Output Formats

SAS 9.3 creates HTML output by default. LISTING output can be requested in addition to or instead of
the default HTML output.
1. Select Tools Options Preferences and click the Results tab.
2. Select or clear the Create Listing and Create HTML boxes to generate both LISTING and HTML
output.
LISTING output is shown throughout this book.
2-16
3. Return to the Editor window for p101d01 and resubmit the program. The output is added to the
Results window, so it now contains two copies of the report. The log also contains information about
the previous and current programs, unless you cleared it before resubmitting the program.
Output Window
1. LISTING output, produced by request only, is displayed in the Output window. Click the Output tab
to view the Output window.
In the SAS windowing environment, the content displayed in the Output window is the last page of
output generated by the most recent program. Scroll up to see the top of the report.
To scroll vertically in the Output window, use the PAGE UP or PAGE DOWN keys on the
keyboard, use the vertical scroll bar, or submit the FORWARD and BACKWARD commands.
You can also use the TOP and BOTTOM commands to scroll vertically in the Output window.
Use the horizontal scroll bar or issue the RIGHT and LEFT commands to scroll horizontally.
2. To clear the contents of the Output window, make sure that window is active and then click
issue the CLEAR command, or select Edit Clear All.
Navigating with the Results Window

You can use the Results window to navigate through your output. It contains bookmarks that can be
expanded and collapsed.
or
2-17
Click a plus sign to expand or collapse individual bookmarks, or right-click the word Results and select
Expand All/Collapse All. Each bookmark contains an icon indicating the file type: HTML or LISTING.
Double-click a bookmark to view the corresponding report in the appropriate viewer.
Accessing SAS Product Documentation on the Web

1. In SAS Enterprise Guide, select Help SAS on the Web SAS Product Documentation. The
SAS Product Documentation web page is displayed.
2. In the SAS windowing environment, select Help SAS on the Web Customer Support Center.
Select Documentation on the Customer Support web page. The SAS Product Documentation web
page is displayed.
2-18

Which environment (or environments) will you use during
this course?
a. SAS windowing environment
b. SAS Enterprise Guide
c. both the SAS windowing environment and
21
Exercises
You can use SAS Enterprise Guide or the SAS windowing environment to complete your exercises.
Select the type of output that you prefer. LISTING output is shown in the exercises.
Level 1
1. Submitting a Program
a. With the appropriate Editor window active, open the SAS program p102e01.
To open the program in SAS Enterprise Guide:
Select Program Open Program, select p102e01, and click Open.
To open the program in SAS:
Select File Open Program and select p102e01.
Windows or UNIX
z/OS (OS/390)
Issue the command: include '.workshop.sascode(p102e01)'.
b. Submit the program for execution. How many rows and columns are in the report?
rows: __________
columns: __________
c. Examine the Log window. Based on the log notes following the DATA step, how many
observations and variables are in the work.country data set?
observations: __________
variables: __________
d. Clear the Log and Output windows (SAS windowing environment only).
2-19
Level 2
2. Exploring Your Environment: SAS Enterprise Guide
a. Customize the appearance and functionality of the Editor by selecting Tools Options
SAS Programs. For example, click Editor Options and click the Appearance tab to modify the
font and font size.
b. Customize the type of output produced by selecting Tools Options Results General. Select
or deselect SAS Report, HTML, PDF, RTF, and Text Output. The default format is
SAS Report. Text output is shown in the course notes.
3. Exploring Your SAS Environment: Windows
a. Customize the appearance and functionality of the Enhanced Editor by selecting Tools
Options Enhanced Editor. For example, click the Appearance tab to modify the font and
font size.
b. Customize the type of output produced by selecting Tools Options Preferences. Click the
Results tab and select or clear Create listing and Create HTML as appropriate. The default
format is HTML. LISTING output is shown in the course notes.
Challenge
4. Enabling and Disabling the Project Log (SAS Enterprise Guide Only)
a. Use the Index tab in SAS Enterprise Guide Help to find information about the project log.
b. Enable and turn on the project log. Run a few programs and view the log to see its contents.
5. Setting Up a Function Key to Clear the Log and Output Windows (SAS Windowing
Environment Only)
a. Issue the KEYS command or select Tools Options Keys to access the KEYS window. The
KEYS window is a secondary window used to browse or change function key definitions.
b. Type the following commands in the Definition column for the F12 key:
clear log; clear output
c. Which key is programmed to submit a KEYS command? _____
d. Close the KEYS window. This saves the key definition in your user profile.
e. Press the F12 key and confirm that the Log and Output windows are cleared.
It is not necessary to clear the Log and Output windows in SAS Enterprise Guide because
this is done automatically each time a program is submitted.
2-20
2.3 SAS Program Syntax

Objectives
Identify the characteristics of SAS statements.

Define SAS syntax rules.
Document a program using comments.
Diagnose and correct a program with errors.
Save the corrected program.
24
24
Business Scenario
Well-formatted, clearly documented SAS programs are
an industry best practice.
25
25
SAS Syntax Rules: Statements

SAS statements
usually begin with an identifying keyword
always end with a semicolon.
Job_Title $ Salary;
run;
run;
var Salary;
run;
p102d01
26
26
2.04 Quiz
How many statements make up this DATA step?
a.
b.
c.
d.
one
three
five
seven
Job_Title $ Salary;
run;
27
27
2-21
2-22
SAS Program Structure

SAS code is free format.
Job_Title $ Salary;run;
proc print data=work.newsalesemps; run;
proc means data =work.newsalesemps;
var Salary;run;
This program is syntactically correct but difficult to read.
p102d02
29
29
SAS Program Structure

Rules for SAS Statements
Statements can begin and end in any column.
A single statement can span multiple lines.
Several statements can appear on the same line.
Unquoted values can be lowercase, uppercase, or
mixed case.
unconventional
data work.newsalesEmps;
formatting
var Salary;run;
30
30
2-23
Recommended Formatting
Begin each statement on a new line.

Use white space to separate words and steps.
Indent statements within a step.
Indent continued lines in multi-line statements.
Job_Title $ Salary;
conventional
run;
formatting

run;
var Salary;
run;
31
31
White space can be blanks, tabs, and new lines. Add them to increase the readability of the code.
Automatic Formatting with SAS Enterprise Guide

p102d02
SAS Enterprise Guide has an auto-formatting tool that applies conventional formatting style to a
SAS program by adding line breaks and indenting each line correctly according to its nesting level.
var Salary;run;
1. Open and submit p102d02. This is the program seen earlier, but with unconventional formatting.
2.
View the log and output to confirm that the program ran successfully.
3. Click the Program tab and select Edit Format Code. The program is formatted automatically.
You can undo the automatic formatting by selecting Edit Undo or pressing the CTRL and
Z keys simultaneously.
4. Select File Save p102d02 to save the modified program file.
To customize the formatting, select Program Editor Option and click the Indenter tab.
2-24
Program Documentation
You can embed comments in a program as explanatory
text.
/* create a temporary data set, newsalesemps */
/* from the text file newemps.csv
*/
/* comment */
*read a comma delimited file;
Job_Title $ Salary;
* comment statement ;
run;
SAS ignores comments during processing but writes them

to the SAS log.
33
* comment;
These comments must be written as separate statements and cannot contain

internal semicolons.
/* comment */ These comments can be any length and can contain semicolons. They cannot be nested.
SAS Comments
This program contains four comments.
*----------------------------------------*
|
This program creates and uses the
|
|
data set called work.newsalesemps.
| n
*----------------------------------------*;
length First_Name $ 12 Last_Name $ 18
Job_Title $ 25;
Job_Title $ Salary /*numeric*/;o
run;
/*
proc print data=work.newsalesemps; p
run;
*/
*var Salary; q
run;
34
p102d03
34
In SAS Enterprise Guide and the Enhanced Editor in SAS, to comment out a block of code using
the /* */ technique, you can highlight the code and then press the CTRL key and the / (forward
slash) key simultaneously. To uncomment a block of code, highlight the block and press the
CTRL, SHIFT, and / keys simultaneously.
2.05 Quiz
Open and examine p102a01. Based on the comments,
which steps do you think will execute and what output will
be generated?
Submit the program. Which steps are executed?
35
Business Scenario
Orion Star programmers must be able to identify and
correct syntax errors in a SAS program.
38
2-25
2-26
Syntax Errors
A syntax error is an error in the spelling or grammar of
a SAS statement. SAS finds syntax errors as it compiles
each SAS statement, before execution begins.
Examples of syntax errors:
misspelled keywords
unmatched quotation marks
missing semicolons
invalid options
39
39
2.06 Quiz
This program includes three syntax errors. One is an
invalid option. What are the other two syntax errors?
daat work.newsalesemps;
Job_Title $ Salary;
run;
proc print data=work.newsalesemps
run;
invalid option
proc means data=work.newsalesemps average min;

var Salary;
run;
40
p102d04
40
Syntax Errors
The Enhanced Editor in SAS and the Program Editor
in SAS Enterprise Guide use the color red to indicate
a potential error in your SAS code.
42
42
Syntax Errors
When SAS encounters a syntax error, it writes a warning or
error message to the log.
ERROR 22-322: Syntax error, expecting one of the following:
a name, a quoted string, (, /, ;, _DATA_, _LAST_,
_NULL_.
WARNING: Data set WORK.TEST was not replaced because this step was
stopped.
You should always check the log to make sure that the
program ran successfully, even if output is generated.
43
43
Diagnosing and Correcting Syntax Errors

p102d04
Submit a SAS program with errors.
Diagnose and correct the errors.
Save the corrected program.
2-27
2-28
Submitting a SAS Program with Errors

infile "&path\newemps.csv" dlm=','
Job_Title $ Salary;
run;
run;
proc means data=work.newsalesemps average max;
var Salary;
run;
1. Open and submit p102d04.
2. Use the vertical scroll bar to view error messages and warnings in the log.
In SAS Enterprise Guide, you can use the Up and Down arrows in the Log menu to find the
previous or next warning or error in the log.
Partial SAS Log

77
---14
WARNING 14-169: Assuming the symbol DATA was misspelled as daat.
78
79
80
81
82
Job_Title $ Salary;
83
run;
NOTE: The infile "s:\workshop\newemps.csv" is:
Filename=s:\workshop\newemps.csv,
NOTE: 71 records were read from the infile "s:\workshop\newemps.csv".
NOTE: The data set WORK.NEWSALESEMPS has 71 observations and 4 variables.
84
85
86
run;
--22
202
ERROR 22-322: Syntax error, expecting one of the following: ;, (, BLANKLINE, DATA, DOUBLE,
HEADING, LABEL, N, NOOBS, OBS, ROUND, ROWS, SPLIT, STYLE, SUMLABEL, UNIFORM,
WIDTH.
2-29
ERROR 202-322: The option or parameter is not recognized and will be ignored.
88
NOTE: The SAS System stopped processing this step because of errors.
89
proc means data=work.newsalesemps average max;
------22
202
ERROR 22-322: Syntax error, expecting one of the following: ;, (, ALPHA, CHARTYPE, CLASSDATA,
CLM, COMPLETETYPES, CSS, CV, DATA, DESCEND, DESCENDING, DESCENDTYPES, EXCLNPWGT,
EXCLNPWGTS, EXCLUSIVE, FW, IDMIN, KURTOSIS, LCLM, MAX, MAXDEC, MEAN, MEDIAN, MIN,
MISSING, MODE, N, NDEC, NMISS, NOLABELS, NONOBS, NOPRINT, NOTHREADS, NOTRAP,
NWAY, ORDER, P1, P10, P20, P25, P30, P40, P5, P50, P60, P70, P75, P80, P90, P95,
P99, PCTLDEF, PRINT, PRINTALL, PRINTALLTYPES, PRINTIDS, PRINTIDVARS, PROBT, Q1,
Q3, QMARKERS, QMETHOD, QNTLDEF, QRANGE, RANGE, SKEWNESS, STACKODS,
STACKODSOUTPUT, STDDEV, STDERR, SUM, SUMSIZE, SUMWGT, T, THREADS, UCLM, USS, VAR,
VARDEF.
ERROR 202-322: The option or parameter is not recognized and will be ignored.
90
var Salary;
91
run;
Diagnosing and Correcting the Errors

The log indicates that SAS
assumed that the keyword DATA was misspelled, issued a warning, and executed the DATA step
interpreted the word RUN as an option in the PROC PRINT statement (because of a missing
semicolon), so the PROC PRINT step did not execute
did not recognize the word AVERAGE as a valid option in the PROC MEANS statement, so the PROC
MEANS step did not execute.
If you are using the Enterprise Guide Editor or the SAS Enhanced Editor, the program remains
in the Editor window. However, if you use the SAS Program Editor, the code disappears with
each submission. Use the RECALL command to recall the program. You can also select Run
Recall Last Submit, or press F4. The program is redisplayed in the Program Editor.
1. Edit the program.

a. Correct the spelling of DATA.
b. Put a semicolon at the end of the PROC PRINT statement.
c. Change the word AVERAGE to MEAN in the PROC MEANS statement.
Job_Title $ Salary;
run;
run;
2-30
proc means data=work.newsalesemps mean max;

var Salary;
run;
2. Submit the program. Verify that it executes without warnings or errors and produces output.
When you modify a program, an asterisk (*) is displayed after the program name in the Editor
window. When you save the program, the asterisk is removed.
Saving the Corrected Program: SAS Enterprise Guide

With the modified program displayed in the Program window, click
program or File Save program As to save your program to a new file.
or select File Save
Saving the Corrected Program: SAS Windowing Environment

With the modified program displayed in the Editor window, click
to save your program to a file.
You can also select File Save or File Save As, or use the FILE command (FILE 'p102d04.sas';).
If the code is not in the Program Editor, recall it before saving the program.
2.07 Quiz
What is the syntax error in this program?
p102d05
45
Correcting Unbalanced Quotation Marks:

p102d05
Submit a SAS program that contains unbalanced quotation marks.
2-31
Diagnose the error.

Correct and resubmit the program.
Submitting a SAS Program That Contains Unbalanced Quotation Marks

In program p102d05, the closing quotation mark for the DLM= option in the INFILE statement is
missing. The color-coding in the editor indicates a very long character string.
Submit the program.
Job_Title $ 25;
infile "&path\newemps.csv" dlm=',;
Job_Title $ Salary;
run;
run;
var Salary;
run;
Diagnosing the Error

View the log. There are no notes in the log because all of the statements after the DLM= option became
part of the quoted string. SAS waits for the closing quotation mark to end the quoted string.
SAS Log
93
94
95
Job_Title $ 25;
96
97
98
Job_Title $ Salary;
99 run;
100
101 proc print data=work.newsalesemps;
102 run;
103
105 proc means data=work.newsalesemps;
106
var Salary;
107 run;
There are no messages of any kind following each step in the log because the steps did not execute. The
absence of messages from SAS typically indicates unbalanced quotation marks. Another indication is the
DATA STEP running message in the banner of the Editor window. This is because the RUN statement
was viewed as part of the character literal and not as a step boundary.
2-32
Stopping a Program
You can stop an executing program in an interactive session.
1. To stop a program in the Windows environment, click
the CTRL and Break keys.
(the break icon) or press
2. Select 1. Cancel Submitted Statements in the Tasking Manager window and click OK.
3. Select Y to cancel submitted statements, and click OK. The program stops executing.
Programmatic Approach to Balancing Quotation Marks

You can balance quotation marks programmatically and eliminate this problem by submitting the
following series of statements:
*';*";run;
SAS maintains a quotation mark counter, and expects an even number of quotation marks. When there is
an odd number, the appropriate quotation mark in the code above serves as the matching quotation mark,
making the count even. Both single quotation marks and double quotation marks are used in this series of
statements, which can fix unbalanced double quotation marks or unbalanced single quotation marks.
Resubmitting a Program
1. Add a closing quotation mark to the DLM= option in the INFILE statement to correct the program.
2. You might choose to insert the statements to balance quotations marks programmatically before your
program code. This is not necessary if you have already stopped the DATA step, but some SAS
programmers include this as the first line of every program that they write. Alternatively, some
programmers add these statements at the end of every program.
*';*";run;
Job_Title $ 25;
Job_Title $ Salary;
run;
run;
var Salary;
run;
3. Resubmit the program and verify that the program ran successfully.
Correcting Unbalanced Quotation Marks: SAS Enterprise

Guide
p102d05
Submit a SAS program that contains unbalanced quotation marks in SAS Enterprise Guide.
Diagnose the error.
Correct and resubmit the program.
1. Open and submit p102d05 in SAS Enterprise Guide.
Job_Title $ 25;
Job_Title $ Salary;
run;
run;
var Salary;
run;
2. View the log. It contains SAS Enterprise Guide wrapper code in addition to the code that you
submitted. Notice the first line of wrapper code.
2-33
2-34
It is very similar to the statement that is entered in SAS to correct the unbalanced quotation marks,
but it also contains a single semicolon, a */ to close a comment, and a QUIT statement. SAS
Enterprise Guide is attempting to fix any potential unbalanced situation from a previously submitted
program.
The program did not execute successfully, but this time there is a warning and an error in the log, and
the DATA step stopped. (Your log might contain different warnings or errors than those shown here.)
WARNING: The quoted string currently being processed has become more than 262 characters long.
You might have unbalanced quotation marks.
36
37
38
39
%LET _CLIENTPROJECTNAME=;
%LET _SASPROGRAMFILE=;
;*';*";*/;quit;run;
____
180
ERROR 180-322: Statement is not valid or it is used out of proper order.
WARNING: The data set WORK.NEWSALESEMPS may be incomplete. When this step was stopped there
were 0 observations and 3 variables.
The first warning above says that a quoted string has exceeded 262 characters. This is not a syntax
error, but the message is an indication that the program might have unbalanced quotation marks.
The error is caused by the QUIT statement. Although it is included to prevent errors, in this case it
caused an error, but the program did not hang as it did in the SAS session.
3. Add a closing quotation mark to the DLM= option in the INFILE statement to correct the program.
4. Resubmit the program, replacing the previous results. Verify that the program runs successfully.
Stopping a Program
SAS Enterprise Guide has built-in features to prevent SAS programs from hanging, but occasionally you
might need to stop or interrupt a program while it is executing. There are two methods:
Method 1: Click the Stop button next to the Run button. It is normally dimmed, but it is red and
selectable while a program is executing.
2-35
Method 2: Click View Task Status to access the Task Status window.
The Task Status window shows the status of all running tasks and programs.
Right-click the task and then select Stop or End SAS Process.
Stop ends the current task while maintaining your connection to the server. If the server is in the
middle of processing a step, it might take a few minutes for the server to stop running the task.
End SAS Process immediately ends the current task and terminates the connection with the server.
Because the connection with the server is terminated, any temporary data is lost.
Exercises
Level 1
6. Diagnosing and Correcting Syntax Errors
b. Submit the program and use the notes in the SAS log to identify the error.
c. Correct the error and resubmit the program.
d. Save the corrected program.
Level 2
2-36
b. Submit the program and use the notes in the SAS log to identify the error.
c. Correct the error and resubmit the program.
d. Save the corrected program.
Challenge
8. Identifying SAS Components
Explore the Internet to find how to list SAS software products licensed at your site.
2.4 Solutions
Solutions to Exercises
1. Submitting a Program
a. Submit the program.
b. The report has 238 rows and 3 columns.
c. The Log window indicates 238 observations and 2 variables.
d. SAS windowing environment only: Click
Clear All in both windows.
in the Log and Output windows, or select Edit
2. Exploring Your Environment: SAS Enterprise Guide

a. Select Tools Options Enhanced Editor. Click the Appearance tab. Select the desired font
size using the Size box in the Font section. Click OK.
b. Select Tools Options Preferences. Click the Results tab. Select the Create listing box in
the Listing section. Select or clear the Create HTML box in the HTML section.
3. Exploring Your SAS Environment: Windows
a. Select Tools Options SAS Programs. Click Editor Options. Click the Appearance tab.
Select the desired font size using the Size box in the Font section. Click OK.
b. Select Tools Options Results - Results General. Select or clear the result formats:
SAS Report, HTML, PDF, and RTF. If SAS Report and HTML are both selected, you can set
one as the default using the Default box.
4. Enabling and Disabling the Project Log (SAS Enterprise Guide Only)
a. Click Help SAS Enterprise Guide Help. Click the Index tab and type Project Log in the text
box. Double-click project log from the list of topics.
b. To enable the Project Log, click the Project Log tab in the workspace, or select View
Project Log. When the Project Log is enabled, you can turn it on and off and clear it using the
buttons on the Project Log toolbar.
2.4 Solutions
2-37
5. Setting Up a Function Key to Clear the Log and Output Windows (SAS Windowing
Environment Only)
a. Submit the Keys command or select Tools Options Keys.
b. Add the commands to the Definition column for the F12 key:
c. Which key is programmed to submit a KEYS command? the F9 key

d. Close the Keys window.
e. Press the F12 key to clear the Log and Output windows.
Only the first three characters are needed for each command, so the following are
alternate ways to define the F12 key: clear log; clear out or cle log; cle out

The program contains a misspelled word. Change prnt to print and resubmit.
The VAR statement is missing a semicolon, so SAS thinks RUN is a variable in the VAR statement.
The error indicates that the variable RUN does not exist. The PROC PRINT step might continue to
run, depending on the environment. If so, terminate the program, type the semicolon, and resubmit.
proc print data=work.country2;
var Country_Code Country_Name;
run;
8. Identifying SAS Components
PROC SETINIT writes the names and expiration dates of all licensed SAS products to the log.
proc setinit;
run;
Partial SAS Log
Product expiration dates:
---Base Product
---SAS/STAT
31DEC2012
31DEC2012
2-38

---SAS/GRAPH
31DEC2012
Solutions to Student Activities (Polls/Quizzes)
2.01 Quiz Correct Answer

How many steps are in this program? three
Job_Title $ Salary;
run;
DATA Step
PROC Step

run;
var Salary;
run;
PROC Step
p102d01
8
8

How does SAS detect the end of each step in this program?
Job_Title $ Salary;
run;
var Salary;
13
The DATA step ends at the RUN statement. The PROC PRINT
step ends at the PROC MEANS statement. The end of the
PROC MEANS step might or might not be detected. A RUN
statement is recommended.
13
2.4 Solutions

How many statements make up this DATA step?
a.
b.
c.
d.
one
three
five
seven
work.NewSalesEmps;
infile "&path\newemps.csv"
'newemps.csv' dlm=',';
dlm=',';
Job_Title $ Salary;
run;
28
28

Open and examine p102a01. Based on the comments,
which steps do you think will execute and what output will
be generated?
Submit the program. Which steps are executed?
The DATA step executes and creates an output

data set.
The PROC PRINT step executes and produces a
report.
The PROC MEANS step is commented out and
therefore does not execute.
36
2-39
2-40

This program includes three syntax errors. One is an
invalid option. What are the other two syntax errors?
misspelled
length keyword
First_Name $ 12
Job_Title $ Salary;
missing
run;
semicolon
run;
invalid option
proc means data=work.newsalesemps average min;

var Salary;
run;
p102d04
41
41

What is the syntax error in this program?
46
The program contains unbalanced quotation marks in

the DLM= option in the INFILE statement.
p102d05
Chapter 3 Accessing Data

3.1
Examining SAS Data Sets .............................................................................................. 3-3

Exercises .............................................................................................................................. 3-12
3.2
Accessing SAS Libraries ............................................................................................. 3-13

Demonstration: Browsing SAS Libraries: SAS Enterprise Guide ........................................ 3-25
Demonstration: Browsing SAS Libraries: SAS Windowing Environment ............................ 3-29
Exercises .............................................................................................................................. 3-33
3.3
Solutions ....................................................................................................................... 3-34

3-2
3.1 Examining SAS Data Sets

Objectives
Define the components of a SAS data set.

Browse the descriptor portion of a SAS data set
using the CONTENTS procedure.
Browse the data portion of a SAS data set using
the PRINT procedure.
Define a SAS variable.
Define a missing value.
Define a SAS date value.
3
3
Business Scenario
Many SAS data sets related to the Orion Star project
already exist. The programmers need to know how to
display the structure and contents of the data sets.
SAS Data Set
Report
4
4
3-3
3-4
What Is a SAS Data Set?

A SAS data set is a specially structured data file that
SAS creates and that only SAS can read. A SAS data set
is a table that contains observations and variables.
observations
variables
SAS Data Set Terminology

SAS Terminology
Database Terminology
SAS Data Set
Table
Observation
Row
Variable
Column
6
6
SAS Data Set Terminology

A SAS data set contains a descriptor portion and
a data portion.
SAS Data Set
descriptor
portion
data
portion
7
7
Descriptor Portion
The descriptor portion contains the following metadata:
general properties (such as data set name and
number of observations)
variable properties (such as name, type, and length)
Partial work.newsalesemps
Data Set Name WORK.NEWSALESEMPS
Engine
V9
Created
Mon, Feb 27, 2012 01:28 PM
Observations
71
Variables
4
...
First_Name
Last_Name
Job_Title Salary
$ 12
$ 18
$ 25
N8
general
properties
variable
properties
3-5
3-6
Browsing the Descriptor Portion

Use PROC CONTENTS to display the descriptor portion
of a SAS data set.
proc contents data=work.newsalesemps;
run;
PROC CONTENTS DATA=SAS-data-set;
RUN;
p103d01
9
9
Viewing the Output

Partial PROC CONTENTS Output
The CONTENTS Procedure
Data Set Name
Member Type
Engine
Created
Last Modified
Protection
Data Set Type
WORK.NEWSALESEMPS
DATA
V9
Mon, Feb 27, 2012 01:28:51 PM
Mon, Feb 27, 2012 01:28:51 PM
Observations
Variables
Indexes
Observation Length
Deleted Observations
Compressed
Sorted
71
4
0
64
0
NO
NO
Engine/Host Dependent Information

...
Alphabetic List of Variables and Attributes
#
Variable
Type
Len
1
3
2
4
First_Name
Job_Title
Last_Name
Salary
Char
Char
Char
Num
12
25
18
8
10
10
The default PROC CONTENTS report has three parts:

general data set properties
engine/host information
variable properties
3.01 Quiz
How many observations are in the data set
work.donations?
Retrieve program p103a01.

After the DATA step, add a PROC CONTENTS step
to view the descriptor portion of work.donations.
Submit the program and review the results.
11
11
Data Portion
The data portion of a SAS data set contains the data
values, which are either character or numeric.
First_Name
Satyakam
Monica
Kevin
Petrea
Last_Name
Denny
Kletschkus
Lyon
Soltau
Job_Title
Sales Rep. II
Sales Rep. IV
Sales Rep. I
Sales Rep. II
character values
Salary
26780
30890
26955
27440
variable
names
data
values
numeric
values
13
13
Variable names are part of the descriptor portion, not the data portion.
3-7
3-8
Browsing the Data Portion

Use PROC PRINT to display the data portion of
a SAS data set.
run;
PROC PRINT DATA=SAS-data-set;
RUN;
p103d02
14
14
Viewing the Output

Partial PROC PRINT Output
Obs
1
2
3
4
5
First_Name
Last_Name
Satyakam
Monica
Kevin
Petrea
Marina
Denny
Kletschkus
Lyon
Soltau
Iyengar
Job_Title
Sales
Sales
Sales
Sales
Sales
Rep.
Rep.
Rep.
Rep.
Rep.
II
IV
I
II
III
Salary
26780
30890
26955
27440
29715
15
15
3-9
SAS Variable Names

SAS variable names
can be 1 to 32 characters long.
must start with a letter or underscore. Subsequent
characters can be letters, underscores, or numerals.
can be uppercase, lowercase, or mixed case.
are not case sensitive.
_score2_
Salary
cust_ID
month1
FirstName
16
16
The same naming rules apply to SAS data set names.
3.02 Multiple Answer Poll

Which variable names are invalid?
a.
b.
c.
d.
e.
f.
g.
data5mon
5monthsdata
data#5
five months data
five_months_data
FiveMonthsData
fivemonthsdata
17
17
A variable name can contain special characters if you place the name in quotation marks and immediately
follow it with the letter N (for example, 'Flight#'n). This is called a SAS name literal. In order to use SAS
name literals as variable names, the VALIDVARNAME= option must be set to ANY.
options validvarname=any;
This setting is the default in SAS Enterprise Guide.
3-10
Data Types
A SAS data set supports two types of variables.
Character variables
can contain any value: letters, numerals, special
characters, and blanks
range from 1 to 32,767 characters in length
have 1 byte per character.
Numeric variables
store numeric values using floating point or binary
representation
have 8 bytes of storage by default
can store 16 or 17 significant digits.
19
Missing Data Values

Missing values are valid values in a SAS data set.
First_Name
Monica
Kevin
Petrea
Last_Name
Kletschkus
Lyon
Soltau
Job_Title
Sales Rep. IV
Sales Rep. I
A blank represents a
missing character value.
Salary
.
26955
27440
A period represents a
missing numeric value.
A value must exist for every variable in every

observation.
20
20
By default, a period is used for a missing numeric value. This default can be altered with the
MISSING= SAS system option.
3-11
SAS Date Values

SAS stores calendar dates as numeric values.
01JAN1959
-365
01JAN1960
store
01JAN1961
366
display
01/01/1959
01/01/1960
01/01/1961
A SAS date value is stored as the number of days

between January 1, 1960, and a specific date.
21
21
SAS can perform calculations on dates starting from 1582 A.D.

SAS can read either two- or four-digit year values. If SAS encounters a two-digit year, the
YEARCUTOFF= system option is used to specify to which 100-year span the two-digit year should be
attributed. For example, by setting the YEARCUTOFF= option to 1950, the 100-year span from 1950 to
2049 is used for two-digit year values.
3.03 Quiz
What is the numeric value for todays date?
Submit program p103a02.
View the output to retrieve the current date as a
SAS date value (that is, a numeric value referencing
January 1, 1960).
22
22
3-12
Exercises
If you restarted your SAS session since the last exercise, open and submit the libname.sas program
found in the data folder.
Level 1
1. Examining the Data Portion of a SAS Data Set
a. Retrieve the starter program p103e01.
b. After the PROC CONTENTS step, add a PROC PRINT step to display all observations,
all variables, and the Obs column for the data set named work.donations.
c. Submit the program to create the PROC PRINT report below. The results contain 124
observations.
Obs
1
2
3
...
123
124
Employee_
ID
Qtr1
Qtr2
Qtr3
Qtr4
Total
120265
120267
120269
.
15
20
.
15
20
.
15
20
25
15
20
25
60
80
121145
121147
35
10
35
10
35
10
35
10
140
40
Level 2
2. Examining the Descriptor and Data Portions of a SAS Data Set
b. After the DATA step, add a PROC CONTENTS step to display the descriptor portion of
work.newpacks.
c. Submit the program and answer the following questions:
How many observations are in the data set? ______________________________________
How many variables are in the data set? _________________________________________
What is the length (byte size) of the variable Product_Name? _____________________
d. After the PROC CONTENTS step, add a PROC PRINT to display the data portion of
work.newpacks.
Submit the program to create the following PROC PRINT report:
Obs
1
Supplier_Name
Top Sports
Supplier_
Country
DK
Product_Name
Black/Black
3.2 Accessing SAS Libraries

2
3
...
14
15
Top Sports
Top Sports
DK
DK
X-Large Bottlegreen/Black
Comanche Women's 6000 Q Backpack. Bark
Luna sastreria S.A.

Miller Trading Inc
ES
US
Hammock Sports Bag

Sioux Men's Backpack 26 Litre.
3-13
Challenge
3. Working with Times and Datetimes
a. Retrieve and submit the starter program p103e03.
b. Notice the values of CurrentTime and CurrentDateTime in the PROC PRINT output.
c. Use the SAS Help facility or product documentation to investigate how times and datetimes are
stored in SAS.
d. Complete the following sentences:
A SAS time value represents the number of __________________________________________.
A SAS datetime value represents the number of _______________________________________.

Objectives
Explain the concept of a SAS library.

State the difference between a temporary library and a
permanent library.
Assign a library reference name to a SAS library using
a LIBNAME statement.
Investigate a SAS library programmatically and
interactively.
Access a data set in a user-created permanent library.
27
27
3-14
Business Scenario
Orion Star programmers need to access existing
SAS data sets, so they need to understand how the
data sets are stored in SAS.
28
SAS Libraries
SAS data sets are stored in SAS libraries. A SAS library is
a collection of SAS files that are referenced and stored as
a unit.
Each file is a member of the library.

29
SAS Libraries
You can think of a SAS library as a drawer in a filing
cabinet and a SAS data set as one of the files in the
drawer.
data set
libraries
30
30
How SAS Libraries Are Defined

When a SAS session starts, SAS automatically creates
one temporary and at least one permanent SAS library
that you can access. These libraries are open and ready
to be used.
temporary library
permanent library
work
sashelp
31
31
3-15
3-16
Assigning a Libref
Regardless of the operating system that you use, you
refer to a SAS library by a logical name called a library
reference name, or libref.
work
libref
sashelp
A libref is a shortcut to the library.
32
32
Temporary Library
Work is a temporary library where you can store and
access SAS data sets for the duration of the SAS session.
It is the default library.
work
SAS deletes the work library and its contents when

the session terminates.
33
Permanent Libraries
Sashelp is a permanent library that contains sample
SAS data sets you can access during your SAS session.
sashelp
34
Permanent Libraries
Sasuser is a permanent library that you can use to store
and access SAS data sets in any SAS session.
sasuser
SAS data sets in permanent libraries are saved after your

SAS session terminates.
35
3-17
3-18
Accessing SAS Data Sets

All SAS data sets have a two-level name that consists of
the libref and the data set name, separated by a period.
libref
data set name
work.newsalesemps
sashelp.class
libref
libref.data-set-name
data set name
When a data set is in the temporary work library, you can

use a one-level name (for example, newsalesemps).
36
Business Scenario
Orion Star programmers need to access and view
SAS data sets that are stored in a permanent
user-defined library.
work
sashelp
orion
37
37
User-Defined Libraries
Users can create their own SAS libraries. A user-defined
library
is permanent. Data sets are stored until the user
deletes them.
is implemented within the operating environments file
system.
is not automatically available in a SAS session.
38
38
User-Defined Libraries
Operating
Environment
A SAS library is
Example
Microsoft Windows
A folder
s:\workshop
UNIX
A directory
~/workshop
z/OS (OS/390)
A sequential file
userid.workshop.sasdata
The user must assign a libref to the user-defined library

to make it available in a SAS session.
39
39
3-19
3-20
Accessing a Permanent Library

Step 1 Identify the location of the library.
In this example, s:\workshop, a Microsoft Windows

folder, is used as the SAS library.
Identify the location of your course data.
40
Accessing a Permanent Library

Step 2 Use a SAS LIBNAME statement to associate the
libref with the physical location of the library.
libref
orion
Associate the libref orion with the Windows folder so that

it is available to your SAS session.
41
Rules for naming a libref:

The name must be 8 characters or less.
The name must begin with a letter or underscore followed by letters, underscores, and numerals.
3-21
LIBNAME Statement
The SAS LIBNAME statement is a global SAS statement.
libname orion "s:\workshop";
LIBNAME libref "SAS-library" <options>;
It is not required to be in a DATA step or PROC step.

It does not require a RUN statement.
It executes immediately.
It remains in effect until changed or canceled, or until
the session ends.
Use the location of your course data in your
LIBNAME statement.
42
In the Microsoft Windows environment, an existing folder is used as a SAS library. The LIBNAME
statement cannot create a new folder.
In the UNIX environment, an existing directory is used as a SAS library. The LIBNAME statement
cannot create a new directory.
In the z/OS environment, a sequential file is used as a SAS library. z/OS (OS/390) users can use a SAS
LIBNAME statement, a DD statement, or a TSO ALLOCATE command. These statements and
commands can create a new library.
Viewing the Log

Partial SAS Log
47
NOTE: Libref ORION was successfully assigned as follows:
Engine:
V9
Physical Name: s:\workshop
43
3-22

Which of the following correctly assigns the libref myfiles
to a SAS library in the c:\mysasfiles folder?
a. libname orion myfiles "c:\mysasfiles";
b. libname myfiles "c:\mysasfiles";
c. libref orion myfiles "c:\mysasfiles";
d. libref myfiles "c:\mysasfiles";
44
Browsing a Library
Step 3 You can browse a library interactively in a
SAS or SAS Enterprise Guide session, or
programmatically using the CONTENTS
procedure.
PROC CONTENTS
orion
46
3-23
Browsing a Library Programmatically

Use PROC CONTENTS with the _ALL_ keyword to
generate a list of all SAS files in a library.
proc contents data=orion._all_ nods;
run;
PROC CONTENTS DATA=libref._ALL_ NODS;
RUN;
_ALL_ requests all files in the library.

The NODS option suppresses the individual data set
descriptor information.
NODS can be used only with the keyword _ALL_.
p103d03
47
47
Viewing the Output

Directory
Libref
Engine
Physical Name
Filename
ORION
V9
S:\workshop
S:\workshop
Name
Member
Type
File
Size
1
2
3
CHARITIES
CONSULTANTS
COUNTRY
COUNTRY
CUSTOMER
CUSTOMER_DIM
DATA
DATA
DATA
INDEX
DATA
DATA
9216
5120
17408
17408
33792
33792
4
5
Last Modified
23Aug12:15:58:39
23Aug12:15:58:39
13Oct10:19:04:39
13Oct10:19:04:39
04Nov11:09:52:27
04Nov11:09:52:27
48
48
A member type of DATA indicates a standard SAS data set. INDEX indicates a file that enables SAS to
access observations in the SAS data set quickly and efficiently.
3-24
Accessing a Permanent Data Set

Step 4 After the libref is assigned, you can access
SAS files in the library.
proc print data=orion.country;
run;
PROC PRINT Output
Obs
1
2
3
4
5
6
7
Country
AU
CA
DE
IL
TR
US
ZA
Country_Name
Population
Australia
Canada
Germany
Israel
Turkey
United States
South Africa
20,000,000
.
80,000,000
5,000,000
70,000,000
280,000,000
43,000,000
Country_
ID
160
260
394
475
905
926
801
Continent_
ID
96
91
93
95
95
91
94
Country_Former
Name
East/West Germany
p103d04
49
Viewing the Log

Partial SAS Log
25
26

run;
NOTE: There were 7 observations read from the data set

ORION.COUNTRY.
The libref orion remains in effect until you change or

cancel it, or until you end your SAS session.
50
50
3-25
Changing or Canceling a Libref

To change a libref, submit a LIBNAME statement with the
same libref but a different path.
libname orion "c:\myfiles";
To cancel a libref, submit a LIBNAME statement with the
CLEAR option.
libname orion clear;
51
Browsing SAS Libraries: SAS Enterprise Guide

libname.sas, p103d03
1. In the Server List, select Servers Local Libraries to display the active libraries.
2. Open the program libname.sas.

The first statement creates a macro variable named path and assigns it a full path to the folder
containing the course data. The second statement, a LIBNAME statement, associates the libref, orion,
with the same data location. The LIBNAME statement is commented out.
3. Modify the program by removing the asterisk from the start of the LIBNAME statement.
4. Save the modified program.
5. Submit the program.
3-26
When working with the course data and programs, be sure to run libname.sas at the start of
every Enterprise Guide session to set the path variable and assign the orion libref.
6. Check the log to confirm that the orion libref was assigned. The physical name reflects the location
of your SAS data files and might differ from the name shown below.
15
16
Engine:
V9
7. In the Server List, click Libraries and then click Refresh. The orion library is in the active library
list.
8. Expand ORION to see the list of data sets.
9. Double-click the COUNTRY data set to open it in a data grid.
10. Click
to close the data grid.
11. In the Server List, right-click COUNTRY and select Properties. The Properties window is displayed
with the General tab selected. Click the Columns tab to see information about the variables.
Click
to close the Properties window.
12. Explore the orion library programmatically using the CONTENTS procedure. Open and submit
program p103d03 to generate a list of library members. Partial output is shown below.
run;
Directory
Libref
Engine
Physical Name
Filename
Name
Member
Type
1
2
3
CHARITIES
CONSULTANTS
COUNTRY
COUNTRY
CUSTOMER
DATA
DATA
DATA
INDEX
DATA
ORION
V9
s:\workshop
s:\workshop
File
Size
9216
5120
17408
17408
33792
Last Modified
23Aug12:15:58:39
23Aug12:15:58:39
13Oct10:19:04:39
13Oct10:19:04:39
04Nov11:09:52:27
13. Open a new Program window. Type and submit the following statement to clear the orion libref.
14. Check the log and the Server List to verify that the orion libref was deassigned.
15
NOTE: Libref ORION has been deassigned.
In the Server List, click Libraries and click Refresh. Verify that orion is no longer defined.
3-27
3-28
Assigning a Library Interactively in SAS Enterprise Guide

1. Select Tools Assign Project Library to invoke the Assign Project Library Wizard.
2. Enter the following values into the wizard windows:

Step 1 of 4: Type the name orion. (The name is displayed in uppercase.) Click Next.
Step 2 of 4: Enter the location of the data files or browse to the location. Click Next.
Step 3 of 4: Do not enter any options. Click Next.
Step 4 of 4: Click Finish.
SAS Enterprise Guide submits a LIBNAME statement based on your wizard entries. Check the log
for success. A sample log is shown below.
15
LIBNAME ORION BASE "S:\workshop";
Engine:
BASE
Physical Name: S:\workshop
3. In the Server List, click Libraries and select Refresh. Expand ORION to see its members.
3.05 Poll
The library display in the Server List updates immediately
when a libref is assigned or cleared using SAS Enterprise
Guide.
True
False
53
Browsing SAS Libraries: SAS Windowing Environment

libname, p103d03
1. Click the Explorer tab to activate the Explorer window or select View Contents Only.
2. Double-click Libraries to show all available libraries. The active libraries are displayed.
3. Open libname.sas.
3-29
3-30
The first statement creates a macro variable named path and assigns it a full path to the folder
containing the course data. The second statement, a LIBNAME statement, associates the libref, orion,
with the same data location. The LIBNAME statement is commented out.
4. Modify the program by removing the asterisk from the start of the LIBNAME statement. Save the
modified program.
5. Submit the program.
When working with the course data and programs, be sure to run libname.sas at the start of
every SAS session to set the path variable and assign the orion libref.
6. Check the log to confirm that the orion libref was assigned. The physical name reflects the location
of your SAS data files and might differ from the name shown below.
1. 412 %let path=s:\workshop;
413 libname orion "s:\workshop";
Engine:
V9
7. In the Explorer window, verify that orion is now displayed as an active library.
8. Double-click the Orion library to show all members of that library.
9. Right-click the Sales data set. Select Properties.

The default view provides general information about the data set. You can view information about the
variables in the data set by clicking the Columns tab at the top of the Properties window.
10. Click
3-31
, OK, or Cancel to close the Properties window.
11. Double-click the Sales data set or right-click the file and select Open. The data set orion.sales opens
in a VIEWTABLE window.
Variable labels are displayed by default. You can display variable names instead of variable labels by
selecting View Column Names. In addition to browsing SAS data sets, you can use the
VIEWTABLE window to edit and create data sets, and to customize your view of a SAS data set.
12. Click
to close the VIEWTABLE window.
13. Explore the orion library programmatically. Open and submit p103d03 to generate a list of library
members.
run;
14. View the output. Partial output is shown below.
Directory
Libref
Engine
Physical Name
Filename
Name
Member
Type
1
2
3
CHARITIES
CONSULTANTS
COUNTRY
COUNTRY
DATA
DATA
DATA
INDEX
ORION
V9
S:\workshop
S:\workshop
File
Size
9216
5120
17408
17408
Last Modified
23Aug12:15:58:39
23Aug12:15:58:39
13Oct10:19:04:39
13Oct10:19:04:39
3-32
15. With the Explorer window active, click
to return to the Active Libraries display.
16. Open a new Editor window. Type and submit the following statement to clear the orion libref.
17. Check the log and the Explorer window to verify that the orion libref was deassigned.
15
NOTE: Libref ORION has been deassigned.
Assigning a Library Interactively in the SAS Windowing Environment

1. From the menu bar, select Tools New Library or click
Library window is displayed.
(the New Library tool). The New
2. Do the following in the New Library window:

a. Type the name orion.
b. Enter a full path to the course data or browse to it.
c. Do not enter any options.
d. Click OK.
3-33
Select Enable at startup to have this library defined automatically when a SAS session
starts.
3. No notes are written to the log when this method is used to assign a library. Use the Explorer window
to verify that orion was assigned.

If you end your SAS or Enterprise Guide session and then
need to access an orion data set again, which of the
following must you do?
a. start a new SAS or Enterprise Guide session
b. use Windows Explorer to locate the files
c. create a new folder to be used as a SAS library
d. submit a LIBNAME statement to define a libref
e. use a two-part name to access the data set
56
Exercises
You must submit a LIBNAME statement before beginning this exercise session.
a. Open libname.sas. Modify the program by removing the asterisk from the LIBNAME statement.
b. Submit the modified program.
c. Verify that the libref orion was successfully assigned.
d. Save the program.
Level 1
4. Accessing a Permanent Data Set
3-34
a. Use an interactive facility to explore the orion library and answer the following questions:
How many observations are in the orion.country data set? _______
How many variables are in the orion.country data set? ______
What is the name of the last country in the data set? _____________________________
b. Submit a PROC CONTENTS step to generate a list of all members in the orion library. What is
the name of the last member listed? __________________________________________
Level 2
5. Viewing General Data Set Properties
a. Examine the general data set properties of orion.staff.
b. What sort information is stored for this data set? ____________________________________
Challenge
6. SAS Autoexec File
Use the Help facility or product documentation to investigate the SAS autoexec file and answer the
following questions.
What is the name of the file? __________________________________________________________
What is its purpose? _________________________________________________________________
How is it created? ___________________________________________________________________
How could this be useful in a SAS session? ______________________________________________
3.3 Solutions
1. Examining the Data Portion of a SAS Data Set
data work.donations;
infile "&path\donation.dat";
input Employee_ID Qtr1 Qtr2 Qtr3 Qtr4;
Total=sum(Qtr1,Qtr2,Qtr3,Qtr4);
run;
proc contents data=work.donations;
run;
proc print data=work.donations;
run;
2. Examining the Descriptor and Data Portions of a SAS Data Set
data work.newpacks;
3.3 Solutions
3-35
input Supplier_Name $ 1-20 Supplier_Country $ 23-24

Product_Name $ 28-70;
datalines;
Top Sports
DK
Black/Black
Top Sports
DK
X-Large Bottlegreen/Black
...
Miller Trading Inc
US
Sioux Men's Backpack 26 Litre.
;
proc contents data=work.newpacks;
run;
proc print data=work.newpacks noobs;
var Product_Name Supplier_Name;
run;
How many observations are in the data set? 13
How many variables are in the data set? 3
What is the length (byte size) of the variable Product_Name? 43
3. Working with Times and Datetimes
A SAS time value represents the number of seconds since midnight of the current day.
A SAS datetime value represents the number of seconds between midnight January 1, 1960, and
an hour/minute/second within a specified date.
4. Accessing a Permanent Data Set
a.
How many observations are in the orion.country data set? 7

How many variables? 6
What is the name of the last country in the data set? South Africa
b. Submit a PROC CONTENTS step to generate a list of all members in the orion library.
run;
What is the name of the last member listed? US_SUPPLIERS
5. Viewing General Data Set Properties
proc contents data=orion.staff;
run;
a. Examine the general data set properties of orion.staff.
b. What sort information is stored for this data set? The General Information section indicates
that the data set is sorted. The Variable section indicates that it is sorted by Employee_ID
using the ANSI character set, and has been validated.
3-36
6. SAS Autoexec File

What is the name of the file? autoexec.sas
What is its purpose? It contains SAS statements that are executed immediately after SAS
initializes. These SAS statements can be used to invoke SAS programs automatically, set up
certain variables for use during your SAS session, or set system options.
How is it created? With any text editor, but the recommended method is to use a SAS text editor
(such as the Enhanced Editor window) and save it using the Save As dialog box.
How could this be useful in a SAS session? It can be used to set the path macro variable and to
automatically submit a LIBNAME statement.

How many observations are in the data set
work.donations? 124 observations
input Employee_ID Qtr1 Qtr2 Qtr3 Qtr4;
run;
proc contents data=work.donations;
run;
12
p103a01s
12
3.3 Solutions
3.02 Multiple Answer Poll Correct Answer

Which variable names are invalid?
a.
b.
c.
d.
e.
f.
g.
data5mon
5monthsdata
data#5
five months data
five_months_data
FiveMonthsData
fivemonthsdata
18
18

What is the numeric value for todays date?
The answer depends on the current date.
Example: If the current date is February 27, 2012, the
numeric value is 19050.
23
23
3-37
3-38
3.04 Multiple Choice Poll Correct Answer

Which of the following correctly assigns the libref myfiles
to a SAS library in the c:\mysasfiles folder?
a. libname orion myfiles "c:\mysasfiles";
b. libname myfiles "c:\mysasfiles";
c. libref orion myfiles "c:\mysasfiles";
d. libref myfiles "c:\mysasfiles";
45
3.05 Poll Correct Answer

The library display in the Server List updates immediately
when a libref is assigned or cleared using SAS Enterprise
Guide.
True
False
The library display might or might not update
immediately. Click Libraries Refresh to update
the list.
54
3.3 Solutions
3.06 Multiple Answer Poll Correct Answers

If you end your SAS or Enterprise Guide session and then
need to access an orion data set again, which of the
following must you do?
a. start a new SAS or Enterprise Guide session
b. use Windows Explorer to locate the files
c. create a new folder to be used as a SAS library
d. submit a LIBNAME statement to define a libref
e. use a two-part name to access the data set
57
3-39
3-40
Chapter 4
4.1
Producing Detail Reports
Subsetting Report Data .................................................................................................. 4-3

Exercises .............................................................................................................................. 4-24
4.2
Sorting and Grouping Report Data ............................................................................. 4-27

Exercises .............................................................................................................................. 4-38
4.3
Enhancing Reports ....................................................................................................... 4-40

Exercises .............................................................................................................................. 4-48
4.4
Solutions ....................................................................................................................... 4-50

4-2
Chapter 4 Producing Detail Reports
4.1 Subsetting Report Data

Objectives
Create a default PROC PRINT report.

Select variables with a VAR statement.
Calculate totals with a SUM statement.
Select observations with a WHERE statement.
Define a date constant.
Identify observations with an ID statement.
3
3
Business Scenario
Orion Star management wants a report that displays the
names, salaries, and a salary total for all sales employees.
orion.sales
Obs
1
2
3
Last_Name
xxxxxxx
xxxxxxx
xxxxxxx
PROC PRINT
First_Name
xxxxxxxxxx
xxxxxxxxxx
xxxxxxxxxx
Salary
99999
99999
99999
----99999
4-3
4-4
PRINT Procedure
By default, PROC PRINT displays all observations, all
variables, and an Obs column on the left side.
proc print data=orion.sales;
run;
Obs
Employee_ID
1
2
3
4
5
120102
120103
120121
120122
120123
First_
Name
Last_Name
Tom
Wilson
Irenie
Christina
Kimiko
Zhou
Dawes
Elvish
Ngan
Hotstone
Gender Salary
M
M
F
F
F
108255
87975
26600
27475
26190
Job_Title
Sales
Sales
Sales
Sales
Sales
Manager
Manager
Rep. II
Rep. II
Rep. I
Country
AU
AU
AU
AU
AU
Birth_
Date
Hire_
Date
3510
-3996
-5630
-1984
1732
10744
5114
5114
6756
9405
Statements and options can be added to the PRINT

procedure to modify the default behavior.
p104d01
5
5
The columns are listed, left to right, in the order the variables are stored in the data set.
VAR Statement
The VAR statement selects variables to include in the
report and specifies their order.

var Last_Name First_Name Salary;
run;
VAR variable(s);

Obs
1
2
3
4
5
Last_Name
First_
Name
Salary
Zhou
Dawes
Elvish
Ngan
Hotstone
Tom
Wilson
Irenie
Christina
Kimiko
108255
87975
26600
27475
26190
p104d01
SUM Statement
The SUM statement calculates and displays report totals
for the requested numeric variables.

sum Salary;
run;
SUM variable(s);
Obs
1
2
3
164
165
Last_Name
Zhou
Dawes
Elvish
...
Capachietti
Lansberry
First_
Name
Salary
Tom
Wilson
Irenie
108255
87975
26600
Renee
Dennis
83505
84260
=======
5141420
p104d01
Viewing the Log

Partial SAS Log
84
85
86
87

sum salary;
run;
NOTE: There were 165 observations read from the data set ORION.SALES.
The order of statements in a SAS procedure is

usually not important.
8
8
4-5
4-6
Business Scenario
Orion Star management wants a report that displays the
names and salaries of the sales employees earning less
than $25,500. Suppress the Obs column.
orion.sales
Last_Name
xxxxxxx
xxxxxxx
xxxxxxx
PROC PRINT
First_Name
xxxxxxxx
xxxxxxxx
xxxxxxxx
Salary
25000
20000
23000
WHERE Statement
The WHERE statement selects observations that meet
the criteria specified in the WHERE expression.

where Salary<25500;
run;
WHERE WHERE-expression;
10
p104d02
10
Viewing the Log

Only 7 of the 165 observations from orion.sales were
selected by the WHERE statement.
295
296
297
298

where Salary<25500;
run;
WHERE Salary<25500;
11
11
Viewing the Output

PROC PRINT Output
Obs
Last_
Name
First_
Name
49
50
85
104
111
131
148
Tilley
Barcoe
Anstey
Voron
Polky
Ould
Buckner
Kimiko
Selina
David
Tachaun
Asishana
Tulsidas
Burnetta
Salary
25185
25275
25285
25125
25110
22710
25390
original observation numbers
12
12
4-7
4-8
Suppressing the Obs Column

Use the NOOBS option in the PROC PRINT statement to
suppress the Obs column.
proc print data=orion.sales noobs;
where Salary<25500;
run;
PROC PRINT DATA=SAS-data-set NOOBS;
PROC PRINT Output

Last_
Name
First_
Name
Tilley
Barcoe
Anstey
Voron
Polky
Ould
Buckner
Kimiko
Selina
David
Tachaun
Asishana
Tulsidas
Burnetta
Salary
25185
25275
25285
25125
25110
22710
25390
p104d02
13
13
WHERE Statement
The WHERE expression defines the condition (or
conditions) for selecting observations.
Operands
character constants
numeric constants
date constants
character variables
numeric variables
Operators
symbols that represent a
comparison, calculation,
or logical operation
>
)
LT
SAS functions
special WHERE operators
14
14
Operands
Constants are fixed values.
Character values are enclosed in quotation marks and
are case sensitive.
Numeric values do not use quotation marks or special
characters.
Variables must exist in the input data set.
where Gender='M';
variable
where Salary>50000;
constant
variable
constant
15
15
SAS Date Constant

A SAS date constant is a date written in the following
form: 'ddmmm<yy>yy'd
Examples
'01JAN2000'd
'31Dec11'D
'1jan04'd
'06NOV2000'D
SAS automatically converts a date constant to a

SAS date value.
16
16
4-9
4-10
Comparison Operators
Comparison operators compare a variable with a value
or with another variable.
Symbol
Mnemonic
Definition
EQ
Equal to
^= = ~=
NE
Not equal to
>
GT
Greater than
<
LT
Less than
>=
GE
Greater than or equal
<=
LE
Less than or equal
IN
Equal to one of a list
17
17
The caret (^), tilde (~), and the not sign ( ) all indicate a logical not. Use the character available on your
keyboard, or use the mnemonic equivalent.
Comparison Operators
Examples
where Gender eq ' ';

where Salary ne .;
where Salary>=50000;
where Hire_Date<'01Jan2000'd;
where Country in ('AU','US');
where Country in ('AU' 'US');
where Order_Type in (1,2,3);
The value list in the IN operator must be enclosed in

parentheses and separated by either commas or blanks.
Character values must be enclosed in quotation marks.
18
18
Setup for the Poll

Program p104a01 contains two WHERE statements.
Open and submit the program.
19

Which of the following is true?
a. The program executes, applying both WHERE
conditions successfully.
b. The program fails and an error message is written to
the log.
c. The program executes, but only the first WHERE
condition is applied.
d. The program executes, but only the second WHERE
20
4-11
4-12
Logical Operators
Logical operators combine or modify WHERE
expressions.
where Country='AU' and
Salary<30000;
run;
WHERE WHERE-expression-1 AND | OR
WHERE-expression-n;
p104d03
22
22
Viewing the Log

Partial SAS Log
67
68
69
70

Salary<30000;
run;
WHERE (Country='AU') and (Salary<30000);
23
23
Logical Operator Priority

The operators can be written as symbols or mnemonics,
and parentheses can be added to modify the order of
evaluation.
Symbol
^
Mnemonic Priority
NOT
&
AND
II
OR
III
The NOT operator modifies a condition by finding the

complement of the specified criteria.
where City not in ('London','Rome','Paris');

24
24
NOT modifies a condition by finding the complement of the specified criteria. Use the character
available on your keyboard, or use the mnemonic equivalent.
AND finds observations that satisfy both conditions.
OR finds observations that satisfy one or both conditions.
Logical Operators
Examples
where Country ne 'AU' and Salary>=50000;
where Gender eq 'M' or Salary ge 50000;
where Country='AU' or Country='US';
where Country in ('AU','US');
where Country not in ('AU','US');
equivalent expressions
You should use only one WHERE statement in a step.

25
25
4-13
4-14
4.02 Quiz
Which WHERE statement correctly subsets the numeric
values for May, June, or July and missing character
names?
a. Answer
where Month in (5-7)
and Names=.;
b. Answer
where Month in (5,6,7)
and Names=' ';
c. Answer
where Month in ('5','6','7')
and Names='.';
26
26
Business Scenario
Orion Star management wants a report that lists only the
Australian sales representatives.
orion.sales
Last_Name
First_
Name
xxxxxxxxxx
xxxxxxxxxx
xxxxxxxxxx
xxxxxxxxxx
xxxxxxx
xxxxxxx
xxxxxxx
xxxxxxx
Country
xx
xx
xx
xx
Job_Title
xxxxxxxxxxxx
xxxxxxxxxxxx
xxxxxxxxxxxx
xxxxxxxxxxxx
28
28
Exploring the Data

var Last_Name First_Name Country
Job_Title;
run;
Plested
Wills
George
Body
Highpoint
Magolan
Desanctis
Ridley
Billy
Matsuoka
Vino
Meera
Harry
Julienne
Scott
Cherda
AU
AU
AU
AU
US
US
US
US
Sales
Sales
Sales
Sales
Chief
Sales
Sales
Sales
Rep. II
Rep. III
Rep. II
Rep. III
Sales Officer
Rep. II
Rep. IV
Rep. IV
p104d04
29
29
Subsetting in a PROC PRINT Step

Include a WHERE statement to subset by Country and
Job_Title.

Job_Title;
Job_Title contains 'Rep';
run;
CONTAINS is a special WHERE operator.
30
p104d04
30
4-15
4-16
CONTAINS Operator
The CONTAINS operator selects observations that
include the specified substring.
Equivalent Statements
where Job_Title contains 'Rep';
where Job_Title ? 'Rep';
? can be used instead of the mnemonic.

The position of the substring within the variables
values is not important.
Comparisons made with the CONTAINS operator are
case sensitive.
31
31
Viewing the Output

Last_Name
First_
Name
Elvish
Ngan
Hotstone
Daymond
Hofmeister
Irenie
Christina
Kimiko
Lucian
Fong
Country
AU
AU
AU
AU
AU
Job_Title
Sales
Sales
Sales
Sales
Sales
Rep.
Rep.
Rep.
Rep.
Rep.
II
II
I
I
IV
32
32
Special WHERE Operators

Special WHERE operators are operators that can be
used only in WHERE expressions.
Operator
Definition
Char
Num
CONTAINS
Includes a substring
BETWEEN-AND
An inclusive range
WHERE SAME AND
Augment a WHERE
expression
IS NULL
A missing value
IS MISSING
A missing value
LIKE
Matches a pattern
33
33
Special WHERE operators cannot be used in IF-THEN or subsetting IF statements.
BETWEEN-AND Operator
The BETWEEN-AND operator selects observations in
which the value of a variable falls within an inclusive
range of values.
Examples
where salary between 50000 and 100000;
where salary not between 50000 and 100000;
where Last_Name between 'A' and 'L';
where Last_Name between 'Baker' and 'Gomez';
34
34
4-17
4-18
BETWEEN-AND Operator
Equivalent Statements
where salary between 50000 and 100000;
where salary>=50000 and salary<=100000;
where 50000<=salary<=100000;
35
35
WHERE SAME AND Operator

Use the WHERE SAME AND operator to add more
conditions to an existing WHERE expression.

where Country='AU' and Salary <30000;
where same and Gender='F';
var First_Name Last_Name Gender
Salary Country;
run;
The WHERE SAME AND condition augments the original
condition.
36
p104d05
Viewing the Log

Partial SAS Log
22
23
where Country='AU' and Salary<30000;
24
where also Gender='F';
NOTE: WHERE clause has been augmented.
25
var First_Name Last_Name Gender Salary Country;
26
run;
WHERE (Country='AU') and (Gender='F') and (Salary<30000);
37
37
Viewing the Output

Obs
First_Name
Last_Name
Gender
Salary
Country
3
4
5
9
14
15
19
20
21
23
Irenie
Christina
Kimiko
Sharryn
Fancine
Petrea
Marina
Shani
Fang
Amanda
Elvish
Ngan
Hotstone
Clarkson
Kaiser
Soltau
Iyengar
Duckett
Wilson
Liebman
F
F
F
F
F
F
F
F
F
F
26600
27475
26190
28100
28525
27440
29715
25795
26810
27465
AU
AU
AU
AU
AU
AU
AU
AU
AU
AU
38
38
4-19
4-20
4.03 Quiz
1. Open p104a01b.
2. Change WHERE SAME AND to WHERE ALSO.
3. Submit the program and view the log.
What message is written to the log?
39
IS NULL Operator
The IS NULL operator selects observations in which a
variable has a missing value.
Examples
where Employee_ID is null;
where Employee_ID is not null;
IS NULL can be used for both character and numeric
variables, and is equivalent to the following statements:
where employee_ID=' ';

where employee_ID=.;
41
41
IS MISSING Operator
The IS MISSING operator selects observations in which a
variable has a missing value.
Examples
where Employee_ID is missing;
where Employee_ID is not missing;
IS MISSING can be used for both character and numeric
variables, and is equivalent to the following statements:
where employee_ID=' ';

where employee_ID=.;
42
42
LIKE Operator
The LIKE operator selects observations by comparing
character values to specified patterns. Two special
characters are used to define a pattern:
A percent sign (%) specifies that any number of
characters can occupy that position.
An underscore (_) specifies that exactly one
character can occupy that position.
Examples
where Name like '%N';
where Name like 'T_m';
where Name like 'T_m%';
43
43
When using the LIKE operator:

Consecutive underscores can be specified.
A percent sign and an underscore can be specified in the same pattern.
The operator is case sensitive.
4-21
4-22
4.04 Quiz
Which WHERE statement returns all the observations that
have a first name starting with the letter M for the given
values?
where Name like '_, M_';

a. Answer
where Name like '%, M%';
b. Answer
c. Answer
where Name like '_, M%';
d. Answer
where Name like '%, M_';
Name
Elvish, Irenie
Ngan, Christina
Hotstone, Kimiko
Daymond, Lucian
Hofmeister, Fong
Denny, Satyakam
Clarkson, Sharryn
Kletschkus, Monica
last name, first name
44
44
Business Scenario
The Sales Manager wants a report that includes only
customers who are 21 years old.
orion.customer_dim
Customer_Age=21
Obs Customer_ID
1
2
3
999
999
999
Customer_Name
Customer_
Gender
XXXXXXXXXXXX
XXXXXXXXXXXX
XXXXXXXXXXXX
X
X
X
Customer_
Country
Customer_Group
Customer_
Age_Group
Customer_
Type
XX
XX
XX
XXXXXXXXXXXX
XXXXXXXXXXXX
XXXXXXXXXXXX
XXXXXXXXX
XXXXXXXXX
XXXXXXXXX
XXXXXXXX
XXXXXXXX
XXXXXXXX
47
Subsetting the Data Set

Display the required rows and variables.
proc print data=orion.customer_dim;

where Customer_Age=21;
var Customer_ID Customer_Name
Customer_Gender Customer_Country
Customer_Group Customer_Age_Group
Customer_Type;
run;
The subsetting variable does not need to be included
in the report.
p104d06
48
48
Viewing the Output

In this output, two lines are used for each observation.
PROC PRINT Output
Obs
37
58
66
...
76
Obs
37
58
66
...
76
Customer_ID
Customer_Name
Customer_
Gender
Customer_
Country
Customer_Group
79
11171
46966
Najma Hicks
Bill Cuddy
Lauren Krasowski
F
M
F
US
CA
CA
Orion Club members

Orion Club Gold members
Orion Club members
70210
Alex Santinello
CA
Orion Club members
Customer_
Age_Group
Customer_Type
15-30 years
15-30 years
15-30 years
Orion Club members medium activity

Orion Club Gold members low activity
Orion Club members high activity
15-30 years
Orion
Club members medium activity
The Obs column helps identify observations in a report

that span multiple lines.
49
Report width and wrapping depends on the LINESIZE system option, the ODS destination, or both.
4-23
4-24
ID Statement
The ID statement specifies the variable or variables
to print at the beginning of each row instead of an
observation number.
proc print data=orion.customer_dim;

where Customer_Age=21;
id Customer_ID;
var Customer_Name Customer_Gender
Customer_Country Customer_Group
Customer_Age_Group Customer_Type;
run;
Choose ID variables that uniquely identify
observations.
p104d07
50
50
Note that the ID variable was removed from the VAR statement. If it were not removed, it would be
displayed twice: once as the leftmost column and again in the position specified by the VAR statement.
Viewing the Output

PROC PRINT Output
Customer_ID
79
11171
46966
70079
70187
70210
Customer_ID
79
11171
46966
70079
70187
70210
Customer_Name
Customer_
Gender
Najma Hicks
Bill Cuddy
Lauren Krasowski
Lera Knott
Soberina Berent
Alex Santinello
Customer_
Age_Group
15-30
15-30
15-30
15-30
15-30
15-30
years
years
years
years
years
years
F
M
F
F
F
M
Customer_
Country
US
CA
CA
CA
CA
CA
Customer_Group
Orion
Orion
Orion
Orion
Orion
Orion
Club
Club
Club
Club
Club
Club
members
Gold members
members
members
members
members
Customer_Type
51
51
Exercises
4-25
Level 1
1. Displaying orion.order_fact with the PRINT Procedure
a. Retrieve the starter program p104e01. Run the program and view the output. Observe that there
are 617 observations. Observations might be displayed over two lines, depending on output
settings.
b. Add a SUM statement to display the sum of Total_Retail_Price. The last several lines of the
report are shown below.
Obs
610
611
612
613
614
615
616
617
Order_
Type
1
1
1
1
1
1
1
1
Product_ID
240700100007
240700100017
240700400003
240800100042
240500200016
240500200122
240700200018
220101400130
Quantity
2
2
2
3
3
2
4
2
Total_Retail_
Price
CostPrice_
Per_Unit
$45.70
$19.98
$24.80
$760.80
$95.10
$48.20
$75.20
$33.80
=============
$100,077.46
$9.30
$11.40
$5.60
$105.30
$14.50
$11.50
$10.30
$5.70
Discount
.
40%
.
.
.
.
.
.
c. Add a WHERE statement to select only the observations with Total_Retail_Price more than 500.
Submit the program. Verify that 35 observations were displayed.
What do you notice about the Obs column? ___________________________________________
Did the sum of Total_Retail_Price change to reflect only the subset? ______________________
d. Add an option to suppress the Obs column. Verify that there are 35 observations in the results.
How can you verify the number of observations in the results? ____________________________
e. Add an ID statement to use Customer_ID as the identifying variable. Submit the program. The
results contain 35 observations.
How did the output change? _______________________________________________________
f. Add a VAR statement to display Customer_ID, Order_ID, Order_Type, Quantity, and
Total_Retail_Price.
What do you notice about Customer_ID? ___________________________________________
g. Modify the VAR statement to address the issue with Customer_ID.
Level 2
2. Displaying orion.customer_dim with the PRINT Procedure
a. Write a PRINT step to display orion.customer_dim.
4-26
b. Modify the program to display a subset of orion.customer_dim by selecting only the

observations for customers between the ages of 30 and 40. Also, suppress the Obs column. The
resulting report should contain 17 observations.
c. Add a statement to use Customer_ID instead of Obs as the identifying column. Submit the
program and verify the results.
d. Add a statement to limit the variables to those shown in the report below.
Customer_ID
Customer_Name
4
9
11
...
54655
70201
James Kvarniq
Cornelia Krahl
Elke Wallstab
...
Lauren Marx
Angel Borwick
Customer_
Age
33
33
33
...
38
38
Customer_Type
Orion Club members low activity
Orion Club Gold members medium activity
...
Internet/Catalog Customers
Challenge
3. Producing a Default Listing Report of orion.order_fact
This exercise assumes you are creating LISTING output in the SAS Windowing
Environment.
a. Produce a default listing report of orion.order_fact. The output might wrap onto a second line.
b. Investigate the use of the LINESIZE= SAS system option to adjust the width of the lines. What
are the minimum and maximum values for the LINESIZE= option? ___________________
Submit an OPTIONS statement with LINESIZE= set to the highest allowed value. Resubmit the
step, and observe the horizontal scroll bar if displayed.
Reset the line size to 96 when finished.
c. Another way to create compact output is to request vertical headings. Investigate the HEADING=
option in the PROC PRINT statement, and then experiment with it to generate vertical headings,
and then horizontal headings.
How do you specify vertical headings? ______________________________________________
How do you specify horizontal headings?_____________________________________________
4. Producing a Default Listing Report of orion.product_dim
This exercise assumes you are creating LISTING output in the SAS Windowing
Environment.
a. Produce a default listing report to display orion.product_dim. Notice that the column width
varies from one page to the next, depending on the width of the values displayed on each page.
b. Investigate the WIDTH= option of the PROC PRINT statement, and modify the program to use
this option with a value of UNIFORM.
4.2 Sorting and Grouping Report Data
4-27
How are the results different when using the WIDTH=UNIFORM option? __________________
______________________________________________________________________________
c. Why might the procedure run more slowly with this option? ______________________________
______________________________________________________________________________
d. How can you save computer resources and still display columns consistently across pages? ____
______________________________________________________________________________

Objectives
Sort the observations in a SAS data set based on the

values of one or more variables.
Display the sorted observations.
Display a data set with report totals and subtotals for
each BY group.
55
55
4-28
Business Scenario
Display observations from orion.sales in ascending order
by the variable Salary.
Employee_ID
999999
999999
999999
Last_Name
xxxxxxxxxx
xxxxxxxxxx
xxxxxxxxxx
Salary
99999
99999
99999
56
56
Creating a Sorted Report

Step 1 Use the SORT procedure to create a new data
set, work.sales, ordering the observations by
the value of Salary.
orion.sales
PROC SORT
work.sales
57
57
Creating a Sorted Report

Step 2 Use the PRINT procedure to display the sorted
data set, work.sales.
work.sales
PROC PRINT
58
58
Step 1: SORT Procedure

The SORT procedure rearranges the observations in the
input data set based on the values of the variable or
variables listed in the BY statement.
proc sort data=orion.sales

out=work.sales;
by Salary;
run;
PROC SORT DATA=input-SAS-data-set

<OUT=output-SAS-data-set>;
BY <DESCENDING> variable<s>;
RUN;
The BY statement in a PROC SORT step specifies the

sort variables and, optionally, the sort order.
59
p104d08
59
4-29
4-30
Viewing the Log

The SORT procedure does not produce a report. Check
the log for errors or warnings.
Partial SAS Log
34
35
36
37

out=work.sales;
by Salary;
run;
ORION.SALES.
NOTE: The data set WORK.SALES has 165 observations and 9
variables.
60
60
Step 2: Viewing the Output

proc print data=work.sales noobs;
var Employee_ID Last_Name Salary;
run;
Employee_ID
121084
121064
121057
...
121143
120102
120261
61
Last_Name
Ould
Polky
Voron
Favaron
Zhou
Highpoint
Salary
22710
25110
25125
95090
108255
243190
p104d08
61
SORT Procedure
The SORT procedure
replaces the original data set or creates a new one
can sort on multiple variables
sorts in ascending (default) or descending order
does not generate printed output.
The input data set is overwritten unless the OUT=
option is used to specify an output data set.
62
62
4.05 Quiz
Which step sorts the observations in a SAS data set and
overwrites the same data set?
a. Answer
proc sort data=work.EmpsAU
out=work.sorted;
by First;
run;
proc sort data=orion.EmpsAU

b. Answer
out=EmpsAU;
by First;
run;
proc sort data=work.EmpsAU;
c. Answer
by First;
run;
63
63
4-31
4-32
Business Scenario
Produce a report that lists sales employees grouped by
Country, in descending Salary order within country.
--------------------------------------Country=AU-------------------------------------------Employee_ID
9999
9999
First_
Name
Last_
Name
xxxx
xxxx
xxxxx
xxxxx
Gender
x
x
Salary
Job_Title
99999
99999
xxxxxx
xxxxxx
Birth_
Date
Hire_
Date
9999
9999
9999
9999
--------------------------------------Country=US-------------------------------------------Employee_ID
First_
Name
Last_
Name
9999
9999
9999
xxxx
xxxx
xxxx
xxxxx
xxxxx
xxxxx
Gender Salary
x
x
x
Job_Title
Birth_
Date
Hire_
Date
xxxxxx
xxxxxx
xxxxxx
9999
9999
9999
9999
9999
9999
99999
99999
99999
66
66
Creating a Grouped Report

Step 1 Use the SORT procedure to group data in a data
set. This scenario requires two variables to be
sorted:
Country
descending Salary within Country
Step 2 Use a BY statement in PROC PRINT to display
the sorted observations grouped by Country.
67
67
Step 1: Sort the Data

Sort the data set to group the observations.

out=work.sales;
by Country descending Salary;
run;
BY <DESCENDING> variable(s);
p104d09
68
68
Specifying Sort Order

The DESCENDING option reverses the sort order for the
variable that immediately follows it. The observations are
sorted from the largest value to the smallest value.
Examples:
by descending Last First;

by Last descending First;
by descending Last descending First;

69
69
4-33
4-34
Specifying Multiple BY Variables
PROC SORT first arranges the data set by the values

of the first BY variable.
PROC SORT
by Country, ascending
PROC SORT then arranges any observations that

have the same value of the first BY variable by the
values of the second BY variable.
PROC SORT
by Salary, descending
This sorting continues for every specified BY variable.
70
70
Step 2: Specify Report Groupings

The BY statement in a PROC PRINT step specifies the
variable or variables to use to form BY groups.

by Country;
run;
BY <DESCENDING> variable(s)>;
71
The variables in the BY statement are called BY

variables.
The observations in the data set must be in order by
the BY variable (or variables).
p104d09
71
Viewing the Output

---------------------------- Country=AU -------------------------Employee_ID
120102
120103
120168
First_Name
Last_Name
Tom
Wilson
Selina
Zhou
Dawes
Barcoe
Gender
Salary
M
M
M
108255
87975
36605
Hire_
Date
...
12205
6575
18567
---------------------------- Country=US -------------------------Employee_ID

120261
121143
121064
First_Name
Last_Name
Gender
Salary
Harry
Louis
Asishana
Highpoint
Favaron
Polky
M
M
M
243190
95090
84260
Hire_
Date
...
11535
15157
13027
72
72
4.06 Quiz
Open and submit p104a02. View the log.
Why did the program fail?
73
4-35
4-36
Business Scenario
Modify the previous report to display selected variables, the
salary subtotal for each country, and the salary grand total.
-------------------- Country=AU -----------------First_Name
Last_Name
XXXX
XXXX
-------Country
XXXXXXX
XXXXXXX
Gender
X
X
Salary
99999
99999
--------999999
-------------------- Country=US ------------------First_Name
Last_Name
XXXXXXX
XXXXXXX
-------Country
XXXXXXX
XXXXXXX
Gender
X
X
Salary
99999
99999
--------999999
=========
9999999
subtotals
grand total
75
75
Generating Subtotals
Use a BY statement and a SUM statement in a PROC
PRINT step.

out=work.sales;
run;
by Country;
sum Salary;
var First_Name Last_Name Gender Salary;
run;
76
p104d10
76
Viewing the Output

----------------- Country=AU ----------------First_Name
Last_Name
Gender
Salary
Tom
Wilson
Daniel
...
Kimiko
-------Country
Zhou
Dawes
Pilgrim
M
M
M
108255
87975
36605
Tilley
25185
------1900015
subtotal for AU
------------------ Country=US -----------------First_Name
Last_Name
Harry
Louis
Dennis
...
Tulsidas
-----------Country
Highpoint
Favaron
Lansberry
M
M
M
243190
95090
84260
Ould
22710
------3241405
=======
5141420
77
Gender
Salary
subtotal for US
grand total
77
Setup for the Poll

Modify the previous report to display only employees
earning less than 25,500. Which WHERE statement (or
statements) will result in the most efficient processing?

out=work.sales;
/* where Salary<25500;*/
run;
by Country;
sum Salary;
/* where Salary<25500;*/
var First_Name Last_Name Gender Salary;
run;
78
p104a03
4-37
4-38

Which WHERE statement (or statements) will result in the
most efficient processing?
a. The WHERE statement in the PROC SORT step.
b. The WHERE statement in the PROC PRINT step.
c. Both WHERE statements are needed.
d. The WHERE statements are equally efficient.
79
Exercises
Level 1
5. Sorting orion.employee_payroll and Displaying the New Data Set
a. Open p104e05. Add a PROC SORT step before the PROC PRINT step to sort
orion.employee_payroll by Salary, placing the sorted observations into a temporary data set
named sort_salary.
b. Modify the PROC PRINT step to display the new data set. Verify that your output matches the
report below.
Obs
Employee_ID
1
2
...
422
423
424
121084
120191
120261
120262
120259
Employee_
Gender
Salary
Birth_
Date
Employee_
Hire_Date
Employee_
Term_Date
Marital_
Status
M
F
22710
24015
3150
1112
12784
17167
.
17347
M
S
3
0
M
M
M
243190
268455
433800
4800
5042
2946
11535
11932
12297
.
.
.
O
M
M
1
2
1
6. Sorting orion.employee_payroll and Displaying Grouped Observations
Dependent
4-39
a. Open p104e06. Add a PROC SORT step before the PROC PRINT step to sort
orion.employee_payroll by Employee_Gender, and within gender by Salary in descending
order. Place the sorted observations into a temporary data set named sort_salary2.
b. Modify the PROC PRINT step to display the new data set with the observations grouped by
Employee_Gender.
-------------------------------------- Employee_Gender=F --------------------------------------
Obs
Employee_ID
Salary
Birth_
Date
Employee_
Hire_Date
Employee_
Term_Date
120260
120719
120661
207885
87420
85495
3258
4770
-400
10532
14641
10227
.
.
17347
M
M
M
2
1
3
120196
120191
24025
24015
10257
1112
17167
17167
17347
17347
S
S
0
0
1
2
3
...
190
191
Marital_
Status
Dependents
-------------------------------------- Employee_Gender=M ---------------------------------------
Obs
Employee_ID
Salary
Birth_
Date
Employee_
Hire_Date
Employee_
Term_Date
Marital_
Status
192
193
...
423
424
120259
120262
433800
268455
2946
5042
12297
11932
.
.
M
M
1
2
120190
121084
24100
22710
10566
3150
17837
12784
18017
.
M
M
2
3
Dependents
Level 2
7. Sorting orion.employee_payroll and Displaying a Subset of the New Data Set
a. Sort orion.employee_payroll by Employee_Gender, and by descending Salary within gender.
Place the sorted observations into a temporary data set named sort_sal.
b. Print a subset of the sort_sal data set. Select only the observations for active employees (those
without a value for Employee_Term_Date) who earn more than $65,000. Group the report by
Employee_Gender, and include a total and subtotals for Salary. Suppress the Obs column.
Display only Employee_ID, Salary, and Marital_Status. The results contain 18 observations.
-------------------------------------- Employee_Gender=F -----------------------------------Marital_
Status
Employee_ID
Salary
120260
120719
...
120677
--------------Employee_Gender
207885
87420
M
M
65555
-----605190
-------------------------------------- Employee_Gender=M ------------------------------------
4-40

Marital_
Status
Employee_ID
Salary
120259
120262
...
120268
--------------Employee_Gender
433800
268455
M
M
76105
------2072410
=======
2677600
Challenge
8. Retaining the First Observation of Each BY Group
a. Sort orion.orders by Customer_ID. Place the sorted observations in a temporary data set.
b. Display the sorted data set. The resulting report should contain 490 observations. Customer_ID
is listed multiple times for customers that placed more than one order.
c. Investigate an option that causes PROC SORT to retain only the first observation in each BY
group.
d. Add the appropriate option to the PROC SORT step to retain only the first observation in each BY
group. The results contain 75 observations with no duplicate values for Customer_ID.
e. Explore the DUPOUT= option to write duplicate observations to a separate output data set.
4.3 Enhancing Reports

Objectives
Include titles and footnotes in a report.

Define descriptive column headings using the
LABEL statement.
Control the use of column headings with the LABEL
and SPLIT= options.
84
84
Business Scenario
Enhance the payroll report by adding titles, footnotes, and
descriptive column headings.
Obs Employee_ID
1
2
3
9999
9999
9999
Last_ Name
xxxxxxxxxx
99999
xxxxxxxxxx
99999
xxxxxxxxxx
99999
Orion Star Sales Staff
Salary Report
Obs Employee ID
1
2
3
Salary
9999
9999
9999
Last Name
Annual Salary
xxxxxxxxxx
xxxxxxxxxx
xxxxxxxxxx
99999
99999
99999
Confidential
85
85
Displaying Titles and Footnotes

Use TITLE and FOOTNOTE statements to enhance the
report.
title1 'Orion Star Sales Staff';

title2 'Salary Report';
footnote1 'Confidential';
TITLEn 'text ';
FOOTNOTEn 'text ';

run;
title;
footnote;
86
p104d11
86
4-41
4-42
Viewing the Output

Salary Report
Obs
Employee_ID
Last_Name
Salary
1
2
3
...
164
165
120102
120103
120121
Zhou
Dawes
Elvish
108255
87975
26600
121144
121145
Capachietti
Lansberry
83505
84260
Confidential
87
87
TITLE Statement
The global TITLE statement specifies title lines for
SAS output.
TITLEn 'text ';
Titles appear at the top of the page.

The default title is The SAS System.
The value of n can be from 1 to 10.
An unnumbered TITLE is equivalent to TITLE1.
Titles remain in effect until they are changed or
canceled, or you end your SAS session.
88
88
FOOTNOTE Statement
The global FOOTNOTE statement specifies footnote lines
for SAS output.
FOOTNOTEn 'text ';
Footnotes appear at the bottom of the page.

No footnote is printed unless one is specified.
The value of n can be from 1 to 10.
An unnumbered FOOTNOTE is equivalent to
FOOTNOTE1.
Footnotes remain in effect until they are changed or
canceled, or you end your SAS session.
89
89
Changing Titles and Footnotes

To change a title line, submit a TITLE statement with the
same number but different text.
replaces a previous title with the same number
cancels all titles with higher numbers
title1 'ABC Company';
title2 'Sales Division';
This statement
changes title 1 and
cancels titles 2 and 3.
Footnotes are changed the same way.
90
90
4-43
4-44
Canceling All Titles and Footnotes
The null TITLE statement cancels all titles.
title;
The null FOOTNOTE statement cancels all footnotes.
footnote;
91
91
Changing and Canceling Titles and Footnotes

PROC PRINT Code
102
Resultant Title(s)
title1 'The First Line';

title2 'The Second Line';
run;
The First Line

The Second Line
title2 'The Next Line';

run;
The First Line

The Next Line
title 'The Top Line';

run;
The Top Line
title3 'The Third Line';

run;
The Top Line

The Third Line
title;
run;
102
4.08 Quiz
Which footnote or footnotes appear in the second
procedure output?
Sales Employees
a. Non Sales Employees
c. Non Confidential
b.
Orion Star
Non Sales Employees
d.
Orion Star
Non Sales Employees
Confidential
footnote1 'Orion Star';

footnote2 'Sales Employees';
run;
footnote2 'Non Sales Employees';
proc print data=orion.nonsales;
run;
103
103
Idea Exchange
Which of the following programs do you prefer and why?
'Orion Star Employees';
a. title
proc print data=orion.staff;
where Gender='F';
var Employee_ID Salary;
run;
105
b.
title 'Orion Star Female Employees';

where Gender='F';
var Employee_ID Salary;
run;
c.
title 'Orion Star Employees';

where Gender='F';
var Employee_ID Gender Salary;
run;
d.
title 'Orion Star Female Employees';

where Gender='F';
var Employee_ID Gender Salary;
run;
105
4-45
4-46
LABEL Statement and Option

Use a LABEL statement and the LABEL option to display
descriptive column headings instead of variable names.
title1 'Orion Star Sales Staff';

proc print data=orion.sales label;
label Employee_ID='Sales ID'
Last_Name='Last Name'
Salary='Annual Salary';
run;
title;
footnote;
LABEL variable-1='label '
variable-n='label ';
p104d12
107
107
LABEL Statement
The LABEL statement assigns descriptive labels to
variables.
A label can be up to 256 characters and include any
characters, including blanks.
Labels are used automatically by many procedures.
The PRINT procedure uses labels only when the
LABEL or SPLIT= option is specified.
108
108
Viewing the Output

Partial PROC PRINT
Output
Orion Star
Sales
Salary Report
Staff
Last Name
Annual
Salary
120102
120103
120121
Zhou
Dawes
Elvish
108255
87975
26600
121144
121145
Capachietti
Lansberry
Obs
Sales ID
1
2
3
...
164
165
83505
84260
Confidential
109
109
SPLIT= Option
The SPLIT= option in PROC PRINT specifies a split
character to control line breaks in column headings.
proc print data=orion.sales split='*';

label Employee_ID='Sales ID'
Last_Name='Last*Name'
Salary='Annual*Salary';
run;
SPLIT='split-character'
The SPLIT= option can be used instead of the LABEL

option in a PROC PRINT step.
110
p104d13
110
4-47
4-48
Viewing the Output

Salary Report
Last
Name
Annual
Salary
120102
120103
120121
Zhou
Dawes
Elvish
108255
87975
26600
121144
121145
Capachietti
Lansberry
Obs
Sales ID
1
2
3
...
164
165
83505
84260
Confidential
111
111
Exercises
Level 1
9. Displaying Titles and Footnotes in a Detail Report
a. Open and submit p104e09 to display all observations for Australian Sales Rep IVs.
b. Add a VAR statement to display only the variables shown in the report below.
c. Add TITLE and FOOTNOTE statements to include the titles and footnotes shown in the report
below.
d. Submit the program and verify the output. The results contain five observations as shown below.
e. Submit a null TITLE and null FOOTNOTE statement to clear all titles and footnotes.
Australian Sales Employees
Senior Sales Representatives
Obs
Employee_ID
7
10
17
41
120125
120128
120135
120159
First_
Name
Last_Name
Fong
Monica
Alexei
Lynelle
Hofmeister
Kletschkus
Platts
Phoumirath
Gender
M
F
M
F
Salary
32040
30890
32490
30765

48
120166
Fadi
Nowd
4-49
30660
Job_Title: Sales Rep. IV
10. Displaying Column Headings in a Detail Report

a. Open and submit p104e10. Modify the program to define and use the following labels:
Variable
Label
Employee_ID
Employee ID
First_Name
First Name
Last_Name
Last Name
Salary
Annual Salary
Submit the program and verify the output.

Entry-level Sales Representatives
Employee ID
121023
121028
121029
...
121138
121140
Gender
Annual
Salary
First Name
Last Name
Shawn
William
Kuo-Chung
Fuller
Smades
Mcelwee
M
M
M
26010
26585
27225
Hershell
Saunders
Tolley
Briggi
M
M
27265
26335
Job_Title: Sales Rep. I
b. Modify the program to use a blank space as the SPLIT= character to generate two-line column
headings. Submit the modified program and verify that two-line column labels are displayed.
Entry-level Sales Representatives
Employee
ID
121023
121028
121029
...
121138
121140
First
Name
Last
Name
Shawn
William
Kuo-Chung
Fuller
Smades
Mcelwee
M
M
M
26010
26585
27225
Hershell
Saunders
Tolley
Briggi
M
M
27265
26335
Gender
Job_Title: Sales Rep. I
Level 2
11. Writing an Enhanced Detail Report
Annual
Salary
4-50
a. Write a program to display a subset of orion.employee_addresses as shown below. The program

should sort the observations by State, City, and Employee_Name and then display the sorted
observations grouped by State. The resulting report should contain 311 observations.
US Employees by State
------------------------------- State=CA --------------------------Employee
ID
120656
120759
121017
121062
121049
Name
Amos, Salley
Apr, Nishan
Arizmendi, Gilbert
Armant, Debra
Bataineh, Perrior
...
City
San
San
San
San
San
Diego
Diego
Diego
Diego
Diego
Zip
Code
92116
92071
91950
92025
92126
4.4 Solutions
1. Displaying orion.order_fact with the PRINT Procedure
proc print data=orion.order_fact noobs;
where Total_Retail_Price>500;
id Customer_ID;
var Order_ID Order_Type Quantity Total_Retail_Price;
sum Total_Retail_Price;
run;
a. Run the program and view the output.
b. Add a SUM statement and verify the resulting sum.
c. What do you notice about the Obs column? The numbers are not sequential. The original
observation numbers are displayed.
Did the sum of Total_Retail Price change to reflect only the subset? Yes
d. If the Obs column is suppressed, how can you verify the number of observations in the results?
Check the log.
e. When the ID statement was added, how did the output change? Customer_ID is the leftmost
column and is displayed on each line for an observation.
f. When the VAR statement is added, what do you notice about Customer_ID? There are two
Customer_ID columns. The first column is the ID field, and a second one is included
because Customer_ID is listed in the VAR statement.
g. Remove the duplicate column by removing Customer_ID from the VAR statement.
2. Displaying orion.customer_dim with the PRINT Procedure
proc print data=orion.customer_dim noobs;
where Customer_Age between 30 and 40;
4.4 Solutions
4-51
id Customer_ID;
var Customer_Name Customer_Age Customer_Type;
run;
3. Producing a Default Listing Report of orion.order_fact (SAS Windowing Environment)
options ls=max;
proc print data=orion.order_fact;
run;
options ls=96;
proc print data=orion.order_fact headings=v;
run;
a. Submit a simple PROC PRINT step.
b. The minimum value for LINESIZE= is 64 and the maximum size is MAX.
Use this statement to reset the line size to 96: options ls=96;
c. HEADINGS=V forces all column headings to display vertically.
HEADINGS=H forces all column headings to display horizontally.
4. Producing a Default Listing Report of orion.product_dim (SAS Windowing Environment)
proc print data=orion.product_dim width=uniform;
run;
a. Submit a simple PROC PRINT step.
b. Add the WIDTH=uniform option. How are the results different? Each column has the same
column width on each page.
c. Why might the procedure run more slowly with this option? With this option, PROC PRINT
must read through the entire data set twice.
d. How can you save computer resources and still display columns consistently across pages? Use a
format on every column to explicitly specify a field width so that PROC PRINT reads the
data only once.
5. Sorting orion.employee_payroll and Displaying the New Data Set
proc sort data=orion.employee_payroll out=work.sort_salary;
by Salary;
run;
proc print data=work.sort_salary;
run;
6. Sorting orion.employee_payroll and Displaying Grouped Observations
proc sort data=orion.employee_payroll out=work.sort_salary2;
by Employee_Gender descending Salary;
run;
4-52
proc print data=work.sort_salary2;

by Employee_Gender;
run;
7. Sorting orion.employee_payroll and Displaying a Subset of the New Data Set
proc sort data=orion.employee_payroll out=work.sort_sal;
by Employee_Gender descending Salary;
run;
proc print data=work.sort_sal noobs;
by Employee_Gender;
sum Salary;
where Employee_Term_Date is missing and Salary>65000;
var Employee_ID Salary Marital_Status;
run;
8. Retaining the First Observation of Each BY Group
proc sort data=orion.orders out=work.custorders nodupkey
dupout=work.duplicates;
by Customer_ID;
run;
title 'Unique Customers';
proc print data=work.custorders;
run;
title 'Duplicate Customer Observations';
proc print data=work.duplicates;
run;
title;
9. Displaying Titles and Footnotes in a Detail Report
title1 'Australian Sales Employees';
title2 'Senior Sales Representatives';
footnote1 'Job_Title: Sales Rep. IV';
where Country='AU' and Job_Title contains 'Rep. IV';
var Employee_ID First_Name Last_Name Gender Salary;
run;
title;
footnote;
10. Display Column Headings in a Detail Report
a.
title 'Entry-level Sales Representatives';
footnote 'Job_Title: Sales Rep. I';
proc print data=orion.sales noobs label;
where Country='US' and Job_Title='Sales Rep. I';
4.4 Solutions
label Employee_ID="Employee ID"

First_Name="First Name"
Last_Name="Last Name"
Salary="Annual Salary";
run;
title;
footnote;
b.
title 'Entry-level Sales Representatives';
footnote 'Job_Title: Sales Rep. I';
proc print data=orion.sales noobs split=' ';
where Country='US' and Job_Title='Sales Rep. I';
label Employee_ID="Employee ID"
First_Name="First Name"
Last_Name="Last Name"
Salary="Annual Salary";
run;
title;
footnote;
11. Writing an Enhanced Detail Report
proc sort data=orion.employee_addresses out=work.address;
where Country='US';
by State City Employee_Name;
run;
title "US Employees by State";
proc print data=work.address noobs split=' ';
var Employee_ID Employee_Name City Postal_Code;
label Employee_ID='Employee ID'
Employee_Name='Name'
Postal_Code='Zip Code';
by State;
run;
4-53
4-54

Which of the following is true?
a. The program executes, applying both WHERE
conditions successfully.
b. The program fails and an error message is written to
the log.
c. The program executes, but only the first WHERE
d. The program executes, but only the second WHERE
182 proc print data=orion.sales;
183
where Country='AU';
184
where Salary<30000;
NOTE: WHERE clause has been replaced.
185 run;
21
WHERE Salary<30000;

Which WHERE statement correctly subsets the numeric
values for May, June, or July and missing character
names?
a. Answer
where Month in (5-7)
and Names=.;
b. Answer
where Month in (5,6,7)
and Names=' ';
c. rwhere Month in ('5','6','7')
and Names='.';
27
27
4.4 Solutions

WHERE ALSO results in the same message:
WHERE clause has been augmented.
27
28
where Country='AU' and Salary<30000;
29
where also Gender='F';
NOTE: WHERE clause has been augmented.
30
var First_Name Last_Name Gender Salary Country;
31
run;
WHERE (Country='AU') and (Gender='F') and (Salary<30000);
40

Which WHERE statement returns all the observations that
have a first name starting with the letter M for the given
values?
where Name like '_, M_';
a. Answer
where Name like '%, M%';
b. Answer
c. Answer
where Name like '_, M%';
d. Answer
where Name like '%, M_';
Name
Elvish, Irenie
Ngan, Christina
Hotstone, Kimiko
Daymond, Lucian
Hofmeister, Fong
Denny, Satyakam
Clarkson, Sharryn
Kletschkus, Monica
last name, first name
45
45
4-55
4-56

Which step sorts the observations in a SAS data set and
overwrites the same data set?
a. Answer
proc sort data=work.EmpsAU
out=work.sorted;
by First;
run;
proc sort data=orion.EmpsAU
b. Answer
out=EmpsAU;
by First;
run;
proc sort data=work.EmpsAU;
c. Answer
by First;
run;
64
64

Open and submit p104a02. View the log.
Why did the program fail?
The input data set was not sorted by Gender.
188
189
190
191

out=work.sorted;
by Country Gender;
run;
NOTE: The data set WORK.SORTED has 165 observations and 9 variables.
192
193
194
195
proc print data=work.sorted;

by Gender;
run;
ERROR: Data set WORK.SORTED is not sorted in ascending sequence. The current
BY group has Gender = M and the next BY group has Gender = F.
NOTE: There were 64 observations read from the data set WORK.SORTED.
74
4.4 Solutions

Which WHERE statement (statements) will result in the
most efficient processing?
a. The WHERE statement in the PROC SORT step.
b. The WHERE statement in the PROC PRINT step.
c. Both WHERE statements are needed.
d. The WHERE statements are equally efficient.
Subsetting in the PROC SORT is more efficient. It
selects and sorts only the required observations.
Be sure to use the OUT= option when subsetting
in a PROC SORT or you will overwrite your
original data set with the subset.
p104a03s
80

Which footnote or footnotes appear in the second
procedure output?
Sales Employees
a. Non Sales Employees
c. Non Confidential
b.
Orion Star
Non Sales Employees
d.
Orion Star
Non Sales Employees
Confidential
footnote1 'Orion Star';

footnote2 'Sales Employees';
run;
104
footnote2 'Non Sales Employees';

run;
104
4-57
4-58
Chapter 5 Formatting Data Values

5.1
Using SAS Formats ........................................................................................................ 5-3

Exercises .............................................................................................................................. 5-10
5.2
User-Defined Formats .................................................................................................. 5-12

Demonstration: Defining and Using a Numeric Format ....................................................... 5-18
Exercises .............................................................................................................................. 5-23
5.3
Solutions ....................................................................................................................... 5-25

5-2
5.1 Using SAS Formats

Objectives
Describe SAS formats.

Apply SAS formats with the FORMAT statement.
Business Scenario
Enhance the appearance of variable values in reports.
Last_Name
First_
Name
Country
Zhou
Dawes
Elvish
Tom
Wilson
Irenie
AU
AU
AU
Last_Name
First_
Name
Country
Zhou
Dawes
Elvish
Tom
Wilson
Irenie
AU
AU
AU
Job_Title
Salary
Hire_
Date
Sales Manager
Sales Manager
Sales Rep. II
108255
87975
26600
12205
6575
6575
Job_Title
Sales Manager
Sales Manager
Sales Rep. II
Salary
Hire_Date
$108,255
$87,975
$26,600
06/01/1993
01/01/1978
01/01/1978
5-3
5-4
SAS Formats
SAS formats can be used in a PROC step to change how
values are displayed in a report.
PROC Step
FORMAT
statement
variable values
5
5
FORMAT Statement
The FORMAT statement associates a format with a variable.

format Salary dollar8. Hire_Date mmddyy10.;
Job_Title Salary Hire_Date;
run;
FORMAT variable(s) format ;
p105d01
Viewing the Output

Last_Name
First_
Name
Zhou
Dawes
Elvish
Ngan
Hotstone
Tom
Wilson
Irenie
Christina
Kimiko
Country
AU
AU
AU
AU
AU
Job_Title
Sales
Sales
Sales
Sales
Sales
Manager
Manager
Rep. II
Rep. II
Rep. I
Salary
Hire_Date
$108,255
$87,975
$26,600
$27,475
$26,190
06/01/1993
01/01/1978
01/01/1978
07/01/1982
10/01/1989
DOLLAR8.
MMDDYY10.
What Is a Format?
A format is an instruction to write data values.
A format changes the appearance of a variables value
in a report.
The values stored in the data set are not changed.
SAS Date
10866
Numeric
5950.35
01/10/1989
10Jan1989
5,950.35
$5,950.35
5-5
5-6
SAS Formats
SAS formats have the following form:
<$>format<w>.<d>
Indicates a character format.
format
Names the SAS format.
Specifies the total format width, including decimal places

and special characters.
Is required syntax. Formats always contain a period (.)

as part of the name.
Specifies the number of decimal places to display in

numeric formats.
SAS Formats
Selected SAS formats:
Format
10
Definition
$w.
Writes standard character data.
w.d
Writes standard numeric data.
COMMAw.d
Writes numeric values with a comma that separates every

three digits and a period that separates the decimal fraction.
DOLLARw.d
Writes numeric values with a leading dollar sign, a comma

that separates every three digits, and a period that
separates the decimal fraction.
COMMAXw.d
Writes numeric values with a period that separates every

three digits and a comma that separates the decimal
fraction.
EUROXw.d
Writes numeric values with a leading euro symbol (), a

period that separates every three digits, and a comma that
separates the decimal fraction.
SAS Format Examples

Selected SAS formats:
Format
$4.
Stored Value
Displayed Value
Programming
Prog
12.
27134.5864
27135
12.2
27134.5864
27134.59
COMMA12.2
27134.5864
27,134.59
DOLLAR12.2
27134.5864
$27,134.59
COMMAX12.2
27134.5864
27.134,59
EUROX12.2
27134.5864
27.134,59
11
SAS Format Examples

If the format width is not large enough to accommodate
a numeric value, the displayed value is automatically
adjusted to fit the width.
Format
Stored Value
Displayed Value
DOLLAR12.2
27134.5864
$27,134.59
DOLLAR9.2
27134.5864
$27134.59
DOLLAR8.2
27134.5864
27134.59
DOLLAR5.2
27134.5864
27135
DOLLAR4.2
27134.5864
27E3
12
One aspect of the adjustment is rounding.
5-7
5-8
5.01 Quiz
Use SAS documentation or the SAS Help Facility
to explore the Zw.d numeric format. What is it used for?
Hint: Search for Zw.d or explore Formats by Category.
13
SAS Date Format Examples

SAS date formats display SAS date values in standard
date forms.
Format
Stored Value
Displayed Value
MMDDYY10.
01/01/1960
MMDDYY8.
01/01/60
MMDDYY6.
010160
DDMMYY10.
365
31/12/1960
DDMMYY8.
365
31/12/60
DDMMYY6.
365
311260
15
5-9
SAS Date Format Examples

Additional date formats:
Format
Stored Value
Displayed Value
DATE7.
-1
31DEC59
DATE9.
-1
31DEC1959
WORDDATE.
January 1, 1960
WEEKDATE.
Friday, January 1, 1960
MONYY7.
JAN1960
YEAR4.
1960
16
5.02 Quiz
Which FORMAT statement creates the output shown
below?
format Birth_Date Hire_Date mmddyy10.
a. Answer
Term_Date monyy7.;
format Birth_Date Hire_Date ddmmyyyy.

b. Answer Term_Date mmmyyyy.;
format Birth_Date Hire_Date ddmmyy10.
c. Answer Term_Date monyy7.;
output
Birth_Date
Hire_Date
Term_Date
21/05/1969
15/10/1992
MAR2007
17
National Language Support (NLS) enables a software product to function properly in every global market
for which the product is targeted. SAS contains NLS features to ensure that SAS applications conform to
local language conventions.
5-10
NLS date formats convert SAS date values to a locale-sensitive date string. The LOCALE= system option
is used to specify the locale, which reflects the local conventions, language, and culture of a geographical
region. For example, a locale value of English_Canada represents the country of Canada with a language
of English. A locale value of French_Canada represents the country of Canada with a language of French.
The LOCALE= system option can be specified in a configuration file, at SAS invocation, or in the
OPTIONS statement. For more information, refer to SAS 9.3 National Language Support Reference
Guide in the SAS documentation.
Exercises
Level 1
1. Displaying Formatted Values in a Detail Report
a. Open p105e01 and submit. Review the output.
b. Modify the PROC PRINT step to display only Employee_ID, Salary, Birth_Date, and
Employee_Hire_Date.
c. Add a FORMAT statement to display Salary in a dollar format, Birth_Date in 01/31/2012 date
style, and Employee_Hire_Date in the 01JAN2012 date style, as shown in the report below.
Obs
Employee_ID
Salary
Birth_Date
Employee_
Hire_Date
1
2
3
...
423
120101
120102
120103
$163,040.00
$108,255.00
$87,975.00
08/18/1980
08/11/1973
01/22/1953
01JUL2007
01JUN1993
01JAN1978
121147
$29,145.00
05/28/1973
01SEP1991

424
121148
$52,930.00
01/01/1973
5-11
01JAN2002
Level 2
a. Write a PROC PRINT step to display the report below using orion.sales as input. Subset the
observations and variables to produce the report shown below. Include titles, labels, and formats.
The results contain 13 observations.
US Sales Employees
Earning Under $26,000
Employee_ID
121036
121038
121044
...
121106
121108
First
Name
Last Name
Teresa
David
Ray
Mesley
Anstey
Abbott
James
Libby
Hilburger
Levi
Salary
Date
Hired
Sales Rep. I
Sales Rep. I
Sales Rep. I
$25,965
$25,285
$25,660
OCT2007
AUG2010
AUG1979
Sales Rep. I
Sales Rep. I
$25,880
$25,930
FEB2000
NOV2010
Title
Challenge
3. Exploring Formats by Category
a. Display orion.sales as shown in the report below. Refer to SAS Help or product documentation to
explore the Dictionary of Formats and investigate SAS Formats by Category. Identify and use
the character format that displays values in uppercase and a format that displays a character value
in quotation marks. The results contain 165 observations.
Employee_ID
120102
120103
120121
...
121144
121145
First_
Name
Last_Name
Job_Title
TOM
WILSON
IRENIE
ZHOU
DAWES
ELVISH
"Sales Manager"
"Sales Manager"
"Sales Rep. II"
RENEE
DENNIS
CAPACHIETTI
LANSBERRY
"Sales Manager"
"Sales Manager"
5-12
5.2 User-Defined Formats

Objectives
Create user-defined formats using the FORMAT
procedure.
Apply user-defined formats using a FORMAT
statement in a report.
Use formats to recode data values.
Use formats to collapse or aggregate data.
22
Business Scenario
Display country names instead of country codes in a
report.
Current Report (partial output)
Obs
Employee_ID
Salary
1
2
3
120102
120103
120121
$108,255
$87,975
$26,600
Country
Birth_
Date
Hire_
Date
AU
AU
AU
AUG1973
JAN1953
AUG1948
JUN1993
JAN1978
JAN1978
Desired Report (partial output)
23
Obs
Employee_ID
Salary
1
2
3
120102
120103
120121
$108,255
$87,975
$26,600
Country
Australia
Australia
Australia
Birth_
Date
Hire_
Date
AUG1973
JAN1953
AUG1948
JUN1993
JAN1978
JAN1978
p105d02
User-Defined Formats: Part 1

Use PROC FORMAT to create a user-defined format.
proc format;
value $ctryfmt
run;
'AU'='Australia'
'US'='United States'
other='Miscoded';
PROC FORMAT;
VALUE format-name range1 = 'label '
range2 = 'label '
...;
RUN;
p105d03
24
User-Defined Formats: Part 2

Use a FORMAT statement in the PROC PRINT step to
apply the format to a specific variable.
var Employee_ID Job_Title Salary
Country Birth_Date Hire_Date;
format Salary dollar10.
Birth_Date Hire_Date monyy7.
Country $ctryfmt.;
run;
25
p105d03
5-13
5-14
Viewing the Output

Obs
Employee_ID
Salary
1
2
3
4
5
120102
120103
120121
120122
120123
$108,255
$87,975
$26,600
$27,475
$26,190
Country
Australia
Australia
Australia
Australia
Australia
Birth_
Date
Hire_
Date
AUG1973
JAN1953
AUG1948
JUL1958
SEP1968
JUN1993
JAN1978
JAN1978
JUL1982
OCT1989
26
VALUE Statement
VALUE format-name range1='label '
range2='label '
...;
A format name
can be up to 32 characters in length
for character formats, must begin with a dollar sign ($),
followed by a letter or underscore
for numeric formats, must begin with a letter or
underscore
cannot end in a number
cannot be given the name of a SAS format
cannot include a period in the VALUE statement.
27
5-15
VALUE Statement
VALUE format-name range1='label '
range2='label '
...;
Each range can be

a single value
a range of values
a list of values.
Labels
can be up to 32,767 characters in length
are enclosed in quotation marks.
28
Enclosing labels in quotation marks is a best practice, and it is required if a label contains internal
blanks.

Which names are invalid for user-defined formats?
a.
b.
c.
d.
e.
$stfmt
$3levels
_4years
salranges
dollar
29
5-16
Defining a Character Format

character
format
name
discrete
character
values
proc format;
value $ctryfmt 'AU'='Australia'
other='Miscoded';
run;
keyword
labels
The OTHER keyword includes all values that do not

match any other value or range.
p105d03
31
Applying a Format
User-defined and SAS formats can be applied in a single
FORMAT statement.
proc print data=orion.sales label;
format Salary dollar10.
Birth_Date Hire_Date monyy7.
Country $ctryfmt.;
run;
A period (for example, at the end of the $CTRYFMT

format) is required when user-defined formats are
used in a FORMAT statement.
32
p105d03
Idea Exchange
The formatting examples shown in this section are
sometimes referred to as translating values.
Can you give an example of where this type of application
might be useful?
33
Business Scenario
An Orion Star manager wants a report showing employee
salaries collapsed into three user-defined groups or tiers.
Current Report
Obs
Employee_ID
1
2
3
4
120102
120103
120121
120122
Last_Name
Salary
Zhou
Dawes
Elvish
Ngan
108255
87975
26600
27475
Last_Name
Salary
Zhou
Dawes
Elvish
Ngan
Tier
Tier
Tier
Tier
Desired Report
Obs
Employee_ID
1
2
3
4
120102
120103
120121
120122
3
2
1
1
35
5-17
5-18
Specifying Ranges of Values

Use PROC FORMAT to specify the salary range for each
tier.
Salary
PROC FORMAT
Value
20,000 to 49,999
Tier1
50,000 to 99,999
Tier2
100,000 to 250,000
Tier3
36
Defining a Numeric Format

numeric
ranges
proc format;
value tiers
run;
0-49999='Tier 1'
50000-99999='Tier 2'
100000-250000='Tier 3';
numeric
format
name
labels
p105d04
37
Defining and Using a Numeric Format

p105d04
This demonstration illustrates the use of a user-defined numeric format.
5-19
proc format;
value tiers 20000-49999 ='Tier 1'
50000-99999 ='Tier 2'
100000-250000='Tier 3';
run;
data work.salaries;
input Name $ Salary;
Original_Salary=Salary;
datalines;
Abi 50000
Mike 65000
Jose 50000.00
Joe 37000.50
Ursula 142000
Lu 49999.99
;
proc print data=work.salaries;
format Salary tiers.;
run;
1. Program p105d04 includes a PROC FORMAT step to create the TIERS format. The DATA step reads
data lines within the program to create a data set containing names and salaries. The PROC PRINT
step displays the new data set, applying the TIERS format. Notice in the program that Salary is
assigned to Original_Salary so that both can be included in the report. The assignment statement is
covered in a later chapter.
2. Look at the data values and predict what label will be displayed when the TIERS format is applied to
Salary. Lus salary falls within a gap.
Name
Salary
Abi
50000
Mike
65000
Jose
50000.00
Joe
37000.50
Ursula
Lu
Coded Salary
142000
49999.99
3. Submit the program. What Salary value is displayed for Lu? ___________________
When a value does not match any of the ranges, PROC PRINT attempts to display the actual value. In
this case, the column width was determined by the width of the formatted values, which is 6. As
mentioned earlier, if the format width is not large enough to accommodate a numeric value, the
displayed value is automatically adjusted to fit in the width. What can you do to correct this?
5-20
4. Specify a width of 8 on the TIERS format in the FORMAT statement and resubmit the program.
proc print data=work.salaries;
format Salary tiers8.;
run;
Now what Salary value is displayed for Lu? ___________________
Defining a Continuous Range

The less than (<) symbol excludes the endpoint from a
range, allowing a continuous range.
Put < after the starting value in a range to exclude it.
Put < before the ending value in a range to exclude it.
Range
Starting Value
Ending Value
50000 - 100000
Includes 50000
Includes 100000
50000 - < 100000
Includes 50000
Excludes 100000
50000 < - 100000
Excludes 50000
Includes 100000
50000 < - < 100000
Excludes 50000
Excludes 100000
39
The < symbol is used to define an exclusive range. The > symbol is not permitted in a VALUE statement.
5.04 Quiz
How will a value of 50000 be displayed if the TIERS
format below is applied to the value?
a.
b.
c.
d.
Tier 1
Tier 2
50000
a missing value
proc format;
value tiers
run;
40
20000-<50000 ='Tier 1'

50000-<100000='Tier 2'
100000-250000='Tier 3';
p105d05
LOW and HIGH Keywords

the lowest
possible value
proc format;
value tiers
run;
low-<50000 ='Tier 1'

50000-<100000='Tier 2'
100000-high
='Tier 3';
the highest
possible value
The LOW keyword

includes missing values for character variables
does not include missing values for numeric variables.
p105d06
42
Applying a Numeric Format

proc format;
value tiers
Part 1
run;
low-<50000 ='Tier 1'

50000-<100000='Tier 2'
100000-high ='Tier 3';

Part 2
format Birth_Date Hire_Date monyy7.
Salary tiers.;
run;
43
p105d06
5-21
5-22
Viewing the Output

Obs
Employee_ID
1
2
3
4
5
120102
120103
120121
120122
120123
Job_Title
Sales
Sales
Sales
Sales
Sales
Manager
Manager
Rep. II
Rep. II
Rep. I
Salary
Tier
Tier
Tier
Tier
Tier
3
2
1
1
1
Country
AU
AU
AU
AU
AU
Birth_
Date
Hire_
Date
AUG1973
JAN1953
AUG1948
JUL1958
SEP1968
JUN1993
JAN1978
JAN1978
JUL1982
OCT1989
44
User-Defined Format Example

Ranges can be specified using lists, ranges, discrete
values, and keywords.
proc format;
value mnthfmt 1,2,3='Qtr 1'
4-6='Qtr 2'
7-9='Qtr 3'
10-12='Qtr 4'
.='missing'
other='unknown';
run;
45
Multiple User-Defined Formats

Multiple VALUE statements can be included in a single
PROC FORMAT step.
proc format;
value $ctryfmt
value tiers
'AU'='Australia'
other='Miscoded';
low-<50000 ='Tier 1'

50000-<100000='Tier 2'
100000-high ='Tier 3';
run;
p105d07
46
Viewing the Output

format Birth_Date Hire_Date monyy7.
Country $ctryfmt.
Salary tiers.;
run;
Obs
Employee_ID
1
2
3
4
5
120102
120103
120121
120122
120123
Job_Title
Sales
Sales
Sales
Sales
Sales
Manager
Manager
Rep. II
Rep. II
Rep. I
Salary
Tier
Tier
Tier
Tier
Tier
3
2
1
1
1
Country
Birth_
Date
Hire_
Date
Australia
Australia
Australia
Australia
Australia
AUG1973
JAN1953
AUG1948
JUL1958
SEP1968
JUN1993
JAN1978
JAN1978
JUL1982
OCT1989
p105d07
47
Exercises
5-23
5-24
Level 1
4. Creating User-Defined Formats
b. Create a character format named $GENDER that displays gender codes as follows:
F
Female
Male
c. Create a numeric format named MNAME that displays month numbers as follows:
1
January
February
March
d. Add a PROC PRINT step to display the data set, applying these two user-defined formats to the
Employee_Gender and BirthMonth variables, respectively.
e. Submit the program to produce the following report. The results contain 113 observations.
Employees with Birthdays in Q1
Obs
Employee_ID
1
2
3
...
112
113
Employee_
Gender
Birth
Month
120103
120107
120108
Male
Female
Female
January
January
February
121142
121148
Male
Male
February
January
Level 2
5. Defining Ranges in User-Defined Formats
b. Create a character format named $GENDER that displays gender codes as follows:
F
Female
Male
Any other value
Invalid code
5.3 Solutions
5-25
c. Create a numeric format named SALRANGE that displays salary ranges as follows:
At least 20,000 but less than 100,000
Below $100,000
At least 100,000 and up to 500,000
$100,000 or more
missing
Missing salary
Any other value
Invalid salary
d. In the PROC PRINT step, apply these two user-defined formats to the Gender and Salary
variables, respectively. Submit the program to produce the following report:
Salary and Gender Values
for Non-Sales Employees
Obs
Employee_ID
1
2
3
4
5
6
7
8
9
10
11
12
13
120101
120104
120105
120106
120107
120108
120108
120110
120111
120112
120113
120114
120115
Job_Title
Director
Administration Manager
Secretary I
Office Assistant II
Office Assistant III
Warehouse Assistant II
Warehouse Assistant I
Warehouse Assistant III
Security Guard II
Security Guard II
Security Manager
Service Assistant I
Salary
$100,000 or more
Below $100,000
Below $100,000
Missing salary
Below $100,000
Below $100,000
Below $100,000
Below $100,000
Below $100,000
Below $100,000
Below $100,000
Below $100,000
Invalid salary
Gender
Male
Female
Female
Male
Female
Female
Female
Male
Male
Female
Female
Invalid code
Male
Challenge
6. Exploring Format Storage Options
User-defined formats are stored in the formats catalog in the work library, work.formats. Use the
SAS Help Facility or product documentation to explore permanent format catalogs in PROC
FORMAT.
What option enables you to store the formats in a permanent library? _______________________
What option causes SAS to look for formats in permanent libraries? _______________________
5.3 Solutions
proc print data=orion.employee_payroll;
var Employee_ID Salary Birth_Date Employee_Hire_Date;
format Salary dollar11.2 Birth_Date mmddyy10.
Employee_Hire_Date date9.;
5-26
run;
a. Submit p105e01
b. Add a VAR statement.
c. Add a FORMAT statement.
title1 'US Sales Employees';
title2 'Earning Under $26,000';
proc print data=orion.sales label noobs;
where Country='US' and Salary<26000;
var Employee_ID First_Name Last_Name Job_Title Salary Hire_Date;
label First_Name='First Name'
Last_Name='Last Name'
Job_Title='Title'
Hire_Date='Date Hired';
format Salary dollar10. Hire_Date monyy7.;
run;
title;
footnote;
3. Exploring Functions by Category
var Employee_ID First_Name Last_Name Job_Title;
format First_Name Last_Name $upcase. Job_Title $quote.;
run;
4. Creating User-Defined Formats
data Q1Birthdays;
set orion.employee_payroll;
BirthMonth=month(Birth_Date);
if BirthMonth le 3;
run;
proc format;
value $gender
'F'='Female'
'M'='Male';
value mname
1='January'
2='February'
3='March';
run;
title 'Employees with Birthdays in Q1';
proc print data=Q1Birthdays;
var Employee_ID Employee_Gender BirthMonth;
format Employee_Gender $gender.
BirthMonth mname.;
5.3 Solutions
run;
title;
5. Defining Ranges in User-Defined Formats
proc format;
value $gender
'F'='Female'
'M'='Male'
other='Invalid code';
value salrange .='Missing salary'
20000-<100000='Below $100,000'
100000-500000='$100,000 or more'
other='Invalid salary';
run;
title1 'Salary and Gender Values';
title2 'for Non-Sales Employees';
var Employee_ID Job_Title Salary Gender;
format Salary salrange. Gender $gender.;
run;
title;
6. Exploring Format Storage Options
What option enables you to store the formats in a permanent library? LIBRARY=
What option causes SAS to look for formats in permanent libraries? FMTSEARCH=
5-27
5-28

Use SAS documentation or the SAS Help Facility
to explore the Zw.d numeric format. What is it used for?
Hint: Search for Zw.d or explore Formats by Category.
The Zw.d format writes standard numeric data with
leading zeros. It is similar to the w.d format except
that Zw.d pads right-aligned output with zeros instead
of blanks.
14

Which FORMAT statement creates the output shown
below?
format Birth_Date Hire_Date mmddyy10.
a. Answer
Term_Date monyy7.;
format Birth_Date Hire_Date ddmmyyyy.
b. Answer Term_Date mmmyyyy.;
format Birth_Date Hire_Date ddmmyy10.
c. Answer Term_Date monyy7.;
output
Birth_Date
Hire_Date
Term_Date
21/05/1969
15/10/1992
MAR2007
18
5.3 Solutions
5.03 Multiple Answer Poll Correct Answer

Which names are invalid for user-defined formats?
a.
b.
c.
d.
e.
$stfmt
$3levels
_4years
salranges
dollar
Character formats must have a dollar sign as the

first character and a letter or underscore as the
second character.
User-defined formats cannot be given the name of a
format provided by SAS.
30

How will a value of 50000 be displayed if the TIERS
format below is applied to the value?
a.
b.
c.
d.
Tier 1
Tier 2
50000
a missing value
proc format;
value tiers
run;
41
20000-<50000 ='Tier 1'

50000-<100000='Tier 2'
100000-250000='Tier 3';
p105d05
5-29
5-30
Chapter 6 Reading SAS Data Sets

6.1
Reading a SAS Data Set ................................................................................................. 6-3

Exercises .............................................................................................................................. 6-15
6.2
Customizing a SAS Data Set ....................................................................................... 6-17

Exercises .............................................................................................................................. 6-39
6.3
Solutions ....................................................................................................................... 6-42

6-2
6.1 Reading a SAS Data Set

Objectives
Define the business scenario that will be used when

reading from a data source to create a SAS data set.
Use a DATA step to create a SAS data set from
an existing SAS data set.
Subset observations with a WHERE statement.
Create a new variable with an assignment statement.
3
3
Business Scenario
Information about Orion Star sales employees resides in
several input sources.
4
4
6-3
6-4
Considerations
Management wants a series of reports for Australian
sales employees. You will read data from various input
sources to create a SAS data set that can be analyzed
and presented.
SAS data set
5
5

Which types of files will you read into SAS?
a.
b.
c.
d.
e.
f.
SAS data sets

Excel worksheets
database tables
raw data files
other
not sure
6
6
Business Scenario: Part 1

Read an existing SAS data set to create a new data set.
The new data set should include only the observations for
the Australian sales representatives.
orion.sales
DATA Step
work.subset1
Rep
7
7
Using a SAS Data Set as Input

data work.subset1;
set orion.sales;
run;
DATA output-SAS-data-set;
SET input-SAS-data-set;
RUN;
p106d01
6-5
6-6
DATA Statement
The DATA statement begins a DATA step and provides
the name of the SAS data set to create.
DATA output-SAS-data-set;
data work.subset1;
set orion.sales;
run;
A DATA step can create temporary or permanent data

sets.
The rules for SAS variable names also apply to
data set names.
p106d01
9
9
SET Statement
The SET statement reads observations from an existing
SAS data set for further processing in the DATA step.
SET input-SAS-data-set;
data work.subset1;
set orion.sales;
run;
10
The SET statement reads all observations and all

variables from the input data set.
Observations are read sequentially, one at a time.
The SET statement can read temporary or permanent
data sets.
p106d01
10
WHERE Statement
The WHERE statement selects observations from
a SAS data set that meet a particular condition.
data work.subset1;
set orion.sales;
run;
The variables named in the WHERE expression must

exist in the input SAS data set.
p106d01
11
11
Using a WHERE statement might improve the efficiency of your SAS programs because SAS only
processes the observations that meet the condition or conditions in the WHERE expression.
Viewing the Log

Partial SAS Log
42
43
44
45
46
data work.subset1;
set orion.sales;
run;
WHERE (Country='AU') and Job_Title contains 'Rep';
NOTE: The data set WORK.SUBSET1 has 61 observations and 9 variables.
SAS read 61 of the 165 observations.
12
12
6-7
6-8
Viewing the Output

proc print data=work.subset1 noobs;
run;
Employee_ID
120121
120122
120123
120124
120125
First_
Name
Last_Name
Irenie
Christina
Kimiko
Lucian
Fong
Elvish
Ngan
Hotstone
Daymond
Hofmeister
Gender
F
F
F
M
M
Salary
26600
27475
26190
26480
32040
Job_Title
Sales
Sales
Sales
Sales
Sales
Rep.
Rep.
Rep.
Rep.
Rep.
II
II
I
I
IV
Country
Birth_
Date
Hire_
Date
AU
AU
AU
AU
AU
-4169
-523
3193
1228
-391
6575
8217
10866
8460
8460
p106d01
13
13
Setup for the Poll

Consider the DATA step below.
data us;
set orion.sales;
where Country='US';
run;
14
p106a01

Considering this DATA step, which statement is true?
a. It reads a temporary data set and creates a permanent
data set.
b. It reads a permanent data set and creates a temporary
data set.
c. It contains a syntax error and will not execute.
d. It will not execute because you cannot work with
permanent and temporary data sets in the same step.
15

Orion Star management wants to give a 10% bonus
to each Australian Sales representative hired before
January 1, 2000.
Bonus
Rep
18
18
6-9
6-10
Considerations
Subsetting is based on Hire_Date, which contains
a SAS date value. How can you compare a SAS date
value to a calendar date?
Use a
SAS date
constant.
19
19
Date Constant
A date constant can be used in any SAS expression,
including a WHERE expression.
data work.subset1;
set orion.sales;
Job_Title contains 'Rep' and
Hire_Date<'01jan2000'd;
run;
A SAS date constant is a date written in the form

'ddmmm<yy>yy'd.
20
p106d02
20
Considerations
Create a data set that includes the new variable, Bonus,
which represents a 10% bonus.
orion.sales
work.subset1
21
21
Assignment Statement
The assignment statement evaluates an expression
and assigns the result to a new or existing variable.
data work.subset1;
set orion.sales;
Bonus=Salary*.10;
run;
variable=expression;
22
p106d02a
22
6-11
6-12
Assignment Statement
The expression consists of operands and operators.
variable=expression;
Operators
Operands
character constants
numeric constants
date constants
character variables
numeric variables
symbols that represent a

calculation or manipulation
+ ** *
) / ||
SAS functions
23
23
The operators can be character or arithmetic operators or SAS functions. A function is a routine that
accepts arguments, performs a calculation or manipulation using the arguments, and returns a single
value.
Sample Assignment Statements

Example
Type
Salary=26960;
Numeric constant
Gender='F';
Character constant
Hire_Date='21JAN1995'd;
Date constant
BonusMonth=month(Hire_Date); SAS function

Bonus=Salary*.10;
Arithmetic expression
24
24
The MONTH function accepts a SAS date and returns the month portion of the date as an integer between
1 and 12. You investigate this and other SAS functions in a later chapter.
Arithmetic Operators
If any operand in an arithmetic expression has a missing
value, the result is a missing value.
Symbol
Definition
Priority
**
Exponentiation
Multiplication
II
Division
II
Addition
III
Subtraction
III
Parentheses can be used to clarify or alter the order of

operations in an arithmetic expression.
25
25
Viewing the Log

Partial SAS Log
214
215
216
217
218
219
220
data work.subset1;
set orion.sales;
Bonus=Salary*.10;
run;
WHERE (Country='AU') and Job_Title contains 'Rep' and
(Hire_Date<'01JAN2000'D);
The input data set has 9 variables, and the new data set
has 10 variables.
26
26
6-13
6-14
Viewing the Output

proc print data=work.subset1 noobs;
var First_Name Last_Name Salary
Job_Title Bonus Hire_Date;
format Hire_Date date9.;
run;
First_Name
Last_Name
Salary
Irenie
Christina
Kimiko
Lucian
Fong
Elvish
Ngan
Hotstone
Daymond
Hofmeister
26600
27475
26190
26480
32040
Job_Title
Sales
Sales
Sales
Sales
Sales
Rep.
Rep.
Rep.
Rep.
Rep.
II
II
I
I
IV
Bonus
Hire_Date
2660.0
2747.5
2619.0
2648.0
3204.0
01JAN1978
01JUL1982
01OCT1989
01MAR1983
01MAR1983
p106d02a
27
27
No format was specified for Bonus, so PROC PRINT uses a BESTw.d format. One decimal position is
sufficient to display the values on this page.
6.03 Quiz
Evaluate the assignment statements below given the
values shown in the PDV.
x
y
.
a.
num=y+z/2;
b.
num=x+z/2;
z
10
28
28
6-15
Exercises
Level 1
1. Creating a SAS Data Set
a. Retrieve and submit the starter program p106e01.
What is the name of the variable that contains gender values? ____________________________
What are the two observed gender values? ____________________________________________
b. Add a DATA step before the PROC PRINT step to create a new data set named work.youngadult
using the data set orion.customer_dim as input. Include a WHERE statement to select only
female customers.
Submit the program and confirm that work.youngadult was created with 30 observations and 11
variables.
c. Modify the program to select female customers whose age is between 18 and 36. Submit the
program and confirm that work.youngadult was created with 15 observations and 11 variables.
d. Modify the program to select 18- to 36-year-old female customers who have the word Gold in
their Customer_Group value. Submit the program and confirm that work.youngadult was
created with 5 observations and 11 variables.
e. Add an assignment statement to the DATA step to create a new variable, Discount, and assign it a
value of .25.
f. Modify the PROC PRINT step to print the new data set as shown below. Use an ID statement to
display Customer_ID instead of the Obs column. Results should contain five observations.
Customer_ID
5
9
45
49
2550
Customer_Name
Sandrina Stephano
Cornelia Krahl
Dianne Patchin
Annmarie Leveille
Sanelisiwe Collier
Customer_
Age
28
33
28
23
19
Customer_
Gender
F
F
F
F
F
Customer_Group
Orion
Orion
Orion
Orion
Orion
Club
Club
Club
Club
Club
Gold
Gold
Gold
Gold
Gold
members
members
members
members
members
Discount
0.25
0.25
0.25
0.25
0.25
Level 2
a. Write a DATA step to create a new data set named work.assistant using the data set orion.staff as
input.
b. The work.assistant data set should contain only the observations where Job_Title contains
Assistant and Salary is less than $26,000.
6-16
c. Create two new variables, Increase and New_Salary.

Increase is Salary multiplied by 0.10.
New_Salary is Salary added to Increase.
d. Generate a detail listing report as shown below. Display Employee_ID as the identifier in place
of the Obs column. The results should contain five observations.
Employee_ID
120685
120688
120690
121010
121011
Job_Title
Service Assistant I
Service Assistant I
Salary
Increase
New_Salary
$25,130.00
$25,905.00
$25,185.00
$25,195.00
$25,735.00
$2,513.00
$2,590.50
$2,518.50
$2,519.50
$2,573.50
$27,643.00
$28,495.50
$27,703.50
$27,714.50
$28,308.50
Challenge
3. Using the SOUNDS-LIKE Operator to Select Observations
a. Write a DATA step to create a new data set named work.tony using orion.customer_dim as
input.
b. Include a WHERE statement in the DATA step to select observations in which the
Customer_FirstName value sounds like Tony.
Documentation on the SOUNDS-LIKE operator can be found in the SAS Help facility or
product documentation by searching for sounds-like operator.
c. Write a PROC PRINT step to create the following report:

Obs
1
2
Customer_
FirstName
Customer_
LastName
Tonie
Tommy
Asmussen
Mcdonald
6.2 Customizing a SAS Data Set

Objectives
Subset variables by using the DROP and KEEP

statements.
Explore the compilation and execution phases of
the DATA step.
Store labels and formats in the descriptor portion
of a SAS data set.
33
33

All Australian sales representatives will get a bonus,
regardless of hire date. The new data set should contain
a subset of the variables from the input data set.
work.subset1
orion.sales
Employee_ID
Gender
Country
Birth_Date
34
34
6-17
6-18
DROP Statement
The DROP statement specifies the variables to exclude
from the output data set.
data work.subset1;
set orion.sales;
Bonus=Salary*.10;
drop Employee_ID Gender Country
Birth_Date;
run;
DROP variable-list;
Partial SAS Log
p106d03
35
35
Viewing the Output

proc print data=work.subset1;
run;
36
Obs
First_
Name
Last_Name
1
2
3
4
5
Irenie
Christina
Kimiko
Lucian
Fong
Elvish
Ngan
Hotstone
Daymond
Hofmeister
Salary
26600
27475
26190
26480
32040
Job_Title
Sales
Sales
Sales
Sales
Sales
Rep.
Rep.
Rep.
Rep.
Rep.
II
II
I
I
IV
Hire_
Date
Bonus
6575
8217
10866
8460
8460
2660.0
2747.5
2619.0
2648.0
3204.0
p106d03
36
6-19
KEEP Statement
The KEEP statement specifies all variables to include in
the output data set.
data work.subset1;
set orion.sales;
Bonus=Salary*.10;
keep First_Name Last_Name Salary
Job_Title Hire_Date Bonus;
run;
KEEP variable-list;
If a KEEP statement is used, it must include every

variable to be written, including any new variables.
p106d04
37
37
When you use a KEEP statement, be sure to name every variable to be written to the new SAS data set,
including any variables created within the step, such as Bonus.
Viewing the Log

Partial SAS Log
38
38
6-20
Viewing the Output

run;
Obs
First_
Name
Last_Name
1
2
3
4
5
Irenie
Christina
Kimiko
Lucian
Fong
Elvish
Ngan
Hotstone
Daymond
Hofmeister
Salary
26600
27475
26190
26480
32040
Hire_
Date
Job_Title
Sales
Sales
Sales
Sales
Sales
Rep.
Rep.
Rep.
Rep.
Rep.
II
II
I
I
IV
6575
8217
10866
8460
8460
Bonus
2660.0
2747.5
2619.0
2648.0
3204.0
p106d04
39
39
Business Scenario: Behind the Scenes

Orion Star programmers need to understand the internal
processing that occurs when a DATA step is submitted.
41
41
DATA Step Processing

SAS processes the DATA step in two phases:
Compilation Phase
Execution Phase
42
42
Compilation Phase
Scans the program for syntax errors;

translates the program into machine
language.
PDV
Name
Salary
Creates the program data vector (PDV)

to hold one observation.
Creates the descriptor portion of the
output data set.
43
43
6-21
6-22
Compilation
data work.subset1;
set orion.sales;
Bonus=Salary*.10;
Birth_Date;
run;
p106d03
...
44
44
Compilation
PDV
Employee_ID
N8
Country
$2
data work.subset1;
set orion.sales;
Bonus=Salary*.10;
Birth_Date;
run;
First_Name
$ 12
Birth_Date
N8
Last_Name Gender
$ 18
$1
Salary
N8
Job_Title
$ 25
Hire_Date
N8
45
45
...
Compilation
data work.subset1;
set orion.sales;
Bonus=Salary*.10;
Birth_Date;
run;
PDV
Employee_ID
N8
Country
$2
First_Name
$ 12
Birth_Date
N8
Last_Name Gender
$ 18
$1
Hire_Date
N8
Salary
N8
Job_Title
$ 25
Bonus
N8
...
46
46
Compilation
data work.subset1;
set orion.sales;
Bonus=Salary*.10;
Birth_Date;
run;
PDV
Employee_ID
N8
Country
$2
First_Name
$ 12
Birth_Date
N8
Last_Name Gender
D
$ 18
$1
Hire_Date
N8
Salary
N8
Job_Title
$ 25
Bonus
N8
47
47
...
6-23
6-24
Compilation
data work.subset1;
set orion.sales;
Bonus=Salary*.10;
Birth_Date;
run;
PDV
Employee_ID
N8
Country
$2
First_Name
$ 12
Birth_Date
N8
Last_Name Gender
D
$ 18
$1
Hire_Date
N8
Salary
N8
Job_Title
$ 25
Bonus
N8
Descriptor Portion of work.subset1

First_Name
Last_Name
Salary
Job_Title
Hire_Date
Bonus
48
48
Execution Phase
Compile the step
Compilation Phase
Success?
No
Next step
Yes
Execution Phase
Initialize PDV to missing

End of
file?
Execute SET statement

Execute other statements
Yes
Next step
No
Output to SAS data set

49
49
Execution
Initialize PDV
Partial orion.sales
Employee
_ID
120121
120122
data work.subset1;
set orion.sales;
Bonus=Salary*.10;
Birth_Date;
run;
Hire_Date
6575
8217
...
120123
10866
120124
8460
PDV
Employee
_ID
.
...
Salary
...
Country
Birth_Date
Hire_Date
.
Bonus
.
work.subset1
First_Name
Last_Name
Salary
Job_Title
Hire_Date
Bonus
...
50
50
Execution
Partial orion.sales
Employee
_ID
120121
120122
data work.subset1;
set orion.sales;
Bonus=Salary*.10;
Birth_Date;
run;
Hire_Date
6575
8217
...
120123
10866
120124
8460
PDV
Employee
_ID
120121
...
Salary
...
26600
Country
AU
Birth_Date
-4169
Hire_Date
Bonus
6575
work.subset1
First_Name
Last_Name
Salary
Job_Title
Hire_Date
Bonus
51
51
...
The WHERE statement is executed before the other executable statements.
6-25
6-26
Execution
Partial orion.sales
Employee
_ID
120121
120122
data work.subset1;
set orion.sales;
Bonus=Salary*.10;
Birth_Date;
run;
Hire_Date
6575
8217
...
120123
10866
120124
8460
PDV
Employee
_ID
120121
...
Salary
...
26600
Country
AU
Birth_Date
Hire_Date
-4169
Bonus
6575
2660
work.subset1
First_Name
Last_Name
Salary
Job_Title
Hire_Date
Bonus
...
52
52
Execution
Partial orion.sales
Employee
_ID
120121
120122
data work.subset1;
set orion.sales;
Bonus=Salary*.10;
Birth_Date;
run;
Implicit OUTPUT;
Hire_Date
6575
8217
...
120123
10866
120124
8460
Implicit RETURN;
PDV
Employee
_ID
120121
...
Salary
...
26600
Country
AU
Birth_Date
-4169
Hire_Date
6575
Bonus
2660
work.subset1
First_Name
Last_Name
Irenie
Elvish
Salary
Job_Title
26600 Sales Rep II.
Hire_Date
6575
Bonus
2660
53
53
...
6-27
Execution
Reinitialize PDV
Partial orion.sales
Employee
_ID
120121
120122
data work.subset1;
set orion.sales;
Bonus=Salary*.10;
Birth_Date;
run;
Hire_Date
6575
8217
...
120123
10866
120124
8460
PDV
Employee
_ID
120121
...
Salary
...
26600
Country
AU
Birth_Date
-4169
Hire_Date
Bonus
6575
work.subset1
First_Name
Last_Name
Irenie
Elvish
Salary
Job_Title
26600 Sales Rep II.
Hire_Date
6575
Bonus
2660
...
54
54
Only the new variables are reinitialized. The variables that come from the input data set are not
reinitialized because they are overwritten when the next observation is read into the PDV. Values in the
PDV are overwritten even if values in the next observation are missing.
Execution
Partial orion.sales
Employee
_ID
120121
120122
data work.subset1;
set orion.sales;
Bonus=Salary*.10;
Birth_Date;
run;
Hire_Date
6575
8217
...
120123
10866
120124
8460
PDV
Employee
_ID
120122
...
Salary
...
27475
Country
AU
Birth_Date
-523
Hire_Date
Bonus
8217
work.subset1
First_Name
Last_Name
Irenie
Elvish
Salary
Job_Title
26600 Sales Rep II.
Hire_Date
6575
Bonus
2660
55
55
...
6-28
Execution
Partial orion.sales
Employee
_ID
120121
120122
data work.subset1;
set orion.sales;
Bonus=Salary*.10;
Birth_Date;
run;
Hire_Date
6575
8217
...
120123
10866
120124
8460
PDV
Employee
_ID
120122
...
Salary
...
27475
Country
AU
Birth_Date
-523
Hire_Date
8217
Bonus
274.75
work.subset1
First_Name
Last_Name
Irenie
Elvish
Salary
Job_Title
Hire_Date
26600 Sales Rep II.
Bonus
6575
2660
...
56
56
Execution
Partial orion.sales
Employee
_ID
120121
120122
data work.subset1;
set orion.sales;
Bonus=Salary*.10;
Birth_Date;
Implicit OUTPUT;
run;
Hire_Date
6575
8217
...
120123
10866
120124
8460
Implicit RETURN;
PDV
Employee
_ID
120122
...
Salary
...
27475
Country
AU
Birth_Date
-523
Hire_Date
8217
Bonus
274.75
work.subset1
First_Name
Last_Name
Salary
Job_Title
Hire_Date
Bonus
Irenie
Elvish
26600 Sales Rep II.
6575
2660.00
Christina
Ngan
27475 Sales Rep II.
8217
2747.50
57
57
...
Execution
Partial orion.sales
Employee
_ID
120121
120122
data work.subset1;
set orion.sales;
where Country='AU'
and until EOF
Continue
Bonus=Salary*.10;
Birth_Date;
run;
Hire_Date
6575
8217
...
120123
10866
120124
8460
PDV
Employee
_ID
120122
...
Salary
...
27475
Country
Birth_Date
AU
Hire_Date
-523
Bonus
8217
work.subset1
First_Name
Last_Name
Irenie
Elvish
Salary
26600 Sales Rep II.
Job_Title
Hire_Date
6575
Bonus
2660.00
Christina
Ngan
27475 Sales Rep II.
8217
2747.50
58
58
Viewing the Output

run;
Obs
1
2
3
4
5
59
First_
Name
Last_Name
Salary
Irenie
Christina
Kimiko
Lucian
Fong
Elvish
Ngan
Hotstone
Daymond
Hofmeister
26600
27475
26190
26480
32040
Job_Title
Sales
Sales
Sales
Sales
Sales
Rep.
Rep.
Rep.
Rep.
Rep.
II
II
I
I
IV
Hire_
Date
Bonus
6575
8217
10866
8460
8460
2660.0
2747.5
2619.0
2648.0
3204.0
p106d03
59
6-29
6-30

Create a data set that contains all Australian employees
whose Bonus is at least $3000.
orion.sales
work.auemps
Bonus
61
61
Selecting Observations
Subsetting is based on the new variable, Bonus, that is
created with an assignment statement.
data work.auemps;
set orion.sales;
where Country='AU';
Bonus=Salary*.10;
Birth_Date;
run;
A WHERE statement is used to subset observations when
the selected variables exist in the input data set.
62
p106d03
6.04 Quiz
Open and submit p106a03. Is the output data set created
successfully?
data work.usemps;
set orion.sales;
Bonus=Salary*.10;
where Country='US' and Bonus>=3000;
run;
p106a03
63
Subsetting IF
The subsetting IF statement tests a condition to determine
whether the DATA step should continue processing the
current observation.
data work.auemps;
set orion.sales;
where Country='AU';
Bonus=Salary*.10;
if Bonus>=3000;
run;
IF condition;
In this program, processing will reach the bottom of the
DATA step and output an observation only if the condition
is true.
65
p106d05
65
6-31
6-32
Viewing the Log

Partial SAS Log
11
12
13
14
15
16
data work.auemps;
set orion.sales;
where Country='AU';
Bonus=Salary*.10;
if Bonus>=3000;
run;
WHERE Country='AU';
NOTE: The data set WORK.AUEMPS has 12 observations and 10 variables.
Of the 165 observations in orion.sales, 63 were read into

the PDV for processing, and only 12 were written to the
new data set.
66
66
Viewing the Output

proc print data=work.auemps;
var First_Name Last_Name Salary Bonus;
run;
PROC PRINT Output
Obs
1
2
3
4
5
6
7
8
9
10
11
12
First_
Name
Last_Name
Salary
Bonus
Tom
Wilson
Fong
Monica
Alvin
Alexei
Viney
Caterina
Daniel
Lynelle
Rosette
Fadi
Zhou
Dawes
Hofmeister
Kletschkus
Roebuck
Platts
Barbis
Hayawardhana
Pilgrim
Phoumirath
Martines
Nowd
108255
87975
32040
30890
30070
32490
30265
30490
36605
30765
30785
30660
10825.5
8797.5
3204.0
3089.0
3007.0
3249.0
3026.5
3049.0
3660.5
3076.5
3078.5
3066.0
67
67
Processing the Subsetting IF Statement

DATA Statement
Read an Observation
False
IF Expression
A subsetting IF statement is
valid only in a DATA step.
True
Continue Processing
the Observation
Output Observation
to SAS Data Set
68
68
Idea Exchange
File p106a04 contains two versions of the previous
program. Submit both programs and compare the output
and number of observations read. What do you notice
about the results?
data work.auemps;
set orion.sales;
Bonus=Salary*.10;
if Country='AU' and Bonus>=3000;
run;
69
69
6-33
6-34
WHERE versus Subsetting IF Statement

Step and Usage
WHERE
IF
Yes
No
SET statement
Yes
Yes
assignment statement
No
Yes
PROC step
DATA step (source of variable)
70
70

Define permanent labels and formats for some of the
variables in the new data set.
labels
work.subset1
formats
72
72
LABEL Statement
The LABEL statement assigns descriptive labels to
variables.
DATA Step
LABEL
statement
Sales Title
Date Hired
work.subset1
First_Name
Last_Name
Salary
Job_Title
Hire_Date
Bonus
73
Defining Permanent Labels

Use a LABEL statement in a DATA step to permanently
assign labels to variables. The labels are stored in the
descriptor portion of the data set.
data work.subset1;
set orion.sales;
Bonus=Salary*.10;
label Job_Title='Sales Title'
Birth_Date;
run;
LABEL variable='label '
<variable='label '>;
74
p106d06
74
6-35
6-36
Viewing the Output

proc contents data=work.subset1;
run;
#
Variable
Type
Len
6
1
5
4
2
3
Bonus
First_Name
Hire_Date
Job_Title
Last_Name
Salary
Num
Char
Num
Char
Char
Num
8
12
8
25
18
8
Label
Date Hired
Sales Title
p106d06
75
75
Viewing the Output: Displaying Labels

To use labels in the PRINT procedure, use the LABEL
option in the PROC PRINT statement.
proc print data=work.subset1 label;
run;
Obs
1
2
3
4
5
76
First_
Name
Last_Name
Salary
Irenie
Christina
Kimiko
Lucian
Fong
Elvish
Ngan
Hotstone
Daymond
Hofmeister
26600
27475
26190
26480
32040
Sales Title
Date
Hired
Bonus
Sales
Sales
Sales
Sales
Sales
6575
8217
10866
8460
8460
2660.0
2747.5
2619.0
2648.0
3204.0
Rep.
Rep.
Rep.
Rep.
Rep.
II
II
I
I
IV
p106d06
76
6-37
Viewing the Output: Splitting Labels

Use the PROC PRINT SPLIT= option to split labels
across lines based on a split character.
proc print data=work.subset1 split=' ';
run;
Obs
1
2
3
4
5
First_
Name
Last_Name
Salary
Irenie
Christina
Kimiko
Lucian
Fong
Elvish
Ngan
Hotstone
Daymond
Hofmeister
26600
27475
26190
26480
32040
Sales
Title
Sales
Sales
Sales
Sales
Sales
Rep.
Rep.
Rep.
Rep.
Rep.
II
II
I
I
IV
Date
Hired
Bonus
6575
8217
10866
8460
8460
2660.0
2747.5
2619.0
2648.0
3204.0
p106d06
77
77
PROC PRINT is the only SAS procedure that requires either the LABEL option or the SPLIT= option to
use custom labels.
6.05 Quiz
What column heading will be displayed for Job_Title in the
program below?
data work.us;
set orion.sales;
where Country='US';
Bonus=Salary*.10;
label Job_Title='Sales Title';
Birth_Date;
run;
label Job_Title='Title';
run;
78
p106a05
6-38
FORMAT Statement
The FORMAT statement associates formats with
variables.
DATA Step
FORMAT
statement
commax8.
ddmmyy10.
commax8.2
work.subset1
First_Name
Last_Name
Irenie
Elvish
Salary
Job_Title
26600 Sales Rep. II
Hire_Date
6575
Bonus
2660.0
80
Defining Permanent Formats

Use a FORMAT statement in a DATA step to permanently
associate formats with variables.
data work.subset1;
set orion.sales;
Bonus=Salary*.10;
format Salary commax8. Bonus commax8.2
Hire_Date ddmmyy10.;
Birth_Date;
run;
FORMAT variable(s) format ;
81
p106d07
81
Viewing the Output

run;
#
Variable
Type
Len
6
1
5
4
2
3
Bonus
First_Name
Hire_Date
Job_Title
Last_Name
Salary
Num
Char
Num
Char
Char
Num
8
12
8
25
18
8
Format
Label
COMMAX8.2
DDMMYY10.
Date Hired
Sales Title
COMMAX8.
p106d07
82
Viewing the Output

run;
Obs
1
2
3
4
5
First_
Name
Last_Name
Salary
Sales Title
Date Hired
Bonus
Irenie
Christina
Kimiko
Lucian
Fong
Elvish
Ngan
Hotstone
Daymond
Hofmeister
26.600
27.475
26.190
26.480
32.040
Sales
Sales
Sales
Sales
Sales
01/01/1978
01/07/1982
01/10/1989
01/03/1983
01/03/1983
2.660,00
2.747,50
2.619,00
2.648,00
3.204,00
Rep.
Rep.
Rep.
Rep.
Rep.
II
II
I
I
IV
p106d07
83
83
Exercises
6-39
6-40
Level 1
4. Subsetting Observations Based on Two Conditions
b. Modify the DATA step to select only the observations with Emp_Hire_Date values on or after
July 1, 2010. Subset the observations as they are being read into the program data vector.
c. In the DATA step, write another statement to select only the observations that have an increase
greater than 3000.
d. The new data set should contain only the following variables: Employee_ID, Emp_Hire_Date,
Salary, Increase, and NewSalary.
e. Add permanent labels for Employee_ID, Emp_Hire_Date, and NewSalary as shown in the
report below.
f. Add permanent formats to display Salary and NewSalary with dollar signs, commas, and two
decimal places, and Increase with commas and no decimal places.
g. Submit a PROC CONTENTS step to verify that the labels and formats are stored in the descriptor
portion of the new data set, work.increase.
#
Variable
Type
3
1
4
5
2
Emp_Hire_Date
Employee_ID
Increase
NewSalary
Salary
Num
Num
Num
Num
Num
Len
8
8
8
8
8
Format
Informat
Label
DATE9.
12.
COMMA5.
DOLLAR10.2
DOLLAR10.2
DATE9.
Hire Date
Employee ID
New Annual Salary
Employee Annual Salary
h. Some variables have labels and formats that were not defined in this program. How were these
created? ____________________________________________________________________
i. Submit the program to create the PROC PRINT report below, with labels split over multiple lines.
Results should contain 10 observations.
Obs
1
2
3
...
9
10
Employee
ID
Employee
Annual
Salary
Hire
Date
120128
120144
120161
$30,890.00
$30,265.00
$30,785.00
01NOV2010
01OCT2010
01OCT2010
3,089
3,027
3,079
$33,979.00
$33,291.50
$33,863.50
121085
121107
$32,235.00
$31,380.00
01JAN2011
01JUL2010
3,224
3,138
$35,458.50
$34,518.00
Increase
Level 2
5. Subsetting Observations Based on Three Conditions
New
Annual
Salary
6-41
a. Write a DATA step to create work.delays using orion.orders as input.

b. Create a new variable, Order_Month, and set it to the month of Order_Date.
Hint: Use the MONTH function.
c. Use a WHERE statement and a subsetting IF statement to select only the observations that
meet all of the following conditions:
Delivery_Date values that are more than four days beyond Order_Date
Employee_ID values that are equal to 99999999
Order_Month values occurring in August
d. The new data set should include only Employee_ID, Customer_ID, Order_Date,
Delivery_Date, and Order_Month.
e. Add permanent labels for Order_Date, Delivery_Date, and Order_Month as shown below.
f. Add permanent formats to display Order_Date and Delivery_Date as MM/DD/YYYY.
g. Add a PROC CONTENTS step to verify that the labels and formats were stored permanently.
#
Variable
Type
2
4
1
3
5
Customer_ID
Delivery_Date
Employee_ID
Order_Date
Order_Month
Num
Num
Num
Num
Num
Len
8
8
8
8
8
Format
Label
12.
MMDDYY10.
12.
MMDDYY10.
Customer ID
Date Delivered
Employee ID
Date Ordered
Month Ordered
h. Write a PROC PRINT step to create the report below. Results should contain nine observations.
Obs
1
2
3
4
5
6
7
8
9
Employee_ID
Customer_ID
Order_Date
Delivery_
Date
99999999
99999999
99999999
99999999
99999999
99999999
99999999
99999999
99999999
70187
52
16
61
2550
70201
9
71
70201
08/13/2007
08/20/2007
08/27/2007
08/29/2007
08/10/2008
08/15/2008
08/10/2009
08/30/2010
08/24/2011
08/18/2007
08/26/2007
09/04/2007
09/03/2007
08/15/2008
08/20/2008
08/15/2009
09/05/2010
08/29/2011
Order_
Month
8
8
8
8
8
8
8
8
8
Challenge
6. Using an IF-THEN/DELETE Statement to Subset Observations
a. Write a DATA step to create work.bigdonations using orion.employee_donations as input.
b. Use the SUM function to create a new variable, Total, which holds the sum of the four quarterly
donations.
6-42
c. Use the N function to create a new variable, NumQtrs, which holds the count of nonmissing
values in Qtr1, Qtr2, Qtr3, and Qtr4. Explore the N function in the SAS Help facility or online
documentation.
d. The new data set should not include the charities or method of payment.
e. The final data set should contain only observations meeting the following two conditions:
Total values greater than or equal to 50
NumQtrs value equal to 4
Use an IF-THEN/DELETE statement to eliminate the observations where the conditions are not
met. Explore the use of IF-THEN/DELETE in the SAS Help facility or online documentation.
f. Store permanent labels in the new data set as shown in the report below.
g. Create the following report to verify that the labels were stored:
#
Variable
Type
1
7
2
3
4
5
6
Employee_ID
NumQtrs
Qtr1
Qtr2
Qtr3
Qtr4
Total
Num
Num
Num
Num
Num
Num
Num
Len
8
8
8
8
8
8
8
Format
Label
12.
Employee ID
First Quarter
Second Quarter
Third Quarter
Fourth Quarter
h. Create the report below. The results should contain 50 observations.

Employee ID
120267
120269
120271
120275
120660
First
Quarter
15
20
20
15
25
Second
Quarter
15
20
20
15
25
Third
Quarter
15
20
20
15
25
Fourth
Quarter
Total
15
20
20
15
25
60
80
80
60
100
Num
Qtrs
4
4
4
4
4
6.3 Solutions
a. What is the name of the variable that contains gender values? Customer_Gender
What are the possible values of this variable? M or F
data work.youngadult;
set orion.customer_dim;
where Customer_Gender='F' and
Customer_Age between 18 and 36 and
Customer_Group contains 'Gold';
Discount=.25;
6.3 Solutions
run;
proc print data=work.youngadult;
var Customer_Name Customer_Age
Customer_Gender Customer_Group Discount;
id Customer_ID;
run;
data work.assistant;
set orion.staff;
where Job_Title contains 'Assistant' and
Salary<26000;
Increase=Salary*.10;
New_Salary=Salary+Increase;
run;
proc print data=work.assistant;
id Employee_ID;
var Job_Title Salary Increase New_Salary;
format Salary Increase New_Salary dollar10.2;
run;
3. Using the SOUNDS-LIKE Operator to Select Observations
data work.tony;
where Customer_FirstName=* 'Tony';
run;
proc print data=work.tony;
var Customer_FirstName Customer_LastName;
run;
4. Subsetting Observations Based on Two Conditions
data work.increase;
set orion.staff;
where Emp_Hire_Date>='01JUL2010'd;
Increase=Salary*0.10;
if Increase>3000;
NewSalary=Salary+Increase;
label Employee_ID='Employee ID'
Salary='Annual Salary'
Emp_Hire_Date='Hire Date'
NewSalary='New Annual Salary';
format Salary NewSalary dollar10.2 Increase comma5.;
keep Employee_ID Emp_Hire_Date Salary Increase NewSalary;
run;
proc print data=work.increase split=' ';
run;
The existing labels and formats were inherited from the input data set.
6-43
6-44
5. Subsetting Observations Based on Three Conditions

data work.delays;
set orion.orders;
where Order_Date+4<Delivery_Date
and Employee_ID=99999999;
Order_Month=month(Order_Date);
if Order_Month=8;
label Order_Date='Date Ordered';
Delivery_Date='Date Delivered'
Order_Month='Month Ordered';
format Order_Date Delivery_Date mmddyy10.;
keep Employee_ID Customer_ID Order_Date Delivery_Date
Order_Month;
run;
proc contents data=work.delays;
run;
proc print data=work.delays;
run;
6. Using an IF-THEN/DELETE Statement to Subset Observations
data work.bigdonations;
set orion.employee_donations;
NumQtrs=n(Qtr1,Qtr2,Qtr3,Qtr4);
if Total<50 or NumQtrs<4 then delete;
label Qtr1='First Quarter'
Qtr2='Second Quarter'
Qtr3='Third Quarter'
Qtr4='Fourth Quarter';
drop Recipients Paid_By;
run;
proc contents data=work.bigdonations;
run;
proc print data=work.bigdonations label noobs;
run;
6.3 Solutions

Considering this DATA step, which statement is true?
a. It reads a temporary data set and creates a permanent
data set.
b. It reads a permanent data set and creates a temporary
data set.
c. It contains a syntax error and will not execute.
d. It will not execute because you cannot work with
permanent and temporary data sets in the same step.
data us;
/* Create a temporary data set */
set orion.sales; /* Read a permanent data set */
where Country='US';
run;
16

Evaluate the assignment statements below given the
values shown in the PDV.
x
y
.
a.
num=y+z/2;
z
10
4+10/2
num=(y+z)/2;
b.
num=x+z/2;
.+10/2
4+5
14/2
9
7
.+5
29
29
6-45
6-46

Open and submit p106a03. Is the output data set created
successfully?
260 data work.usemps;
261
set orion.sales;
262
Bonus=Salary*.10;
263
where Country='US' and Bonus>=3000;
ERROR: Variable Bonus is not on file ORION.SALES.
264 run;
NOTE: The SAS System stopped processing this step because of
errors.
WARNING: The data set WORK.USEMPS may be incomplete. When
this step was stopped there were 0 observations and 10
variables.
No. Bonus cannot be used in a WHERE statement

because it is not in the input data set. It is a new
variable created in this DATA step.
64

What column heading will be displayed for Job_Title in the
program below?
data work.us;
set orion.sales;
where Country='US';
Bonus=Salary*.10;
label Job_Title='Sales Title';
Birth_Date;
run;
label Job_Title='Title';
run;
The column heading will be Title. Labels and formats in
PROC steps override permanent labels and formats.
79
Chapter 7 Reading Spreadsheet

and Database Data
7.1
Reading Spreadsheet Data ............................................................................................ 7-3

Demonstration: Reading Excel Data in SAS Enterprise Guide............................................ 7-10
Demonstration: Reading Excel Data in the SAS Windowing Environment.......................... 7-13
Demonstration: Creating a SAS Data Set ............................................................................ 7-16
Exercises .............................................................................................................................. 7-17
7.2
Reading Database Data ................................................................................................ 7-20
7.3
Solutions ....................................................................................................................... 7-24

7-2
Chapter 7 Reading Spreadsheet and Database Data
7.1 Reading Spreadsheet Data

Objectives
Assign a libref to a Microsoft Excel workbook using a

SAS/ACCESS LIBNAME statement.
Access an Excel worksheet using a SAS two-level
name.
Create a SAS data set using a subset of worksheet
data.
3
3
Business Scenario
The Sales Manager has requested a report about Orion
Star sales employees from Australia and the United
States.
The input data is in an Excel workbook.
sales.xls
4
4
7-3
7-4
Business Scenario
Use SAS/ACCESS Interface to PC Files to read the
worksheets within the sales.xls workbook as if they were
SAS data sets.
sales.xls
5
5
Examine the Workbook

Partial sales.xls
two worksheets
cells formatted as dates
6
6
7-5
SAS/ACCESS LIBNAME Statement

When the bit count of SAS and Microsoft Office are the
same (both are 32-bit or both are 64-bit), you can use the
default SAS/ACCESS Excel engine.
libref
engine name
libname orionx excel "&path\sales.xls";

LIBNAME libref <engine> "workbook-name" <options>;
When the bit counts differ, use the PC Files Server.

libref
engine name
libname orionx pcfiles path="&path\sales.xls";

LIBNAME libref <engine> <PATH=>"workbook-name" <options>;
7
7
SAS/ACCESS provides data connectivity and integration between SAS and third-party data sources,
including Microsoft Excel workbooks and various databases. SAS/ACCESS uses data access engines to
read, write, and update data regardless of the data source or platform.
Both SAS and Microsoft Office offer 32-bit and 64-bit versions. Different SAS/ACCESS engines are
needed based on the products bitness. If the bitness of both products is the same, use the default
SAS/ACCESS Excel engine, If the bitness differs, use the PC Files Server engine and specify PATH=
in front of workbook-name.
The table below summarizes the possible bit combinations and the appropriate engine to use for each.
SAS
Microsoft
Office
SAS/ACCESS Engine
9.4
32-bit
PC Files Server
9.4
64-bit
Default
9.3 32-bit
32-bit
Default
9.3 32-bit
64-bit
PC Files Server
9.3 64-bit
32-bit
PC Files Server
9.3 64-bit
64-bit
Default
SAS 9.2 or earlier
32-bit
Default
SAS 9.2 or earlier
64-bit
PC Files Server
For more information, see the usage note Installing SAS 9.3 PC Files Server and using it to convert
32-bit Microsoft Office files to SAS 64-bit files: http://support.sas.com/kb/43/802.html.
7-6
SAS Explorer Window

SAS treats the workbook as a library, and each worksheet
as a SAS data set.
A named range might exist for each worksheet.

Worksheet names end with a dollar sign.
Named ranges do not end with a dollar sign.
8
8
The named ranges might or might not exist, depending on how the Excel worksheets were created. They
are included here to show the difference between a named range and a worksheet name.
CONTENTS Procedure
proc contents data=orionx._all_;
run;
Directory
Libref
Engine
Physical Name
Schema/Owner
ORIONX
PCFILES
s:\workshop\sales.xls
.
Name
Member
Type
DBMS Member
Type
1
2
3
4
Australia
Australia$
UnitedStates
UnitedStates$
DATA
DATA
DATA
DATA
TABLE
SYSTEM TABLE
TABLE
SYSTEM TABLE
p107d01
7-7

Data Set Name
Member Type
Engine
Created
Last Modified
Protection
Data Set Type
Label
Data Representation
Encoding
ORIONX.'Australia$'n
DATA
PCFILES
.
.
Observations
Variables
Indexes
Observation Length
Compressed
Sorted
.
9
0
0
0
NO
NO
Default
Default

#
Variable
Type
Len
Format
Informat
Label
8
7
1
2
4
9
6
3
5
Birth_Date
Country
Employee_ID
First_Name
Gender
Hire_Date
Job_Title
Last_Name
Salary
Num
Char
Num
Char
Char
Num
Char
Char
Num
8
2
8
10
1
8
14
12
8
DATE9.
$2.
DATE9.
$2.
$10.
$1.
DATE9.
$14.
$12.
$10.
$1.
DATE9.
$14.
$12.
Birth Date
Country
Employee ID
First Name
Gender
Hire Date
Job Title
Last Name
Salary
10
10
The column headings are used to create variable names. In the SAS windowing environment, embedded
spaces in column names are replaced with underscores. In SAS Enterprise Guide, the column headings
are used without modification because special characters are allowed in variable names. Set the
VALIDVARNAME=V7 option in SAS Enterprise Guide to cause Enterprise Guide to behave the same as
the windowing environment.
Some fields are missing in the PROC CONTENTS output because Excel metadata is incomplete.

Data Set Name
Member Type
Engine
Created
Last Modified
Protection
Data Set Type
Label
Data Representation
Encoding
ORIONX.UnitedStates$'n
DATA
PCFILES
.
.
Observations
Variables
Indexes
Observation Length
Compressed
Sorted
.
9
0
0
0
NO
NO
Default
Default

#
Variable
Type
Len
Format
Informat
Label
8
7
1
2
4
9
6
3
5
Birth_Date
Country
Employee_ID
First_Name
Gender
Hire_Date
Job_Title
Last_Name
Salary
Num
Char
Num
Char
Char
Num
Char
Char
Num
8
2
8
10
1
8
14
12
8
DATE9.
$2.
DATE9.
$2.
$10.
$1.
DATE9.
$14.
$12.
$10.
$1.
DATE9.
$14.
$12.
Birth Date
Country
Employee ID
First Name
Gender
Hire Date
Job Title
Last Name
Salary
11
11
7-8
SAS Name Literals

A SAS name literal is a string within quotation marks,
followed by the letter n.
orionx.'Australia$'n
SAS name literal
SAS name literals permit special characters in data set

names.
12
12
Printing an Excel Worksheet

proc print data=orionx.'Australia$'n;
run;

Employee_
Obs
ID
1
2
3
4
5
13
120102
120103
120121
120122
120123
First_Name Last_Name
Tom
Wilson
Irenie
Christina
Kimiko
Zhou
Dawes
Elvish
Ngan
Hotstone
Gender Salary Job_Title

M
M
F
F
F
108255
87975
26600
27475
26190
Sales
Sales
Sales
Sales
Sales
Manager
Manager
Rep. II
Rep. II
Rep. I
Country
AU
AU
AU
AU
AU
Birth_
Date
Hire_Date
11AUG1973
22JAN1953
02AUG1948
27JUL1958
28SEP1968
01JUN1993
01JAN1978
01JAN1978
01JUL1982
01OCT1989
p107d01
13
Subsetting Worksheet Data

You can select a subset of the worksheet data.

proc print data=orionx.'Australia$'n noobs;
where Job_Title ? 'IV';
var Employee_ID Last_Name Job_Title Salary;
run;
p107d01
14
14
Viewing the Output

PROC PRINT Output
Employee_
ID
120125
120128
120135
120159
120166
Last_Name
Job_Title
Hofmeister
Kletschkus
Platts
Phoumirath
Nowd
Sales
Sales
Sales
Sales
Sales
Rep.
Rep.
Rep.
Rep.
Rep.
Salary
IV
IV
IV
IV
IV
32040
30890
32490
30765
30660
15
15
7-9
7-10
Disassociating a Libref
If SAS has a libref assigned to an Excel workbook, the
workbook cannot be opened in Excel. To disassociate the
libref, use a LIBNAME statement with the CLEAR option.

/* program to access the worksheets */
libname orionx clear;
SAS disconnects from the data source and closes any
resources associated with the connection.
p107d01
16
16
7.01 Quiz
Which PROC PRINT step displays the worksheet
containing employees from the United States?
a. proc print data=orionx.'UnitedStates';
run;
b. proc print data=orionx.'UnitedStates$';

run;
c. proc print data=orionx.'UnitedStates'n;

run;
d. proc print data=orionx.'UnitedStates$'n;

run;
17
17
Reading Excel Data in SAS Enterprise Guide

p107d01
1. In the Server List window, select Servers Local Libraries. If ORION is expanded, click the
minus sign to collapse it.
7-11
2. Open p107d01 and submit only the LIBNAME statement.

3. Check the log for success.
16
NOTE: Libref ORIONX was successfully assigned as follows:
Engine:
PCFILES
Physical Name: s:\workshop\sales.xls
4. In the Server List window, select Libraries and click Refresh to refresh the list. Verify that
ORIONX is displayed as an active library. Expand ORIONX to see its contents. The named ranges
might not exist.
5. In the Server List window, right-click Australia$ and select Properties. Then click the Column tab
to see the variable information. Notice that the variable names contain embedded blanks. SAS
Enterprise Guide allows special characters in variable names without the need for a name literal.
7-12
Close the Properties window.

6. Return to the Program window and submit the PROC CONTENTS step to see the same information
in a report. Again, notice that there are embedded blanks in the variable names.
7. The VALIDVARNAME= option instructs SAS Enterprise Guide to follow the same rules as the SAS
windowing environment in regard to variable names. To set this option, return to the Program
window, and uncomment and submit the OPTIONS statement.
options validvarname=v7;
8. Resubmit the PROC CONTENTS step and observe that the embedded blanks have been replaced
with underscores.
9. In the Server List window, double-click Australia$ to see the data displayed in a data grid.
7-13
Close the data grid.

10. Return to the Program window and submit the PROC PRINT step to see the same information in a
report.
11. Return to the Program window and submit the final LIBNAME statement to clear the libref, releasing
the spreadsheet. Check the log or the Server List window for success. You might need to refresh the
Server List window.
To restore the default variable name behavior, submit another OPTIONS statement to set
VALIDVARNAME=ANY.
Reading Excel Data in the SAS Windowing Environment

p107d01
1. Activate the Explorer window and double-click Libraries to display the active libraries.
2. Open p107d01 and submit only the LIBNAME statement.
64
NOTE: Libref ORIONX was successfully assigned as follows:
Engine:
PCFILES
Physical Name: s:\workshop\sales.xls
4. Use the Explorer window to verify that orionx is active. Double-click orionx to drill into it to see its
contents. The named ranges might not exist.
7-14
5. In the Explorer window, right-click Australia$ and select View Columns to see the variable
information. Notice that the variable names contain underscores in place of embedded blanks, and the
Label column contains the original spreadsheet column headings.
Close the Properties window.

6. Return to the Editor and submit the PROC CONTENTS step to see the same information in a report.
#
Variable
Type
Len
Format
Informat
Label
8
7
1
2
4
9
6
3
5
Birth_Date
Country
Employee_ID
First_Name
Gender
Hire_Date
Job_Title
Last_Name
Salary
Num
Char
Num
Char
Char
Num
Char
Char
Num
8
2
8
12
1
8
20
18
8
DATE9.
$2.
DATE9.
$2.
$12.
$1.
DATE9.
$20.
$18.
$12.
$1.
DATE9.
$20.
$18.
Birth Date
Country
Employee ID
First Name
Gender
Hire Date
Job Title
Last Name
Salary
In the Explorer window, double-click Australia$ to see the data portion. The labels are displayed as
column headings. Close the table.
7. Return to the Editor and submit the PROC PRINT step to see the same information in a report.
ID
120125
120128
120135
120159
120166
Last_Name
Job_Title
Hofmeister
Kletschkus
Platts
Phoumirath
Nowd
Sales
Sales
Sales
Sales
Sales
7-15
Employee_
Salary
Rep.
Rep.
Rep.
Rep.
Rep.
IV
IV
IV
IV
IV
32040
30890
32490
30765
30660
8. Submit the final LIBNAME statement to clear the libref, releasing the spreadsheet. Check the log or
Explorer window for success.
Business Scenario
Create a SAS data set using a Microsoft Excel workbook
as input.
sales.xls
20
20
Considerations
Use a SAS/ACCESS LIBNAME statement to read the
Australia$ worksheet and create a temporary data set.
Australia$
DATA Step
work.subset2
Rep
The new data set should include the following:
only the employees with Rep in their job title
a Bonus variable that is 10% of Salary
permanent labels and formats
21
21
7-16
7.02 Poll
A DROP or KEEP statement can be used to control which
worksheet columns are written to the new data set.
True
False
22
Creating a SAS Data Set

p107d02
If you are using SAS Enterprise Guide, remember to uncomment and submit the OPTIONS
statement to set VALIDVARNAME=V7.
libname orionxls pcfiles path="&path\sales.xls";
*options validvarname=v7;
/* Needed for SAS Enterprise Guide */
data work.subset2;
set orionxls.'Australia$'n;
where Job_Title contains 'Rep';
Bonus=Salary*.10;
keep First_Name Last_Name Salary Bonus
Job_Title Hire_Date;
format Salary comma10. Hire_Date mmddyy10.
Bonus comma8.2;
run;
run;
run;
libname orionxls clear;
7-17
1. Open p107d02 and submit the LIBNAME statement.

2. Use the Server List window or SAS Explorer to explore the active libraries and drill into orionxls.
Observe that the two spreadsheets, Australia$ and UnitedStates$, are displayed as if they were SAS
data sets.
3. Submit the DATA step.
4. Check the log. Verify that 61 observations were written to work.subset2.
If you are using SAS Enterprise Guide, the following error occurs if the VALIDVARNAME=V7
option was not set:
ERROR: Variable Job_Title is not on file ORIONXLS.'Australia$'n.
Job_Title is not found in Australia$. Job Title is found instead, because SAS Enterprise Guide
permits blanks and special characters in variable names. Setting the VALIDVARNAME=v7 option
causes SAS Enterprise Guide to use the same variable naming rules as the SAS windowing
environment. Be sure to set the option as explained in the note above.
5. Run the PROC CONTENTS step. Verify that the formats and labels were stored in the descriptor
portion of work.subset2. Notice that all columns have labels, not just the columns that were assigned
labels in the DATA step.
#
Variable
Type
Len
6
1
5
4
2
3
Bonus
First_Name
Hire_Date
Job_Title
Last_Name
Salary
Num
Char
Num
Char
Char
Num
8
10
8
14
12
8
Format
Informat
Label
COMMA8.2
$10.
MMDDYY10.
$14.
$12.
COMMA10.
$10.
DATE9.
$14.
$12.
First Name
Date Hired
Sales Title
Last Name
Salary
6. Submit the PROC PRINT step and verify that the results contain 61 observations and that the labels
were displayed and formats applied.
First Name
Last Name
Salary
Sales Title
Date Hired
Bonus
Irenie
Christina
Kimiko
...
Alban
Alena
Elvish
Ngan
Hotstone
26,600
27,475
26,190
Sales Rep. II
Sales Rep. II
Sales Rep. I
01/01/1978
07/01/1982
10/01/1989
2,660.00
2,747.50
2,619.00
Kingston
Moody
28,830
26,205
Sales Rep. III

Sales Rep. II
10/01/1996
09/01/2010
2,883.00
2,620.50
7. Submit the LIBNAME statement to clear the resources.

Exercises
7-18
Level 1
1. Accessing an Excel Worksheet
b. Add a LIBNAME statement before the PROC CONTENTS step to create a libref named
CUSTFM that references the Excel workbook, custfm.xls.
If you are using SAS Enterprise Guide, you need to set the VALIDVARNAME=V7
option if not already set in the session.
c. Submit the LIBNAME statement and the PROC CONTENTS step to create the following partial
PROC CONTENTS report:
Part 1 of 3
Directory
Libref
Engine
Physical Name
User
CUSTFM
EXCEL
custfm.xls
Admin
Name
Member
Type
1
2
Females$
Males$
DATA
DATA
DBMS
Member
Type
TABLE
TABLE
d. Add a SET statement in the DATA step to read the worksheet containing the male data.
e. Add a KEEP statement in the DATA step to include only the First_Name, Last_Name, and
Birth_Date variables in the new data set.
f. Add a FORMAT statement in the DATA step to display Birth_Date as a four-digit year.
g. Add a LABEL statement to change the column heading of Birth_Date to Birth Year.
h. Submit the program including the final LIBNAME statement and create the report below. Results
should contain 47 observations.
i. Verify that the worksheet was released.
Obs
1
2
3
4
5
First Name
Last Name
Birth
Year
James
David
Markus
Ulrich
Jimmie
Kvarniq
Black
Sepke
Heyde
Evans
1974
1969
1988
1939
1954
7-19
Level 2
a. Write a LIBNAME statement to create a libref named PROD that references the Excel workbook
products.xls.
b. Write a PROC CONTENTS step to view all of the contents of PROD.
c. Submit the program to determine the names of the four worksheets in products.xls.
d. Write a DATA step that reads the worksheet containing sports data and creates a new data set
named work.golf.
The data set work.golf should
include only the observations where Category is equal to Golf
not include the Category variable
include a label of Golf Products for the Name variable.
e. Write a LIBNAME statement to clear the PROD libref.
f. Create the report below. The results should contain 56 observations.
Obs
1
2
3
4
5
Golf Products
Ball Bag
Red/White/Black Staff 9 Bag
Tee Holder
Bb Softspikes - Xp 22-pack
Bretagne Performance Tg Men's Golf Shoes L.
Challenge
3. Creating an Excel Spreadsheet
a. Open p107e03. Insert a LIBNAME statement to associate the libref out with the Excel workbook
employees.xls in the default data folder.
b. Modify the program so that it creates a spreadsheet named salesemps in the employees.xls
workbook.
c. Submit the SAS program and verify that it created the data set out.salesemps with 71
observations and 4 variables as shown in the partial SAS log below.
NOTE: 71 records were read from the infile "s:\workshop\newemps.csv".
The minimum record length was 28.
The maximum record length was 47.
NOTE: The data set OUT.salesemps has 71 observations and 4 variables.
The program will fail if the workbook already exists. Use Windows Explorer to navigate
to the data folder and delete employees.xls.
7-20
4. Using the IMPORT and EXPORT Procedures

Use the SAS Help facility or online documentation to investigate the IMPORT and EXPORT
procedures. Describe the chief differences from using the SAS/ACCESS LIBNAME statement.
7.2 Reading Database Data

Objectives
Assign a libref to an Oracle database using a

Access an Oracle table using a SAS two-level name.
Create a SAS data set that contains a subset of an
Oracle table.
28
28
Business Scenario
The Northeast Sales Manager requested a report listing
supervisors from New York and New Jersey. The input
data is in an Oracle database.
29
29
7-21
Business Scenario
Use SAS/ACCESS to read the tables within the database
as if they were SAS data sets.
30
30
SAS/ACCESS LIBNAME Statement

The SAS/ACCESS LIBNAME statement assigns a libref
to a relational database.
libref
engine name
options
libname oralib oracle user=edu001 pw=edu001

path=dbmssrv schema=educ;
LIBNAME libref engine <SAS/ACCESS options>;
This example uses the LIBNAME statement supported by
the SAS/ACCESS Interface to Oracle.
31
31
The engine name, such as Oracle or DB2, is the SAS/ACCESS component that reads and writes to your
DBMS. The engine name is required.
USER= specifies an optional Oracle user name. USER= must be used with PASSWORD=.
PASSWORD= (or PW=) specifies an optional Oracle password associated with the Oracle user name.
PATH= specifies the Oracle driver, node, and database. SAS/ACCESS uses the same Oracle path
designation that you use to connect to Oracle directly.
7-22
SCHEMA= enables you to read database objects, such as tables and views, in the specified schema. If this
option is omitted, you connect to the default schema for your DBMS.
Accessing an Oracle Database

The Oracle database is treated like a SAS library.
Any table in this Oracle database can be referenced

using a SAS two-level name.
32
32

Which of the following statements will read the Oracle
table supervisors as if it were a SAS data set?
a. input oralib.supervisors;
b. input 'oralib.supervisors$'n;
c. set oralib.supervisors;
d. set 'oralib.supervisors$'n;
33
Printing an Oracle Table

libname oralib oracle
user=edu001 pw=edu001
path=dbmssrv schema=educ;
proc print data=oralib.supervisors;
where state in ('NY' 'NJ');
run;
libname oralib clear;
Remember to release the database and associated
resources by submitting a LIBNAME statement that
contains the CLEAR option.
35
35
Viewing the Output

Obs
1
2
3
4
5
EMPID
STATE
JOBCATEGORY
1834
1433
1983
1420
1882
NY
NJ
NY
NJ
NY
BC
FA
FA
ME
ME
36
36
7-23
7-24
Creating a SAS Data Set

libname oralib oracle user=edu001
pw=edu001 path=dbmssrv schema=educ;
data nynjsup;
set oralib.supervisors;
where state in ('NY' 'NJ');
run;
proc print data=nynjsup;
run;
libname oralib clear;
37
37
Viewing the Output

PROC PRINT Output
Obs
1
2
3
4
5
EMPID
STATE
JOBCATEGORY
1834
1433
1983
1420
1882
NY
NJ
NY
NJ
NY
BC
FA
FA
ME
ME
38
38
7.3 Solutions
libname custfm pcfiles path="&PATH\custfm.xls";
/* needed for SAS Enterprise Guide */
7.3 Solutions
proc contents data=custfm._all_;

run;
data work.males;
set custfm.'Males$'n;
keep First_Name Last_Name Birth_Date;
format Birth_Date year4.;
label Birth_Date='Birth Year';
run;
proc print data=work.males label;
run;
libname custfm clear;
libname prod pcfiles path="&PATH\products.xls";
/* SAS Enterprise Guide only */
proc contents data=prod._all_;
run;
data work.golf;
set prod.'Sports$'n;
where Category='Golf';
drop Category;
label Name='Golf Products';
run;
libname prod clear;
proc print data=work.golf label;
run;
3. Creating an Excel Spreadsheet
/* delete employees.xls from the data directory if it exists */
libname out pcfiles path="&path\employees.xls";
/* SAS Enterprise Guide only */
data out.salesemps;
Job_Title $ 25;
Job_Title $ Salary;
run;
libname out clear;
4. Using the IMPORT and EXPORT Procedures
7-25
7-26
Explore the IMPORT and EXPORT procedures. Describe the chief differences from using the
When you use the SAS/ACCESS LIBNAME statement, you are
accessing the most recent data in the workbook or database.
not necessarily making a SAS copy of the data. You can use PROC PRINT (or other
procedures) directly on a sheet or table through a dynamic "pipeline."
When you import, you are
duplicating storage, in a sense
creating a static SAS copy of the data, which might get out of date unless you remember to
re-import
allowing for easier programming (for example, no name literals).

Which PROC PRINT step displays the worksheet
containing employees from the United States?
a. proc print data=orionx.'UnitedStates';
run;
b. proc print data=orionx.'UnitedStates$';

run;
c. proc print data=orionx.'UnitedStates'n;

run;
d. proc print data=orionx.'UnitedStates$'n;

run;
18
18
7.3 Solutions
7.02 Poll Correct Answer

A DROP or KEEP statement can be used to control which
worksheet columns are written to the new data set.
True
False
23

Which of the following statements will read the Oracle
table supervisors as if it were a SAS data set?
a. input oralib.supervisors;
b. input 'oralib.supervisors$'n;
c. set oralib.supervisors;
d. set 'oralib.supervisors$'n;
34
7-27
7-28
Chapter 8 Reading Raw Data Files

8.1
Introduction to Reading Raw Data Files ....................................................................... 8-3
8.2
Reading Standard Delimited Data ................................................................................. 8-7

Demonstration: Examining Data Errors ............................................................................... 8-33
Exercises .............................................................................................................................. 8-34
8.3
Reading Nonstandard Delimited Data ........................................................................ 8-37

Demonstration: Using List Input: Importance of Colon Format Modifier .............................. 8-48
Exercises .............................................................................................................................. 8-51
8.4
Handling Missing Data ................................................................................................. 8-54

Exercises .............................................................................................................................. 8-60
8.5
Solutions ....................................................................................................................... 8-63

8-2
8.1 Introduction to
o Reading Raw
w Data Files
8.1 Introd
ductio
on to Readiing Ra
aw Da
ata
Files
Ob
bjectives
Identify type
es of raw data
a files and in
nput styles.
Define the te
erms standard and nonsttandard data
a.
3
3
Copyright 2013, SAS Institute Inc., Cary, Nortth Carolina, USA. ALL RIGHTS RE
ESERVED.
8-3
8-4
Raw Data Files

A raw data file is also known as a flat file.
They are text files that contain one record per line.
A record typically contains multiple fields.
Flat files do not have internal metadata.
External documentation, known as a record layout,
should exist.
A record layout describes the fields and locations
within each record.
5
5
Raw Data Files

Fields in a raw data file can be delimited or arranged in
fixed columns.
Delimited File
120102,Tom,Zhou,M,108255,Sales Manager,AU,11AUG1973,06/01/1993
120103,Wilson,Dawes,M,87975,Sales Manager,AU,22JAN1953,01/01/1978
120121,Irenie,Elvish,F,26600,Sales Rep. II,AU,02AUG1948,01/01/1978
120122,Christina,Ngan,F,27475,Sales Rep. II,AU,27JUL1958,07/01/1982
Fixed Column File

1
1
2
2
3
3
4
4
5
5
6
1---5----0----5----0----5----0----5----0----5----0----5----0-120102Tom
Zhou
Sales Manager
108255AU
120103Wilson
Dawes
Sales Manager
87975AU
120121Irenie
Elvish
Sales Rep. II
26600AU
120122Christina Ngan
Sales Rep. II
27475AU
6
6
8.1 Introduction to Reading Raw Data Files
Fields in Raw Data Files

In order for SAS to read a raw data file, you must specify
the following information about each field:
the location of the data value in the record
the name of the SAS variable in which to store the
data
the type of the SAS variable
7
7
Reading Raw Data Files

There are different techniques, or input styles, for reading
raw data files in SAS.
Input Style
Column Input
Used for Reading

Standard data in fixed columns
Formatted Input Standard and nonstandard data in fixed columns

List Input
Standard and nonstandard data separated by

blanks or some other delimiter
8
8
8-5
8-6
Chapter 8 Read
ding Raw Data Files
8.001 Multiplee Answerr Poll

What type of raw
w data files do
d you read?
?
a.
b.
c.
d.
delimited
mn
fixed colum
both delimitted and fixed
d column
I do not rea
ad raw data files.
f
10
10
ESERVED.
8.2 Reading
g Standard Dellimited Data
8.2 Reading Sttanda

ard De
elimite
ed Datta
Ob
bjectives
Use list inpu

ut to create a SAS data set
s from a de
elimited
raw data file
e.
Examine the
e compilation
n and executtion phases of
o the
DATA step when
w
reading
g a raw data
a file.
Explicitly deffine the leng
gth of a variable.
Examine behavior when a data errorr is encounte
ered.
12
12
ESERVED.
8-7
8-8
List Input
Use list input to read delimited raw data files.
Partial sales.csv
120123,Kimiko,Hotstone,F,26190,Sales Rep. I,AU,28SEP1968,10/01/1989
SAS considers a space (blank) to be the default

delimiter.
Both standard and nonstandard data can be read.
Fields must be read sequentially, left to right.
14
14
8.02 Quiz
Which fields in this file can be read as standard numeric
values?
Partial sales.csv
15
8.2 Reading
8-9
DLM= is an alias for DELIMITER=

D
=delimiter(s). It is used to specify an altternate delimiiter (other thaan a
blank) to be
b used for list input. Dellimiter(s) is on
ne or more chharacters encllosed in quotaation marks. If
more than
n one delimiteer is specified
d, each is treatted as a delim
miter. The deliimiter is case sensitive.
To speccify a tab delimiter on Win
ndows or UNIIX, type dlm=
=09x.
To speccify a tab delimiter on z/OS
S (OS/390), ty
ype dlm=055x.
A %LET statement was submitted in

n the libnamee.sas program
m to create thee user-definedd macro variaable,
path. Thee INFILE statement referen
nces the valuee of the macroo variable by placing an am
mpersand beffore
its name. The macro vaariable will ex
xist for the du
uration of the SAS session unless it is chhanged or delleted.
ESERVED.
8-10
Chapter 8 Read
ding Raw Data Files
Vieewing the Log

Partial SAS Log
249
250
251
252
253
da
ata work.subset;
d
infile "&path\sales.csv" dlm=',';
yee_ID First_Na
ame $ Last_Nam
me $
input Employ
Gender
r $ Salary Job_
_Title $ Count
try $;
un;
ru
NOTE: T
The infile "s:\workshop\sale
es.csv" is:
F
Filename=s:\wo
orkshop\sales.c
csv,
R
RECFM=V,LRECL=
=256,File Size (bytes)=11340
0
NOTE: 1
165
T
The
T
The
NOTE: The
T
20
records were read from the

t
infile "s:\workshop\sales.csv".
minimum record length wa
as 61.
maximum record length wa
as 80.
data set WORK.SUBSET
W
has
s 165 observat
tions and 7 variables.
p108d01
20
ESERVED.
8.2 Reading
ESERVED.
8-11
8-12
Chapter 8 Read
ding Raw Data Files
Co
ompilation
n
data work.subs
set;
in
nfile "&pa
ath\sales.
.csv" dlm
m=',';
in
nput Emplo
oyee_ID Fi
irst_Name $ Last_N
Name $
Gende
er $ Salar
ry Job_Title $ Cou
untry $;
run;
24
p108d01
24
ESERVED.
...
8.2 Reading
The defau
ult length of th
he input buffeer depends on
n the operatin g system. It ccan be modified using the
LRECL= option in the INFILE stateement.
ESERVED.
8-13
8-14
Chapter 8 Read
ding Raw Data Files
ESERVED.
8.2 Reading
8.003 Multiplee Choice Poll

Which statemen
nt is true?
a. An input bu
uffer is create
ed only if you
u are reading
g data
from a raw data file.
b. The PDV att compile tim
me holds the variable nam
me,
type, byte size,
s
and initiial value.
c. The descrip
ptor portion iss the first item that is cre
eated
at compile time.
t
30
30
ESERVED.
8-15
8-16
Chapter 8 Read
ding Raw Data Files
ESERVED.
8.2 Reading
ESERVED.
8-17
8-18
Chapter 8 Read
ding Raw Data Files
ESERVED.
8.2 Reading
ESERVED.
8-19
8-20
Chapter 8 Read
ding Raw Data Files
There is usually
u
no dellimiter after th
he last field in
n a record, soo SAS stops reeading when it encounters an
end-of-reccord marker.
ESERVED.
8.2 Reading
Execution
Herre is the outp
put data set after
a
the firstt iteration of the
DAT
TA step.
work.subset
Emplo
oyee
_ID
D
First
e
_Name
12
20102 Tom
Last
G
Gender
Salary
y
_Name
Zhou
Job
_Title
108255 Sales Ma
Country
AU
42
42
ESERVED.
...
8-21
8-22
Chapter 8 Read
ding Raw Data Files
ESERVED.
8.2 Reading
ESERVED.
8-23
8-24
Chapter 8 Read
ding Raw Data Files
ESERVED.
8.2 Reading Standard Delimited Data

Which statement is true of a DATA step when reading
from a raw data file?
a. Data is read from the raw data file into the PDV.
b. The size of the input buffer adjusts automatically
based on the length of the input record.
c. At the bottom of the DATA step, the contents of the
PDV are output to the output SAS data set.
50
50
Viewing the Output

proc print data=work.subset noobs;
run;
Employee_
ID
120102
120103
120121
120122
120123
120124
120125
First_
Name
Last_
Name
Tom
Wilson
Irenie
Christin
Kimiko
Lucian
Fong
Zhou
Dawes
Elvish
Ngan
Hotstone
Daymond
Hofmeist
Gender
Salary
M
M
F
F
F
M
M
108255
87975
26600
27475
26190
26480
32040
Job_
Title
Sales
Sales
Sales
Sales
Sales
Sales
Sales
Ma
Ma
Re
Re
Re
Re
Re
Country
AU
AU
AU
AU
AU
AU
AU
Some character values are truncated.

52
p108d01
52
Some character values contain unnecessary trailing blanks, although this is not obvious from a
PROC PRINT report.
8-25
8-26
Chapter 8 Read
ding Raw Data Files
The LENG
GTH statemeent is used primarily for ch
haracter variabbles.
Co
ompilation
n
data work.subs
set;
ength Firs
st_Name $ 12 Last_N
Name $ 18
le
Gend
der $ 1 Jo
ob_Title $ 25
Coun
ntry $ 2;
in
nfile "&pa
ath\sales.
.csv" dlm=
=',';
in
nput Emplo
oyee_ID Fi
irst_Name $ Last_Na
ame $
Gende
er $ Salar
ry Job_Tit
tle $ Coun
ntry $;
run;
PDV
First_
Name
$ 12
Attributes are bas

sed on the LEN
NGTH stateme
ent.
Last
_Name
$ 18
Gender
$1
Job_T
Title
$ 25
Countrry
$2
54
...
54
The namee, type, and leength of a variiable are deteermined at thee variables firrst use. Thesee specificationns
can be in a LENGTH statement
s
or th
he INPUT staatement, whicchever appearrs first in the DATA step. T
The
name is used exactly ass specified at first use, inclluding the casse.
ESERVED.
Compilation
data work.subset;
Gender $ 1 Job_Title $ 25
Country $ 2;
input Employee_ID First_Name $ Last_Name $
Gender $ Salary Job_Title $ Country $;
run;
PDV
First_
Name
$ 12
Last
_Name
$ 18
Gender
$1
Job_Title
$ 25
Country
$2
Employee_
ID
N8
Salary
N8
55
55
Viewing the Output

proc print data=work.subset noobs;
run;
First_
Name
Last_Name
Tom
Wilson
Irenie
Christina
Kimiko
Zhou
Dawes
Elvish
Ngan
Hotstone
Gender Job_Title
M
M
F
F
F
Sales
Sales
Sales
Sales
Sales
Manager
Manager
Rep. II
Rep. II
Rep. I
Country
AU
AU
AU
AU
AU
Employee_
ID
Salary
120102
120103
120121
120122
120123
108255
87975
26600
27475
26190
The character values are no longer truncated, but the order

of the variables has changed.
56
p108d02
56
8-27
8-28
8.05 Quiz
Suppose you want the order of the variables to match the
order of the fields. You can include the numeric variables
in the LENGTH statement. Which of the following
produces the correct results?
a. length Employee_ID First_Name $ 12
Last_Name $ 18 Gender $ 1
Salary Job_Title $ 25
Country $ 2;
b. length Employee_ID 8 First_Name $ 12
Salary 8 Job_Title $ 25
Country $ 2;
57
Using a LENGTH Statement

The LENGTH statement identifies the character variables,
so dollar signs can be omitted from the INPUT statement.
data work.subset;
length Employee_ID 8 First_Name $ 12
Country $ 2;
input Employee_ID First_Name Last_Name
Gender Salary Job_Title Country;
run;
59
p108d03
Viewing the Output

Display the variables in creation order.
proc contents data=work.subset varnum;

run;
Variables in Creation Order
#
Variable
Type
Len
1
2
3
4
5
6
7
Employee_ID
First_Name
Last_Name
Gender
Salary
Job_Title
Country
Num
Char
Char
Char
Num
Char
Char
8
12
18
1
8
25
2
60
60
Viewing the Output

Employee_
ID
120102
120103
120121
120122
120123
120124
120125
First_
Name
Last_Name
Gender
Salary
Job_Title
Tom
Wilson
Irenie
Christina
Kimiko
Lucian
Fong
Zhou
Dawes
Elvish
Ngan
Hotstone
Daymond
Hofmeister
M
M
F
F
F
M
M
108255
87975
26600
27475
26190
26480
32040
Sales
Sales
Sales
Sales
Sales
Sales
Sales
Manager
Manager
Rep. II
Rep. II
Rep. I
Rep. I
Rep. IV
Country
AU
AU
AU
AU
AU
AU
AU
61
61
8-29
8-30
Chapter 8 Read
ding Raw Data Files
8.006 Quiz
What problems do you see with the data
a values for the
t
lastt two data fie
elds, Salary and
a Country
y?
Partial sales3in
nv.csv
120
0102,Tom,Zhou,Manager,108255
5,AU
120
0103,Wilson,Daw
wes,Manager,87
7975,AU
120
0121,Irenie,Elvissh,Rep. II,26600,AU
120
0122,Christina,N
Ngan,Rep. II,n/a,A
AU
120
0123,Kimiko,Hottstone,Rep. I,261
190,AU
120
0124,Lucian,Dayymond,Rep. I,26480,12
120
0125,Fong,Hofm
meister,Rep. IV,32
2040,AU
64
64
ESERVED.
Reading a Raw Data File with Data Errors

data work.sales;
infile "&path\sales3inv.csv" dlm=',';
input Employee_ID First $ Last $
Job_Title $ Salary Country $;
run;
proc print data=work.sales;
run;
Salary is defined as numeric and Country as character.
p108d04
66
66
Viewing the Output

Obs
1
2
3
4
5
6
7
Employee_
ID
120102
120103
120121
120122
120123
120124
120125
First
Last
Tom
Wilson
Irenie
Christina
Kimiko
Lucian
Fong
Zhou
Dawes
Elvish
Ngan
Hotstone
Daymond
Hofmeister
Job_
Title
Salary
Manager
Manager
Rep. II
Rep. II
Rep. I
Rep. I
Rep. IV
108255
87975
26600
.
26190
26480
32040
Country
AU
AU
AU
AU
AU
12
AU
A missing value was stored in Salary for the input

value n/a.
The value 12 was successfully stored in Country.
A data error occurred on observation 4 but not on
observation 6.
67
8-31
8-32
Viewing the Log

Partial SAS Log
480
481
482
483
484
data work.sales;
run;
NOTE: The infile "s:\workshop\sales3inv.csv" is:

Filename=s:\workshop\sales3inv.csv,
RECFM=V,LRECL=256,File Size (bytes)=1972,
NOTE: Invalid data for Salary in line 4 31-33.
RULE:
----+----1----+----2----+----3----+----4----+----54
120122,Christina,Ngan,Rep. II,n/a,AU 36
Employee_ID=120122 First=Christin Last=Ngan Job_Title=Rep. II Salary=.
Country=AU _ERROR_=1 _N_=4
NOTE: 50 records were read from the infile "s:\workshop\sales3inv.csv".
NOTE: The data set WORK.SALES has 50 observations and 6 variables.
A data error occurs when a data value does not match the
field specification.
68
Even though these are referred to as data errors, they generate notes, not error messages. Syntax errors
stop the DATA step, whereas data errors allow processing to continue.
Data Errors
When this kind of data error occurs, the following
information is written to the SAS log:
a note describing the error
a column ruler
the input record
the contents of the PDV
RULE:
----+----1----+----2----+----3----+----4----+----54
A missing value is assigned to the corresponding

variable, and execution continues.
69
69
8-33
Data Errors
Two temporary variables are created during the
processing of every DATA step:
_N_ is the DATA step iteration counter.
_ERROR_ indicates data error status.
0 indicates that no data error occurred on that
record.
1 indicates that one or more data errors occurred
on that record.
RULE:
----+----1----+----2----+----3----+----4----+----54
70
70
_ERROR_ is reset to 0, and _N_ is incremented by 1 at the beginning of each iteration.
Examining Data Errors

p108d04
data work.sales;
run;
proc print data=work.sales;
run;
1. Open the raw data file sales3inv.csv.
In SAS Enterprise Guide, select File Open Data. Navigate to your data folder.
In the SAS windowing environment, select File Open Program. Select All Files (*.*) from the
File of type menu.
Select sales3inv.csv from the list of files.

2. View the Salary values. There are three invalid values and one missing value.
8-34

4. Examine the log. The DATA step creates an output data set, work.sales, with 50 observations. The
three invalid Salary values were identified in the log, and the corresponding observations have a
missing value for Salary. The missing Salary value was not reported as a data error because missing
is a valid value in SAS.
5. Examine the output. Notice that there are four observations with missing Salary values.

Submit program p108a01 and examine the log.
Which statement best describes the reason for the error?
a. The data in the raw data file is invalid.
b. The programmer incorrectly read the data.
72
72
Use the SAS system option ERRORS=n to specify the maximum number of observations for which error
messages about data input errors are printed.
options errors=5;
Exercises
Level 1
1. Reading a Comma-Delimited Raw Data File
a. Open p108e01. Add the appropriate LENGTH, INFILE, and INPUT statements to read the
comma-delimited raw data file named the following:
Windows
"&path\newemps.csv"
UNIX
"&path/newemps.csv"
z/OS (OS/390)
"&path..rawdata(newemps)"
Partial Raw Data File

Satyakam,Denny,Sales Rep. II,26780
Monica,Kletschkus,Sales Rep. IV,30890
Kevin,Lyon,Sales Rep. I,26955
Petrea,Soltau,Sales Rep. II,27440
Marina,Iyengar,Sales Rep. III,29715
b. Read the following fields from the raw data file:

Name
Type
Length
First
Character
12
Last
Character
18
Title
Character
25
Salary
Numeric
c. Submit the program to create the report below. The results should contain 71 observations.
Obs
1
2
3
4
5
First
Last
Satyakam
Monica
Kevin
Petrea
Marina
Denny
Kletschkus
Lyon
Soltau
Iyengar
Title
Sales
Sales
Sales
Sales
Sales
Rep.
Rep.
Rep.
Rep.
Rep.
Salary
II
IV
I
II
III
26780
30890
26955
27440
29715
Level 2
2. Reading a Space-Delimited Raw Data File
a. Write a DATA step to create a new data set named work.qtrdonation, reading the spacedelimited raw data file named the following:
Windows
"&path\donation.dat"
UNIX
"&path/donation.dat"
z/OS (OS/390)
"&path..rawdata(donation)"

120265
120267
120269
120270
120271
. . .
15 15
20 20
20 10
20 20
25
15 15
20 20
5 .
20 20
8-35
8-36

Name
Type
Length
IDNum
Character
Qtr1
Numeric
Qtr2
Numeric
Qtr3
Numeric
Qtr4
Numeric
c. Write a PROC PRINT step to create the report below. The results contain 124 observations.
Obs
IDNum
1
2
3
4
5
120265
120267
120269
120270
120271
Qtr1
Qtr2
Qtr3
Qtr4
.
15
20
20
20
.
15
20
10
20
.
15
20
5
20
25
15
20
.
20
Challenge
3. Reading a Tab-Delimited Raw Data File
a. Create a temporary data set, managers2, using the tab-delimited raw data file named the
following:
Windows
"&path\managers2.dat"
UNIX
"&path/ managers2.dat"
z/OS (OS/390)
"&path..rawdata(managers2)"
Raw Data File

120102
120103
120261
121143
121144
121145
Tom
Wilson
Harry
Louis
Renee
Dennis
Zhou
M
Dawes
M
Highpoint
Favaron M
Capachietti
Lansberry
108255
87975
M
95090
F
M
Sales Manager
Sales Manager
243190 Chief Sales Officer
Senior Sales Manager
83505
Sales Manager
84260
Sales Manager

Name
Type
ID
Numeric
First
Character
Last
Character
8.3 Reading Nonstandard Delimited Data
Name
Type
Gender
Character
Salary
Numeric
Title
Character
c. The new data set should contain only First, Last, and Title.
d. Generate the report below. The results should contain six observations.
Obs
1
2
3
4
5
6
First
Tom
Wilson
Harry
Louis
Renee
Dennis
Last
Zhou
Dawes
Highpoint
Favaron
Capachietti
Lansberry
Title
Sales Manager
Sales Manager
Chief Sales Officer
Sales Manager
Sales Manager

Objectives
Use informats to read character data.

Use informats to read nonstandard data.
Subset observations and add permanent attributes.
77
77
8-37
8-38
Chapter 8 Read
ding Raw Data Files
Co
onsideratio
ons
Use
e modified lisst input to rea
ad all the fields from sale
es.csv.
Store the date fields
f
as SAS
S dates.
Partial sales.csv
120102,
,Tom,Zhou,M,10
08255,Sales Manager,AU,11AUG
G1973,06/01/19
993
120103,
,Wilson,Dawes,M,87975,Sales Manager,AU,22
2JAN1953,01/01
1/1978
120121,
,Irenie,Elvish
h,F,26600,Sales Rep. II,AU,0
02AUG1948,01/0
01/1978
120122,
,Christina,Nga
an,F,27475,Sales Rep. II,AU,27JUL1958,07/
/01/1982
120123,
,Kimiko,Hotsto
one,F,26190,Sales Rep. I,AU,28SEP1968,10/
/01/1989
79
ESERVED.
8..3 Reading No
onstandard Dellimited Data
ESERVED.
8-39
8-40
Reading Nonstandard Data

An informat is required to read nonstandard numeric
data.
Partial sales.csv
In this example, informats are needed to specify the style

of the date fields so that they can be read and converted
to SAS dates.
82
8.08 Quiz
A format is an instruction that tells SAS how to display
data values. What formats would you specify to display a
SAS date in the styles shown below?
a) 01JAN2000
b) 01/16/2000
83
8..3 Reading No
ESERVED.
8-41
8-42
SAS Informats
Selected SAS Informats for Nonstandard Numeric Values
Informat
COMMA.
DOLLAR.
Definition
Reads nonstandard numeric data and removes
embedded commas, blanks, dollar signs, percent signs,
and dashes.
COMMAX. Reads nonstandard numeric data and removes

embedded non-numeric characters; reverses the roles of
DOLLARX. the decimal point and the comma.
EUROX.
Reads nonstandard numeric data and removes

embedded non-numeric characters in European currency.
$CHAR.
Reads character values and preserves leading blanks.
$UPCASE. Reads character values and converts them to uppercase.

87
87
SAS Informats
Informats are used to read and convert raw data.
Informat
Raw Data Value
COMMA.
DOLLAR.
$12,345
12345
COMMAX.
DOLLARX.
$12.345
12345
EUROX.
12.345
12345
$CHAR.
##Australia
$UPCASE. au
SAS Data Value
##Australia
AU
The character # represents a blank space.
88
88
SAS Informats
Use date informats to read and convert dates to SAS date
values.
Informat
Raw Data Value
SAS Data Value
MMDDYY.
010160
01/01/60
01/01/1960
1/1/1960
DDMMYY.
311260
31/12/60
31/12/1960
365
DATE.
31DEC59
31DEC1959
-1
89
89
8.09 Quiz
Use the SAS Help facility or documentation to investigate
the DATEw. informat and answer the following questions:
a) What does the w represent?
b) What is the default width of this informat?
90
8-43
8-44
Chapter 8 Read
ding Raw Data Files
ESERVED.
8..3 Reading No
Vieewing the Log

37
38
39
40
41
42
da
ata work.sales;
d
ame :$12. Last
t_Name :$18.
input Employee_ID First_Na
J
:$25
5. Country :$2.
Gender :$1. Salary Job_Title
_Date :date. Hi
ire_Date :mmdd
dyy.;
Birth_
un;
ru
NOTE: T
The infile "s:\workshop\sale
es.csv" is:
F
Filename=s:\wo
rkshop\sales.c
csv,
R
RECFM=V,LRECL=
256,File Size (bytes)=11340
0,
NOTE: 1
165 records were read from the
t
infile "s:\workshop\sale
es.csv".
NOTE: The
T
data set WORK.SALES
W
has 165 observati
ions and 9 vari
iables.
94
p108d06
94
ESERVED.
8-45
8-46

A new data set should contain observations only for
Australian employees. Which of the following can be used
to subset the data?
a. where Country='AU';
b. if Country='AU';
c. either a or b
d. You cannot subset when reading from a raw data file.
96
Additional SAS Statements

Additional SAS statements can be added to perform
further processing in the DATA step.
data work.sales;
input Employee_ID First_Name :$12. Last_Name :$18.
Gender :$1. Salary Job_Title :$25. Country :$2.
Birth_Date :date. Hire_Date :mmddyy.;
if Country='AU';
keep First_Name Last_Name Salary
Job_Title Hire_Date;
format Salary dollar12. Hire_Date monyy7.;
run;
98
p108d07
98
Viewing the Output

proc print data=work.sales label;
run;
Obs
1
2
3
4
5
First_
Name
Last_Name
Sales Title
Tom
Wilson
Irenie
Christina
Kimiko
Zhou
Dawes
Elvish
Ngan
Hotstone
Sales
Sales
Sales
Sales
Sales
Manager
Manager
Rep. II
Rep. II
Rep. I
Salary
Date
Hired
$108,255
$87,975
$26,600
$27,475
$26,190
JUN1993
JAN1978
JAN1978
JUL1982
OCT1989
p108d07
99
99

Step and Usage
PROC step
WHERE
IF
Yes
No
SET statement
Yes
Yes
No
Yes
INPUT statement
No
Yes
100
100
8-47
8-48
Chapter 8 Read
ding Raw Data Files
Using
g List Input: Imporrtance off Colon Fo
ormat Mo
odifier
p108a02
2
1. Open p108a02 and
d examine thee INPUT stateement.
he INFILE staatement does not
n contain DLM=
D
becausse the file is sppace delimiteed.
Th
HiireDate and Salary
S
are non
nstandard num
meric fields, sso an informaat is needed.
2. In thee SAS window
wing environm
ment, select File
F Open Program, chhange the valuue for Files off
type to
t Data Files (*.dat), and select salary..dat.
Files with a .dat extensiion can be im

mported in SA S Enterprise G
Guide but not viewed in
an editor.
Partiaal salary.dat
Donny 5MAY2008 25 FL $43,132.50
Margaret 20FEB2008 43 NC 65,150
Dan 1J
JUN2008 27 FL $40,000.00
0
Subash 2FEB2008 45 NC 75,750
Antonio 25MAY2008 35 FL $43,5
500.50
3. Subm
mit part 1 and view
v
the log and output. The
T expected ooutput is show
wn below.
/*
/ Part 1 - using colon
c
for
rmat modif
fiers*/
data
a work.sal
laries;
infile "&p
path\salar
ry.dat";
input Name
e $ HireDa
ate :date. Age Sta
ate $ Sala
ary :comm
ma.;
run;
;
proc
c print da
ata=work.s
salaries;
run;
;
ESERVED.
8..3 Reading No
8-49
Partiaal PROC PRIN

NT Output
Obs
1
2
3
4
5
Name
Donny
Margaret
Dan
Subash
Antonio
Hire
Date
Age
A
17657
17582
17684
17564
17677
25
43
27
45
35
State
FL
NC
FL
NC
FL
Salary
43132.5
65150.0
40000.0
75750.0
43500.5
4. Now lets see whatt happens wheen a colon forrmat modifierr is omitted fr
from Salary. S
Submit part 22.
/*
* Part 2 omit the
e colon format
f
mod
difier for Salary */
data
a work.sal
laries;
infile
i
"&p
path\salar
ry.dat";
input
i
Name
e $ HireDa
ate :date
e. Age Sta
ate $ Salary comma
a.;
run;
;
proc
c print da
ata=work.s
salaries;
run;
;
5. Exam
mine the log. There
T
are no errors
e
or warn
nings.
923
924
925
926
927
/* Part 2 omit the colon

c
format modifier for
r Salary */
data work.salaries;
infile "&path\salary
y.dat";
input Name $ HireDate :date. Age
e State $ Sal
lary comma.;
run;
NOTE: The infile "s:\workshop

p\salary.dat" is:
Filename=s:\workshop\sa
alary.dat,
NOTE: 8 records were
w
read fro
om the infile
e "s:\worksho
op\salary.dat
t".
NOTE: The data set WORK.SALAR
RIES has 8 ob
bservations a
and 5 variabl
les.
6. Exam
mine the outpu
ut.
Obs
1
2
3
4
5
Name
Donny
Margaret
Dan
Subash
Antonio
Date
Age
A
17657
17582
17684
17564
17677
25
43
27
45
35
State
FL
NC
FL
NC
FL
Salary
.
6
.
7
.
The Salary
S
values are incorrectt. Why are thee values eitheer missing or oonly one digitt in length?
Thee comma. info
format has a default
d
width of
o 1, so SAS reads 1 colum
mn from the input file. Thiis
reads the first ch
haracter of thee Salary valu
ue.
hen the first column contains a dollar sig
gn, a missingg value is assiggned to the nuumeric variabble.
Wh
Thiis does not caause a data errror because th
he comma infformat removves non-numerric characterss,
including the do
ollar sign. Wh
hen the first co
olumn containns a digit, thaat digit becom
mes the value oof the
varriable.
c
omitting a colon resu
ults in missinng or invalid vvalues. In otheer cases, it results
In some cases,
in data errrors.
ESERVED.
8-50
Chapter 8 Read
ding Raw Data Files
ESERVED.
Viewing the Output

proc print data=work.newemps;
run;
PROC PRINT Output
Obs
First_
Name
Last_
Name
Job_
Title
1
2
3
Steven
Merle
Marta
Worton
Hieds
Bamberge
Auditor
Trainee
Manager
Salary
40450
24025
32000
p108d08
105
Exercises
Level 1
4. Reading Nonstandard Data from a Comma-Delimited Raw Data File
a. Open p108e04. Add the appropriate LENGTH, INFILE, and INPUT statements to read the
comma-delimited raw data file named the following:
Windows
"&path\custca.csv"
UNIX
"&path/custca.csv"
z/OS (OS/390)
"&path..rawdata(custca)"

Bill,Cuddy,11171,M,16/10/1986,21,15-30 years
Susan,Krasowski,17023,F,09/07/1959,48,46-60 years
Andreas,Rennie,26148,M,18/07/1934,73,61-75 years
Lauren,Krasowski,46966,F,24/10/1986,21,15-30 years
Lauren,Marx,54655,F,18/08/1969,38,31-45 years
8-51
8-52
Read the following fields:

Name
Type
Length
First
Character
20
Last
Character
20
ID
Numeric
Gender
Character
BirthDate
Numeric
Age
Numeric
AgeGroup
Character
12
b. Use FORMAT and DROP statements in the DATA step to create a data set that results in the
report below when displayed with a PROC PRINT step. Include an appropriate title. The results
Canadian Customers
Obs
1
2
3
4
5
First
Last
Bill
Susan
Andreas
Lauren
Lauren
Cuddy
Krasowski
Rennie
Krasowski
Marx
Gender
M
F
M
F
F
AgeGroup
15-30
46-60
61-75
15-30
31-45
years
years
years
years
years
Birth
Date
OCT1986
JUL1959
JUL1934
OCT1986
AUG1969
Level 2
5. Reading a Delimited Raw Data File with Nonstandard Data Values
a. Write a DATA step to create a temporary data set, prices, reading the delimited raw data file
named the following:
Windows
"&path\pricing.dat"
UNIX
"&path/ pricing.dat"
z/OS (OS/390)
"&path..rawdata(pricing)"
All data fields are numeric.

210200100009*09JUN2011*31DEC9999*$15.50*$34.70
210200100017*24JAN2011*31DEC9999*$17.80*22.80
210200200023*04JUL2011*31DEC9999*$8.25*$19.80
210200600067*27OCT2011*31DEC9999*$28.90*47.00
210200600085*28AUG2011*31DEC9999*$17.85*$39.40
b. Generate the report below. The results should contain 16 observations.
8-53

2011 Pricing
Obs
ProductID
StartDate
EndDate
Cost
Sales
Price
1
2
3
4
5
210200100009
210200100017
210200200023
210200600067
210200600085
06/09/2011
01/24/2011
07/04/2011
10/27/2011
08/28/2011
12/31/9999
12/31/9999
12/31/9999
12/31/9999
12/31/9999
15.50
17.80
8.25
28.90
17.85
34.70
22.80
19.80
47.00
39.40
Challenge
6. Reading In-Stream Delimited Data
a. Open p108e06. Write a DATA step to read the delimited in-stream data shown below.
An INFILE statement is required. Use SAS Help or online documentation to explore the
use of DATALINES as a file specification in an INFILE statement.
120102/Tom/Zhou/M/108,255/Sales Manager/01Jun1993
120103/Wilson/Dawes/M/87,975/Sales Manager/01Jan1978
120261/Harry/Highpoint/M/243,190/Chief Sales Officer/01Aug1991
121143/Louis/Favaron/M/95,090/Senior Sales Manager/01Jul2001
121144/Renee/Capachietti/F/83,505/Sales Manager/01Nov1995
121145/Dennis/Lansberry/M/84,260/Sales Manager/01Apr1980
b. Generate the report below:

Orion Star Management Team
First
Last
Title
Tom
Wilson
Harry
Louis
Renee
Dennis
Zhou
Dawes
Highpoint
Favaron
Capachietti
Lansberry
Sales Manager
Sales Manager
Chief Sales Officer
Sales Manager
Sales Manager
ID
Gender
Salary
HireDate
120102
120103
120261
121143
121144
121145
M
M
M
M
F
M
108255
87975
243190
95090
83505
84260
06/01/1993
01/01/1978
08/01/1991
07/01/2001
11/01/1995
04/01/1980
8-54
Chapter 8 Read
ding Raw Data Files
8.4 Handling Missin

M
ng Datta
Ob
bjectives
Use the DSD

D option to re
ead consecu
utive delimite
ers as
missing valu
ues.
Use the MIS
SSOVER opttion to recognize missing
g
v
values
at the
e end of a re
ecord.
109
109
ESERVED.
8.4 Handling Missing Data
Missing Values in the Middle of the Record

The records in phone2.csv have a contact name, phone
number, and a mobile number. The phone number is
missing from some of the records.
phone2.csv
Missing data is indicated by

consecutive delimiters.
1
1
2
2
3
3
4
4
1---5----0----5----0----5----0----5----0----5
James Kvarniq,(704) 293-8126,(701) 281-8923
Sandrina Stephano,, (919) 271-4592
Cornelia Krahl,(212) 891-3241,(212) 233-5413
Karen Ballinger,, (714) 644-9090
Elke Wallstab,(910) 763-5561,(910) 545-3421
111
111
8.11 Quiz
Open and submit p108a03.

Examine the SAS log. How many input records were
read and how many observations were created?
Examine the report. Does it look correct?
data work.contacts;
length Name $ 20 Phone Mobile $ 14;
infile "&path\phone2.csv" dlm=',';
input Name $ Phone $ Mobile $;
run;
proc print data=work.contacts noobs;
run;
112
p108a03
112
8-55
8-56
Chapter 8 Read
ding Raw Data Files
Co
onsecutivee Delimiteers in Listt Input
Listt input treats two or more
e consecutive
e delimiters as a
sing
gle delimiter and not as a missing value.
pho
one2.csv
1
1
2
2
3
3
4
4
1---5----0----5----0----5----0----5----0----5
5
Jam
mes Kvarniq,(704) 293-8126,(701) 281-8923
San
ndrina Stephano,, (919
9) 271-4592
Cor
rnelia Krahl,(212) 891
1-3241,(212)
) 233-5413
Kar
ren Ballinger,, (714) 644-9090
Elk
ke Wallstab,(910) 763-5561,(910) 545-3421
114
When there is missing

m
data in a record, SAS does th
he
follo
owing:
loads the ne
ext record to finish the ob
bservation
w
a note
e to the log
writes
114
ESERVED.
8
Vieewing the Output

Add
ding the DSD
D option give
es the correcct results.
PRO
OC PRINT Output
O
Name
Ph
hone
James Kvarniq
hano
Sandrina Steph
Cornelia Krahl
l
Karen Ballinge
er
Elke Wallstab
(704) 293-8126
(212) 891-3241
(910) 763-5561
Mobile
(701)
919)
(9
(2
212)
(714)
910)
(9
281-8923
271-4592
233-5413
644-9090
545-3421
Parrtial SAS Log

g
NOTE
E: 5 records were
w
read from the infile "S
S:\workshop\phone2.csv".
The minimum
m record length was 31.
The maximum
m record length was 44.
NOTE
E: The data se
et WORK.CONTACT
TS has 5 obser
rvations and 3
vari
iables.
117
117
ESERVED.
8-57
8-58
Chapter 8 Read
ding Raw Data Files
Misssing Values at thee End of a Record

The
e raw data file phone.csv
v contains missing
m
value
es at
the end of some
e records.
pho
one.csv
1
1
2
2missing
3 valu
3
4
4
ues
1---5----0----5----0----5----0----5----0----5
5
Jam
mes Kvarniq,(704) 293-8126,(701) 281-8923
San
ndrina Stephano,(919) 871-7830
Cor
rnelia Krahl,(212) 891
1-3241,(212)
) 233-5413
Kar
ren Ballinger,(714) 34
44-4321
Elk
ke Wallstab,(910) 763-5561,(910) 545-3421
The
e DSD option
n is not appro
opriate beca
ause the misssing
data
a is not markked by conse
ecutive delim
miters.
120
120
ESERVED.
8
Vieewing the Output

Parrtial SAS Log
g
NOTE
E: 5 records were read from the infile "S
S:\workshop\pho
one.csv".
h was 31.
The minimum record length
h was 44.
The maximum record length
NOTE
E: The data set WORK.CONTACT
TS has 5 obser
rvations and 3
vari
iables.
PRO
OC PRINT Output
O
Name
James Kvarniq
ano
Sandrina Stepha
Cornelia Krahl
r
Karen Ballinger
Elke Wallstab
Pho
one
(704)
(919)
(212)
(714)
(910)
293-8126
2
8
871-7830
8
891-3241
3
344-4321
7
763-5561
Mobile
(70
01) 281-8923
(21
12) 233-5413
(91
10) 545-3421
122
122
ESERVED.
8-59
8-60
Chapter 8 Read
ding Raw Data Files
Exerc
cises
If you restarted your SA
AS session siince the last exercise,
e
openn and submit tthe libname.ssas program
found in the
t data folder.
Level 1
7. Readiing a Comma
a-Delimited File
F with Misssing Values
a. Open p108e07 and insert IN
NFILE and IN
NPUT statemeents to read thhe comma-dellimited raw ddata
naamed the following:
Windowss
"&path\dona
"
ation.csv"
UNIX
"&path/dona
"
ation.csv"
z/OS (OS/39
90)
"&p
path..rawdatta(donation)"
"
Partial Raw Daata File

12
20265,,,,25
12
20267,15,15,15,15
12
20269,20,20,2
20,20
12
20270,20,10,5
5
12
20271,20,20,2
20,20
b. Th
here might bee missing dataa in the middlle or at the ennd of a recordd. Read the following fields
frrom the raw data
d file:
ESERVED.
Name
Type
Employee ID
Numeric
Quarter 1
Numeric
Quarter 2
Numeric
Quarter 3
Numeric
Quarter 4
Numeric
8-61
c. Write a PROC PRINT step to generate the report below. The results should include 124
observations.
Obs
EmpID
Q1
Q2
Q3
Q4
1
2
3
4
5
120265
120267
120269
120270
120271
.
15
20
20
20
.
15
20
10
20
.
15
20
5
20
25
15
20
.
20
Level 2
8. Reading a Delimited File with Missing Values
a. Write a DATA step to create a temporary data set, prices, using the asterisk-delimited raw data
file named the following:
Windows
"&path\prices.dat"
UNIX
"&path/prices.dat"
z/OS (OS/390)
"&path..rawdata(prices)"

210200100009*09JUN2007*31DEC9999*$15.50*$34.70
210200100017*24JAN2007*31DEC9999*$17.80
210200200023*04JUL2007*31DEC9999*$8.25*$19.80
210200600067*27OCT2007*31DEC9999*$28.90
210200600085*28AUG2007*31DEC9999*$17.85*$39.40
There might be missing data at the end of some records. Read the following fields from the raw
data file:
Name
Type
Length
ProductID
Numeric
StartDate
Numeric
EndDate
Numeric
UnitCostPrice
Numeric
8-62
Name
Type
Length
UnitSalesPrice
Numeric
b. Define labels and formats in the DATA step to create a data set that generates the following output
when used in the PROC PRINT step. The results should contain 259 observations.
2007 Prices
Obs
1
2
3
4
5
Product ID
Start of
Date Range
End of
Date Range
210200100009
210200100017
210200200023
210200600067
210200600085
06/09/2007
01/24/2007
07/04/2007
10/27/2007
08/28/2007
12/31/9999
12/31/9999
12/31/9999
12/31/9999
12/31/9999
Cost Price
per Unit
15.50
17.80
8.25
28.90
17.85
Sales
Price per
Unit
34.70
.
19.80
.
39.40
Challenge
9. Reading a Delimited File with Missing Values and Embedded Delimiters
a. Write a DATA step to create a temporary data set, salesmgmt, using the raw data file named the
following:
Windows
"&path\managers.dat"
UNIX
"&path/managers.dat"
z/OS (OS/390)
"&path..rawdata(managers)"

120102/Tom/Zhou/M//Sales Manager/AU/11AUG1969/'06/01/1989'
120103/Wilson/Dawes/M/87975/Sales Manager/AU/22JAN1949/'01/01/1974'
120261/Harry/Highpoint/M/243190//US/21FEB1969/'08/01/1987'
121143/Louis/Favaron/M/95090/Senior Sales Manager/US/26NOV1969/'07/01/1997'
121144/Renee/Capachietti/F/83505/Sales Manager/US/28JUN1964
121145/Dennis/Lansberry/M/84260/Sales Manager/US/22NOV1949/'04/01/1976'
b. ID is a numeric value. The salesmgmt data set should contain only the variables shown in the
report below.
c. Write a PROC PRINT step to generate the report below. The results should contain six
observations.
PROC PRINT Output
Orion Star Managers
Obs
ID
1
2
3
4
120102
120103
120261
121143
Last
Title
Zhou
Dawes
Highpoint
Favaron
Sales Manager
Sales Manager
HireDate
Salary
01JUN1989
01JAN1974
01AUG1987
01JUL1997
.
87975
243190
95090
8.5 Solutions
5
6
121144
121145
Capachietti
Lansberry
Sales Manager
Sales Manager
.
01APR1976
8.5 Solutions
1. Reading a Comma-Delimited Raw Data File
data work.newemployees;
length First $ 12 Last $ 18 Title $ 25;
input First $ Last $ Title $ Salary;
run;
proc print data=work.newemployees;
run;
2. Reading a Space-Delimited Raw Data File
data work.qtrdonation;
length IDNum $ 6;
input IDNum $ Qtr1 Qtr2 Qtr3 Qtr4;
run;
proc print data=work.qtrdonation;
run;
3. Reading a Tab-Delimited Raw Data File
data work.managers2;
length First Last $ 12 Title $ 25;
infile "&path\managers2.dat" dlm='09'x;
input ID First $ Last $ Gender $ Salary Title $;
keep First Last Title;
proc print data=work.managers2;
run;
4. Reading Nonstandard Data from a Comma-Delimited Raw Data File
data work.canada_customers;
length First Last $ 20 Gender $ 1 AgeGroup $ 12;
infile "&path\custca.csv" dlm=',';
input First $ Last $ ID Gender $
BirthDate :ddmmyy. Age AgeGroup $;
format BirthDate monyy7.;
drop ID Age;
run;
title 'Canadian Customers';
proc print data=work.canada_customers;
run;
title;
83505
84260
8-63
8-64
5.
Reading a Delimited Raw Data File with Nonstandard Values

data work.prices;
infile "&path\pricing.dat" dlm='*';
input ProductID StartDate :date. EndDate :date.
Cost :dollar. SalesPrice :dollar.;
format StartDate EndDate mmddyy10.
Cost SalesPrice 8.2;
run;
title '2011 Pricing';
proc print data=work.prices;
run;
title;
6. Reading In-Stream Delimited Data

data work.managers;
infile datalines dlm='/';
input ID First :$12. Last :$12. Gender $ Salary :comma.
Title :$25. HireDate :date.;
datalines;
120102/Tom/Zhou/M/108,255/Sales Manager/01Jun1993
120103/Wilson/Dawes/M/87,975/Sales Manager/01Jan1978
120261/Harry/Highpoint/M/243,190/Chief Sales Officer/01Aug1991
121143/Louis/Favaron/M/95,090/Senior Sales Manager/01Jul2001
121144/Renee/Capachietti/F/83,505/Sales Manager/01Nov1995
121145/Dennis/Lansberry/M/84,260/Sales Manager/01Apr1980
;
title 'Orion Star Management Team';
proc print data=work.managers noobs;
format HireDate mmddyy10.;
run;
title;
7. Reading a Comma-Delimited File with Missing Values
infile "&path\donation.csv" dsd missover;
input EmpID Q1 Q2 Q3 Q4;
run;
proc print data=work.donations;
run;
8. Reading a Delimited File with Missing Values
data work.prices;
infile "&path\prices.dat" dlm='*' missover;
input ProductID StartDate :date. EndDate :date.
UnitCostPrice :dollar. UnitSalesPrice :dollar.;
label ProductID='Product ID'
StartDate='Start of Date Range'
EndDate='End of Date Range'
8.5 Solutions
UnitCostPrice='Cost Price per Unit'

UnitSalesPrice='Sales Price per Unit';
format StartDate EndDate mmddyy10.
UnitCostPrice UnitSalesPrice 8.2;
run;
title '2007 Prices';
proc print data=work.prices label;
run;
title;
9. Reading a Delimited File with Missing Values and Embedded Delimiters
data work.salesmgmt;
length First Last $ 12 Gender $ 1 Title $ 25 Country $ 2;
format BirthDate HireDate date9.;
infile "&path\managers.dat" dsd dlm='/' missover;
input ID First Last Gender Salary Title Country
BirthDate :date. HireDate :mmddyy.;
run;
title 'Orion Star Managers';
proc print data=work.salesmgmt;
var ID Last Title HireDate Salary ;
run;
title;

Which fields in this file can be read as standard numeric
values?
The employee ID and salary. The date fields are
nonstandard and require special processing.
Partial sales.csv
16
8-65
8-66

Which statement is true?
a. An input buffer is created only if you are reading data
from a raw data file.
b. The PDV at compile time holds the variable name,
type, byte size, and initial value.
c. The descriptor portion is the first item that is created
at compile time.
31
31

Which statement is true of a DATA step when reading
from a raw data file?
a. Data is read from the raw data file into the PDV.
b. The size of the input buffer adjusts automatically
based on the length of the input record.
c. At the bottom of the DATA step, the contents of the
PDV are output to the output SAS data set.
51
51
8.5 Solutions

Suppose you want the order of the variables to match the
order of the fields. You can include the numeric variables
in the LENGTH statement. Which of the following
produces the correct results?
a. length Employee_ID First_Name $ 12
Salary Job_Title $ 25
Country $ 2;
b. length Employee_ID 8 First_Name $ 12
Country $ 2;
58

What problems do you see with the data values for the
last two data fields, Salary and Country?
Partial sales3inv.csv
120102,Tom,Zhou,Manager,108255,AU
120103,Wilson,Dawes,Manager,87975,AU
120121,Irenie,Elvish,Rep. II,26600,AU
120122,Christina,Ngan,Rep. II,n/a,AU
120123,Kimiko,Hotstone,Rep. I,26190,AU
120124,Lucian,Daymond,Rep. I,26480,12
120125,Fong,Hofmeister,Rep. IV,32040,AU
65
65
8-67
8-68
Chapter 8 Read
ding Raw Data Files
8.008 Quiz Correct

C
A
Answer
A fo
ormat is an instruction
i
th
hat tells SAS
S how to disp
play
data
a values. Wh
hat formats would
w
you sp
pecify to disp
play a
SAS
S date in the
e styles show
wn below?
a) 01JAN2000 DATE9
9.
b) 01/16/2000
MMDDY
YY10.
84
ESERVED.
8.5 Solutions

Use the SAS Help facility or documentation to investigate
the DATEw. informat and answer the following questions:
a) What does the w represent?
the width of the input field
b) What is the default width of this informat?
The default width is 7.
91

A new data set should contain observations only for
Australian employees. Which of the following can be used
to subset the data?
a. where Country='AU';
b. if Country='AU';
c. either a or b
d. You cannot subset when reading from a raw data file.
97
8-69
8-70
How many input records were read and how many

observations were created? five read, three created
Examine the report. Does it look correct? no
NOTE: 5 records were read from the infile "S:\workshop\phone2.csv".

The minimum record length was 31.
The maximum record length was 44.
NOTE: SAS went to a new line when INPUT statement reached past the
end of a line.
NOTE: The data set WORK.CONTACTS has 3 observations and 3 variables.
Name
Phone
Mobile
James Kvarniq
Sandrina Stephano
Karen Ballinger
(704) 293-8126
(919) 871-7830
(714) 344-4321
(701) 281-8923
Cornelia Krahl
Elke Wallstab
113
113
Chapter 9 Manipulating Data

9.1
Using SAS Functions ..................................................................................................... 9-3

Exercises .............................................................................................................................. 9-10
9.2
Conditional Processing ............................................................................................... 9-12

Exercises .............................................................................................................................. 9-37
9.3
Solutions ....................................................................................................................... 9-40

9-2
9.1 Using SAS Functions

Objectives
Create data values using SAS functions.
3
3
Business Scenario
Orion Star management plans to give a $500 bonus to
each employee in his or her hire month.
500
4
4
9-3
9-4
Considerations
Create a new data set with three new variables:
Bonus, which is a constant 500
Compensation, which is the sum of Salary and
Bonus
BonusMonth, which is the month in which the
employee was hired
Compensation
Bonus
BonusMonth
work.comp
orion.sales
5
5
Considerations
Partial orion.sales
Employee
First
_ID
_Name
120102 Tom
Birth_
Date
3510
Hire_
Date
10744
AU
-3996
5114
AU
-5630
5114
Last
_ Name
Zhou
Gender
Salary
108255 Sales Manager
AU
120103 Wilson
Dawes
87975 Sales Manager
120121 Irenie
Elvish
26600 Sales Rep. II
Job_ Title
Country
Partial work.comp
Employee
_ID
First
_Name
Last
_ Name
Bonus
Compensation
Bonus
Month
120102 Tom
Zhou
500
108755
120103 Wilson
Dawes
500
88475
120121 Irenie
Elvish
500
27100
Drop Gender, Salary, Job_Title, Country, Birth_Date, and

Hire_Date from work.comp.
6
6

Which of the following statements creates a numeric
variable Bonus with a value of 500?
a. Bonus=$500;
b. Bonus=500;
c. label Bonus='500';
d. format Bonus 500.;
Partial work.comp
Bonus
Month
Bonus
Compensation
500
108755
500
88475
500
27100
SAS Functions
SAS functions can be used in an assignment statement. A
function is a routine that accepts arguments and returns a
value.
variable=function-name(argument1, argument2, );
Some functions manipulate character values, compute
descriptive statistics, or manipulate SAS date values.
Arguments are enclosed in parentheses and separated
by commas.
A function can return a numeric or character result.
9
9
9-5
9-6
SUM Function
Use the SUM function to create Compensation. The
SUM function is a descriptive statistics function that
returns the sum of its arguments.
Compensation=sum(Salary,Bonus);
SUM(argument1,argument2, ...)
The arguments must be numeric.

Missing values are ignored by SUM and other
descriptive statistics functions.
10
10
MONTH Function
Use the MONTH function to extract the month of hire from
Hire_Date.
BonusMonth=month(Hire_Date);
MONTH(SAS-date)
Other date functions can do the following:
extract information from SAS date values
create SAS date values
11
11
Date Functions: Extracting Values

Syntax
Description
YEAR(SAS-date)
Extracts the year from a SAS date and

returns a four-digit year.
QTR(SAS-date)
Extracts the calendar quarter from a SAS

date and returns a number from 1 to 4.
MONTH(SAS-date)
Extracts the month from a SAS date and

returns a number from 1 to 12.
DAY(SAS-date)
Extracts the day of the month from a SAS

date and returns a number from 1 to 31.
Extracts the day of the week from a SAS

WEEKDAY(SAS-date) date and returns a number from 1 to 7,
where 1 represents Sunday.
12
12
Date Functions: Creating SAS Dates

Syntax
Description
TODAY()
DATE()
Returns the current date as a SAS date

value.
MDY(month,day,year)
Returns a SAS date value from numeric

month, day, and year values.
Examples
CurrentDate=today();
y2k=mdy(01,1,2000);
NewYear=mdy(Mon,Day,2013);
13
13
9-7
9-8
Using SAS Functions

A function call can be used alone in an assignment
statement.
AnnivBonus=mdy(BonusMonth,15,2008);
A function call can be part of any SAS expression.
if month(Hire_Date)=12;
A function call can be an argument to another function.
AnnivBonus=mdy(month(Hire_Date),15,2012);
14
14
Using SAS Functions

Create Bonus, Compensation, and BonusMonth.
data work.comp;
set orion.sales;
Bonus=500;
run;
175
176
177
178
179
180
data work.comp;
set orion.sales;
Bonus=500;
run;
orion.sales has
nine variables.
NOTE: The data set WORK.COMP has 165 observations and 12 variables.
15
p109d01
15
Viewing the Output

proc print data=work.comp noobs;
var Employee_ID First_Name Last_Name
Bonus Compensation BonusMonth;
run;
Employee_ID
120102
120103
120121
120122
120123
First_
Name
Last_Name
Tom
Wilson
Irenie
Christina
Kimiko
Zhou
Dawes
Elvish
Ngan
Hotstone
Bonus
Compensation
Bonus
Month
500
500
500
500
500
108755
88475
27100
27975
26690
6
1
1
7
10
p109d01
16
16
9.02 Quiz
A DROP statement has been added to this DATA step.
Will the program calculate Compensation and
BonusMonth correctly?
data work.comp;
set orion.sales;
drop Gender Salary Job_Title Country
Birth_Date Hire_Date;
Bonus=500;
run;
17
p109a01
17
9-9
9-10
Viewing the Output

proc print data=work.comp noobs;
run;
First_
Employee_ID Name
120102
120103
120121
120122
120123
Tom
Wilson
Irenie
Christina
Kimiko
Last_Name
Zhou
Dawes
Elvish
Ngan
Hotstone
Bonus
500
500
500
500
500
Compensation
108755
88475
27100
27975
26690
Bonus
Month
6
1
1
7
10
p109a01
19
19
Exercises
Level 1
1. Creating New Variables
b. In the DATA step, create three new variables:
Increase, which is Salary multiplied by 0.10
NewSalary, which is Salary added to Increase
BdayQtr, which is the quarter in which the employee was born
c. The new data set should include only Employee_ID, Salary, Birth_Date, and the three new
variables.
d. Store permanent formats to display Salary, Increase, and NewSalary with commas.
e. Modify the program to create the report below, including labels. The results should contain 424
observations.
9-11

Obs
Employee ID
Employee
Annual
Salary
1
2
3
4
5
120101
120102
120103
120104
120105
163,040
108,255
87,975
46,230
27,110
Employee
Birth Date
Increase
NewSalary
16,304
10,826
8,798
4,623
2,711
179,344
119,081
96,773
50,853
29,821
18AUG1980
11AUG1973
22JAN1953
11MAY1958
21DEC1978
Bday
Qtr
3
3
1
2
4
Level 2
a. Write a DATA step that reads orion.customer to create work.birthday.
b. In the DATA step, create three new variables: Bday2012, BdayDOW2012, and Age2012.
Bday2012 is the combination of the month of Birth_Date, the day of Birth_Date, and the
constant of 2012 in the MDY function.
BdayDOW2012 is the day of the week of Bday2012.
Age2012 is the age of the customer in 2012. Subtract Birth_Date from Bday2012 and divide
the result by 365.25.
c. Include only the following variables in the new data set: Customer_Name, Birth_Date,
Bday2012, BdayDOW2012, and Age2012.
d. Format Bday2012 to display in the form 01Jan2012. Age2012 should be formatted to display
with no decimal places.
e. Write a PROC PRINT step to create the report below. The results should contain 77 observations.
Obs
1
2
3
4
5
Customer_Name
James Kvarniq
Sandrina Stephano
Cornelia Krahl
Karen Ballinger
Elke Wallstab
Birth_
Date
Bday2012
27JUN1978
09JUL1983
27FEB1978
18OCT1988
16AUG1978
27JUN2012
09JUL2012
27FEB2012
18OCT2012
16AUG2012
Bday
DOW2012
4
2
2
5
5
Age2012
34
29
34
24
34
Challenge
3. Using the CATX and INTCK Functions to Create Variables
a. Write a DATA step that reads orion.sales to create work.employees.
In the DATA step, create a new variable, FullName, which is the combination of First_Name, a
space, and Last_Name. Use the CATX function. Documentation on CATX can be found in the
SAS Help facility or in the online documentation.
9-12
In the DATA step, create a new variable, Yrs2012, which is the number of years between
January 1, 2012, and Hire_Date. Use the INTCK function. Documentation on INTCK
can be found in the SAS Help facility or in the online documentation.
b. Format Hire_Date to display in the form 01/31/2012.
c. Give Yrs2012 a label of Years of Employment in 2012.
d. Create the report shown below. The results should contain 165 observations.
Obs
1
2
3
4
5
FullName
Tom Zhou
Wilson Dawes
Irenie Elvish
Christina Ngan
Kimiko Hotstone
Hire_Date
Years of
Employment
in 2012
01/06/1993
01/01/1978
01/01/1978
01/07/1982
01/10/1989
19
34
34
30
23
9.2 Conditional Processing

Objectives
Process data conditionally using IF-THEN/ELSE

statements.
Execute multiple statements conditionally using
DO and END statements.
Control the length of character variables using
the LENGTH statement.
23
23
Business Scenario
Orion Star management plans to give each sales
employee a bonus based on his or her job title.
Bonus
24
24
Considerations
Create a new data set, work.comp, using orion.sales as
input. Include a new variable, Bonus, with a value based
on Job_Title.
Job_Title
Bonus
Sales Rep. IV
1000
Sales Manager
1500
2000
Chief Sales Officer
2500
25
25
9-13
9-14
IF-THEN Statements
The IF-THEN statement executes a SAS statement for
observations that meet a specific condition.
data work.comp;
set orion.sales;
if Job_Title='Sales Rep. IV' then
Bonus=1000;
...
run;
IF expression THEN statement;
expression defines a condition.

statement can be any executable SAS statement.
If expression is true, then statement executes.
26
26
Conditional Processing
The value assigned to Bonus is determined by testing for
various values of Job_Title.
Sales Rep. IV
Yes
Bonus=1000
Yes
Bonus=1500
No
Sales Manager
No
Yes
Bonus=2000
No
Chief Sales Manager

27
Yes
Bonus=2500
No
27
PDV
Employee_ID Last_Name
120102 Zhou
Job_Title
Sales Manager
Bonus
.
p109d02
...
28
28
false
PDV
120102 Zhou
Job_Title
Sales Manager
Bonus
29
29
.
...
9-15
9-16
true
PDV
120102 Zhou
Job_Title
Sales Manager
Bonus
.
...
30
30
PDV
120102 Zhou
Job_Title
Sales Manager
Bonus
1500
31
31
...
false
PDV
120102 Zhou
Job_Title
Sales Manager
Bonus
1500
...
32
32
false
PDV
120102 Zhou
Job_Title
Sales Manager
Bonus
1500
33
33
...
9-17
9-18
Implicit OUTPUT;
Implicit RETURN;
PDV
120102 Zhou
Job_Title
Bonus
1500
Job_Title
Bonus
1500
Sales Manager
34
34
Continue until EOF
PDV
120102 Zhou
Sales Manager
35
35
Viewing the Output

proc print data=work.comp;
var Last_Name Job_Title Bonus;
run;
Obs
Last_Name
Job_Title
1
2
3
4
5
6
7
8
9
10
11
12
Zhou
Dawes
Elvish
Ngan
Hotstone
Daymond
Hofmeister
Denny
Clarkson
Kletschkus
Roebuck
Lyon
Sales
Sales
Sales
Sales
Sales
Sales
Sales
Sales
Sales
Sales
Sales
Sales
Manager
Manager
Rep. II
Rep. II
Rep. I
Rep. I
Rep. IV
Rep. II
Rep. II
Rep. IV
Rep. III
Rep. I
Bonus
1500
1500
.
.
.
.
1000
.
.
1000
.
.
p109d02
36
36

In the previous program, is it possible for more than one
condition to be true for a single observation?
a. Yes, more than one condition can be true.
b. No, the conditions are mutually exclusive, so only one
condition can be true.
37
9-19
9-20
Using the ELSE Statement

Use the ELSE statement when testing mutually exclusive
conditions.
.
data
work.comp;
set orion.sales;
Bonus=1000;
else if Job_Title='Sales Manager' then
Bonus=1500;
else if Job_Title='Senior Sales Manager'
then Bonus=2000;
else if Job_Title='Chief Sales Officer'
then Bonus=2500;
run;
<ELSE IF expression THEN statement;>
p109d03
39
39
When an expression is true, the associated statement is
executed and subsequent ELSE statements are skipped.
Sales Rep. IV
Yes
Bonus=1000
Yes
Bonus=1500
No
Sales Manager
No
Yes
Bonus=2000
No
Chief Sales Manager
Yes
Bonus=2500
No
40
40
IF-THEN Statements
data work.comp;
set orion.sales;
Bonus=1000;
Bonus=1500;
then Bonus=2000;
then Bonus=2500;
run;
PDV
120102 Zhou
Job_Title
Sales Manager
Bonus
.
p109d03
...
41
41
IF-THEN Statements
data work.comp;
false
set orion.sales;
Bonus=1000;
Bonus=1500;
then Bonus=2000;
then Bonus=2500;
run;
PDV
120102 Zhou
Job_Title
Sales Manager
Bonus
42
42
.
...
9-21
9-22
IF-THEN Statements
data work.comp;
set orion.sales;
true
Bonus=1000;
Bonus=1500;
then Bonus=2000;
then Bonus=2500;
run;
PDV
120102 Zhou
Job_Title
Sales Manager
Bonus
.
...
43
43
IF-THEN Statements
data work.comp;
set orion.sales;
Bonus=1000;
Bonus=1500;
then Bonus=2000;
then Bonus=2500;
run;
PDV
120102 Zhou
Job_Title
Sales Manager
Bonus
1500
44
44
...
IF-THEN Statements
data work.comp;
set orion.sales;
Bonus=1000;
Bonus=1500;
then Bonus=2000;
then Bonus=2500;
run;
Implicit OUTPUT;
Implicit RETURN;
PDV
120102 Zhou
Job_Title
Sales Manager
Bonus
1500
45
45
IF-THEN Statements
data work.comp;
Continue
until EOF
set
orion.sales;
Bonus=1000;
Bonus=1500;
then Bonus=2000;
then Bonus=2500;
run;
PDV
120102 Zhou
Job_Title
Sales Manager
Bonus
1500
46
46
9-23
9-24
Viewing the Output

run;
Obs
Last_Name
1
2
3
4
5
6
7
8
9
10
Zhou
Dawes
Elvish
Ngan
Hotstone
Daymond
Hofmeister
Denny
Clarkson
Kletschkus
Job_Title
Sales
Sales
Sales
Sales
Sales
Sales
Sales
Sales
Sales
Sales
Bonus
Manager
Manager
Rep. II
Rep. II
Rep. I
Rep. I
Rep. IV
Rep. II
Rep. II
Rep. IV
1500
1500
.
.
.
.
1000
.
.
1000
p109d03
47
47

Orion Star management wants to modify the bonus plan
as defined below.
Job_Title
Bonus
Sales Rep. III
1000
Sales Rep. IV
1000
Sales Manager
1500
2000
Chief Sales Officer
2500
All other titles
500
48
48
Using Conditional Processing

data work.comp;
compound
set orion.sales;
condition
if Job_Title='Sales Rep. III' or
Job_Title='Sales Rep. IV' then
Bonus=1000;
Bonus=1500;
then Bonus=2000;
then Bonus=2500;
else Bonus=500;
run;
<>
<ELSE statement;>
p109d04
49
49
An optional final ELSE statement gives an alternative
action if none of the conditions are true.
Sales Rep. III or IV
Yes
Bonus=1000
No
Sales Manager
Yes
Bonus=1500
No
Yes
Bonus=2000
No
Chief Sales Manager
Yes
Bonus=2500
No
Bonus=500
50
50
9-25
9-26
Viewing the Output

run;
Obs
Last_Name
1
2
3
4
5
6
7
8
9
10
Zhou
Dawes
Elvish
Ngan
Hotstone
Daymond
Hofmeister
Denny
Clarkson
Kletschkus
Job_Title
Sales
Sales
Sales
Sales
Sales
Sales
Sales
Sales
Sales
Sales
Manager
Manager
Rep. II
Rep. II
Rep. I
Rep. I
Rep. IV
Rep. II
Rep. II
Rep. IV
Bonus
1500
1500
500
500
500
500
1000
500
500
1000
p109d04
51
51
Business Scenario
Orion Star managers are considering a country-based
bonus. Create a new SAS data set named work.bonus
using orion.sales as input. The value of the new variable,
Bonus, is based on Country.
$500
$300
52
52
9-27
IF-THEN/ELSE Statements
If orion.sales has been validated and only includes the
Country values US and AU, the conditional clause can be
omitted from the ELSE statement.
data work.bonus;
set orion.sales;
if Country='US' then Bonus=500;
else Bonus=300;
run;

ELSE statement;
All observations not equal to US get a bonus of 300.
p109d05
53
53
This technique should be used only when you know that the final ELSE statement must be executed for
all other observations.
Viewing the Output

proc print data=work.bonus;
var First_Name Last_Name Country Bonus;
run;
54
Obs
First_Name
Last_Name
60
61
62
63
64
65
66
67
68
69
Billy
Matsuoka
Vino
Meera
Harry
Julienne
Scott
Cherda
Priscilla
Robert
Plested
Wills
George
Body
Highpoint
Magolan
Desanctis
Ridley
Farren
Stevens
Country
AU
AU
AU
AU
US
US
US
US
US
US
Bonus
300
300
300
300
500
500
500
500
500
500
p109d05
54
9-28
9.04 Quiz
Program p109a02 reads orion.nonsales, a non-validated
data set. Open and submit the program and review the
results. Why is Bonus set to 300 in observations 125,
197, and 200?
data work.bonus;
set orion.nonsales;
else Bonus=300;
run;
55
55
Testing for Invalid Data

You can test for multiple values of Country.
data work.bonus;
set orion.nonsales;
if Country in ('US','us')
then Bonus=500;
else Bonus=300;
run;
You can use the UPCASE function in the expression.
data work.bonus;
set orion.nonsales;
if upcase(Country)='US'
then Bonus=500;
else Bonus=300;
run;
57
p109a02s
57
Cleaning Invalid Data

You can clean the data before checking the value.
data work.bonus;
set orion.nonsales;
Country=upcase(Country);
if Country='US'
then Bonus=500;
else Bonus=300;
run;
It is a best practice to clean the data at the source,
but in some cases, that is not possible. With this
method, you are creating a clean data set.
p109d06
58
58
Business Scenario
Orion Star employees will receive a bonus once or twice a
year. In addition to Bonus, add a new variable, Freq, that
is equal to the following:
Once a Year for United States employees
Twice a Year for Australian employees
$500
Once a Year
$300
Twice a Year
60
60
9-29
9-30
IF-THEN/ELSE Statements
Only one executable statement is allowed in IF-THEN
and ELSE statements.

ELSE IF expression THEN statement;
ELSE statement;
For this business scenario, two statements must be
executed for each true expression.
Bonus=500;
if Country='US' then
Freq='Once a Year';
61
61
DO Group
Multiple statements are permitted in a DO group.
data work.bonus;
set orion.sales;
if Country='US' then do;
Bonus=500;
Freq='Once a Year';
DO group
end;
else if Country='AU' then do;
Bonus=300;
Freq='Twice a Year';
end;
run;
Each DO group ends with an END statement.
62
p109d07
62
IF-THEN DO/ELSE DO Statements

Multiple statements are also permitted in an ELSE DO
group.
IF expression THEN DO;

executable statements
END;
ELSE IF expression THEN DO;
executable statements
END;
ELSE DO;
executable statements;
END;
DO group
DO group
DO group
63
63
Viewing the Output

var First_Name Last_Name Country Bonus
Freq;
run;
64
Obs
First_Name
Last_Name
Country
Bonus
Freq
60
61
62
63
64
65
66
67
68
69
Billy
Matsuoka
Vino
Meera
Harry
Julienne
Scott
Cherda
Priscilla
Robert
Plested
Wills
George
Body
Highpoint
Magolan
Desanctis
Ridley
Farren
Stevens
AU
AU
AU
AU
US
US
US
US
US
US
300
300
300
300
500
500
500
500
500
500
Twice a Yea
Twice a Yea
Twice a Yea
Twice a Yea
Once a Year
Once a Year
Once a Year
Once a Year
Once a Year
Once a Year
truncation
p109d07
64
9-31
9-32
Compilation
data work.bonus;
set orion.sales;
Bonus=500;
Freq='Once a Year';
end;
Bonus=300;
end;
run;
PDV
Employee_ID
N8
First_Name
$ 12
...
Hire_Date
N8
p109d07
...
65
65
Compilation
data work.bonus;
set orion.sales;
Bonus=500;
Freq='Once a Year';
end;
Bonus=300;
end;
run;
PDV
Employee_ID
N8
First_Name
$ 12
...
Hire_Date
N8
Bonus
N8
66
66
...
Compilation
data work.bonus;
set orion.sales;
Bonus=500;
Freq='Once a Year';
end;
else if Country='AU'
then do;
11 characters
Bonus=300;
end;
run;
PDV
Employee_ID
N8
First_Name
$ 12
...
Hire_Date
N8
Bonus
N8
Freq
$ 11
...
67
67
Compilation
data work.bonus;
set orion.sales;
Bonus=500;
Freq='Once a Year';
end;
Bonus=300;
end;
length does
run;
12 characters
not change
PDV
Employee_ID
N8
First_Name
$ 12
...
Hire_Date
N8
Bonus
N8
Freq
$ 11
68
68
9-33
9-34
9.05 Quiz
How would you prevent Freq from being truncated?
69
69
Defining Character Variables

Set the length of the variable Freq to avoid truncation.
data work.bonus;
set orion.sales;
length Freq $ 12;
Bonus=500;
Freq='Once a Year';
end;
Bonus=300;
end;
run;
LENGTH variable(s) <$> length;
It is a good practice to use a LENGTH statement
any time you create a new character variable.
71
p109d08
71
The LENGTH statement is usually placed at or near the top of the DATA step.
Compilation
data work.bonus;
set orion.sales;
length Freq $ 12;
Bonus=500;
Freq='Once a Year';
end;
Bonus=300;
end;
run;
PDV
Employee_ID
N8
First_Name
$ 12
...
Hire_Date
N8
p109d08
...
72
72
Compilation
data work.bonus;
set orion.sales;
length Freq $ 12;
Bonus=500;
Freq='Once a Year';
end;
Bonus=300;
end;
run;
PDV
Employee_ID
N8
First_Name
$ 12
...
Hire_Date
N8
Freq
$ 12
73
73
...
9-35
9-36
Compilation
data work.bonus;
set orion.sales;
length Freq $ 12;
Bonus=500;
Freq='Once a Year';
end;
Bonus=300;
end;
run;
PDV
Employee_ID
N8
First_Name
$ 12
...
Hire_Date
N8
Freq
C
$ 12
Bonus
N8
...
74
74
Compilation
data work.bonus;
set orion.sales;
length Freq $ 12;
Bonus=500;
Freq='Once a Year';
end;
Bonus=300;
end;
run;
length does
not change
PDV
Employee_ID
N8
First_Name
$ 12
...
Hire_Date
N8
Freq
C
$ 12
Bonus
N8
75
75
Viewing the Output

var First_Name Last_Name Country
Bonus Freq;
run;
Obs
First_Name
Last_Name
Country
Bonus
60
61
62
63
64
65
66
67
68
69
Billy
Matsuoka
Vino
Meera
Harry
Julienne
Scott
Cherda
Priscilla
Robert
Plested
Wills
George
Body
Highpoint
Magolan
Desanctis
Ridley
Farren
Stevens
AU
AU
AU
AU
US
US
US
US
US
US
300
300
300
300
500
500
500
500
500
500
Freq
Twice a Year
Twice a Year
Twice a Year
Twice a Year
Once a Year
Once a Year
Once a Year
Once a Year
Once a Year
Once a Year
p109d08
76
76
Exercises
Level 1
4. Using Conditional Processing
b. In the DATA step, create a new variable, Method, and assign a value based on Order_Type.
If Order_Type is equal to 1 then Method equals Retail.
If Order_Type is equal to 2 then Method equals Catalog.
If Order_Type is equal to 3 then Method equals Internet.
For any other values or Order_Type, Method equals Unknown.
c. Modify the PROC PRINT step to display the report below. The results should contain 490
observations.
9-37
9-38

Obs
Order_ID
1
2
3
4
5
1230058123
1230080101
1230106883
1230147441
1230315085
Order_
Type
1
2
2
1
1
Method
Retail
Catalog
Catalog
Retail
Retail
5. Using Conditional Processing with DO Groups

b. In the DATA step, create three new variables: Discount, DiscountType, and Region.
If Country is equal to CA or US,
Discount is equal to 0.10
DiscountType is equal to Required
Region is equal to North America.
If Country is equal to any other value,
Discount is equal to 0.05
DiscountType is equal to Optional
Region is equal to Not North America.
c. The new data set should include only Supplier_Name, Country, Discount, DiscountType, and
Region.
d. Submit the program to create the report below. The results should contain 52 observations.
Obs
1
2
3
4
5
Supplier_Name
Scandinavian Clothing A/S
Petterson AB
Prime Sports Ltd
Top Sports
AllSeasons Outdoor Clothing
Country
NO
SE
GB
DK
US
Region
Not North America
Not North America
Not North America
Not North America
North America
Discount
Discount
Type
0.05
0.05
0.05
0.05
0.10
Optional
Optional
Optional
Optional
Required
Level 2
6. Creating Multiple Variables in Conditional Processing
a. Write a DATA step that reads orion.customer_dim to create work.season.
b. Create two new variables: Promo and Promo2.
The value of Promo is based on the quarter in which the customer was born.
If the customer was born in the first quarter, then Promo is equal to Winter.
If the customer was born in the second quarter, then Promo is equal to Spring.
If the customer was born in the third quarter, then Promo is equal to Summer.
If the customer was born in the fourth quarter, then Promo is equal to Fall.
The value of Promo2 is based on the customers age:
For young adults, whose age is between 18 and 25, set Promo2 equal to YA.
For seniors, aged 65 or older, set Promo2 equal to Senior.
Promo2 should have a missing value for all other customers.
c. The new data set should include only Customer_FirstName, Customer_LastName,
Customer_BirthDate, Customer_Age, Promo, and Promo2.
d. Create the report below. The results should include 77 observations.
Obs
1
2
3
4
5
6
7
8
Customer_
FirstName
Customer_
LastName
Customer_
BirthDate
Promo
James
Sandrina
Cornelia
Karen
Elke
David
Markus
Ulrich
Kvarniq
Stephano
Krahl
Ballinger
Wallstab
Black
Sepke
Heyde
27JUN1978
09JUL1983
27FEB1978
18OCT1988
16AUG1978
12APR1973
21JUL1992
16JAN1943
Spring
Summer
Winter
Fall
Summer
Spring
Summer
Winter
Customer_
Age
33
28
33
23
33
38
19
68
Promo2
YA
YA
Senior
7. Creating Variables Unconditionally and Conditionally

a. Write a DATA step that reads orion.orders to create work.ordertype.
b. Create a new variable, DayOfWeek, that is equal to the weekday of Order_Date.
c. Create the new variable Type, which is equal to
Retail Sale if Order_Type is equal to 1
Catalog Sale if Order_Type is equal to 2
Internet Sale if Order_Type is equal to 3.
d. Create the new variable SaleAds, which is equal to
Mail if Order_Type is equal to 2
Email if Order_Type is equal to 3.
e. Do not include Order_Type, Employee_ID, and Customer_ID in the new data set.
f. Create the report below. The results should contain 490 observations.
Obs
Order_ID
Order_
Date
1
2
3
1230058123
1230080101
1230106883
11JAN2007
15JAN2007
20JAN2007
Delivery_
Date
11JAN2007
19JAN2007
22JAN2007
Type
Sale
Ads
Day
Of
Week
Retail Sale
Catalog Sale
Catalog Sale
Mail
Mail
5
2
7
9-39
9-40

4
5
1230147441
1230315085
28JAN2007
27FEB2007
28JAN2007
27FEB2007
Retail Sale
Retail Sale
1
3
Challenge
8. Using WHEN Statements in a SELECT Group to Create Variables Conditionally
a. Write a DATA step that reads orion.nonsales to create work.gifts.
b. Create two new variables, Gift1 and Gift2, using a SELECT group with WHEN statements.
Documentation about the SELECT group with WHEN statements can be found in the SAS Help
facility or in the online documentation.
If Gender is equal to F,
Gift1 is equal to Scarf
Gift2 is equal to Pedometer.
If Gender is equal to M,
Gift1 is equal to Gloves
Gift2 is equal to Money Clip.
If Gender is not equal to F or M,
Gift1 is equal to Coffee
Gift2 is equal to Calendar.
c. The new data set should include only Employee_ID, First, Last, Gender, Gift1, and Gift2.
d. Create the report below. The results should contain 235 observations.
Employee_ID
120101
120104
120105
120106
120107
First
Last
Patrick
Kareen
Liz
John
Sherie
Lu
Billington
Povey
Hornsey
Sheedy
Gender
M
F
F
M
F
Gift1
Gift2
Gloves
Scarf
Scarf
Gloves
Scarf
Money Clip
Pedometer
Pedometer
Money Clip
Pedometer
9.3 Solutions
data work.increase;
set orion.staff;
Increase=Salary*0.10;
NewSalary=sum(Salary,Increase);
/* alternate statement is */
/* NewSalary=Salary+Increase; */
BdayQtr=qtr(Birth_Date);
keep Employee_ID Birth_Date Salary Increase NewSalary BdayQtr;
9.3 Solutions
format Salary Increase NewSalary comma8.;

run;
proc print data=work.increase label;
run;
data work.birthday;
set orion.customer;
Bday2012=mdy(month(Birth_Date),day(Birth_Date),2012);
BdayDOW2012=weekday(Bday2012);
Age2012=(Bday2012-Birth_Date)/365.25;
keep Customer_Name Birth_Date Bday2012 BdayDOW2012 Age2012;
format Bday2012 date9. Age2012 3.;
run;
proc print data=work.birthday;
run;
3. Using the CATX and INTCK Functions to Create Variables
data work.employees;
set orion.sales;
FullName=catx(' ',First_Name,Last_Name);
Yrs2012=intck('year',Hire_Date,'01JAN2012'd);
format Hire_Date ddmmyy10.;
label Yrs2012='Years of Employment in 2012';
run;
proc print data=work.employees label;
var FullName Hire_Date Yrs2012;
run;
4. Using Conditional Processing
data work.ordertype;
set orion.orders;
length Method $ 8;
if Order_Type=1 then Method='Retail';
else if Order_Type=2 then Method='Catalog';
else if Order_type=3 then Method='Internet';
else Method='Unknown';
run;
proc print data=work.ordertype;
var Order_ID Order_Type Method;
run;
5. Using Conditional Processing with DO Groups
data work.region;
set orion.supplier;
length Region $ 17;
if Country in ('CA','US') then do;
9-41
9-42
Discount=0.10;
DiscountType='Required';
Region='North America';
end;
else do;
Discount=0.05;
DiscountType='Optional';
Region='Not North America';
end;
keep Supplier_Name Country
Discount DiscountType Region;
run;
proc print data=work.region;
run;
6. Creating Multiple Variables in Conditional Processing
data work.season;
length Promo2 $ 6;
Quarter=qtr(Customer_BirthDate);
if Quarter=1 then Promo='Winter';
else if Quarter=2 then Promo='Spring';
else if Quarter=3 then Promo='Summer';
else if Quarter=4 then Promo='Fall';
if Customer_Age>=18 and Customer_Age<=25 then Promo2='YA';
else if Customer_Age>=65 then Promo2='Senior';
keep Customer_FirstName Customer_LastName Customer_BirthDate
Customer_Age Promo Promo2;
run;
proc print data=work.season;
var Customer_FirstName Customer_LastName Customer_BirthDate Promo
Customer_Age Promo2;
run;
7. Creating Variables Unconditionally and Conditionally
data work.ordertype;
set orion.orders;
length Type $ 13 SaleAds $ 5;
DayOfWeek=weekday(Order_Date);
if Order_Type=1 then
Type='Retail Sale';
else if Order_Type=2 then do;
Type='Catalog Sale';
SaleAds='Mail';
end;
else if Order_Type=3 then do;
Type='Internet Sale';
SaleAds='Email';
end;
9.3 Solutions
drop Order_Type Employee_ID Customer_ID;

run;
proc print data=work.ordertype;
run;
8. Using WHEN Statements in a SELECT Group to Create Variables Conditionally
data work.gifts;
set orion.nonsales;
length Gift1 $ 6 Gift2 $ 10;
select(Gender);
when('F') do;
Gift1='Scarf';
Gift2='Pedometer';
end;
when('M') do;
Gift1='Gloves';
Gift2='Money Clip';
end;
otherwise do;
Gift1='Coffee';
Gift2='Calendar';
end;
end;
keep Employee_ID First Last Gender Gift1 Gift2;
run;
proc print data=work.gifts noobs;
run;
9-43
9-44

Which of the following statements creates a numeric
variable Bonus with a value of 500?
a. Bonus=$500;
b. Bonus=500;
c. label Bonus='500';
d. format Bonus 500.;
Partial work.comp
Bonus
Month
Bonus
Compensation
500
108755
500
88475
500
27100
You use an assignment statement to set the value of

the variable Bonus equal to 500. Numeric constants
do not include commas or currency symbols.

A DROP statement has been added to this DATA step.
Will the program calculate Compensation and
BonusMonth correctly?
23
data work.comp;
24
set orion.sales;
25
drop Gender Salary Job_Title Country
26
Birth_Date Hire_Date;
27
Bonus=500;
28
29
30
run;
NOTE: The data set WORK.COMP has 165 observations and 6 variables.
Yes. A drop flag is set for the dropped variables, but

the variables are in the PDV and therefore available
for processing. DROP is a compile-time only
statement.
18
18
9.3 Solutions

In the previous program, is it possible for more than one
condition to be true for a single observation?
a. Yes, more than one condition can be true.
b. No, the conditions are mutually exclusive, so only one
condition can be true.
For each observation there is only one value for
Job_Title. If that value matches one of the
conditions, then it cannot match any other condition.
38

Program p109a02 reads orion.nonsales, a non-validated
data set. Open and submit the program and review the
results. Why is Bonus set to 300 in observations 125,
197, and 200?
data work.bonus;
set orion.nonsales;
else Bonus=300;
run;
The Country variable has some mixed case values in
orion.nonsales. Observations with a country value of
US are assigned 500. All others are assigned 300,
including us.
56
56
9-45
9-46

How would you prevent Freq from being truncated?
Possible solutions:
Pad the first occurrence of the Freq value with
blanks to be the length of the longest possible
value.
Switch conditional statements to place the
longest value of Freq in the first conditional
statement.
Add a LENGTH statement to declare the byte size
of the variable up front.
70
70
Chapter 10 Combining Data Sets

10.1 Concatenating Data Sets ............................................................................................. 10-3
Exercises ............................................................................................................................ 10-22
10.2 Merging Data Sets One-to-One .................................................................................. 10-25

10.3 Merging Data Sets One-to-Many ............................................................................... 10-30
Exercises ............................................................................................................................ 10-47
10.4 Merging Data Sets with Non-Matches ....................................................................... 10-49

Exercises ............................................................................................................................ 10-71
10.5 Solutions ..................................................................................................................... 10-73

Solutions to Exercises ........................................................................................................ 10-73
Solutions to Student Activities (Polls/Quizzes) ................................................................... 10-77
10-2
10.1 Concatenating Data Sets

Objectives
Concatenate two or more SAS data sets using

the SET statement in a DATA step.
Change the names of variables using the RENAME=
data set option.
3
3
Business Scenario
You have been asked to combine the data sets containing
information about Orion Star employees from Denmark
and France into a new data set.
empsdk
empsall1
empsfr
4
4
10-3
10-4
Considerations
Concatenate like-structured data sets empsdk and
empsfr to create a new data set named empsall1.
empsdk
First
Lars
Kari
Jonas
Gender
M
F
M
Country
Denmark
Denmark
Denmark
empsfr
First
Pierre
Sophie
Gender Country
M
France
F
France
empsall1
First
Lars
Kari
Jonas
Pierre
Sophie
Gender
M
F
M
M
F
Country
Denmark
Denmark
Denmark
France
France
Both data sets contain the same variables.

5
5
Using a DATA Step

Use a DATA step to concatenate the data sets. List the
data sets in the SET statement.
data empsall1;
set empsdk empsfr;
run;
SET SAS-data-set1 SAS-data-set2 . . .;
The SET statement reads observations from each

data set in the order in which they are listed.
Any number of data sets can be included in the SET
statement.
p110d01
Compilation
empsdk
First
Lars
Kari
Jonas
empsfr
Gender Country
M
Denmark
F
Denmark
M
Denmark
First
Pierre
Sophie
Gender Country
M
France
F
France
data empsall1;
set empsdk empsfr;
run;
PDV
First
empsall1
Gender
Country
First
Gender
Country
p110d01
...
7
7
Execution
empsdk
First
Lars
Kari
Jonas
Gender Country
M
Denmark
F
Denmark
M
Denmark
empsfr
First
Pierre
Sophie
Gender Country
M
France
F
France
Initialize PDV
data empsall1;
set empsdk empsfr;
run;
PDV
First
empsall1
Gender
Country
First
Gender
Country
8
8
...
10-5
10-6
Execution
empsdk
First
Lars
Kari
Jonas
empsfr
Gender Country
M
Denmark
F
Denmark
M
Denmark
First
Pierre
Sophie
Gender Country
M
France
F
France
data empsall1;
set empsdk empsfr;
run;
PDV
First
Lars
empsall1
Gender Country
M
Denmark
First
Gender
Country
...
9
9
Execution
empsdk
First
Lars
Kari
Jonas
Gender Country
M
Denmark
F
Denmark
M
Denmark
empsfr
First
Pierre
Sophie
Gender Country
M
France
F
France
data empsall1;
set empsdk empsfr;
Implicit OUTPUT;
run;
PDV
First
Lars
Implicit RETURN;
empsall1
Gender Country
M
Denmark
First
Lars
Gender Country
M
Denmark
10
10
...
Execution
empsdk
First
Lars
Kari
Jonas
empsfr
Gender Country
M
Denmark
F
Denmark
M
Denmark
First
Pierre
Sophie
Gender Country
M
France
F
France
data empsall1;
set empsdk empsfr;
run;
PDV
First
Lars
empsall1
Gender Country
M
Denmark
First
Lars
Gender Country
M
Denmark
Data set variables are not reinitialized.

...
11
11
Execution
empsdk
First
Lars
Kari
Jonas
Gender Country
M
Denmark
F
Denmark
M
Denmark
empsfr
First
Pierre
Sophie
Gender Country
M
France
F
France
data empsall1;
set empsdk empsfr;
run;
PDV
First
Kari
empsall1
Gender Country
F
Denmark
First
Lars
Gender Country
M
Denmark
12
12
...
10-7
10-8
Execution
empsdk
First
Lars
Kari
Jonas
empsfr
Gender Country
M
Denmark
F
Denmark
M
Denmark
First
Pierre
Sophie
Gender Country
M
France
F
France
data empsall1;
set empsdk empsfr;
Implicit OUTPUT;
run;
PDV
First
Kari
Implicit RETURN;
empsall1
Gender Country
F
Denmark
First
Lars
Kari
Gender Country
M
Denmark
F
Denmark
...
13
13
Execution
empsdk
First
Lars
Kari
Jonas
Gender Country
M
Denmark
F
Denmark
M
Denmark
empsfr
First
Pierre
Sophie
Gender Country
M
France
F
France
data empsall1;
set empsdk empsfr;
run;
PDV
First
Kari
empsall1
Gender Country
F
Denmark
First
Lars
Kari
Gender Country
M
Denmark
F
Denmark
PDV is not reinitialized.

14
14
...
Execution
empsdk
First
Lars
Kari
Jonas
empsfr
Gender Country
M
Denmark
F
Denmark
M
Denmark
First
Pierre
Sophie
Gender Country
M
France
F
France
data empsall1;
set empsdk empsfr;
run;
PDV
First
Jonas
empsall1
Gender Country
M
Denmark
First
Lars
Kari
Gender Country
M
Denmark
F
Denmark
...
15
15
Execution
empsdk
First
Lars
Kari
Jonas
Gender Country
M
Denmark
F
Denmark
M
Denmark
empsfr
First
Pierre
Sophie
Gender Country
M
France
F
France
data empsall1;
set empsdk empsfr;
Implicit OUTPUT;
run;
PDV
First
Jonas
Implicit RETURN;
empsall1
Gender Country
M
Denmark
First
Lars
Kari
Jonas
Gender Country
M
Denmark
F
Denmark
M
Denmark
16
16
...
10-9
10-10
Execution
empsdk
First
Lars
Kari
Jonas
empsfr
Gender Country
M
Denmark
F
Denmark
M
Denmark
First
Pierre
Sophie
Gender Country
M
France
F
France
data empsall1;
set empsdk empsfr;
run;
PDV
First
Jonas
empsall1
Gender Country
M
Denmark
First
Lars
Kari
Jonas
Gender Country
M
Denmark
F
Denmark
M
Denmark
...
17
17
Execution
empsdk
First
Lars
Kari
EOF
Jonas
Gender Country
M
Denmark
F
Denmark
M
Denmark
empsfr
First
Pierre
Sophie
Gender Country
M
France
F
France
data empsall1;
set empsdk empsfr;
run;
PDV
First
Jonas
empsall1
Gender Country
M
Denmark
First
Lars
Kari
Jonas
Gender Country
M
Denmark
F
Denmark
M
Denmark
18
18
...
10-11
Execution
empsdk
First
Lars
Kari
Jonas
empsfr
Gender Country
M
Denmark
F
Denmark
M
Denmark
First
Pierre
Sophie
Gender Country
M
France
F
France
data empsall1; Reinitialize PDV

set empsdk empsfr;
run;
PDV
First
empsall1
Gender Country
PDV is reinitialized before

processing the next data set.
First
Lars
Kari
Jonas
Gender Country
M
Denmark
F
Denmark
M
Denmark
...
19
19
When concatenating data sets, SAS reinitializes the entire PDV before it begins reading the next data set
listed in the SET statement.
Execution
empsdk
First
Lars
Kari
Jonas
Gender Country
M
Denmark
F
Denmark
M
Denmark
empsfr
First
Pierre
Sophie
Gender Country
M
France
F
France
data empsall1;
set empsdk empsfr;
run;
PDV
First
Pierre
empsall1
Gender Country
M
France
First
Lars
Kari
Jonas
Gender Country
M
Denmark
F
Denmark
M
Denmark
20
20
...
10-12
Execution
empsdk
First
Lars
Kari
Jonas
empsfr
Gender Country
M
Denmark
F
Denmark
M
Denmark
First
Pierre
Sophie
Gender Country
M
France
F
France
data empsall1;
set empsdk empsfr;
Implicit OUTPUT;
run;
Implicit RETURN;
PDV
First
Pierre
empsall1
Gender Country
M
France
First
Lars
Kari
Jonas
Pierre
Gender
M
F
M
M
Country
Denmark
Denmark
Denmark
France
...
21
21
Execution
empsdk
First
Lars
Kari
Jonas
Gender Country
M
Denmark
F
Denmark
M
Denmark
empsfr
First
Pierre
Sophie
Gender Country
M
France
F
France
data empsall1;
set empsdk empsfr;
run;
PDV
First
Pierre
empsall1
Gender Country
M
France
First
Lars
Kari
Jonas
Pierre
Gender
M
F
M
M
Country
Denmark
Denmark
Denmark
France
22
22
...
Execution
empsdk
First
Lars
Kari
Jonas
empsfr
Gender Country
M
Denmark
F
Denmark
M
Denmark
First
Pierre
Sophie
Gender Country
M
France
F
France
data empsall1;
set empsdk empsfr;
run;
PDV
First
Sophie
empsall1
Gender Country
F
France
First
Lars
Kari
Jonas
Pierre
Gender
M
F
M
M
Country
Denmark
Denmark
Denmark
France
...
23
23
Execution
empsdk
First
Lars
Kari
Jonas
Gender Country
M
Denmark
F
Denmark
M
Denmark
empsfr
First
Pierre
Sophie
Gender Country
M
France
F
France
data empsall1;
set empsdk empsfr;
Implicit OUTPUT;
run;
PDV
First
Sophie
24
Implicit RETURN;
empsall1
Gender Country
F
France
First
Lars
Kari
Jonas
Pierre
Sophie
Gender
M
F
M
M
F
Country
Denmark
Denmark
Denmark
France
France ...
24
10-13
10-14
Execution
empsdk
First
Lars
Kari
Jonas
empsfr
Gender Country
M
Denmark
F
Denmark
M
Denmark
First
Pierre
Sophie
Gender Country
M
France
F
France
data empsall1;
set empsdk empsfr;
run;
PDV
First
Sophie
empsall1
Gender Country
F
France

25
First
Lars
Kari
Jonas
Pierre
Sophie
Gender
M
F
M
M
F
Country
Denmark
Denmark
Denmark
France
France ...
25
Execution
empsdk
First
Lars
Kari
Jonas
Gender Country
M
Denmark
F
Denmark
M
Denmark
empsfr
First
Pierre
EOFSophie
Gender Country
M
France
F
France
data empsall1;
set empsdk empsfr;
run;
PDV
First
Sophie
26
empsall1
Gender Country
F
France
First
Lars
Kari
Jonas
Pierre
Sophie
Gender
M
F
M
M
F
Country
Denmark
Denmark
Denmark
France
France
26
Viewing the Log

Partial SAS Log
145
146
147
data empsall1;
set empsdk empsfr;
run;
NOTE: There were 3 observations read from the data set WORK.EMPSDK.
NOTE: There were 2 observations read from the data set WORK.EMPSFR.
NOTE: The data set WORK.EMPSALL1 has 5 observations and 3
variables.
27
27
Unlike-Structured Data Sets

Concatenate empscn and empsjp to create a new data
set named empsall2.
empscn
First
Chang
Li
Ming
empsjp
Gender
M
M
F
Country
China
China
China
First
Cho
Tomi
Gender Region
F
Japan
M
Japan
The data sets do not contain the same variables.

data empsall2;
set empscn empsjp;
run;
28
p110d02
28
10-15
10-16
10.01 Quiz
How many variables will be in empsall2 after
concatenating empscn and empsjp?
empscn
First
Chang
Li
Ming
empsjp
Gender
M
M
F
Country
China
China
China
First
Cho
Tomi
Gender Region
F
Japan
M
Japan
data empsall2;
set empscn empsjp;
run;
29
29
Compilation
empscn
First
Chang
Li
Ming
Gender
M
M
F
Country
China
China
China
empsjp
First
Cho
Tomi
Gender Region
F
Japan
M
Japan
data empsall2;
set empscn empsjp;
run;
PDV
First
31
Gender Country
p110d02
31
...
Compilation
empscn
First
Chang
Li
Ming
empsjp
Gender
M
M
F
Country
China
China
China
First
Cho
Tomi
Gender Region
F
Japan
M
Japan
data empsall2;
set empscn empsjp;
run;
PDV
First
Gender Country Region
32
32
Final Results
empsall2
First
Chang
Li
Ming
Cho
Tomi
Gender
M
M
F
F
M
Country Region
China
China
China
Japan
Japan
Region has missing values due to PDV initialization.

Country has missing values due to PDV
reinitialization before processing the second data set.
33
33
10-17
10-18
Business Scenario
Rename variables in one or more data sets to align
columns.
empscn
First
Chang
Li
Ming
empsjp
Gender
M
M
F
Country
China
China
China
First
Cho
Tomi
Gender Region
F
Japan
M
Japan
34
34
RENAME= Data Set Option

The RENAME= data set option changes the name
of a variable.
data empsall2;
set empscn empsjp(rename=(Region=Country));
run;
SAS-data-set (RENAME=(old-name-1=new-name-1
old-name-2=new-name-2
...
old-name-n=new-name-n))
The RENAME= option must be specified in parentheses

immediately after the appropriate SAS data set name.
The name change affects the PDV and the output data
set. It has no effect on the input data set.
35
35
RENAME= Data Set Option

Multiples variables can be renamed in one or more data
sets.
set empscn(rename=(Country=Region))
empsjp;
set empscn(rename=(First=Fname
Country=Region))
empsjp(rename=(First=Fname));
set empscn
empsjp(rename=(Region=Country));
36
36
10.02 Quiz
Which statement has correct syntax?
a. Answer
set empscn(rename(Country=Location))
empsjp(rename(Region=Location));
b. Answer
set empscn(rename=(Country=Location))
empsjp(rename=(Region=Location));
c. Answer
set empscn rename=(Country=Location)
empsjp rename=(Region=Location);
37
37
10-19
10-20
Compilation
empscn
First
Chang
Li
Ming
empsjp
Gender
M
M
F
Country
China
China
China
First
Cho
Tomi
Gender Region
F
Japan
M
Japan
data empsall2;
run;
PDV
First
Gender Country
p110d03
...
39
39
Compilation
empscn
First
Chang
Li
Ming
Gender
M
M
F
Country
China
China
China
empsjp
First
Cho
Tomi
Gender Region
F
Japan
M
Japan
data empsall2;
run;
PDV
First
Gender Country
40
40
...
Compilation
empscn
First
Chang
Li
Ming
empsjp
Gender
M
M
F
Country
China
China
China
First
Cho
Tomi
Gender Region
F
Japan
M
Japan
data empsall2;
run;
PDV
First
Gender Country
...
41
41
Compilation
empscn
First
Chang
Li
Ming
Gender
M
M
F
Country
China
China
China
empsjp
First
Cho
Tomi
Gender Region
F
Japan
M
Japan
data empsall2;
run;
PDV
First
Gender Country
42
42
...
10-21
10-22
Compilation
empscn
empsjp
First
Chang
Li
Ming
Gender
M
M
F
Country
China
China
China
First
Cho
Tomi
Gender Region
F
Japan
M
Japan
data empsall2;
run;
PDV
First
Gender Country
43
43
Final Results
The Region values are stored in Country.
empsall2
First
Chang
Li
Ming
Cho
Tomi
Gender
M
M
F
F
M
Country
China
China
China
Japan
Japan
44
44
Exercises
10-23
Level 1
1. Concatenating Like-Structured Data Sets
a. Write and submit a DATA step to concatenate orion.mnth7_2011, orion.mnth8_2011, and
orion.mnth9_2011 to create a new data set, work.thirdqtr.
How many observations in work.thirdqtr are from orion.mnth7_2011? ____________
b. Write a PROC PRINT step to create the report below. The results should contain 32 observations.
Obs
Order_ID
1
2
3
4
5
1242691897
1242736731
1242773202
1242782701
1242827683
Order_
Type
2
1
3
3
1
Employee_ID
Customer_ID
Order_
Date
Delivery_
Date
99999999
121107
99999999
99999999
121105
90
10
24
27
10
02JUL2011
07JUL2011
11JUL2011
12JUL2011
17JUL2011
04JUL2011
07JUL2011
14JUL2011
17JUL2011
17JUL2011
2. Concatenating Unlike-Structured Data Sets

Open p110e02. Submit the two PROC CONTENTS steps or explore the data sets interactively to
compare the variables in the two data sets. What are the names of the two variables that are
different in the two data sets?
orion.sales
orion.nonsales
a. Add a DATA step after the PROC CONTENTS steps to concatenate orion.sales and
orion.nonsales to create a new data set, work.allemployees.
Use a RENAME= data set option to change the names of the different variables in
orion.nonsales.
The new data set should include only Employee_ID, First_Name, Last_Name, Job_Title, and
Salary.
b. Add a PROC PRINT step to create the report below. The results should contain 400 observations.
Obs
Employee_ID
1
2
3
4
5
120102
120103
120121
120122
120123
First_
Name
Last_Name
Salary
Tom
Wilson
Irenie
Christina
Kimiko
Zhou
Dawes
Elvish
Ngan
Hotstone
108255
87975
26600
27475
26190
Job_Title
Sales
Sales
Sales
Sales
Sales
Manager
Manager
Rep. II
Rep. II
Rep. I
10-24
Level 2
3. Concatenating Data Sets with Variables of Different Lengths and Types
a. Open p110e03. Submit the PROC CONTENTS steps or explore the data sets interactively to
complete the table below by filling in attribute information for each variable in each data set.
Code
Type
Company
Length
Type
Length
ContactType
Type
Length
orion.charities
orion.us_suppliers
orion.consultants
b. Write a DATA step to concatenate orion.charities and orion.us_suppliers, creating a temporary
data set, contacts.
c. Submit a PROC CONTENTS step to examine work.contacts. From which input data set were the
variable attributes assigned? ___________________________________________
d. Write a DATA step to concatenate orion.us_suppliers and orion.charities, creating a temporary
data set, contacts2. Note that these are the same data sets as the previous program, but they are in
reverse order.
e. Submit a PROC CONTENTS step to examine work.contacts2. From which input data set were
the variable attributes assigned?
_____________________________________________________________
f. Write a DATA step to concatenate orion.us_suppliers and orion.consultants, creating a
temporary data set, contacts3.
Why did the DATA step fail? ________________________________________________
10.2 Merging Data Sets One-to-One

Objectives
Prepare data sets for merging using the SORT

procedure.
Merge SAS data sets one-to-one based on a
common variable.
48
48
Match-Merging
A B C
1
2
3
C D E
1
2
3
One-to-One
A single observation in one data set is
related to exactly one observation in
another data set based on the values of
one or more selected variables.
A B C
1
2
C D E
1
1
2
One-to-Many
A single observation in one data set is
related to more than one observation in
A B C
1
2
4
C D E
2
3
4
Non-matches
At least one observation in one data set
is unrelated to any observation in
49
49
10-25
10-26
Match-Merging: Sorting the Data Sets

The data sets in a match-merge must be sorted by the
common variable or variables being matched.
PROC SORT DATA=input-SAS-data-set
<OUT=output-SAS-data-set>;
BY <DESCENDING> by-variable(s);
RUN;
50
50

Which of the following BY statements correctly sorts by
descending salary within gender?
a.
b.
c.
d.
by descending salary within gender;

by descending salary gender;
by gender descending salary;
by gender salary descending;
51
Business Scenario
Merge the Australian employee data set with a phone
data set to obtain each employees home phone number,
storing the results in a new data set.
empsau
First
phoneh
Gender EmpID
EmpID
Phone
empsauh
empsauh
First
Gender
EmpID Phone
53
53
Match-Merging
The MERGE statement in a DATA step joins observations
from two or more SAS data sets into single observations.
data empsauh;
merge empsau phoneh;
by EmpID;
run;
MERGE SAS-data-set1 SAS-data-set2 . . .;
BY <DESCENDING> BY-variable(s);
A BY statement indicates a match-merge and lists the
variable or variables to match.
54
p110d04
54
10-27
10-28
MERGE and BY Statements

Requirements for match-merging:
Two or more data sets are listed in the MERGE
statement.
The variables in the BY statement must be common
to all data sets.
The data sets must be sorted by the variables listed in
the BY statement.
55
55
One-to-One Merge
One observation in empsau matches exactly one
observation in phoneh.
empsau
First
Togar
Kylie
Birin
phoneh
Gender EmpID
M
121150
F
121151
M
121152
EmpID
121150
121151
121152
Phone
+61(2)5555-1793
+61(2)5555-1849
+61(2)5555-1665
The data sets are sorted by EmpID.
56
56
Final Results
empsau
First
Togar
Kylie
Birin
phoneh
EmpID
Phone
121150 +61(2)5555-1793
121151 +61(2)5555-1849
121152 +61(2)5555-1665
Gender EmpID
M
121150
F
121151
M
121152
empsauh
First
Togar
Kylie
Birin
Gender
M
F
M
EmpID
121150
121151
121152
Phone
+61(2)5555-1793
+61(2)5555-1849
+61(2)5555-1665
57
57
10.04 Quiz
Retrieve program p110a01.

Complete the program to match-merge the sorted
SAS data sets referenced in the PROC SORT steps.
Submit the program. Correct and resubmit, if
necessary.
What are the modified, completed statements?
58
58
10-29
10-30
10.3 Merging Data Sets One-to-Many

Objectives
Merge SAS data sets one-to-many based on a

common variable.
62
62
Business Scenario
Merge the Australian employee information data set with
the phone data set to obtain the phone numbers for each
employee.
phones
empsau
empphones
63
63
Considerations
In this one-to-many merge, one observation in empsau
matches one or more observations in phones.
phones
empsau
First
Togar
Kylie
Birin
Gender EmpID
M
121150
F
121151
M
121152
EmpID
121150
121150
121151
121152
121152
121152
Type
Home
Work
Home
Work
Home
Cell
Phone
+61(2)5555-1793
+61(2)5555-1794
+61(2)5555-1849
+61(2)5555-1850
+61(2)5555-1665
+61(2)5555-1666
64
64
Match-Merging
Merge the two data sets by EmpID and create a new data
set named empphones.
data empphones;
merge empsau phones;
by EmpID;
run;
65
p110d05
65
10-31
10-32
Match-Merging
BY group
phones
EmpID
121150
121150
121151
121152
121152
121152
empsau
First
Togar
Kylie
Birin
Gender EmpID
M
121150
F
121151
M
121152
Type
Home
Work
Home
Work
Home
Cell
Phone
+61(2)5555-1793
+61(2)5555-1794
+61(2)5555-1849
+61(2)5555-1850
+61(2)5555-1665
+61(2)5555-1666
The common variable is called the BY variable.

The value of the BY variable is the BY value.
A group of observations with the same BY value is a
BY group.
66
66
Execution
empsau
First
Togar
Kylie
Birin
Gender EmpID
M
121150
F
121151
M
121152
phones
EmpID
121150
121150
121151
121152
121152
121152
Type
Home
Work
Home
Work
Home
Cell
Phone
+61(2)5555-1793
+61(2)5555-1794
+61(2)5555-1849
+61(2)5555-1850
+61(2)5555-1665
+61(2)5555-1666
Initialize PDV
data empphones;
by EmpID;
run;
PDV
First
67
Gender EmpID
.
Type
Phone
p110d05
67
...
Execution
empsau
First
Togar
Kylie
Birin
Gender EmpID
M
121150
F
121151
M
121152
phones
EmpID
121150
121150
121151
121152
121152
121152
data empphones;
by EmpID;
run;
PDV
First
Gender EmpID
.
Type
Home
Work
Home
Work
Home
Cell
Phone
+61(2)5555-1793
+61(2)5555-1794
+61(2)5555-1849
+61(2)5555-1850
+61(2)5555-1665
+61(2)5555-1666
Do the EmpID values match?

Yes
Type
Phone
...
68
68
Execution
empsau
First
Togar
Kylie
Birin
Gender EmpID
M
121150
F
121151
M
121152
phones
EmpID
121150
121150
121151
121152
121152
121152
data empphones;
by EmpID;
run;
PDV
First
Togar
Gender EmpID Type

M
121150 Home
Type
Home
Work
Home
Work
Home
Cell
Phone
+61(2)5555-1793
+61(2)5555-1794
+61(2)5555-1849
+61(2)5555-1850
+61(2)5555-1665
+61(2)5555-1666
Reads both observations

into the PDV.
Phone
+61(2)5555-1793
69
69
...
10-33
10-34
Execution
empsau
First
Togar
Kylie
Birin
Gender EmpID
M
121150
F
121151
M
121152
phones
EmpID
121150
121150
121151
121152
121152
121152
Type
Home
Work
Home
Work
Home
Cell
Phone
+61(2)5555-1793
+61(2)5555-1794
+61(2)5555-1849
+61(2)5555-1850
+61(2)5555-1665
+61(2)5555-1666
data empphones;
by EmpID;
run;
Implicit OUTPUT;
PDV
First
Togar
Implicit RETURN;
Gender EmpID Type
M
121150 Work
Phone
+61(2)5555-1793
...
70
70
Execution
empsau
First
Togar
Kylie
Birin
Gender EmpID
M
121150
F
121151
M
121152
phones
EmpID
121150
121150
121151
121152
121152
121152
Type
Home
Work
Home
Work
Home
Cell
Phone
+61(2)5555-1793
+61(2)5555-1794
+61(2)5555-1849
+61(2)5555-1850
+61(2)5555-1665
+61(2)5555-1666
data empphones;
by EmpID;
run;
PDV
First
Togar
Data set variables are not reinitialized.

Gender EmpID Type
M
121150 Work
Phone
+61(2)5555-1793
71
71
...
Execution
empsau
First
Togar
Kylie
Birin
Gender EmpID
M
121150
F
121151
M
121152
phones
EmpID
121150
121150
121151
121152
121152
121152
data empphones;
by EmpID;
run;
PDV
First
Togar
Gender EmpID Type

M
121150 Home
Type
Home
Work
Home
Work
Home
Cell
Phone
+61(2)5555-1793
+61(2)5555-1794
+61(2)5555-1849
+61(2)5555-1850
+61(2)5555-1665
+61(2)5555-1666

No
Phone
+61(2)5555-1793
...
72
72
Execution
empsau
First
Togar
Kylie
Birin
Gender EmpID
M
121150
F
121151
M
121152
phones
EmpID
121150
121150
121151
121152
121152
121152
data empphones;
by EmpID;
run;
PDV
First
Togar
Gender EmpID Type

M
121150 Home
Type
Home
Work
Home
Work
Home
Cell
Phone
+61(2)5555-1793
+61(2)5555-1794
+61(2)5555-1849
+61(2)5555-1850
+61(2)5555-1665
+61(2)5555-1666
Did EmpID change?

No
Phone
+61(2)5555-1793
73
73
...
10-35
10-36
Execution
phones
EmpID
121150
121150
121151
121152
121152
121152
empsau
First
Togar
Kylie
Birin
Gender EmpID
M
121150
F
121151
M
121152
data empphones;
by EmpID;
run;
PDV
First
Togar
Gender EmpID Type

M
121150 Work
Type
Home
Work
Home
Work
Home
Cell
Phone
+61(2)5555-1793
+61(2)5555-1794
+61(2)5555-1849
+61(2)5555-1850
+61(2)5555-1665
+61(2)5555-1666
Reads the matching

observation into the PDV.
Phone
+61(2)5555-1794
...
74
74
Execution
phones
empsau
First
Togar
Kylie
Birin
Gender EmpID
M
121150
F
121151
M
121152
EmpID
121150
121150
121151
121152
121152
121152
Type
Home
Work
Home
Work
Home
Cell
Phone
+61(2)5555-1793
+61(2)5555-1794
+61(2)5555-1849
+61(2)5555-1850
+61(2)5555-1665
+61(2)5555-1666
data empphones;
merge empsau phonehW;
by EmpID;
run;
PDV
First
Togar
Implicit OUTPUT;
Implicit RETURN;
Gender EmpID Type

M
121150 Work
Phone
+61(2)5555-1794
75
75
...
Execution
empsau
First
Togar
Kylie
Birin
Gender EmpID
M
121150
F
121151
M
121152
phones
EmpID
121150
121150
121151
121152
121152
121152
data empphones;
by EmpID;
run;
PDV
First
Togar
Gender EmpID Type

M
121150 Work
Type
Home
Work
Home
Work
Home
Cell
Phone
+61(2)5555-1793
+61(2)5555-1794
+61(2)5555-1849
+61(2)5555-1850
+61(2)5555-1665
+61(2)5555-1666

Yes
Phone
+61(2)5555-1794
...
76
76
Execution
empsau
First
Togar
Kylie
Birin
Gender EmpID
M
121150
F
121151
M
121152
phones
EmpID
121150
121150
121151
121152
121152
121152
data empphones;
by EmpID;
run;
PDV
First
Togar
Gender EmpID Type

M
121150 Work
Type
Home
Work
Home
Work
Home
Cell
Phone
+61(2)5555-1793
+61(2)5555-1794
+61(2)5555-1849
+61(2)5555-1850
+61(2)5555-1665
+61(2)5555-1666
Did EmpID change?

Yes
Phone
+61(2)5555-1794
77
77
...
10-37
10-38
Execution
empsau
First
Togar
Kylie
Birin
Gender EmpID
M
121150
F
121151
M
121152
phones
EmpID
121150
121150
121151
121152
121152
121152
Type
Home
Work
Home
Work
Home
Cell
Phone
+61(2)5555-1793
+61(2)5555-1794
+61(2)5555-1849
+61(2)5555-1850
+61(2)5555-1665
+61(2)5555-1666
data empphones;
by EmpID;
run;
Reinitialize PDV
PDV
First
Gender EmpID
.
Type
Phone
...
78
78
SAS reinitializes the entire program data vector before processing a different BY group.
Execution
empsau
First
Togar
Kylie
Birin
Gender EmpID
M
121150
F
121151
M
121152
phones
EmpID
121150
121150
121151
121152
121152
121152
data empphones;
by EmpID;
run;
PDV
First
Kylie
Gender EmpID Type

F
121151 Home
Type
Home
Work
Home
Work
Home
Cell
Phone
+61(2)5555-1793
+61(2)5555-1794
+61(2)5555-1849
+61(2)5555-1850
+61(2)5555-1665
+61(2)5555-1666

into the PDV.
Phone
+61(2)5555-1849
79
79
...
Execution
phones
empsau
First
Togar
Kylie
Birin
Gender EmpID
M
121150
F
121151
M
121152
EmpID
121150
121150
121151
121152
121152
121152
Type
Home
Work
Home
Work
Home
Cell
Phone
+61(2)5555-1793
+61(2)5555-1794
+61(2)5555-1849
+61(2)5555-1850
+61(2)5555-1665
+61(2)5555-1666
data empphones;
by EmpID;
run;
PDV
First
Kylie
Implicit OUTPUT;
Implicit RETURN;
Gender EmpID Type

F
121151 Home
Phone
+61(2)5555-1849
...
80
80
Execution
empsau
First
Togar
Kylie
Birin
Gender EmpID
M
121150
F
121151
M
121152
phones
EmpID
121150
121150
121151
121152
121152
121152
data empphones;
by EmpID;
run;
PDV
First
Kylie
Gender EmpID Type

F
121151 Home
Type
Home
Work
Home
Work
Home
Cell
Phone
+61(2)5555-1793
+61(2)5555-1794
+61(2)5555-1849
+61(2)5555-1850
+61(2)5555-1665
+61(2)5555-1666

Yes
Phone
+61(2)5555-1849
81
81
...
10-39
10-40
Execution
empsau
First
Togar
Kylie
Birin
Gender EmpID
M
121150
F
121151
M
121152
phones
EmpID
121150
121150
121151
121152
121152
121152
data empphones;
by EmpID;
run;
PDV
First
Kylie
Gender EmpID Type

F
121151 Home
Type
Home
Work
Home
Work
Home
Cell
Phone
+61(2)5555-1793
+61(2)5555-1794
+61(2)5555-1849
+61(2)5555-1850
+61(2)5555-1665
+61(2)5555-1666
Did EmpID change?

Yes
Phone
+61(2)5555-1849
...
82
82
Execution
empsau
First
Togar
Kylie
Birin
Gender EmpID
M
121150
F
121151
M
121152
phones
EmpID
121150
121150
121151
121152
121152
121152
Type
Home
Work
Home
Work
Home
Cell
Phone
+61(2)5555-1793
+61(2)5555-1794
+61(2)5555-1849
+61(2)5555-1850
+61(2)5555-1665
+61(2)5555-1666
Reinitialize PDV
data empphones;
by EmpID;
run;
PDV
First
Gender EmpID
.
Type
Phone
83
83
...
Execution
phones
empsau
First
Togar
Kylie
Birin
Gender EmpID
M
121150
F
121151
M
121152
EmpID
121150
121150
121151
121152
121152
121152
data empphones;
by EmpID;
run;
PDV
First
Birin
Gender
M
EmpID Type
121152 Work
Type
Home
Work
Home
Work
Home
Cell
Phone
+61(2)5555-1793
+61(2)5555-1794
+61(2)5555-1849
+61(2)5555-1850
+61(2)5555-1665
+61(2)5555-1666

into the PDV.
Phone
+61(2)5555-1850
...
84
84
Execution
phones
empsau
First
Togar
Kylie
Birin
Gender EmpID
M
121150
F
121151
M
121152
EmpID
121150
121150
121151
121152
121152
121152
Type
Home
Work
Home
Work
Home
Cell
Phone
+61(2)5555-1793
+61(2)5555-1794
+61(2)5555-1849
+61(2)5555-1850
+61(2)5555-1665
+61(2)5555-1666
data empphones;
by EmpID;
run;
PDV
First
Birin
Implicit OUTPUT;
Implicit RETURN;
Gender
M
EmpID Type
121152 Work
Phone
+61(2)5555-1850
85
85
...
10-41
10-42
Execution
phones
empsau
First
Togar
Kylie
Birin
Gender EmpID
M
121150
F
121151
M
121152
EOF
EmpID
121150
121150
121151
121152
121152
121152
data empphones;
by EmpID;
run;
PDV
First
Birin
Gender
M
EmpID Type
121152 Work
Type
Home
Work
Home
Work
Home
Cell
Phone
+61(2)5555-1793
+61(2)5555-1794
+61(2)5555-1849
+61(2)5555-1850
+61(2)5555-1665
+61(2)5555-1666
Did EmpID change?

No
Phone
+61(2)5555-1850
...
86
86
Execution
phones
empsau
First
Togar
Kylie
Birin
Gender EmpID
M
121150
F
121151
M
121152
EOF
EmpID
121150
121150
121151
121152
121152
121152
data empphones;
by EmpID;
run;
PDV
First
Birin
Gender
M
Type
Home
Work
Home
Work
Home
Cell
Phone
+61(2)5555-1793
+61(2)5555-1794
+61(2)5555-1849
+61(2)5555-1850
+61(2)5555-1665
+61(2)5555-1666
Reads the matching

EmpID Type
Phone
121152 Home +61(2)5555-1665
87
87
...
Execution
phones
empsau
First
Togar
Kylie
Birin
Gender EmpID
M
121150
F
121151
M
121152
EOF
EmpID
121150
121150
121151
121152
121152
121152
Type
Home
Work
Home
Work
Home
Cell
Phone
+61(2)5555-1793
+61(2)5555-1794
+61(2)5555-1849
+61(2)5555-1850
+61(2)5555-1665
+61(2)5555-1666
data empphones;
by EmpID;
run;
PDV
First
Birin
Implicit OUTPUT;
Implicit RETURN;
Gender
M
EmpID Type
Phone
121152 Home +61(2)5555-1665
...
88
88
Execution
phones
empsau
First
Togar
Kylie
Birin
Gender EmpID
M
121150
F
121151
M
121152
EOF
EmpID
121150
121150
121151
121152
121152
121152
data empphones;
by EmpID;
run;
PDV
First
Birin
Gender
M
EmpID Type
121152 Home
Type
Home
Work
Home
Work
Home
Cell
Phone
+61(2)5555-1793
+61(2)5555-1794
+61(2)5555-1849
+61(2)5555-1850
+61(2)5555-1665
+61(2)5555-1666
Did EmpID change?

No
Phone
+61(2)5555-1665
89
89
...
10-43
10-44
Execution
phones
empsau
First
Togar
Kylie
Birin
Gender EmpID
M
121150
F
121151
M
121152
EOF
EmpID
121150
121150
121151
121152
121152
121152
data empphones;
by EmpID;
run;
PDV
First
Birin
Gender
M
EmpID Type
121152 Cell
Type
Home
Work
Home
Work
Home
Cell
Phone
+61(2)5555-1793
+61(2)5555-1794
+61(2)5555-1849
+61(2)5555-1850
+61(2)5555-1665
+61(2)5555-1666
Reads the matching

Phone
+61(2)5555-1666
...
90
90
Execution
phones
empsau
First
Togar
Kylie
Birin
Gender EmpID
M
121150
F
121151
M
121152
EOF
EmpID
121150
121150
121151
121152
121152
121152
Type
Home
Work
Home
Work
Home
Cell
Phone
+61(2)5555-1793
+61(2)5555-1794
+61(2)5555-1849
+61(2)5555-1850
+61(2)5555-1665
+61(2)5555-1666
data empphones;
by EmpID;
run;
PDV
First
Birin
Implicit OUTPUT;
Implicit RETURN;
Gender
M
EmpID Type
121152 Cell
Phone
+61(2)5555-1666
91
91
...
Execution
empsau
First
Togar
Kylie
Birin
EOF
phones
EmpID
121150
Gender EmpID
121150
M
121150
121151
F
121151
121152
M
121152
121152
EOF 121152
Type
Home
Work
Home
Work
Home
Cell
Phone
+61(2)5555-1793
+61(2)5555-1794
+61(2)5555-1849
+61(2)5555-1850
+61(2)5555-1665
+61(2)5555-1666
data empphones;
by EmpID;
run;
PDV
First
Birin
Gender
M
EmpID Type
121152 Cell
Phone
+61(2)5555-1666
...
92
92
Final Results
empphones
First
Togar
Togar
Kylie
Birin
Birin
Birin
Gender
M
M
F
M
M
M
EmpID
121150
121150
121151
121152
121152
121152
Type
Home
Work
Home
Work
Home
Cell
Phone
+61(2)5555-1793
+61(2)5555-1794
+61(2)5555-1849
+61(2)5555-1850
+61(2)5555-1665
+61(2)5555-1666
93
93
10-45
10-46
10.05 Quiz
In a one-to-many merge, does it matter which data set is
listed first in the MERGE statement?
Open p110a02 and submit
Reverse the order of the data sets and submit again.
Observe the results. How are they different?
95
Many-to-One Merge
One or more rows in one data set match exactly one row
in the other data set.
phones
EmpID
121150
121150
121151
121152
121152
121152
Type
Home
Work
Home
Work
Home
Cell
Phone
+61(2)5555-1793
+61(2)5555-1794
+61(2)5555-1849
+61(2)5555-1850
+61(2)5555-1665
+61(2)5555-1666
data phones;
merge phones empsau;
by EmpID;
run;
97
empsau
EmpID
First
121150 Togar
121151 Kylie
121512 Birin
Gender
M
F
M
p110d06
97
10-47
Viewing the Output

PROC PRINT Output
Obs
EmpID
Type
Phone
First
Gender
1
2
3
4
5
6
121150
121150
121151
121152
121152
121152
Home
Work
Home
Work
Home
Cell
+61(2)5555-1793
+61(2)5555-1794
+61(2)5555-1849
+61(2)5555-1850
+61(2)5555-1665
+61(2)5555-1666
Togar
Togar
Kylie
Birin
Birin
Birin
M
M
F
M
M
M
The results are the same as the one-to-many merge.

The order of variables is different.
98
98
Exercises
Level 1
4. Merging Two Sorted Data Sets in a One-to-Many Merge
b. Submit the two PROC CONTENTS steps or explore the data sets interactively to determine the
common variable among the two data sets.
c. Add a DATA step after the two PROC CONTENTS to merge orion.orders and orion.order_item
by the common variable to create a new data set, work.allorders. A sort is not required because
the data sets are already sorted by the common variable.
d. Submit the program and confirm that work.allorders was created with 732 observations
and 12 variables.
e. Add a statement to subset the variables. The new data set should contain six variables: Order_ID,
Order_Item_Num, Order_Type, Order_Date, Quantity, and Total_Retail_Price.
f. Write a PROC PRINT step to create the report below. Include only observations with a value for
Order_Date in the fourth quarter of 2011. The results should contain 35 observations.
10-48
Order_ID
1243515588
1243515588
1243568955
1243643970
1243644877
Order_
Type
1
1
1
1
3
Order_
Date
01OCT2011
01OCT2011
07OCT2011
16OCT2011
16OCT2011
Order_
Item_Num
1
2
1
1
1
Quantity
Total_Retail_
Price
1
1
1
1
1
$251.80
$114.20
$172.50
$101.50
$14.60
Level 2
5. Merging a Sorted Data Set and an Unsorted Data Set in a One-to-Many Merge
a. Sort orion.product_list by Product_Level to create a new data set, work.product_list.
b. Merge orion.product_level with the sorted data set. Create a new data set, work.listlevel, which
includes only Product_ID, Product_Name, Product_Level, and Product_Level_Name.
c. Create the report below, including only observations with Product Level equal to 3. The results
Product_
Level
3
3
3
3
3
Product_Level_
Name
Product
Product
Product
Product
Product
Category
Category
Category
Category
Category
Product_ID
210100000000
210200000000
220100000000
220200000000
230100000000
Product_Name
Children Outdoors
Children Sports
Clothes
Shoes
Outdoors
Challenge
6. Using the MERGENOBY Option
a. Use the SAS Help facility or online documentation to explore the MERGENOBY system option.
b. What is the purpose of this option and why is it used? __________________________________
_________________________________________________________________________________
c. Complete the following table to include the values that this option can assume:
Value
Description
Default (Y/N)
10.4 Merging Data Sets with Non-Matches
10.4 Merging Data Sets with NonMatches
Objectives
Control the observations in the output data set by

using the IN= option.
102
102
Business Scenario
An Orion Star manager in Australia requested an inventory
of company phone numbers.
103
103
10-49
10-50
Merge with Non-Matches

There are observations in empsau that do not have a
match in phonec, and some in phonec that do not match
any in empsau.
empsau
First
Togar
Kylie
Birin
phonec
Gender EmpID
M
121150
F
121151
M
121152
EmpID
Phone
121150 +61(2)5555-1795
121152 +61(2)5555-1667
121153 +61(2)5555-1348
104
104
Match-Merging
Merge empsau and phonec by EmpID to create
a new data set named empsauc.
data empsauc;
merge empsau phonec;
by EmpID;
run;
105
p110d07
105
Execution
empsau
First
Togar
Kylie
Birin
phonec
Gender EmpID
M
121150
F
121151
M
121152
EmpID
Phone
121150 +61(2)5555-1795
121152 +61(2)5555-1667
121153 +61(2)5555-1348
Initialize PDV
data empsauc;
by EmpID;
run;
PDV
First
Gender EmpID
.
Phone
...
106
106
Execution
empsau
First
Togar
Kylie
Birin
phonec
Gender EmpID
M
121150
F
121151
M
121152
EmpID
Phone
121150 +61(2)5555-1795
121152 +61(2)5555-1667
121153 +61(2)5555-1348
data empsauc;
by EmpID;
Yes
run;
PDV
First
Gender EmpID
.
Phone
107
107
...
10-51
10-52
Execution
empsau
First
Togar
Kylie
Birin
phonec
Gender EmpID
M
121150
F
121151
M
121152
EmpID
Phone
121150 +61(2)5555-1795
121152 +61(2)5555-1667
121153 +61(2)5555-1348
data empsauc;
by EmpID;
run;
PDV
First
Togar

into the PDV.
Gender EmpID
Phone
M
121150 +61(2)5555-1795
...
108
108
Execution
empsau
First
Togar
Kylie
Birin
phonec
Gender EmpID
M
121150
F
121151
M
121152
EmpID
Phone
121150 +61(2)5555-1795
121152 +61(2)5555-1667
121153 +61(2)5555-1348
data empsauc;
by EmpID;
run;
Implicit OUTPUT;
Implicit RETURN;
PDV
First
Togar
Gender EmpID
Phone
M
121150 +61(2)5555-1795
109
109
...
Execution
empsau
First
Togar
Kylie
Birin
phonec
Gender EmpID
M
121150
F
121151
M
121152
EmpID
Phone
121150 +61(2)5555-1795
121152 +61(2)5555-1667
121153 +61(2)5555-1348
data empsauc;
by EmpID;
No
run;
PDV
First
Togar
Gender EmpID
Phone
M
121150 +61(2)5555-1795
...
110
110
Execution
empsau
First
Togar
Kylie
Birin
phonec
Gender EmpID
M
121150
F
121151
M
121152
EmpID
Phone
121150 +61(2)5555-1795
121152 +61(2)5555-1667
121153 +61(2)5555-1348
data empsauc;
by EmpID;
run;
PDV
First
Togar
Did EmpID change?

Yes
Gender EmpID
Phone
M
121150 +61(2)5555-1795
111
111
...
10-53
10-54
Execution
empsau
First
Togar
Kylie
Birin
phonec
Gender EmpID
M
121150
F
121151
M
121152
EmpID
Phone
121150 +61(2)5555-1795
121152 +61(2)5555-1667
121153 +61(2)5555-1348
Reinitialize PDV
data empsauc;
by EmpID;
run;
PDV
First
Gender EmpID
.
Phone
...
112
112
Execution
empsau
First
Togar
Kylie
Birin
phonec
Gender EmpID
M
121150
F
121151
M
121152
EmpID
Phone
121150 +61(2)5555-1795
121152 +61(2)5555-1667
121153 +61(2)5555-1348
data empsauc;
by EmpID;
run;
PDV
First
Gender EmpID
.
Which EmpID value

sequentially comes first?
121151
Phone
113
113
...
Execution
empsau
First
Togar
Kylie
Birin
phonec
Gender EmpID
M
121150
F
121151
M
121152
EmpID
Phone
121150 +61(2)5555-1795
121152 +61(2)5555-1667
121153 +61(2)5555-1348
data empsauc;
by EmpID;
run;
PDV
First
Kylie
Gender EmpID
F
121151
Reads that observation

into the PDV.
Phone
...
114
114
Execution
empsau
First
Togar
Kylie
Birin
phonec
Gender EmpID
M
121150
F
121151
M
121152
EmpID
Phone
121150 +61(2)5555-1795
121152 +61(2)5555-1667
121153 +61(2)5555-1348
data empsauc;
by EmpID;
run;
Implicit OUTPUT;
Implicit RETURN;
PDV
First
Kylie
Gender EmpID
F
121151
Phone
115
115
...
10-55
10-56
Execution
empsau
First
Togar
Kylie
Birin
phonec
Gender EmpID
M
121150
F
121151
M
121152
EmpID
Phone
121150 +61(2)5555-1795
121152 +61(2)5555-1667
121153 +61(2)5555-1348
data empsauc;
by EmpID;
Yes
run;
PDV
First
Kylie
Gender EmpID
F
121151
Phone
...
116
116
Execution
empsau
First
Togar
Kylie
Birin
phonec
Gender EmpID
M
121150
F
121151
M
121152
EmpID
Phone
121150 +61(2)5555-1795
121152 +61(2)5555-1667
121153 +61(2)5555-1348
data empsauc;
by EmpID;
run;
PDV
First
Kylie
Gender EmpID
F
121151
Did EmpID change?

Yes
Phone
117
117
...
Execution
empsau
First
Togar
Kylie
Birin
phonec
Gender EmpID
M
121150
F
121151
M
121152
EmpID
Phone
121150 +61(2)5555-1795
121152 +61(2)5555-1667
121153 +61(2)5555-1348
Reinitialize PDV
data empsauc;
by EmpID;
run;
PDV
First
Gender EmpID
.
Phone
...
118
118
Execution
empsau
First
Togar
Kylie
Birin
phonec
Gender EmpID
M
121150
F
121151
M
121152
EmpID
Phone
121150 +61(2)5555-1795
121152 +61(2)5555-1667
121153 +61(2)5555-1348
data empsauc;
by EmpID;
run;
PDV
First
Birin

into the PDV.
Gender EmpID
Phone
M
121152 +61(2)5555-1667
119
119
...
10-57
10-58
Execution
empsau
First
Togar
Kylie
Birin
phonec
Gender EmpID
M
121150
F
121151
M
121152
EmpID
Phone
121150 +61(2)5555-1795
121152 +61(2)5555-1667
121153 +61(2)5555-1348
data empsauc;
by EmpID;
run;
Implicit OUTPUT;
Implicit RETURN;
PDV
First
Birin
Gender EmpID
Phone
M
121152 +61(2)5555-1667
...
120
120
Execution
empsau
First
Togar
Kylie
Birin
phonec
Gender EmpID
M
121150
F
121151
M
121152
EmpID
Phone
121150 +61(2)5555-1795
121152 +61(2)5555-1667
121153 +61(2)5555-1348
EOF
data empsauc;
by EmpID;
run;
PDV
First
Birin
Did EmpID change?

Yes
Gender EmpID
Phone
M
121152 +61(2)5555-1667
121
121
...
Execution
empsau
First
Togar
Kylie
Birin
phonec
Gender EmpID
M
121150
F
121151
M
121152
EOF
EmpID
Phone
121150 +61(2)5555-1795
121152 +61(2)5555-1667
121153 +61(2)5555-1348
Reinitialize PDV
data empsauc;
by EmpID;
run;
PDV
First
Gender EmpID
.
Phone
...
122
122
Execution
empsau
First
Togar
Kylie
Birin
phonec
Gender EmpID
M
121150
F
121151
M
121152
EmpID
Phone
121150 +61(2)5555-1795
121152 +61(2)5555-1667
121153 +61(2)5555-1348
EOF
data empsauc;
by EmpID;
run;
PDV
First
Reads the observation

into the PDV.
Gender EmpID
Phone
121153 +61(2)5555-1348
123
123
...
10-59
10-60
Execution
empsau
First
Togar
Kylie
Birin
phonec
Gender EmpID
M
121150
F
121151
M
121152
EmpID
Phone
121150 +61(2)5555-1795
121152 +61(2)5555-1667
121153 +61(2)5555-1348
EOF
data empsauc;
by EmpID;
run;
Implicit OUTPUT;
Implicit RETURN;
PDV
First
Gender EmpID
Phone
121153 +61(2)5555-1348
...
124
124
Execution
empsau
First
Togar
Kylie
Birin
EOF
phonec
Gender EmpID
M
121150
F
121151
M
121152
EmpID
Phone
121150 +61(2)5555-1795
121152 +61(2)5555-1667
121153 +61(2)5555-1348
EOF
data empsauc;
by EmpID;
run;
PDV
First
Gender EmpID
Phone
121153 +61(2)5555-1348
125
125
Final Results
empsauc
First
Togar
Kylie
Birin
Gender EmpID
Phone
M
121150 +61(2)5555-1795
F
121151
M
121152 +61(2)5555-1667
121153 +61(2)5555-1348
The final results include both matches and non-matches.
126
126
10.06 Quiz
Which data set contributed information to the last
observation in the output data set?
a.
b.
c.
d.
empsau
phonec
both empsau and phonec
There is insufficient information.
empsauc
First
Togar
Kylie
Birin
Gender EmpID
Phone
M
121150 +61(2)5555-1795
F
121151
M
121152 +61(2)5555-1667
121153 +61(2)5555-1348
127
127
10-61
10-62
Business Scenario
An Orion Star manager requested three phone inventory
reports.
employees with company phones

employees without company phones
phones with an invalid employee ID
130
130
IN= Data Set Option

The IN= data set option creates a variable that
indicates whether the data set contributed to
building the current observation.
MERGE SAS-data-set (IN=variable)
variable is a temporary numeric variable that has
two possible values:
0
Indicates that the data set did not contribute

to the current observation.
Indicates that the data set did contribute to

the current observation.
131
131
IN= Data Set Option

MERGE statement examples:
merge empsau(in=Emps)
phonec(in=Cell);
merge empsau(in=E)
phonec(in=P);
merge empsau(in=AU)
phonec;
132
132
Execution
empsau
First
Togar
Kylie
Birin
phonec
Gender EmpID
M
121150
F
121151
M
121152
EmpID
Phone
121150 +61(2)5555-1795
121152 +61(2)5555-1667
121153 +61(2)5555-1348
data empsauc;
phonec(in=Cell);
by EmpID;
run;
PDV
First
Togar
133
Gender EmpID
M
121150
match
Emps
Phone
1 +61(2)5555-1795
Cell
1
p110d08
133
...
10-63
10-64
Execution
empsau
First
Togar
Kylie
Birin
phonec
Gender EmpID
M
121150
F
121151
M
121152
EmpID
Phone
121150 +61(2)5555-1795
121152 +61(2)5555-1667
121153 +61(2)5555-1348
data empsauc;
phonec(in=Cell);
by EmpID;
run;
PDV
First
Kylie
Gender EmpID
F
121151
Emps
1
non-match
Phone
Cell
0
...
134
134
Execution
empsau
First
Togar
Kylie
Birin
phonec
Gender EmpID
M
121150
F
121151
M
121152
EmpID
Phone
121150 +61(2)5555-1795
121152 +61(2)5555-1667
121153 +61(2)5555-1348
data empsauc;
phonec(in=Cell);
by EmpID;
run;
PDV
First
Birin
Gender EmpID
M
121152
match
Emps
Phone
1 +61(2)5555-1667
Cell
1
135
135
...
10.07 Quiz
What are the values of Emps and Cell?
empsau
EOF
First
Togar
Kylie
Birin
phonec
Gender EmpID
M
121150
F
121151
M
121152
EmpID
Phone
121150 +61(2)5555-1795
121152 +61(2)5555-1667
121153 +61(2)5555-1348
data empsauc;
phonec(in=Cell);
by EmpID;
run;
PDV
First
136
Gender EmpID
121153
Emps
Phone
+61(2)5555-1348
Cell
Emps
Phone
1 +61(2)5555-1795
1
1 +61(2)5555-1667
0 +61(2)5555-1348
Cell
1
0
1
1
136
PDV Results
PDV
First
Togar
Kylie
Birin
Gender EmpID
M
121150
F
121151
M
121152
121153
The variables created with the IN= data set option are
only available during DATA step execution.
They are not written to the SAS data set.
Their value can be tested using conditional logic.
138
138
10-65
10-66
Matches Only
Add a subsetting IF statement to select the employees
that have company phones.
data empsauc;
phonec(in=Cell);
by EmpID;
if Emps=1 and Cell=1;
run;
empsauc
First
Togar
Birin
Gender EmpID
Phone
M
121150 +61(2)5555-1795
M
121152 +61(2)5555-1667
p110d08
139
139
Non-Matches from empsau

Select the employees that do not have company phones.
data empsauc;
phonec(in=Cell);
by EmpID;
run;
empsauc
First
Kylie
140
Gender EmpID
F
121151
Phone
p110d08
140
Non-Matches from phonec

Select the phones associated with an invalid employee ID.
data empsauc;
phonec(in=Cell);
by EmpID;
run;
empsauc
First
Gender EmpID
Phone
121153 +61(2)5555-1348
p110d08
141
141
All Non-Matches
data empsauc;
phonec(in=Cell);
by EmpID;
if Emps=0 or Cell=0;
run;
empsauc
First
Kylie
Gender EmpID
Phone
F
121151
121153 +61(2)5555-1348
Use the OR operator, not the AND operator.
142
p110d08
142
10-67
10-68
Alternate Syntax
When checking a variable for a value of 1 or 0 as in the
previous scenario, you can use the following syntax:
Instead of
You can use
if Emps and Cell;
if Emps and not Cell;
if not Emps and Cell;
if Emps=0 or Cell=0;
If not Emps or not Cell;
143
143
Alternate Syntax
Both programs create a report of employees without cell
phones.
data empsphone;
merge empsact(in=inEmps)
phoneact(in=inCell);
by EmpID;
if inEmps=1 and inCell=0;
run;
data empsphone;
merge empsact(in=inEmps)
phoneact(in=inCell);
by EmpID;
if inEmps and not inCell;
run;
144
p110d09
Business Scenario
Merge Orion Star customer information with customer
type data to obtain a customer description. The new data
set should include only US customers.
orion.
customer_type
orion.customer
customers
146
146
Considerations
The orion.customer data set is not sorted by
Customer_Type_ID, the common variable. The subsetting
variable, Country, is defined in only one data set.
orion.customer
Customer_
ID
Country
Customer_
Name
Birth_
Date
Customer_Type_
ID
orion.customer_type
Customer_
Group
Customer_Group_ Customer_
ID
Type
Customer_Type_
ID
147
147
10-69
10-70
10.08 Quiz
Open and submit p110a03. Correct the program and
resubmit. What change is needed to correct the error?
proc sort data=orion.customer
out=cust_by_type;
by Customer_Type_ID;
run;
data customers;
merge cust_by_type orion.customer_type;
where Country='US';
run;
148
Subsetting IF
Use a subsetting IF when the subsetting variable is not in
all data sets named in the MERGE statement.
out=cust_by_type;
run;
data customers;
if Country='US';
run;
150
p110a03s
150
Viewing the Output

Partial SAS Log
407 proc sort data=orion.customer
408
out=cust_by_type;
409
410 run;
NOTE: There were 77 observations read from the data set ORION.CUSTOMER.
NOTE: The data set WORK.CUST_BY_TYPE has 77 observations and 12 variables.
411
412
413
414
415
416
data customers;
if Country='US';
run;
NOTE: There were 77 observations read from the data set WORK.CUST_BY_TYPE.
NOTE: There were 8 observations read from the data set ORION.CUSTOMER_TYPE.
NOTE: The data set WORK.CUSTOMERS has 28 observations and 15 variables.
151
151

Step and Usage
PROC step
WHERE
IF
Yes
No
SET statement
Yes
Yes
No
Yes
INPUT statement
No
Yes
SET/MERGE statement (multiple data sets)

Variable in ALL data sets
Yes
Variable not in ALL data sets
No
Yes
Yes
152
152
Exercises
10-71
10-72
Level 1
7. Merging Using the IN= Option
a. Retrieve the starter program p110e07. Add a DATA step after the PROC SORT step to merge
work.product and orion.supplier by Supplier_ID to create a new data set, work.prodsup.
b. Submit the program and confirm that work.prodsup was created with 556 observations
and 10 variables.
c. Modify the DATA step to output only observations that are in work.product but
not orion.supplier. A subsetting IF statement that references IN= variables in the MERGE
statement must be added.
d. Submit the program and confirm that work.prodsup was created with 75 observations
and 10 variables.
e. Submit the PROC PRINT step to create the report below. The results should contain 75
observations.
Supplier information will be missing in the output.
Obs
Product_ID
1
2
3
4
5
210000000000
210100000000
210100100000
210200000000
210200100000
Product_Name
Supplier_ID
Children
Children Outdoors
Outdoor things, Kids
Children Sports
A-Team, Kids
Supplier_
Name
.
.
.
.
.
Level 2
8. Merging Using the IN= and RENAME= Options
a. Write a PROC SORT step to sort orion.customer by Country to create a new data set,
work.customer.
b. Write a DATA step to merge the resulting data set with orion.lookup_country
by Country to create a new data set, work.allcustomer.
In the orion.lookup_country data set, rename Start to Country and rename Label to
Country_Name.
Include only four variables: Customer_ID, Country, Customer_Name, and Country_Name.
c. Create the report below. The results should contain 308 observations.
Obs
1
2
...
306
Customer_ID
.
.
3959
Country
Customer_Name
AD
AE
ZA
Country_Name
Andorra
United Arab Emirates
Rita Lotz
South Africa
10.5 Solutions
307
308
.
.
ZM
ZW
10-73
Zambia
Zimbabwe
d. Modify the DATA step to store only the observations that contain both customer information
and country information. A subsetting IF statement that references IN= variables in the MERGE
statement must be added.
e. Submit the program to create the report below. The results should contain 77 observations.
Obs
Customer_ID
1
2
3
4
5
29
41
53
111
171
Country
AU
AU
AU
AU
AU
Customer_Name
Candy Kinsey
Wendell Summersby
Dericka Pockran
Karolina Dokter
Robert Bowerman
Country_
Name
Australia
Australia
Australia
Australia
Australia
Challenge
9. Merging and Outputting to Multiple Data Sets
a. Write a PROC SORT step to sort orion.orders by Employee_ID to create a new data set,
work.orders.
b. Write a DATA step to merge orion.staff and work.orders by Employee_ID and create two new
data sets: work.allorders and work.noorders.
work.allorders should include all observations from work.orders, regardless of matches or
non-matches from the orion.staff data set.
work.noorders should include only the observations from orion.staff that do not have a
match in work.orders.
Both new data sets should include only Employee_ID, Job_Title, Gender, Order_ID,
Order_Type, and Order_Date.
c. Submit the program and confirm that work.allorders was created with 490 observations
and 6 variables and work.noorders was created with 324 observations and 6 variables.
d. Create a detailed listing report for each new data set with an appropriate title.
10.5 Solutions
1. Concatenating Like-Structured Data Sets
How many observations in work.thirdqtr are from orion.mnth7_2011? 10 observations
data work.thirdqtr;
set orion.mnth7_2011 orion.mnth8_2011 orion.mnth9_2011;
10-74
run;
proc print data=work.thirdqtr;
run;
2. Concatenating Unlike-Structured Data Sets
What are the names of the two variables that are different in the two data sets?
orion.sales
orion.nonsales
First_Name
First
Last_Name
Last
proc contents data=orion.sales;

run;
proc contents data=orion.nonsales;
run;
data work.allemployees;
set orion.sales
orion.nonsales(rename=(First=First_Name Last=Last_Name));
keep Employee_ID First_Name Last_Name Job_Title Salary;
run;
proc print data=work.allemployees;
run;
3. Concatenating Data Sets with Variables of Different Lengths and Types
proc contents data=orion.charities;
run;
proc contents data=orion.us_suppliers;
run;
proc contents data=orion.consultants;
run;
data work.contacts;
set orion.charities orion.us_suppliers;
run;
proc contents data=work.contacts;
run;
data work.contacts2;
set orion.us_suppliers orion.charities;
run;
proc contents data=work.contacts2;
run;
data work.contacts3;
10.5 Solutions
set
run;
10-75
orion.us_suppliers orion.consultants;
c. Submit a PROC CONTENTS step to examine work.contacts. From which input data set were the
variable attributes assigned? the first data set in the set statement, orion.charities
e. Submit a PROC CONTENTS step to examine work.contacts2. From which input data set were
the variable attributes assigned? the first data set in the set statement, orion.us_suppliers
f. Write a DATA step to concatenate orion.us_suppliers and orion.consultants, creating a
temporary data set, contacts3.
Why did the DATA step fail? ContactType has been defined as both character and numeric.
4. Merging Two Sorted Data Sets in a One-to-Many Merge
proc contents data=orion.orders;
run;
proc contents data=orion.order_item;
run;
data work.allorders;
merge orion.orders
orion.order_item;
by Order_ID;
keep Order_ID Order_Item_Num Order_Type
Order_Date Quantity Total_Retail_Price;
run;
proc print data=work.allorders noobs;
where Order_Date between '01Oct2011'd and '31Dec2011'd;
run;
/* alternate solution */
where Order_Date>='01Oct2011'd and Order_Date<='31Dec2011'd;
run;
where qtr(Order_Date)=4 and year(Order_Date)=2011;
run;
5. Merging a Sorted Data Set and an Unsorted Data Set in a One-to-Many Merge
proc sort data=orion.product_list
out=work.product_list;
by Product_Level;
run;
data work.listlevel;
merge orion.product_level work.product_list ;
by Product_Level;
keep Product_ID Product_Name Product_Level Product_Level_Name;
10-76
run;
proc print data=work.listlevel noobs;
where Product_Level=3;
run;
6. Using the MERGENOBY Option
b. What is the purpose of this option and why is it used? This option is used to issue a warning or
an error when a BY statement is omitted from a merge. Performing a merge without a BY
statement merges the observations based on their position. This is almost never done
intentionally and can lead to unexpected results.
c. Complete the following table to include the values that this option can assume.
Value
Description
Default (Y/N)
NOWARN
Performs the positional merge without warning.
WARN
Performs the positional merge but writes a warning message to the

log.
ERROR
Writes an error message to the log and the DATA step terminates.
7. Merging Using the IN= Option

proc sort data=orion.product_list
out=work.product;
by Supplier_ID;
run;
data work.prodsup;
merge work.product(in=P)
orion.supplier(in=S);
by Supplier_ID;
if P=1 and S=0;
run;
proc print data=work.prodsup;
var Product_ID Product_Name Supplier_ID Supplier_Name;
run;
8. Merging Using the IN= and RENAME= Options
out=work.customer;
by Country;
run;
data work.allcustomer;
merge work.customer(in=Cust)
orion.lookup_country(rename=(Start=Country
Label=Country_Name) in=Ctry);
by Country;
keep Customer_ID Country Customer_Name Country_Name;
10.5 Solutions
10-77
if Cust=1 and Ctry=1;

run;
proc print data=work.allcustomer;
run;
9. Merging and Outputting to Multiple Data Sets
proc sort data=orion.orders
out=work.orders;
by Employee_ID;
run;
data work.allorders work.noorders;
merge orion.staff(in=Staff) work.orders(in=Ord);
by Employee_ID;
if Ord=1 then output work.allorders;
else if Staff=1 and Ord=0 then output work.noorders;
/* alternate statement */
/* else output work.noorders; */
keep Employee_ID Job_Title Gender Order_ID Order_Type Order_Date;
run;
proc print data=work.allorders;
run;
proc print data=work.noorders;
run;

How many variables will be in empsall2 after
concatenating empscn and empsjp?
empscn
First
Chang
Li
Ming
empsjp
Gender
M
M
F
Country
China
China
China
First
Cho
Tomi
Gender Region
F
Japan
M
Japan
Four variables: First, Gender, Country, and Region
30
30
10-78

Which statement has correct syntax?
a. Answer
set empscn(rename(Country=Location))
empsjp(rename(Region=Location));
b. Answer
set empscn(rename=(Country=Location))
empsjp(rename=(Region=Location));
c.
set empscn rename=(Country=Location)

empsjp rename=(Region=Location);
38
38

Which of the following BY statements correctly sorts by
descending salary within gender?
a.
b.
c.
d.
by descending salary within gender;

by descending salary gender;
by gender descending salary;
by gender salary descending;
The keyword descending is placed before the variable

to which it applies.
52
10.5 Solutions

What are the modified, completed statements?
proc sort data=orion.employee_payroll
out=work.payroll;
by Employee_ID;
run;
proc sort data=orion.employee_addresses
out=work.addresses;
by Employee_ID;
run;
data work.payadd;
merge work.payroll work.addresses;
by Employee_ID;
run;
p110a01s
59
59

In a one-to-many merge, does it matter which data set is
listed first in the MERGE statement?
Open p110a02 and submit
Reverse the order of the data sets and submit again.
Observe the results. How are they different?
The results are the same, but the order of the
variables is different.
96
p110a02s
10-79
10-80

Which data set contributed information to the last
observation in the output data set?
a.
b.
c.
d.
empsau
phonec
both empsau and phonec
There is insufficient information.
empsauc
First
Togar
Kylie
Birin
Gender EmpID
Phone
M
121150 +61(2)5555-1795
F
121151
M
121152 +61(2)5555-1667
121153 +61(2)5555-1348
128
128

What are the values of Emps and Cell?
empsau
EOF
First
Togar
Kylie
Birin
phonec
Gender EmpID
M
121150
F
121151
M
121152
EmpID
Phone
121150 +61(2)5555-1795
121152 +61(2)5555-1667
121153 +61(2)5555-1348
data empsauc;
phonec(in=Cell);
by EmpID;
run;
PDV
First
137
Gender EmpID
121153
non-match
Emps
Phone
0 +61(2)5555-1348
Cell
1
137
10.5 Solutions

What change is needed to correct the error?
397 proc sort data=orion.customer
398
out=cust_by_type;
399
400 run;
NOTE: There were 77 observations read from the data set ORION.CUSTOMER.
NOTE: The data set WORK.CUST_BY_TYPE has 77 observations and 12 variables.
401
402 data customers;
403
404
405
where Country='US';
ERROR: Variable Country is not on file ORION.CUSTOMER_TYPE.
406 run;
WARNING: The data set WORK.CUSTOMERS may be incomplete. When this step was
stopped there were 0 observations and 15 variables.
Country is not defined in both data sets. Replace the

WHERE statement with a subsetting IF.
149
10-81
10-82
Chapter 11 Creating Summary

Reports
11.1 The FREQ Procedure ................................................................................................... 11-3
Exercises ............................................................................................................................ 11-24
11.2 The MEANS and UNIVARIATE Procedures............................................................... 11-28

Exercises ............................................................................................................................ 11-40
11.3 Using the Output Delivery System ............................................................................ 11-44

Demonstration: Using the Output Delivery System ........................................................... 11-55
Exercises ............................................................................................................................ 11-62
11.4 Solutions ..................................................................................................................... 11-65

Solutions to Exercises ........................................................................................................ 11-65
Solutions to Student Activities (Polls/Quizzes) ................................................................... 11-69
11-2
Chapter 11 Creating Summary Reports
11.1 The FREQ Procedure

Objectives
Produce one-way and two-way frequency tables with

the FREQ procedure.
Enhance frequency tables with options.
Use PROC FREQ to validate data in a SAS data set.
Business Scenario
Orion Star management wants to know the number of
male and female sales employees in Australia.
11-3
11-4
Considerations
Use the FREQ procedure to analyze the Gender variable
in a subset of orion.sales.
The FREQ Procedure
Gender
Frequency
Percent
F
XX
XX.XX
M
XX
XX.XX
FREQ Procedure
The FREQ procedure produces a one-way frequency
table for each variable named in the TABLES statement.
proc freq data=orion.sales;

tables Gender;
where Country='AU';
run;
PROC FREQ DATA=SAS-data-set;

<TABLES variable(s) </ options>> ;
RUN;
If the TABLES statement is omitted, a one-way

frequency table is produced for every variable in the
data set. This can produce voluminous output and is
seldom desired.
6
p111d01
PROC FREQ displays a report by default. The output can be saved in a SAS data set.
The procedure can compute the following:
chi-square tests for one-way to n-way tables
tests and measures of agreement for contingency tables
tests and measures of association for contingency tables and more
11-5
Viewing the Output

A one-way frequency table was created for Gender. It lists
the discrete values found in the data set and the number
of observations in which the variable has that value.
The FREQ Procedure
Cumulative
Cumulative
Gender
Frequency
Percent
Frequency
Percent
F
27
42.86
27
42.86
M
36
57.14
63
100.00
The default output includes frequency and percentage

values, including cumulative statistics.
Options to Suppress Statistics

Use options in the TABLES statement to suppress the
display of selected default statistics.
TABLES variable(s) / options ;
Option
Description
NOCUM
Suppresses the cumulative statistics.
NOPERCENT
Suppresses the percentage display.
Cumulative frequencies and percentages are useful when there are at least three levels of a variable in an
ordinal relationship. When this is not the case, the NOCUM option produces a simpler, less confusing
report.
11-6

NOCUM
suppresses
The FREQ Procedure
Cumulative
Cumulative
Gender
Frequency
Percent
Frequency
Percent
F
27
42.86
27
42.86
M
36
57.14
63
100.00
NOPERCENT
suppresses
11.01 Quiz
Open and submit p111a01. Review the log to determine
the cause of the error. Correct the program and resubmit.
What change was needed?

tables country nocum nopercent;
run;
10
Idea Exchange
This step creates a table for every variable in the data set:
run;
Employee_ID
First_Name
Last_Name
Gender
Salary
Job_Title
Country
Birth_Date
Hire_Date
Which variables are most

appropriate for a frequency
analysis? Why?
12
Business Scenario
Orion Star management wants to know how many sales
employees are in each country, as well as the count of
males and females.
14
11-7
11-8
TABLES Statement
You can list multiple variables in a TABLES statement. A
separate table is produced for each variable.
tables Gender Country;
run;
PROC FREQ Output
The FREQ Procedure
Cumulative
Cumulative
Gender
Frequency
Percent
Frequency
Percent
F
68
41.21
68
41.21
M
97
58.79
165
100.00
Cumulative
Cumulative
Country
Frequency
Percent
Frequency
Percent
AU
63
38.18
63
38.18
US
102
61.82
165
100.00
p111d02
15
BY Statement
The BY statement is used to request separate analyses
for each BY group.
proc sort data=orion.sales out=sorted;
by Country;
run;
proc freq data=sorted;
tables Gender;
by Country;
run;
The data set must be sorted or indexed by the variable (or
variables) named in the BY statement.
16
p111d02
Viewing the Output

Each group displays on a separate page with a BY line.
------------------------- Country=AU -------------------------The FREQ Procedure
Cumulative
Cumulative
Gender
Frequency
Percent
Frequency
Percent
F
27
42.86
27
42.86
M
36
57.14
63
100.00
-------------------------- Country=US -----------------------The FREQ Procedure
Cumulative
Cumulative
Gender
Frequency
Percent
Frequency
Percent
F
41
40.20
41
40.20
M
61
59.80
102
100.00
17
Crosstabulation Table
An asterisk between two variables generates a two-way
frequency table, or crosstabulation table.
tables Gender*Country;
run;
rows
columns
A two-way frequency table generates a single table with

statistics for each distinct combination of values of the
selected variables.
18
p111d02
11-9
11-10
Viewing the Output

PROC FREQ Output
The FREQ Procedure
Table of Gender by Country
Gender
Country
Frequency
Percent
Row Pct
Col Pct AU
US
Total
27
41
68
16.36 24.85 41.21
39.71 60.29
42.86 40.20
36
61
97
21.82 36.97 58.79
37.11 62.89
57.14 59.80
Total
63
102
165
38.18
61.82
100.00
19

Use options in the TABLES statement to suppress the
display of selected default statistics.
TABLES variable(s) / options ;
Option
Description
NOROW
Suppresses the display of the row percentage.
NOCOL
Suppresses the display of the column percentage.
NOPERCENT
NOFREQ
Suppresses the percentage display.

Suppresses the frequency display.
20

Gender
Country
Frequency
Percent
Row Pct
Col Pct AU
US
Total
27
41
68
16.36 24.85 41.21
39.71 60.29
42.86 40.20
36
61
97
21.82 36.97 58.79
37.11 62.89
57.14 59.80
Total
63
102
165
38.18
61.82
100.00
21

Gender
Country
Frequency
Percent
Row Pct
Col Pct AU
US
Total
27
41
68
16.36 24.85 41.21
39.71 60.29
42.86 40.20
36
61
97
21.82 36.97 58.79
37.11 62.89
57.14 59.80
Total
63
102
165
38.18
61.82
100.00
22
11-11
11-12
LIST and CROSSLIST Options

You can use the LIST and CROSSLIST options in the
TABLES statement to flatten the output.
Cumulative
Cumulative
Gender
Country
Frequency
Percent
Frequency
Percent
F
AU
27
16.36
27
16.36
F
US
41
24.85
68
41.21
M
AU
36
21.82
104
63.03
M
US
61
36.97
165
100.00
LIST
option
Row
Column
Gender
Country
Frequency
Percent
Percent
Percent
F
AU
27
16.36
39.71
42.86
US
41
24.85
60.29
40.20
Total
68
41.21
100.00
---------------------------------------------------------------M
AU
36
21.82
37.11
57.14
US
61
36.97
62.89
59.80
Total
97
58.79
100.00
---------------------------------------------------------------Total
AU
63
38.18
100.00
US
102
61.82
100.00
Total
165
100.00
23
Business Scenario
A new data set, orion.nonsales2, must be validated. It
contains information on non-sales employees and might
include invalid and missing values.
Partial orion.nonsales2
Employee_
First
ID
120101 Patrick
Last
Gender
Salary
Job_Title
163040 Director
Country
Lu
AU
120104 Kareen
Billington
46230 Admin Mgr
au
120105 Liz
Povey
27110 Secretary I
AU
120106 John
Hornsey
. Office Asst II
120107 Sherie
Sheedy
30475 Office Asst II
AU
120108 Gladys
Gromek
27660 Warehouse Asst II
AU
AU
25
Considerations
Use the FREQ procedure to screen for invalid, missing,
and duplicate data values.
Requirements of non-sales employee data:
Employee_ID values must be unique and not missing.
Gender must be F or M.
Job_Title must not be missing.
Country must have a value of AU or US.
Salary values must be in the numeric range of
24000 to 500000.
26
11.02 Quiz
What problems exist with the data in this partial data set?
Hint: There are seven data problems.

27
11-13
11-14
FREQ Procedure for Data Validation

The FREQ procedure lists all discrete values for a
variable and reports missing values.
proc freq data=orion.nonsales2;
tables Gender Country / nocum nopercent;
run;
p111d03
29
Viewing the Output

PROC FREQ Output
The FREQ Procedure
Gender
Frequency
F
110
G
1
M
123
Frequency Missing = 1
Country
Frequency
AU
33
US
196
au
3
us
3
30
NLEVELS Option
The NLEVELS option displays a table that provides the
number of distinct values for each analysis variable.
proc freq data=orion.nonsales2 nlevels;
tables Gender Country / nocum nopercent;
run;
PROC FREQ DATA=SAS-data-set NLEVELS;
TABLES variable(s) ;
RUN;
p111d03
31
Viewing the Output

PROC FREQ Output
The FREQ Procedure
Number of Variable Levels
Missing
Nonmissing
Variable
Levels
Levels
Levels
Gender
4
1
3
Country
4
0
4
Gender
Frequency
F
110
G
1
M
123
Country
Frequency
AU
33
US
196
au
3
us
3
32
11-15
11-16
Check for Uniqueness

The values of Employee_ID must be unique and not
missing. PROC FREQ can be used to check for duplicate
or missing values.
proc freq data=orion.nonsales2 order=freq;
tables Employee_ID / nocum nopercent;
run;
The ORDER=FREQ option displays the results in
descending frequency order.
p111d04
33
Viewing the Output

Partial PROC FREQ Output
The FREQ Procedure
Employee_ID
Frequency
120108
2
120101
1
120104
1
120105
1
120106
1
121134
121141
121142
121146
121147
121148
1
1
1
1
1
1
34
p111d04
NLEVELS Option
NLEVELS can also be used to identify duplicates, when
the number of distinct values is known.
tables Employee_ID / noprint;
run;
This example uses the NOPRINT option to suppress the
frequency table. Only the Number of Variable Levels table
is displayed.
p111d04
35
Viewing the Output

The FREQ Procedure
Missing
Nonmissing
Variable
Levels
Levels
Levels
Employee_ID
234
1
233
There are 235 employees, but there are only 234 distinct
Employee_ID values. Therefore, there is one duplicate
value and one missing value for Employee_ID.
36
11-17
11-18
NLEVELS Option
The _ALL_ keyword with the NOPRINT option displays
the number of levels for all variables without displaying
frequency counts.
tables _all_ / noprint;
run;
p111d04
37
Viewing the Output

PROC FREQ Output
The FREQ Procedure
Missing
Nonmissing
Variable
Levels
Levels
Levels
Employee_ID
234
1
233
First
204
0
204
Last
228
0
228
Gender
4
1
3
Salary
230
1
229
Job_Title
125
1
124
Country
4
0
4
No frequency tables were displayed.
38
11.03 Quiz
Modify p111a02 to analyze Job_Title. Display the
NLEVELS table listing the frequency counts in decreasing
order.
How many unique, nonmissing job titles exist?
Which job title occurs most frequently?
What is the frequency of missing job titles?
39
Identifying Observations with Invalid Data

PROC FREQ has uncovered the existence of invalid data
values for Gender, Country, and Employee_ID. Use
PROC PRINT to display the observations with invalid
values.
proc print data=orion.nonsales2;
where Gender not in ('F','M') or
Country not in ('AU','US') or
Job_Title is null or
Employee_ID is missing or
Employee_ID=120108;
run;
41
11-19
11-20
Viewing the Output

PROC PRINT Output
Obs
Employee_ID
2
6
7
10
12
14
84
87
101
125
197
200
120104
120108
120108
120112
120114
.
120695
120698
120723
120747
120994
120997
First
Last
Gender
Kareen
Gladys
Gabriele
Ellis
Jeannette
Austen
Trent
Geoff
Deanna
Zashia
Danelle
Mary
Billington
Gromek
Baker
Glattback
Buddery
Ralston
Moffat
Kistanna
Olsen
Farthing
Sergeant
Donathan
F
F
F
F
G
M
M
M
F
F
F
Salary
46230
27660
26495
26550
31285
29250
28180
26160
33950
43590
31645
27420
Job_Title
Administration Manager
Security Manager
Service Assistant II
Corp. Comm. Specialist II
Financial Controller I
Office Administrator I
Shipping Administrator I
Country
au
AU
AU
AU
AU
AU
au
au
US
us
us
us
original
observation
numbers
42
Business Scenario
The manager of Human Resources has requested
a report showing the number and percent of sales
employees hired each year.
44
Using Formats in PROC FREQ

A FORMAT statement can be used in PROC FREQ to
format data values.
tables Hire_Date / nocum;
format Hire_Date date9.;
run;
The FREQ Procedure
Hire_Date
Frequency
Percent
01JAN1978
17
10.30
01FEB1978
2
1.21
01APR1978
1
0.61
01JUL1978
1
0.61
01AUG1978
1
0.61
many discrete values,

and not what the
manager requested
45
p111d05
Using Formats in PROC FREQ

A FORMAT statement can also be used in PROC FREQ
to group the data.
tables Hire_Date / nocum;
format Hire_Date year4.;
run;
The FREQ Procedure
Hire_Date
Frequency
Percent
1978
23
13.94
1979
2
1.21
1980
4
2.42
1981
3
1.82
1982
7
4.24
46
fewer discrete values
p111d05
If a format is permanently assigned to a variable, PROC FREQ automatically groups

the report by the formatted values.
11-21
11-22
11.04 Quiz
Open and submit p111a03 and view the output. Add a
statement to apply the TIERS format to Salary and
resubmit.
Can user-defined formats be used to group data?
47
FORMAT Statement
User-defined formats can also be used to display levels
with alternate text in a frequency table.
tables Gender*Country;
format Country $ctryfmt.
Gender $gender.;
run;
49
p111d06
11-23
Viewing the Output

Gender
Country
Labels are
Frequency
wrapped.
Percent
Row Pct
Col Pct AustraliUnited S Total
a
tates
Female
27
41
68
16.36 24.85 41.21
39.71 60.29
42.86 40.20
Male
36
61
97
21.82 36.97 58.79
37.11 62.89
57.14 59.80
Total
63
102
165
38.18
61.82
100.00
50
The default format for cell statistics is 8.d, where d is 0 for frequencies and 2 for percentages. Eight print
positions are sometimes not enough for column headings.
FORMAT= Option
Use the FORMAT= option in the TABLES statement to
format the frequency value and to change the width of the
column.
tables Gender*Country / format=12.;
format Country $ctryfmt.
Gender $gender.;
run;
51
p111d06
11-24
Viewing the Output

PROC FREQ Output
The FREQ Procedure
Gender
Country
Columns are 12
characters wide.
Frequency
Percent
Row Pct
Col Pct Australia
United States
Total
Female
27
41
68
16.36
24.85
41.21
39.71
60.29
42.86
40.20
Male
36
61
97
21.82
36.97
58.79
37.11
62.89
57.14
59.80
Total
63
102
165
38.18
61.82
100.00
52
The FORMAT= option applies only to the frequency. Percentage values are always displayed with two
decimal places.
Exercises
Level 1
1. Counting Levels of a Variable with PROC FREQ
b. Submit the program without making changes to analyze Customer_ID and Employee_ID in
orion.orders. Would you expect to see frequencies of 1 for customers and sales employees? ____
c. Modify the program to produce two separate reports.
1) Display the number of distinct levels of Customer_ID and Employee_ID for retail orders.
a) Use a WHERE statement to limit the report to retail sales (Order_Type=1).
b) Do not display the frequency count tables.
c) Display the title Unique Customers and Salespersons for Retail Sales.
d) Submit the program to produce the following report:
PROC FREQ Output

Unique Customers and Salespersons for Retail Sales
The FREQ Procedure
Variable
Label
Levels
Customer_ID
Customer ID
31
Employee_ID
Employee ID
100
2) Display the number of distinct levels for Customer_ID for catalog and Internet orders.
a) Use a WHERE statement to limit the report to catalog and Internet sales by selecting
observations with Order_Type values other than 1.
b) Specify an option to display the results in decreasing frequency order.
c) Specify an option to suppress the cumulative statistics.
d) Display the title Catalog and Internet Customers.
e) Submit the program to produce the following report:
Catalog and Internet Customers
The FREQ Procedure
Customer ID
Customer_ID
Frequency
Percent
16
15
6.52
29
9
3.91
5
8
3.48
...
26148
1
0.43
70059
1
0.43
2. Validating orion.shoes_tracker with PROC FREQ

b. Complete the PROC FREQ step to create one-way frequency tables for Supplier_Name and
Supplier_ID in orion.shoes_tracker. Include the NLEVELS option.
The data in orion.shoes_tracker should meet the following requirements:
Supplier_Name must be 3Top Sports or Greenline Sports Ltd.
Supplier_ID must be 2963 or 14682.
What invalid data exists for Supplier_Name and Supplier_ID?
11-25
11-26
Level 2
3. Producing Frequency Reports with PROC FREQ
b. Add statements to the PROC FREQ step to produce three frequency reports.
1) Number of orders in each year: Apply a format to the Order_Date variable to combine all
orders within the same year.
2) Number of orders of each order type: Apply the ORDERTYPES. format defined in the starter
program to the Order_Type variable. Suppress the cumulative frequency and percentages.
3) Number of orders for each combination of year and order type: Suppress all percentages that
normally appear in each cell of a two-way table.
c. Submit the program to produce the following output:
PROC FREQ Output
Order Summary by Year and Type
The FREQ Procedure
Date Order was placed by Customer
Cumulative
Cumulative
Order_Date
Frequency
Percent
Frequency
Percent
2007
104
21.22
104
21.22
2008
87
17.76
191
38.98
2009
70
14.29
261
53.27
2010
113
23.06
374
76.33
2011
116
23.67
490
100.00
Order Type
Order_
Type
Frequency
Percent
Retail
260
53.06
Catalog
132
26.94
Internet
98
20.00
Table of Order_Date by Order_Type

Order_Date(Date Order was placed by Customer)
Order_Type(Order Type)
FrequencyRetail Catalog Internet
2007
45
41
18
2008
51
20
16
Total
104
87
2009
27
23
20
2010
67
33
13
2011
70
15
31
Total
260
132
98
11-27
70
113
116
490
4. Validating orion.qtr2_2011 with PROC FREQ

Write a PROC FREQ step to validate the data in orion.qtr2_2011.
a. Create frequency tables for Order_ID and Order_Type. Include the Number of Variable Levels
table.
b. Submit the program.
The data in orion.qtr2_2011 should meet the following requirements:
Order_ID must be unique (36 distinct values) and not missing.
Order_Type must have a value of 1, 2, or 3.
What invalid data exists for Order_ID and Order_Type?
Challenge
5. Creating an Output Data Set with PROC FREQ
Write a program to perform a frequency analysis on Product_ID in orion.order_fact.
a. Create an output data set containing the frequency counts based on Product_ID. Explore the SAS
Help facility or online documentation for information about creating an output data set of counts
from PROC FREQ results.
b. Combine the output data set with orion.product_list to obtain the Product_Name value for each
Product_ID code. Output only products that have been ordered.
c. Sort the combined data so that the most frequently ordered products appear first in the resulting
data set. Print the first five observationsthat is, those that represent the five products ordered
most often. Use the OBS= data set option to limit the number of observations displayed.
d. Submit the program to produce the following report:
PROC PRINT Output
Top Five Products by Number of Orders
Orders
6
6
5
5
4
Product
Number
230100500056
230100600030
230100600022
240400300035
230100500082
Product
Knife
Outback Sleeping Bag, Large,Left,Blue/Black
Expedition10,Medium,Right,Blue Ribbon
Smasher Shorts
Lucky Tech Intergal Wp/B Rain Pants
11-28
11.2 The MEANS and UNIVARIATE

Procedures
Objectives
Calculate summary statistics and multilevel

summaries with the MEANS procedure.
Enhance summary tables with options.
Identify extreme and missing values with the
UNIVARIATE procedure.
56
Business Scenario
The payroll manager would like to see the average salary
for all employees.
57
11.2 The MEANS and UNIVARIATE Procedures
MEANS Procedure
The MEANS procedure produces summary reports with
descriptive statistics.
proc means data=orion.sales;
run;
PROC MEANS DATA=input-data-set <options statistics>;

<VAR analysis-variable(s);>
<CLASS classification-variable(s);>
RUN;
Analysis variables are the numeric variables for which

statistics are to be computed.
Classification variables are variables whose values
define subgroups for the analysis. They can be
character or numeric.
p111d07
58
Viewing the Output

PROC MEANS Output
The MEANS Procedure
Variable
N
Mean
Std Dev
Minimum
Maximum
Employee_ID
165
120713.90
450.0866939
120102.00
121145.00
Salary
165
31160.12
20082.67
22710.00
243190.00
Birth_Date
165
3622.58
5456.29
-5842.00
10490.00
Hire_Date
165
12054.28
4619.94
5114.00
17167.00
Default statistics are displayed for all numeric variables.
59
11-29
11-30
VAR Statement
The VAR statement identifies the analysis variable
(or variables) and their order in the output.
var Salary;
run;
VAR variable(s);
The MEANS Procedure

Analysis Variable : Salary
N
Mean
Std Dev
Minimum
Maximum
165
31160.12
20082.67
22710.00
243190.00
p111d07
60
Business Scenario
Analyze Salary by Country within Gender.
female
male
61
CLASS Statement
The CLASS statement identifies variables whose values
define subgroups for the analysis.
var Salary;
class Gender Country;
run;
CLASS classification-variable(s);
Classification variables are character or numeric.

They typically have few discrete values.
The data set does not need to be sorted or indexed by
the classification variables.
p111d08
62
Viewing the Output

Statistics are produced for each combination of values of
the classification variables.
The MEANS Procedure
N
Gender Country
Obs
N
Mean
Std Dev
Minimum
Maximum
F
AU
27
27
27702.41
1728.23
25185.00
30890.00
63
US
41
41
29460.98
8847.03
25390.00
83505.00
AU
36
36
32001.39
16592.45
25745.00
108255.00
US
61
61
33336.15
29592.69
22710.00
243190.00
N Obs the number of observations with each unique

combination of class variables
N the number of observations with nonmissing values
of the analysis variable (or variables)
11-31
11-32
11.05 Quiz
For a given data set, there are 63 observations with
a Country value of AU. Of those 63 observations,
only 61 observations have a value for Salary.
Which output is correct?
a.

N
Country
Obs
N
AU
63
61
b.

N
Country
Obs
N
AU
61
63
64
Business Scenario
Analyze Salary by Country within Gender. Generate a
report that includes the number of missing Salary values,
as well as the minimum, maximum, and sum of salaries.
female
male
66
PROC MEANS Statistics

Use options in the PROC MEANS statement to request
specific statistics.
proc means data=orion.sales nmiss min max sum;
var Salary;
class Gender Country;
run;
The requested statistics override the default statistics.
p111d09
67
PROC MEANS Statistics

The statistics are displayed in the order in which they are
requested.
The MEANS Procedure
N
N
Gender
Country
Obs
Miss
Minimum
Maximum
Sum
F
AU
27
0
25185.00
30890.00
747965.00
US
41
25390.00
83505.00
1207900.00
AU
36
25745.00
108255.00
1152050.00
US
61
0
22710.00
243190.00
2033505.00
68
11-33
11-34
PROC MEANS Statement Options

Options can also be placed in the PROC MEANS
statement.
Option
Description
MAXDEC=
Specifies the number of decimal places to display.
NONOBS
Suppresses the N Obs column.
69
MAXDEC= Option
MAXDEC=0
The MEANS Procedure
N
Country
Obs
N
Mean
Std Dev
Minimum
Maximum
AU
63
63
30159
12699
25185
108255
US
102
102
31778
23556
22710
243190
MAXDEC=1
The MEANS Procedure
N
Country
Obs
N
Mean
Std Dev
Minimum
Maximum
AU
63
63
30159.0
12699.1
25185.0
108255.0
US
70
102
102
31778.5
23555.8
22710.0
243190.0
p111d10
NONOBS Option
N Obs included by default
The MEANS Procedure
N
Country
Obs
N
Mean
Std Dev
Minimum
Maximum
AU
63
63
30158.97
12699.14
25185.00
108255.00
US
102
102
31778.48
23555.84
22710.00
243190.00
NONOBS option
The MEANS Procedure
Country
N
Mean
Std Dev
Minimum
Maximum
AU
63
30158.97
12699.14
25185.00
108255.00
US
102
31778.48
23555.84
22710.00
243190.00
p111d10
71
Other PROC MEANS Statistics

Descriptive Statistic Keywords
CLM
CSS
CV
LCLM
MAX
MEAN
MIN
MODE
NMISS
KURTOSIS
RANGE
SKEWNESS
STDDEV
STDERR
SUM
SUMWGT
UCLM
USS
VAR
Quantile Statistic Keywords

MEDIAN |
P50
P1
P5
P10
Q1 | P25
Q3 | P75
P90
P95
P99
QRANGE
Hypothesis Testing Keywords

PROBT
72
11-35
11-36
Idea Exchange
Which PROC MEANS statistics would you request when
validating numeric variables?
73
Business Scenario
Validate salary data in orion.nonsales2. Salary must be
in the numeric range of 24000 to 500000.
Partial orion.nonsales2
Employee
First
_ID
120101 Patrick
Last
Gender
Salary
Job_Title
163040 Director
Country
Lu
120104 Kareen
Billington
46230 Admin Mgr
au
AU
120105 Liz
Povey
27110 Secretary I
AU
120106 John
Hornsey
. Office Asst II
120107 Sherie
Sheedy
30475 Office Asst II
AU
120108 Gladys
Gromek
27660 Warehouse Asst II
AU
AU
75
UNIVARIATE Procedure
PROC UNIVARIATE displays extreme observations,
missing values, and other statistics for the variables
included in the VAR statement.
proc univariate data=orion.nonsales2;
var Salary;
run;
PROC UNIVARIATE DATA=SAS-data-set;
<VAR variable(s);>
RUN;
If the VAR statement is omitted, PROC UNIVARIATE

analyzes all numeric variables in the data set.
p111d11
76
Viewing the Output: Extreme Observations

The Extreme Observations section includes the five
lowest and five highest values for the analysis variable
and the corresponding observation numbers.
Partial PROC UNIVARIATE Output
Extreme Observations
-----Lowest----
77
-----Highest----
Value
Obs
Value
Obs
2401
2650
24025
24100
24390
20
13
25
19
228
163040
194885
207885
268455
433800
1
231
28
29
27
Obs is the observation number, not the count of

observations with that value.
11-37
11-38
NEXTROBS= Option
The NEXTROBS= option specifies the number of extreme
observations to display.
proc univariate data=orion.nonsales2
nextrobs=3;
var Salary;
run;
The UNIVARIATE Procedure
Variable: Salary
-----Lowest----
-----Highest----
Value
Obs
Value
Obs
2401
2650
24025
20
13
25
207885
268455
433800
28
29
27
78
p111d11
The default value for NEXTROBS= is 5, and n can range between 0 and half the maximum number of
observations. You can specify NEXTROBS=0 to suppress the table of extreme observations.
ID Statement
The ID statement displays the value of the identifying
variable (or variables) in addition to the observation
number.
proc univariate data=orion.nonsales2;
var Salary;
id Employee_ID;
run;
ID variable(s);
79
p111d11
Viewing the Output

Variable: Salary
------------Lowest----------
------------Highest----------
Value
Employee_ID
Obs
Value
Employee_ID
Obs
2401
2650
24025
24100
24390
120191
120115
120196
120190
121132
20
13
25
19
228
163040
194885
207885
268455
433800
120101
121141
120260
120262
120259
1
231
28
29
27
80
Viewing the Output: Missing Values Section

The Missing Values section displays the number and
percentage of observations with missing values of the
analysis variable.
Missing Values
Missing
Value
Count
-----Percent Of----Missing
All Obs
Obs
0.43
100.00
81
11-39
11-40
11.06 Quiz
PROC UNIVARIATE identified two observations with
Salary values less than 24,000.
What procedure can be used to display the observations
containing the invalid values?
82
Exercises
Level 1
6. Creating a Summary Report with PROC MEANS
b. Display only the SUM statistic for the Total_Retail_Price variable.
c. Add a CLASS statement to display separate statistics for each combination of Order_Date and
Order_Type.
d. Apply the ORDERTYPES format so that the order types are displayed as text descriptions.
Apply the YEAR4. format so that order dates are displayed as years.
e. Submit the program to produce the following report:
Partial PROC MEANS Output
Revenue from All Orders
The MEANS Procedure
Analysis Variable : Total_Retail_Price Total Retail Price for This Product
Date
11-41
Order
was
placed
by
Order
N
Customer
Type
Obs
Sum
2007
Retail
53
7938.80
2008
Catalog
52
10668.08
Internet
23
4124.05
Retail
63
9012.22
Catalog
23
3494.60
Internet
22
3275.70
7. Validating orion.price_current with the UNIVARIATE Procedure

b. Add a VAR statement to analyze Unit_Sales_Price and Factor and submit the program.
c. View the Extreme Observations output.
How many values of Unit_Sales_Price are over the maximum of 800? __________________
How many values of Factor are under the minimum of 1? ____________________________
How many values of Factor are over the maximum of 1.05? __________________________
Level 2
8. Analyzing Missing Numeric Values with PROC MEANS
b. Display the number of missing values and the number of nonmissing values present in the
Birth_Date, Emp_Hire_Date, and Emp_Term_Date variables.
c. Add a CLASS statement to display separate statistics for each value of Gender.
d. Suppress the column that displays the total number of observations in each classification group.
e. Submit the program to produce the following report:
PROC MEANS Output
Number of Missing and Non-Missing Date Values
The MEANS Procedure
Employee
N
Gender
Variable
Label
Miss
N
F
Birth_Date
Employee Birth Date
0
191
Emp_Hire_Date
Employee Hire Date
0
191
11-42

Emp_Term_Date
Employee Termination Date
139
52
Birth_Date
Employee Birth Date
0
233
Emp_Hire_Date
Employee Hire Date
0
233
Emp_Term_Date
Employee Termination Date
169
64
9. Validating orion.shoes_tracker with the UNIVARIATE Procedure

a. Write a PROC UNIVARIATE step to validate Product_ID of orion.shoes_tracker. A valid
Product_ID must have exactly twelve digits.
b. Submit the program and view the Extreme Observations output.
How many values of Product_ID are too small? _____________________________________
How many values of Product_ID are too large? _____________________________________
Challenge
10. Creating an Output Data Set with PROC MEANS
b. Modify the PROC MEANS step to create an output data set containing the sum of
Total_Retail_Price values for each value of Product_ID. Creating an output data set from
PROC MEANS results is discussed in the SAS Help facility and in the online documentation.
c. Combine the output data set with orion.product_list to obtain the Product_Name value for each
Product_ID code.
d. Sort the combined data so that the products with higher revenues appear at the top of the resulting
data set.
e. Apply the OBS= data set option in a PROC PRINT step to display the first five observations
that is, those that represent the five products with the most revenue.
f. Display the revenue values with a leading euro symbol (), a period that separates every three
digits, a comma that separates the decimal fraction, and two decimal places.
g. Submit the program to produce the following report:
PROC MEANS Output
Top Five Products by Revenue
Obs
1
2
3
4
5
Revenue
3.391,80
3.080,30
2.250,00
1.937,20
1.796,00
Product
Number
230100700009
230100700008
230100700011
240200100173
240200100076
Product
Family Holiday 6
Family Holiday 4
Hurricane 4
Proplay Executive Bi-Metal Graphite
Expert Men's Firesole Driver
11-43
11. Selecting Only the Extreme Observations Output from the UNIVARIATE Procedure
a. Write a PROC UNIVARIATE step to validate Product_ID in orion.shoes_tracker.
b. Before the PROC UNIVARIATE step, add the following ODS statement:
ods trace on;
c. After the PROC UNIVARIATE step add, the following ODS statement:
ods trace off;
d. Submit the program and notice the trace information in the SAS log.
What is the name of the last output added in the SAS log? ______________________________
e. Add an ODS SELECT statement immediately before the PROC UNIVARIATE step to select only
the Extreme Observation output object. Documentation about the ODS TRACE and ODS
SELECT statements can be found in the SAS Help facility and in the online documentation.
f. Submit the program to create the following PROC UNIVARIATE report:
Variable: Product_ID (Product ID)
--------Lowest-------
-------Highest------
Value
Obs
Value
Obs
2.20200E+10
2.20200E+11
2.20200E+11
2.20200E+11
2.20200E+11
4
1
2
3
5
2.2020E+11
2.2020E+11
2.2020E+11
2.2020E+11
2.2020E+12
6
7
9
10
8
11-44
11.3 Using the Output Delivery System

Objectives
Define the Output Delivery System and ODS

destinations.
Create report output in PDF, RTF, and HTML format.
Specify a style definition using the STYLE= option.
Create report output that can be viewed in Microsoft
Excel.
87
Business Scenario
Generate reports in various formats for distribution within
Orion Star.
88
SAS Procedure Output

Historically, SAS procedures generate plain text output
using monospace fonts.
Obs
1
2
3
4
5
6
7
8
9
10
First_Name
Last_Name
Satyakam
Monica
Kevin
Petrea
Marina
Shani
Fang
Michael
Amanda
Vincent
Denny
Kletschkus
Lyon
Soltau
Iyengar
Duckett
Wilson
Minas
Liebman
Eastley
Job_Title
Sales
Sales
Sales
Sales
Sales
Sales
Sales
Sales
Sales
Sales
Rep.
Rep.
Rep.
Rep.
Rep.
Rep.
Rep.
Rep.
Rep.
Rep.
II
IV
I
II
III
I
II
I
II
III
Salary
26780
30890
26955
27440
29715
25795
26810
26970
27465
29695
89
Output in SAS Enterprise Guide

SAS Enterprise Guide produces SAS Report output by
default.
The output from each program is written to a specially
formatted file.
You can request the following output styles by
selecting Tools Options Results General:
PDF
RTF
HTML
LISTING (plain text)
The PDF, RTF, and HTML files can be saved and
opened outside of Enterprise Guide.
90
11-45
11-46
Output in the SAS Windowing Environment

SAS 9.3 produces HTML output by default.
All procedure output accumulates in a single file.
LISTING output can be requested by selecting
Tools Options Preferences Results.
LISTING output is the default for prior releases and for
SAS 9.3 in batch mode.
91
Output Delivery System (ODS)

Use the Output Delivery System to create different output
formats by directing output to various ODS destinations.
LISTING
HTML
Procedure Output
RTF
PDF
ODS statements and destinations can be used regardless

of the mode, operating environment, or platform.
92
11-47
ODS Statements
Open an ODS destination, submit one or more
procedures that generate output, and then close the
destination.
ods html file="&path\myreport.html";
tables Country;
run;
ods html close;
ODS destination FILE="filename" <options>;

<SAS code to generate the report>
ODS destination CLOSE;
p111d12
93
The Output Delivery System works in all operating environments. Use the RS=NONE option when you
create HTML and RTF files on z/OS (OS/390).
/* z/OS example */
ods html file="&path..report(example)" rs=none;
tables Country;
run;
ods html close;
Viewing the Output

HTML output can be viewed in the interactive
environment or using a browser.
94
11-48
ODS Destinations
ODS creates various types of output based on the
specified destinations and file types.
Destination
Type of File
Extension
Viewed In
LISTING
Plain text
HTML
Hypertext Markup
Language
html
Web Browsers such as

Internet Explorer
PDF
Portable
Document Format
pdf
Adobe products such

as Acrobat Reader
RTF
Rich Text Format
rtf
Word processors such

as Microsoft Word
SAS Output window
95
Multiple Destinations and Procedures

Output from multiple procedures can be sent to multiple
ODS destinations.
ods pdf file="&path\example.pdf";
ods rtf file="&path\example.rtf";
tables Country;
run;
var Salary;
run;
ods pdf close;
ods rtf close;
96
p111d14
Closing All Destinations

Use _ALL_ in the ODS CLOSE statement to close all
open destinations. This includes the LISTING destination.
ods pdf file="&path\example.pdf";
ods rtf file="&path\example.rtf";
tables Country;
run;
ods _all_ close;
p111d15
97
No Open Destinations
Be sure to have at least one destination open.
2723
2724
2725
2726
ods _all_ close;

tables Country;
run;
Output is generated
but not displayed.
WARNING: No output destinations active.

ORION.SALES.
2727
2728
2729
2730
2731
2732
ods listing;
tables Country;
run;
Output is displayed.
ORION.SALES.
98
p111d15
11-49
11-50
11.07 Quiz
What is the problem with this program?
run;
ods close;
p111a05
99
STYLE= Option
Use a STYLE= option in the ODS statement to specify
a style definition.
ODS destination FILE="filename"

STYLE=style-definition;
A style definition sets presentation aspects including

colors and fonts.
STYLE= has no effect in the LISTING destination.
101
HTML Examples
STYLE=DEFAULT
STYLE=SASWEB
p111d12
102
PDF and RTF Examples

STYLE=PRINTER
STYLE=JOURNAL
STYLE=RTF
STYLE=OCEAN
103
p111d17
11-51
11-52
Available Styles
Use the TEMPLATE procedure to see the available
styles.
Partial Output
Listing of: SASHELP.TMPLMST
proc template;
Path Filter is: Styles
list styles;
Sort by: PATH/ASCENDING
run;
Obs
Path
Type
1
Styles
Dir
2
Styles.Analysis
Style
3
Styles.Astronomy
Style
4
Styles.Banker
Style
5
Styles.BarrettsBlue
Style
6
Styles.Curve
Style
7
Styles.Default
Style
8
Styles.Dtree
Style
9
Styles.EGDefault
Style
10
Styles.Education
Style
p111d13
104
Business Scenario
Create SAS reports that can be opened in Microsoft Excel
using the CSVALL, MSOFFICE2K, and EXCELXP
destinations.
csvall
msoffice2k
excelxp
106
Destinations Used with Excel

Destination
Type of File
Extension
Viewed In
CSVALL
Comma-Separated
Value
csv
Editor or
Microsoft Excel
MSOFFICE2K
Hypertext Markup
Language
html
Web browser or
Microsoft Word
or Microsoft
Excel
TAGSETS.
EXCELXP
Extensible Markup
Language
xml
Microsoft Excel
107
11.08 Quiz
Complete the ODS statements below to send the output
to a CSVALL destination.
ods
file="&path\myexcel.
";

tables Country;
run;
var Salary;
run;
ods
close;
108
11-53
11-54
CSVALL Destination
CSVALL does not include any style information.
110
MSOFFICE2K Destination
MSOFFICE2K keeps the style information, including
spanning headers.
111
EXCELXP Destination
EXCELXP keeps the style information. Output from each
procedure is on a separate sheet.
112
Keep in Mind
The file you are creating is not an Excel file.
CSVALL
MSOFFICE2K
EXCELXP
113
Using the Output Delivery System

Creating PDF, RTF, and HTML Files
p111d18
This demo can be run in SAS Enterprise Guide and in the SAS windowing environment.
11-55
11-56
Open and submit p111d18. It creates HTML, PDF, and RTF output.
ods listing close;
ods pdf file="&path\myreport.pdf";
ods rtf file="&path\myreport.rtf";
title 'Report 1';
tables Country;
run;
title 'Report 2';
var Salary;
run;
title 'Report 3';
var First_Name Last_Name
Job_Title Country Salary;
where Salary > 75000;
run;
ods _all_ close;
ods listing;
title;
For UNIX, the following ODS statements are used:
ods listing close;
ods html file="&path/myreport.html";
ods pdf file="&path/myreport.pdf";
ods rtf file="&path/myreport.rtf";
For z/OS (OS/390), the following ODS statements are used:
ods listing close;
ods html file="&path..report(myhtml) rs=none";
ods pdf file="&path..report(mypdf) rs=none";
ods rtf file="&path..report(myrtf) rs=none";
Viewing the Output in SAS Enterprise Guide

The results are displayed on individual tabs within the workspace.
Click the appropriate tab to see the results within Enterprise Guide. If a download prompt is displayed,
select the download.
HTML Results
PDF Results
RTF Results
After the RTF file is downloaded, Microsoft Word opens automatically to display the RTF results.
The RTF tab is removed from the workspace.
The RTF file shortcut is available in the process flow.
11-57
11-58
Viewing the Output in the SAS Windowing Environment

In the SAS windowing environment on the Windows operating platform, the Results Viewer window can
be used to view the HTML file.
1. After submitting the program, go to the Results window.
2. Right-click the word Results within the Results window and select Expand All.
3. The program created three output files. They are listed in the Results window with an appropriate
icon based on file type. Double-click an icon to view the results.
HTML File in the Results Viewer
PDF File in Adobe Acrobat
RTF File in Microsoft Word
Creating Files That Open in Excel

p111d19
This demo can be run in SAS Enterprise Guide and in the SAS windowing environment.

ods listing close;
ods csvall file="&path\myexcel.csv";
ods msoffice2k file="&path\myexcel.html";
11-59
11-60
ods tagsets.excelxp file="&path\myexcel.xml";

title 'Report 1';
tables Country;
run;
title 'Report 2';
var Salary;
run;
title 'Report 3';
var First_Name Last_Name
Job_Title Country Salary;
where Salary > 75000;
run;
ods _all_ close;
ods listing;
title;
For UNIX, the following ODS statements are used:
ods listing close;
ods csvall file="&path/myexcel.csv";
ods msoffice2k file="&path/myexcel.html";
ods tagsets.excelxp file="&path/myexcel.xml";
For z/OS (OS/390), the following ODS statements are used:
ods listing close;
ods csvall file="&path..report(mycsv)" rs=none;
ods msoffice2k file="&path..report(myhtml)" rs=none;
ods tagsets.excelxp file="&path..report(myxml)" rs=none;
2. You might be prompted to open or save the CSV file. Click Cancel.
3. Launch Microsoft Excel to view the three files created by this program.
4. Within Excel, select File Open and navigate to your data folder. In the Filename box, type
myexcel and select myexcel.csv to view the CSV file created through the CSVALL destination.
11-61
5. Within Excel, open myexcel.html to view the HTML file created through the MSOFFICE2K
destination.
6. Within Excel, open myexcel.xml to view the XML file created through the EXCELXP destination.
7. Close Microsoft Excel.
11-62
Exercises
Level 1
12. Directing Output to PDF and RTF Destinations
Open p111e12. Create a PDF version of the PROC PRINT report by adding ODS statements. Use the
following naming convention when creating the PDF file:
Windows
"&path\p111s12p.pdf"
UNIX
"&path/p111s12p.pdf"
z/OS (OS/390)
"&path...report(p111s12p)"
a. Submit the program and view the output in Adobe Reader.

PROC PRINT Output
Compare this PDF output to the equivalent report that appears in the Output window.
Modify your ODS statements to create the RTF version of the PROC PRINT report. Use the
following naming convention when creating the RTF file.
Windows
"&path\p111s12r.rtf"
UNIX
"&path/p111s12r.rtf"
z/OS (OS/390)
"&path..report(p111s12r)"
b. Submit the program and view the RTF output in Microsoft Word:
11-63
What happens if the RUN statement is moved to the end of the program?
c. Add the STYLE= option to the ODS RTF statement to use the Curve style definition.
d. Submit the program and view output in Microsoft Word.
Level 2
13. Creating ODS Output Compatible with Microsoft Excel
a. Open p111e13. Add ODS statements to send the report to a file that can be viewed in Microsoft
Excel. Choose the ODS destination and file extension based on whether you want the following:
1) style information stored in the report output
2) the reports in a single worksheet or multiple worksheets
If selecting a destination that supports style information, specify the LISTING style definition.
Use the following naming convention when creating the file. For Windows or UNIX, choose
an appropriate extension for the file depending on the type of file that is created.
Windows
"&path\p111s13.xxx"
UNIX
"&path/p111s13.xxx"
z/OS (OS/390)
"&path..report(p111s13)"
b. Submit the program to produce the output file.
11-64
c. Open the file with Microsoft Excel. The report should resemble the following results. Your
output will look different depending on the ODS destination that you choose.
Challenge
14. Adding HTML-Specific Features to ODS Output
a. Open p111e14. Create an HTML version of the PROC PRINT report by adding ODS statements.
Use the following naming convention when creating the HTML file:
Windows
"&path\p111s14.html"
UNIX
"&path/p111s14.html"
z/OS (OS/390)
"&path..report(p111s14)"
b. Customize the title so that it becomes a clickable hyperlink when displayed in a web browser.
The hyperlink should point to the URL http://www.sas.com (the SAS home page).
An option must be added to the TITLE statement to make it an active hyperlink. Explore
TITLE statement options in the SAS Help facility or online documentation.
c. Submit the program to produce the following HTML output:

15. Implementing Cascading Style Sheets with ODS Output

a. Open p111e15. Modify the ODS HTML statement so that the output generated by the program
uses a cascading style sheet. This CSS definition applies a yellow background to data cells in the
ODS output.
11.4 Solutions
11-65
Use the following cascading style sheet:

Windows
"&path\p111e15c.css"
UNIX
"&path/p111e15c.css"
z/OS (OS/390)
"&path..report(p111e15c)"
Explore the ODS HTML statement in the SAS Help facility or online documentation to
investigate the syntax required to reference an existing CSS file. Follow the
documentation links for the STYLESHEET= option and its URL= suboption.
b. Submit the program to produce the following HTML output:

PROC PRINT Output
11.4 Solutions
1. Counting Levels of a Variable with PROC FREQ
Would you expect to see frequencies of 1 for customers and sales employees? No. This file contains
all customer orders, and the ID of the employee who helped with the sale. Employee_ID is likely
to have higher frequency counts, as would frequent customers.
title1 'Unique Customers and Salespersons for Retail Sales';
proc freq data=orion.orders nlevels;
where Order_Type=1;
tables Customer_ID Employee_ID / noprint;
run;
title;
title1 'Catalog and Internet Customers';
proc freq data=orion.orders order=freq;
where Order_Type ne 1;
tables Customer_ID / nocum;
run;
title;
2. Validating orion.shoes_tracker with PROC FREQ
proc freq data=orion.shoes_tracker nlevels;
tables Supplier_Name Supplier_ID;
run;
11-66
What invalid data exists for Supplier_Name and Supplier_ID?

two observations with invalid values for Supplier_Name (3op Sports)
one observation with a missing value for Supplier_ID
3. Producing Frequency Reports with PROC FREQ
proc format;
value ordertypes
1='Retail'
2='Catalog'
3='Internet';
run;
title 'Order Summary by Year and Type';
proc freq data=orion.orders ;
tables Order_Date;
tables Order_Type / nocum;
tables Order_Date*Order_Type / nopercent norow nocol;
format Order_Date year4. Order_Type ordertypes.;
run;
title;
4. Validating orion.qtr2_2011 with PROC FREQ
proc freq data=orion.qtr2_2011 nlevels;
tables Order_ID Order_Type;
run;
What invalid data exists for Order_ID and Order_Type?
two observations with missing values for Order_ID
one observation with a value of 0 for Order_Type
one observation with a value of 4 for Order_Type
5. Creating an Output Data Set with PROC FREQ
proc freq data=orion.order_fact noprint;
tables Product_ID / out=product_orders;
run;
data product_names;
merge product_orders orion.product_list;
by Product_ID;
keep Product_ID Product_Name Count;
run;
proc sort data=product_names;
by descending Count;
run;
title 'Top Five Products by Number of Orders';
proc print data=product_names(obs=5) label noobs;
var Count Product_ID Product_Name;
label Product_ID='Product Number'
11.4 Solutions
Product_Name='Product'
Count='Orders';
run;
title;
6. Creating a Summary Report with PROC MEANS
proc format;
value ordertypes
1='Retail'
2='Catalog'
3='Internet';
run;
title 'Revenue from All Orders';
proc means data=orion.order_fact sum;
var Total_Retail_Price;
class Order_Date Order_Type;
format Order_Date year4. Order_Type ordertypes.;
run;
title;
7. Validating orion.price_current with the UNIVARIATE Procedure
proc univariate data=orion.price_current;
var Unit_Sales_Price Factor;
run;
Find the Extreme Observations output.
How many values of Unit_Sales_Price are over the maximum of 800? one (5730)
How many values of Factor are under the minimum of 1? one (0.01)
How many values of Factor are over the maximum of 1.05? two (10.20 and 100.00)
8. Analyzing Missing Numeric Values with PROC MEANS
title 'Number of Missing and Non-Missing Date Values';
proc means data=orion.staff nmiss n nonobs;
var Birth_Date Emp_Hire_Date Emp_Term_Date;
class Gender;
run;
title;
9. Validating orion.shoes_tracker with the UNIVARIATE Procedure
proc univariate data=orion.shoes_tracker;
var Product_ID;
run;
How many values of Product_ID are too small? one (2.20200E+10)
How many values of Product_ID are too large? one (2.2020E+12)
10. Creating an Output Data Set with PROC MEANS
proc means data=orion.order_fact noprint nway;
class Product_ID;
11-67
11-68
var Total_Retail_Price;
output out=product_orders sum=Product_Revenue;
run;
data product_names;
merge product_orders orion.product_list;
by Product_ID;
keep Product_ID Product_Name Product_Revenue;
run;
proc sort data=product_names;
by descending Product_Revenue;
run;
title 'Top Five Products by Revenue';
proc print data=product_names(obs=5) label;
var Product_Revenue Product_ID Product_Name;
label Product_ID='Product Number'
Product_Name='Product'
Product_Revenue='Revenue';
format Product_Revenue eurox12.2;
run;
title;
11. Selecting Only the Extreme Observations Output from the UNIVARIATE Procedure
ods trace on;
ods select ExtremeObs;
proc univariate data=orion.shoes_tracker;
var Product_ID;
run;
ods trace off;
What is the name of the last output added in the SAS log? ExtremeObs
12. Directing Output to PDF and RTF Destinations
ods pdf file="&path\p111s12p.pdf";
ods rtf file="&path\p111s12r.rtf" style=curve;
title 'July 2011 Orders';
proc print data=orion.mnth7_2011;
run;
ods pdf close;
ods rtf close;
13. Creating ODS Output Compatible with Microsoft Excel
ods csvall file="&path\p111s13.csv";
ods msoffice2k file="&path\p111s13.html" style=Listing;
ods tagsets.excelxp file="&path\p111s13.xml" style=Listing;
title 'Customer Type Definitions';
proc print data=orion.customer_type;
run;
11.4 Solutions
title 'Country Definitions';

run;
title;
ods csvall close;
ods msoffice2k close;
ods tagsets.excelxp close;
14. Adding HTML-Specific Features to ODS Output
ods html file="&path\p111s14.html";
proc print data=orion.customer;
title link='http://www.sas.com' 'Customer Information';
run;
ods html close;
15. Implementing Cascading Style Sheets with ODS Output
ods html file="&path\p111s15.html"
stylesheet=(url="&path\p111e15c.css");
proc print data=orion.customer_type;
title 'Customer Type Definitions';
run;
ods html close;

What change was needed? A slash is required before
the options in the TABLES statement.
31
32
tables country nocum nopercent;
ERROR: Variable NOCUM not found.
ERROR: Variable NOPERCENT not found.
33
run;

tables country / nocum nopercent;
run;
The FREQ Procedure
Country
Frequency
AU
63
US
102
11
p111a01s
11-69
11-70

What problems exist with the data in this partial data set?
Hint: There are seven data problems.

28

How many unique, nonmissing job titles exist? 124
Which job title occurs most frequently? Trainee
What is the frequency of missing job titles? 1
proc freq data=orion.nonsales2 nlevels
order=freq;
tables Job_Title /nocum nopercent;
run;
40
p111a02s
11.4 Solutions

Can user-defined formats be used to group data? yes
The FREQ Procedure
Cumulative
Cumulative
Salary
Frequency
Percent
Frequency
Percent
Tier1
1
0.61
1
0.61
Tier2
158
95.76
159
96.36
Tier3
4
2.42
163
98.79
Tier4
2
1.21
165
100.00
p111a03s
48

For a given data set, there are 63 observations with
a Country value of AU. Of those 63 observations,
only 61 observations have a value for Salary.
Which output is correct?
a.

N
Country
Obs
N
AU
63
61
b.

N
Country
Obs
N
AU
61
63
65
11-71
11-72

PROC UNIVARIATE identified two observations with
Salary values less than 24,000.
What procedure can be used to display the observations
containing the invalid values? PROC PRINT
proc print data=orion.nonsales2;
where Salary<24000;
run;
PROC PRINT Output
Obs
Employee_ID
4
13
20
120106
120115
120191
First
Last
John
Hugh
Jannene
Hornsey
Nichollas
Graham-Rowe
Gender
M
M
F
Salary
.
2650
2401
Job_Title
Office Assistant II
Service Assistant I
Trainee
Country
AU
AU
AU
p111a04s
83

What is the problem with this program?
run;
ods html close;
100
p111a01s
11.4 Solutions

Complete the ODS statements below to send the output
to a CSVALL destination.
ods csvall file="&path\myexcel.csv";
tables Country;
run;
var Salary;
run;
ods csvall close;
109
p111a06s
11-73
11-74
Chapter 12 Learning More

12.1 Introduction................................................................................................................... 12-3
12-2
12.1 Introduction
12.1 Introduction
Objectives
Identify the areas of support that SAS offers.

Identify the next steps after the completion of this
course.
2
2
Customer Support
SAS provides a variety of resources to help customers.
http://support.sas.com/resourcekit/
12-3
12-4
Education
SAS Education provides comprehensive training, including
more than 200 course offerings

world-class instructors
multiple delivery methods
worldwide training centers.
http://support.sas.com/training/
4
4
SAS Global Certification Program

SAS Education also provides
globally recognized certifications
preparation materials
practice exams.
http://support.sas.com/certify/
5
5
12.1 Introduction
Networking
Social media channels and user group organizations
enable you to
interact with other SAS users and SAS staff

learn new programming tips and tricks
get exclusive discounts.
For training-specific information:

http://support.sas.com/training/socialmedia
6
6
Icons are from top left to right, and then bottom left to right:
Twitter, RSS, Myspace, YouTube, Facebook, Technorati, and sasCommunity.org
SAS Publishing
SAS Publications offers a complete selection
of publications, including
e-books
CD-ROM
hard-copy books
books written by outside authors.
http://support.sas.com/publishing/
7
7
12-5
12-6
Beyond This Course

To grow your SAS skills, remember to activate the
extended learning page for this course.
Individual learning software licenses are available

through SAS OnDemand for Professionals:
Enterprise Guide.
http://support.sas.com/learn/ondemand/professionals
8
8
Next Steps
To learn more about this:
Enroll in this course:
Programming techniques, including

DATA and PROC steps to manipulate
data
SAS Programming 2: Data

Manipulation Techniques
How to use SQL to manipulate and

merge data files
SAS SQL 1: Essentials
Creating reports using the REPORT and SAS Report Writing 1: Essentials
TABULATE procedures, plus the Output
Delivery System (ODS)
Performing statistical analysis using
SAS/STAT software
Statistics 1: Introduction to ANOVA,

Regression, and Logistic Regression
How to use many of the most popular

SAS functions
SAS Functions by Example
A variety of techniques to find errors in

your data
Data Cleaning Techniques
SAS9 Business Intelligence

capabilities
Overview of SAS Business Intelligence

and Data Integration Applications
9
9

LWPRG1

Uploaded by

Copyright:

Available Formats

LWPRG1

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

LWPRG1

Uploaded by

Copyright:

Available Formats

SAS Programming 1:

For Your Information

Introduction .......................................................................................... 1-1

Overview of SAS Foundation .......................................................................................... 1-3

Course Logistics .............................................................................................................. 1-6

Course Data Files ........................................................................................................... 1-14

SAS Programs ..................................................................................... 2-1

Introduction to SAS Programs ......................................................................................... 2-3

Submitting a SAS Program .............................................................................................. 2-7

SAS Program Syntax ..................................................................................................... 2-20

Solutions ........................................................................................................................ 2-36

For Your Information

Solutions to Student Activities (Polls/Quizzes) ....................................................... 2-38

Accessing Data ..................................................................................... 3-1

Examining SAS Data Sets ............................................................................................... 3-3

Accessing SAS Libraries ............................................................................................... 3-13

Solutions ........................................................................................................................ 3-34

Producing Detail Reports..................................................................... 4-1

Subsetting Report Data .................................................................................................... 4-3

Sorting and Grouping Report Data ................................................................................ 4-27

Enhancing Reports ......................................................................................................... 4-40

Solutions ........................................................................................................................ 4-50

Formatting Data Values........................................................................ 5-1

Using SAS Formats.......................................................................................................... 5-3

User-Defined Formats .................................................................................................... 5-12

For Your Information

Solutions ........................................................................................................................ 5-25

Reading SAS Data Sets ...................................................................... 6-1

Reading a SAS Data Set .................................................................................................. 6-3

Customizing a SAS Data Set ......................................................................................... 6-17

Solutions ........................................................................................................................ 6-42

Reading Spreadsheet and Database Data .......................................... 7-1

Reading Spreadsheet Data ............................................................................................... 7-3

Reading Database Data .................................................................................................. 7-20

Solutions ........................................................................................................................ 7-24

Reading Raw Data Files ....................................................................... 8-1

Introduction to Reading Raw Data Files .......................................................................... 8-3

Reading Standard Delimited Data.................................................................................... 8-7

For Your Information

Reading Nonstandard Delimited Data ........................................................................... 8-37

Handling Missing Data .................................................................................................. 8-54

Solutions ........................................................................................................................ 8-63

Manipulating Data................................................................................. 9-1

Using SAS Functions ....................................................................................................... 9-3

Conditional Processing .................................................................................................. 9-12

Solutions ........................................................................................................................ 9-40

Combining Data Sets.......................................................................... 10-1

10.1 Concatenating Data Sets ................................................................................................ 10-3

For Your Information

Solutions to Exercises ........................................................................................... 10-73

Creating Summary Reports ............................................................... 11-1

11.1 The FREQ Procedure ..................................................................................................... 11-3

Learning More ..................................................................................... 12-1

12.1 Introduction .................................................................................................................... 12-3

For Your Information

For Your Information

For Your Information

Overview of SAS Foundation ........................................................................................ 1-3

Course Logistics ............................................................................................................ 1-6

Course Data Files ......................................................................................................... 1-14