Handbook of SAS DATA Step Programming

The document summarizes and reviews the book "Handbook of SAS DATA Step Programming" by Arthur Li. It provides an overview of the book's contents and structure. The book is a thorough introduction to SAS DATA step programming and focuses exclusively on the DATA step without using macros or SQL. It explains how the DATA step works through examples demonstrating data manipulation, summary statistics, and the underlying program data vector (PDV) mechanism. The review concludes that the book helps statisticians improve their programming efficiency in SAS DATA step.

Uploaded by

Evelyn Lin

Available Formats

Download as PDF, TXT or read online on Scribd

0% found this document useful (0 votes)

65 views

Handbook of SAS DATA Step Programming

Uploaded by

Evelyn Lin

Available Formats

Download as PDF, TXT or read online on Scribd

You are on page 1/ 3

JSS Journal of Statistical Software

January 2014, Volume 56, Book Review 1. http://www.jstatsoft.org/

Reviewer: Xiangxiang Meng, Wayne Thompson

SAS Institute Inc.

Handbook of SAS ® DATA Step Programming

Arthur Li
Chapman & Hall/CRC, Boca Raton, FL, 2013.
ISBN 978-1-46655-239-5. 275 pp. USD 59.95.
http://www.crcpress.com/product/isbn/9781466552388

The SAS DATA step is one of the most widely used statistical programming languages for
data manipulation, data set management and statistical summary. “Handbook of SAS DATA
Step Programming” by Arthur Li provides a thorough introduction to the statements and
functionalities of the SAS DATA step. The book can be used as an introductory tutorial for
beginning SAS programmers, or as a reference book for experienced users.
Unlike other tutorials for SAS programming, the book focuses on DATA step. Without using
any SAS macro language or PROC SQL, the book shows the comprehensive capabilities of
DATA step to manipulate and summarize data, including data input/output, table merge and
concatenation, column modification and derivation, and calculation of summary statistics.
The book not only shows how to program in a DATA step, but also explains in details how
the DATA step works when it executes a sequence of statements through an implicit loop of
observations. What happens inside the SAS program data vector (PDV) is explained in full
detail for many important elements of the DATA step, such as the RETAIN statement and the
BY processing.
The first two chapters of the book serve as a tutorial for beginners. Instead of introducing
DATA step as a statistical or data manipulation tool, Chapter 1 describes it as a programming
language, and well explains the building blocks and basic syntax of SAS programming lan-
guage and SAS DATA step. The most widely used SAS statements and SAS/BASE procedures
are then introduced with examples that are simple and easy to use. Chapter 2 follows the
style of Chapter 1 and continues to introduce some commonly used features of DATA step,
such as IF-THEN-ELSE and DO-END for conditional and group processing of SAS statements. In
general, the functionality of DATA step is very comprehensive and complex. Instead of covering
every statement of DATA step, these two chapters try to describe what a DATA step can do by
using examples that are frequently used by a SAS programmer to process real data, such as
clinical trial data from pharmaceutical industry.
For users who have programed in SAS DATA step and want to better understand how it works,
Chapter 3 is an excellent overview of the underlying mechanism of DATA step. DATA step
executes the statements in an implicit loop through observations, with one observation loaded
2 Handbook of SAS ® DATA Step Programming

into a program data vector after another. The underlying mechanism of DATA step cannot be
well described without explaining how the data is processed in PDV. In this chapter, the data
flow in PDV is well explained with two statements, RETAIN and SUM. These two statements
are actually the most basic but commonly used statements that requires a PDV to save some
information from one observation to be used by the next observation. And they are the
building blocks for many complex data step applications, such as last-value-carry-forward
and data aggregation. Debugging using PUT statement and SAS debugger are also introduced
in this chapter. We recommend a careful reading of the PDV examples in this chapter (such
as Figure 3.5, 3.6, 3.7 for RETAIN statement). Understanding how the PDV works is the best
way to improve programming efficiency using SAS DATA step.
Chapter 4 to 6 introduce three important concepts in SAS DATA step: by-group processing,
do-loop, and variable array. Each of these concepts are explained with examples to show the
underlying data processing in the program data vector. Most of the examples in these chapters
are well known tips and tricks for DATA step, such as example 4.3 for removing duplicated
data, and example 5.5 for random number generation.
The next four chapters cover the details of some important DATA step statements and con-
cepts that have been briefly introduced in first two chapter, such as SET/MERGE for table
concatenation and joint (Chapter 7), the INFILE/INPUT statements for reading external data
source (Chapter 8), common SAS functions and call routines (Chapter 9), and some impor-
tant SAS/BASE procedures (Chapter 10). Although reading data directly from a flat table
becomes a standard way to import data source into SAS, reading data stored in a non-flat or
compressed type of source file is still common, and Section 8.2 provides an overview of the
INPUT statement for reading such text files. SAS date/time functions, character functions, and
INPUT/PUT functons are powerful add-ons to DATA step, and they are introduced in Section 9.2
to 9.4. the TRANSPOSE procedure is another important feature that requires careful reading.
Readers are recommended to read through the six examples in this section to understand how
this procedure can transpose a data from one shape to another. Last but not least, SAS user-
defined format is a unique SAS feature to transform or translate the information in a column
without duplicating the column. For example, example 10.17 illustrates how to use grouping
formats to categorize two continuous variables AGE and INCOME, and how to use the formatted
continuous variables in the FREQ procedure to produce cross-tabulations. The examples in
this section well describe SAS formats as a derivation or mapping tool to transform variable
efficiently.
In summary, we applaud Anthur for introducing SAS DATA step as a programming language,
and explaining how things work behind the grammar of DATA step. Many users of DATA step
have a strong statistical background, but not all of them really try to understand DATA step
as a programming language. The “Handbook of SAS DATA Step Programming” covers this
gap, and helps statisticians improving their programming efficiency using SAS DATA step.

Reviewer:
Xiangxiang Meng, Wayne Thompson
SAS Institute Inc.
Journal of Statistical Software – Book Reviews 3

100 SAS Campus Drive

Cary, NC 27513, United States of America
E-mail: [email protected], [email protected]

Journal of Statistical Software http://www.jstatsoft.org/

published by the American Statistical Association http://www.amstat.org/
Volume 56, Book Review 1 Published: 2014-01-06
January 2014