Previous Page

|

Next Page

Introduction to DATA Step Processing

The SAS Data Set: Your Key to the SAS System

Understanding the Function of the SAS Data Set

SAS enables you to solve problems by providing methods to analyze or to process your data in some way. You need to first get the data into a form that SAS can recognize and process. After the data is in that form, you can analyze it and generate reports. The following figure shows this process in the simplest case.

From Raw Data to Final Analysis

You begin with raw data, that is, a collection of data that has not yet been processed by SAS. You use a set of statements known as a DATA step to get your data into a SAS data set. Then you can further process your data with additional DATA step programming or with SAS procedures.

In its simplest form, the DATA step can be represented by the three components that are shown in the following figure.

From Raw Data to a SAS Data Set

SAS processes input in the form of raw data and creates a SAS data set.

When you have a SAS data set, you can use it as input to other DATA steps. The following figure shows the SAS statements that you can use to create a new SAS data set.

Using One SAS Data Set to Create Another

Understanding the Structure of the SAS Data Set

Think of a SAS data set as a rectangular structure that identifies and stores data. When your data is in a SAS data set, you can use additional DATA steps for further processing, or perform many types of analyses with SAS procedures.

The rectangular structure of a SAS data set consists of rows and columns in which data values are stored. The rows in a SAS data set are called observations, and the columns are called variables. In a raw data file, the rows are called records and the columns are called fields. Variables contain the data values for all of the items in an observation.

For example, the following figure shows a collection of raw data about participants in a health and fitness club. Each record contains information about one participant.

Raw Data from the Health and Fitness Club

The following figure shows how easily the health club records can be translated into parts of a SAS data set. Each record becomes an observation. In this case, each observation represents a participant in the program. Each field in the record becomes a variable. The variables represent each participant's identification number, name, team name, and weight at the beginning and end of a 16-week program.

How Data Fits into a SAS Data Set

In a SAS data set, every variable exists for every observation. What if you do not have all the data for each observation? If the raw data is incomplete because a value for the numeric variable EndWeight was not recorded for one observation, then this missing value is represented by a period that serves as a placeholder, as shown in observation 6 in the previous figure. (Missing values for character variables are represented by blanks. Character and numeric variables are discussed later in this section.) By coding a value as missing, you can add an observation to the data set for which the data is incomplete and still retain the rectangular shape necessary for a SAS data set.

Along with data values, each SAS data set contains a descriptor portion, as illustrated in the following figure:

Parts of a SAS Data Set

The descriptor portion consists of details that SAS records about a data set, such as the names and attributes of all the variables, the number of observations in the data set, and the date and time that the data set was created and updated.

Operating Environment Information:   Depending on your operating environment and the engine used to write the SAS data set, SAS may store additional information about a SAS data set in its descriptor portion. For more information, refer to the SAS documentation for your operating environment.   

Temporary versus Permanent SAS Data Sets

Creating and Using Temporary SAS Data Sets

When you use a DATA step to create a SAS data set with a one-level name, you normally create a temporary SAS data set, one that exists only for the duration of your current session. SAS places this data set in a SAS data library referred to as WORK. In most operating environments, all files that SAS stores in the WORK library are deleted at the end of a session.

The following is an example of a DATA step that creates the temporary data set WEIGHT_CLUB.

data weight_club;
   input IdNumber Name $ 6-20 Team $ 22-27 StartWeight EndWeight;
   datalines;
1023 David Shaw      red    189 165
1049 Amelia Serrano  yellow 145 124
1219 Alan Nance      red    210 192
1246 Ravi Sinha      yellow 194 177
1078 Ashley McKnight red    127 118
1221 Jim Brown       yellow 220 .
;

run;

The preceding program code refers to the temporary data set as WEIGHT_CLUB. SAS. However, it assigns the first-level name WORK to all temporary data sets, and refers to the WEIGHT_CLUB data set with its two-level name, WORK.WEIGHT_CLUB. The following output from the SAS log shows the name of the temporary data set.

SAS Log: The WORK.WEIGHT_CLUB Temporary Data Set

162  data weight_club;
163     input IdNumber Name $ 6-20 Team $ 22-27 StartWeight EndWeight;
164     datalines;
NOTE: The data set WORK.WEIGHT_CLUB has 6 observations and 5 variables.

Because SAS assigns the first-level name WORK to all SAS data sets that have only a one-level name, you do not need to use WORK. You can refer to these temporary data sets with a one-level name, such as WEIGHT_CLUB.

To reference this SAS data set in a later DATA step or in a PROC step, you can use a one-level name:

proc print data = weight_club;
run;

Creating and Using Permanent SAS Data Sets

To create a permanent SAS data set, you must indicate a SAS data library other than WORK. (WORK is a reserved libref that SAS automatically assigns to a temporary SAS data library.) Use a LIBNAME statement to assign a libref to a SAS data library on your operating environment's file system. The libref functions as a shorthand way of referring to a SAS data library. Here is the form of the LIBNAME statement:

LIBNAME libref 'your-data-library';

where

libref

is a shortcut name to where your SAS files are stored. libref must be a valid SAS name. It must begin with a letter or an underscore, and it can contain uppercase and lowercase letters, numbers, or underscores. A libref has a maximum length of 8 characters.

'your-data-library'

must be the physical name for your SAS data library. The physical name is the name that is recognized by the operating environment.

Operating Environment Information:   Additional restrictions can apply to librefs and physical file names under some operating environments. For more information, refer to the SAS documentation for your operating environment.  

The following is an example of the LIBNAME statement that is used with a DATA step:

libname saveit 'your-data-library'; 1 
data saveit.weight_club; 2 
   ...more SAS statements...
;

proc print data = saveit.weight_club; 3 
run;

The following list corresponds to the numbered items:

	The LIBNAME statement associates the libref SAVEIT with your-data-library, where your-data-library is your operating environment's name for a SAS data library.
	To create a new permanent SAS data set and store it in this SAS data library, you must use the two-level name SAVEIT.WEIGHT_CLUB in the DATA statement.
	To reference this SAS data set in a later DATA step or in a PROC step, you must use the two-level name SAVEIT.WEIGHT_CLUB in the PROC step.

For more information, see Understanding SAS Data Libraries.

Conventions That Are Used in This Documentation

Data sets that are used in examples are usually shown as temporary data sets specified with a one-level name:

data fitness;

In rare cases in this documentation, data sets are created as permanent SAS data sets. These data sets are specified with a two-level name, and a LIBNAME statement precedes each DATA step in which a permanent SAS data set is created:

libname saveit 'your-data-library';
data saveit.weight_club;

Previous Page

|

Next Page

|

Top of Page

Copyright © 2012 by SAS Institute Inc., Cary, NC, USA. All rights reserved.