Introduction

The SAS System was originally developed to provide a collection of statistical procedures to facilitate the analysis of data. Over the years, it has expanded to include a broad spectrum of capabilities: matrix manipulation, macro programming language, and data base management. It is now used extensively for report writing and data warehousing in addition to it data analysis capabilities.

For information on the availability of SAS on different hardware platforms at Virginia Tech, see:

• VT Available SAS Products

The SAS System provides tools for strategic application development and provides SAS language portability from one environment to another (see Transferring SAS Data Sets Among Operating Systems). Portability was made possible through what SAS Institute's Multi-Vendor Architecture (MVA). By rewriting SAS in C, instead of PL/I, and implementing MVA with minimal hardware specific code, SAS Institute is able to implement changes more rapidly than in the past and to make updates to supported platforms within the same time frame.

This provides uniformity of the SAS product across hardware platforms; e.g., Macintosh, PC, UNIX, VM/CMS, MVS, AOS/VS, PRIMOS, or VMS (VAX). Thus if you learn SAS

The SAS System is designed for both the sophisticated and novice programmer; it is easy to learn and to use. You, however, are responsible for selecting the appropriate procedures and for interpreting the results; Virginia Tech SAS Support describes the types of support which are available at Virginia Tech. Although SAS is capable of analyzing simple and complex statistical designs, implementing matrix operations (PROC IML), and performing time-series analysis (SAS/ETS), it is not restricted to statistical applications. SAS can also be used for tasks unrelated to statistical analysis, such as report writing and managing data files.

In routine problems, you will typically perform only one analysis for each data set. However, a data set can be processed by any number of procedures in a single job stream. The system handles all the work involved in matching up a data set with a procedure.

New data sets can be created from existing data sets. These can be simple subsets of the data observations or subsets with transformed data elements. The union of several data sets is also possible.

Since many jobs are repeated executions of the same types of analysis, SAS provides a macro facility to reduce the labor in making control statements for each run. This allows the user to make only one set of control statements for a series of runs.