Chapter 15: Statistical Graphics Using ODS (Experimental)

Overview

Graphics are indispensable for modern statistical analysis. They enrich the analysis by revealing patterns, identifying differences, and expressing uncertainty that would not be readily apparent in tabular output. Effective graphics also add visual clarity to an analytical presentation, and they provoke questions that would not otherwise be raised, stimulating deeper investigation.

In SAS 9.1, a number of SAS/STAT procedures have been modified to use an experimental extension to the Output Delivery System (ODS) that enables them to create statistical graphics as automatically as tables. This facility is referred to as ODS Statistical Graphics (or ODS Graphics for short), and it is invoked when you provide the experimental ODS GRAPHICS statement prior to your procedure statements. Any procedures that use ODS Graphics then create graphics, either by default or when you specify procedure options for requesting specific graphs.

With ODS Graphics, a procedure creates the graphs that are most commonly needed for a particular analysis. In many cases, graphs are automatically enhanced with useful statistical information or metadata, such as sample sizes and p -values, which are displayed in an inset box. Using ODS Graphics eliminates the need to save numerical results in an output data set, manipulate them with a DATA step program, and display them with a graphics procedure.

The SAS/STAT procedures that use ODS Graphics in SAS 9.1 are listed on page 348. The plots produced by each procedure and any corresponding options are described in the procedure chapter. See the ODS Graphics subsection in the Details section of each procedure chapter for additional information.

In many ways, creating graphics with ODS is analogous to creating tables with ODS. You use

• procedure options and defaults to determine which graphs are created

• ODS destination statements (such as ODS HTML) to specify the output destination for graphics

Additionally, you can use

• graph names in ODS SELECT and ODS EXCLUDE statements to select or exclude graphs from your output

• ODS styles to control the general appearance and consistency of all graphs

• ODS templates to control the layout and details of individual graphs . A default template is provided by SAS for each graph.

In SAS 9.1, the ODS destinations that support ODS Graphics include HTML, LATEX, PRINTER, and RTF. These are discussed on page 326.

Both tables and graphs are saved in the ODS output file produced for a destination. However, individual graphs can also be saved in files, which are produced in a specific graphics image file type, such as GIF or PostScript. This enables you to access individual graphs for inclusion in a document. For example, you can save graphs in PostScript files to include in a paper that you are writing with LATEX. Likewise, you can save graphs in GIF files to include in an HTML document. With the HTML destination, you can also request an image map format that supports tool tip displays, which appear when you move a mouse over certain features of the graph.

In common applications of procedures that use ODS Graphics, the default graphs should suffice. However, when modifications become necessary, you can customize a particular graph by changing its template, or you can make consistent changes to all your graphs by selecting a different ODS style or by modifying an existing ODS style definition:

• As with table definitions, you can access graph template definitions and modify them with the TEMPLATE procedure. Graph template definitions are written in an experimental graph template language, which has been added to the TEMPLATE procedure in SAS 9.1. This language includes statements for specifying plot types (such as scatter plots and histograms), plot layouts, and text elements (such as titles and insets ). It also provides support for built-in computations (such as histogram binning ) and evaluation of computational expressions. Options are available for specifying colors, marker symbols, and other aspects of plot features.

• ODS style definitions include a number of graph elements that correspond to general features of statistical graphics, such as titles and fitted lines. The attributes of these elements, such as fonts and colors, provide the defaults for options in graph templates provided by SAS. Consequently, you can change all of your graphs in a consistent manner by simply selecting a different style. For example, by specifying the Journal style, you can create gray-scale graphs and tables that are suitable for publication in statistical journals.

Note: Statistical graphics created with ODS are experimental in this release, meaning that both their appearance and their syntax are subject to change in a future release.

This chapter illustrates the use of ODS Graphics, and it provides general information on managing your graphics. If you are unfamiliar with ODS, you will find it helpful to read Chapter 14, Using the Output Delivery System. For complete documentation on the Output Delivery System, refer to the SAS Output Delivery System User s Guide .

How to Use This Chapter

If you are trying out ODS Graphics for the first time, begin by reading the section Getting Started on page 321, which provides the essentials. Additional examples are given in the chapters for procedures that use ODS Graphics in SAS 9.1.

To take full advantage of ODS Graphics, you will need to learn more about ODS destinations, output files, and image file types for graphics, as well as ways to access and include individual graphs in reports and presentations. This is explained in the section Managing Your Graphics on page 326, the section Graphics Image Files on page 334, and the section Examples beginning on page 352.

If you need to customize a graph by modifying its template, read the section Customizing Graphics with Templates on page 338 and the series of examples beginning on page 363.

If you need to customize a style definition read the section Styles for Graphics on page 344 and the series of examples beginning on page 374.