From Data to Graphs: Harnessing SAS/GRAPH for Effective Analysis

SAS/GRAPH, a component of SAS software, is frequently used in clinical trials to create insightful and visually appealing graphical representations of clinical data. These graphs are crucial in data analysis and presentation within the pharmaceutical and healthcare industries. 

SAS/GRAPH enables researchers and statisticians to generate charts, such as scatter plots, bar charts, survival curves, and more, to visualize trends, treatment effects, and safety profiles. These visualizations aid in interpreting clinical trial results, support regulatory submissions, and facilitate informed decision-making by clinical teams and regulatory authorities. 

SAS/GRAPH helps transform complex clinical data into clear and actionable insights, ultimately enhancing the efficiency and effectiveness of the clinical trial process.

What is the purpose of SAS/GRAPH?

Using SAS/GRAPH in clinical trials is leveraging its robust graphing and visualization capabilities to communicate and analyze complex clinical data effectively. SAS/GRAPH helps researchers and professionals in the pharmaceutical and healthcare industries to transform raw data into visually meaningful representations. These graphs and charts serve multiple purposes, including identifying treatment efficacy, safety profiles, and adverse events. They also assist in tracking patient outcomes over time, comparing different treatment groups, and presenting data to regulatory agencies.

SAS/GRAPH facilitates data-driven decision-making, enhances the clarity of clinical trial results, and plays a pivotal role in ensuring that critical findings are communicated accurately and comprehensively, ultimately contributing to advancing medical research and patient care.

In clinical trials, it is used for two purposes: Proc Gchart and Proc Gplot.

 Proc Gchart 2. Proc Gplot. These two procedures belong to SAS/Graph Product. 

1. Proc Gchart Charts like Hbar, Vbar, Pie, Star, Donut Etc., 

2. Proc Gplot like Scatter plot, Line plot, Odds-Ratio plot, Box-plot, Kaplan-Meier plot (KM Plot)

Proc Gchart (GCHART Procedure)

The GCHART procedure in SAS, also known as Proc Gchart, is a powerful tool for creating a wide variety of high-quality graphical charts and plots. It is primarily used for visualizing data and generating statistical graphics. Proc Gchart can produce bar charts, pie charts, scatter plots, histograms, and more, making it a versatile data analysis and presentation solution. The GCHART Procedure has vertical and horizontal bar charts, block charts, pie charts, star charts, donut charts, Vbar3d, Hbar3d, and Pie3d charts.

Researchers, analysts, and data scientists use Proc Gchart to explore data distributions, trends, and relationships, enabling them to gain deeper insights and make informed decisions. With customization options for labels, colors, and annotations, Proc Gchart allows users to create visually appealing and informative graphs, making it an essential component in the SAS software suite for data visualization and analysis.

The general form of the PROC GCHART statement

Proc Gchart Data= YLP. Dmads; 

Vbar Race; 

Run; 

Quit;

A bar or slice is generated for every distinct variable value in a character-based chart. If otherwise noted, it is assumed that the variables in a numerical chart are continuous. Midpoints are used to identify and automatically calculate intervals. Every midpoint has a single bar or slice built for it.

In the HBAR, VBAR, or PIE statement, use the DISCRETE option to override the default behavior for numeric chart variables. The values are no longer regarded as intervals when using the DISCRETE option; instead, a bar or slice is produced for each distinct numeric variable value.

The values are no longer regarded as intervals when using the DISCRETE option, which instead creates a bar or slice for each distinct numeric variable value.

Proc Gchart Data= YLP. Dmads;

Vbar Race/Discrete; 

Run; 

Quit;

By default, the statistic used to determine each bar's length or height or the pie slice's size is a frequency count (N). To override the default frequency count, use the following HBAR, VBAR, or PIE statement options:

Proc Gchart Data= YLP. Dmads; 

Vbar Race/ Discrete Sumvar=Ageyr; 

Run; 

Quit;

SUMVAR= analysis-variable TYPE=MEAN | SUM 

SUMVAR=SUMVAR stores the sum of a set of variables or values in a statistical analysis or programming language. It helps calculate the total or cumulative sum of the variables for further analysis or computations.

TYPE = Specifies that the height or length of the bar or size of the slice represents a mean or sum of the analysis-variable values.

To Present the Average Age by Race

 Proc Gchart Data= YLP. Dmads; 

Vbar Race/ Discrete Sumvar=Ageyr Type=Mean; 

Run; 

Quit;

If we want to present Race-wise Age in the 3D-bar chart, we want only Two Bars for Race1, 3. 

Proc Gchart data=YLP. Dmads; 

Vbar3d race/ Sumvar=Ageyr Midpoints=1, 3; 

Run; 

Quit;

When creating Pie or Pie3d Charts, we can specify additional options such as Fill=, Explode=, etc. The FILL= option allows you to decide whether the pie slices should be filled in a solid (FILL=S) or crosshatched (FILL=X) Pattern. Individual pie chart slices can be highlighted by dragging them away from the rest of the pie using the EXPLODE= option.

Proc Gchart data=YLP. Dmads;

Pie Race / discrete sumvar=Ageyr type=Mean

Fill=X Explode=2 Slice=Inside; 

Run;

Quit;

Proc Gchart Data=YLP. Dmads; 

Pie Race / Discrete Sumvar=Ageyr Type=Mean 

Fill=S Explode=2 Slice=Arrow Coutline=Green; 

Run; 

Quit;

GPLOT Procedure

The GPLOT procedure in SAS is a versatile and powerful tool for creating various graphical plots and charts. It is primarily used for producing two-dimensional plots, including scatter plots, line graphs, bar charts, etc. Proc GPLOT allows data analysts and researchers to visualize relationships and patterns in data, making it an essential component of data exploration and presentation.

General form of a PROC GPLOT step: 

PROC GPLOT DATA=SAS-data-set;

PLOT vertical-variable*horizontal-variable </options>; RUN;

QUIT;

While creating Charts/Plots, we use the following statements to create the Figure/ Graph as per the mock shell: 

♦ Goptions (Graphical Options) Statement

 ♦ Axis statement 

♦ Legend Statement 

♦ Symbol 

♦ Pattern 

♦ Title Statement

 ♦ Footnote Statement

GOPTIONS Statement

The GOPTIONS statement in SAS is used to customize the appearance and behavior of graphics output. It allows users to control aspects like colors, fonts, and scaling of the graphical production created by SAS procedures like GPLOT and GCHART.

General form of the GOPTIONS statement: 

GOPTIONS graphics-options; 

To reveal the devices provided by SAS: 

Proc GDevice;

Run;

General form of the GOPTIONS statement: 

GOPTIONS graphics-options; 

To reveal the devices provided by SAS: 

Proc GDevice;

Run; 

We can use the DEV= Option to utilize a particular device to comply with standards or customer requirements. The graphics device is specified by the GOPTIONS statement's DEVICE= (or DEV=) graphics option. 

Goptions Dev=JPEG (We generally use PNG, CGMOF97L, JPEG, GIF)

AXIS statement

The AXIS statement in SAS is used to customize and control the appearance and labeling of axes in graphical plots. It allows users to specify axis titles, labels, tick marks, and other attributes to make the presentation of data on graphs more informative and tailored to specific needs.

Different options we use in Axis statement: 

A. Axis scale options:- Order= (value-list) 

B. Appearance options: COLOR=Axis-color LENGTH=axis-length <units > 

OFFSET= (n1,n2) 

WIDTH= thickness-factor 

C. Tick mark options: Major=,

 Minor= D. Text options: LABEL= (angle= height= ) SPLIT=”split-char” 

VALUE= (text-argument(s) )

Legend Statement

The LEGEND statement in SAS is used within various graphical procedures to create legends or keys that explain the meaning of different graphical elements, such as colors, symbols, or lines, used in a plot. It helps viewers interpret the graph by clearly representing data categories or groups. Users can customize the legend's appearance, position, and content to enhance the clarity and readability of their graphs.

A. Appearance Options 

♦ ACROSS= Number-of-columns 

♦ CBORDER=Frame-color 

♦ FRAME

B. Position Options POSITION = (<BOTTOM | MIDDLE | TOP> 

<LEFT | CENTER | RIGHT> 

<INSIDE | OUTSIDE>)

C. Text Options 

♦ LABEL= (text-argument(s)) 

♦ ORDER= (value-list) 

♦ VALUE= (text-argument(s)) 

SYMBOL statement

The SYMBOL statement in SAS is used within procedures like GPLOT and GCHART to define symbols for data points in graphical plots. It specifies the shape, size, and color of symbols to represent data on the graph. 

General form of the SYMBOL statement: 

SYMBOL n options; 

The value of n can range from 1 to 99. If n is omitted, Default is 1. 

SYMBOL statements

Global: Once defined, they are in effect until they are changed or the SAS session ends. 

Additive: Determining the value of one option does not impact the values of other options.

The VALUE= option allows you to choose a specific plotting symbol representing your data points.

VALUE=symbol | V=symbol 

The default symbol value is plus; symbols include diamond, star, triangle, square, etc. You can use the I= option in the SYMBOL statement to draw lines between the data points. (I=interpolation) 

Selected interpolation values:

 ♦ JOIN ---Joins the points with straight lines.

 ♦ SPLINE -- Joins the points with a smooth line. 

 ♦ NEEDLE -- Draws vertical lines from the points to the horizontal axes.

 Symbol value=dot Interpol=needle;

Modifying the SYMBOL Statement: Set the attributes for SYMBOL1. Symbol1 C=blue V=diamond: You can cancel a SYMBOL statement by submitting a null SYMBOL statement. 

To cancel all SYMBOL statements:

 Goptions reset=symbol; 

PATTERN Statement: sets the Fill and Color of the patterns that are applied to different areas of plots and charts.

TITLE Statement: Adds Titles to graphics output 

FOOTNOTE Statement: Adds Footnotes to graphics output.

Adding Titles and Footnotes: Text string properties can be altered with the TITLE and FOOTNOTE statement options. 

♦ COLOR=color | C=color 

♦ FONT= type-font | F=type-font 

♦ HEIGHT= n | H=n

In conclusion, SAS/GRAPH is a powerful and versatile component of the SAS software suite that empowers users to create a wide range of informative and visually appealing graphical representations. From scatter plots to bar charts, SAS/GRAPH provides the tools to explore data, visualize trends, and communicate insights effectively. 

Its customization options, scalability, and integration capabilities make it a valuable resource for data analysts, researchers, and professionals across various industries. With SAS/GRAPH, data visualization becomes a dynamic and indispensable tool for decision-making and data-driven storytelling, contributing to more informed and impactful outcomes.