Sanjay Matange is R&D Director in the Data Visualization Division responsible for the development and support of the ODS Graphics system, including the Graph Template Language (GTL), Statistical Graphics (SG) procedures, ODS Graphics Designer and related software. Sanjay has co-authored a book on SG Procedures with SAS/PRESS.
The ODS Graphics software, first released with SAS 9.2, supported creating graphs directly from statistical procedures. Prior to this, very few statistical procedures created graphs on their own, and in most cases creating graphs was a post process or creating the graphs from the saved data using SAS/GRAPH procedures. With
The SAS 9.4M6 software includes a new SGPIE procedure (preproduction) as introduced in the recent article - The SGPIE Procedure. In that article, I described the basic features of the two statements supported in the procedure, the PIE and the DONUT, with some examples. It is my humble opinion that
The 6th maintenance release of SAS 9.4 is now available for users. Along with it comes a new SG procedure - the SGPIE procedure (pre-production). The primary audience for the SG procedures has been the analytical user, for creating effective displays of analytical data. However, the rich feature set of these
A few days back I published an article on including more subject data in a Waterfall graph. This was motivated by the desire from many users to view more data in the same graph, thus avoiding the need to refer to different graphs for the relevant information. In this case,
Previously, I have written articles on how we could display more subject response data along with the traditional WaterFall graph showing the best "Change from Baseline". This is in response to the desire to see data like treatment duration for each subject in the same graph. Otherwise, investigators have to
The TEXT plot was introduced with SAS 9.4M2 to facilitate placement of text strings in a graph. This replaces the MARKERCHAR feature of the SCATTER plot statement, which is still available, but it is better to use TEXT plot in most cases. The syntax is: text x=column y=column text=column </
Recently a colleague was requested to create a bar chart showing the performance of students in a school compared to county and state wide schools. The solution using SGPLOT was simple as shown below. /*--Create data--*/ data school; input Year Group $ Value; format Value percent5.1; datalines;
It seems only a few months back I posted an article on creating Pie Charts using a GTL based macro. Well, looking back, that was almost 6 years ago!! Recently, a colleague here at SAS needed to create Pie Charts in his report along with other plots created using SGPLOT
The PharmaSUG 2018 conference was held two weeks ago, with great enthusiasm and a record attendance. The highlight for me was the large spread on visual presentation of data, from papers to poster presentations. I will provide a more detailed report on all the exciting graphical presentations. One topic of
A while back a user requested to create a 3D WaterFall chart as presented by E Castanon Alvarez et. al. in "3D waterfall plots: a better graphical representation of tumor response in oncology" Annals of Oncology, Volume 28, Issue 3, 1 March 2017, Pages 454–456. I posted a blog article titled
This article is motivated by a recent question on the SAS Communities board. This user wants to create a series or spline plot where the attributes of the line (color, thickness) can be changed based on another variable. In this case it may be a binary variable with "0" and
A common request on the communities page is to place data labels on horizontal bar charts. Often users want to display stacked horizontal bar charts, with the values displayed for each segment and the overall value of the bar itself as shown in the example below. In this example, the
Survival plots are automatically created by the LIFETEST procedure. These graphs are most often customized to fit the needs of SAS users. One way to create the customized survival plot is to save the generated data from the LIFETEST procedure, and then use the SGPLOT procedure to create your custom
A dual response axis chart is useful when the data type for the multiple measures are not compatible. For example, when overlaying measures like "This Year" sales with "Last Year" sales, the format and magnitudes of the two measures (or values for two groups) may be compatible, and it is
A few months ago, a user inquired about a chart that showed tumor response and treatment duration for each subject on 2 different planes of a 3D view. The data was really 2D, with one independent variable (the subject id) and two or more response values. I had provided an
This is the 9th installment of the "Getting Started" series, and the audience is the user who is new to the SG Procedures. It is quite possible that an experienced users may also find some useful nuggets here. In this article, we will cover the basics of the BUBBLE plot.
In my previous post, I described a new options to control the widths of the caps for Whiskers, Error and Limit bars. This topic could have been titled "Little things go a long way", as such details really make for a good graph. In a similar manner, another detail issue
The SG procedures and GTL statements do a lot of work for us to display the data using the specified statements. This includes setting many details such as arrow heads, line patterns etc, including caps. Often, such details have a fixed design according to what seems reasonable for most use
In the previous post, I discussed creating a 2D grid of spark lines by Year and Claim Type. This graph was presented in the SESUG conference held last week on SAS campus in the paper ""Methods for creating Sparklines using SAS" by Rick Andrews. This grid of sparklines was actually the
The 25th annual SESUG conference was held at in the SAS campus this week. I had the opportunity to meet and chat with many users and attend many excellent presentations. I will write about those that stood out (graphically) in my view. One excellent presentation was on "Methods for creating
The South East SAS Users Group meeting wound up yesterday. The 25th anniversary conference was held on SAS Campus and it provided a great opportunity to meet with many enthusiastic SAS users and attend many informative presentations. More on this in a follow-up article. During one of these presentations, Mary
Once in a while you run into a pesky situation that is hard to overcome without resorting to major surgery. Such a situation occurs when you have a stacked bar chart with a discrete legend positioned vertically on the side of the graph. A simple example is shown below. title
When a plot is classified by one or more variables, the different classes values are displayed in the graph either by position or by using different plot attributes such as color, marker shape or line pattern. For plots that display the visual by a filled area (bar, bin, band, bubble,
Plot statements included in the graph definition can contribute to the legend(s). This can happen automatically, or can be customized using the KEYLEGEND statement. For plot statements that are classified by a group variable, all of the unique group values are displayed in the legend, along with their graphical representation
Last year, a user asked about creating a "Turnip Plot" as used in this study of Caesarian Section Rates. Primarily, this is similar to a histogram on the y-axis for each unique value on the y-axis. A marker is drawn for each occurrence, starting from the center. Back then, I
The previous post on Multiple Blank Categories showed how to include multiple blank categories on the axis. But, given the purpose for this was to separate different segments in the data, I also included ideas on how to segmented a discrete axis using reference lines or Block Plot. A similar idea
Off and on, users have expressed the need to include multiple blank categories on a discrete axis. Often, this is desirable to separate groups of bars (or categories) in a graph due to some difference their definition. Such a case was discussed in this blog article on using non breaking
ODS Graphics procedures primarily strive towards the following goal: "Make simple graphs easy and complex graphs possible". SGPLOT procedure allows you create simple graphs with a single plot statement, and create complex graphs by layering together or combining multiple plot statements. Generally, the appearance follows the guidelines set by industry
SAS 9.4 maintenance release 5 was released on Sept 19, 2017. This release includes many new items including integration with SAS Viya and SAS Studio, a web application for SAS development. Also Included with this release are some cool new features in the graphics domain, some of which were requested
On a recent visit to an In-House Users Group meeting at a Pharmaceutical company, I presented a 1/2 day seminar on creating Clinical Graphs using SG Procedures. Polling the audience for their experience with these procedures indicated that many SAS users are not familiar with these new ways to create graphs. So,
