Featured news from SAS.


View as a web page
SAS The Power to Know logo

I’ve been on the road a bit this fall, and it’s always great to meet users, get their feedback, and tell them about our latest software. The Western Users of SAS® Software organization put on another great conference in Long Beach, California, and the Boston Area SAS Users Group hosted a closed-out SAS Analytics speakers day at the end of September that was great fun. If you live in these areas, I’d encourage you to attend one of their events. If not, check out the regional, state, or company SAS users groups that are convenient for you. The tips, current information, and user connections are well worth the effort. And if you are thinking about providing professional presentations yourself to share your knowledge, these events are a great place to start! Mentoring is often available, as are scholarships for students and junior SAS professionals.

The SouthEast SAS Users Group (SESUG) is holding its annual conference at the SAS campus in Cary, November 5–7, so those of you in the Triangle area have no excuse!

In addition, note that SAS will be represented at the 50th Winter Simulation Conference in December, so come find us if you attend!

Today’s newsletter provides information about the recently released SAS® 9.4M5 and a number of new videos, along with numerous tips on how to get the most out of your SAS analytical software.

Until the next newsletter!

Maura Stokes

Senior R&D Director, Statistical Applications


Technical Highlights


SAS 9.4M5 Now Available

SAS 9.4M5 was recently made available and includes the new SAS/STAT® 14.3 and SAS/ETS® 14.3 releases. SAS/STAT 14.3 includes a new procedure for causal mediation, along with cause-specific proportional hazards regression for competing-risks data in the PHREG procedure, as well as other new features. SAS/ETS 14.3 provides the new TMODEL procedure, which incorporates high-performance techniques and other new features that enhance the functionality of the MODEL procedure. Learn more about the highlights of all the updated analytical products


Video Camera - Icon

Growing Video Library

Here’s a plug for our ever-growing library of videos, mostly slide presentations with voiceover, that provide up-to-date information about features and applications of the SAS analytical products. We’re adding new videos all the time, and we list recent additions later in this newsletter. The videos are also available on YouTube, on the Statistics and Operations Research YouTube Channel and the Data Mining and Text Analytics YouTube Channel. View, share, and repeat!


Blue blog icon

The DO Loop

There are many ways to think about correlation. Distinguished Research Statistician Developer Rick Wicklin describes seven of them in this interesting post. Wicklin also discusses how to use the EFFECT statement to construct polynomial effects in regression and how to use singular value decomposition to construct a low-rank approximation to an image.


Blue blog icon

Graphically Speaking

In recent posts, Distinguished Research Statistician Developer Warren Kuhfeld discusses new means and LSMEANS plots, as well as answering a customer’s question about how to put a graph within a graph.


Technical Papers


May I Direct Your Attention to: The FASTQUAD Option in the GLIMMIX Procedure

One issue that often comes up with users is what to do when you have multilevel models in the GLIMMIX procedure where the number of random effects and/or nesting makes the model fitting very slow and sometimes impossible. The FASTQUAD option in the GLIMMIX procedure, available beginning in SAS/STAT 14.1, solves this problem in many cases by using an approximation method in the quadrature algorithm. The example in the documentation illustrates an application.


Heat Maps: Graphically Displaying Big Data and Small Tables

Heat maps use colors to communicate numeric data by varying the underlying values that represent red, green, and blue (RGB) as a linear function of the data. You can use heat maps to display spatial data, plot big data sets, and enhance tables. You can use colors on the spectrum from blue to red to show population density in a US map. In fields such as epidemiology and sociology, colors and maps are used to show spatial data, such as how rates of disease or crime vary with location. With big data sets, patterns that you would hope to see in scatter plots are hidden in dense clouds of points. In contrast, patterns in heat maps are clear, because colors are used to display the frequency of observations in each cell of the graph. Heat maps also make tables easier to interpret. For example, when displaying a correlation matrix, you can vary the background color from white to red to correspond to the absolute correlation range from 0 to 1. You can shade the cell behind a value, or you can replace the table with a shaded grid. This paper shows you how to make a variety of heat maps by using PROC SGPLOT, the Graph Template Language, and SG annotation.


Five Things You Should Know about Quantile Regression

The increasing complexity of data in research and business analytics requires versatile, robust, and scalable methods of building explanatory and predictive statistical models. Quantile regression meets these requirements by fitting conditional quantiles of the response with a general linear model that assumes no parametric form for the conditional distribution of the response; it gives you information that you would not obtain directly from standard regression methods. Quantile regression yields valuable insights in applications such as risk management, where answers to important questions lie in modeling the tails of the conditional distribution. Furthermore, quantile regression is capable of modeling the entire conditional distribution; this is essential for applications such as ranking the performance of students on standardized exams. This expository paper explains the concepts and benefits of quantile regression, and it introduces you to the appropriate procedures in SAS/STAT software.


Advanced Hierarchical Modeling with the MCMC Procedure

Hierarchical models, also known as random-effects models, are widely used for data that consist of collections of units and are hierarchically structured. Bayesian methods offer flexibility in modeling assumptions that enable you to develop models that capture the complex nature of real-world data. These flexible modeling techniques include choice of likelihood functions or prior distributions, regression structure, multiple levels of observational units, and so on. This paper shows how you can fit these complex, multilevel hierarchical models by using the MCMC procedure in SAS/STAT software. PROC MCMC easily handles models that go beyond the single-level random-effects model, which typically assumes the normal distribution for the random effects and estimates regression coefficients. The paper shows how you can use PROC MCMC to fit hierarchical models that have varying degrees of complexity, from frequently encountered conditional independent models to more involved cases of modeling intricate interdependence. Examples include multilevel models for single and multiple outcomes, nested and non-nested models, autoregressive models, and Cox regression models with frailty. Also discussed are repeated measurement models, latent class models, spatial models, and models with nonnormal random-effects prior distributions.


Tech Support Points Out

Tech Support Points Out icon

Fitting and Assessing Proportional Odds Models to Ordinal Responses

For an ordinal, multinomial response (such as low, medium, high), a set of cumulative response functions are simultaneously modeled. In an ordinal logistic model, the fully unrestricted model has a different parameter vector for each of the cumulative logit response functions. This is the nonproportional odds model. The much simpler proportional odds model restricts the parameters on each model effect to be equal so that only the intercepts vary across the parameter vectors. This is the model that PROC LOGISTIC fits by default. In between lie partial proportional odds models that restrict the parameters of only some effects to be equal. Two notes are available to help you assess whether the proportional odds restriction should apply to each model effect. Both graphical and test-based methods are discussed. “Plots to assess the proportional odds assumption in an ordinal logistic model” introduces code and a macro that uses the observed data to produce plots for assessing the assumption. Tests of the assumption are presented in “The PROC LOGISTIC proportional odds test and fitting a partial proportional odds model.” In addition to testing the effects in a specified model, the note shows you how to use model selection methods to find a model that selects effects with or without the proportional odds restriction based on tests.


Talks and Tutorials


Southeast SAS Users Group (SESUG) 
November 5–7, 2017
Cary, NC

Winter Simulation Conference
December 3–6, 2017
Las Vegas, NV 



Quick Links


SAS® Statistics and Operations Research News

Don't miss important updates from SAS! Please add sas.com as a domain in your safe sender list.
Want to get more out of your relationship with SAS? Update your SAS profile

About this e-mail:
If you do not wish to receive future SAS® Statistics and Operations Research News editions, please unsubscribe.
SAS places great value in fair information practices and in connection with the management of our contact database, we would like to remind you that SAS Institute Inc. may use your personal contact details for marketing purposes, as stated in its privacy policy. To contact SAS via postal mail: SAS, SAS Campus Drive, Cary NC 27513 USA. ATTN: Legal Division/Privacy Manager.

SAS and all other SAS Institute Inc. product or service names are registered trademarks or trademarks of SAS Institute Inc. in the USA and other countries. ® indicates USA registration. Other brand and product names are trademarks of their respective companies. Copyright © SAS Institute Inc. All rights reserved.