Featured news from SAS.


View as a web page
SAS The Power to Know logo
SAS Statistics and Operations Research News

I hope your holiday season is off to a good start. To put us in the mood here in Cary, we had a serious snowfall yesterday—almost nine inches in my backyard!

We also have gifts for you: new features in the 15.1 analytical release of SAS/STAT®, SAS/ETS®, and SAS/QC® software, which was made available at the end of November with the M6 maintenance release of Base SAS® 9.4. And as they say, it will be the gift that keeps on giving, as you learn more about the capabilities that can help you with your data analysis. (See below for more information.) Of course we will present papers about this new work, as well as about new SAS® Viya® work, at the next SAS® Global Forum in Dallas in April.

In particular, check out the upcoming Global Forum tutorials, a great way to dig in a bit more than you can with a paper presentation.  

Finally, note that I will soon be retiring from SAS and will hand over the newsletter reins to Phil Gibbs, a longtime manager in Analytical Technical Support and a longtime comrade in arms. If you have recently attended the Joint Statistical Meetings, you probably met Phil in the booth—he was the one who knew the most stuff!

It’s been a real pleasure to talk with you this way, and I hope that the information and pointers to resources and tips have been useful. Customers are the reason we do what we do, and please keep in mind that it’s a two-way street. Let us know what you need for your work, and we’ll make it our work. Just this past release we were able to incorporate a suggestion in SAS/QC software made by a user at the Dartmouth Area Users Group conference that I addressed in July. And if you need help with our software, please contact Tech Support. We are fortunate to have the best, most experienced support crew for analytical software that exists on this planet. I can’t speak for the galaxy, but I’ll bet it’s no different. Take advantage!

Thanks, all, and here’s to a wonderful holiday season and a great 2019.

All the best,


Senior R&D Director, Statistical Applications


Technical Highlights


Highlights of the SAS/STAT 15.1 Release

This new release really does have something for everyone. Right—we say that all the time, but . . . The BGLIMM procedure provides full Bayesian inference for generalized linear mixed models, using the mixed model syntax you already know. The CAUSALGRAPH procedure extends our offerings for causal inference to include examining the structure of graphical causal models. And the RMSTREG procedure analyzes time-to-event data by using regression with response to the restricted mean survival time (RMST). You can also fit the semiparametric proportional hazards model to interval-censored data (ICPHREG), perform Bayesian analysis of the proportional hazards spline mode, and do counterfactual analysis using quantile regression.

For more information, see the chapter “What’s New in SAS/STAT 15.1” in the online documentation.


blog icon

The DO Loop

Distinguished Research Statistician Developer Rick Wicklin shows you how to resample residuals to obtain bootstrap regression estimates, fit the Pareto distribution in SAS®, and create and compare ROC curves for any predictive model.


SAS® Enterprise Miner Gradient Boosting Tip

Note that the SAS Communities offer many active analytical forums for getting help from other SAS customers and from SAS staff. Explore and see if they could be a useful resource for your work.

The Gradient Boosting node, available on the Model tab of the SAS Enterprise Miner toolbar, enables you to train a gradient boosting model, which is created by a sequence of decision trees that together form a single predictive model. A tree in the sequence is fit to the residuals of the predictions from the earlier trees in the sequence. The residuals are calculated in terms of the derivative of a loss function. The resulting ensemble model, which averages together the predictions from the decision trees, often outperforms (in terms of prediction accuracy) other machine learning algorithms, making gradient boosting extremely popular.


Pointing Out References—SAS for Mixed Models: Introduction and Basic Applications

Keep an eye out for this new book by George Milliken, Elizabeth Claassen, Walter Stroup, and Russ Wolfinger on mixed models analysis with SAS, published this month. The book covers the latest capabilities of a variety of applications featuring the SAS GLIMMIX and MIXED procedures. Written for instructors of statistics, graduate students, scientists, statisticians in business or government, and other decision makers, SAS for Mixed Models: Introduction and Basic Applications is the perfect entry point for those with a background in two-way analysis of variance, regression, and intermediate-level use of SAS. Topics include random-effects-only and random-coefficients models; multilevel, split-plot, multilocation, and repeated measures models; hierarchical models with nested random effects; analysis-of-covariance models; and generalized linear mixed models.



Technical Papers


SAS/STAT 14.3: Roundup: Modern Methods for the Modern Statistician

Before you dive into the 15.1 release of SAS/STAT software, you might want to catch up on what you have now. The 14.3 release of SAS/STAT software has things that everyone can use. The new CAUSALMED procedure performs causal mediation analysis for observational data, enabling you to obtain unbiased estimates of the direct causal effect. You can now fit compartment models for pharmacokinetic analysis with the NLMIXED and MCMC procedures. In addition, variance estimation by the bootstrap method is available in the survey data analysis procedures, and the PHREG procedure provides cause-specific proportional hazards analysis for competing-risks data. Several other procedures have been enhanced as well. Learn about the latest methods available in SAS/STAT software that can modernize your statistical practice.


Propensity Score Methods for Causal Inference with the PSMATCH Procedure

Probably the most popular new procedure we’ve introduced in some time is the PSMATCH procedure, which has drawn huge interest in the last year. This paper from 2017 introduces the procedure.

In observational studies, special statistical approaches that adjust for the covariate confounding are required in order to obtain unbiased estimation of causal treatment effects. One strategy for correctly estimating the treatment effect is based on the propensity score, which is the conditional probability of the treatment assignment given the observed covariates. Prior to the analysis, you use propensity scores to adjust the data by weighting observations, stratifying subjects that have similar propensity scores, or matching treated subjects to control subjects. This paper reviews propensity score methods for causal inference and introduces PROC PSMATCH, which was new in SAS/STAT 14.2. The procedure provides methods of weighting, stratification, and matching. Matching methods include greedy matching, matching with replacement, and optimal matching. The procedure assesses covariate balance by comparing distributions between the adjusted treated and control groups.


Optimization Modeling with Python and SAS Viya

Python has become a popular programming language for both data analytics and mathematical optimization. With SAS Viya and its Python interface, Python programmers can use the state-of-the-art optimization solvers that SAS provides. This paper demonstrates an approach for Python programmers to naturally model their optimization problems, solve them by using SAS® Optimization solver actions, and view and interact with the results. The common tools for using the optimization solvers in SAS for these purposes are the OPTMODEL and IML procedures, but programmers more familiar with Python might find this alternative approach easier to grasp.


Tech Support Points Out

Tech Support Points Out icon

Tech Support was in a bonus mood!

Usage Note 60335: Choice of Continuous Response Distribution in Log-Linked GLMs

There are several response distributions available when you are fitting generalized linear models (GLMs) in procedures such as PROC GENMOD, PROC GLIMMIX, and others. The possible distributions are the normal, Poisson, gamma, inverse Gaussian, and Tweedie, among others. One important aspect of how these distributions differ is in their relationship between the mean and the variance. When you are selecting a distribution to use in modeling a response, it is important to choose a distribution that matches the observed mean-variance relationship. This note discusses some tools for finding a suitable response distribution if it is not already known, including graphics and an analytical method that assumes that the variance is proportional to a power of the mean, V(y)=φμp. Also discussed is the use of the Tweedie distribution, which can accommodate a range of powers, p.

Usage Note 63275: Combining the Results from PROC SURVEYPHREG in PROC MIANALYZE

When there are CLASS variables in the model, the ODS table ParameterEstimates produced by PROC SURVEYPHREG does not match any of the available CLASSVAR= formats that PROC MIANALYZE requires. However, with minimal code in a DATA step, the data set can be converted to a format that PROC MIANALYZE can readily use.


Talks and Tutorials


SAS Global Forum 2019

The following tutorials by R&D staff are on tap for SAS Global Forum 2019:

  • Introduction to Logistic Regression
  • Survey of Missing Data Analysis Using SAS
  • Advanced Methods for Survival Analysis Using SAS
  • Causal Analysis with Observational Data: Methods and Applications
  • Quantile Regression in Practice
  • Introduction to Modern Machine Learning Techniques in SAS® Visual Data Mining and Machine Learning

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.