You now have convenient web access to the sample programs that are distributed with SAS® software. Currently, the SAS/STAT® and SAS/ETS® samples are uploaded and available through links on their respective product documentation pages. Examples are listed by procedure. We plan to make the samples for other products available soon.

SAS Global Forum 2013, to be held April 28–May 1 in San Francisco, offers a wealth of opportunities in the analytics arena. Statistics and Data Analysis section co-chair Tyler Smith says, “This year’s section offers over 40 papers in two concurrent sessions that keep you on the cutting edge.” To whet your appetite, consider these papers from SAS: “Introducing the New ADAPTIVEREG Procedure for Adaptive Regression” by Senior Manager Warren Kuhfeld, “Unleashing the Power of Unified Text Analytics to Categorize Call Center Data” by Software Developer Arila Barnes, “The Hospital Game: Optimizing Scheduling to Save Resources and to Save Lives” by Senior Business Solutions Manager Andrew Pease, and “Introducing the SAS/IML® Language: A Hands-On Workshop” by Research Statistician Developer Rick Wicklin. Four statistical tutorials set the stage on Sunday morning

This new interface facilitates time series exploration and segmentation prior to forecasting. Learn what it can do for you in this two-part video presentation from Principal Analytical Consultant Udo Sglavo, who provides an overview and a demonstration in two separate presentations.

Product Manager Ed Hughes provides an overview of the new features in SAS/OR 12.1 in this video presentation. This release provides updates in the areas of nonlinear optimization, linear optimization, mixed integer optimization, and network optimization, among others.

Bayesian analysis in SAS/STAT software is available through several standard procedures as well as through the general Bayesian modeling tool, the MCMC procedure. Using the BAYES statement in the GENMOD procedure gives you convenient access to Bayesian methods for techniques such as logistic regression, Poisson regression, and linear regression. This video describes the use of this method in the GENMOD procedure with a basic example.

Finite mixture models provide a parametric modeling approach to one-dimensional cluster analysis. This approach uses the fitted component distributions and the estimated mixing probabilities to compute a posterior probability of component membership. An observation is assigned membership to the component that has the maximum posterior probability. A benefit of using a model-based approach to clustering is that it permits estimation and hypothesis testing within the framework of standard statistical theory. This presentation illustrates an example of model-based clustering that uses the FMM procedure.

After the finite mixture model is fit to estimate the model parameters and the posterior probabilities of population membership, each observation is assigned membership to the population for which it has the highest estimated posterior probability of belonging. As with any method of cluster analysis, the practitioner faces the problem of assessing the accuracy of the population membership allocations that are obtained, because in practice, population membership is not observed. This example shows how to produce bias-adjusted allocation rate estimates by using the FMM procedure and a little SAS programming.

Rick Wicklin has pulled together the 12 most popular postings from his blog in 2012. These topics come from the areas of multivariate statistics, simulation, matrix computations, and data analysis.

Since I am a co-author of this one, I thought I would go first . . . I've never had so much fun as I did revising this book. It was last published in 2000, which was a very long time before the International Year of Statistics. The content has been updated to use SAS/STAT 9.3, and it also displays several features from SAS/STAT 12.1, all presented in a lovely format thanks to the adoption of a suitable ODS template. ODS graphics are another new feature, and you will see plots ranging from effect plots to agreement plots. New topics include exact Poisson regression, partial proportional odds models, Newcombe confidence intervals, and incidence density ratios. It might be worth a look if you work with categorical area.

No doubt, some of you have made a New Year's resolution to dig deeper into ODS statistical graphics. I can't think of a better way to learn about the statistical graphics (SG) procedures and the Graph Template Language (GTL) than to read this book. Written by Warren Kuhfeld, an insider who helped shape the use of ODS Graphics in the statistical division at SAS, this book teaches you about GTL by illustrating how you can generate the same plots by using an SG procedure or your own graph template. Thus, you learn about the capabilities of the SG procedures as well as how to produce more customized graphs whenever you want. The book is organized around graph types, such as scatter plots, regression fit plots, histograms, density plots, needle plots, and so on. I like the side-by-side plots, which make the comparisons easy. One chapter focuses on panels, and another one shows you how to modify ODS styles. A final chapter and appendices provide more details about the ins and outs of using GTL. Written for the data analyst, this book deserves a spot on your SAS shelf.

The following tutorials will be presented on April 28 at the 2013 SAS Global Forum in San Francisco.

• Modeling Categorical Response Data (Maura Stokes)
• Model Selection with SAS/STAT Software (Funda Güneş)
• Introduction to the MCMC Procedure in SAS/STAT Software (Fang Chen)
• Creating Statistical Graphics in SAS (Warren Kuhfeld)

Learn how to use the TTEST procedure to run several t tests for each pair of groups when you have more than two groups.