 |
|
|
 |
Dear Reader
I’m writing this between trips to SAS regional conferences. (OK, I didn’t make it to NESUG as I stated in the last newsletter, but some cartilage needed repair. The good news is that the crutches are history!) I enjoyed speaking at SESUG in Birmingham, AL, where I promoted SAS/STAT® 9.2 and was happy to see that more than half of the audience had access to our latest software at their site. I also provided an overview of our Bayesian capabilities. It’s always great to reconnect with some of the energetic organizers and meet new users, including students from Emory University and Georgia State University. Next is SCSUG, and then comes a local users group meeting in Columbus, OH, in December.
It’s rare that I attend one of these regional events and don’t come away with a new idea for inclusion in our software. My bet is that most users also pick up new ideas for using SAS products. Check out our Web site for a regional or local users group that is convenient to you.
Of course, the big event each year is SAS Global Forum, held in Seattle next April. As a member of the internal SAS planning committee, I’ve had my hands full of SAS abstracts since August. There simply isn’t room for everything we want to present, and the same goes for users. Indeed, I hear that user abstract submissions are up more than 11 percent from 2009. The tweeting has already begun, and #sgf10 is the magic wand for you social networkers.
This newsletter includes more information about SAS/IML® Studio, SAS Simulation Studio and an interview with Senior Research Statistician Michele Trovero.
Enjoy the material, and feel free to share it with your friends and relatives on Thanksgiving!
Maura
P.S. All comments are welcome, particularly those about the types of information you want to see in upcoming editions of this newsletter.
|
|
An Interview with Michele Trovero, Senior Research Statistician Developer »
Michele Trovero is a Senior Research Statistician Developer in the Economics Technology Department in SAS Research and Development. Michele hails from Italy, and he received a PhD in statistics from the University of North Carolina at Chapel Hill. He is one of the forecasting experts at SAS, and he works on SAS High-Performance Forecasting and SAS Forecast Studio software.
Read this interview and learn more about one of our econometrics developers.
|
|
Checking In with SAS/IML® Studio
Last issue, we ran an interview with Rick Wicklin, and hidden in there was the fact that Rick and co-author Robert Allison won the 2009 Data Expo poster competition sponsored by the ASA Statistical Graphics section. We thought you might be interested in learning more about this work and the graphics produced with SAS/IML Studio, so Rick wrote an
article
for us.
Rick has also provided an example of using
bootstrap methods in SAS/IML Studio
for the Samples & SAS Notes section of the SAS Customer Support Web site. This particular example illustrates how to use bootstrap methods to compute both a bootstrap estimate for a statistic and an estimate for the standard error of the statistic.
Finally,
documentation
for the R language interface available in SAS/IML Studio 3.2 is available for downloading.
|
|
Maintenance Release Updates »
This fall, the second maintenance for SAS® 9.2 was released to users. While primarily a vehicle for fixes, this release also provides production versions of the following SAS/STAT procedures, which were released as experimental in the original 9.2 release: MCMC, SEQTEST and SEQDESIGN. The documentation for these procedures has also been updated, and it can be obtained from the SAS/STAT documentation product page.
|
|
SAS Talks Webinar Series Launch
Last month, Bob Rodriguez was the brave first speaker of the new SAS Talks Webinar Series, which will offer technical presentations on a monthly basis. The live Webinar will be offered through the SAS users group organizations, and then the WebEx talk will be made available on-demand through the SAS Customer Support Web site. Bob gave a 45-minute presentation on “Getting Started with ODS Statistical Graphics in SAS 9.2,” which was viewed by more than 300 users. You can register to view Bob’s presentation now. Coming up are talks on SAS/IML Studio and SAS® Enterprise Guide® software, so stay tuned.
|
|
Group Sequential Analysis »
Group sequential trials provide for interim analyses before the end of a clinical trial. Thus, a group sequential trial is useful for preventing unnecessary exposure of patients to an unsafe new drug or to a placebo treatment if a new drug shows significant improvement.
New SAS/STAT procedures provide capabilities for group sequential analysis. The SEQDESIGN procedure creates group sequential designs by computing boundary values with a variety of methods, including the O’Brien-Fleming, Whitehead and error spending methods; it also provides required sample sizes. The SEQTEST procedure compares the test statistic with the boundary values at each stage so that the trial can be stopped to reject or accept the hypothesis; it also computes parameter estimates, confidence limits and p-values after the trial stops.
|
|
More Bayesian Examples That Use the MCMC Procedure »
More examples of Bayesian analyses implemented with PROC MCMC have been added to the Resources section of the Statistics and Operations Research Focus Area. New applications include hierarchical Poisson regression, power priors for binomial data, multivariate prior on linear regression and a multinomial model for ordinal response data. There are other useful existing examples for applications such as zero-inflated Poisson models, exponential mixture models, standardized regression to improve convergence, and missing data models. All of the Bayesian Web examples are now available in HTML and PDF format, along with the SAS program code.
|
|
Introducing SAS® Simulation Studio »
SAS Simulation Studio is a Windows 32-bit application based on Java software for building and working with discrete event simulation models. Its graphical user interface provides a full set of tools for building, executing and analyzing the results of discrete event simulation models. SAS Simulation Studio is designed to integrate with both SAS and JMP® software for statistical analysis of simulation results; it also can interface with JMP software to generate experimental designs.
|
|
Documentation for SAS® Add-Ins for JMP® Software »
Readers with access to JMP software might already know that you can submit code to the SAS System through a JMP interface. What you might not know is that JMP 8 includes SAS add-ins that enable you to use SAS procedures to analyze your data. Dialog boxes are available for performing some of the analyses that are available in the LOESS, QUANTREG, ROBUSTREG and TPSPLINE procedures in SAS/STAT software, and also for the AUTOREG procedure in SAS/ETS software. The results are displayed in a JMP window. The documentation also includes information about creating your own SAS add-in.
|
|
How to Set Reference Levels for the CLASS Predictor Variables in Your Model »
Some procedures (such as GENMOD, GLMSELECT and PHREG) provide options in their CLASS statements (or in other statements) that allow you to specify reference levels for the CLASS variables. However, many procedures (such as GLM, MIXED, LIFEREG and GLIMMIX) do not provide this functionality for CLASS variables. This usage note illustrates how to set reference levels for these different types of SAS procedures.
|
|
The Sunday Statistical Tutorials have already been set for SAS Global Forum in Seattle in 2010. Consider attending this event for the following tutorials in addition to many papers by users and SAS developers:
Introduction to Bayesian Analysis Using SAS Software:
Fang Chen
Introduction to the MCMC Procedure in SAS/STAT Software:
Fang Chen
Advanced Tutorial on Logistic Regression:
Maura Stokes
An Introduction to SAS/IML Studio for SAS/STAT Users:
Rick Wicklin
|
|
Contact Us
Have a question, comment or suggestion? Want to see something covered in the next SAS Statistics and Operations News? Simply e-mail the editor. |
| |
| Tech Support | Privacy Statement |
|
