 |
|
|
|
Dear Readers,
August is busting out all over, and the Tour de France has ended, so it must be time for another newsletter! Following the tour this year meant keeping up with blogs and tweets and RSS feeds, and it reminded me that communication modes are ever-changing. Our new SAS discussion forums for several analytical areas are seeing increasing traffic and our RSS feeds are also popular with many users. Check them out and see what you think!
I am not presently using Twitter, although I suppose I could tweet from the upcoming NESUG and SCSUG conferences. But I’d rather my conference time go to as many in-person conversations with users as possible! If you are attending one of these conferences, please find me and let me know what you think about SAS/STAT® software. Save me from my Blackberry!
This issue completes a full year of newsletters. Welcome to the new readers from the 2009 JSM who found out about the newsletter at the SAS booth in Washington, DC. Our interview is with Rick Wicklin, one of the principal developers of SAS/IML® Studio software, which now enables you to submit R code and view the results in the SAS/IML Studio environment. Also learn about upcoming statistical tutorials at the SAS fall regional conferences as well as new Bayesian examples.
Enjoy!
Maura
P. S. All comments are welcome, particularly those about the types of information you want to see in upcoming editions.
|
|
An Interview with Rick Wicklin, Senior Research Statistician Developer »
Rick Wicklin is a Senior Research Statistician Developer in the Advanced Analytics Department at SAS. He began working here in 1997. Before his move to North Carolina, Rick received a PhD in applied mathematics from Cornell University in 1993. He also did a postdoctorate at the University of Minnesota, where he later worked as an assistant professor of mathematics. Rick is one of the principal developers of SAS/IML Studio, and he has also contributed to SAS software in the areas of numerical computations and graphics.
Read this interview about the life of a SAS/IML Studio developer.
|
|
R Interface Now Available in SAS/IML® Studio 3.2 »
SAS/IML Studio 3.2, formerly known as SAS Stat Studio, is now available. If you have a license for Base SAS®, SAS/STAT and SAS/IML software, you can download it directly from the SAS site. The SAS/IML Studio interface enables you to integrate R functionality with SAS/IML or SAS programs. You can also exchange data between the SAS System and R as data sets or matrices.
|
|
Call for SAS® Global Forum Papers »
Readers should consider submitting an abstract for the upcoming SAS Global Forum, to be held in Seattle, WA, April 11-14, 2010. It’s a great opportunity to present your work as well as meet other SAS users interested in the same areas. Attendees love to see statistical methodology put into practice with case studies, in addition to innovative uses of SAS software. Tyler Smith and Jennifer Waller are chairing the Statistics and Data Analysis section this year, and they are looking forward to the fast-approaching conference. They are particularly interested in papers on survival analysis, Bayesian analysis and statistical methods in clinical trials.
“We're in the process of finalizing our keynote speaker,” says Tyler. “Our attendees will be happy with our choice. Stay tuned!”
Abstracts will be accepted through Monday, Oct. 26, 2009.
|
|
New Discussion Forum »
Besides new discussion forums, “SAS Statistical Procedures” and “SAS/IML and SAS/IML Studio,” we’ve also added a discussion forum for you data miners: “SAS Data Mining and Text Mining.” Check it out!
|
|
New Product Pages on the Support Site »
A recent addition to support.sas.com is the availability of product pages, which are the starting places for a number of SAS software products, such as Base SAS, SAS/STAT, SAS/OR® and so on. The first edition of these pages includes 10 products, from SAS® Enterprise Guide® to SAS Text Miner. They provide single-page shopping for information related to that particular product: news, documentation, training, usage notes, samples and discussion forums. Quick links to installation information and available focus areas are also provided. Take a look and let us know what you think; the page comes complete with a survey!
|
|
Congratulations to Guixian Lin, New PhD!
Guixian Lin first came to SAS as the winner of the SAS Statistical Summer Fellowship in 2008, and he joined us as a full-time employee last winter. Guixian successfully defended his dissertation on censored quantile regression at the University of Illinois at the end of July. Guixian develops software in the area of nonparametric regression. Congratulations again, Guixian!
|
|
Ten Great Reasons to Upgrade to SAS/ETS® 9.2 »
We continue our top 10 list series with the top 10 reasons for upgrading to SAS/ETS software in SAS 9.2. Many customer sites are still making decisions about whether to move to SAS 9.2 and here are the top 10 reasons why we think econometricians can’t do without it!
|
|
Learn About Problems That SAS® Forecast Server Can Solve »
Often, time series data is intermittent (discontinuous or interrupted). The data points are mostly zero (the base value), with occasional departures. Commonly used forecasting techniques are continuous time series methods that are often inadequate for these problems. In contrast, intermittent demand methods (IDM) forecast the average demand per period, which is more appropriate. This paper discusses the advantages of IDM and demonstrates a technique and system of large-scale automatic forecasting of intermittent demand series using SAS Forecast Server.
|
|
HPMIXED Procedure for Analyzing Large Mixed Models »
SAS/STAT software provides a number of specialized features for analyzing data with mixed models, including the MIXED, GLIMMIX and NLMIXED procedures. These features provide a rich toolset for a wide variety of problems, small to large. However, it is common for applications of mixed models to go beyond large, becoming what can only be described as huge. In these situations, the mixed model equations might be too large to fit into computer memory, much less be solvable in a reasonable amount of time. However, the new HPMIXED procedure can enable you to perform these analyses with its special techniques that include sparse matrix methods. It complements the existing mixed model tools by supporting a subset of the models that can be fit with these techniques, while providing considerably better performance in terms of memory requirements and computational speed.
|
|
More Bayesian Examples That Use the MCMC Procedure »
A number of new examples of Bayesian analyses implemented with PROC MCMC have been added to the Resources section of the Statistics and Operations Research focus area. Look under “SAS/STAT Software” for applications such as hierarchical Poisson regression models, binomial model with power prior, missing data models and the exponential mixture models. The downloadable PDF files also include a link to the source code.
|
|
An Introduction to Similarity Analysis Using SAS® Software »
Web sites and transactional databases can collect large amounts of time-stamped data related to an organization’s customers or suppliers or both. Traditional data mining techniques can be hampered because each set of transactions might be quite large. An alternative approach is large-scale similarity analysis, which uses similarity measures combined with automatic time series analysis and decomposition. Data mining techniques can then be applied to the similarity results along with other profile data. This new paper demonstrates how to use these techniques with SAS/ETS software.
|
|
Using SAS/OR® Software to Solve the Student Assignment Problem »
School districts in areas with significant population growth encounter student assignment challenges each year. Even assuming availability of accurate, multiyear student population forecasts, these districts face the difficult task of balancing the often-conflicting goals of reducing overcrowding, travel time and socioeconomic imbalances among neighborhood schools. These goals must be achieved while adhering to numerous constraints, which include limits on student demographics by region and school-specific capacities, and the observance of predetermined special assignments. This paper describes the development of an optimization-based engine to solve this problem for the Wake County school district in North Carolina using the SAS/OR OPTMODEL procedure. Also, this paper explores the formulations and solutions developed using this approach, enabling school administrators to quickly develop alternative solutions and scenarios, thus helping policy makers see the impact of various strategies on student assignment.
|
|
Get the Goods on ODS Statistical Graphics »
Bob Rodriguez, Senior R&D Director of Statistical Development at SAS, has recently revised his 2008 SAS Global Forum paper Introducing ODS Statistical Software for the SAS 9.2 release. It includes updated information about the ODS Graphics Editor as well as a new gallery of graphs. It provides an excellent overview of the various ODS Graphics capabilities in SAS software. Download while it’s still hot off the press!
|
|
Getting Started Resources for SAS/STAT® 9.2 »
We’ve been getting good feedback on this resource for customers who are beginning to work with SAS 9.2, so I decided to leave this link here again for the newer subscribers. These pages include a collection of useful information and papers concerning important new capabilities in SAS/STAT 9.2.
|
|
Multiple Comparison Approaches for When a Pearson Chi-Square Statistic Is Significant »
No multiple comparison techniques are available in the FREQ procedure when your Pearson chi-square statistics indicate association in your table. However, you can pursue other strategies to understand the association, including correspondence analysis, performing chi-square tests on subtables and using PROC GENMOD to fit equivalent models and make the comparisons with CONTRAST statements.
|
|
The fall SAS users group conferences are coming up fast. Consider attending one near you. These conferences attract from 150 to 700 attendees and provide you with big conference information in a small conference setting. I have presented at all of them, and each year I really miss the ones not on my schedule. I particularly like getting to know attendees over the conference days, unlike at bigger conferences where connections are often drive-by in nature. The organizers do a great job, and we always get great feedback from our customers. This year, the statistical staff will be presenting various papers as well as these tutorials:
Creating Statistical Graphics with ODS in SAS 9.2:
Bob Rodriguez
San Jose, CA, WUSS, Sept. 1-3
Introduction to Bayesian Analysis Using SAS Software:
Maura Stokes
Burlington, VT, NESUG, Sept. 13-16
Creating Statistical Graphics with ODS in SAS 9.2:
Bob Rodriguez
Portland, OR, PNWSUG, Sept. 27-29
Introduction to Logistic Regression:
Bob Derr
Cleveland, OH, MWSUG, Oct. 11-13
An Introduction to SAS/IML Studio for SAS/STAT Users:
Rick Wicklin
Birmingham, AL, SESUG, Oct. 25-27
Introduction to Bayesian Analysis Using SAS Software:
Maura Stokes
San Antonio, TX, SCSUG, Nov. 9-10
|
|
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 |
|
