 |
|
|
 |
|
Preparing for a Busy 2012 - New Analytical Product Releases, Conference Registration, and More!
It’s hard to believe it’s the end of the year. I hope yours is wrapping up nicely and that you have some rest and relaxation lined up. We’ve been busy here planning for the next releases of the analytical products, and development, testing, and documentation are well under way.
We’ve also planned our conferences for next year, and some of the upcoming new software features will be showcased at the 2012 SAS® Global Forum in Orlando, FL, in the spring. In February also in Orlando, we’ll be exhibiting at the new Statistical Practice conference, sponsored by the American Statistical Association. The conference is designed to provide statistical practitioners with best practices in statistical design, analysis, programming, and consulting. Bob Rodriguez, ASA president-elect and Senior SAS R&D Director, will be giving their keynote address.
This newsletter points you to yet another video on new features in SAS/STAT® 9.3, tips on using ODS Graphics, and a discussion of similarity analysis.
Happy Holidays!
Maura
R&D Research Director, Statistical Applications
|
Making Use of Incomplete Observations in the Analysis of Structural Equation »
The full information maximum likelihood (FIML) method of the CALIS procedure in SAS/STAT 9.22 and later enables you to use all the available information in your data to estimate your structural equation models. This paper illustrates how you can use PROC CALIS to deal with random missing values in the following data-analytic situations: (1) estimating means and covariances, (2) regression analysis, and (3) structural equation or path modeling. This paper also illustrates some new features of PROC CALIS for analyzing missing patterns and data coverage.
|
An Introduction to Similarity Analysis Using SAS® »
Web sites and transactional databases collect large amounts of time-stamped data related to an organization's suppliers and/or customers over time. Mining these time-stamped data can help business leaders make better decisions by enabling them to listen to their suppliers or customers via their transactions collected over time. A business can have many suppliers and/or customers and might have a set of transactions associated with each one. However, each set of transactions might be quite large, making it difficult to perform many traditional data mining tasks. This paper proposes techniques, using SAS/ETS® software, for large-scale similarity analysis that uses similarity measures combined with automatic time series analysis and decomposition. After similarity analysis, traditional data mining techniques can then be applied to the similarity analysis results along with other profile data.
|
Data Simulations Using SAS® »
If you are doing simulations in SAS without taking advantage of BY-group processing, you are dealing with some slow processing times. Long-time SAS user David Cassell’s paper on doing it the right way is a must-read for anyone who performs simulations in SAS. He shows how to perform bootstrapping, jackknifing, cross-validation, and simulations from established populations.
|
Modifying Your Survival Plot »
|
While ODS Graphics produces hundreds of graphs with little more than a glance (okay, sometimes you have to supply an option), you can customize the resulting graphs by making modifications to the underlying template. This post from the new Graphically Speaking SAS blog describes how to modify the default survival plot from the LIFETEST procedure by moving the "At Risk" table from inside the plot area to below the plot area. |
This blog provides insights on working with the Graphical Template Language and using Statistical Graphics procedures such as SGPLOT. Check it out!
|
Working with Statistical Distributions in SAS® »
|
|
Rick Wicklin's post from his The DO Loop SAS blog reviews the operations you use most when you work with statistical distributions in SAS. He discusses the PDF, CDF, QUANTILE, and RAND functions, and he provides useful examples of these functions in action. |
|
Fitting Bayesian Random Effects Using PROC MCMC – The Video! »
The MCMC procedure, first released in SAS/STAT 9.2, provides a flexible environment for fitting a wide range of Bayesian statistical models. Key enhancements in SAS/STAT 9.22 and 9.3 offer additional functionality and improved performance. The RANDOM statement provides a convenient way to specify linear and nonlinear random-effects models along with substantially improved performance. This video describes the new RANDOM statement in SAS/STAT 9.3 and illustrates its use with an example.
|
Funnel Plots for Proportions »
I also really liked Wicklin's blog post about creating funnel plots for proportions, a follow-up to a general discussion of these plots. Wicklin describes the use of SAS/IML® software and the SGPLOT procedure to create a display like this one:
|
|
|
SAS Global Forum 2012 »
SAS Global Forum 2012 is just around the corner. Several two-hour statistical tutorials will be held on Sunday morning:
• Introduction to Bayesian Analysis Using SAS Software
• Creating Statistical Graphics with ODS in SAS
• SAS Procedures for Analyzing Survey Data
• Data Simulation for Evaluating Statistical Methods in SAS
|
|
Fitting Truncated Poisson and Negative Binomial Models »
Count data in which zero counts cannot be observed is called truncated count data. Such data can be modeled using truncated versions of the Poisson or negative binomial distributions. Models for truncated data using the truncated Poisson and truncated negative binomial distributions can be fit using PROC NLMIXED by specifying just the log likelihood function. This note illustrates this use of PROC NLMIXED with two examples. Users with access to the experimental FMM procedure in SAS/STAT 9.3 might want to try out the truncated Poisson distribution using that procedure.
|
|
|
