Icebreaker
Keynote Talk: Contrasting programming techniques for
summarizing voluminous SAS output using the SAS Output Delivery System (ODS)
(PROC FREQ as an example)
by: Stuart Long, Westat
Abstract: SAS® ODS provides programmers with the ability to extract selected information from a procedure and store it in datasets. Such datasets can then be combined to summarize the results from numerous procedures. The SAS Macro facility can be used to execute and extract information from repetitively called SAS “Macro Variable Arrays” can simplify the extraction of information from SAS procedures when identical or similar code is executed repeatedly on a multitude of datasets and/or variables. This paper describes Macro Variable Arrays and contrasts these with other programming methods to obtain summary datasets from ODS output.. Although PROC FREQ is used in this paper, these techniques can serve as a blueprint for extracting and summarizing output from other SAS procedures. This is an advanced tutorial intended for SAS programmers knowledgeable with the SAS Macro facility.
Networking Session
Where to Get SAS® Help When the Local Goo-Roo Is Out
Talk by: Earl Westerlund, University of Rochester
Abstract: Most SAS shops have one person who seems to know a bit more than the rest of the programmers. The go-to person who can get you over, through or around the wall when you hit it. The Goo-Roo. This begs the question: Where does the Goo-Roo go for help when s/he is stuck? This talk will survey the variety of SAS help sources, from SAS's own website to email communities. Who knows: if you avail yourself of enough of these resources, you may become a Goo-Roo yourself!
Macros that
do Windows, Data and other tricks.
Talk by: Michael E. Tomb,
Informatica ECS, Inc.
Abstract: This is an overview of some of the advanced macro techniques that are enabled using the %SYSFUNC function. But while the code may be considered advanced, the talk will demonstrate several tools that will be shared with the audience and can be taken home and run without hiring a macro guru. The unusual access methods highlighted will include reading SAS dataset observations directly from macro code, calling windows libraries and code that dynamically reacts to option settings.
Refreshment Break
Ask the Experts
Univariate Outlier Detection Using a Macro based on the IML Routines LTS and LMS
Talk by: John Hixon, Eastman Kodak Company
Abstract: There are SAS macros available on the web for Multivariate outlier detection. One example is by Professor Michael Friendly of York University. But, last year, when I needed a macro to scrub datasets containing many Univariate Responses, I could not find a useful macro anywhere on the web. This led me to write a macro to perform Univariate outlier detection. This was before the advent of SAS9 with its new Proc RobustReg, which may be another good choice. Though the macro has its limitations, it has served me well in practice. It will reliably detect multiple outliers in a single pass, which is its key feature. The macro relies on the Proc IML functions "LMS" (Least Median of Squares) and "LTS" (Least Trimmed Squares). These methods are mainly attributed to the work of Rousseeuw, P.J. The macro includes diagnostic plots to allow the user to review the results of the outlier deletion, as well as an output dataset that includes the raw and 'scrubbed' data. Examples using real as well as contrived data will be shown. The examples will illustrate the macro's usefulness (as well as its limitations)
Optional: SAS and SUDS and
ZEBB'S
7:00 PM (or so)
Zebb's Rochester
1890 South Clinton Avenue
Brighton , New York
see map
Click Here for Meeting Registration...
Last modified: March 1, 2005