Date: Fri, 7 Sep 2001 10:25:40 -0400
Reply-To: NYASUG@pace.edu
Sender: "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From: NYASUG <NYASUG@PACE.EDU>
Subject: NYASUG Meeting -- Reminder # 1
Content-Type: text/plain; charset=us-ascii
Reminder #1 -- Have A Great Weekend !
NYASUG Meeting Announcement !
For members of SAS-L in the Tri-State area who are also
members of the New York Area SAS Users Group (NYASUG):
The next meeting of the New York Area SAS Users Group
will be on Wednesday, September 12, 2001. This will be
an all day meeting with 7 scheduled presentations and
will be a joint effort with the NY Chapter of The
American Statistical Association.
This meeting is being held at the Bank of New York, and
we thank them for hosting the meeting ! Directions to
the BONY site are at the end of this text.
The following are abstracts of the scheduled
presentations:
-------------------------------------------------------
Hidden Passageways in SAS/GRAPH Software
By Earl Westerlund, Kodak
The procedures in SAS/GRAPH software have many
features; they have many shortcomings as well. The
good news is that the software product also provides us
with ways to get around those shortcomings. The bad
news is that these ways around are often quite obscure
and difficult to find. This talk will discuss a random
collection of bits of not-so-obvious information that
the author has stumbled across while trying to get
SAS/GRAPH software to do things nature (or at least
SAS) never thought of.
An Introduction to the DATA Step Graphics Interface
By Earl Westerlund, Kodak
The DATA Step Graphics Interface (DSGI) is a component
of SAS/GRAPH software that enables you to generate
graphics output from within a DATA step, macro, or SCL
application. It consists of low-level graphics
routines that allow you to develop a custom graph or to
add elements to an existing graph. This talk will
describe the steps involved in creating a graph in the
DATA step, from opening a session to storing a graphic
"segment" in a catalog. A brief comparison of DSGI
with other options for creating graphics will be given.
Biography
Earl Westerlund's first SAS program was keypunched on
Hollerith cards and run on an IBM 370 computer;
SAS/GRAPH hadn't been invented yet. Earl is a
Technical Associate at Eastman Kodak Company in
Rochester, and has a BS in math from Rochester
Institute of Technology. His areas of SAS expertise
are SAS/GRAPH and the macro language. He has presented
at numerous SUGI and NESUG conferences. One of his
macro papers was selected for inclusion in the SAS
Institute publication "SAS Macro Facility Tips &
Techniques." He is also experienced in developing
applications with Visual Basic, Excel VBA, JMP
Scripting Language, Perl and JavaScript.
-------------------------------------------------------
Hidden Nuggets in Version 8: New Informats, Formats,
and Functions
By Mike Rhoads, Westat
Much of the interest in Versions 7 and 8 of the SAS.
system has been in major new features such as the
Output Delivery System and long variable and table
names. Hidden within the Changes and Enhancements
documentation, however, are many other useful new
items. This presentation summarizes several
enhancements to SAS informats, formats, and functions,
and describes how they can make life easier for
programmers who know about them.
Some Practical Ways to Use the New SAS Pattern-Matching
Functions
By Mike Rhoads, Westat
The 6.11 and 6.09E releases of SAS. software
introduced, in experimental form, a powerful new set of
pattern-matching functions, which can be used in DATA
steps, macros, or SCL code. This presentation provides
a summary description of these new routines and
includes several examples of how their use can
significantly reduce the coding effort required to
perform common tasks.
Mike Rhoads has served on DCSUG's steering committee
for the past several years and chaired its Discover DC
committee for NESUG '99. He has also contributed to
various SUGI and NESUG conferences as a paper presenter
and session coordinator. The year 2001 marks the 22nd
anniversary of his initiation into the wonderful world
of SAS software. He is currently a Vice President in
the Computer Systems and Applications group at Westat,
where he still manages to sneak in an occasional DATA
step or two when not overburdened by his other
responsibilities.
-------------------------------------------------------
PROTEOMIC COWS: Using SAS to Manage and Run Parallel
Clusters for Gene Searching
By Haftan Eckholdt, Albert Einstein College of Medicine
Using SAS to manage and run a LINUX / WINDOWS-NT based
cluster of workstations (COW) to simulate a parallel
processor puts the price of supercomputing well within
most research budgets. The criteria for using COWs
will be discussed in terms of their job symmetry and
intermediate solution timing. Special problems and
their solutions are discussed, including the fact that
all COWs need: lots of smoothed electricity, lots of
cool breezes; and lots of hard drives. Using three
layers of file monitoring, job distribution, and data
analysis, SAS can provide clever and cost effective
paths to COW management and usage [X command, %MACRO,
DATA, STAT].
Parsing the problem, distributing it symmetrically
across the COW, and executing the solution, are
discussed in the context of an example from genetics: a
search for a family a proteins in the entire genetic
sequence of a worm, Caenorhabditis elegans, initially
referenced from the fly, Drosophila melanogaster.
Using a modified distance profile algorithm to assess
amino acid periodicities, 12,000 proteins were
characterized and ranked to yield 23 familial
candidates.
Simulating Attrition and Visual Models of Type I and
Type II Errors
By Haftan Eckholdt Albert, Einstein College of Medicine
Longitudinal studies are plagued by attrition. While
statisticians can address these problems, researchers
have trouble understanding the impact of attrition.
Methods described here are designed to help scientists
understand and appreciate the impact of attrition on
their hypotheses. Starting with the essential
hypotheses, the approach begins by characterizing the
sample at each stage of research to date: population of
origin, recruitment, baseline, first follow up, etc.
[BASE & STAT: PROC FREQ, PROC TABULATE, etc,]. The
next step executes 1000's of simulations of attrition
through the random as well as systematic deletion of
cases [%MACRO]. In the next step, models are run
[STAT: PROC GLM, PROC MIXED, etc.] and relevant
coefficients are saved [ODS]. Finally, distributions
of coefficients (observed, expected, random deletion,
and systematic deletion) are compared [GRAPH: PROC
GPLOT].
This process will help scientists see how their
interpretations of hypothesis tests are likely to
change over time, assuming that attrition (1) never
occurred, (2) occurred at random, (3) occurred in some
systematic way. This demonstration will include data
from several longitudinal studies assessing relevant
changes in women and the peri-menopause.
Lost in Space: Recovering Unmeasured Dimensions in
Spatial Studies
By Haftan Eckholdt, Albert Einstein College of Medicine
Research on spatiotemporal events are often missing one
or more dimensions due to oversight or expense. In
some cases, the recovery of those dimensions addresses
existing hypotheses, and in some cases dimensional
recovery brings about entirely new and unforeseen
hypotheses. SAS can be used to recover space and time
through a combination of geometry and fuzzy merges
[DATA STEP] that produces 3-D objects and 4-D
animations from 2- and 3-dimensional recordings.
This approach begins with an elaborate version of the
original event: conceptualizing perspectives and
reviewing the role of time. Logical geometric
reconstructions of missing dimensions can be built upon
information from existing data in measured dimensions
[BASE & STAT]. Similarly, objects can be recombined
through the use of fuzzy spatial merges in the new
geometry. Animation techniques can then be used to
build, melt, and surgurize the new data object in
virtual ways [GRAPH: gifanim].
This presentation will include examples from
neuroscience of (1) data on bodies in motion recorded
from one camera providing data in 2-D over time used to
recover the third dimension for assessing full body
orientation, and (2) data on anatomical structures from
electron micrographs providing data in 2-D used to
recover the third dimension that could then be
surgurized in new dimensions.
Both examples come from laboratories where scientists
begin with a priori hypotheses that are enhanced by the
addition of unmeasured dimensions. In the end,
researchers come to understand new classes of
hypotheses that could not be formulated, much less
tested, prior to dimensional reconstruction.
Biography
Haftan Eckholdt has degrees in human development and
biostatistics and is Director of the Biometry Core of
the Rose F. Kennedy Center for Research in Human
Development at the Albert Einstein College of Medicine
of Yeshiva University in Bronx, New York. Last year
Dr. Eckholdt also built a research laboratory with
DayTrends LLC to study simulations of complex events.
His most recent project is an effort to design a rapid,
high throughput computer lab to digitize the
neuroanatomy of the worm c. elegans from electron
micrographs, for connectivity mapping as well as
virtual surgery. This will aid researchers in
assessing the impact of their discoveries in genetic
mutation on neurological development. You can view his
current research, favorite recipes, etc. at
http://www.thismind.com/haftan. Born and raised in
Baltimore, Maryland, he has been a resident of
Beautiful Park Slope in Brooklyn since 1990 and
considers SAS to be his first language, and English his
second. You be the judge!
-------------------------------------------------------
The agenda for the September 12th meeting is:
08:30 - 09:00 Coffee and Bagels
09:00 - 10:00 Proteomic COWS: Using SAS to
manage and run parallel clusters for
gene searching
Haftan Eckholdt
10:00 - 10:30 Break and Random Access
10:30 - 11:30 Hidden passageways in SAS/Graph
Software
Earl Westerlund
11:30 - 12:00 Hidden nuggets in Version 8: New
Informats, Formats, and Functions.
Mike Rhoads
12:00- 12:30 Some practical ways to use the new
SAS pattern-matching functions.
Mike Rhoads
12:30- 01:30 Lunch
01:30- 02:30 Simulating attrition and visual models
of Type I and Type II erors.
Haftan Eckholdt
02:30- 02:45 Break
02:45- 03:45 An Introduction to the DATA Step
Graphics Interface.
Earl Westerlund
03:45-04:45 Lost in Space: Recovering unmeasured
dimensions in spatial studies.
Haftan Eckholdt
-------------------------------------------------------
Directions
Bank of New York
10th Floor Auditorium
101 Barclay Street
New York, NY 10286
The Bank of New York is located one block north of the
World Trade Center between Greenwich and West Streets,
Barclay and Park Place. The main entrance is on
Greenwich/Park Place. For longtime members, this is
just north of our old location at the World Financial
Center.
By SUBWAY
Take the IND (A, C, E) to Chambers Street/World Trade
Center stop exit at Park Place, walk 2 blocks west to
Greenwich. West side IRT (1 or 9) to Chambers Street,
walk 1 block west to Greenwich and three blocks south
to Park Place, IRT (2 or 3) to Park Place, walk 2
blocks west to Park Place. East side IRT (4 or 5) to
Broadway Nassau, walk north to Park Place and west to
Greenwich. BMT (R or N) to Cortlandt Street/World
Trade Center. Walk west to Greenwich and north to Park
Place.
By BUS
M9, M10 or M22 to Battery Park City.
By CAR
Take West Street (West Side Highway) and turn onto
Vesey Street (1 block south of Barclay). Parking is
available at the World Trade Center.
>From NEW JERSEY
Take the Path train to World Trade Center and then walk
north.
-------------------------------------------------------
For further information about this meeting, or the New
York Area SAS Users Group, please contact:
Henny Wolland
Phone: (212) 573-1007
E-Mail: HWAssocites@Pop.MindSpring.Com
__________________________________________________________________
____
Sent via the Pace University Mail system at fsmail.pace.edu
