| Date: | Mon, 11 Nov 2002 08:33:24 -0500 |
| Reply-To: | Charles Patridge <Charles_S_Patridge@PRODIGY.NET> |
| Sender: | "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU> |
| From: | Charles Patridge <Charles_S_Patridge@PRODIGY.NET> |
| Subject: | HASUG MEETING ANNOUNCEMENT - Thursday, November 14,
2002 9:00am to 11:30am |
|---|
HASUG MEETING ANNOUNCEMENT - Thursday, November 14, 2002 9:00am to 11:30am
The fourth quarter HASUG meeting is to be held on the Second Thursday of
November, November 14, 2002, 9:00 a.m. to 11:30 a.m. at CIGNA in
Bloomfield, Connecticut in the South Building, Rooms 2B&C. Refreshments
will be available at 8:30 a.m. Please pre-register for this meeting by
completing the short pre-registration form on the www.HASUG.org web site.
If questions, contact our host, Barbara Moss at 860-226-1504, or by email
at Barbara.Moss@CIGNA.com.
The first presentation will be "Enterprise Guide Version 2.0" on Enterprise
Guide by James Blaha, Pace University. James is an Internet
Programmer/Analyst at Pace University as well as a member of the HASUG
Steering Committee. The abstract for his presentation follows:
"Enterprise Guide is a tool that if used properly can dramatically increase
user productivity and assist in increasing your technology ROI. Rapid
deployment of critical information in multiple formats (e.g. HTML, PDF,
RTF) helps empower those power users who know their data to generate
crucial ad hoc SAS reports on the fly without programmer intervention. The
presentation will be dealing with simple data and will be focusing on
getting the concepts of Enterprise Guide understood. The presentation will
include a demonstration of Enterprise Guide 2.0"
The second presentation will be "Danger: MERGE Ahead! Warning: BY
Variables with Multiple Lengths! by Bob Virgile, Robert Virgile Associates,
Inc. The abstract for Bob's presentation follows:
"Normally, when a data step merges two data sets, any common variables will
have the same length in both sources of data. When a variable has different
lengths in the incoming data sets, and when that variable is also a by
variable, the merge can produce truly bizarre results. For example, change
the order of the data sets in the merge statement and the data step
generates a different number of observations. Or merge two data sets that
have a unique sorted order and the resulting data set still contains
multiple observations for some by values. Or merge two data sets that are
in sorted order and the data step issues an error message claiming that the
data sets are not in sorted order. The results are so unusual, and
difficult to decipher, that current releases of the SAS software issue a
warning message:"
"WARNING: Multiple lengths were specified for the BY variable [variable
name] by input data sets. This may cause unexpected results." "This paper
examines the situation, some of its manifestations, and solutions to the
problem."
Those of you who plan on attending this meeting are encouraged to review
Bob Virgile's FOUR merge programs in preparation for his presentation. At
the bottom of the HASUG NEXT MEETING page (
http://www.hasug.org/next_mtg.htm ) after the directions and map, are four
merge programs that should be reviewed prior to attending Bob Virgile's
presentation.
Regards,
Charles Patridge
HASUG Steering Committee Member
Email: Charles_S_Patridge@prodigy.net
|