NA Digest Sunday, February 2, 1992 Volume 92 : Issue 5

Today's Editor:

Cleve Moler
The MathWorks, Inc.

Submissions for NA Digest:

Mail to

Information about NA-NET:

Mail to


From: Bill Rosener <>
Date: Mon, 27 Jan 92 13:47:25 -0500
Subject: New NA-Net Welcome Message

To help encourage and inform new members about the NA-NET whitepages
and netlib, the brief message shown below is now returned when joining
the na-net.

You have been successfully joined to the NA-NET list.

In the future mail sent to:
will be forwarded to:

If you have not done so yet, we recommend that you send mail to to receive more information on the NA-NET.
Included in the message that you will receive is a description of
the NA-NET whitepages directory service. This facility is
available to everyone. The whitepages allow users to find out
more information about other members. This facility is available
to everyone. However, we would especially encourage members of
the NA-NET to join.

The NA-NET is running on the same machine that serves Netlib is a separate facility that
distributes mathematical software via electronic mail.
For more information on netlib send the one-line message
"send index" to


From: Tim <>
Date: Mon, 27 Jan 92 10:40:27 +0000
Subject: Re: Fortran Tools

Last week, Fred Fritsch asked:

> I am interested in knowing of the existence of portable software for
> manipulating Fortran programs. Specific abilities we will need
> include:
> (1) Converting nH... Hollerith strings to Fortran 77 '...' strings.
> (2) Producing a double precision version of a single precision routine.

The tools from Toolpack do these jobs.

They are available from NAG in two forms

1) As a public domain release tape which contains around 50 tools but is
very much a do-it-yourself job and, unless someone has already implemented
it on the machine you are interested, usually painful!
The cost is around \pounds 200 -- a handling charge + documentation (which
is also supplied in machine readable form).

2) As part of their NAGWARE package -- this does not include all the tools
but is a supported product. You'd need to check prices with Nag but I
reckon you can count on a factor of 10 or so over option 1) above.

We have all the tools running on Suns and they are well worth getting. The
pretty printer which does job (1) above is a terrific tools and the
precision transformer does a really good job. There is also a declaration
standardiser and a Fortran 77 verifier (PFORT) plus loads of other useful
little tools like Fortran name changers and Parameter statement
consistency checkers.

Several of the more popular tools are available as executables for Sun
Sparc architecutes from the netlib server at Kent but only via the UK's
incompatible ftp service. We are soon to be connected to the Internet and
I'll post more details when this finally happens.



From: Josep Fortiana Gregori <>
Date: Mon, 27 Jan 92 09:42:46 CET
Subject: High Accuracy Probability Distribution Functions

I need to compute to high accuracy (REAL*16) some of
the usual probability distribution functions: Normal,
Student's t, Chi Square, F, Beta, Gamma ...
Currently I'm using routines at REAL*8 precision, from
NETLIB, but as the the output undergoes long iterative
computations, precision is lost.
In books of tables and formulas I could find only
double precision.
Can someone help me to find the algorithms? (i.e.
how to obtain to REAL*16 accuracy the constants
in the rational approximations / series / continued
fractions which are used to evaluate the functions)


Josep Fortiana
Departament d'Estadistica
(Facultat de Biologia) Phone : 34 - 3 - 4021561
Universitat de Barcelona E-mail: ubaesq01@ebcesca1.bitnet
Av. Diagonal 645
08028 - Barcelona (also


From: Mustafa Pinar <>
Date: Wed, 29 Jan 92 15:45:12 EST
Subject: Article by Pietrzykowsky Sought

I am trying to find an article from a Polish Journal which
does not seem to be in the Libraries in the area here. Any help
would be appreciated.

The article is:

T.Pietrzykowsky, "On the steepest ascent method in linear programming",
Prace ZAM, Ser. A, No.11, Warsaw (1961).

I assume it is a Polish publication. Since I cannot read Polish, I would
need the English translation. I was not even able to get the Librarians
to track down this journal anywhere.

My e-mail address is :

Mustafa Pinar.


From: Allison Bogardo <>
Date: Mon, 27 Jan 92 11:51:40 EST
Subject: SIAM Student Paper Competition

The annual SIAM Student Paper Competition will be held during the 1992
SIAM Annual Meeting.

The student authors of the three best papers in applied and computational
mathematics submitted to SIAM will be invited to attend the 1992 annual
meeting in Los Angeles, July 19-24. Each winner must present his/her paper
at the conference and will receive up to $750 to offset expenses as well as
gratis registration to the meeting. The winners will be awarded a calligraphed
certificate at a special prize ceremony at the meeting. Papers must be singly
authored to be eligible for consideration. To qualify, authors must be students
in good standing who have not received their PhDs at the time of submission.

In submitting their work for publication, authors are asked to consider the
SIAM journals.

Submissions must be received by SIAM on or before March 15, 1992.

Submissions can be sent by regular mail or fax. Each submission must
include (1) an extended abstract NOT LONGER THAN 5 PAGES (including
bibliography), double-spaced, in English; (2) the signature of the author on
the submission; (3) a statement by the student's faculty advisor that the paper
has been prepared by the author indicated and that the author is a student in
good standing; and (4) a short biography of the student. It is not necessary
to send the full paper.

Each submission must also include a letter of recommendation from
the student's advisor or the department chair. Submissions will be judged
on the basis of originality, applicability, and clarity of exposition.

The winners will be notified by May 30, 1992.

If you have any questions, please contact:

Allison Bogardo
3600 University City Science Center
Philadelphia, PA 19104-2688
Telephone: (215) 382-9800
E-mail to
Fax to (215) 386-7999


From: Pat Worley <worley@haven.EPM.ORNL.GOV>
Date: Tue, 28 Jan 92 10:48:18 EST
Subject: PICL Upgrade Now Available on Netlib

A new upgrade of PICL (Portable Instrumented Communication Library)
is now available from This backward compatible upgrade
includes support for hostless programming and new implementations of PICL
for version 3.0 of the Ncube 2 software, for the Intel Delta, and for the
iPSC/2 simulator mpsim.

The upgrade has been reasonably well tested on the current generation
of Intel and Ncube machines, but less well tested on the Ncube/ten
and the Cogent. In consequence, the source for the original version
will be retained on netlib, with an "orig" prefix to the file names.
Also note that the documentation has not yet been upgraded.
The original documentation (in creference.shar and userguide.shar)
is still essentially correct, but the new features are not described.
Short descriptions of the new features are contained in the files
"upgrade.1.91" and "upgrade.1.92". PICL source can be obtained by sending
the following message to

send index from picl

The next upgrade to PICL will be available in 2-3 months, and will
include a redesigned trace file format. With the exception of additional
commands supporting user instrumentation, this next upgrade will not effect
most users, other than decreasing the overhead of tracing.
In particular, the next release will be coordinated with similar modifications
to ParaGraph, so that users will still be able to examine trace files with

Pat Worley


From: David Gates <gates@bigboote.Berkeley.EDU>
Date: Thu, 30 Jan 92 10:36:17 -0800
Subject: Question about Distributed-Memory Sparse Matrix Packages

I am a PhD student at UC Berkeley in the EECS department.
My thesis topic is concurrent mixed-level circuit and device
simulation. I am attempting to modify an existing simulator
written in C to exploit parallelism during the device simulation
phase. During the device simulations, large, slightly asymmetric
systems of linear equations must be solved. These systems can be
either real or complex. The sequential version uses
Ken Kundert's SPARSE1.3 (available from Netlib, developed here at
Berkeley) to perform this task. I am currently unaware of any
package which meets all my needs that is available for distributed
memory machines, specifically the Intel iPSC/860.
I am trying to find out if anyone out there is working on,
close to finishing, or even just starting work in this area.
I'm pretty confident I haven't missed anything obvious in the
published literature. If you can find the time, please send me
a brief summary of your research on this topic. I'd
like to know what the current state-of-the-art is before I take
my doctoral qualifying exam next week.

Specifically, I'm looking for a package that has all the following
1. Preferrably written in C, or with a C interface to Fortran.
2. Handles real and complex, asymmetric matrices.
3. Can store multiple matrices at once.
4. Can solve multiple matrices on different sections of a machine.
(i.e. user-defined space-sharing)
5. Can deallocate or reallocate matrix memory. (not essential)
SPARSE1.3 can do all these except (4) on a sequential machine. (3) and
(4) are the keys to exploiting parallelism in this application. I'm
not really expecting to find such a package yet, but want to make sure
before I go to the trouble of developing one myself. It would also
be nice if I could start from something close to this ideal if I do.

Thanks for any help you can give,
Dave Gates


From: Rob Corless <>
Date: Thu, 30 Jan 92 11:52:11 EST
Subject: Southern Ontario NA-Day '92

Tenth Annual Southern Ontario Numerical Analysis Day

Preliminary Announcement

Saturday, May 2, 1992
Room 240, Laurene O. Paterson Bldg (Western Science Centre)
University of Western Ontario
London, Ontario

Featured Speakers:

Keith Geddes, Symbolic Computation Group,
University of Waterloo

Cleve Moler, The Mathworks (as circumstances permit)

Henning Rasmussen, Computer Simulation Western,
University of Western Ontario

This is a preliminary announcement and call for contributions
to NA-Day '92.

In addition to the invited speakers, there will be seven to
nine contributed talks of 15 to 25 minutes each. Any
interested persons are invited to submit abstracts for consideration.
In keeping with tradition, we encourage graduate students to
give contributed talks. Submit abstracts to:

Robert Corless Charles Dunham
Dept. Applied Mathematics Dept. Computer Science
University of Western Ontario University of Western Ontario
London, Ontario, London, Ontario
N6A 5B7 N6A 5B7

Electronic submission is also possible, to

We will wait only until we have "enough" submissions, so submit early!
Talks in all areas of numerical analysis will be considered.

e-mail to for further details.


From: Willard Miller <>
Date: Tue, 28 Jan 92 15:53:27 CST
Subject: IMA Workshop on Iterative Methods

IMA Workshop on
February 24 -- February 29, 1992

Organizers: Gene Golub, Anne Greenbaum, and Mitchell Luskin

Large systems of matrix equations arise frequently in applications and
they have the property that they are sparse and/or structured.
Important applications await techniques for solving large nonsymmetric
systems of linear equations and eigenvalue problems. The purpose of this
workshop is to bring together researchers in numerical analysis and various
application areas to discuss where such problems arise and possible methods
of solution. Problems involving both convection and dissipation as well as
statistical applications lead to nonsymmetric and indefinite linear systems
and eigenvalue problems. We intend to explore recent developments in the
nonsymmetric Lanczos/Loewy algorithm, domain decomposition and substructuring,
conjugate gradient - like methods, preconditioners for nonsymmetric linear
systems, and other new ideas in this area. Methods that are particularly
efficient on modern computer architectures will be emphasized.

Sponsored jointly with the Minnesota Supercomputer Institute

The last two days of this workshop will be a celebration dedicated
to Gene Golub on the occasion of his sixtieth birthday (Feb. 29, 1992).

Most of the workshop talks will be held in Conference Hall 3-180 on the
entry floor of the Electrical Engineering/Computer Science Building.
This building is located on the corner of Washington Avenue and Union Street,
a block from the IMA Main Office. The conference hall is on the Ethernet and
has a projection system for display of computer output.

Monday, February 24

G. W. Stewart
Iterative aggregation for nearly completely decomposable Markov chains

Michael Eirmann
Is the optimal omega best for the SOR iteration method?

Youcef Saad
To be announced

Axel Ruhe
Arnoldi and rational Krylov algorithms for nonsymmetric eigenvalue problems
John Shadid
A comparison of preconditioned nonsymmetric Krylov methods on a large-scale
MIMD machine

Gerhard Opfer
Some structure simplifications oflinear systems arising in spline construction
Tuesday, February 25

Thomas A. Manteuffel
Large approximating the spectrum using the Ritz values and the roots
of the residual polynomials

Anne Greenbaum
Matrices that generate the same Krylov varieties

Henk A. van der Vorst
Minimum residual modifications to Bi-CG and to preconditioners

Olof B. Widlund
Schwarz algorithms for some nonsymmetric and indefinite problems

Martin H. Gutknecht
Stable algorithms for Pade approximation and their applications in
numerical linear algebra

Mary Wheeler
To be announced

Martin Hanke
Iterative methods for discrete ill-posed problems

Steve Vavasis
Multigrid for mixed boundary integral equations

Michael Berry
SVDPACK - A Fortran-77 library of routines for computing the sparse
singular value decomposition

Zdenek Strakos
Open problems in the convergence analysis of the symmetric Lanczos

Thursday, February 27

Howard Elman
Ordering effects on relaxation methods applied to the discrete
convection-diffusion equation

Roland W. Freund
Transpose-free quasi-minimal residual methods for non-Hermitian linear systems

Lloyd N. Trefethen
Lemniscates in the complex plane for Lanczos, Arnoldi, CR, and GMRES

Gerhard Starke
A hybrid Arnoldi-Faber iterative method for nonsymmetric systems of linear

Daniela Calvetti
An adaptive Chebyshev iterative method for nonsymmetric linear systems based
on modified moments

Lothar Reichel
Iterative methods for symmetric linear systems based on spectral density

Friday, February 28

Tony F. Chan
A composite step biconjugate gradient algorithm for nonsymmetric linear systems

Randolph E. Bank
An analysis of the composite step biconjugate gradient method

L. Yu. Kolotilina
To be announced

Wilhelm Niethammer
Iterative methods on massively parallel computers: Some remarks

Pal Rozga
On generalized band matrices and their inverses

Saturday, February 29

The program today is a celebration dedicated to Gene Golub on the occasion
of his sixtieth birthday (Feb. 29, 1992)

Organized by Dan Boley, Jack Dongarra and Paul Van Dooren

The meeting features four survey talks on topics of linear algebra
influenced by Gene's work. All other speakers will give short informal
talks addressing Gene's unique style of interacting with people, his
travel habits, anecdotes a.s.o. Some of the informal speakers may still change.

1. Pete Stewart, Roger Horn, Tony Chan, Mike Heath and Charles Van Loan.
2. Ake Bjorck, Frank Luk, Walter Gander, Nick Trefethen and Bob Plemmons.
3. David Young, Dick Varga, Ahmed Sameh, Cleve Moler and Jim Varah.
4. Dianne P. O'Leary, Henk van der Vorst, Howard Elman, Iain Duff
and John de Pillis.

A detailed schedule for the Saturday sessions is available from
Dan Boley,

Willard Miller

University of Minnesota
514 Vincent Hall
206 Church Street S.E.
Minneapolis, Minnesota 55455

FAX (612) 626-7370 telephone (612) 624-6066


From: Bernard Tourancheau <btouranc%frensl61.bitnet@CUNYVM.CUNY.EDU>
Date: Mon, 27 Jan 92 13:34:51 +0100
Subject: CONPAR92-VAPP V Conference on Parallel Processing


September 1-4, 1992

Laboratoire de l'Informatique du Parallelisme
Ecole Normale Superieure de Lyon, France

Sponsored by BCS-PPSG, CNRS, GI-PARS, Institut IMAG,
Programme de Recherches Coordonnees C3

in cooperation with IFIP WG10.3, IEEE, ACM, AFCET, SI-PARS, INRIA

The past decade has seen the emergence of two highly successful
series of CONPAR and of VAPP conferences on the subject of parallel
The Vector and Parallel Processors in Computational Science meetings
were held in Chester (VAPP I, 1981), Oxford (VAPP II, 1984)
and Liverpool (VAPP III, 1987).
The International Conferences on Parallel Processing took place
in Nurnberg (CONPAR 81), Aachen (CONPAR 86) and Manchester (CONPAR 88).
Thereafter the two series joined and the CONPAR 90 - VAPP IV
conference was organised in Zurich.

The next event in the series CONPAR 92 - VAPP V will be organised
in 1992 at the Ecole Normale Superieure de LYON (FRANCE)
from September 1, 1992 to September 4, 1992.

The format of the joint meeting will follow the pattern set by its
predecessors. It is intended to review hardware and architecture
developments together with languages and software tools for
supporting parallel processing and to highlight advances in models,
algorithms and applications software on vector
and parallel architectures.

It is expected that the program will cover:

languages / software tools
automatic parallelization and mapping
hardware / architecture
performance analysis
models / semantics
paradigms for concurrency
testing and debugging

The Proceedings will be published by Springer Verlag in the series
Lecture Notes in Computer Science.


Return Form of Intent Now please !
Submission of paper before Feb 29,1992
Notification of acceptance May 15,1992
Final Program June 15,1992

Secretariat CONPAR92/VAPPV

Ecole Normale Superieure de Lyon
Laboratoire de l'Informatique du Parallelisme
46, allee d'Italie - 69364 Lyon Cedex 07 - France
+33/72/72/80/37 - fax +33/72/72/80/80
e-mail conpar92@frensl61.bitnet


From: Alvaror <>
Date: Thu, 30 Jan 1992 16:31 GMT-0200
Subject: Brazilian Workshop on Inverse Problems


The meeting was held at the Institute for Pure and Applied
Mathematics (IMPA) in Rio de Janeiro from january 8 to 10,1992.
The main objective was bringing together people from numerical
analysis and optimization having particular Inverse Problems
as principal source of applications,as well as people working
on general theory for IP,both using iterative methods as a
principal tool.
Twelve invited talks (tutorials and new research results)
and several contributed papers were presented.A selection of
these papers will possibly be published in a forthcoming
Procedings volume.
It follows a list of the invited speakers and the titles of
their talks:

-A.Bjorck(Linkoping University);Lanczos method for solving ill-
posed problems.
-A.Dax(Israel Hydrological Service);Duality in regularized linear
programming and related problems.
-H.Engl(J.Kepler University);Stopping rules for iterative methods
for solving ill-posed problems with optimal convergence rates.
-R.Plato(Technical University of Berlin);Stopping rules for some
projection-iteration methods for solving linear,nonnecessarily well
posed equations.
-Y.Censor(Haifa University);Iterative methods for radiation thera-
py planning.
-M.Neumann(University of Connecticut);Convergence of sequential
and asyncronous nonlinear paracontractions.
-A.Neubauer(Linz University);An iterative solution of a nonlinear
inverse problem for industry:computer aided design of reflectors.
-A.De Pierro(University of Campinas);The EM algorithm and its
application in emission computed tomography.
-M.Teboulle(University of Maryland);Duality and iterative methods
for inverse problems.
-E.Schock(Kaiserslautern University);Implicit iterative methods
for ill-posed problems.
-I.Koltracht(University of Connecticut);Seismic deconvolution via
discrete inverse scattering;condition of the problem and stability
of the algorithms.
-A.Iusem(IMPA);Image reconstruction from noisy projections:a
regularized dual based iterative method.


From: Bobby Schnabel <>
Date: Thu, 30 Jan 1992 15:08:39 -0700
Subject: Boulder Distributed Language Workshop

Boulder Distributed Language Workshop -- First Announcement

The second Workshop on Languages, Compilers, and Run-Time Environments
for Distributed Memory Multiprocessors will be held in Boulder, Colorado
on Sept. 30 - Oct. 2, 1992. The first workshop was held at ICASE in
Hampton, VA in May 1990, and was judged to be very successful by the
participants. The second workshop will address similar topics to the
first, including : machine independent parallel programming; support for
regular and irregular data parallel computations as well as for coarse grain
parallel computations; compiler and run time communications optimizations;
compiler optimizations to expose parallelism; and experience with applying
parallel programming languages and techniques to applications.

The format of the workshop will include 30-minute research presentations,
with ample time for discussion, and group sessions that discuss the
challenges and future directions for research in this area.
Attendance and presentation of papers is open to all interested parties,
although it may be necessary to limit attendance to retain the workshop
atmosphere. Anyone interested in presenting a paper at the workshop
should submit a 3 page extended abstract by May 8, 1992 to
Carolyn Mich
Department of Computer Science
Campus Box 430
University of Colorado at Boulder
Boulder, CO 80309-0430
Electronic submissions also are welcome and should be sent to in LaTex, troff, or postscript form.
Submissions will be reviewed by the organizing committee consisting
of Ken Kennedy (Rice), Chuck Koelbel (Rice), Piyush Mehrotra (ICASE),
Joel Saltz (ICASE), Bobby Schnabel (Colorado) and Bob Weaver (Colorado).
Authors will be informed about the status of their papers by
June 20, 1992. Workshop proceedings will be published;
final papers will be at the workshop.

Anyone interested in attending the workshop, whether presenting a paper
or not, is asked to contact In case it is
necessary to limit attendance, preference will be given to those responding
by May 8, 1992. Registration materials for the workshop will be available in
early July 1992.

Carolyn Mich
Department of Computer Science
Campus Box 430
University of Colorado at Boulder
Boulder, CO 80309-0430


From: Steven J. Leon <>
Date: Sat, 1 Feb 1992 15:52 EST
Subject: Summer Workshops on Teaching Linear Algebra

ATLAST an NSF-ILAS Project to
Augment the Teaching Linear Algebra
through the use of Software Tools

The ATLAST Project will offer ten faculty workshops on the use of
software in teaching linear algebra. The workshops will last three
days. They will be held at regional sites across the United States
during the summers of 1992 and 1993. Each workshop will have the
same format and content. The project was conceived by the Education
Committee of the International Linear Algebra Society (ILAS).
Steven J. Leon of the ILAS Education Committee is serving as the
ATLAST Project Director and the Assistant Director is Richard Faulkenberry.
Both are in the Mathematics Department of the University of Massachusetts
Dartmouth. The ATLAST project is funded by a National Science Foundation
Faculty Enhancement grant.

Workshop participants will learn about existing commercial linear algebra
software packages and will be trained in the use of the MATLAB software
package. Attendees will learn how to effectively incorporate computer
exercises and laboratories into undergraduate linear algebra courses.

The ATLAST Project provides room and board for participants attending the
workshops. Participants will learn to design computing exercises or projects
at a level suitable for assigning to an undergraduate linear algebra
class. These exercises will be class-tested during the school year following
the workshop and then submitted to the project director for inclusion in a
database. Participants will each receive a $200 stipend for their

The databases from each of the summer workshops will be edited and printed
as manuals which will be distributed to the workshop attendees. The best
exercises will be selected for inclusion in an ATLAST Project Book. This
book will be published by one of the mathematics societies with the provision
that its contents will be public domain. Participants' contributions will be
acknowledged in both the database and the project book.

Summer 1992 ATLAST Workshops

West Valley College, Saratoga, California
June 4-6, 1992
Dr. Jane Day, San Jose State University

Auburn University, Auburn, Alabama
June 11-13
Dr. Kermit Sigmon, University of Florida

University of Wisconsin, Madison Wisconsin
June 18-20, 1992
Dr. Steven J. Leon, UMass Dartmouth

University of Wyoming, Laramie Wyoming
June 25-27, 1992
Dr. Eugene Herman, Grinnell College

University of Maryland, College Park, Maryland
July 22-25, 1992
Dr. David Hill, Temple University

All teachers of undergraduate linear algebra courses at colleges or
universities in the USA are invited to apply for the ATLAST workshops.
The deadline for applications is April 1, 1992. Late applications
will be accepted on a space available basis. Each workshop will be
limited to thirty participants. A screening committee will review
applications in mid-April and notify applicants of its decisions
by May 1, 1991. For further information, including an electronic
application form, contact

Steven J. Leon
ATLAST Project Director
Department of Mathematics
University of Massachusetts Dartmouth
Dartmouth, MA 02747
Telephone: (508) 999-8320
FAX: (508) 999-8901


From: John Hutchinson <>
Date: Tue, 28 Jan 92 16:09:44 EST
Subject: Postion at Australian National University

Faculty of Science
Department of Mathematics
Australian National University
Research Associate in Computational Mathematics and Geometry

Applications are invited for a research associate position (level one,
currently $36,285 p.a.), funded by the Australian Research Council,
in a grant to Dr Gerhard Huisken, Dr John Hutchinson and Dr Stephen Roberts.
The position will support a project investigating the computational and
theoretical aspects of geometric and other highly non-linear partial
differential equations. The successful applicant will be expected to
become a full collaborator in the project, and undertake responsibility for
the implementation and analysis of appropriate numerical algorithms, together
with appropriate theoretical investigations. Preference will be given to
applicants who have experience of computational methods for partial
differential equations or differential geometry. A PhD or equivalent
qualification is necessary. The position is available for up to three
years in the first instance. Consideration will also be given to
well-qualified applicants interested in short term appointments.

For further information contact:
Dr J. Hutchinson, tel: 61-(0)6-249-4042

Dr S. Roberts,
tel 61-(0)6-249-4445


From: Tom Coleman <>
Date: Tue, 28 Jan 92 09:05:55 -0500
Subject: Cornell Postdoc in Optimization

Postdoctoral Research Position in Optimization at Cornell University

The Cornell Computational Optimization Project (CCOP) has an opening
for a postdoctoral research position, beginning Fall 92. This
is a 1-year position with possible extension to two years.

CCOP spans two departments at Cornell (CS and OR) and is concerned
with all aspects of optimization, continuous and discrete. Currently we are
comprised of several faculty, research associates, and postdoctoral
researchers (and PhD students) covering many facets of computational
and theoretical optimization. Computing resources include a large network of
Sun workstations and access to several parallel computers in the Cornell Theory
Center (32-processor Kendall Square Computer, 32-processor IPSC/860, a ring of
RS6000's, and a 12-processor IBM 3090). Access to the Computer Science
parallel machines (an 8K CM-200 and a 62-node Butterfly) may
also be possible.

If you will have received a PhD in an optimization-related
area by Fall 92 and you are interested, please send your
application, with a vita listing three references, to:

Professor Mike Todd,
School of OR&IE,
229 ETC Bldg,
Cornell University,
Ithaca, NY 14853.

Inquiries can be sent via email to:


From: Tim Kelley <>
Date: Tue, 28 Jan 92 11:30:19 EST
Subject: Position at North Carolina State

Assistant Professor of Mathematics with postdoctoral and/or beginning
level faculty experience. This position also carries
responsibilities in the Center for Research in Scientific Computation
and in addition to established excellence in instruction,
there should be evidence of a strong research program. Consideration
will be given to the ability to pursue independent research
and participate in joint interdisciplinary research projects, and
to the potential of securing outside funding to help support
mathematical and computational aspects of interdisciplinary
research projects. Planned research directions of the Center include
distributed parameter systems (theoretical and computational
aspects of partial differential equations), deterministic and
stochastic methods for parameter estimation, optimization,
control of infinite dimensional systems, and research in applied
areas such as biomathematics, physical and/or chemical processes,
fluids, elasticity and material sciences. Applications, including
a resume and at least 3 letters of recommendation should be sent
to Professor H. T. Banks, Director: CRSC, Box 8205; North
Carolina State University; Raleigh, NC 27695-8205. Women and
minorities are especially encouraged to apply. NCSU is an EO/AAE.


From: SIAM <>
Date: Mon, 27 Jan 92 14:49:50 EST
Subject: Contents, SIAM Mathematical Analysis

Contents, SIAM J. Math. Anal.
May 1992, Vol. 23, No. 3

On Two Conjectures Concerning the Multiplicity of Solutions of a Dirichlet
Hans G. Kaper and Man Kam Kwong

Variational Formulations for the Determination of Resonant States in Scattering
M. Lenoir, M. Vullierme-Ledard, and C. Hazard

Global Solutions to the Compressible Navier-Stokes Equations for a Reacting
Gui-Qiang Chen

Solution of the Cauchy Problem for a Conservation Law with a Discontinuous Flux
Tore Gimse and Nils Henrik Risebro

Linearized Stability and Irreducibility for a Functional Differential Equation
Mary E. Parrott

Analytical and Numerical Results for the Age-Structured S-I-S Epidemic Model
with Mixed Inter-Intra-Cohort Transmission
M. Iannelli, F.A. Milner, and A. Pugliese

Periodic Solutions of Single-Species Models with Periodic Delay
H.I. Freedman and Jianhong Wu

Homoclinic Solutions for Autonomous Dynamical Systems in Arbitrary Dimension
Joseph Gruendler

Differential Operators and the Laguerre Type Polynomials
W.N. Everitt, A.M. Krall, L.L. Littlejohn, and V.P. Onyango-Otieno

On Orthogonal Polynomials of Sobolev Type: Algebraic Properties and Zeros
M. Alfaro, F. Marcellan, M.L. Rezola, and A. Ronveaux

Multidimensional q-Beta Integrals
Ronald J. Evans

On Refinement Equations Determined by Polya Frequency Sequences
T.N.T. Goodman and Charles A. Micchelli

Weighted Fourier Transform Inequalities for Radially Decreasing Functions
C. Carton-Lebrun and H.P. Heinig

Some Trace Theorems in Anisotropic Sobolev Spaces
Patrick Joly


End of NA Digest