NA Digest Sunday, October 30, 1994 Volume 94 : Issue 44

Today's Editor:
Cleve Moler
The MathWorks, Inc.
moler@mathworks.com

Submissions for NA Digest:

Mail to na.digest@na-net.ornl.gov.

Information about NA-NET:

Mail to na.help@na-net.ornl.gov.

-------------------------------------------------------

From: NA Digest <na.digest@na-net.ornl.gov>
Date: Sun Oct 30 14:33:33 EST 1994
Subject: NA Digest Calendar

NA Digest Calendar
Date Topic Place NA Digest #

Nov. 7- 9 Adaptive Grid Methods Hampton, VA 26
Nov. 12 Matrix Theory Salt Lake City, UT 43
Nov. 18 Parallell Computing Applications Utrecht, Nethelands 24
Nov. 14-18 SuperComputing '94 Washington, DC 6
Nov. 29 Upstate Numerical Analysis Day Ithaca, NY 44
Nov. 30... Computational Methods in Engineering Belo Horizonte, Brazil 11

Dec. 10-20 Knots&Nodes@Huia Huia, New Zealand 33
Dec. 15-16 Benchmarking of Parallel Systems Coventry, England 44
Dec. 16-19 Methods and Applications of Analysis Hong Kong 25

1995

Jan 3- 6 Computational Computer Design Maui, Hawaii 17
Jan. 9-10 Conference Honoring Ake Bjorck Linkoping, Sweden 37
Jan. 15-18 Bond Graph Modeling and Simulation Las Vegas, NV 4
Jan. 16-18 Markov Chains Raleigh, NC 7
Jan. 20-22 Numerical Treatment of Coupled Systems Kiel, Germany 44
Jan. 23-27 Optimal Design and Control Newport News, VA 31

Feb. 6- 9 Massively Parallel Computation McLean, VA 26
Feb. 8-10 Computational Issues in the Geosciences San Antonio, TX 27
Feb. 15-17 SIAM Conference on Parallel Processing San Francisco, CA
Feb. 19-25 Applications of Interval Computations El Paso, TX 16

Mar. 2- 8 Scattered Data Fitting Cancun, Mexico 39
Mar. 17-18 Krylov Subspace Methods Raleigh, NC 29
Mar. 20-21 Nonconvex Energy Functions Rutgers, NJ 20
Mar. 20-22 Steltjes Conference Toulouse, France 39
Mar. 22-24 Nonlinear Partial Differential Eqns. Knoxville, TN 41
Mar. 24-25 Southeastern-Atlantic Section of SIAM Charleston, SC 42
Mar. 28... Sci. Computation & Diff. Eqns. Stanford, CA 27
Mar. 29-31 Systems and Control Houthalen, Belgium 43

Apr. 2- 7 Multigrid Methods Copper Mountain, CO 30
Apr. 3- 5 Applied Mathematical Programming London, England 41
Apr. 3- 6 Numerical Methods for Fluid Dynamics Oxford, England 43
Apr. 20-22 Computational and Applied Mathematics Austin, TX 42
Apr. 24-28 3rd INRIA-SIAM Wave Propagation Conf. Juan-les-Pins, France 19
Apr. 26-30 Numerical Analysis & Applied Math. Constanza,Romania 26
Apr. 27-29 Control and Its Applications St. Louis, MO 36

May 9-12 Agriculture & Bio-Industries Brussels, Belgium 42
May 15-19 Graphics Interface '95 Quebec City, Canada 36
May 21-24 Object-Oriented Numerics Conference Jackson Lake, WY 32

June 5- 9 Spectral And High Order Methods Houston, TX 44
June 5- 9 Control and Information Shatin NT, Hong Kong 33
June 6-10 Inertial Manifolds Xi'an, China 21
June 13-21 Nonlinear Optimization Erice, Sicily, Italy 44
June 17-20 Iterative Methods in Linear Algebra Blagoevgrad, Bulgaria 39
June 20-24 Honoring G. I. Marchuk Novosibirsk, Russia 28
June 27-30 Numerical Analysis Dundee, Scotland 37

July 3- 7 ICIAM, Int'l Cong. Indust. Appl. Math. Hamburg, Germany 23
July 10-12 Linear Algebra and Its Applications Manchester, England 39
July 10-21 Nonlinear Waves Sapporo, Japan 43
July 16... Mathematics of Numerical Analysis Park City, UT 5
July 17-21 Distributed Parameter Systems Warsaw, Poland 25
July 17-21 Modelling and Optimization Warsaw, Poland 32
July 31... Rocky Mountain Numerical Analysis Salt Lake City, UT 32

Aug. 21-23 Workshop on Conservation Laws Trondheim, Norway 44
Aug. 27-31 Circuit Theory and Design Istanbul, Turkey 42
Aug. 29... Parallel Statistics and Economics Trier-Mainz, Germany 41

Sep. 17-21 Acoustics of Submerged Structures Boston, MA 22
Sep. 26-29 Validated Numerics Wuppertal, Germany 37

Nov. 19-22 Pure and Applied Mathematics Isa Town, Bahrain 38


------------------------------

From: Dirk Laurie <WSKDPL@puknet.puk.ac.za>
Date: Tue, 25 Oct 94 13:26
Subject: Orthogonal Polynomials

Many NA-Netters responded to my question, and since a desire for more
answers was expressed here recently, I've summarized the replies below.

First, I should apologize for stating the question badly, and worse,
making a mistake in the statement. (This did not prevent the right
people from answering the right question!) So I'll rephrase the question
as I should have asked it, and introduce it with some supporting
material.

Introduction:

Classical orthogonal polynomials are defined for the intervals [-1,1],
[0,\infty] or [-\infty,\infty]. Orthogonality only determines the
polynomial up to a nonzero constant factor, so they are standardized in
various ways: Legendre and first kind Chebyshev polynomials are
standardized to have the value 1 at x=1, second kind Chebyshev to have
value n+1 at x=1, etc.
In Table 22.2 of the Handbook of Mathematical Functions, some polynomial
families are standardized in more than one way. Interestingly enough,
only one of the preferred standardizations (Laguerre) leads to
orthonormal polynomials.
For computational purposes, the most convenient standardization (used
by Walter Gautschi in his ORTHPOL package, ACM TOMS 1994) is to use monic
polynomials.
When doing actual computations with orthogonal polynomials, several
numbers are of importance, of which the leading coefficient and the norm
are the most prominent.

The problem:

Take for example the Chebyshev polynomials. As usually standardized, they
have norm pi/2, except T_0 which has norm pi. Nice numbers, not big, not
small. But the leading coefficients go 1, 1, 2, 4, 8, 16, ... If you're
using single precision on an IEEE floating point processor, you run into
overflow at n=128. If we use leading coefficient 1 as Gautschi does,
then the norms go down by a factor 2 each time, and we run into underflow
at n=128. (Or even at n=64, if you need to use norm-squared.)

The question:

Somebody has discussed this problem and noted that the solution is to
scale the interval to [-2,2]. Who? Where?

Why is that solution correct?

There is a whole theory of 'capacity' of a set, important enough to be
one of the 450 subjects selected as main articles in the Encyclopedic
Dictionary of Mathematics, Second Edition. (But unknown to me until
kindly pointed out by some NA-Netters.) It's rather technical to define.
Henrici (Applied and Computational Complex Analysis, Volume 3, p374) has
the easiest definition in the case when the boundary of the set is a Jordan
curve G:
you map the exterior of G onto the exterior of the unit circle by a
function f such that f(\infty)=\infty, and with Laurent series of the
form
f(z) = a_1 z + a_0 + a_{-1}z^{-1} + ... with a_1>0.
Then the capacity of the set bounded by G is 1/a_1.
The best explanation of what effect capacity has, came from Nick Trefethen:
"But the phenomenon you mention is surely tied to the capacity, in the
sense of complex analysis, of the set you're dealing with. [-2,2], like
the unit disk, is a set of capacity 1. Presumably in general, if you do
Faber expansions on a set in the complex plane, you want to scale it to have
capacity 1 if exponentially growing or shrinking coefficients are to be
avoided. Taylor series on the disk and Chebyshev series on [-2,2] are
special cases."

The reference:

I received the following pointers:

L. Reichel, Newton interpolation at Leja points, BIT 30 (1990), 332-346.

L. Reichel, Fast QR decomposition of Vandermonde-like matrices and
polynomial least squares approximation, SIAM J. Matrix Anal. Appl.,
12 (1991), pp. 552-564.

L. Reichel and G. Opfer, Chebyshev-Vandermonde systems, Math. Comp., 57
(1991), pp. 703-721.

D. Calvetti and L. Reichel, Fast inversion of Vandermonde-like matrices
involving orthogonal polynomials, BIT, 33 (1993), pp. 473-484.

D. Funaro, Computational Aspects of Pseudospectral Laguerre Approximations,
Applied Numerical Math. 6 (1989/90) 447-457

I can't swear that any of these is the one I saw. But they certainly
use the trick mentioned, or something similar. Perhaps the first would
explain the sense of leja vu that I felt when encountering the phenomenon.

Other useful hints:

Several respondents made me realize how badly I phrased the question by
gently reminding me that one should evaluate orthogonal polynomials by
the recurrence formula, and series involving these polynomials by
Clenshaw's algorithm. All were tactful enough to refrain from pointing
out that I should have said 2^n instead of 4^n.

Thanks to all who responded.

*** NA-NET is great! ***

Dirk Laurie


------------------------------

From: Jeffrey Dunn <dunn@neptune.nrl.navy.mil>
Date: Thu, 27 Oct 1994 08:06:19 -0400
Subject: New Address for Jeffrey Dunn

Colleagues:

My new address is:

J. H. Dunn
Code 7214
Naval Research Lab.
Washington, DC 20375-5351

email: dunn@neptune.nrl.navy.mil
Voice: (202) 767-3870 FAX: (202) 404-7453


------------------------------

From: Alan Hindmarsh <alanh@daphne.llnl.gov>
Date: Wed, 26 Oct 94 14:18:53 PDT
Subject: LSODE Reference Manual Available

A comprehensive reference manual for the widely used initial value ODE
solver LSODE is now available:

Krishnan Radhakrishnan and Alan C. Hindmarsh, Discription and Use of
LSODE, the Livermore Solver for Ordinary Differential Equations,
NASA Reference Publication 1327, and LLNL Report UCRL-ID-113855,
March 1994.

This 108-page booklet includes documentation of the LSODE algorithms,
their implementation, the code organization, and its usage. Copies
are available on request from either
Alan C. Hindmarsh, alanh@llnl.gov
or
Krishnan Radhakrishnan, fsradhak@perch.lerc.nasa.gov

Alan Hindmarsh
Lawrence Livermore National Lab.
and
Krishnan Radhakrishnan
NASA - Lewis Research Center


------------------------------

From: Randall J. LeVeque <rjl@amath.washington.edu>
Date: Fri, 28 Oct 94 12:53:18 -0700
Subject: CLAWPACK, Conservation LAWs software PACKage

CLAWPACK -- Conservation LAWs software PACKage

Version 1.0 of CLAWPACK is now available from netlib. This is a package of
Fortran subroutines for solving hyperbolic systems of conservation laws in one
and two space dimensions. (Future versions should include 3D.) High resolution
flux-limiter methods are used, based on solving one-dimensional Riemann
problems. In two dimensions, multi-dimensional wave-propagation methods are
used. The current version supports only uniform Cartesian grids in
rectangular domains, and is intended primarily as a research and teaching
tool. The modular form should make it easy to modify and experiment with
other methods, as well as to apply it to new problems.

The code requires that the user supply a Riemann solver for the problem
being solved and also a subroutine that implements the boundary conditions.
Source terms can also be handled (via Strang splitting) in which case an ODE
solver for the source terms must also be supplied. Several examples of
different Riemann solvers are included with the package, including e.g.,
advection equation, Burgers' equation, Euler equations, isothermal
equations, shallow water equations. Various different boundary conditions
are also demonstrated, such as periodic, inflow, nonreflecting, and solid
walls. Numerous example drivers are included to demonstrate the use of
these subroutines. Matlab m-files are included to graphically display the
output, or they can be used as a model for writing graphics routines in
other languages. Documentation is included in the package, including the
postscript file for an introductory paper on the package.

The package can be obtained from netlib, where it resides in the library
pdes/claw. This directory currently contains:
index the index
clawpack the basic package, with 1d, 2d routines and many examples
doc documentation (postscript files of papers, slides)
advection applications of clawpack to 2d advection equations, with
examples from a recent paper on this subject (in doc).
nozzle application of clawpack to the quasi-1d nozzle problem.
vBurgers viscous Burgers' equation, with diffusion equation as source.
L-domain L-shaped domains, such as the forward facing step problem.
My hope is that a library of more sophisticated applications
will be gradually built up, and contributions from other users will be
gratefully accepted.

The package may be obtained by anonymous ftp from netlib.att.com, where it
resides in netlib/pdes/claw. The file netlib/pdes/claw.tar is a tar file of
the entire directory. Files can also be obtained by sending e-mail to
netlib@research.att.com. Send the message "help" to this address for more
information. Alternatively, you can obtain files by ftp through Mosaic using
the URL ftp://netlib.att.com/netlib/pdes/claw/index.html.
It is also possible to browse through the library using Mosaic from
ftp://amath.washington.edu/pub/leveque/programs/clawpack.html.

Randall J. LeVeque, University of Washington, Seattle
On sabbatical through July, 1995 at NCAR and the University of Colorado
Mail: Scientific Computing Division, NCAR, P.O. Box 3000, Boulder, CO 80307-3000
email: rjl@amath.washington.edu


------------------------------

From: Guess <guessab@iprvs1.univ-pau.fr>
Date: Thu, 27 Oct 1994 11:50:31 EDT
Subject: New Book, Topics in Polynomials

BOOK ANNOUNCEMENT:

TOPICS IN POLYNOMIALS: Extremal Problems, Inequalities, Zeros
G.V. Milovanovic, D.S. Mitrinovic, Th.M. Rassias
World Scientific, Singapore - New Jersey - London - Hong Kong, 1994
pp. xiv+822, US$ 115; ISBN 981-02-0499-X

The book contains some of the most important results on the analysis of
polynomials and their derivatives. Besides the fundamental results which are
treated with their proofs, the book also provides an account of the most
recent developments concerning extremal properties of polynomials and their
derivatives in various metrics with an extensive analysis of inequalities for
trigonometric sums and algebraic polynomials, as well as their zeros. The
final chapter provides some selected applications of polynomials in
approximation theory and computer aided geometric design (CAGD). One can also
find in this book several new research problems and conjectures with
sufficient information concerning the results obtained to date towards the
investigation of their solution.

Contents: Preface; General Concepts of Algebraic Polynomials; Selected
Polynomial Inequalities; Zeros of Polynomials; Inequalities Connected
with Trigonometric Sums; Extremal Problems for Polynomials; Extremal Problems
of Markov-Bernstein Type; Some Applications of Polynomials; Symbol Index;
Name Index; Subject Index.

For Orders and information contact:

USA World Scientific Publishing Co. Inc.
1060 Main Street, River Edge NJ 07661, USA
Toll-free: 1-800-227-7562 Fax: 1-201-487-9656 Tel: 1-201-487-9655


------------------------------

From: Sean McKee <caas29@vms.strath.ac.uk>
Date: Thu, 27 Oct 1994 11:51:51 EDT
Subject: IMA Journal of Mathematics Applied in Business and Industry

IMA Journal of Mathematics Applied in Business and Industry

Dear Colleague,

You will, I am sure, all be familiar with the two Oxford University Press
journals: IMA Journal of Numerical Analysis and Applied Mathematics.
You may, however, not know of the IMA Journal of Mathematics Applied
in Business and Industry or what its aims are.

The IMA Journal of Mathematics Applied in Business and Industry was founded
9 years ago and is now well established publishing high quality papers in
Operational Research, Decision Theory, Queueing Theory and other aspects
of Business applications.

I was asked to join Professor Thomas as joint Managing Editor to
enhance ``Mathematics Applied in Industry''. Thus I wish to stress that,
in addition to ``Business Applications'', this journal is interested and
has always been interested in publishing papers on Industrial Applications.
In particular, I am seeking papers on industrial case studies, mathematical
modelling of devices and processes, and using mathematical models as an
aid to the design of new products. Relevant papers could, for example,
be in the area of the mathematics underpinning computer aided design,
computer aided manufacture, or discrete event simulation. I would
particularly welcome papers which describe interdisciplinary projects in
which mathematics played a significant, or even crucial role; papers in
which mathematics was a contributory factor in a problem solving exercise;
or papers which contained the novel application of even standard mathematics
to a new area of engineering or science which hithertofore had never
received a mathematical treatment.

I would be very grateful if you could mention the industrial aspect of
this journal to your colleagues in mathematics, engineering mathematics,
physics and electrical engineering.

Yours sincerely,

Professor Sean McKee


------------------------------

From: Dianne O'Leary <oleary@cs.umd.edu>
Date: Mon, 24 Oct 1994 09:34:12 -0400
Subject: Householder XIII Symposium

The next Householder Symposium on Numerical Algebra (Householder XIII)
will be held June 17-21, 1996, at the Hotel Kronenhof in Pontresina,
in the Swiss Alps (1800m). The hotel is 4 hours from Zurich in the
southeast corner of Switzerland, and the main meeting room holds 125.

The local organizers are Walter Gander and Martin Gutknecht.
Further information for those who wish to attend will be available
in September 1995.

Dianne O'Leary
Chair, Program Committee


------------------------------

From: Andrew Conn <arconn@watson.ibm.com>
Date: Sun, 23 Oct 1994 23:01:37 -0400
Subject: SIAM Conference on Optimization

Planning Announcement of
SIAM Conference on Optimization
May 20-22, 1996
Victoria, British Columbia, Canada

This meeting is sponsored by the SIAM Activity Group on
Optimization, and will occur approximately midway between
the triennial International Symposia on Mathematical Programming.
SIAM is fortunate to have secured meeting and hotel space during
May, which is an ideal time to visit Victoria.

The organizing committee consists of:

Dr. Professor Georg Bock Dr. John Betts
IWR Boeing Computer Services
Universitaet Heidelberg P.O. Box 24346, MS 7L-21
Im Neuenheimer Feld 368 USA
6900 Heidelberg, Germany email: betts@espresso.boeing.com
email: bock@iwr.uni-heidelberg.de

Professor John Birge Professor Jonathan M. Borwein
Department of Industrial and Mathematics and Statistics Department
Operations Engineering Simon Fraser University
University of Michigan Burnaby, BC V5A 1S6
Ann Arbor, MI 48109 Canada
USA email: jborwein@cecm.sfu.ca
email: jrbirge@engin.umich.edu

Professor Albert Buckley Dr. Andrew R. Conn
Department of Mathematics Thomas J. Watson Research Center
Royal Roads Military College P.O. Box 218
FMO Victoria, B. C. V0S 1B0 Yorktown Heights, NY 10598
Canada USA
email: bbuckley@post.RoyalRoads.ca email: arconn@watson.ibm.com

Dr. Margaret H. Wright
AT & T Bell Laboratories, Room 2C-462
Murray Hill, NJ 07974
USA
email: mhw@research.att.com

Our intent is that the meeting will include core areas of optimization,
applications of optimization, and connections between optimization
and other areas of applied mathematics and scientific computing.

The organizing committee welcomes and encourages your suggestions for themes,
speakers, organisational aspects, format etc.


------------------------------

From: Anders Szepessy <szepessy@nada.kth.se>
Date: Mon, 24 Oct 94 10:24:47 +0100
Subject: Workshop on Conservation Laws

Workshop on Conservation Laws
Trondheim, Norway
August 21 - 23, 1995

Call for papers

This workshop will be an informal forum for presentation of new and recent
results concerning conservation laws. The main focus of the workshop will
be on mathematical analysis and numerical methods, but presentations concerning
applications are also welcome. There will be a series of longer lectures given
by invited speakers, as well as shorter contributed sessions. The invited
speakers will have time to give a broad introduction to a particular field.
The contributed sessions are expected to last 20 minutes including discussion.
For the contributed talks, an abstract should be submitted to the scientific
committee before March 1, 1995. There will be no proceedings from this
workshopbut copies of accepted abstracts will be distributed at the workshop.

Invited speakers:
A. Bressan, Trieste
B. Engquist, Los Angeles/Stockholm
T. Gimse, Oslo
J. Goodman, New York (to be confirmed)
D. Hoff, Bloomington
C. Johnson, Gothenburg
B. Plohr, Stony Brook
E. Tadmor, Tel Aviv
R. Winther, Oslo

Deadlines:

March 1, 1995: Submission of abstracts.
May 1, 1995: Registration.

Registration fee:

There is no registration fee for registrations received before April 1, 1995.
For registrations received after this date there is a fee of USD 100.

Abstracts:

For the contributed talks, an abstract not exceeding one typewritten page
should be submitted to the scientific committee before March 1, 1995.

Accommodation:

The workshop is held at the Britannia Hotel, Trondheim. The workshop will start
at 09:15 Monday August 21st and end at 16:00 Wednesday August 23rd.

Scientific Committee:

H. Holden, Norwegian Institute of Technology, University of Trondheim
N.H. Risebro, University of Oslo
A. Szepessy, Royal Institute of Technology, Stockholm
A. Tveito, University of Oslo

Registration forms and further information:

Further information can be obtained from world wide web
"http://www.imf.unit.no/conferences/cons-law"
or by sending email to holden@imf.unit.no


------------------------------

From: Andrew Ilin <ilin@sina.tcamc.uh.edu>
Date: Mon, 24 Oct 1994 12:38:24 -0500 (CDT)
Subject: New Dates of ICOSAHOM'95

DATE CHANGE ANNOUNCEMENT

In order to avoid conflicts with other international meetings ....

International Conference On Spectral And High Order Methods
ICOSAHOM'95

will be held

June 5--9, 1995
Houston, Texas

instead of the previously announced week in May, 1995.

The deadline for submission of abstracts has also been extended to
November 15, 1994. For more information, see NA Digest, V. 94, # 36
or contact

Leigh Ann Jacks, Conference Manager
ICOSAHOM'95
Department of Mathematics
University of Houston
Houston, TX 77204-3476.

E-mail submissions, requests for registration materials or requests
for further information to the following address are also welcome:

ICOSAHOM@math.uh.edu

IMPORTANT DEADLINES

Submission of abstracts: November 15, 1994.

Authors to receive notification regarding acceptance
and full instructions for submission of papers: December 15, 1994.

Full papers due: February 1, 1995.


------------------------------

From: Aad van der Steen <A.vanderSteen@cc.ruu.nl>
Date: Tue, 25 Oct 1994 15:11:09 +0100 (MET)
Subject: Benchmarking and Performance Evaluation Workshop

2nd Call for Participation
Workshop on Performance Evaluation
& Benchmarking of Parallel Systems

A joint workshop of various parties that are involved in
the performance evaluation of high-performance computing systems
will be held at the University of Warwick (Coventry, England) on
December 15--16, 1994. Participants will include EuroBen, PARKBENCH,
High Performance SPEC, and PEPS. Topics include, but are not limited
to:

- Structure of benchmarks.
- Machine characterisation.
- Monitoring systems.
- Recent results of the various benchmarks.
- New benchmark initiatives and opportunities for cooperation.
- Simulation.
- Interpretation of results.
- (Re)presentation of performance results.

The maximum number of participants is 35 to enable a high degree of
interaction and discussions to take place between participants.

The formal program will not be announced until shortly before the
workshop takes place to allow for maximum flexibility.

For further information, or registration please contact:

Prof. Graham R. Nudd or; Aad J. van der Steen
(PEPS Workshop) EuroBen
Dept. of Computer Science c/o Academic Computing
University of Warwick Centre Utrecht
Coventry CV4 7AL Budapestlaan 6
England 3584 CD Utrecht
Tel +44--203--523193 The Netherlands
Fax +44--203--525714 Tel +31--30--531444
Email conf@dcs.warwick.ac.uk Fax +31--30--531633


------------------------------

From: Gianni Di Pillo <erice@peano.dis.uniroma1.it>
Date: Wed, 26 Oct 1994 13:48:44 -0500
Subject: Workshop on Nonlinear Optimization

Ettore Majorana Centre for Scientific Culture
International School of Mathematics "G. Stampacchia"
Erice, Sicily, Italy
FIRST ANNOUNCEMENT

A workshop on

NONLINEAR OPTIMIZATION AND APPLICATIONS

will be held in Erice between 13 - 21 June 1995.

The workshop aims to review and discuss recent advances and promising
research trends in the field of Nonlinear Optimization, concerning
theory, algorithms and innovative applications. Both the finite and the
infinite dimensional cases will be of interest.

As usual, the course will be structured to include invited lectures
and contributed lectures. Proceedings including the invited lectures and a
selection of contributed lectures will be published.

The following is the list of invited lecturers:

Abadie J. Pang J. S.
Demyanov V. Polak E.
Evtushenko Y. G. Qi L.
Fukushima M. Robinson S. M.
Grippo L. Rockafellar R.T.
Judice J. J. Schnabel R.
Mangasarian O. L. Spedicato E.
More' J. J. Toint Ph.
Nocedal J.

Persons wishing to attend the workshop should write to:

Prof. Gianni Di Pillo
Dipartimento di Informatica e Sistemistica
Universita' di Roma "La Sapienza"
via Buonarroti 12
00185 Roma, Italy
E-mail: erice@peano.dis.uniroma1.it

Application by e-mail is strongly encouraged.

The total fee, which includes full board and lodging (arranged by the School),
should be approximatively US $ 800; the exact amount will be specified
in the second announcement.

Closing date for application: 30 April 1995


------------------------------

From: Jens Burmeister <jb@informatik.uni-kiel.d400.de>
Date: Wed, 26 Oct 1994 13:42:44 +0100
Subject: Numerical Treatment of Coupled Systems

Second Announcement

The GAMM Committee "Efficient numerical methods for pde" in cooperation
with the Christian-Albrechts-Universit"at Kiel organizes the

11th GAMM-Seminar Kiel
on
Numerical Treatment of Coupled Systems

Invited lectures will be given by Ronald H.W. Hoppe (Munich, Germany) and
Ulrich Langer (Linz, Austria).


Chairmanship: W. Hackbusch (Kiel), G. Wittum (Stuttgart)
Date: January 20 to 22, 1995
Location: Institut f"ur Informatik und Praktische Mathematik,
Universit"at Kiel (Germany)
Topics: Solving coupled systems of pde, coupling of different
physical processes e.g. multiphase flow or flow around
elastic structures, coupling of different discretiza-
tions (e.g. FEM/BEM, FEM-spectral methods), coupling of
separate local refinements.
Abstracts: Please send abstracts (10-20 lines) of your lecture by
Nov. 15, 1994. Notice of acceptance will be given by
Nov. 30. All participants, whether giving a talk or not,
have the possibility of sending an abstract of their
work on the topic of the conference. The collection
of abstracts will be available at the conference.

Conference fee: DM 50,- (to be paid after arrivel)

The first ten GAMM-Seminars were held at Kiel in 1984, 1986-
1994. The corresponding proceedings have been published
in the series "Notes on Numerical Fluid Mechanics"
(Vieweg, Braunschweig; vols. 10, 16, 21, 23, 30, 31, 41).

Please give notice to

by post : Prof. Dr. W. Hackbusch
Praktische Mathematik, CAU
Olshausenstr. 40-60
24098 Kiel, Germany
or by e-mail : jb@informatik.uni-kiel.d400.de
or by fax : ++49-431-880-4054


------------------------------

From: Nick Trefethen <lnt@cs.cornell.edu>
Date: Fri, 28 Oct 94 13:39:05 -0400
Subject: Upstate New York Numerical Analysis Day

2ND ANNUAL UPSTATE NUMERICAL ANALYSIS DAY
Cornell University
Ithaca, NY
Tuesday, 29 November 1994

On Tuesday, 29 November, the Cornell Computer Science Department and
the Cornell Theory Center will host an informal get-together for
numerical analysts, as we did last year at about this time. The
following is the current list of speakers (not yet complete):

Kevin Burrage University of Queensland
Anne Greenbaum New York University and Cornell University
Leslie Greengard New York University
Yuying Li Cornell University
Anita Mayo IBM TJ Watson Research Center
Lars Wahlbin Cornell University
Nick Trefethen Cornell University
Margaret Wright AT&T Bell Labs [not yet confirmed]

There will be a social event in the evening. If you're in the area,
or feel like visiting the area, please come! Please contact me if
you are likely to attend, so that we can get an idea of numbers,
and also if you need hotel or other information.

Nick Trefethen
LNT@cs.cornell.edu
(607) 255-4222


------------------------------

From: Stefano Foresti <stefano@osiris.usi.utah.edu>
Date: Fri, 28 Oct 1994 14:00:40 -0600
Subject: BibNet: Project Announcement

BibNet
A Bibliography Network Project

Announcement and solicited contributions

BibNet is a public-domain bibliography data base in BibTeX format.
The data base contains two types of bibliographies:

- by author (i.e. an author's list of publications), and
- by subject.

This initiative is a step toward sharing information electronically,
and it allows scientists to:

- provide complete and updated information on their own work,
- find information about ongoing research or specific topics,
- have an efficient pointer to electronically accessible material, and
- simplify the work of preparing publications.

The BibNet data base is maintained at the University of Utah, on
ftp.math.utah.edu, where the files are stored in /pub/bibnet. It is
mirrored nightly from there to a few other Internet archive sites.
BibNet is accessible through anonymous ftp, Netlib, Mosaic, Gopher and
e-mail. E-mail to the address:

bibnet-info@math.utah.edu

with the subject or body

send readme from bibnet

will return information and instructions on:

- features of the BibNet bibliography database,
- access and retrieval of bibliographies, and
- contributing new or updated bibliographies.

WE SOLICIT CONTRIBUTIONS !!

New or updated contributions will be standardized and checked and the
bibliographies will be incorporated in the BibNet data base.

We hope that the NaNet community will understand the convenience
and power of a bibliography database.

We anticipate that the success of this initiative will likely
generate motivation to render the work as automatable as possible,
to increase the features of this service, and to scale the project
to other fields.

Stefano Foresti
Utah Supercomputing Institute
University of Utah

Nelson H. F. Beebe
Department of Mathematics
University of Utah

Eric Grosse
AT&T Bell Labs


------------------------------

From: Michael Singer <singer@schubert.math.ncsu.edu>
Date: Mon, 24 Oct 94 07:35:38 EDT
Subject: Position at North Carolina State University

NORTH CAROLINA STATE UNIVERSITY
Department of Mathematics


The Department of Mathematics of North Carolina State
University invites applications for a tenure track
appointment in Symbolic Computation, beginning in the
Fall of 1995. Applicants at all levels will be
considered. Candidates should have a strong ongoing research
program and a demonstrated competence in teaching.
Applications should send a vita and letters of
reference to:

Symbolic Computation Search Committee,
Mathematics Department,
Box 8205, North Carolina State University,
Raleigh, N. C., 27695-8205.

Applications received by January 15,1995 will
be given full consideration. North Carolina State
University is an equal opportunity / affirmative
action employer. In its committment to diversity and equity, NCSU seeks applications from women, minorities, and the disabled.


------------------------------

From: Farid Alizadeh <alizadeh@rutcor.rutgers.edu>
Date: Mon, 24 Oct 1994 08:01:27 -0400
Subject: Positions at Rutgers University

Rutgers University - School of Business, New Brunswick

The Department of Management Science and Information Systems
has a tenure track assistant professor position opening beginning
September 1995. Candidates must have a strong quantitative
educational background, should be capable of doing outstanding
quantitative research in information systems and the management
sciences and must demonstrate a strong potential for scholarly
excellence.

Candidates are expected to be capable of teaching a least two out
of the three courses: Management Information Systems, Operations
Management, Statistics, or should have the necessary educational
background and experience for developing this capability. Affiliation
with RUTCOR - the Rutgers Center for Operations Research may be possible
but is not assured.

Curriculum vitae and three (3) confidential reference letters should
be sent to Benjamin Avi-Itzhak, Chairman, Department of MSIS, School
of Business, Rutgers University, P.O. Box 5062, New Brunswick, NJ 08903-
5062. Applications must be received no later than December 10, 1994.

Equal Opportunity/Affirmative Action Employer.


------------------------------

From: Peter Knabner <knabner@fauam3b.am.uni-erlangen.de>
Date: Wed, 26 Oct 1994 20:05:34 +0100
Subject: Positions at the University of Erlangen-Nuremberg

Starting January,1st,1995 there will be a major 3 year project to develop a
highly efficient software package for the
SIMULATION OF DENSITY INDUCED SUBSURFACE FLOWS
in three dimensions, based on the method of finite volumes and the multigrid
method, on parallel machines.
The participating research groups are from the universities of Stuttgart (Prof.
G.Wittum, coordinator), Heidelberg (Prof. W. Kinzelbach), Erlangen (Prof. P.
Knabner), Freiburg (Prof.D.Kroener), Prof. H. Zielke (Hannover).
The software development will be based on the package UG developed by the
Stuttgart group.

The tasks of the Erlangen group are:

A) GRID GENERATION

B) DISCRETIZATION AND ERROR ESTIMATORS

The following positions are available

-- one POSTDOCTORAL RESEARCH POSITION

The ideal candidate will have a doctoral degree in numerical analysis or
computational mathematics and be familiar with nonlinear partial differential
equations and related boundary value problems, and computer simulation
techniques. He will be responsible for the fulfillment of the project tasks of
subproject B). The salary will be according to the German BAT2A. It depends on
age and family status and is e.g. for a 27year old unmarried DM 65,832
(appr.$44,000) per annum.

-- two Ph.D. STUDENTS POSITIONS

The ideal candidate will have a MSc degree (or equivalent) in numerical
analysis or computational mathematics and be familiar with nonlinear partial
differential
equations and related boundary value problems, and computer simulation
techniques, but also someone familiar with the analysis of partial differential
equations or with a background in modelling of (subsurface) flow problems would
be eligible. The students have to perform work within the subprojects A) or B),
respectively. There are no tuition fees at German universities. The salary will
be according to the German BAT2A/2. It depends on age and family status and is
e.g. for a 27year old unmarried DM 32,916 (appr.$22,000) per annum.

Knowledge of the German language is not expected, and only necessary as far as
everyday life is concerned.

Candidates should contact contact me as soon as possible, and provide me with
CV, list of publications (for the postdoctoral position) and a desciption of
their scientific background.

Prof. Dr. Peter Knabner
Institute for Applied Mathematics
University of Erlangen-Nuremberg
Martensstrasse 3
D 91058 Erlangen
Germany

Tel. +49 9131 857015
+49 9131 857016
Fax +49 9131 857670
email: knabner@am.uni-erlangen.de


------------------------------

From: David Walker <walker@rios2.EPM.ORNL.GOV>
Date: Wed, 26 Oct 94 15:50:47 -0500
Subject: Position in CFD at Oak Ridge National Lab

CFD PROJECT LEADER

The Mathematical Sciences Section of the Engineering Physics & Math
Division at the Oak Ridge National Laboratory (ORNL) is seeking a senior
project leader in computational fluid dynamics (CFD) to contribute to
current and planned research projects in areas related to the U.S. Department
of Energy objectives. Current areas of interest include groundwater
modeling, environmental remediation, atmospheric modeling, ocean
modeling, enhanced gas and oil recovery, and automobile design.

The successful candidate should anticipate expanding the CFD program effort
to establish a world-class research team within a 2-3 year time frame.
Therefore, the project leader must be an internationally recognized CFD expert
with a demonstrated record in scholarship, funding, and successful research
project management.

ORNL is managed by Martin Marietta Energy Systems, Inc., for the
U.S. Department of Energy. For immediate consideration, send your resume to:
J. K. Johnson, PhD Employment, Oak Ridge National Laboratory, Dept. WWW,
P.O. Box 2008, Oak Ridge, TN 37831-6216.

ORNL is an equal opportunity employer committed to building and maintaining a
diverse workforce.


------------------------------

From: Colin Wright <WRIGHT@gauss.cam.wits.ac.za>
Date: Thu, 27 Oct 1994 15:44:56 GMT+0200
Subject: Positions at the University of the Witwatersrand

UNIVERSITY OF THE WITWATERSRAND, JOHANNESBURG -- SOUTH AFRICA

Assistant Professors -- Computational & Applied Mathematics

The Department of Computational and Applied Mathematics invites applications
from suitably qualified persons for appointment to tenure track positions at
the assistant professor level, to begin January 1995 or later.

The successful candidates will be expected to contribute towards the supervision
of Masters and Doctoral students in Computational and Applied Mathematics as
well as to the Department's undergraduate teaching programme. Preference, in
at least one of the positions, will be given to applicants working in the area
of Optimization.

Active research areas in the department are: optimization; theory and control
of dynamical systems; financial modelling; numerical analysis and computational
mathematics; chaos theory; classical mechanics; continuum mechanics;
differential equations; general relativity and cosmology; observational
astrophysics. Close ties also exist with the Departments of Computer Science,
Mathematics, and Statistics and Actuarial Science.

To apply submit a detailed CV with the names and addresses of three
professional referees + certified copies of degrees/diplomas to: Personnel
Office (Academic), University of the Witwatersrand, Private Bag 3, Wits 2050,
South Africa or fax +27-11-339-2223. The closing date is 15 November 1994, late
applications will however be considered until the positions are filled. E-mail
enquiries may be addressed to Professor CJ Wright,
wright@gauss.cam.wits.ac.za.

The University is an Affirmative Action and Equal Opportunity Employer.


------------------------------

From: Jens Burmeister <jb@informatik.uni-kiel.d400.de>
Date: Fri, 28 Oct 1994 15:02:42 +0100
Subject: Position at Christian-Albrechts-University

At the department of mathematics at the Christian-Albrechts-University in Kiel,
Germany, there is an open position as

C 4-Professur f"ur Angewandte Mathematik
(Wissenschaftliches Rechnen)

For your information, here is the official text written in German:

In der Mathematisch-Naturwissenschaftlichen Fakult"at der
Christian-Albrechts-Universit"at zu Kiel ist am Mathematischen Seminar
wiederzubesetzen:

In Kooperation mit der Technischen Fakult"at ab sofort eine

C 4-Professur f"ur Angewandte Mathematik
(Wissenschaftliches Rechnen)
Nachfolge Ahrens

Bewerberinnen und Bewerber m"ussen habilitiert sein oder gleichwertige
wissenschaftliche Leistungen aufweisen und sollten sich durch Forschungs-
arbeiten auf dem Gebiet der Entwicklung und Analyse computerorientierter
mathematischer Methoden zur Bearbeitung komplexer Probleme auf Hoch-
leistungsrechnern ausgewiesen haben. Sie sollten "uber Erfahrung bei der
Behandlung konkreter physikalisch-ingenieurwissenschaftlicher Probleme
verf"ugen.

Es wird eine Beteiligung an der Ausbildung der Studierenden der Mathematik,
Informatik, Physik und Ingenieurwissenschaften und die Bereitschaft zur
Mitwirkung beim Aufbau einer Studienrichtung Technomathematik erwartet.

Die Christian-Albrechts-Universit"at zu Kiel ist bestrebt, den Anteil
der Professorinnen und Dozentinnen zu erh"ohen und fordert deshalb
entsprechend qualifizierte Frauen nachdr"ucklich auf, sich zu bewerben.

Schwerbehinderte Bewerberinnen und Bewerber werden bei entsprechender
Eignung bevorzugt ber"ucksichtigt.

Bewerbungen mit den "ublichen Unterlagen (darunter eine kurz gefa"ste
Forschungsperspektive) sind bis zum 15. November 1994 zu richten an
den Dekan der Mathematisch-Naturwissenschaftlichen Fakult"at der
Christian-Albrechts-Universit"at zu Kiel, Olshausenstra"se 40, 24098 Kiel.


------------------------------

End of NA Digest

**************************
-------