NA Digest Sunday, December 10, 1995 Volume 95 : Issue 49

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.

URL for the World Wide Web: http://www.netlib.org/na-net/na_home.html -------------------------------------------------------

From: Allison Bogardo <bogardo@siam.org>
Date: Wed, 06 Dec 95 10:33:36 EST
Subject: SIAM Election Results

SIAM Election Results

Following are the results of the recent SIAM election. The terms are
effective January 1, 1996.

President-Elect: John Guckenheimer (will assume the presidency
January 1, 1997; Margaret Wright continues as
president through the end of 1996)

VP at Large: Thomas A. Manteuffel

Secretary: Peter E. Castro

Board of Trustees: Margaret Cheney
James M. Hyman
Cleve Moler

Council: Kathryn Eleda Brenan
Eric Grosse
Nicholas John Higham
Lloyd N. Trefethen


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

From: Alfred Gautschy <gautschy@deimos.astro.unibas.ch>
Date: Thu, 7 Dec 1995 15:53:29 +0100 (MET)
Subject: Query on Block Tridiagonal Systems

Dear NA specialists,

in a recent attempt to recode stellar structure and evolution calculations
we came across a problem which we could not successfully overcome by now.
As I could not find any answer to it in texts on numerical methods
I decided to bother this forum:

Looking for solutions of the stellar structure equations requires to
solve a non-linear boundary value problem. (Funnily enough the
textbooks always do the same second order problem which causes *no*
problems). Classical approaches in astrophysics approximate the
spatial differential equations by finite difference equations and
solve them by Newton-Raphson iterations. When dealing also with
diffusion/mixing of chemical elements in the star one ends up with a
block tridiagonal matrix which has to be inverted for the
NR-iteration. Let A_i, B_i, and C_i be the subdiagonal, diagaonal,
and superdiagonal blocks for a certain gridpoint i. Unfortunately, we
cannot guarantee that Norm(A_i) + Norm(C_i) .LE. Norm(B_i) for some
sort of a matrix norm. Hence, this naive generalization of diagonal
dominance in a block system does not necessarily obtain anymore since
the entries in A_i...C_i depend on the magnitudes of the Jacobians of
the structure equations.

Our simplistic approach of using a `text-book' block-generalization
of a tridiagonal solver does not give stable numerical results.
(Possibly because we only perform partial pivoting on the block
level?). Now the question: Are there any block tridiagonal solvers
available in the NA community that could deal with a problem as
described above? If there are none, is it guaranteed that we would
get stable results when using some solvers for banded matrices such
as available in LAPACK?

I apologize if I should have bothered you with trivialities, but it might
at least reflect the level of knowledge in other sciences.

Alfred Gautschy
Astronomisches Institut der Uni Basel
Venusstrasse 7
CH-4102 Binningen/Switzerland
gautschy@astro.unibas.ch


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

From: Hermann Haaf <hhaaf@slb.cube.de>
Date: Wed, 06 Dec 1995 16:12:48 +0000
Subject: BLAS Routines in C

Does anyone know of a C-code version of the BLAS (Basic Linear Algebra
Subroutines) ideally for NEXTSTEP on Sun SPARCstations or anything which
comes close to it.

Many thanks in advance.

Hermann Haaf
Risk Management dept.
Suedwestdeutsche Landesbank
Stuttgart, Germany
fax ++ 49 711 127 3829


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

From: Tim Davis <davis@cis.ufl.edu>
Date: Fri, 8 Dec 1995 11:08:24 -0500
Subject: New Version of UMFPACK for Sparse Linear Systems

UMFPACK is a set of ANSI Fortran-77 routines for factorizing general sparse
matrices (unsymmetric, both in value and nonzero pattern), and solving sparse
linear systems. UMFPACK has very competitive performance on RISC workstations
and vector computers. The current version in NETLIB (2.0) has both single and
double precision codes. I've just released Version 2.1(Beta), with COMPLEX and
(non-ANSI) COMPLEX*16 versions. It's available at my home page
(http://www.cis.ufl.edu/~davis), or via anonymous ftp at ftp.cis.ufl.edu in
pub/umfpack/UMFPACK2.1Beta.tar.gz. Licensing restrictions apply (see the
README file for details). The COMPLEX and COMPLEX*16 codes have been fully
tested, but my test set of complex matrices is somewhat limited - thus the
"Beta" status. Once the complex codes see some usage, I'll place Version 2.1
in NETLIB. Please send me email if you use Version 2.1(Beta), whether or not
you find any bugs.

Tim Davis
University of Florida


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

From: Scott Hutchinson <sahutch@cs.sandia.gov>
Date: Thu, 7 Dec 1995 13:00:21 -0700
Subject: Aztec: A Parallel Iterative Package

SOFTWARE ANNOUNCEMENT

Aztec: A parallel iterative package for the solving linear
systems arising in Newton-Krylov Methods

Authors: Ray S. Tuminaro, John N. Shadid, Scott A. Hutchinson
Sandia National Laboratories
Albuquerque, New Mexico, USA

Aztec is an iterative library that greatly simplifies the
parallelization process when solving a sparse linear system of
equations Ax = b where A is a user supplied nxn sparse matrix, b is a
user supplied vector of length n and x is a vector of length n to be
computed. Aztec is intended as a software tool for users who want to
avoid cumbersome parallel programming details but who have large
sparse linear systems which require an efficiently utilized parallel
processing system. A collection of data transformation tools are
provided that allow for easy creation of distributed sparse
unstructured matrices for parallel solution. Once the distributed
matrix is created, computation can be performed on any of the parallel
machines running Aztec: nCUBE 2, IBM SP2 and Intel Paragon, MPI
platforms as well as standard serial and vector platforms.

Aztec includes a number of Krylov iterative methods such as conjugate
gradient (CG), generalized minimum residual (GMRES) and stabilized
biconjugate gradient (BiCGSTAB) to solve systems of equations. These
Krylov methods are used in conjunction with various preconditioners
such as polynomial or domain decomposition methods using LU or
incomplete LU factorizations within subdomains. Although the matrix A
can be general, the package has been designed for matrices arising
from the approximation of partial differential equations (PDEs).

Also, Aztec supports two different sparse matrix notations - either a
point-entry modified sparse row (MSR) format or a block-entry variable
block row (VBR) format. These two formats have been generalized for
parallel implementation and the library includes highly optimized
matrix-vector multiply kernels and preconditioners for both types of
data structures.

Aztec is publicly available through a research license. The code is
distributed along with technical documentation, example C and Fortran
driver routines and sample input files via the internet. It may be
obtained by contacting one of the authors below:

Ray S. Tuminaro
tuminaro@cs.sandia.gov
(505) 845-7298

John N. Shadid
jnshadi@cs.sandia.gov
(505) 845-7876

Scott A. Hutchinson
sahutch@cs.sandia.gov
(505) 845-7996

Please visit our Aztec Web page with postscript papers available at

http://www.cs.sandia.gov/HPCCIT/aztec.html

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

From: Jesse Barlow <barlow@cse.psu.edu>
Date: Sun, 3 Dec 1995 15:01:46 -0500
Subject: Temporary Address Change for Jesse Barlow

I am going on sabbatical from January through August 1996. My addresses
are below

January 1 - June 31

Department of Mathematics
University of Manchester
Manchester M13 9PL England

July 1 - August 17

Department of Mathematics
University of Linkoping
S-58183 Linkoping Sweden

Jesse Barlow


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

From: Y. F. Hu <Y.F.Hu@dl.ac.uk>
Date: Tue, 5 Dec 1995 11:54:41 GMT
Subject: Report on Dynamic Load Balancing Meeting

Dynamic Load Balancing on MPP Systems:
Progress, Challenge and Issues

This 1-day meeting on dynamic load balancing was held at Daresbury
Laboratory on November 27th, 1995. The meeting attracted around 70 academic
and industrial participants, from a wide range of disciplines and
interests, from destinations throughout the United Kingdom,
mainland Europe and the US. Although the main focus was towards
systems with homogeneous processor configurations, such as MPP
architectures, time was also devoted to issues on workstation
clusters and shared memory machines. Invited talks includes

"Can Static Load Balancing Algorithms Be Appropriate in a Dynamic
Setting?" Bruce Hendrickson (Sandia National Laboratory)

"Dynamic Mesh Partitioning" Chris Walshaw (University of Greenwich)

"Long range forces AND hydrodynamics" Frazer Pearce (University of Sussex)

"Dynamic Load Balancing in the UKMO's Operational Weather Forecasting
and Climate Prediction Model" Rupert Ford (CNC. Manchester)

"Dynamic Load Balancing Issues in the MPP Version of Long-Range Forcasting
System" Fritz Wollenweber (ECMWF)

"Perspectives on Dynamic Load Balancing on Parallel Tasks on
Workstation Clusters" Bill McMillan (University of Glaskow)

"Variants of the Spectral Bisection Technique Suited for Dynamic
Load Balancing" Dirk Roose and Raf Van Driessche (KU Leuven, Belgium)

"Some Algorithms for Dynamic Load Balancing" Yifan Hu (Daresbury
Laboratory)

"Towards Parallel Dynamic Load Balancing for Adaptive PDE solvers"
Peter Jimack (University of Leeds)

The compressed postscript files of all the talks are now available at
the URL:
http://www.dl.ac.uk/TCSC/Staff/Hu_Y_F/MEETING/meeting.html


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

From: Mike Powell <M.J.D.Powell@damtp.cam.ac.uk>
Date: Mon, 4 Dec 95 12:54:29 GMT
Subject: Cambridge Approximation Day

Cambridge Approximation Day: March 21st, 1996

Anyone who wishes to attend will be welcome at the Cambridge
Approximation Day on Thursday, March 21st, 1996. No fees will be
charged and there is no way of registering in advance, but please
send an e-mail to mjdp@amtp.cam.ac.uk if you will attend. Coffee
will be available at 10.00 am in the Common Room of DAMTP,
University of Cambridge (directions are available from mjdp).
Then the following talks will be presented in Seminar Room A.

M.D. Buhmann (ETH, Zurich)
"On a recovery problem with radial basis functions"

G. Goodsell and M.J.D. Powell (University of Cambridge)
"A new iterative method for thin plate spline
interpolation to scattered data"

B.J.C. Baxter (Imperial College, London)
Title to be announced

A.K. Kushpel and J. Levesley (University of Leicester)
"Approximation on spheres"

J.C. Mason (University of Huddersfield)
"Approximation problems in neural networks"

P.R. Graves-Morris (University of Bradford)
"A new approach to acceleration of convergence of
vector sequences"

G.M. Phillips (University of St Andrews)
"Generalized Bernstein polynomials"

L.N. Trefethen (Cornell University)
"The Chebyshev polynomial of a matrix"

M.J.D. Powell


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

From: Jeanne C. Butler <jeanne@tc.cornell.edu>
Date: Mon, 4 Dec 1995 13:41:25 -0500
Subject: Supercomputing Program for Undergraduate Research

SUPERCOMPUTING
PROGRAM FOR
UNDERGRADUATE
RESEARCH
1996 (SPUR)
June 2 - August 2, 1996

ABOUT SPUR

This program offers undergraduate students the opportunity to pursue a
computational science research project while developing skills in the use
of high performance computing technologies. Students apply to work on a
specific research project under the guidance of a faculty or staff member
at Cornell University. The proposed projects explore current research
problems in areas such as acoustics, chemistry, social dynamics, earthquake
modeling, pollution remediation, and fractals. Several of the projects
include a strong visualization component.

Successful applicants will attend a nine-week training and research program
at Cornell University and will pursue the research projects using the high
performance computing resources of the Cornell Theory Center. Students
will receive a $2,000 stipend, travel allowance, room (shared dorm room),
and partial board (dinner allowance at campus dining facilities).

ELIGIBILITY REQUIREMENTS

Applicants must be undergraduate students (graduating not before December
1996) who are U.S. citizens or permanent residents. Students must have
relevant coursework for their research areas, as well as coursework or
programming experience in FORTRAN or C. Students who participated in
previous SPURs are not eligible in 1996.

Women, minorities, and persons with disabilities are strongly encouraged to
apply. Students from four-year colleges with limited research facilities
are also encouraged to apply.

TO APPLY

Applicants must submit a completed application form, two letters of
recommendation, and college transcript(s). Applications must be received
by February 28, 1996. Students will be notified of their acceptance no
later than March 22. For more information or an application form (email or
hard copy), contact:

Jeanne Butler, Conference Assistant
Cornell Theory Center
427 Rhodes Hall
Ithaca, NY 14853-380
Email: spur@tc.cornell.edu
Telephone: 607/254-8813
Fax: 607/254-8888

Information and application also available via the World Wide Web at
http://www.tc.cornell.edu/Edu/SPUR

It is anticipated that this program will be sponsored by the National
Science Foundation; the program is dependent upon approval of funding
through the Researach Experiences for Undergraduates Program.


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

From: A. Bellen <bellen@univ.trieste.it>
Date: Tue, 5 Dec 1995 9:33:03 +0100 (MET)
Subject: Scientific Computing and Differential Equations

SciCADE97:
SCIENTIFIC COMPUTING AND DIFFERENTIAL EQUATIONS.
Grado (Italy) September 15-19, 1997.
FIRST ANNOUNCEMENT

The third SciCADE conference, organized by the Department of
Mathematical Sciences of the University of Trieste, will take place in
Grado (Italy) from 15th to 19th September 1997.

The conference will consist of plenary talks, mimisymposia and short
communications.

PLENARY SPEAKERS:

o Chris Budd (Bath University )
o Peter Crouch (Arizona State University)
o Ernst Hairer (Universite de Geneve)
o Marlis Hochbruck (Universitat Tuebingen)
o Karel in 't Hout (Rijksuniversitet Leiden)
o Roswitha Maerz (Universitat Berlin)
o Olavi Nevanlinna (Helsinki University of Technology)
o Stan Osher (University of Califonia at Los Angeles)
o Linda Petzold (University of Minnesota)
o Jesus Maria Sanz-Serna (Universidad de Valladolid)
o Daniel Stoffer (ETH Zurich) [tentatively]
o Andrew Stuart (Stanford University)

MINISYMPOSIA.

Minisymposia are scheduled on the following topics:

Boundary Value Problems
Differnential Algebraic Equations
Delay Differential Equations
Parallel ODE methods
Applications of ODEs
Waveform Ralaxation methods
Partial Differential Equations
ODE software
Generalizations of Runge Kutta methods
Continuous ODE methods
Nonlinear stability
Hamiltonians
Numerical methods on manifolds
Krylov space methods for ODEs.

NEW TALENT DISCOVERY PROGRAMME

The Organizing Committee decided to reserve two plenary talks for
young researchers, under 30 years of age. Our intention is to provide
an opportunity for young colleagues, whether research students or
postdoctoral workers, to present innovative and important research in a
plenary forum. The speakers will be chosen by the Organising Committee
on a basis of extended abstracts.

More details on minisymposia, new talent discovery programme, communications,
registration and accomodation will be forthcoming in a further announcement.

The Organizing Committee,

A. Bellen (University of Trieste, Italy)
A. Iserles (University of Cambridge, England)
S.P. Norsett (Norwegian Institute of Technology, Trondheim, Norway)
M. Zennaro (University of Trieste, Italy)


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

From: Leonid Faybusovich <lfaybuso@toda.math.nd.edu>
Date: Tue, 5 Dec 95 18:01:05 EST
Subject: Current and Future Directions in Applied Mathematics

The University of Notre Dame Symposium on
Current and Future Directions in Applied Mathematics
April 19 - 21, 1996

Program:
8 invited scholars will deliver a one hour lecture about current and
future trends in their research field. The lectures will be complemented
by a number of Mini-Sessions which will focus on a specific research
area in applied mathematics. A panel discussion is planned for
Saturday, April 20 about the role of applied mathematics in the
next decade. Because of the special scope of this symposium,
the organizers encourage the participation of graduate students.

The following principal speakers agreed to participate:

- Roger Brockett, Harvard University
- Christopher Byrnes, Washington University St. Louis
- Nicholas Ercolani, University of Arizona
- Richard Ewing, Texas A and M University
- Avner Friedman, IMA University of Minnesota
- David McLaughlin, Courant Institute of Mathematical Sciences
- Jerrold Marsden, Caltech
- Mike Todd, Cornell University

Contributed Talks:
The organizers are soliciting contributed talks and small cohesive
sessions focusing on new or emerging topics in applied mathematics.
Prospective contributors and organizers of sessions should submit
an extended abstract of 2 pages for each speaker to the organizing
committee by February 15, 1995. Submissions should be sent either
electronically to sym@kenna.math.nd.edu, or by regular mail to:

Applied Mathematics Symposium,
Department of Mathematics,
University of Notre Dame,
Notre Dame, Indiana, 46556-5683.

Organizing Committee and Sponsors:

The symposium is organized by the Applied Mathematics Group in
the Department of Mathematics at Notre Dame:

Mark.Alber@nd.edu Gerard.Misiolek@nd.edu
Leonid.Faybusovich@nd.edu Joachim.Rosenthal@nd.edu
Bei.Hu@nd.edu Hong-Ming.Yin@nd.edu

The symposium is sponsored by:
Hewlett Packard
and
the Department of Mathematics, the Center for Applied Mathematics
and the College of Science at Notre Dame.

Further Information:
All request about the symposium should be sent to sym@kenna.math.nd.edu.
Further Information can be obtained via the world-wide-web address:

http://www.science.nd.edu/math/symposium.html


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

From: Mayer Humi <mhumi@WPI.EDU>
Date: Thu, 7 Dec 1995 10:01:43 -0500 (EST)
Subject: Department Head Position at Worcester Polytechnic Institute

Worcester Polytechnic Institute (WPI) is an innovative college of science,
engineering and management with an enrollment of 2,500 undergraduates and
400 full-time graduate students, located in central Massachusetts.

The WPI Mathematical Sciences Department, currently with 22 full-time faculty,
provides undergraduate and graduate education through the Ph.D. Active areas of
faculty research include applied mathematic, optimal control, discrete
mathematics, statistics, operations research, scientific computation and
pedagogy. Its teaching/research facilities include networked workstation
laboratories and offices in addition to well supported campus-wide facilities.

The department seeks a dynamic individual who can promote growth in the
department's nationally recognized research program by attracting several
outstanding faculty to fill anticipated openings. The new department head
will be expected to take a leadership role in educational innovation.
He/she will have the opportunity to expand and develop new research and
educational programs, including an industrial project program for graduate
and undergraduate students.

Nominations for and applications from persons holding a Ph.D. should be
directed to the Director of Human Resources, Mathematical Sciences
Department Head Search Committee, Dept. A, Office of Human Resources,
Worcester Polytechnic Institute, 100 Institute Road, Worcester, MA 01609-2280.
Priority will be given to applications received before February 1, 1996.

The department www page is at:
http://www.wpi.edu/Academics/Depts/Math/

Prof. Mayer Humi tel:(508)831-5213
Fax:(508)831-5824
Dept. of Math, WPI, e-mail:mhumi@wpi.wpi.edu
100 Institute Rd.
Worcester,MA 01609-2247,USA


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

From: Florida State <dak@scri.fsu.edu>
Date: Tue, 5 Dec 1995 10:09:27 -0500
Subject: Positions at Florida State University

The Florida State University Computational Science and Engineering Program

Applications are being sought for two faculty positions in the College
of Arts and Sciences at the Florida State University, to begin in August,
1996. These positions are to be in conjunction with FSU's initiative in
Computational Science and Engineering, which is an interdisciplinary
program of graduate education and research founded upon the recognition
that fundamental progress in science and engineering depends more and more
on progress in the science of computation, itself. Information about the
Florida State University can be found at http://www.fsu.edu.

The first position is tenure-track at the assistant professor level
in the Department of Mathematics. Candidates must have a Ph.D. in Applied
Mathematics, Mathematics or related field. The successful candidate is
expected to do research in an area of computational mathematics, numerical
analysis and algorithms that complements the existing research program in
the department. Teaching at both the graduate and undergraduate levels will
be required. Experience beyond the doctorate is preferred, as is
experience with high performance computing and parallel algorithms.

The Department of Mathematics has a large applied mathematics
component, and has recently secured the new Thinking Machines Corporation
Eminent Scholar Chair in High Performance Computing. Information about the
department can be found at http://www.math.fsu.edu.

The second position is open with regard to both rank and
department. It is expected to be awarded in an area of computational
mathematics or applied computer science or computation-intensive
discipline. The successful applicant must have
a Ph.D. and a significant academic or research record in a field closely
associated with modern high-performance computing.

Applicants should send a resume and arrange for five letters of
recommendation to be sent to: David A. Kopriva, Chair, Search Committee,
Department of Mathematics, The Florida State University, Tallahassee,
FL 32306-3027 (kopriva@math.fsu.edu). The deadline for applications is
February 15, 1996.

The Florida State University is an equal opportunity/affirmative
action employer, and especially encourages applications from women and
minorities.


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

From: John Maddocks <jhm@Glue.umd.edu>
Date: Tue, 5 Dec 1995 16:57:45 -0500 (EST)
Subject: Positions at University of Maryland

University of Maryland
Institute for Physical Science and Technology
College Park, MD
Faculty Positions in Applied Mathematics

Applications are invited for tenure and/or tenure-track positions in
applied mathematics (subject to availability of resources and
administrative approval) commencing Fall 1996 at competitive salaries.
Candidates must already have outstanding research records in applied
mathematics. Candidates at all levels will be considered but research
will be evaluated primarily on the basis of significant achievements
and only secondarily on potential. A major selection criterion will
be the ability to communicate on a technical level with scientists and
engineers. Fields to be considered include numerical analysis,
scientific computation, computational analysis and applied dynamics.
The position may be joint with other departments and will have reduced
teaching responsibilities. While a candidate's research need not be
computational, the person must be able to teach graduate or
undergraduate courses that have a significant computational component.
To apply send a letter of application, curriculum vitae (signed and
dated) and a list of suggested names and addresses for letters of
recommendation. Candidates wishing to be considered for an Assistant
Professor position should arrange to have 3 or more letters of
recommendation sent directly. Send applications to Chair, Faculty
Search Committee, University of Maryland, Institute for Physical
Science and Technology, B-224, Rm. 4203, College Park, MD 20742-2431.
Applications should be received by January 20, 1996 for full
consideration. Minority and women candidates are encouraged to apply.
The University of Maryland is an Affirmative Action/Equal Opportunity
Employer.


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

From: Peter Monk <monk@math.udel.edu>
Date: Mon, 4 Dec 1995 09:26:39 -0500
Subject: Postdoctoral Position at University of Delaware

University of Delaware
Center for the Mathematics of Waves
Department of Mathematical Sciences

MURI Post-Doctoral Position in Computational Electromagnetics

Applications are invited for a post-doctoral position in
computational electromagnetics to begin September 1,1996. This position
is part of a Multidisciplinary University Research Initiative grant
in computational electromagnetics awarded to the Center by the Air Force
Office of Scientific Research. The term of the award is one year and the
stipend is $32,000.

Candidates should have a Ph.D. and demonstrated research potential
in applied mathematics with an emphasis in scientific computing. Preference
will be given to candidates who have the potential to contribute to ongoing
research projects in one or more of the following areas:

* direct scattering and radiation
* inverse scattering
* wave propagation (acoustic and electromagnetic)
* numerical methods for integral equations
* antenna optimization

Applicants should send a curriculum vitae, reprints and preprints,
and arrange to have 3 letters of reference sent to MURI Post-Doc Search
Committee, Center for the Mathematics of Waves,Rees Hall, University
of Delaware, Newark, DE 19716 by March 1,1996 for full consideration.

The University of Delaware is an equal opportunity employer which
encourages applications from qualified minority group members and
women.

Peter Monk
Department of Mathematical Sciences Phone: 302-831-1873
University of Delaware FAX : 302-831-4511
Newark, DE 19716
USA


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

From: Mike Field <Mike.Field@comlab.ox.ac.uk>
Date: Thu, 7 Dec 95 16:52:09 GMT
Subject: Postdoctoral Position at Oxford University

OXFORD UNIVERSITY COMPUTING LABORATORY

Postdoctoral Research Fellowship in Numerical Analysis

The Computing Laboratory is a full academic department of the
University and at the present time there are twenty-seven academic
staff, nearly fifty research officers and sixty-five doctoral students,
engaged in teaching and carrying out research in computing science and
numerical analysis.

Applications are invited for the above post, tenable for three years
from 1 January 1996 or as soon as possible thereafter, to work with
Dr E Suli and Professor K W Morton on a research project that concerns
the adaptive finite element approximation of multi-dimensional
hyperbolic systems. The project is funded by the Engineering and
Physical Sciences Research Council.

The aim of the project is to develop a comprehensive theory of reliable
and efficient a posteriori error analysis for adaptive finite element
and finite volume approximations of hyperbolic problems. At the
theoretical end, the key objective is the derivation of sharp a
posteriori bounds on the discretisation error in terms of the
computable finite element residual. The connection between dispersion
and error transmission in discretisations of hyperbolic problems will
be studied exploiting tools from harmonic analysis and micro-local
analysis of hyperbolic operators. The a posteriori error bounds will be
implemented into adaptive finite element and finite volume algorithms
for multi-dimensional hyperbolic systems.

Candidates should have strong competence in finite element theory and
modern theory of partial differential equations; research interest and
experience in theoretical and/or computational aspects of areas such as
a posteriori error estimation for finite element methods and the
development of adaptive finite element algorithms would be an
advantage.

The successful candidate will join an active research group whose
interests cover the whole range of numerical analysis: the main thrust
at present is in the numerical solution of partial differential
equations by finite element, finite volume, finite difference and
spectral methods, together with applications in computational fluid
dynamics and electromagnetics.

Salary will be at the lower end of the experience related RS1A scale,
currently GBP 14,317 to GBP 21,519 per annum (GBP = pounds sterling).
Further details are available by telephoning +44 1865 273898, or from
the World-Wide Web at URL
http://www.comlab.ox.ac.uk/oucl/jobs/na-rf.html

Applications, stating the post being applied for, should be in the form
of a full curriculum vitae together with the names and addresses of two
referees. They should be sent to arrive before the closing date of 22nd
December 1995, to The Administrator, Oxford University Computing
Laboratory, Wolfson Building, Parks Road, Oxford, OX1 3QD.
Alternatively, they may be sent via email to Mike.Field@comlab.ox.ac.uk

Oxford University is an Equal Opportunities Employer.


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

From: Wei Cai <wcai@newton.math.ucsb.edu>
Date: Tue, 5 Dec 1995 15:44:43 -0800
Subject: Postdoctoral Positions at UC Santa Barbara

Two Postdoctoral Positions are available starting from early 1996.
Candidates are expected to work on an ARPA funded research project
on the development of computer aided electrical design tools for
mixed digital/analog circuits packaging. Applicants should be highly
self-motivated and have demonstrated excellence in research. The
positions require strong experiences in some of the following areas:
numerical techniques for solving differential and integral equations;
electrical analysis of electronic packaging; development of computer
aided software tools. One position will be at the University of
California at Santa Barbara, and the other at the State University
of New York at Binghamton. Both positions are renewable up to three
years. Annual stipend is around $30k+. Applicants should send a detailed
curriculum vitae, selected publications, and a list of three references
to: Prof. Wei Cai, Department of Mathematics, University of California,
Santa Barbara, CA 93106. Position will remain open until filled.

UCSB is an affirmative action/equal opportunity employer.


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

From: Baltzer Science Publishers <publish@baltzer.nl>
Date: Mon, 4 Dec 1995 11:33:28 +0100
Subject: Contents, Advances in Computational Mathematics

CONTENTS:
Advances in Computational Mathematics
Volume 4, No. III, 1995
ISSN 1019 7168
Editors-in-Chief: John C. Mason & Charles A. Micchelli

pp. 207-260; T. Sauer and Y. Yu, On multivariate Hermite interpolation

pp. 261-282; S.M. Gomes, Convergence estimates for the wavelet-Galerkin
method: superconvergence at the node points

pp. 283-292; Z. Wu, Compactly supported positive definite radial functions


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

From: Baltzer Science Publishers <publish@baltzer.nl>
Date: Thu, 7 Dec 1995 11:44:24 +0100
Subject: Contents, Numerical Algorithms

CONTENTS:
Numerical Algorithms, Volume 10, No. 3-4, 1995
ISSN 1017 1398
Editor-in-Chief: Claude Brezinski

Volume 10, No. 3-4, 1995

pp. 203-224: G.L.G. Sleijpen and H.A. van der Vorst, Maintaining
convergence properties of BiCGstab methods in finite precision arithmetic

pp. 225-244: A.W. Bojanczyk, R.P. Brent and F.R. de Hoog, Stability
analysis of a general Toeplitz systems solver

pp. 245-260: X. Yan, Multistep methods for differential algebraic equation

pp. 261-288: M.C. Bartholomew-Biggs and S. Zakovic, Using Markov's interval
arithmetic to evaluate Bessel-Ricatti functions

pp. 289-336: A. Bultheel and M. Van Barel, Formal orthogonal polynomials
and Hankel/Toeplitz duality

pp. 337-362: S. Paszkowski, Quasipower and hypergeometric series --
construction and evaluation

pp. 363-378: S.H. Lui and G.H. Golub, Homotopy method for the numerical
solution of the eigenvalue problem of self-adjoint partial differential
operators

pp. 379-400: J. Choi, J.J. Dongarra and D.W. Walker, The design of a
parallel dense linear algebra software library: Reduction to Hessenberg,
tridiagonal, and bidiagonal form

pp. 401-420: J. Levesley, Convolution kernels based on thin-plate splines

pp. 421: P. Ciarlini and F. Pavese, Erratum to ``Application of special
reduction procedures to metrological data''

p. 423-424: Book reviews

p. 425-426: Author Index

Submissions of articles and proposals for special issues are to be
addressed to the Editor-in-Chief:

Claude Brezinski
Laboratoire d'Analyse Numerique et d'Optimisation
UFR IEEA - M3
Universite des Sciences et Technologies de Lille
59655 Villeneuve d'Ascq Cedex
France
E-mail: brezinsk@omega.univ-lille1.fr

postal address:
Paris Drouot BP 18
75433 Paris Cedex 09
France

Requests for FREE SPECIMEN copies and orders for Numerical Algorithms are
to be sent to: E-mail: publish@baltzer.nl or see our homepage at
http://www.NL.net/~baltzer/

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

From: Richard Brualdi <brualdi@math.wisc.edu>
Date: Mon, 4 Dec 1995 17:15:03 -0600 (CST)
Subject: Contents, Linear Algebra and its Applications

LINEAR ALGEBRA AND ITS APPLICATIONS
Contents Volume 233, January 15, 1996

Call for Challenges in Matrix Theory 1

Roberto Bevilacqua (Pisa, Italy) and
Paolo Zellini (Roma, Italy)
Structure of Algebras of Commutative Matrices 5

M. A. Dehghan and M. Radjabalipour (Kerman, Iran)
On Products of Unbounded Collections of Matrices 43

Karla Rost (Chemnitz, Germany) and
Zdenek Vavrin (Praha, Czech Republic)
Recursive Solution of Lo@a2wner-Vandermonde Systems of Equations. I 51

Reinhard Nabben (Bielefeld, Germany)
A Note on Comparison Theorems for Splittings and Multisplittings
of Hermitian Positive Definite Matrices 67

Yonghong Chen (Storrs, Connecticut), Stephen J. Kirkland
(Regina, Saskatchewan, Canada), and Michael Neumann
(Storrs, Connecticut)
Nonnegative Alternating Circulants Leading to
M-Matrix Group Inverses 81

Frieder Knuppel (Kiel, Germany)
Products of Similar Matrices 99

H. Valiaho (Helsinki, Finland)
Criteria for Sufficient Matrices 109

Jose A. Hermida-Alonso (Leon, Spain), M. Pilar Perez,
and Tomas Sanchez-Giralda (Valladolid, Spain)
Brunovsky's Canonical Form for Linear Dynamical Systems
over Commutative Rings 131

Zeev Nutov and Michal Penn (Haifa, Israel)
On Non-{0, 1/2, 1} Extreme Points of the Generalized
Transitive Tournament Polytope 149

Richard Varro (Montepellier, France)
Dupliquee d'Une Train Algebre et d'Une Algebre de
Bernstein Periodique 161

Suk-Geun Hwang (Taegu, Republic of Korea), Si-Ju Kim
(Kyungpook, Republic of Korea), and Seok-Zun Song
(Cheju, Republic of Korea)
On Convertible Complex Matrices 167

Wen Li and Mou-Cheng Zhang (Guangzhou,
People's Republic of China)
On Upper Triangular Block Weak Regular Splittings
of a Singular M-Matrix 175

Hans Gradl (Munchen, Germany)
A Result on Exponents of Finite-Dimensional Simple
Lie Algebras and Its Application to Kac-Moody Algebras 189

Chen Xu-Zhou and Robert E. Hartwig (Raleigh, North Carolina)
The Hyperpower Iteration Revisited 207

Shu-An Hu (Chattanooga, Tennessee) and Eugene Spiegel
(Storrs, Connecticut)
The Similarity Class of a Matrix 231

S. Gonzalez, J. C. Gutierrez, and C. Martinez
(Oviedo, Spain)
Second Order Bernstein Algebras of Dimension 4 243

Author Index 275


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

From: SIAM <poulson@siam.org>
Date: Thu, 07 Dec 95 10:39:36 EST
Subject: Contents, SIAM Discrete Mathematics

SIAM Journal on Discrete Mathematics
February 1996, Volume 9, Number 1
CONTENTS

A Characterization of Nonnegative Box-Greedy Matrices
Ulrich Faigle, Alan J. Hoffman, and Walter Kern

Efficient Gossiping by Packets in Networks With Random Faults
Krzysztof Diks and Andrzej Pelc

Short Random Walks on Graphs
Greg Barnes and Uriel Feige

The Biased Coin Problem
Ravi B. Boppana and Babu O. Narayanan

Spanners of Hypercube-Derived Networks
Marie-Claude Heydemann, Joseph G. Peters, and Dominique Sotteau

Lower Bounds on Representing Boolean Functions as Polynomials in Zm
Shi-Chun Tsai

On Convex Subsets in Tournaments
David J. Haglin and Marty J. Wolf

Horizontal Principal Structure of Layered Mixed Matrices: Decomposition of
Discrete Systems by Design-Variable Selections
Satoru Iwata and Kazuo Murota

Data Security Equals Graph Connectivity
Ming-Yang Kao

On Linear Recognition of Tree-Width At Most Four
Daniel P. Sanders

Chip-Firing Games on Mutating Graphs
Kimmo Eriksson

Linear Algorithms for Partitioning Embedded Graphs of Bounded Genus
L. Aleksandrov and H. Djidjev

Nowhere-Zero 4-Flows and Cayley Graphs on Solvable Groups
Brian Alspach, Yi-Ping Liu, and Cun-Quan Zhang

Tight Bounds for Dynamic Storage Allocation
Michael G. Luby, Joseph (Seffi) Naor, and Ariel Orda


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

End of NA Digest

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