NA Digest Sunday, July 12, 1992 Volume 92 : Issue 27

Today's Editor:

Cleve Moler
The MathWorks, Inc.

Submissions for NA Digest:

Mail to

Information about NA-NET:

Mail to


From: Murli Gupta <>
Date: Mon, 06 Jul 92 08:18:09 EDT
Subject: Temporary Change of Address for Murli Gupta

I am leaving for Australia on my sabbatical leave. During
July 20- November 20, my address will be:

Murli Gupta
School of Mathematics
Queensland University of Technology
Brisbane, Qld. 4001 Australia

My e-mail address:


From: Beth Ong <>
Date: Tue, 7 Jul 92 11:07:28 -0700
Subject: New Address for Beth Ong

I moved to San Diego July 1, 1992.
My new address and phone numbers are:

University of California, San Diego
Department of Mathematics
9500 Gilman Drive
La Jolla, CA 92093-0112
phone: 619-534-4899
619-534-3590 (secretary)

Please send email to

- Beth Ong


From: Peter Brown, George Byrne, Alan Hindmarsh <GDBYRNE%ERENJ.BITNET@pucc.Princeton.EDU>
Date: Tue, 07 Jul 92 11:27:52 EDT
Subject: New ODE Solvers Available

The ordinary differential equation (ODE) solvers VODE and VODPK can be
obtained from netlib in both single and double precision Fortran. The
solvers feature backward differentiation (BDF) methods for stiff ODEs
and implicit Adams methods for non-stiff ODEs. They combine features
of the earlier codes EPISODE and LSODE and have a calling sequence
very similar to LSODE and LSODPK, respectively. Both use a
fixed-leading coefficient version of the variable step BDF and Adams
methods. VODE uses LU factorization, allows banded and dense
Jacobians, and has an option that lets the user keep a copy of the
full Jacobian matrix, as well as the LU factored Newton matrix. For
many problems this can increase performance considerably. VODPK uses
a preconditioned Krylov method for the numerical linear algebra. The
packages documentation in the preamble, and examples of use.

These codes may be obtained free of charge as follows:

send mail to

with the message:

send vode.f from ode to obtain double precision VODE
send svode.f from ode to obtain single precision VODE
send vodpk.f from ode to obtain double precision VODPK
send svodpk.f from ode to obtain single precision VODPK

These codes make use of LINPACK and the Basic Linear Algebra
Subroutines (BLAS), and the machine constant routines R1MACH/D1MACH,
which must be obtained separately from netlib or from some other
source .


1. P. N. Brown, G. D. Byrne, and A. C. Hindmarsh, VODE: A Variable-
Coefficient ODE Solver, SIAM J. Sci. Stat. Comput., 10 (1989),

2. G. D. Byrne, Pragmatic Experiments with Krylov Methods in the
Stiff ODE Setting, in Numerical ODEs (Approximate Title), J. Cash and
I. Gladwell, eds., Oxford University Press, to appear.

3. G. D. Byrne, The Solution of a Co-Polymerization Model with VODE,
in Recent Developments in Numerical Methods and Software for
ODEs/DAEs/PDEs, G. D. Byrne and W. E. Schiesser, eds., World
Scientific, River Edge, NJ, 1992, pp. 137-197.


From: Wim Sweldens, Bjorn Jawerth <>
Date: Sat, 11 Jul 92 16:34:53 EDT
Subject: Wavelets Mailing List

We are happy to announce you a

W A V E L E T D I G E S T.

Nowadays, wavelets are one of the most rapidly developing research areas
both in pure/applied mathematics and signal/image processing.
More and more people are becoming interested in using wavelets
for a broad spectrum of applications.
For researchers it is essential to stay informed on the latest
developments. Hence we are planning, here at the University of
South Carolina, to start an edited "wavelet digest."

Topics we would like to cover in our digest are:

- Announcements of:
- future wavelet conferences, call for papers, etc.
- talks on wavelets
- wavelet courses
- new preprints (and how to get them, e.g. by ftp)
- wavelet articles appearing in journals
- new books on wavelets
- wavelet-related job opportunities
- Questions (and answers) concerning:
- mathematical background of wavelets
- technical aspects of wavelets
- references
- addresses
- Regularly updated reference lists
- Reviews of recently published books on wavelets
- Posting of open problems
- Available wavelet software, where and how to get it

If you would like to subscribe to this mailing list,
send a message with "subscribe" as the subject to

If you want to submit something to the digest,
send a message with "submit" as the subject to

Please pass this message on to others who you think might be interested.

We are hoping to present the first wavelet digest to you soon.

Wim Sweldens
Bjorn Jawerth

Department of Mathematics,
University of South Carolina.


From: Ardhendu Pathak <NXJ101@URIACC.URI.EDU>
Date: Wed, 8 Jul 1992 16:10:47 EDT
Subject: Matrix Inverse

I am looking for a reliable algorithm (/MATLAB function file) to invert
a very ill-conditioned 4x4 matrix. Two rows of the matrix have exponentially
large elements and rest have exponentially small elements. This matrix occurs
in certain fluid-structure interaction problems, where elements have physical
significance in terms of exponentially varying elastic and acoustic fields.

Thanking you in advance,

Ardhendu Pathak.


From: Bernard Tourancheau <>
Date: Tue, 7 Jul 92 16:25:49 +0200
Subject: Conference on Vector and Parallel Processing

Joint Conference on Vector and Parallel Processing

September 1 - 4, 1992
Ecole Normale Superieure de Lyon
Laboratoire de l'Informatique du Parallelisme
Lyon, France

Sponsored by LIP, ENS Lyon, Institut IMAG, Programme de Recherches
Coordonnees C3, CNRS, Digital Equipment Corporation, Archipel

In Cooperation with BCS-PPSG, GI-PARS, IFIP WG 10.3, IEEE, ACM,

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.

More than 230 papers written by authors from 35 countries have been
submitted for consideration to CONPAR 92-VAPP V. Among these papers,
the Program Committee has selected 36 full papers, 58 short papers
and several posters. Five invited presentations, four tutorials and
various exhibitions will complete the Conference. I am pleased to thank
the Program Committee Members and the 600 referees (each paper has been
evaluated by four referees), for their excellent job in building the
scientific content of the Conference.

For registration and complete program please write to the Conference

Secretariat CONPAR92 / VAPP V
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
Bitnet : conpar92@FRENSL61
Internet :

During the first day of the conference, the following tutorials will be
presented by international experts :

* Prof. Jean-Luc Gaudiot (University of Southern California - USA)
Data-Driven and Multithreaded Architectures for High-Performance Computing

* Prof. Ron H. Perrott (Belfast - U.K.)
Parallel Programming Languages

* Prof. Paul Feautrier (IBP-Masi-Versailles - France)
& Dr.Patrice Quinton (Irisa - Rennes - France)
Automatic Parallelization Techniques

* Dr. Heinz Mulhenbein (Augustin 1 - FRG)
Genetic Algorithms, Artificial Neural Networks and Parallel Processing


Invited Presentations

Dr. Gerard Berry (INRIA Sophia Antipolis - France)
The ESTEREL language
Prof. Jack Dongarra (Knoxville - USA)
Heterogeneous parallel computing
Prof. Jose Fortes (Purdue University - USA)
On the adaptative of algorithms and architectures
Prof. Dr. Friedemann Mattern (Saarbrucken - FRG)
Observing distributed computations
Prof. Charles Leiserson ( MIT-USA)
The network architecture of the connection machine CM-5

Exhibition, Demonstrations and Posters
In addition to the tutorials and regular sessions of the conference, there
will be many exhibitions.
- Exhibition by vendors of hardware and software
- Application of parallel systems in industrial products
- Poster sessions

Organising committee
Professor Luc BOUGE LIP-IMAG - ENS Lyon
Professor Michel COSNARD LIP-IMAG - ENS Lyon (Chairman)
Professor Yves ROBERT LIP-IMAG - ENS Lyon
Professor Denis TRYSTRAM LMC-IMAG - Grenoble


From: P.J. Laurent <>
Date: Fri, 10 Jul 92 08:28:03 +0200
Subject: Conference in France on Curves and Surfaces

Second international conference organized by A.F.A.


JUNE 10-16, 1993

The "Association Francaise d'Approximation" (A.F.A.) groups
together academic and industrial people concerned with approximation and
representation of curves and surfaces. It organizes an International
Conference at Chamonix-Mont-Blanc (France) from June 10 to 16, 1993 in
collaboration with "Universite Joseph Fourier (Grenoble)", "Universite de
Nantes", "Universite de Pau et des Pays de l'Adour", "Universite Paul
Sabatier (Toulouse)", "Universite de Valenciennes", "Ecole Nationale
Superieure des Telecommunications (Paris)", "Instituts Nationaux des
Sciences Appliquees de Rennes et de Toulouse", "Universite de Rennes".

P.J. Laurent (Grenoble, France),
A. Le Mehaute (Nantes, France),
L.L. Schumaker (Nashville, USA).

Representation of curves and surfaces (splines, finite elements,
subdivision,...), computer aided geometric design, interpolation and
smoothing, variational theory of splines, wavelets and splines, as well as
relevant applications like image analysis, CAD-CAM,...

C. Brezinski (France), M. D. Buhmann (U.K.), C.K. Chui (U.S.A.),
W. Dahmen (Germany), T. DeRose (U.S.A.), R. DeVore (U.S.A.), N. Dyn
(Israel), W. Light (U.K.), C.A. Micchelli (U.S.A.), H.P. Seidel (Canada).

A. Cohen (France), J.C. Fiorot (France), J. Hoschek (Germany), F.
Utreras (Chile).

D. Apprato, M. Atteia, J.M. Chassery, P. Chenin, J.P. Dedieu, J.C.
Fiorot, J. Gaches, B. Lacolle, M.L. Mazure, Y. Lafranche, C. Rabut, P.
Sablonniere, C. Vercken.

For more details, please contact the organizers at your earliest
convenience but not after NOV. 15, 1992.

For submitting a research talk, please send your title and
abstract before JAN. 31, 1993. The abstract, typed double-spaced, of one
page maximum, will be directly reproduced in the booklet handed out to all

Curves and Surfaces,
LMC-IMAG, Universite Joseph Fourier,
BP 53, 38041- GRENOBLE (France).

E-mail: Fax: (33) 76 63 12 63

Pierre-Jean LAURENT
Universite Joseph Fourier
BP 53X
38041 GRENOBLE (France)

Tel : (33) 76 51 46 11
Fax : (33) 76 51 45 55


From: Jim Bunch <>
Date: Fri, 10 Jul 92 15:59:27 -0700
Subject: Parlett/Kahan Meeting at Berkeley



A conference will be held at MSRI in Berkeley on October 17, 1992,
in honor of the 60th birthdays of Beresford Parlett (July 4, 1992) and
William Kahan (June 5, 1993).
The Organizing Committee consists of:
James Bunch, UCSD,
James Demmel, UC Berkeley,
Horst Simon, NASA Ames,

The speakers at the conference will be:

James Bunch, University of California, San Diego, "Three Decades of
Numerical Linear Algebra at Berkeley".

G. W. Stewart, University of Maryland, "On the Perturbation of
Matrix Factorizations".

John Reid, Rutherford Appleton Laboratory, England, "Taking Advantage
of Sparsity within 2x2 Pivots When Solving Symmetric Indefinite Sets of
Linear Equations".

Horst Simon, NASA Ames, "Spectral Algorithms--A New Approach to Some
Discrete Optimization Problems in Scientific Computing".

Larry Nazareth, Washington State University, "The Newton and Cauchy
Perspectives on Computational Nonlinear Optimization".

James Ortega, University of Virginia, "Solution of Nonlinear
Poisson-type Equations".

Anne Greenbaum, Courant Institute, NYU, "Matrices that Generate
the Same Krylov Residual Spaces".

Bahram Nour-Omid, Scopus, Berkeley, "Ordered Modified Gram-Schmidt".

David Scott, Intel, "A High Performance Out-of Core Dense Equation
Solver for the Intel Parallel Supercomputer".

Peter Tang, Argonne National Laboratory, "Recent Advances in Rank-
Revealing QR Factorization".

Scott Baden, University of California, San Diego, "The Role of
Heuristics in Parallel Computation of Scientific Problems".

James Demmel, University of California, Berkeley, "Recent Progress
in Parallel Eigenroutines".

There will be a banquet on the evening of October 17. The banquet
speaker will be Richard Lau, Office of Naval Research.

There will also be a special issue of the Journal of Numerical Linear
Algebra with Applications (JNLAA) dedicated to Parlett and Kahan. The
deadline for manuscripts will be October 17, 1992. James Bunch will be the
editor of the special issue; anyone interested in submitting a manuscript
should get a copy of "Guidelines for Contributors" from him.

Hotel Information

We have reserved 20 rooms at the Durant Hotel (only singles available)
and 50 rooms at the Marriott Hotel. You must make your own reservation
directly with the hotel. You should make your reservation soon if you
have not done so already.

Durant Hotel, 2600 Durant Ave, Berkeley, south side of campus,
(510) 845-8981, $75/single, mention the conference.

Marriott Hotel, 200 Marina Blvd, Berkeley, 1.5 miles west of campus
on SF Bay, (510) 548-7920. $85 flat rate for single, double, or triple,
mention "UC Berkeley Computer Sciences".


From: P. Aston <>
Date: Tue, 7 Jul 92 16:34:34 BST
Subject: Two Positions at the University of Surrey

Dept of Mathematical and Computing Sciences
Guildford, Surrey, U.K.


Applications are invited for two posts of Lecturer in Mathematics.
Both pure and applied mathematicians with research interests in areas
such as dynamical systems, bifurcation theory, chaos and associated
numerical methods or other branches of nonlinear pure and applied
mathematics will be considered. Candidates should possess, or expect
to gain shortly, a PhD in an appropriate discipline.

Salary will be on the Lecturer Grade A/B scale (12860-23739 pounds)
depending on age, qualifications and experience. Applications in the
form of a Curriculum Vitae (3 copies), including the names and
addresses of three referees,

University of Surrey
Surrey GU2 5XH

quoting reference number 094 by 31st July 1992. Further particulars
can also be obtained by writing to the above address or by
telephoning 0483-509279. Informal enquiries may be made to
Prof. R. Shail on 0483-509196.


From: Beth Gallagher <>
Date: Thu, 09 Jul 92 16:18:09 EST
Subject: Contents: SIAM Matrix Analysis and Applications

SIAM Journal on Matrix Analysis and Applications
October 1992 Volume 13, Number 2

Exponential Convergence Properties of Autocovariance Matrix
Inverses and Latent Vector Prediction Coefficients
John L. Eltinge

A Bisection Method for Solving Two-Parameter Eigenvalue Problems
Xingzhi Ji

Algorithms for Minimum Trace Factor Analysis
G. A. Watson

Estimating the Largest Eigenvalue By the Power and Lanczos
Algorithms with a Random Start
J. Kuczynski and H. Wozniakowski

Explicit Solutions of the Matrix Equation dA^XD_i= C
Harald K. Wimmer

Least-Index Resolution of Degeneracy In Linear
Complimentary Problems with Sufficient Matrices
Richard W. Cottle and Yow-Yieh Chang

Computing the Structured Singular Value
G. A. Watson

Inverse Eigenvalue Problems for Symmetric Toeplitz Matrices
Shmuel Friedland

Equality Cases in Matrix Exponential Inequalities
Wasin So

Backward Error Analysis for a Pole Assignment Algorithm II: The
Complex Case
Christopher L. Cox and William F. Moss

A Singular Value Decomposition Updating Algorithm for Subspace
Marc Moonen, Paul van Dooren, and Joos Vandewalle

Circulative Matrices of Degree theta
Hsin-Chu Chen

Parallel Solution of Large Lyapunov Equations
A. Scottedward Hodel and Kameshwar Poolla

Jacobi's Method is More Accurate than QR
James Demmel and Kresimir Veselic

A Tricyclic Diagonal Equation Solver
David S. Dodson and Stewart A. Levin

A Projective Method for Rescaling a Diagonally Stable Matrix to be
Positive Definite
H. Hu

Adaptive Condition Estimation for Rank-One Updates of QR
Gautam M. Shroff and Christian H. Bischof

Generalizations of the Singular Value and QR Decompositions
Bart de Moor and Paul van Dooren

Rank Detection Methods for Sparse Matrices
Jesse L. Barlow and Udaya B. Vemulapati

Modifying the QR Decomposition to Constrained and Weighted inear
Least Squares
Mrten Gulliksson and Per-ke Wedin


End of NA Digest