## NA Digest Sunday, April 25, 2004 Volume 04 : Issue 17

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

### Submissions for NA Digest:

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

Information via e-mail about NA-NET: Mail to na.help@na-net.ornl.gov.

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

Date: Tue, 20 Apr 2004 11:27:13 -0700 (PDT)
Subject: A Complex Symmetric Tridiagonal Eigenvalue Problem

Background

We use (x, y, z) do denote position in three coordinates. The
inhomogeneous Helmholtz equation is
[ d^2/dx^2 + d^2/dy^2 + d^2/dz^2 ] F (x, y, z) + k(x, y, z)^2 F (x, y, z) = 0

It is often useful to concentrate on the dependence of k on one spatial
coordinate x , leading to a mode eigenvalue equation
d^2/dx^2 f(x) + k0(x)^2 f(x) = b f(x)
where k0 resembles k in some appropriate way.

For the problems we are interested in, there is absorption of the wave,
which means that k0(x)^2 is complex valued. If one makes the standard
centered second difference approximation for d^2/dx^2 f(x) in a bounded
domain with homogeneous (e. g., f = 0 ) boundary conditions, one is led to
a matrix eigenvalue problem. The matrix is tridiagonal and symmetric with
complex elements along the diagonal and real elements above and below the
diagonal. The matrix it is not Hermitian; in fact we are interested in the
case where it is not even normal. Thus we are lead to the following
question:

Question

Suppose that D is an n by n real diagonal matrix, S is an n by n real
symmetric tridiagonal matrix, and D S is not equal to S D (the matrices do
not commute). What is the best available routine for finding the
eigenvalues and eigenvectors of the matrix S + i D when the dimension n is
between 1 and 1000 ?

It appears that the following article describes
a method that is appropriate:

A QL procedure for computing the eigenvalues of complex symmetric
tridiagonal matrices SIAM J. Matrix Anal App Vol 17, No 1, pp 83-109 1996.

It also appears that the authors have implemented this method in the
routine CMTQL1. Where can one download a copy of this routine ?

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

From: Martin Nigsch <nigsch@tph.tuwien.ac.at>
Date: Sun, 18 Apr 2004 20:42:17 +0200 (CEST)
Subject: Reference Values for Spheroidal Functions

Hello,

I'm looking for reference values for spheroidal functions. In the book
"Spheroidal Wave Functions in Electromagnetic Theory" (Li, Kang, Leong
2002) the authors state:

"It is also found that some of the previous computed results of spheroidal
eigenvalues and spheroidal harmonics are not accurate [...], although they
have been used elsewhere for many years [...]."

The references where that have been used for many years are 1) Flammer,
"Spheroidal Wave Functions" (1957) and 2) Abramovitz, Stegun, "Handbook of
Mathematical Functions" (1972).

My question now is: what reference values are in use nowadays to validate
implementations of spheroidal wave functions? Are there still numerical
discrepancies?

best regards,
Martin Nigsch
Institute for theoretical Physics
TU Vienna

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

From: Garry Tee <tee@math.auckland.ac.nz>
Date: Mon, 19 Apr 2004 10:48:53 +1200
Subject: Origin of Packed Format for Arrays

In last week's NA Digest, Fred Gustavson asked

I am trying to find out who first proposed packed format for symmetric and
triangular arrays. It appears that many people used packed format early on
and it then got adopted in early linear algebra codes such as the Handbook
of Automatic Computation, Eispack and Linpack. I do know that Velvel Kahan
used this format for a matrix package he developed in 1958.

The first software manual was issued by the University of Cambridge Mathematical
Laboratory in September 1950. The Director Dr M. V. Wilkes edited the
"Report on the Preparation of Programmes for the EDSAC and the Use of the
Library of Subroutines". That first edition was produced on a spirit duplicator
(in violet and crimson), and hence less than 100 copies could be printed.

The subroutines were carefully chosen and very skillfully written. But only
two subroutines deal with linear algebra: V2 Addition and subtraction of
n dimensional vectors, and V1 Multiplication of vector by symmetric matrix.
The specification for V1 states that "Given a symmetric n-by-n matrix of
which only (1/2)n(n+1) elements are stored" (Part 2, page 26), and the complete
subroutine (in EDSAC machine code) is given in Part 3, pages 38 & 39.

A revised version of that Cambridge manual was then published as "The
Preparation of Programs for an Electronic Digital Computer, with special
reference to the EDSAC, and the use of a library of subroutines", by Maurice
V. Wilkes, David J. Wheeler and Stanley Gill, Addison-Wesley Press, Inc.,
Cambridge 42, Mass. USA, 1951 (with many later editions). That book had
immense and beneficial influence upon computer programming. Subroutine V1
is printed (in black and white, not violet and crimson) on pages 102 & 103.

Garry J. Tee,
Department of Mathematics, University of Auckland,
Private Bag 92019, Auckland, New Zealand

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

From: Armin Iske <iske@ma.tum.de>
Date: Mon, 19 Apr 2004 13:41:37 +0200
Subject: New Book on Multiresolution Methods in Scattered Data Modelling

A. Iske: "Multiresolution Methods in Scattered Data Modelling",
Lecture Notes in Computational Science and Engineering, Vol. 37,
Springer-Verlag, Heidelberg, 2004, 182 p., ISBN 3-540-20479-2.

This application-oriented work concerns the design of efficient,
robust and reliable algorithms for the numerical simulation of
multiscale phenomena. To this end, various modern techniques from
scattered data modelling, such as splines over triangulations and
strategies. The resulting multiresolution methods are thinning
algorithms, multilevel approximation schemes, and meshfree
discretizations for transport equations. The utility of the
algorithmic approach taken in this research is supported by the
wide range of applications, including image compression,
hierarchical surface visualization, and multiscale flow simulation.
Special emphasis is placed on comparisons between the various
numerical algorithms developed in this work and comparable
state-of-the-art methods.

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

From: Connie Young <cyoung@siam.org>
Date: Fri, 23 Apr 2004 13:39:34 -0400
Subject: SIAM Annual Meeting in Portand

Conference Name: 2004 SIAM Annual Meeting

Location: Portland, Oregon

Dates: July 12-16, 2004

Program and registration information is available at:
http://www.siam.org/meetings/an04/

Meeting Registration Deadline: Monday, June 14, 2004
Hotel Reservation Deadline: Friday, June 11, 2004

For additional information visit http://www.siam.org/meetings/an04/ or
contact SIAM Conference Department at meetings@siam.org.

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

From: Iain Duff <I.Duff@rl.ac.uk>
Date: Thu, 22 Apr 2004 20:49:01 +0100 (BST)
Subject: Sparse Days at CERFACS

Sparse Days at CERFACS ... June 2-3, 2004

Here are more details about the Sparse Days at CERFACS the dates of which were
announced in a recent NA Digest.

The meeting will commence at 11.00 on Wednesday June 2 and will continue
until the afternoon of Thursday June 3, the exact time being dependent on the
number of talks.

There is no registration fee for the meeting but the cost of the conference
dinner on Wednesday evening will be about 30 euros.

Some low cost accommodation is available at the Meteo residences and
this may be waived for students or other needy cases. This accommodation is
somewhat limited, so you should let us know soon whether you want to stay
there. Communications for this and for hotel information should be made with
Brigitte Yzel (Brigitte.Yzel@cerfacs.fr). If you wish our assistance with
accommodation you must let Brigitte know before May 12th.

Although we are particularly interested by data assimilation applications,
any contribution involving sparse matrix research or applications is welcome.
A title and short abstract should be sent to Martin van Gijzen
(gijzen@cerfacs.fr) before May 12th. You should indicate the time, t, that you
would ideally like for your presentation, 15 min <= t <= 45 min and we
will try to accommodate this request within the constraint of having no
parallel sessions and finishing by 17.00 on June 3rd.

Regularly updated information will shortly be available by following the
indicated web links from the CERAFCS Parallel Algorithms Group web page
http://www.cerfacs.fr/algor/. In particular the tentative schedule will be
available there on May 15th.

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

From: Joab Winkler <joab@dcs.shef.ac.uk>
Date: Mon, 19 Apr 2004 07:50:47 +0100 (BST)
Subject: Workshop at University of Sheffield on Machine Learning

SHEFFIELD MACHINE LEARNING WORKSHOP
Sheffield, United Kingdom
Septemeber 7-10, 2004
www.dcs.shef.ac.uk/ml/workshop

The Machine Learning Research Group in The Department of Computer
Science at The University of Sheffield is organising a workshop on
deterministic and statistical methods in machine learning, with a
strong emphasis on mathematical and numerical methods.

REGISTRATION NOW OPEN

Confirmed invited speakers:

Chris Bishop (Microsoft Research, UK)
Andrew Blake (Microsoft Research, UK)
Herve Bourlard (IDIAP, Switzerland)
Andrzej Cichocki (RIKEN, Japan)
Jerry Eriksson (Umea, Sweden)
Zoubin Ghahramani (University College London, UK)
Gene Golub (Stanford, USA)
Josef Kittler (Surrey University, UK)
David Lowe (Aston University, UK)
David MacKay (Cambridge University, UK)
Ian Nabney (Aston University, UK)
Manfred Opper (Southhampton University, UK)
John Platt (Microsoft Research, USA)
Stephen Roberts (Oxford University, UK)
Bernard Schoelkopf (Max Planck Institute, Germany)
Michael Tipping (Microsoft Research, UK)
Chris Williams (Edinburgh University, UK)

Springer in the series Lecture Notes in Computer Science. The workshop
is supported by grants from the Engineering and Physical
Sciences Research Council, the London Mathematical Society,
Sheffield University, and the PASCAL European Network of
Excellence.

Registration is now open and early registration is encouraged
because (1) there is a limited number of places, and (b) there
is an early bird discount. Registration is available on-line at
www.dcs.shef.ac.uk/ml/workshop/.

For more details, contact the organisers:

Joab Winkler
Mahesan Niranjan
Neil Lawrence

E-mail: mlworkshop@dcs.shef.ac.uk

Department of Computer Science
The University of Sheffield
United Kingdom

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

From: Ron Boisvert <boisvert@nist.gov>
Date: Mon, 19 Apr 2004 11:41:44 -0400
Subject: Workshop in Washington on the Changing Face of Mathematical Software

Workshop on the The Changing Face of Mathematical Software
Washington, DC, USA
June 3-4, 2004

The IFIP Working Group 2.5 (Numerical Software) will be holding
one of its series of regional workshops on June 3-4, 2004 in
Washington, DC on the topic of "The Changing Face of Mathematical
Software". Participants will consider how math software architecture
and interfaces are responding to recent changes in the underlying
platforms, application contexts, as well as in the user community
itself. Speakers include Brian Ford, Ian Gladwell, William Mitchell,
Roldan Pozo, and Pete Stewart. The workshop will be limited to 40
participants. For further details see the workshop web site at
http://math.nist.gov/workshops/wg25-2004/.

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

From: Stefan Koerkel <stefan@koerkel.de>
Date: Thu, 22 Apr 2004 15:14:56 +0200 (CEST)
Subject: Workshop in Heidelberg on Parameter Estimation

Second Call for Participation and Papers
PARAOPE 2004

International Workshop on
Parameter Estimation and Optimal Design of Experiments
Numerical Methods and Applications

June 30 - July 2, 2004
University of Heidelberg, Germany

TOPICS include (but are not restricted to):

Theory of inverse problems
Theory of optimum experimental design
Model validation for dynamic processes
Large scale optimization
Numerical methods
Application problems

INVITED SPEAKERS

U. Ascher (University of British Columbia, Vancouver, Canada)
M. R. Osborne (Australian National University, Canberra, Australia)
V. H. Schulz (University of Trier, Germany)
H. Schwetlick (Technical University of Dresden, Germany)

CONTRIBUTED TALKS

Contributed talks (25 minutes) from researchers working on the
subjects Parameter Estimation and Optimal Design of Experiments will
be presented. Anyone interested in participating is invited to submit
a short abstract (1/2 to 1 page of length).

REGISTRATION

Registration is possible online on the website
www.iwr.uni-heidelberg.de/~agbock/conferences/2004/paraope/

Conference Fee: 100 Euro
to be paid in cash at the conference site.

IMPORTANT DATES

Registration: April 30, 2004
Notification of Acceptance: May 17, 2004
Workshop: June 30 - July 2, 2004

CONFERENCE PUBLICATIONS

It is planned to publish selected papers presented at the Workshop
as a special issue of Optimization Methods & Software (OMS) Journal
(http://www.tandf.co.uk/journals/titles/10556788.asp).

INFORMATION

Visit the conference website
http://www.iwr.uni-heidelberg.de/~agbock/conferences/2004/paraope/

ORGANIZATION

This workshop is organized by the Sonderforschungsbereich (SFB) 359
"Reactive Flows, Diffusion and Transport" and by the Postgraduate
Program "Complex Processes: Modelling, Simulation and Optimization" at
the Interdisciplinary Center (IWR) of the University of Heidelberg

Organizers:
H. G. Bock (IWR, University of Heidelberg)
S. Koerkel (IWR, University of Heidelberg)
E. Kostina (IWR, University of Heidelberg)
J. P. Schloeder (IWR, University of Heidelberg)

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

From: Georgios Stavroulakis <gestavr@cc.uoi.gr>
Date: Tue, 20 Apr 2004 12:50:33 +0300
Subject: Advanced Short Course in Italy on Nonsmooth Mechanics of Solids

International Centre for Mechanical Sciences
Centre International des Sciences Mecaniques

NONSMOOTH MECHANICS OF SOLIDS

J. Haslinger, and G.E. Stavroulakis

CISM, Udine, Italy, October 4 - 8, 2004
More details: in http://www.cism.it

Lecturers:

M. Fremond, Ch. Glocker, J. Haslinger, Z. Naniewicz, G.E. Stavroulakis

Short Description:

In many real-life problems coming from engineering or economics one can
encounter nondifferentiable or discontinuous functions and set-valued
mappings. A deep study of the properties of these maps including a
certain generalized differential calculus is the subject of nonsmooth
analysis. We shall focus on some problems in mechanics of solids which

The classical mechanics (statics and dynamics) of solids provide a large
number of nonsmooth effects: contact problems, collisions, stick-slip
motions connected with friction,
delaminations in composites. All these effects can be mathematically
described by means of differential inclusions. The mathematical research
in this area began at sixties assuming
multivalued parts to be represented firstly by maximal monotone
mappings, i.e. the case leading to variational inequalities. The
monotonicity assumption however turns out to be very restrictive. In
practice, we meet a lot of problems whose basic constitutive laws are no
longer monotone. At the beginning of eighties Prof. P.D. Panagiotopoulos
used tools of nonsmmoth analysis and introduced what he called
hemivariational inequalities (HE). HE's represent an appropriate
mathematical tool enabling us to involve nonmonotone multivalued
relations into the model. Due to HE's, the range of problems which can
be now rigorously treated is enlarged.

The goal of this course is to illustrate the potential of nonsmooth
analysis in modelling of various problems in mechanics of solids. The
emphasis will be laid on the completeness and mathematical
correctness of the presentation, although several industrial
applications will be presented. It will cover the following topics:
nonsmooth modelling of problems in mechanics of solids, the mathematical
theory of variational and hemivariational inequalities, approximation of
variational and hemivariational inequalities by finite element and
boundary element methods, the numerical realization (including smoothing
and regularisation techniques), algorithms and applications from civil
and mechanical engineering and related optimal design and identification
problems.

A number of well-known experts and active researchers in the field,
including mathematicians and engineers, will report on classical and new
results covering all the above mentioned topics. The presentation of all
these topics will be carefully balanced between theory, numerical
students, PhD candidates and young faculty members in mathematics,
physical sciences and engineering.

Engineers working on advanced applications of computational mechanics
and modelling of highly nonlinear and nonsmooth effects such as contact
and friction problems in industry (civil, aerospace, automotive) as well
as applied mathematicians and computer scientists (dealing with
nonsmooth analysis, optimisation, calculus of variations, computational
mechanics) will benefit from the course.

For more details and application forms, please vistit the web page of CISM
http://www.cism.it

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

From: Warren R Smith <smithwar@for.mat.bham.ac.uk>
Date: Fri, 23 Apr 2004 10:56:36 +0100 (BST)
Subject: Faculty Positions at the University of Birmingham

University of Birmingham
Lectureship in Applied Mathematics (Two Posts)

Applications are invited from strong researchers in any area of
to the Applied Mathematics Group, particularly in the area
of numerical and computational mathematics.

Informal enquiries can be made to Professor J R Blake of the School of
Mathematics and Statistics e-mail: j.r.blake@bham.ac.uk or
telephone: 0121 414 6577 or Professor G R Robinson, Head
of School of Mathematics and Statistics, e-mail: g.r.robinson@bham.ac.uk
or telephone: 0121 414 6581.

Starting salary on scale 22,191 - 33,679 pounds a year depending on
experience and qualifications. In addition, a 9,000 pounds "Golden Hello"
supplement may be available. The posts are available from 1 October 2004.

Application forms (returnable by 14 May 2004) and details from
Personnel Services, The University of Birmingham, Edgbaston,
Birmingham B15 2TT. Tel: 0121 415 9000, web: www.punit.bham.ac.uk/vacancies

Working towards equal opportunities.

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

From: Nick Higham <higham@maths.man.ac.uk>
Date: Mon, 19 Apr 2004 08:47:13 +0100
Subject: Faculty Position at University of Manchester

Lectureship in Applied Mathematics, University of Manchester (Ref. 431/04)

As part of its commitment to high quality research and teaching the
School of Mathematics at the University of Manchester invites
applications for a Lectureship in Applied Mathematics, available from
September 1, 2004 or an agreed date thereafter.

Interest in any branch of Applied Mathematics will be considered but
preference may be given to applicants who complement or reinforce the
School's existing strengths in continuum mechanics (solid and fluid
mechanics, waves, granular materials, combustion), dynamical systems,
industrial and biological modelling, inverse problems, financial
mathematics, numerical analysis, astrophysics and geophysics, or who
will enhance collaborative links with other groups within or outside
the University. Candidates should have, or be about to obtain, a
doctorate in an area of applied mathematics.

Salary will be in the range GBP22,191 - GBP25,451 per annum (Lecturer
Grade A) or GBP26,270 - GBP33,679 per annum (Lecturer Grade B) (under review)

Informal enquiries may be made to the Head of the School of
Mathematics: Professor Paul Glendinning tel: +44 (0)161-200-8972;
e-mail: p.a.glendinning@umist.ac.uk
or the current Heads of Applied Mathematics: Professor David
Abrahams, tel: +44(0)-161-275-5901; e-mail: i.d.abrahams@ma.man.ac.uk
Professor John Dold tel: +44(0)-161-200-3654;
e-mail: John.Dold@umist.ac.uk

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

Date: Fri, 23 Apr 2004 18:02:48 +0100 (BST)
Subject: Staff Position at the University of Manchester

We have a vacancy for a High Performance Computing Consultant to join the
HPC team at Manchester Computing, University of Manchester.

Manchester Computing supports world class research and teaching in all
disciplines, providing both a local and national supercomputing service
and undertakes research and development in high performance computing,
visualization and e-science.

We are seeking a HPC consultant who relishes the challenge of solving
complex problems on the latest state-of-the-art high performance
computers, and has a desire to provide the best possible service to the
user community.

The successful applicant will specialise in one of the following areas:
Computational Chemistry, Parallel Numerical Software or Computational
Mechanics/Fluid Dynamics

A good degree in an appropriate discipline is required. Experience of
multiprocessor computers, at least as a user, is essential, as is
familiarity with Unix/Linux and a high level programming language.
Experience in code optimisation and parallelisation would be an advantage.
Posts are initially for a period of two years, but with the possibility of
extensions. Starting salary will be in the range GBP 18,265 - GBP 33,679
pa (under review), depending on qualifications and experience.

For an application form and for further information, see
http://www.man.ac.uk/news/vacancies/tech.html#438

The closing date for applications is May 6th.

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

From: W. B. Liu <W.B.Liu@kent.ac.uk>
Date: Sun, 25 Apr 2004 09:30:38 +0100 (BST)
Subject: Vacancies at University of Kent

THE UNIVERSITY OF KENT - CANTERBURY BUSINESS SCHOOL

Professor/Reader - computational management science, mathematical
Programming, OR, stochastic optimisation, Systems (ref.A04/60)
Two Lecturers (ref. A04/62&63)

Location: Kent
Salary: Professor: minimum L40,841, Reader: L35,251 - L44,084
Senior Lecturer: L35,251 - L39,958pa; Lecturer: L22,191 - L33,679 pa;

Date posted: 20 Apr 04
Closing date: 21 May 04

Job description
undertaking a major development programme. This year we have made three
senior appointments - Martyn Jones as Director, Professor John Mingers as
Director of Research and Professor Paul Phillips as Deputy Director - as
well as six new lecturerships. We are helping to shape the new Transmanche
University in collaboration with four French institutions and are also
part of the University's establishment of a new campus at the Medway. The
MBA Programme is being redeveloped to include a modular version and a
public sector MBA. We have recently been awarded ESRC 1+3 PhD quotas and
the University received a greater increase in HEFCE funding than any
other.

The Business School is situated in the historic city of Canterbury and has
excellent links with London and Europe.

We now wish to move into the second phase of development to strengthen our
research profile with these appointments, mainly at senior levels. We
would be interested to hear from applicants across the field of management
who can demonstrate a strong commitment to research, but are particularly
interested in making appointments in the following areas. Applicants for
senior positions will have a track record of successful publication in

Particular areas of interest are mathematical programming/DEA, simulation,
soft OR/systems, statistics/forecasting, CRM/yield management, heuristics.
The senior post would involve an active role in course development and
Prof. John Mingers (J.Mingers@kent.ac.uk) or Prof. (Steve) WB Liu
(W.B.Liu@kent.ac.uk). Further particulars will soon be available on
http://www.kent.ac.uk/registry/personnel/vacancies.htm

Professor: minimum L40,841 pa;
Senior Lecturer: L35,251 - L39,958 pa;
Lecturer: L22,191 - L33,679 pa;

Some of the post carry an additional salary supplement for new lecturers
who meet certain criteria.

Closing date for receipt of applications is Friday 21 May 2004.
Interviews are likely be held in June/July.

We actively promote equal opportunity in education and employment and
welcome applicants from all sections of the community.

UNIVERSITY OF KENT - POSITIVE ABOUT DISABLED PEOPLE

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

From: Piet Wesseling <P.Wesseling@ewi.tudelft.nl>
Date: Tue, 20 Apr 2004 10:19:33 +0200
Subject: Assistant Professorship at Delft University of Technology

Assistant Professor in Numerical Analysis (permanent position)
Department of Applied Analysis
Faculty of Electrical Engineering, Mathematics and Computer Science
Delft University of Technology

The Department of Applied Analysis of Delft University of Technology in The
Netherlands has an open permanent position for an Assistant Professor
(Universitair Docent) in Numerical Analysis. The Assistant Professor will
belong to the Numerical Group, chaired by P. Wesseling. The Numerical Group
engages in teaching and research in the field of numerical analysis and
applications in the engineering sciences. The Group participates in the
teaching programs for the BSc and MSc degrees in Applied Mathematics, and for
engineering degrees in the University. Research is carried out in the field of
computational science and engineering. An important part of the research is
done by PhD students, is externally financed and is performed in cooperation
with partners from other disciplines and industry. Some of the current
problems we are working on include free boundary problems in materials
science, computational and mathematical modelling of optical recording,
efficient methods for large sparse systems of equations, and unstructured grid
methods in computational fluid dynamics.

Requirements include a PhD degree in applied mathematics or a subject in the
computational sciences, and preferably a few years of postdoctoral research
experience; research skills shown by publications of high quality; affinity
and experience with or aptitude for teaching; good communication skills and
interest in interdisciplinary work.

Tasks will include both research and teaching on a fifty-fifty basis.

For further information one may contact Professor P. Wesseling, phone
+3115 2783631, email P.Wesseling@ewi.tudelft.nl, or Ms G.M. Mense-Baars, phone
+3115 2783351, email G.M.Mense-Baars@ewi.tudelft.nl

Applications are to be sent to the Faculty of Electrical Engineering,
Mathematics and Computer Science, Afdeling P&O, Mekelweg 4, 2628 CD Delft, The
Netherlands. Please include the vacancy number EWI 2004.13 in the upper left

Closing date: May 15 2004

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

From: Ron Boisvert <boisvert@nist.gov>
Date: Mon, 19 Apr 2004 15:47:30 -0400
Subject: Contents, ACM Transactions on Mathematical Software

ACM Transactions on Mathematical Software
March 2004, Volume 30, Number 1

to full text, see http://www.acm.org/toms/V30.html.

S. Gonzalez-Pinto, J.I. Montijano, and S. Perez-Rodriguez
Two-Step Error Estimators for Implicit Runge-Kutta Methods Applied
to Stiff Systems
1-18

The Design and Implementation of a New Out-of-Core Sparse Cholesky
Factorization Method
19-46

A. Ismael F. Vaz, Edite M. G. P. Fernandez, and M. Paula S. F. Gomez
SIPAMPL: Semi-Infinite Programming with AMPL.
47-61

Roscoe A. Bartlett, Bart G. Van Bloemen Waanders, and Michael A. Heroux
Vector Reduction/Transformation Operators.
62-85

R. J. Hanson and Tim Hopkins
Algorithm 830: Another Visit with Standard and Modified Givens
Transformations and a Remark on Algorithm 539.
86-94

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

From: Science Direct <sciencedirect@prod.lexis-nexis.com>
Date: Thu, 22 Apr 2004 07:26:11 -0400 (EDT)
Subject: Contents, Nonlinear Analysis

Nonlinear Analysis Volume 57, Issue 1, Pages 1-152 (April 2004)

On the second part of Hilbert's 16th problem, Page 1
E. Oxenhielm

The number of bifurcation points of a periodic one-parameter ODE with at most two periodic solutions, Pages 3-22
J.L.Jose L. Bravo, Manuel Fernandez and Antonio Tineo

Random fixed points of uniformly Lipschitzian mappings, Pages 23-34
P.L.P. Lorenzo Ramirez

An alternating projection that does not converge in norm, Pages 35-61
H.S.Hein S. Hundal

Comparison principles for parabolic differential-functional initial-value problems, Pages 63-84
Agnieszka Bartomiejczyk and Henryk Leszczyski

On characterization of convexity for regularly locally Lipschitz functions, Pages 85-97
Duan Bednaik and Karel Pastor

The natural best approximant in Orlicz spaces of Young measures, Pages 99-110
C.C.Cristian Constantin Popa

Global asymptotic stability of a class of nonautonomous integro-differential systems and applications, Pages 111-135
Meng Fan and Xingfu Zou

Wave breaking for a shallow water equation, Pages 137-152
Yong Zhou

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

From: Claude Brezinski <claude.brezinski@univ-lille1.fr>
Date: Thu, 22 Apr 2004 13:55:35 +0200
Subject: Contents, Numerical Algorithms

CONTENTS
Numerical Algorithms
April 2004, Volume 35 (Issue: 2-4) ISSN: 1017-1398

Theory and Practice in Optimization
Guest Editors: Jose Mario Martinez, Jin Yun Yuan

137 Preface
J.M. Martinez, Jin Yun Yuan

X. Wu, B.P.B. Silva, J.Y. Yuan

155 A Computational Study of Global Algorithms for Linear Bilevel
Programming
C.H.M. de Saboia, M. Campelo, S. Scheimberg

175 Some Inexact Hybrid Proximal Augmented Lagrangian Algorithms
C. Humes Jr., P.J.S. Silva, B.F. Svaiter

185 Properties of the Central Points in Linear Programming Problems
C.C. Gonzaga, M. Cardia

205 Error Estimates for Two Filters Based on Polynomial
Interpolation for Recovering a Function from Its Fourier Coefficients
Musheng Wei, A.R. De Pierro, Jiahong Yin

233 A New Approach to a Multicriteria Optimization Problem
W. Sosa, F.M.P. Raupp

249 On Restart Procedures for the Conjugate Gradient Method
Yu-Hong Dai, Li-Zhi Liao, Duan Li

261 Recent Applications and Numerical Implementation of
Quasi-Newton Methods for Solving Nonlinear Systems of Equations
R. Perez, V.L. Rocha Lopes

287 A Function to Test Methods Applied to Global Minimization of
Potential Energy of Molecules
C. Lavor, N. Maculan

301 L-Curve and Curvature Bounds for Tikhonov Regularization
D. Calvetti, L. Reichel, A. Shuibi

315 Using a Conic Formulation for Finding Steiner Minimal Trees
M. Fampa, N. Maculan

331 An Acceleration Scheme for Solving Convex Feasibility Problems
Using Incomplete Projection Algorithms
N. Echebest, M.T. Guardarucci, H. Scolnik, M.C. Vacchino

351 A Class of Nested Iteration Schemes for Linear Systems with a
Coefficient Matrix with a Dominant Positive Definite Symmetric Part
O. Axelsson, Zhong-Zhi Bai, Shou-Xia Qiu

373 A Genetic Algorithm for Solving a Capacitated p-Median Problem
E. Santos Correa, M.T.A. Steiner, A.A. Freitas, C. Carnieri

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

End of NA Digest

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