NA Digest Tuesday, July 28, 1987 Volume 87 : Issue 63

This weeks Editor: Cleve Moler

Today's Topics:

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


Posted-From: The MITRE Corp., Bedford, MA
To: [email protected]
Subject: Symbolic Manipulators
Date: Fri, 24 Jul 87 09:21:22 EDT
From: James R. Van Zandt <[email protected]>

> ... does anyone care to volunteer opinions concerning microcomputer-based
> symbolic manipulation software (such as mu-math)?
>
> Baker Kearfott
> [email protected]

I've spent some time with muMATH. I think it needs work in three areas:

1. The command language is hard to work with. Control variables
determine whether, for example, expressions are distributed or
factored. Positive values cause distribution while negative values
cause factoring. The different types of expressions (numerical,
sums, others) are each assigned a prime number, and the value
assigned to a variable is the product of the corresponding set of
primes. This is even worse than the usual scheme, of summing the
correct set of powers of 2.

2. Expressions are displayed in a linear format similar to FORTRAN
rather than a 2D format with built-up fractions, etc. However, the
latest Soft Warehouse newsletter had an ad for CALC-87, available
in source code, which is supposed to include this. (source: R.
Freese or D. Stegenga, Dept. of Mathematics, U of Hawaii,
Honolulu HI 96822, $35. Caveat emptor - I haven't tried it.)

3. There are no convenient ways to select a subexpression, perform
some operations on it, and substitute it back. I've made some
progress on this one.

One would expect size limitations on a microcomputer system, and in
fact the documentation includes a fair amount of discussion on space
optimization. However, the documentation (and program, for the most
part) dates from 1983, when 640K machines were less common. (muMATH
can use up to 320K.) I've run out of space only due to programming
errors (infinite recursion).

Unfortunately, I've had no significant experience with other symbolic
manipulation systems, so I can't give many comparisons. However, I
notice that in MATHLAB (a precursor of MACSYMA) you could specify which
variables were more important so that, for example, in a polynomial the
like powers of a given variable would be collected. muMATH doesn't let
you do this.

Has anybody else been working on extending muMATH? Would the na library
accept contributions of muMATH source code?

- Jim Van Zandt

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

Date: Fri, 24 Jul 87 09:32:08 MDT
From: Roland Sweet <[email protected]>
To: [email protected]
Subject: Whereabouts of one Bernadette Miara

Does anyone know the whereabouts of Bernadette Miara. Several months ago she
sent me a message, but my reply to her via nanet has been returned as undeliv-
erable. I would appreciate any help in determining her address--electronic or
postal. Thanks.

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

Date: Sun, 26 Jul 87 16:18 MST
From: Juan-Carlos Lerman <[email protected]>
Subject: Position Wanted
To: [email protected]

From: JUAN-CARLOS LERMAN <[email protected]>
To: EVERYONE on the NA NETWORK
Re: POSITION WANTED
M.S. (Buenos Aires), Ph.D. (Paris), I am currently looking for a position
in research or/and teaching or/and service.
My expertise is in the field of numerical and computational experiments and
simulations, including computer techniques, numerical analysis, exploratory
statistics, time-series and data analysis. I have published applications to
diverse areas such as geophysics and geochemistry, and physiology.
Recently contributed to the development of new methods to study spacial and
temporal data sequences, of application also to biomedical areas of clinical
importance. This work was supported by a Senior Fellowship award from the NIH
(National Institutes of Health).
This award allowed me to work during several years in a multidisciplinary
field relating mathematics, dynamical systems, computations, physics, biology,
and medicine.
I can provide a curriculum vitae, reviewers comments, publications lists,
and other information by e-mail.
Phones: (602) 795-6534 at home, and (602) 621-4664 at the Applied Math
Program, Univ. of Arizona, Tucson, AZ 85721.
E-mail addresses: na.lerman in the NA NETWORK
and also at: newell1@arizrvax in BITNET
(THESE ARE MULTIUSER ACCOUNTS. PLEASE MENTION MY NAME IN MESSAGE'S HEADINGS.)

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

Date: Tue, 28 Jul 87 20:25:33 CDT
From: "Douglas N. Arnold" <arnold%[email protected]>
To: [email protected]
Subject: Conference Honoring Jim Douglas

***** Second Announcement and registration form *****

ADVANCES IN COMPUTATIONAL MODELLING AND NUMERICAL ANALYSIS

University of Chicago

September 10, 11, and 12, 1987

A conference in honor of Jim Douglas, Jr.
on the occasion of his sixtieth birthday.


The conference will focus on recent advances in techniques of
scientific computation, especially the numerical solution of
differential equations, and their mathematical analysis. The
conference will consist of invited lectures with ample time for
discussion. Lectures will be presented by

Douglas Arnold Jacques-Louis Lions
Ivo Babuska Mitchell Luskin
James Bramble Jean-Claude Nedelec
Franco Brezzi Joachim Nitsche
Craig Douglas Vidar Thomee
Roland Glowinski Mary Wheeler
Herbert Keller

REGISTRATION: Conference registration is $10 if received by
August 15. If paid after August 15 or at the conference,
registration is $15. A banquet in honor of Professor Douglas
will be held on Friday, September 11, at the Quadrangle Club.

ACCOMODATIONS: Rooms are available at the Hyde Park Hilton at a
special conference rate of $65 for a single or double room. To
receive these rates your reservations must be received by the
conference secretary by August 15. Reservation requests
received thereafter will be subject to room availability and
will not benefit from the reduced conference rate.

CONFERENCE SUPPORT: The conference is supported by the National
Science Foundation, the Air Force Office of Scientific Research,
and the Office of Naval Research.


To register use the form below. Make checks payable to the
University of Chicago. For more information write:

Conference Secretary
Department of Mathematics
University of Chicago
5734 Univerisity Avenue
Chicago, Illinois 60637
phone: (312) 702-7100

--------- detach here and return to the Conference Secretary --------

Name: ___________________________________________ Phone: ____________

Address: ___________________________________________________________

___________________________________________________________

___________________________________________________________

_____ Yes, I plan to attend. Enclosed please find my registration
fee of $10 ($15 after August 15).

_____ I will attend the banquet. An additional $20 is enclosed.

----- Please reserve a single room for me. I attend to arrive on
______________ and depart on ______________.

_____ Please reserve a double room for me and a companion. We intend
to arrive on ______________ and depart on ______________.

_____ Please reserve a double room for me and _______________________
who will also attend the conference. We intend to arrive on
______________ and depart on ______________.

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

End of NA Digest

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