Summary: Wanted: Software for population dynamics modeling
This is the promised summary, which unfortunately doesn't enclose more
than half a dozen messages, from the following persons which I want to
thank for their replies and interest (I haven't been able to follow all
pointers yet):
apk at Athena.MIT.EDUbernasconi1 at urz.unibas.ch, Giorgina Bernasconi
fuellen at Athena.MIT.EDU, Georg Fuellen
GBZA51 at VMS3.GLASGOW.ac.uk, Niall Barr
skjetnep at nobipol.unit.no, Paal
Let me elaborate a bit on what I've been doing so far. Using C, I've
written a basic version with a concrete population with 1000 hosts and
+/- infinite number of parasites. This version runs fine as it is, and
I've used it now for the first complete assignment with more than 12'000
runs.
The pseudo random number generator I've used was provided by an assistant
of the local statistics institute, and I've just changed a few lines.
The major problem with these subroutines is that they are written in
Pascal, whereas all my routines are in plain ANSI C. The only Macintosh
specific routines are responsible for run-time linking of the Pascal
routines. Before trying my code on a mainframe computer to see if it
could pay to shift to another operating system and compiler, I'd like to
convert the Pascal part to C, and will try to find a book called
"Numerical Recipes in C - The art of scientific computing"
William H. Press, Saul A. Teukolsky, William T. Vetterling
and Brian P. Flannery,
recommended by skjetnep at nobipol.unit.no. Another approach would be to use
one of the available public domain source code collections, e.g.
"mrandom", which can be downloaded from theory.lcs.mit.edu (copy of the
announcement enclosed at the end of this message). I've seen other
sources, but haven't yet tried to use any of them on the mac. Also, Niall
Barr (email address see above) is working on a teaching package, using
C++.
At the local university, I also found a copy of Mathematica to play with,
and up to now I can't see any advantages of using the built-in
programming language for my purpose. It seems then that I'm going to
follow my original course of writing my own programs, with all the
compromises and inevitable debugging sessions.
If you have further suggestions, ideas and insights you want to share
with me, just write me. - Mark (frischknecht at esh.unibe.ch)
Announcement for "mrandom" package:
> Newsgroups: sci.math.stat
> From: cthombor at theory.lcs.mit.edu (Clark D. Thomborson)
> Subject: mrandom version 3.0 available
> Organization: MIT Lab for Computer Science
> Date: Thu, 10 Jun 1993 00:30:40 GMT
>> Version 3.0 of mrandom is now available for alpha-test by anon ftp
> from theory.lcs.mit.edu, directory pub/cthombor/Mrandom/V3.0. It has
> made successfully on my Sun SS1+ and on an Athena DecStation. It
> doesn't make on an HP workstation running HP's non-bsd Unix. Maybe
> I'll put a non-bsd option in version 3.1...
>> Like its predecessors (version 2.3 of 9/92 and version 1.0 of 12/91),
> mrandom 3.0 is designed to aid the development of portable C-language
> codes for pseudorandom experimentation under 4.3bsd Unix. Version 3.0
> gives a standardized and efficient interface to 9 different RNGs,
> selectable at runtime, plus a number of utility routines. It is
> designed to make it easy for anyone to "plug in" new RNGs.
>> Comments & bug reports to cthombor at lcs.mit.edu, please.
--
Markus Frischknecht Internet: frischknecht at esh.unibe.ch
Universitaet Bern, Wohlenstrasse 50a, CH-3032 Hinterkappelen, Switzerland
"Idealism precedes experience; cynicism follows." David T. Wolf