SCS-C(1) EDUCATIONAL TOOLS SCS-C(1)
NAME
scsc - a simple classifier system in C
SYNOPSIS
scsc [ options ]
[-N, --no-noise-in-auction]
[-S, --no-specificity-in-auction]
[-b, --batch]
[-h, --help]
[-V, --version]
[-W, --warranty]
[-c, --classifier-data <classifier data file name>]
[-d, --detector-data <detector data file name>]
[-e, --environmental-
data <environmental data file name>]
[-g, --genetics-data <GA data file name>]
[-r, --reinforcement-
data <reinforcement data file name>]
[-t, --timekeeper-data <timekeeper data file name>]
[-l, --log-file <log file name>]
[-p, --plot-file <plot file name>]
[-s, --seed-for-random <seed value>]
[-T, --time-steps <number>]
DESCRIPTION
scsc implements the simple classifier system SCS-C; a C
translation of Goldberg's Pascal SCS (refer to the SEE ALSO
section below).
scsc's input and output streams, and some of it's default
parameters may be changed using various options (see the
OPTIONS section below).
Version 1.3 Last change: 23 October 1994 1
SCS-C(1) EDUCATIONAL TOOLS SCS-C(1)
OPTIONS
-N, --no-noise-in-aution
Turn off the noise generating function during the
apportionment of credit.
-S, --no-specificity-in-auction
Turn off default hierarchy seeking (you can also assign
zero values to classifier-data parameters bid2 and
ebid2.)
-T, --time-steps number
Run the system for number of timesteps. (Note: on
non-UNIX machines the user may interrupt scsc by hit-
ting a key on the keyboard. He is then prompted if he
want's to abort execution, or go on: Halt (y/n)? On
UNIX systems this facility is currently not imple-
mented, we use the shell's job control mechanism.)
-b, --batch
Run scsc in batch mode, ie. don't prompt the user for
any input; eg. the random number generator's seed is
initialized with a default value (0.02021965), unless
specified otherwise with the `-s, --seed-for-random'
flag.
-c, --classifier-data filename
The classifier data will be read in from filename
(defaults to scs.cls).
-d, --detector-data filename
The detector data will be read in from filename
(defaults to scs.det). (Unused in the contributed 6-
bit multiplexer example.)
-e, --environmental-data filename
The environmental data will be read in from filename
(defaults to scs.env).
-g, --genetics-data filename
The genetic algorithm parameters will be read in from
filename (defaults to scs.gen).
-h, --help
Print a usage message containing all options.
-l, --log-file filename
Logging information will go to filename (defaults to
scs.log).
-p, --plot-file filename
Plot data will go to filename (defaults to scs.plt).
Version 1.3 Last change: 23 October 1994 2
SCS-C(1) EDUCATIONAL TOOLS SCS-C(1)
-r, --reinforcement-data filename
The reinforcement parameters will be read in from
filename (defaults to scs.rfc).
-s, --seed-for-random seed value
Set the random number generator's seed with the real
seed value.
-t, --timekeeper-data filename
The time keeper's parameters will be read in from
filename (defaults to scs.tim).
-V, --version
Print a version and copyright message; you will also be
informed which random number generator is currently
compiled into scsc.
-W, --warranty
Print a warranty information message.
SEE ALSO
David E. Goldberg, Genetic Algorithms in Search, Optimiza-
tion, and Machine Learning, Addison-Wesley, Reading MA,
1989.
csh(1), jazz(1)
AUTHOR
Copyright (C) 1993, 1994 by Joerg Heitkoetter
Systems Analysis Group, University of Dortmund, Germany.
Research and Development Group, EUnet Deutschland GmbH.
Send bugs, comments, etc., to <joke@Germany.EU.net>
AVAILABILITY
This work is protected by the terms of the GNU General Pub-
lic License. Please refer to the LICENSE file accompanying
the sources of this software package for a lengthy, boring,
but absolute complete description.
SCS-C is available via ENCORE in it's classifier systems
section as file ../EC/CFS/src/scsc-1.3j.tar.gz
THANKS
To all of those who showed their interest and/or helped
debugging the program and sent fixes, in the order of
appearance in my mailbox:
Version 1.3 Last change: 23 October 1994 3
SCS-C(1) EDUCATIONAL TOOLS SCS-C(1)
Uwe Schnepf<Uwe.Schnepf@gmd.de>
Looser Bjoern<ia90loos@htl-bw.ch>
Marc Levitt<Marc.Levitt@Eng.Sun.COM>
Amin Hammad<c42394a@nucc.cc.nagoya-u.ac.jp>
Bill Gillock<whg@houston.geoquest.slb.com>
Wajid Khan<wajid@ee.umr.edu>
Friedrich Krebs<fritz@theory.informatik.uni-kassel.de>
Ms. Wan Hua<a92351@cs.ait.ac.th>
Sergio<sergio@cic.unb.br>
Jacques Bonneville<bonnev@lisisun.univ-lyon1.fr>
Loia Vincenzo<loia@udsab.dia.unisa.it>
Joey Baumgartner<jbaumgar@ringer.cs.utsa.edu>
Pankaj Kanwar<pankaj@magenta.me.fau.edu>
Joseph Taylor Potts<potts@cse.uta.edu>
Martyn Amos<martyn@dcs.warwick.ac.uk>
John Prior<jprior@soleil.esa.lanl.gov>
Frank Hoffmann<hoefi@hp1.ang-physik.uni-kiel.d400.de>
Marc M. Lankhorst<lankhors@cs.rug.nl>
Martin Schuetz<schuetz@edna.informatik.uni-dortmund.de>
D. Poulet<csc270@cent1.lancs.ac.uk>
Kwok-Ching Tsui<ktsui@dcs.kcl.ac.uk>
Jorge Gabriel Ponce de Leon Sanchez<al702782@academ01.mty.itesm.mx>
Andreas Frick<afr@aifb.uni-karlsruhe.de>
Version 1.3 Last change: 23 October 1994 4
