readme.ant

标准的GP源代码,由Andy Singleton维护

*************************************************************************
GPDATA Version 3 release Easter 1997 W.B.Langdon for the Artificial
Ant problem $Revision: 1.5 $. This updates 
GPQUICK
A simple Genetic Programming system in C++
Version 2, released 2/12/94
*************************************************************************

 Copyright Andy Singleton, 1993,1994
 This code is released for non-commercial use only
 For questions or upgrades contact:
    Andy Singleton
    Creation Mechanics, Inc.
    PO Box 248, Peterborough, NH 03458
    Internet: [e-mail]
    Compuserve: 73313,757
    Phone: [phone]

 Further documentation appears in the associated Byte Magazine article
 "Genetic Programming with C++", February, 1994

The sample problem ANT will attempt to evolve an artificial ant
which can traverse the trail of food.  The ant moves across a 32x32 grid,
and it can sense food immediately ahead, move forward, or turn.  As the
problem runs, it will display the best ant by the amount of food found.
After the ant finds all of the food (89 pieces on the Santa Fe trail), or
after 25,000 generates, the problem terminates and prints the path of the
best ant to file "RESULT.TXT".  This is a replication of the problem
described by John Koza in "A-Life II" and "Genetic Programming".

*************************************************************************
Instructions
*************************************************************************

Two versions of the ant problem are supplied. The original antprob
from Andy Singleton and one integrated into the GPxxx framework.  The
GPxxx framework is designed for batch runs but gives much greater
control of run parameters.  (The configuration macro WITHIN_GP must be
defined to compile in the code to produce GPant). GPant is currently
defined so that runs continue until pGenerationLimit is reached. To
stop on producing a solution change max fitness from 90 to 89. If the
configuration macro KOZA_ANT is defined, GPxxx uses a torriodal grid
and is more in keeping with that used in Koza's GP1 book.

To Make antprob
	Create a symbolic link from antprob.h to prob.h 
	(eg ln -s antprob.h prob.h)

	If gcc is available, make using antprob.make
	If running on an alpha, make using antprob.make.alpha 


To Make GPant:
	Create a symbolic link from ant.h to prob.h (eg ln -s ant.h prob.h)
	Either use default for KOZA_ANT or edit prob.h so as not to define it.

	If gcc is available, make with either ant.make 
	Eg make -f ant.make

	If running on an alpha, make with either ant.make.alpha

To RUN antprob:
	antprob can be run as it stands

To run GPant:
	copy either ant.ini or ant.koza.ini to gp.ini Create an
	empty file called gp.test Then run GPant

	GPant takes optionals command line arguments, these allow the
	user to override the paramter settings in gp.ini Eg

		GPant pPopSeed:100

	Evolved solutions can be loaded and their execution traced by
	creating a file gp.load containing the program. Three examples
	from GP are provide in files  

		gp.load.ant.avoider
		gp.load.ant.quilter
		gp.load.ant.solution