Algorithm-Evolutionary-Fitness

 view release on metacpan or  search on metacpan

Changes  view on Meta::CPAN

	* lib/Algorithm/Evolutionary/Op/VectorCrossover.pm (apply): Fixed
	bug: the incoming parent was modified along with the offspring. 

2012-07-10  Juan J. Merelo Guervós  <jjmerelo@gmail.com>

	* MANIFEST: Eliminating pod.t also in production code.

2012-07-09  Juan J. Merelo Guervós  <jjmerelo@gmail.com>

	* MANIFEST: Eliminated pod-coverage.t, which is not needed for
	production and causes errors in some systems. 

2012-07-08  Juan J. Merelo Guervós  <jjmerelo@gmail.com>

	* MANIFEST: Adding missing test files

	* lib/Algorithm/Evolutionary.pm: Starting a maintenance version,
	which will include the bugs fixed since last summer. 
	Since there's a bug fix, I think I'm going for full 0.78...

2012-05-21  Juan J. Merelo Guervós  <jjmerelo@gmail.com>

Changes  view on Meta::CPAN

	* lib/Algorithm/Evolutionary/Wheel.pm (spin): Changed spin
	implementation to make it faster and more efficient. It was
	sucking time out of mastermind EAs

2010-09-25  Juan J. Merelo  <jmerelo@sheldon>

	* lib/Algorithm/Evolutionary/Fitness/Rastrigin.pm (Rastrigin):
	Fixed formula

	* lib/Algorithm/Evolutionary.pm: Starting 0.74_1 with cosmetic
	changes and an attempt to fix sporadic test errors.  

2010-09-24  Juan Julian Merelo Guervos  <jmerelo@usuario-desktop>

	* lib/Algorithm/Evolutionary/Fitness/Rastrigin.pm (Rastrigin): Added first floating-point fitness func.

	* lib/Algorithm/Evolutionary/Utils.pm (random_number_array): Added this function

	* lib/Algorithm/Evolutionary/Individual/BitString.pm (from_string): Added from_string ctor

	* lib/Algorithm/Evolutionary/Individual/String.pm (from_string): Changed name and slightly modified

Changes  view on Meta::CPAN

	Starting documentation revision for developer releases.

	* lib/Algorithm/Evolutionary/Op/EDA_step.pm (apply): Added new
	module implementing estimation of distribution algorithms; added
	test and example in /examples

2009-07-30  Juan Julian Merelo Guervos  <jmerelo@geneura.ugr.es>

	* lib/Algorithm/Evolutionary.pm: Moved a GUI example to the
	scripts dir so that it's installed automatically. Expanded
	documentation to reflect that. Fixed also error (population
	explosion) in Easy_MO.

2009-07-28  Juan Julian Merelo Guervos  <jmerelo@geneura.ugr.es>

	* lib/Algorithm/Evolutionary.pm: Starting 0.70_1 with the
	classical documentation fixes. 

	* MANIFEST: 0.70 fixes a bug in Easy.pm and those derived from it:
	last population element was not eliminated. An animated GIF output
	has been added, and a bug in Makefile.PL that caused problems with

Changes  view on Meta::CPAN

	* lib/Algorithm/Evolutionary.pm (import): 0.69_1 fixed some
	problems with missing prerrequisites in the test programs, and
	some copy/paste stuff left from creating new modules. 0.70 starts
	with the intention of finally having a GUI.

2009-07-24    <jmerelo@localhost.localdomain>

	* lib/Algorithm/Evolutionary.pm: 0.68 was shortlived, but 0.69
	more or less the same. 0.69 included missing file and fixed some
	docs errors; 0.69_1 starts with the same objective, let's see how
	it ends.

2009-07-24  Juan Julian Merelo Guervos  <jmerelo@geneura.ugr.es>

	* TODO: 0.68 ends with several new test functions added (Trap,
	ECC), and finally a multiobjective evolutionary algorithm, not
	very good, but in working order.

2009-03-29    <jmerelo@localhost.localdomain>

lib/Algorithm/Evolutionary/Fitness/ECC.pm  view on Meta::CPAN

    my $p_peaks = Algorithm::Evolutionary::Fitness::ECC->new( $number_of_codewords, $min_distance );

=head1 DESCRIPTION

Extracted from article "Effects of scale-free and small-world topologies on binary coded self-adaptive CEA", by Giacobini et al [Ga]. Quoting:

"                                                    The ECC problem was presented in
[MW]. We will consider a three-tuple (n, M, d), where n is the length of each codeword
(number of bits), M is the number of codewords, and d is the minimum Hamming
distance between any pair of codewords. Our objective will be to find a code which
has a value for d as large as possible (reflecting greater tolerance to noise and errors),
given previously fixed values for n and M . The problem we have studied is a simplified
version of that in [MW]. In our case we search half of the codewords (M/2) that will
compose the code, and the other half is made up by the complement of the codewords
computed by the algorithm"

[Ga] Mario Giacobini, Mike Preuss, Marco Tomassini: Effects of Scale-Free and Small-World Topologies on Binary Coded Self-adaptive CEA. EvoCOP 2006: 86-98.

[MW] F. J. MacWilliams and N. J. A. Sloane. The Theory of Error-Correcting Codes. North-
    Holland, Amsterdam, 1977.