``````use strict;
use warnings;

Algorithm::Evolutionary::Op::DeltaTerm - Termination condition for an algorithm; checks that
the difference of the best to a target is less than a delta

my \$target = 1;
my \$epsilon = 0.01;
my \$dt = new Algorithm::Evolutionary::Op::DeltaTerm \$target, \$epsilon;
#\$dt->apply( \@pop ) when the best fitness is 1 plus/minus 0.1

L<Algorithm::Evolutionary::Op::Base|Algorithm::Evolutionary::Op::Base>

Termination condition for evolutionary algorithm loops; the C<apply>
method returns false when the first element in the array is as close
to the target as the differente indicated.

=cut

package Algorithm::Evolutionary::Op::DeltaTerm;
use Carp;

our (\$VERSION) = ( '\$Revision: 3.0 \$ ' =~ / (\d+\.\d+)/ ) ;

use base 'Algorithm::Evolutionary::Op::Base';

=head2 new( \$target[, \$delta] )

Creates a new terminator. Takes as parameters the target and the
epsilon (or delta, whatever you want to call it):

my \$target = 1;
my \$epsilon = 0.01;
my \$dt = new Algorithm::Evolutionary::Op::DeltaTerm \$target, \$epsilon;

Delta can be 0, which means that application of this operator will
return true only when the first element fitness is the same as the
target. Use this judiciously when your fitness is a floating
point number.

=cut

sub new {
my \$class = shift;
my \$target = shift || croak "No target here!";
my \$delta = shift ; #  Could be 0
my \$hash = { target => \$target,
delta =>  \$delta };
my \$self = Algorithm::Evolutionary::Op::Base::new( \$class, 1, \$hash );
return \$self;
}

=head2 apply( \$population )

Will return true while the difference between the fitness of the first element
in the population and the target is less than C<\$delta>, true otherwise

\$dt->apply( \@pop ) == 1

if the target has not been reached. Population must be sorted before this.

=cut

sub apply (\$) {
my \$self = shift;
my \$pop = shift;

return abs( \$pop->[0]->Fitness() -  \$self->{_target} ) > \$self->{_delta};

}

L<Algorithm::Evolutionary::Op::FullAlgorithm> needs an object of this class to check
for the termination condition. It's normally used alongside "generation-type"
objects such as L<Algorithm::Evolutionary::Op::Easy>.

There are other options for termination conditions: L<Algorithm::Evolutionary::Op::NoChangeTerm|Algorithm::Evolutionary::Op::NoChangeTerm> and
L<Algorithm::Evolutionary::Op::GenerationalTerm>.