individual.h File Reference

#include "kernel/individual.h"
#include "kernel/decision_vector.h"
#include "kernel/problem.h"
#include "kernel/random.h"
#include "kernel/gp/gene.h"
#include "utility/matrix.h"
#include "utility/misc.h"
#include <iosfwd>
#include <iterator>
#include <set>
#include "kernel/gp/individual_exon_view.tcc"
#include "kernel/individual_format.h"

Go to the source code of this file.

Classes
class	ultra::gp::exon_view
class	ultra::gp::const_exon_view
struct	ultra::gp::decision_coordinate
class	ultra::gp::individual

Namespaces
namespace	ultra

Typedefs
using	ultra::gp::decision_vector = ultra::decision_vector<decision_coordinate>
	GP-specific decision-vector type used for numerical optimisation.

Functions
unsigned	ultra::gp::active_slots (const individual &)
individual	ultra::gp::crossover (const problem &, const individual &, const individual &)
unsigned	ultra::gp::distance (const individual &, const individual &)
decision_vector	ultra::gp::extract_decision_vector (const individual &ind)
locus	ultra::gp::random_locus (const individual &prg)

Detailed Description

Remarks

This file is part of ULTRA.

Copyright

Copyright (C) 2024 EOS di Manlio Morini.

License

This Source Code Form is subject to the terms of the Mozilla Public License, v. 2.0. If a copy of the MPL was not distributed with this file, You can obtain one at http://mozilla.org/MPL/2.0/

Function Documentation

extract_decision_vector()

decision_vector ultra::gp::extract_decision_vector ( const individual & ind )

nodiscard

Extracts the decision vector associated with an individual.

Returns

a valid decision vector

The returned decision_vector provides a flat, optimiser-friendly view of all tunable scalar parameters (currently D_DOUBLE and D_INT) appearing in the active portion of the genome.

Note

Only parameters in active genes (exons) are included. Inactive code is ignored.

Integer parameters are converted to double for optimisation. The original type information is preserved in decision_vector::coordinate::kind and restored when applying the vector.

The ordering of parameters is deterministic for a fixed individual and reflects the traversal order of its active genes. The returned vector is therefore compatible with apply_decision_vector() as long as the structure of the individual is unchanged.