desdeo_emo.recombination

This module provides implementations of various crossover and mutation operators.

Submodules

Package Contents

Classes

BioGP_mutation

BioGP_xover

BP_mutation

EvoDN2Recombination

EvoNNRecombination

SBX_xover

Simulated binary crossover.

SinglePoint_Xover

Simple single point crossover and mutation.

SinglePoint_Mutation

Simple single point crossover and mutation.
class desdeo_emo.recombination.BioGP_mutation(probability_mutation: float)[source]
do(offspring)
class desdeo_emo.recombination.BioGP_xover(probability_crossover: float = 0.9, probability_standard: float = 0.5)[source]
do(pop, mating_pop_ids)
class desdeo_emo.recombination.BP_mutation(lower_limits: numpy.ndarray, upper_limits: numpy.ndarray, ProM: float = None, DisM: float = 20)[source]
do(offspring: numpy.ndarray)

Conduct bounded polynomial mutation. Return the mutated individuals.

Parameters:

offspring (np.ndarray) – The array of offsprings to be mutated.

Returns:

The mutated offsprings

Return type:

np.ndarray

class desdeo_emo.recombination.EvoDN2Recombination(evolver: BaseEA, ProC: float = 0.8, ProM: float = 0.3, mutation_strength: float = 1.0)[source]
do(pop, mating_pop_ids: list = None)
class desdeo_emo.recombination.EvoNNRecombination(evolver: BaseEA, ProC: float = 0.8, ProM: float = 0.3, mutation_strength: float = 1.0, mutation_type: str = 'gaussian')[source]
do(pop, mating_pop_ids: list = None)
class desdeo_emo.recombination.SBX_xover(ProC: float = 1, DisC: float = 30)[source]

Simulated binary crossover.

Parameters
ProCfloat, optional

[description], by default 1

DisCfloat, optional

[description], by default 30

do(pop: numpy.ndarray, mating_pop_ids: list = None) → numpy.ndarray
Consecutive members of mating_pop_ids are crossed over

in pairs. Example: if mating_pop_ids = [0, 2, 3, 6, 5] then the individuals are crossover as: [0, 2], [3, 6], [5, 0]. Note: if the number of elements is odd, the last individual is crossed over with the first one.

Parameters:

  • pop (np.ndarray) – Array of all individuals

  • mating_pop_ids (list, optional) –

    Indices of population members to mate, by default None, which shuffles and

    mates whole population

Returns:

The offspring produced as a result of crossover.

Return type:

np.ndarray

class desdeo_emo.recombination.SinglePoint_Xover(ProC: float = 1, DisC: float = 30)[source]

Simple single point crossover and mutation.

Parameters
ProCfloat, optional

[description], by default 1

DisCfloat, optional

[description], by default 30

crossover(ind_0, ind_1)

Single point crossover. :param ind_0: individual_0 :param ind_1: individual_1

Ret:

new_0, new_1: the individuals generatd after crossover.

mutation(indi)

Simple mutation. Arg:

indi: individual to mutation.

do(pop: numpy.ndarray, mating_pop_ids: list = None) → numpy.ndarray
Consecutive members of mating_pop_ids are crossed over

in pairs. Example: if mating_pop_ids = [0, 2, 3, 6, 5] then the individuals are crossover as: [0, 2], [3, 6], [5, 0]. Note: if the number of elements is odd, the last individual is crossed over with the first one.

Parameters:

  • pop (np.ndarray) – Array of all individuals

  • mating_pop_ids (list, optional) –

    Indices of population members to mate, by default None, which shuffles and

    mates whole population

Returns:

The offspring produced as a result of crossover.

Return type:

np.ndarray

class desdeo_emo.recombination.SinglePoint_Mutation(ProC: float = 1, DisC: float = 30)[source]

Simple single point crossover and mutation.

Parameters
ProCfloat, optional

[description], by default 1

DisCfloat, optional

[description], by default 30

mutation(indi)

Simple mutation. Arg:

indi: individual to mutation.

do(offspring: numpy.ndarray)
Consecutive members of mating_pop_ids are crossed over

in pairs. Example: if mating_pop_ids = [0, 2, 3, 6, 5] then the individuals are crossover as: [0, 2], [3, 6], [5, 0]. Note: if the number of elements is odd, the last individual is crossed over with the first one.

Parameters:

  • pop (np.ndarray) – Array of all individuals

  • mating_pop_ids (list, optional) –

    Indices of population members to mate, by default None, which shuffles and

    mates whole population

Returns:

The offspring produced as a result of crossover.

Return type:

np.ndarray