evolutionary_optimization.genotype.implemented_genotypes package
Submodules
evolutionary_optimization.genotype.implemented_genotypes.binary_list_genotype module
- class BinaryListGenotype(genotype=None, mutation_probability=0.5, ratio_of_population_for_crossover=0.5, number_of_genes=32, value_range=(0, 1))[source]
Bases:
AbstractGenotype- __init__(genotype=None, mutation_probability=0.5, ratio_of_population_for_crossover=0.5, number_of_genes=32, value_range=(0, 1))[source]
Initialise instance of AbstractGenotype.
- Parameters
genotype (
Optional[List[int]]) – genotype used for mutation, crossover and to calculate phenotype_value in binary form.mutation_probability (
float) – probability of a gene mutating.ratio_of_population_for_crossover (
float) – ratio of population used for crossover when updating population.number_of_genes (
int) – number of genes in the genotype.value_range (
Optional[Tuple[int,int]]) – minimum and maximum values of a gene.
Warning
This genotype can only be used for single value optimization e.g. for a ParabolaPhenotype, and would not work for BoothPhenotype.
- classmethod build_random_genotype(number_of_genes=32, value_range=(0, 1), mutation_probability=0.1, ratio_of_population_for_crossover=0.5)[source]
Build random genotype attribute based on requirements.
- Parameters
number_of_genes (
int) – number of genes in the genotype.value_range (
Tuple[int,int]) – minimum and maximum values of a gene.mutation_probability (
Optional[float]) – probability of a gene mutating.ratio_of_population_for_crossover (
Optional[float]) – ratio of population used for crossover when updating population.
- Return type
- Returns
Genotype object with updated genotype attribute.
- crossover(parent_2_genotype)[source]
Performs single point crossover operation for 1 set of parents.
A random integer is generated to split the genotype of the two individuals - this is the gene slice index. Then two child genotypes are generated with the complementary parts of the parent genotypes. If the parent’s genotype length is 1, crossover is impossible so the parent instances are returned.
Example
parent_1.genotype = [1, 2, 3, 4] parent_2.genotype = [A, B, C, D] gene_slice_index = 1
child_1.genotype = [1, B, C, D] child_2.genotype = [A, 2, 3, 4]
- Parameters
parent_2_genotype (
BinaryListGenotype) – Individual which will be used to create an offspring.- Return type
Tuple[BinaryListGenotype,BinaryListGenotype]- Returns
Tuple of AbstractGenotype, representing two children genotypes that are a combination of the parents.
- classmethod from_genotype(base_genotype, new_genotype)[source]
Create a new genotype using the parameters of an existing genotype.
- Return type
- property genotype
Genotype value used for evaluation of phenotype.
Warning
This is the integer form of the self.binary_genotype.
evolutionary_optimization.genotype.implemented_genotypes.float_list_genotype module
- class FloatListGenotype(genotype=None, mutation_probability=0.5, ratio_of_population_for_crossover=0, number_of_genes=1, value_range=(-10000, 10000))[source]
Bases:
object- __init__(genotype=None, mutation_probability=0.5, ratio_of_population_for_crossover=0, number_of_genes=1, value_range=(-10000, 10000))[source]
Initialise instance of AbstractGenotype.
- Parameters
genotype (
Optional[List[float]]) – genotype used for mutation, crossover and to calculate phenotype_value.mutation_probability (
float) – probability of a gene mutating.ratio_of_population_for_crossover (
float) – ratio of population used for crossover when updating population.number_of_genes (
int) – number of genes in the genotype.value_range (
Tuple[int,int]) – minimum and maximum values of a gene.
- classmethod build_random_genotype(number_of_genes=1, value_range=(-10000, 10000), mutation_probability=0.5, ratio_of_population_for_crossover=0)[source]
Build random genotype attribute based on class parameters.
- Parameters
mutation_probability (
float) – probability of a gene mutating.ratio_of_population_for_crossover (
float) – ratio of population used for crossover when updating population.
- Return type
- Returns
AbstractGenotype object with a randomly generated genotype attribute.
- crossover(parent_2_genotype)[source]
Perform crossover between two phenotypes.
Combines a portion of this object’s genotype with that of parent_2 to return 2 new phenotypes based on the combined genotypes. The new genotype length is the same as of the parents.
- Parameters
parent_2_genotype (AbstractGenotype) – a genotype of the same class whose genotype will be mixed with
- Return type
Tuple[AbstractGenotype, AbstractGenotype]- Returns
Two new phenotype instances based on the combined genotypes of the two parents.
- classmethod from_genotype(base_genotype, new_genotype)[source]
Create a new genotype using the parameters of an existing genotype.
- Return type
- property genotype
Genotype value used for evaluation of phenotype.
evolutionary_optimization.genotype.implemented_genotypes.integer_list_genotype module
- class IntegerListGenotype(genotype=None, mutation_probability=0.1, ratio_of_population_for_crossover=0.5, number_of_genes=1, value_range=(-10000, 10000))[source]
Bases:
AbstractGenotype- __init__(genotype=None, mutation_probability=0.1, ratio_of_population_for_crossover=0.5, number_of_genes=1, value_range=(-10000, 10000))[source]
Initialise instance of AbstractGenotype.
- Parameters
genotype (
Optional[List[int]]) – genotype used for mutation, crossover and to calculate phenotype_value.mutation_probability (
float) – probability of a gene mutating.ratio_of_population_for_crossover (
float) – ratio of population used for crossover when updating population.number_of_genes (
int) – number of genes in the genotype.value_range (
Tuple[int,int]) – minimum and maximum values of a gene.
- classmethod build_random_genotype(number_of_genes=1, value_range=(-10000, 10000), mutation_probability=0.5, ratio_of_population_for_crossover=0.5)[source]
Builds random genotype attribute based on requirements.
- Parameters
number_of_genes (
int) – number of genes in the genotype.value_range (
Tuple[int,int]) – minimum and maximum values of a gene.mutation_probability (
Optional[float]) – probability of a gene mutating.ratio_of_population_for_crossover (
Optional[float]) – ratio of population used for crossover when updating population.
- Return type
- Returns
Genotype object with updated genotype attribute.
- crossover(parent_2_genotype)[source]
Performs single point crossover operation for 1 set of parents.
A random integer is generated to split the genotype of the two individuals - this is the gene slice index. Then two child genotypes are generated with the complementary parts of the parent genotypes. If the parent’s genotype length is 1, crossover is impossible so the parent instances are returned.
Example
parent_1.genotype = [1, 2, 3, 4] parent_2.genotype = [A, B, C, D] gene_slice_index = 1
child_1.genotype = [1, B, C, D] child_2.genotype = [A, 2, 3, 4]
- Parameters
parent_2_genotype (
IntegerListGenotype) – Individual which will be used to create an offspring.- Return type
- Returns
Tuple of AbstractGenotype, representing two children genotypes that are a combination of the parents.
- classmethod from_genotype(base_genotype, new_genotype)[source]
Create a new genotype using the parameters of an existing genotype.
- Return type
- property genotype
Genotype value used for evaluation of phenotype.