pyccea.initialization package
- class pyccea.initialization.RandomBinaryInitialization(data: DataLoader, subcomp_sizes: list, subpop_sizes: list, collaborator, fitness_function)[source]
Bases:
SubpopulationInitializationRandomly initialize subpopulations with binary representation.
Methods
build_subpopulations()Initialize individuals from all subpopulations.
evaluate_individuals()Evaluate all individuals from all subpopulations.
- class pyccea.initialization.RandomContinuousInitialization(data: DataLoader, subcomp_sizes: list, subpop_sizes: list, collaborator, fitness_function, bounds: tuple = (0, 1))[source]
Bases:
SubpopulationInitializationRandomly initialize subpopulations with continuous representation.
For certain Evolutionary Algorithms, like Differential Evolution, which operate on continuous solutions, an appropriate initialization is required based on this representation.
Methods
build_subpopulations()Initialize individuals from all subpopulations.
evaluate_individuals()Evaluate all individuals from all subpopulations.
- class pyccea.initialization.SubpopulationInitialization(data: DataLoader, subcomp_sizes: list, subpop_sizes: list, collaborator, fitness_function)[source]
Bases:
ABCAn abstract class for subpopulation initialization.
- Attributes:
- subpopslist
Individuals from all subpopulations.
- fitnesslist
Evaluation of all context vectors from all subpopulations.
- context_vectors: list
Complete problem solutions that were randomly initialized.
Methods
Initialize individuals from all subpopulations.
Evaluate all individuals from all subpopulations.
Submodules
pyccea.initialization.binary module
- class pyccea.initialization.binary.RandomBinaryInitialization(data: DataLoader, subcomp_sizes: list, subpop_sizes: list, collaborator, fitness_function)[source]
Bases:
SubpopulationInitializationRandomly initialize subpopulations with binary representation.
Methods
build_subpopulations()Initialize individuals from all subpopulations.
evaluate_individuals()Evaluate all individuals from all subpopulations.
pyccea.initialization.build module
- class pyccea.initialization.build.SubpopulationInitialization(data: DataLoader, subcomp_sizes: list, subpop_sizes: list, collaborator, fitness_function)[source]
Bases:
ABCAn abstract class for subpopulation initialization.
- Attributes:
- subpopslist
Individuals from all subpopulations.
- fitnesslist
Evaluation of all context vectors from all subpopulations.
- context_vectors: list
Complete problem solutions that were randomly initialized.
Methods
Initialize individuals from all subpopulations.
Evaluate all individuals from all subpopulations.
pyccea.initialization.continuous module
- class pyccea.initialization.continuous.RandomContinuousInitialization(data: DataLoader, subcomp_sizes: list, subpop_sizes: list, collaborator, fitness_function, bounds: tuple = (0, 1))[source]
Bases:
SubpopulationInitializationRandomly initialize subpopulations with continuous representation.
For certain Evolutionary Algorithms, like Differential Evolution, which operate on continuous solutions, an appropriate initialization is required based on this representation.
Methods
build_subpopulations()Initialize individuals from all subpopulations.
evaluate_individuals()Evaluate all individuals from all subpopulations.