hyponic.optimizers.swarm_based package

hyponic.optimizers.swarm_based.ABC module

class hyponic.optimizers.swarm_based.ABC.ABC(epoch: int = 10, population_size: int = 10, minmax: str = None, limits=25, verbose: bool = False, mode: str = 'single', n_workers: int = 4, early_stopping: int | None = None, **kwargs)

Bases: BaseOptimizer

Artificial Bee Colony (ABC) algorithm

Hyperparameters:

• limits(int), default=25: the number of trials before abandoning food source

Example

>>> from hyponic.optimizers.swarm_based.ABC import ABC
>>> import numpy as np
>>>
>>> def sphere(x):
>>>     return np.sum(x ** 2)
>>>
>>> problem_dict = {
>>>     'fit_func': sphere,
>>>     'lb': [-5.12, -5, -14, -6, -0.9],
>>>     'ub': [5.12, 5, 14, 6, 0.9],
>>>     'minmax': 'min'
>>> }
>>>
>>> abc = ABC(epoch=40, population_size=100, verbose=True, early_stopping=4)
>>> abc.solve(problem_dict)
>>> print(abc.get_best_score())
>>> print(abc.get_best_solution())

evolve(epoch)

Evolve the population for one epoch

Parameters:

current_epoch – current epoch number

get_best_score()

Get the best score of the current population

Returns:

best score of the fitness function

get_best_solution()

Get the best solution of the current population

Returns:

coordinates of the best solution

get_current_best_score()

Get the best score of the current population

Returns:

current best score of the fitness function

get_current_best_solution()

Get the best solution of the current population

Returns:

current coordinates of the best solution

initialize(problem_dict)

Initialize the optimizer with the problem dictionary

Parameters:

problem_dict – dictionary containing the problem definition

hyponic.optimizers.swarm_based.ACO module

class hyponic.optimizers.swarm_based.ACO.ACO(epoch: int = 10, population_size: int = 10, minmax: str = None, alpha: float = 1, beta: float = 1, rho: float = 0.5, q: float = 1, verbose: bool = False, mode: str = 'single', n_workers: int = 4, early_stopping: int | None = None, **kwargs)

Bases: BaseOptimizer

Ant Colony Optimization (ACO)

Hyperparameters:

• alpha(float), default=1: the importance of pheromone

• beta(float), default=1: the importance of heuristic information

• rho(float), default=0.5: the pheromone evaporation rate

• q(float), default=1: the pheromone intensity

Example

>>> from hyponic.optimizers.swarm_based.ACO import ACO
>>> import numpy as np
>>>
>>> def sphere(x):
>>>     return np.sum(x ** 2)
>>>
>>> problem_dict = {
>>>     'fit_func': sphere,
>>>     'lb': [-5.12, -5, -14, -6, -0.9],
>>>     'ub': [5.12, 5, 14, 6, 0.9],
>>>     'minmax': 'min'
>>> }
>>>
>>> alpha = 1
>>> beta = 1
>>> rho = 0.5
>>> q = 1
>>>
>>> aco = ACO(epoch=40, population_size=100, verbose=True, early_stopping=4, alpha=alpha, beta=beta, rho=rho, q=q)
>>> aco.solve(problem_dict)
>>> print(aco.get_best_score())
>>> print(aco.get_best_solution())

evolve(epoch)

Evolve the population for one epoch

Parameters:

current_epoch – current epoch number

get_best_score()

Get the best score of the current population

Returns:

best score of the fitness function

get_best_solution()

Get the best solution of the current population

Returns:

coordinates of the best solution

get_current_best_score()

Get the best score of the current population

Returns:

current best score of the fitness function

get_current_best_solution()

Get the best solution of the current population

Returns:

current coordinates of the best solution

initialize(problem_dict)

Initialize the optimizer with the problem dictionary

Parameters:

problem_dict – dictionary containing the problem definition

hyponic.optimizers.swarm_based.CS module

class hyponic.optimizers.swarm_based.CS.CS(epoch: int = 10, population_size: int = 10, minmax: str = None, verbose: bool = False, pa: float = 0.25, alpha: float = 0.5, k: float = 1, mode: str = 'single', n_workers: int = 4, early_stopping: int | None = None, **kwargs)

Bases: BaseOptimizer

Cuckoo Search (CS) algorithm

Hyperparameters:

• pa(float), default=0.25: probability of cuckoo’s egg to be discovered

• alpha(float), default=0.5: step size

• k(float), default=1: Levy multiplication coefficient

Example

>>> from hyponic.optimizers.swarm_based.CS import CS
>>> import numpy as np
>>>
>>> def sphere(x):
>>>     return np.sum(x ** 2)
>>>
>>> problem_dict = {
>>>     'fit_func': sphere,
>>>     'lb': [-5.12, -5, -14, -6, -0.9],
>>>     'ub': [5.12, 5, 14, 6, 0.9],
>>>     'minmax': 'min'
>>> }
>>>
>>> cs = CS(epoch=40, population_size=100, verbose=True, early_stopping=4)
>>> cs.solve(problem_dict)
>>> print(cs.get_best_score())
>>> print(cs.get_best_solution())

evolve(current_epoch)

Evolve the population for one epoch

Parameters:

current_epoch – current epoch number

get_best_score()

Get the best score of the current population

Returns:

best score of the fitness function

get_best_solution()

Get the best solution of the current population

Returns:

coordinates of the best solution

get_current_best_score()

Get the best score of the current population

Returns:

current best score of the fitness function

get_current_best_solution()

Get the best solution of the current population

Returns:

current coordinates of the best solution

initialize(problem_dict)

Initialize the optimizer with the problem dictionary

Parameters:

problem_dict – dictionary containing the problem definition

hyponic.optimizers.swarm_based.GWO module

class hyponic.optimizers.swarm_based.GWO.GWO(epoch: int = 10, population_size: int = 10, minmax: str = None, verbose: bool = False, mode: str = 'single', n_workers: int = 4, early_stopping: int | None = None, **kwargs)

Bases: BaseOptimizer

Grey Wolf Optimization (GWO) algorithm

Example

>>> from hyponic.optimizers.swarm_based.GWO import GWO
>>> import numpy as np
>>>
>>> def sphere(x):
>>>     return np.sum(x ** 2)
>>>
>>> problem_dict = {
>>>     'fit_func': sphere,
>>>     'lb': [-5.12, -5, -14, -6, -0.9],
>>>     'ub': [5.12, 5, 14, 6, 0.9],
>>>     'minmax': 'min'
>>> }
>>>
>>> gwo = GWO(epoch=40, population_size=100, verbose=True, early_stopping=4)
>>> gwo.solve(problem_dict)
>>> print(gwo.get_best_score())
>>> print(gwo.get_best_solution())

evolve(current_epoch)

Evolve the population for one epoch

Parameters:

current_epoch – current epoch number

get_best_score()

Get the best score of the current population

Returns:

best score of the fitness function

get_best_solution()

Get the best solution of the current population

Returns:

coordinates of the best solution

get_current_best_score()

Get the best score of the current population

Returns:

current best score of the fitness function

get_current_best_solution()

Get the best solution of the current population

Returns:

current coordinates of the best solution

initialize(problem_dict)

Initialize the optimizer with the problem dictionary

Parameters:

problem_dict – dictionary containing the problem definition

hyponic.optimizers.swarm_based.PSO module

class hyponic.optimizers.swarm_based.PSO.IWPSO(epoch: int = 10, population_size: int = 10, minmax: str = None, a1: float = 0.5, a2: float = 0.5, w: float = 0.8, verbose: bool = False, mode: str = 'single', n_workers: int = 4, early_stopping: int | None = None, **kwargs)

Bases: PSO

Inertia Weight Particle Swarm Optimization

Hyperparameters:

• a1(float), default=0.5: acceleration parameter

• a2(float), default=0.5: acceleration parameter

• w(float), default=0.5: inertia weight

Example

>>> from hyponic.optimizers.swarm_based.PSO import IWPSO
>>> import numpy as np
>>>
>>> def sphere(x):
>>>     return np.sum(x ** 2)
>>>
>>> problem_dict = {
>>>     'fit_func': sphere,
>>>     'lb': [-5.12, -5, -14, -6, -0.9],
>>>     'ub': [5.12, 5, 14, 6, 0.9],
>>>     'minmax': 'min'
>>> }
>>>
>>> a1, a2 = 0.8, 0.4
>>> w = 0.3
>>> iwpso = IWPSO(epoch=40, population_size=100, verbose=True, early_stopping=4, a1=a1, a2=a2, w=w)
>>> iwpso.solve(problem_dict)
>>> print(iwpso.get_best_score())
>>> print(iwpso.get_best_solution())

class hyponic.optimizers.swarm_based.PSO.PSO(epoch: int = 10, population_size: int = 10, minmax: str = None, a1: float = 0.5, a2: float = 0.5, verbose: bool = False, mode: str = 'single', n_workers: int = 4, early_stopping: int | None = None, **kwargs)

Bases: BaseOptimizer

Particle Swarm Optimization (PSO) algorithm

Hyperparameters:

• a1(float), default=0.5: acceleration parameter

• a2(float), default=0.5: acceleration parameter

Example

>>> from hyponic.optimizers.swarm_based.PSO import PSO
>>> import numpy as np
>>>
>>> def sphere(x):
>>>     return np.sum(x ** 2)
>>>
>>> problem_dict = {
>>>     'fit_func': sphere,
>>>     'lb': [-5.12, -5, -14, -6, -0.9],
>>>     'ub': [5.12, 5, 14, 6, 0.9],
>>>     'minmax': 'min'
>>> }
>>>
>>> a1, a2 = 0.8, 0.4
>>> pso = PSO(epoch=40, population_size=100, verbose=True, early_stopping=4, a1=a1, a2=a2)
>>> pso.solve(problem_dict)
>>> print(pso.get_best_score())
>>> print(pso.get_best_solution())

evolve(epoch)

Evolve the population for one epoch

Parameters:

current_epoch – current epoch number

get_best_score()

Get the best score of the current population

Returns:

best score of the fitness function

get_best_solution()

Get the best solution of the current population

Returns:

coordinates of the best solution

get_current_best_score()

Get the best score of the current population

Returns:

current best score of the fitness function

get_current_best_solution()

Get the best solution of the current population

Returns:

current coordinates of the best solution

initialize(problem_dict)

Initialize the optimizer with the problem dictionary

Parameters:

problem_dict – dictionary containing the problem definition