pyswarms.base package

The pyswarms.base module implements base swarm classes to implement variants of particle swarm optimization.

pyswarms.base module

Base class for single-objective Particle Swarm Optimization implementations.

All methods here are abstract and raise a NotImplementedError when not used. When defining your own swarm implementation, create another class,

>>> class MySwarm(SwarmBase):
>>>     def __init__(self):
>>>        super(MySwarm, self).__init__()

and define all the necessary methods needed.

As a guide, check the global best and local best implementations in this package.

Note

Regarding options, it is highly recommended to include parameters used in position and velocity updates as keyword arguments. For parameters that affect the topology of the swarm, it may be much better to have them as positional arguments.

See also

pyswarms.single.global_best

global-best PSO implementation

pyswarms.single.local_best

local-best PSO implementation

pyswarms.single.general_optimizer

a more general PSO implementation with a custom topology

class pyswarms.base.base_single.SwarmOptimizer(n_particles, dimensions, options, bounds=None, velocity_clamp=None, center=1.0, ftol=-inf, init_pos=None)[source]

Bases: abc.ABC

__init__(n_particles, dimensions, options, bounds=None, velocity_clamp=None, center=1.0, ftol=-inf, init_pos=None)[source]

Initialize the swarm

Creates a Swarm class depending on the values initialized

n_particles

number of particles in the swarm.

Type

int

dimensions

number of dimensions in the space.

Type

int

options

a dictionary containing the parameters for the specific optimization technique

  • c1float

    cognitive parameter

  • c2float

    social parameter

  • wfloat

    inertia parameter

Type

dict with keys {'c1', 'c2', 'w'}

bounds

a tuple of size 2 where the first entry is the minimum bound while the second entry is the maximum bound. Each array must be of shape (dimensions,).

Type

tuple of numpy.ndarray, optional

velocity_clamp

a tuple of size 2 where the first entry is the minimum velocity and the second entry is the maximum velocity. It sets the limits for velocity clamping.

Type

tuple, optional

center

an array of size dimensions

Type

list, optional

ftol

relative error in objective_func(best_pos) acceptable for convergence. Default is -np.inf.

Type

float, optional

_abc_impl = <_abc_data object>
_populate_history(hist)[source]

Populate all history lists

The cost_history, mean_pbest_history, and neighborhood_best is expected to have a shape of (iters,),on the other hand, the pos_history and velocity_history are expected to have a shape of (iters, n_particles, dimensions)

Parameters

hist (collections.namedtuple) – Must be of the same type as self.ToHistory

abstract optimize(objective_func, iters, n_processes=None, **kwargs)[source]

Optimize the swarm for a number of iterations

Performs the optimization to evaluate the objective function objective_func for a number of iterations iter.

Parameters
  • objective_func (function) – objective function to be evaluated

  • iters (int) – number of iterations

  • n_processes (int) – number of processes to use for parallel particle evaluation Default is None with no parallelization

  • kwargs (dict) – arguments for objective function

Raises

NotImplementedError – When this method is not implemented.

reset()[source]

Reset the attributes of the optimizer

All variables/atributes that will be re-initialized when this method is defined here. Note that this method can be called twice: (1) during initialization, and (2) when this is called from an instance.

It is good practice to keep the number of resettable attributes at a minimum. This is to prevent spamming the same object instance with various swarm definitions.

Normally, swarm definitions are as atomic as possible, where each type of swarm is contained in its own instance. Thus, the following attributes are the only ones recommended to be resettable:

  • Swarm position matrix (self.pos)

  • Velocity matrix (self.pos)

  • Best scores and positions (gbest_cost, gbest_pos, etc.)

Otherwise, consider using positional arguments.

Base class for single-objective discrete Particle Swarm Optimization implementations.

All methods here are abstract and raises a NotImplementedError when not used. When defining your own swarm implementation, create another class,

>>> class MySwarm(DiscreteSwarmOptimizer):
>>>     def __init__(self):
>>>        super(MySwarm, self).__init__()

and define all the necessary methods needed.

As a guide, check the discrete PSO implementations in this package.

Note

Regarding options, it is highly recommended to include parameters used in position and velocity updates as keyword arguments. For parameters that affect the topology of the swarm, it may be much better to have them as positional arguments.

See also

pyswarms.discrete.binary

binary PSO implementation

class pyswarms.base.base_discrete.DiscreteSwarmOptimizer(n_particles, dimensions, binary, options, velocity_clamp=None, init_pos=None, ftol=-inf)[source]

Bases: abc.ABC

__init__(n_particles, dimensions, binary, options, velocity_clamp=None, init_pos=None, ftol=-inf)[source]

Initialize the swarm.

Creates a numpy.ndarray of positions depending on the number of particles needed and the number of dimensions. The initial positions of the particles depends on the argument binary, which governs if a binary matrix will be produced.

n_particles

number of particles in the swarm.

Type

int

dimensions

number of dimensions in the space.

Type

int

binary

a trigger to generate a binary matrix for the swarm’s initial positions. When passed with a False value, random integers from 0 to dimensions are generated.

Type

boolean

options

a dictionary containing the parameters for the specific optimization technique

  • c1float

    cognitive parameter

  • c2float

    social parameter

  • wfloat

    inertia parameter

Type

dict with keys {'c1', 'c2', 'w'}

velocity_clamp

a tuple of size 2 where the first entry is the minimum velocity and the second entry is the maximum velocity. It sets the limits for velocity clamping.

Type

tuple, optional

options

a dictionary containing the parameters for a specific optimization technique

Type

dict

_abc_impl = <_abc_data object>
_populate_history(hist)[source]

Populate all history lists

The cost_history, mean_pbest_history, and neighborhood_best is expected to have a shape of (iters,),on the other hand, the pos_history and velocity_history are expected to have a shape of (iters, n_particles, dimensions)

Parameters

hist (collections.namedtuple) – Must be of the same type as self.ToHistory

abstract optimize(objective_func, iters, n_processes=None, **kwargs)[source]

Optimize the swarm for a number of iterations

Performs the optimization to evaluate the objective function objective_func for a number of iterations iter.

Parameters
  • objective_func (callable) – objective function to be evaluated

  • iters (int) – number of iterations

  • n_processes (int) – number of processes to use for parallel particle evaluation Default is None with no parallelization

  • kwargs (dict) – arguments for objective function

Raises

NotImplementedError – When this method is not implemented.

reset()[source]

Reset the attributes of the optimizer

All variables/atributes that will be re-initialized when this method is defined here. Note that this method can be called twice: (1) during initialization, and (2) when this is called from an instance.

It is good practice to keep the number of resettable attributes at a minimum. This is to prevent spamming the same object instance with various swarm definitions.

Normally, swarm definitions are as atomic as possible, where each type of swarm is contained in its own instance. Thus, the following attributes are the only ones recommended to be resettable:

  • Swarm position matrix (self.pos)

  • Velocity matrix (self.pos)

  • Best scores and positions (gbest_cost, gbest_pos, etc.)

Otherwise, consider using positional arguments.