metaheuristic_designer.simple.local_search module

Ready-to-run Local Search wrappers.

local_search_binary(objfunc, mutated_bits=1, samples_per_iteration=100, encoding=None, rng=None, **kwargs)

Local Search for binary-coded vectors.

Return type:

Algorithm

Parameters:

objfuncObjectiveFunc

The objective function to optimize.

mutated_bitsint, optional

Number of bits flipped per mutation (default 1).

samples_per_iterationint, optional

Number of samples evaluated per iteration (default 100).

encodingEncoding, optional

Encoding; defaults to TypeCastEncoding (int → bool).

rngRNGLike, optional

Random seed or generator.

**kwargs

Forwarded to Algorithm.

Parameters:

• objfunc (ObjectiveFunc)

• mutated_bits (int)

• samples_per_iteration (int)

• encoding (Encoding | None)

• rng (int | Generator | None)

local_search_permutation(objfunc, swapped_positions=2, samples_per_iteration=100, encoding=None, rng=None, **kwargs)

Local Search for permutation-coded vectors.

Return type:

Algorithm

Parameters:

objfuncObjectiveFunc

The objective function to optimize.

swapped_positionsint, optional

Number of positions swapped per mutation (default 2).

samples_per_iterationint, optional

Number of samples evaluated per iteration (default 100).

encodingEncoding, optional

Encoding applied to the genotype.

rngRNGLike, optional

Random seed or generator.

**kwargs

Forwarded to Algorithm.

Parameters:

• objfunc (ObjectiveFunc)

• swapped_positions (int)

• samples_per_iteration (int)

• encoding (Encoding | None)

• rng (int | Generator | None)

local_search_discrete(objfunc, resampled_components=1, samples_per_iteration=100, encoding=None, rng=None, **kwargs)

Local Search for integer-coded vectors.

Return type:

Algorithm

Parameters:

objfuncObjectiveFunc

The objective function to optimize.

resampled_componentsint, optional

Number of components resampled per mutation (default 1).

samples_per_iterationint, optional

Number of samples evaluated per iteration (default 100).

encodingEncoding, optional

Encoding applied to the genotype.

rngRNGLike, optional

Random seed or generator.

**kwargs

Forwarded to Algorithm.

Parameters:

• objfunc (ObjectiveFunc)

• resampled_components (int)

• samples_per_iteration (int)

• encoding (Encoding | None)

• rng (int | Generator | None)

local_search_real(objfunc, mutation_strength=0.01, mutated_components=1, samples_per_iteration=100, encoding=None, rng=None, **kwargs)

Local Search for real-coded vectors.

Return type:

Algorithm

Parameters:

objfuncObjectiveFunc

The objective function to optimize.

mutation_strengthfloat, optional

Standard deviation of Gaussian mutation (default 1e-2).

mutated_componentsint, optional

Number of components mutated per individual (default 1).

samples_per_iterationint, optional

Number of samples evaluated per iteration (default 100).

encodingEncoding, optional

Encoding applied to the genotype.

rngRNGLike, optional

Random seed or generator.

**kwargs

Forwarded to Algorithm.

Parameters:

• objfunc (ObjectiveFunc)

• mutation_strength (float)

• mutated_components (int)

• samples_per_iteration (int)

• encoding (Encoding | None)

• rng (int | Generator | None)