src.functions_sampler module

class src.functions_sampler.LatinHypercube(bounds: list[tuple[float, float]], precision: float | None = None, seed: int | None = None, constraint_fns: list[Callable[[ndarray], bool]] | None = None, logger: Logger | None = None, silent: bool = True, log_dev: bool = False)

Bases: object

Optimized Latin Hypercube Sampling with constraints and precision handling.

bounds

List of (min, max) tuples for each parameter.

Type:

List[Tuple[float, float]]

precision

Decimal precision (e.g., 0.1 for 1-decimal-place precision).

Type:

float

seed

Random seed for reproducibility.

Type:

Optional[int]

constraint_fns

List of constraint functions. Each function takes a sample as input and returns True if the sample satisfies the constraint.

Type:

Optional[List[Callable[[np.ndarray], bool]]]

sampler

Instance of the Latin Hypercube sampler from scipy.stats.qmc.

Type:

qmc.LatinHypercube

generate_samples(n_samples

int) -> np.ndarray: Generate Latin Hypercube Samples within specified bounds, ensuring the samples have the desired precision and satisfy constraints.

extend_samples(existing_samples

np.ndarray, n_samples: int, only_new: bool = False) -> np.ndarray: Extend the existing samples with new, unique samples.

_apply_constraints(samples: ndarray) → ndarray

_scale_and_round(raw_samples: ndarray) → ndarray

extend_samples(existing_samples: ndarray, n_samples: int, only_new: bool = False) → ndarray

Extend the existing samples with new, unique samples.

Parameters:

• existing_samples (np.ndarray) – Existing samples to extend.

• n_samples (int) – Number of new samples to generate.

• only_new (bool) – If True, return only the new samples.

Returns:

Array of unique, constrained, and precision-rounded samples.

Return type:

np.ndarray

generate_samples(n_samples: int, samples: ndarray | None = None, stop_on_fail: bool = False) → ndarray

Generate constrained, unique samples using Latin Hypercube Sampling.

Parameters:

• n_samples (int) – Number of samples to generate.

• samples (np.array) – Existing samples to extend.

• stop_on_fail (bool) – If True, raise an error if the desired number of samples cannot be generated.

Returns:

Array of unique, constrained, and precision-rounded samples.

Return type:

np.ndarray

class src.functions_sampler.SOBOL_sampler(n_dim: int, bounds: list[tuple[float, float]], precision: float | None = None, seed: int | None = None)

Bases: object

Sobol Sampler for generating low-discrepancy samples within specified bounds.

n_dim

Number of dimensions for the samples.

Type:

int

bounds

List of (min, max) tuples for each parameter.

Type:

List[Tuple[float, float]]

precision

Decimal precision (e.g., 0.1 for 1-decimal-place precision).

Type:

float

seed

Random seed for reproducibility.

Type:

Optional[int]

generate_samples(n_samples

int) -> np.ndarray: Generate low-discrepancy samples within specified bounds and precision.

generate_samples(n_samples: int) → ndarray

class src.functions_sampler.Uniform_sampler(n_dim: int, bounds: list[tuple[float, float]], precision: float | None = None, seed: int | None = None)

Bases: object

Uniform Sampler for generating random samples within specified bounds.

n_dim

Number of dimensions for the samples.

Type:

int

bounds

List of (min, max) tuples for each parameter.

Type:

List[Tuple[float, float]]

precision

Decimal precision (e.g., 0.1 for 1-decimal-place precision).

Type:

float

seed

Random seed for reproducibility.

Type:

Optional[int]

generate_samples(n_samples

int) -> np.ndarray: Generate uniformly distributed samples within specified bounds and precision.

generate_samples(n_samples: int) → ndarray