easy_vic_build.tools.calibrate_func.sampling

Sampling utilities for calibration and parameter exploration.

This module provides commonly used random and quasi-random sampling routines:

• uniform and Gaussian sampling,

• Latin Hypercube Sampling (LHS),

• Sobol and Halton low-discrepancy sequences,

• discrete sampling with optional constraints.

Functions

mixed_sampling(n_samples)	Reserved interface for mixed sampling strategies.
sampling_Halton(n_samples, bounds)	Generate random samples using the Halton sequence method within the specified bounds.
sampling_LHS_1(n_samples, n_dimensions, bounds)	Generate random samples using Latin Hypercube Sampling (LHS) method, variant 1, within the specified bounds.
sampling_LHS_2(n_samples, bounds[, seed])	Generate random samples using Latin Hypercube Sampling (LHS) method, variant 2, within the specified bounds.
sampling_Sobol(n_samples, bounds)	Generate random samples using the Sobol sequence method within the specified bounds.
sampling_discrete(discrete_values, n_samples)	Generate random samples from a set of discrete values, optionally with weights.
sampling_discrete_constrained(...)	Generate random samples from discrete values, with the constraint that the sum of the samples equals target_sum.
sampling_gaussian(n_samples, mean, std)	Generate random samples from a Gaussian (normal) distribution with specified mean and standard deviation.
sampling_gaussian_clip(n_samples, mean, std)	Generate random samples from a Gaussian distribution with specified mean and standard deviation, and clip the samples to the specified range [low, up].
sampling_uniform(n_samples, bounds)	Generate random samples from a uniform distribution within the specified bounds.
sampling_uniform_int(n_samples, bounds)	Generate random integer samples from a uniform distribution within the specified bounds.

easy_vic_build.tools.calibrate_func.sampling.sampling_uniform(n_samples, bounds)

Generate random samples from a uniform distribution within the specified bounds.

Parameters:

• n_samples (int) – The number of samples to generate.

• bounds (tuple of float) – The lower and upper bounds for the uniform distribution.

Returns:

samples – The generated random samples.

Return type:

list of float

easy_vic_build.tools.calibrate_func.sampling.sampling_uniform_int(n_samples, bounds)

Generate random integer samples from a uniform distribution within the specified bounds.

Parameters:

• n_samples (int) – The number of integer samples to generate.

• bounds (tuple of int) – The lower and upper bounds for the uniform distribution.

Returns:

samples – The generated random integer samples.

Return type:

list of int

easy_vic_build.tools.calibrate_func.sampling.sampling_gaussian(n_samples, mean, std)

Generate random samples from a Gaussian (normal) distribution with specified mean and standard deviation.

Parameters:

• n_samples (int) – The number of samples to generate.

• mean (float) – The mean of the Gaussian distribution.

• std (float) – The standard deviation of the Gaussian distribution.

Returns:

samples – The generated random samples.

Return type:

list of float

easy_vic_build.tools.calibrate_func.sampling.sampling_gaussian_clip(n_samples, mean, std, low=None, up=None)

Generate random samples from a Gaussian distribution with specified mean and standard deviation, and clip the samples to the specified range [low, up].

Parameters:

• n_samples (int) – The number of samples to generate.

• mean (float) – The mean of the Gaussian distribution.

• std (float) – The standard deviation of the Gaussian distribution.

• low (float, optional) – The lower bound for clipping the samples.

• up (float, optional) – The upper bound for clipping the samples.

Returns:

samples – The generated random samples, clipped to the specified range.

Return type:

numpy.ndarray

easy_vic_build.tools.calibrate_func.sampling.sampling_LHS_1(n_samples, n_dimensions, bounds)

Generate random samples using Latin Hypercube Sampling (LHS) method, variant 1, within the specified bounds.

Parameters:

• n_samples (int) – The number of samples to generate.

• n_dimensions (int) – The number of dimensions for each sample.

• bounds (list of tuple) – A list of tuples specifying the lower and upper bounds for each dimension.

Returns:

samples – The generated Latin Hypercube samples.

Return type:

numpy.ndarray

easy_vic_build.tools.calibrate_func.sampling.sampling_LHS_2(n_samples, bounds, seed=None)

Generate random samples using Latin Hypercube Sampling (LHS) method, variant 2, within the specified bounds.

Parameters:

• n_samples (int) – The number of samples to generate.

• bounds (list of tuple) – A list of tuples specifying the lower and upper bounds for each dimension.

• seed (int, optional) – Seed used by scipy.stats.qmc.LatinHypercube.

Returns:

The generated Latin Hypercube samples, scaled to the specified bounds.

Return type:

numpy.ndarray

Raises:

ValueError – If any bound does not satisfy min < max.

easy_vic_build.tools.calibrate_func.sampling.sampling_Sobol(n_samples, bounds)

Generate random samples using the Sobol sequence method within the specified bounds.

Parameters:

• n_samples (int) – The number of samples to generate.

• bounds (list of tuple) – A list of tuples specifying the lower and upper bounds for each dimension.

Returns:

The generated Sobol samples, scaled to the specified bounds.

Return type:

numpy.ndarray

easy_vic_build.tools.calibrate_func.sampling.sampling_Halton(n_samples, bounds)

Generate random samples using the Halton sequence method within the specified bounds.

Parameters:

• n_samples (int) – The number of samples to generate.

• bounds (list of tuple) – A list of tuples specifying the lower and upper bounds for each dimension.

Returns:

The generated Halton samples, scaled to the specified bounds.

Return type:

numpy.ndarray

easy_vic_build.tools.calibrate_func.sampling.sampling_discrete(discrete_values, n_samples, weights=None)

Generate random samples from a set of discrete values, optionally with weights.

Parameters:

• discrete_values (array-like) – A list or array of discrete values to sample from.

• n_samples (int) – The number of samples to generate.

• weights (array-like, optional) – The weights associated with the discrete values. If None, the values are assumed to be equally likely.

Returns:

samples – The generated discrete samples.

Return type:

numpy.ndarray

easy_vic_build.tools.calibrate_func.sampling.sampling_discrete_constrained(discrete_values, target_sum, n_samples)

Generate random samples from discrete values, with the constraint that the sum of the samples equals target_sum.

Parameters:

• discrete_values (array-like) – A list or array of discrete values to sample from.

• target_sum (int) – The target sum for the generated samples.

• n_samples (int) – The number of samples to generate.

Returns:

samples – The generated discrete samples with the constraint on their sum.

Return type:

numpy.ndarray

easy_vic_build.tools.calibrate_func.sampling.mixed_sampling(n_samples)

Reserved interface for mixed sampling strategies.

Parameters:

n_samples (int) – Requested number of samples.

Returns:

This function currently has no implementation.

Return type:

None

Notes

This function is intentionally left as a placeholder for future extension.