queens.parameters package

Parameters.

Modules grouping deterministic and continuous random variables or random fields.

Subpackages

• queens.parameters.random_fields package
  • Submodules
  • queens.parameters.random_fields.fourier module
    • DimensionMethods1D
      • DimensionMethods1D.calculate_basis()
      • DimensionMethods1D.calculate_covariance()
      • DimensionMethods1D.get_dim()
    • DimensionMethods2D
      • DimensionMethods2D.calculate_basis()
      • DimensionMethods2D.calculate_covariance()
      • DimensionMethods2D.get_dim()
    • DimensionMethods3D
      • DimensionMethods3D.calculate_basis()
      • DimensionMethods3D.calculate_covariance()
      • DimensionMethods3D.get_dim()
    • Fourier
      • Fourier.mean
      • Fourier.std
      • Fourier.corr_length
      • Fourier.variability
      • Fourier.trunc_threshold
      • Fourier.basis_dimension
      • Fourier.latent_index
      • Fourier.covariance_index
      • Fourier.covariance
      • Fourier.basis
      • Fourier.coordinates
      • Fourier.field_dimension
      • Fourier.number_expansion_terms
      • Fourier.dimension
      • Fourier.convex_hull_size
      • Fourier.check_convergence()
      • Fourier.draw()
      • Fourier.expanded_representation()
      • Fourier.grad_logpdf()
      • Fourier.latent_gradient()
      • Fourier.logpdf()
  • queens.parameters.random_fields.karhunen_loeve module
    • KarhunenLoeve
      • KarhunenLoeve.nugget_variance
      • KarhunenLoeve.explained_variance
      • KarhunenLoeve.std
      • KarhunenLoeve.corr_length
      • KarhunenLoeve.cut_off
      • KarhunenLoeve.mean
      • KarhunenLoeve.cov_matrix
      • KarhunenLoeve.eigenbasis
      • KarhunenLoeve.eigenvalues
      • KarhunenLoeve.eigenvectors
      • KarhunenLoeve.dimension
      • KarhunenLoeve.calculate_covariance_matrix()
      • KarhunenLoeve.draw()
      • KarhunenLoeve.eigendecomp_cov_matrix()
      • KarhunenLoeve.expanded_representation()
      • KarhunenLoeve.grad_logpdf()
      • KarhunenLoeve.latent_gradient()
      • KarhunenLoeve.logpdf()
  • queens.parameters.random_fields.piece_wise module
    • PieceWise
      • PieceWise.distribution
      • PieceWise.latent_1d_distribution
      • PieceWise.create_latent_distribution()
      • PieceWise.draw()
      • PieceWise.expanded_representation()
      • PieceWise.grad_logpdf()
      • PieceWise.latent_gradient()
      • PieceWise.logpdf()

Submodules

queens.parameters.parameters module

Parameters.

class Parameters(**parameters)

Bases: object

Parameters class.

dict

Random variables and random fields stored in a dict.

Type:

dict

parameters_keys

List of keys for all parameter members.

Type:

list

num_parameters

Number of (truncated) parameters.

Type:

int

random_field_flag

Specifies if random fields are used.

Type:

bool

names

Parameter names.

Type:

list

draw_samples(num_samples)

Draw samples from all parameters.

Parameters:

num_samples (int) – The number of samples to draw for each parameter.

Returns:

samples (np.ndarray) – Drawn samples

expand_random_field_realization(truncated_sample)

Expand truncated representation of random fields.

Parameters:

truncated_sample (np.ndarray) – Truncated representation of sample

Returns:

sample_expanded (np.ndarray) – Expanded representation of sample

grad_joint_logpdf(samples)

Evaluate the gradient of the joint logpdf w.r.t. the samples.

Parameters:

samples (np.ndarray) – Samples for which to evaluate the gradient of the joint logpdf. Each row represents a sample and each column corresponds to a parameter dimension.

Returns:

grad_logpdf (np.ndarray) – Gradient of the joint logpdf w.r.t. the samples

inverse_cdf_transform(samples)

Transform samples to unit interval.

Parameters:

samples (np.ndarray) – Samples that should be transformed.

Returns:

transformed_samples (np.ndarray) – Transformed samples

joint_logpdf(samples)

Evaluate the logpdf summed over all parameters.

Parameters:

samples (np.ndarray) – Samples for which to evaluate the joint logpdf. Each row represents a sample and each column corresponds to a parameter dimension.

Returns:

logpdf (np.ndarray) – logpdf summed over all parameters

latent_grad(upstream_gradient)

Gradient of the rvs and rfs w.r.t. latent variables.

Parameters:

upstream_gradient (np.array) – Upstream gradient

Returns:

gradient (np.ndarray) – Gradient of the joint rvs/rfs w.r.t. the samples

sample_as_dict(sample)

Return sample as a dict.

Parameters:

sample (np.ndarray) – A single sample

Returns:

• sample_dict (dict) – Dictionary containing sample members and the corresponding parameter

• keys

to_distribution_list()

Return the distributions of the parameters as list.

Returns:

distribution_list (list) – List of distributions of parameters

to_list()

Return parameters as list.

Returns:

parameter_list (list) – List of parameters

from_config_create_parameters(parameters_options, pre_processor=None)

Create a QUEENS parameter object from config.

Parameters:

• parameters_options (dict) – Parameters description

• pre_processor (obj, optional) – Pre-processor object to read coordinates of random field discretization