gemseo / algos

parameter_space module¶

Parameter space including both deterministic and uncertain parameters¶

Overview¶

The ParameterSpace class describes a set of parameters of interest which can be either deterministic or uncertain. This class inherits from DesignSpace.

Capabilities¶

The DesignSpace.add_variable() aims to add deterministic variables from:

a variable name,
a variable size (default: 1),
a variable type (default: float),
a lower bound (default: - infinity),
an upper bound (default: + infinity),
a current value (default: None).

The add_random_variable() aims to add uncertain variables (a.k.a. random variables) from:

a variable name,
a distribution name (see get_available_distributions()),
a variable size,
distribution parameters (parameters set as a tuple of positional arguments for OTDistribution or a dictionary of keyword arguments for SPDistribution, or keyword arguments for standard probability distribution such as OTNormalDistribution and OTNormalDistribution).

The ParameterSpace also provides the following methods:

get_cdf(): evaluate the cumulative density function for the different variables and their different
get_composed_distribution(): returns the probability distribution of an uncertain variable,
get_marginal_distributions(): returns the marginal probability distributions of an uncertain variable,
get_range() returns the numerical range of the different uncertain parameters,
ParameterSpace.get_sample(): returns several sample of the uncertain variables,
get_support(): returns the mathematical support of the different uncertain variables,
is_uncertain(): checks if a parameter is uncertain,
is_deterministic(): checks if a parameter is deterministic.

class gemseo.algos.parameter_space.ParameterSpace(print_decimals=2, shorten=True, copula='independent_copula')[source]¶

Bases: gemseo.algos.design_space.DesignSpace

Parameter space.

Constructor

Parameters

print_decimals (int) – number of decimals to print. Default: 2.
shorten (bool) – if True, simplify the expressions of variable transformations. Default: True.
copula (str) – copula name. Default: ComposedDistribution.INDEPENDENT_COPULA.

BLANK = ''¶

INITIAL_DISTRIBUTION = 'Initial distribution'¶

MEAN = 'Mean'¶

PARAMETER_SPACE = 'Parameter space'¶

RANGE = 'Range'¶

STANDARD_DEVIATION = 'Standard deviation'¶

SUPPORT = 'Support'¶

TRANSFORMATION = 'Transformation'¶

add_random_variable(name, distribution, size=1, **parameters)[source]¶

Add a random variable from a distribution

Parameters

name (str) – name of the random variable.
distribution (str) – distribution name.
size (int) – variable size.
parameters – parameters of the distribution.

get_cdf(value, inverse=False)[source]¶

Get the inverse Cumulative Density Function: values of the different marginals.

Parameters: value (dict(array)) – values
Returns: (inverse) CDF values
Return type: dict(array)

get_composed_distribution(variable)[source]¶

Get the composed distribution of a random variable.

Parameters: variable (str) – variable name.

get_marginal_distributions(variable)[source]¶

Get the marginal distributions of a random variable.

Parameters: variable (str) – variable name.

get_range(variable)[source]¶

Get the numerical range of a random variable.

Parameters: variable (str) – variable name.

get_sample(n_samples=1, as_dict=False)[source]¶

Get sample.

Parameters

n_samples (int) – number of samples.
as_dict (bool) – return a dictionary.

Returns

samples

Return type

list(array) or list(dict)

get_support(variable)[source]¶

Get the mathematical support of a random variable.

Parameters: variable (str) – variable name.

is_deterministic(variable)[source]¶

Check if a variable is deterministic.

Parameters: variable (str) – variable name.

is_uncertain(variable)[source]¶

Check if a variable is uncertain.

Parameters: variable (str) – variable name.

normalize_vect(x_vect, minus_lb=True, use_dist=False)[source]¶

Normalizes a vector of the design space. Unbounded variables are not normalized.

Parameters

x_vect (array) – design variables.
minus_lb (bool) – if True, remove lower bounds at normalization.
no_check (bool) – if True, don’t check that values are in [0,1].
use_dist (bool) – if True, rescale wrt the stats law.

Returns

normalized vector

Return type

array

remove_variable(name)[source]¶

Remove a variable from the probability space.

Parameters: name (str) – variable name.

set_dependence(variables, copula, **options)[source]¶

Set dependence relation between random variables.

Parameters

variables (list(str)) – list of variables names.
copula (str) – copula name.
options – copula options.

unnormalize_vect(x_vect, minus_lb=True, no_check=False, use_dist=True)[source]¶

Inverse transformation from a unit design vector. Unnormalizes a normalized vector of the design space.

Parameters

x_vect (array) – design variables.
minus_lb (bool) – if True, remove lower bounds at normalization.
no_check (bool) – if True, don’t check that values are in [0,1].
use_dist (bool) – if True, rescale wrt the stats law.

Returns

normalized vector

Return type

array