cosapp.drivers.montecarlo

Classes

MonteCarlo(name[, owner])	This driver execute a MonteCarlo simulation on its system.
RandomVariable(system, varname[, distribution])	Class representing a random variable in a system of interest

class cosapp.drivers.montecarlo.MonteCarlo(name: str, owner: System | None = None, **options)

Bases: AbstractSetOfCases

This driver execute a MonteCarlo simulation on its system.

A: numpy.ndarray

X0: numpy.ndarray

Y0: numpy.ndarray

add_random_variable(name, distribution: Distribution | None = None) → None

add_random_variable(specs: dict[str, Distribution], /) → None

add_random_variable(varname: str, distribution: Distribution | None = None) → None

add_random_variable(varnames: Iterable[str], distribution: Distribution | None = None) → None

Add variable to be randomly perturbated, following an optional distribution.

The perturbation distribution is defined by the variable distribution details.

Parameters:

• varname (Union[str, Iterable[str], dict[str, Distribution]]) – Name of the variable to be perturbed, or iterable thereof. If the argument is a dict, it is interpreted as a {name: distribution} specification

• distribution (cosapp.utils.distributions.Distribution | None) – Distribution of the variable(s) to be perturbed. If None (default), distributions are deduced from variable names. This argument is only meaningful when a variable name or a list thereof is provided.

Examples

>>> from cosapp.utils.distributions import Normal, Uniform
>>>
>>> montecarlo = system.add_driver(MonteCarlo("montecarlo"))
>>>
>>> montecarlo.add_random_variable("port.v")
>>> montecarlo.add_random_variable("port.v", Normal(...))
>>>
>>> montecarlo.add_random_variable(["x", "port.v"])
>>> montecarlo.add_random_variable(["x", "port.v"], Normal(...))
>>>
>>> montecarlo.add_random_variable({
>>>     "x": Normal(...),
>>>     "port.v": Uniform(...),
>>> })

add_response(name: str | Iterable[str]) → None

Add a variable for which the statistical response will be calculated.

Parameters:

name (Union[str, Iterable[str]]) – List of variable names to add

clear_random_variables() → None

Purge all random variables

draws

linear

perturbations

random_variable_data() → dict[str, Distribution]

Builds and returns the dictionary associating distributions to random variable names

property random_variable_names: Iterator[str]

Generator yielding all random variable names

property random_variables: Iterator[RandomVariable]

Generator yielding all random variable names

reference_case_solution: dict[str, float]

response_varnames: list[str]

setup_run() → None

Actions performed prior to the compute call.

class cosapp.drivers.montecarlo.RandomVariable(system: System, varname: str, distribution: Distribution | None = None)

Bases: object

Class representing a random variable in a system of interest

add_noise(quantile: float | None = None) → float

connector: SystemConnector | None

distribution: Distribution | None

draw(quantile: float | None = None) → float

name

set_perturbation(value) → None

property value