PythonFieldModel¶

class persalys.PythonFieldModel(*args)¶

Create a physical model defined by Python code and a mesh.

Parameters

namestr: Name
inputssequence of Input: Input variables (optional)
outputssequence of Output: Output variables (optional)
codestr: Code string (optional)

Notes

The code string must contain a function named ‘_exec’ that takes float arguments and return sequences of floats. The mesh is set as a global variable of type openturns.Mesh named ‘mesh’. The code can use this variable and retrieve the mesh vertices as follows: $aVar = getMesh().getVertices()$

Examples

>>> import openturns as ot
>>> import persalys

Create the mesh model:

>>> meshModel = persalys.GridMeshModel([persalys.Variable('x', 'aDescription')], ot.Interval(0., 12.), [10])

Create the model:

>>> z0 = persalys.Input('z0', 100)
>>> v0 = persalys.Input('v0', 55)
>>> m = persalys.Input('m', 80)
>>> c = persalys.Input('c', 15)
>>> z = persalys.Output('z')
>>> code = 'from math import exp\n\ndef _exec(z0,v0,m,c):\n    g = 9.81\n    zmin = 0.\n    tau = m / c\n    vinf = -m * g / c\n\n    # mesh nodes\n    t = getMesh().getVertices()\n\n    z = [max(z0 + vinf * t_i[0] + tau * (v0 - vinf) * (1 - exp(-t_i[0] / tau)), zmin) for t_i in t]\n    return z'
>>> model = persalys.PythonFieldModel('aModel', meshModel, [z0, v0, m, c], [z], code)

>>> y = model.getPointToFieldFunction()([100, 55, 80, 15])

Methods

`addInput`(input)	Add an input variable.
`addOutput`(output)	Add an output variable.
`clearInputs`()	Remove all the input variables.
`clearOutputs`()	Remove all the output variables.
`exportStandalonePythonScript`(fileName)	Export the model into a Python script.
`getClassName`()	Accessor to the object's name.
`getCode`()	Code accessor.
`getCopula`()	Copula accessor.
`getDistribution`()	Distribution accessor.
`getEvalTime`()	Evaluation time accessor.
`getFunction`(*args)	Accessor to the underlying function.
`getInputByName`(*args)	Input variable accessor.
`getInputDimension`()	Number of input variables accessor.
`getInputNames`()	Input variable names accessor.
`getInputRandomVector`()	Input random vector accessor.
`getInputs`()	Input variables accessor.
`getMeshModel`()	Mesh model accessor.
`getName`()	Accessor to the object's name.
`getOutputByName`(*args)	Output variable accessor.
`getOutputDimension`()	Number of output variables accessor.
`getOutputNames`()	Output variable names accessor.
`getOutputRandomVector`(outputNames)	Output random vector accessor.
`getOutputs`()	Output variables accessor.
`getPointToFieldFunction`(*args)	Accessor to the underlying function.
`getProcessNumber`()	Number of concurrent processes.
`getPythonScript`()	Python script accessor.
`getRestrictedFunction`(*args)	Accessor to the function restricted to its stochastic inputs.
`getRestrictedPointToFieldFunction`(outputNames)	Accessor to the field function restricted to its stochastic inputs.
`getSelectedOutputsNames`()	Accessor to the selected outputs names.
`getStochasticInputNames`()	Accessor to the names of the stochastic input variables.
`hasInputNamed`(inputName)	Check if an input has the given name.
`hasMesh`()	Whether the model has a mesh.
`hasName`()	Test if the object is named.
`hasOutputNamed`(outputName)	Check if an output has the given name.
`hasStochasticInputs`()	Whether it contains stochastic variables.
`isParallel`()	Whether the evaluations of the model are parallelized.
`isValid`()	Whether it is valid.
`removeInput`(inputName)	Remove an input variable.
`removeOutput`(outputName)	Remove an output variable.
`selectOutput`(outputName, selected)	Select output.
`setCode`(code)	Code accessor.
`setCopula`(inputNames, copula)	Copula accessor.
`setDistribution`(inputName, distribution)	Input distribution accessor.
`setDistributionParametersType`(inputName, ...)	Input distribution parameter type accessor.
`setEvalTime`(evalTime)	Evaluation time accessor.
`setFiniteDifferenceStep`(inputName, step)	Finite difference step accessor.
`setInputDescription`(inputName, description)	Input description accessor.
`setInputName`(inputName, newName)	Input variable name accessor.
`setInputStochastic`(inputName, stoch)	Whether the input variable is stochastic.
`setInputValue`(inputName, value)	Input variable value accessor.
`setInputs`(inputs)	Input variables accessor.
`setMeshModel`(meshModel)	Mesh model accessor.
`setName`(name)	Accessor to the object's name.
`setOutputDescription`(outputName, description)	Output description accessor.
`setOutputName`(outputName, newName)	Output variable name accessor.
`setOutputValue`(outputName, value)	Output variable name accessor.
`setOutputs`(outputs)	Output variables accessor.
`setParallel`(flag)	Whether the evaluations of the model are parallelized.
`setProcessNumber`(processNumber)	Number of concurrent processes.

__init__(*args)¶

addInput(input)¶

Add an input variable.

Parameters

inputInput: New variable

addOutput(output)¶

Add an output variable.

Parameters

outputOutput: New variable

clearInputs()¶: Remove all the input variables.

clearOutputs()¶: Remove all the output variables.

exportStandalonePythonScript(fileName)¶

Export the model into a Python script.

In the case of metamodels the script only depends on OpenTURNS and is meant to be used out of Persalys.

Parameters

fileNamestr: Path of the exported script

getClassName()¶

Accessor to the object’s name.

Returns

class_namestr: The object class name (object.__class__.__name__).

getCode()¶

Code accessor.

Returns

codestr: Code text

getCopula()¶

Copula accessor.

Parameters

copulaopenturns.Copula: The copula

getDistribution()¶

Distribution accessor.

Parameters

Distributionopenturns.JointDistribution: The composed distribution (marginals and dependence)

getEvalTime()¶

Evaluation time accessor.

Returns

evalTimefloat: Evaluation time

getFunction(*args)¶

Accessor to the underlying function.

Parameters

outputNamessequence of str: Names of the outputs to be evaluated (optional)

Returns

functionopenturns.Function: Function evaluating the outputs

getInputByName(*args)¶

Input variable accessor.

Parameters

namestr: Variable name

Returns

inputInput: Input variable

getInputDimension()¶

Number of input variables accessor.

Returns

dimensionint: Number of input variables

getInputNames()¶

Input variable names accessor.

Returns

namesopenturns.Description: Input variable names

getInputRandomVector()¶

Input random vector accessor.

Returns

namesopenturns.RandomVector: Input random vector

getInputs()¶

Input variables accessor.

Returns

inputssequence of Input: Input variables

getMeshModel()¶

Mesh model accessor.

Returns

meshMeshModel: Mesh model

getName()¶

Accessor to the object’s name.

Returns

namestr: The name of the object.

getOutputByName(*args)¶

Output variable accessor.

Parameters

namestr: Variable name

Returns

inputOutput: Output variable

getOutputDimension()¶

Number of output variables accessor.

Returns

dimensionint: Number of output variables

getOutputNames()¶

Output variable names accessor.

Returns

namesopenturns.Description: Output variable names

getOutputRandomVector(outputNames)¶

Output random vector accessor.

Returns

namesopenturns.RandomVector: Output random vector

getOutputs()¶

Output variables accessor.

Returns

outputssequence of Output: Output variables

getPointToFieldFunction(*args)¶

Accessor to the underlying function.

Parameters

outputNamessequence of str: Names of the outputs to be evaluated (optional)

Returns

functionopenturns.PointToFieldFunction: Function evaluating the outputs along the mesh nodes

getProcessNumber()¶

Number of concurrent processes.

Returns

process_numberint: Maximum number of processes

getPythonScript()¶

Python script accessor.

Returns

scriptstr: Python script to replay the model

getRestrictedFunction(*args)¶

Accessor to the function restricted to its stochastic inputs.

Parameters

outputNamessequence of str: Output variables (optional)

Returns

restrictedopenturns.Function: Stochastic function

getRestrictedPointToFieldFunction(outputNames)¶

Accessor to the field function restricted to its stochastic inputs.

Parameters

outputNamessequence of str: Output variables (optional)

Returns

restrictedopenturns.PointToFieldFunction: Stochastic field function (for model with mesh)

getSelectedOutputsNames()¶

Accessor to the selected outputs names.

Returns

namesopenturns.Description: Selected outputs names

getStochasticInputNames()¶

Accessor to the names of the stochastic input variables.

Returns

namesopenturns.Description: Names of the stochastic input variables

hasInputNamed(inputName)¶

Check if an input has the given name.

Parameters

namestr: Variable name

Returns

hasbool: Whether an input has the given name

hasMesh()¶

Whether the model has a mesh.

Returns

hasMeshbool: Whether the model has a mesh

hasName()¶

Test if the object is named.

Returns

hasNamebool: True if the name is not empty.

hasOutputNamed(outputName)¶

Check if an output has the given name.

Parameters

namestr: Variable name

Returns

hasbool: Whether an output has the given name

hasStochasticInputs()¶

Whether it contains stochastic variables.

Returns

hasbool: Whether it contains stochastic variables

isParallel()¶

Whether the evaluations of the model are parallelized.

Returns

isParallelbool: Whether the evaluations of the model are parallelized (available only for the Python model)

isValid()¶

Whether it is valid.

Returns

isValidbool: Whether it is valid

removeInput(inputName)¶

Remove an input variable.

Parameters

namestr: Variable name

removeOutput(outputName)¶

Remove an output variable.

Parameters

namestr: Variable name

selectOutput(outputName, selected)¶

Select output.

Parameters

namestr: Output name
isSelectedbool: Whether it is selected

setCode(code)¶

Code accessor.

Parameters

codestr: Code text

setCopula(inputNames, copula)¶

Copula accessor.

Parameters

copulaopenturns.Copula: The copula

setDistribution(inputName, distribution)¶

Input distribution accessor.

Parameters

namestr: Variable name
distributionopenturns.Distribution: Variable distribution

setDistributionParametersType(inputName, distributionParametersType)¶

Input distribution parameter type accessor.

Parameters

namestr: Variable name
parametersTypeint: Distribution parameter type accessor

setEvalTime(evalTime)¶

Evaluation time accessor.

Parameters

evalTimefloat: Evaluation time

setFiniteDifferenceStep(inputName, step)¶

Finite difference step accessor.

Parameters

namestr: Variable name
stepfloat: Finite difference step used to define the gradient of the model’s function

setInputDescription(inputName, description)¶

Input description accessor.

Parameters

namestr: Variable name
descriptionstr: Description text

setInputName(inputName, newName)¶

Input variable name accessor.

Parameters

namestr: Variable name
newNamestr: New name

setInputStochastic(inputName, stoch)¶

Whether the input variable is stochastic.

Parameters

namestr: Variable name
isStochasticbool: Whether the input variable is stochastic

setInputValue(inputName, value)¶

Input variable value accessor.

Parameters

namestr: Variable name
valuefloat: New value

setInputs(inputs)¶

Input variables accessor.

Parameters

inputssequence of Input: Input variables

setMeshModel(meshModel)¶

Mesh model accessor.

Parameters

meshMeshModel: Mesh model

setName(name)¶

Accessor to the object’s name.

Parameters

namestr: The name of the object.

setOutputDescription(outputName, description)¶

Output description accessor.

Parameters

namestr: Variable name
descriptionstr: Description text

setOutputName(outputName, newName)¶

Output variable name accessor.

Parameters

namestr: Variable name
newNamestr: New name

setOutputValue(outputName, value)¶

Output variable name accessor.

Parameters

namestr: Variable name
valuefloat: New value

setOutputs(outputs)¶

Output variables accessor.

Parameters

outputssequence of Output: Output variables

setParallel(flag)¶

Whether the evaluations of the model are parallelized.

Parameters

isParallelbool: Whether the evaluations of the model are parallelized

Notes

This is only implemented for the Python model. Also, on Windows platforms, ProcessPoolExecutor startup time being much higher, you may want to evaluate enough points in batch rather than many small blocks. Refer to the setBlockSize method of the various algorithms.

setProcessNumber(processNumber)¶

Number of concurrent processes.

Parameters

process_numberint: Maximum number of processes

PythonFieldModel¶

Previous topic

Next topic

This Page