StandardSpaceCrossEntropyImportanceSampling

class StandardSpaceCrossEntropyImportanceSampling(*args)

Standard Space Cross-Entropy Importance Sampling.

Warning

This class is experimental and likely to be modified in future releases. To use it, import the openturns.experimental submodule.

Parameters:
eventThresholdEvent

Event we are computing the probability of.

quantileLevelfloat 0<quantileLevel<1

Intermediate quantile level. The default number can be tweaked with the CrossEntropyImportanceSampling-DefaultQuantileLevel key from ResourceMap.

Notes

The following explanations are given for a failure event defined as g(\vect{X})<T with \vect{X} a random vector following a joint PDF f_\vect{X}, T a threshold and g a limit state function, without loss of generality.

The Importance Sampling (IS) probability estimate \widehat{P}^\text{IS} is given by:

\widehat{P}^\text{IS}=\frac{1}{N} \sum_{i=1}^{N} {\vect{1}}_{g(\vect{x}_i)<T} \frac{f_\vect{X}(\vect{x}_i)}{h(\vect{x}_i)},

with h the auxiliary PDF of Importance Sampling, N the number of independent samples generated with h and {\vect{1}}_{g(\vect{x}_i)<T} the indicator function of the failure domain.

The optimal density h_{opt} minimizing the variance of the estimator h_{opt} is defined as:

h_{opt}=\frac{{\vect{1}}_{g(\vect{x})<T}f_\vect{X}}{P},

with P the failure probability which is inaccessible in practice since this probability is the quantity of interest and unknown.

The Standard Space Cross-Entropy Importance Sampling [rubinstein2017] uses a parametric auxiliary distribution h_\vect{\lambda} in order to optimize its parameters to compute the probability of interest with accuracy. This process is performed in the standard space using a Normal distribution as the auxiliary distribution. Indeed, in the standard space, the optimal value of the mean and standard deviations are given analytically, see [morio2015] for details.

We note G(\vect{u}) the limit state function in standard space and g(\vect{x}) its counterpart in the physical space. f_\vect{U} stands for the initial distribution in standard space, i.e. the Normal law. The algorithm is as follows:

  1. k=1, set the quantile level \rho \in [0,1] and h_0 = f_{\vect{U}}

  2. Generate the population \vect{u}_1^{(k)},...,\vect{u}_N^{(k)} according to the PDF h_{k-1}, apply the function G in order to have y_1^{(k)}=G(\vect{u}_1^{(k)}),...,y_N^{(k)} = G(\vect{u}_N^{(k)})

  3. Compute the \rho-quantile q_k=\max(T,y^{(k)}_{\left \lfloor \rho N \right\rfloor}) using the floor of \rho N.

  4. If q_k>T, go to Step 7

  1. Update the mean and standard deviation of the auxiliary distribution:

\mu_{k}= \frac{\frac{1}{N} \sum_{i=1}^{N}{\vect{1}}_{G(\vect{u}_i^{(k)}) \leq q_k} \frac{f_\vect{U}(\vect{u}_i^{(k)})}{h_{\vect{\lambda}_{k-1}}(\vect{u}_i^{(k)})} \vect{u}_i^{(k)}}{\frac{1}{N} \sum_{i=1}^{N}{\vect{1}}_{G(\vect{u}_i^{(k)}) \leq q_k} \frac{f_\vect{U}(\vect{u}_i^{(k)})}{h_{\vect{\lambda}_{k-1}}(\vect{u}_i^{(k)})} }

\sigma_{k}= \frac{\frac{1}{N} \sum_{i=1}^{N}{\vect{1}}_{G(\vect{u}_i^{(k)}) \leq q_k} \frac{f_\vect{U}(\vect{u}_i^{(k)})}{h_{\vect{\lambda}_{k-1}}(\vect{u}_i^{(k)})} (\vect{u}_i^{(k)}-\mu_k)^2}{\frac{1}{N} \sum_{i=1}^{N}{\vect{1}}_{G(\vect{u}_i^{(k)}) \leq q_k} \frac{f_\vect{U}(\vect{u}_i^{(k)})}{h_{\vect{\lambda}_{k-1}}(\vect{u}_i^{(k)})} }

  1. k\leftarrow k+1, go to Step 2

  2. Estimate the probability \widehat{P}^{CE}(g(\vect{\vect{X}})<T)=\widehat{P}^{CE}(G(\vect{\vect{U}})<T)=\frac{1}{N}\displaystyle \sum_{i=1}^{N} \vect{1}_{G(\vect{u}_i^{(k)})<T} \frac{f_\vect{U}(\vect{u}_i^{(k)})}{h_{\vect{\lambda}_{k-1}(\vect{u}_i^{(k)})}}

Examples

>>> import openturns as ot
>>> import openturns.experimental as otexp
>>> ot.RandomGenerator.SetSeed(0)
>>> distribution_R = ot.LogNormalMuSigma(300.0, 30.0, 0.0).getDistribution()
>>> distribution_F = ot.Normal(75e3, 5e3)
>>> marginals = [distribution_R, distribution_F]
>>> distribution = ot.JointDistribution(marginals)
>>> # create the model
>>> model = ot.SymbolicFunction(['R', 'F'], ['R - F / (pi_ * 100.0)'])  
>>> # create the event 
>>> vect = ot.RandomVector(distribution)
>>> g = ot.CompositeRandomVector(model, vect)
>>> event = ot.ThresholdEvent(g, ot.Less(), 0.0)
>>> myIS = otexp.StandardSpaceCrossEntropyImportanceSampling(event, 0.3)
>>> # run StandardSpaceCrossEntropyImportanceSampling
>>> myIS.run()
>>> myResult = myIS.getResult()

Methods

drawProbabilityConvergence(*args)

Draw the probability convergence at a given level.

getBlockSize()

Accessor to the block size.

getClassName()

Accessor to the object's name.

getConvergenceStrategy()

Accessor to the convergence strategy.

getEvent()

Accessor to the event.

getInputSample(*args)

Input sample accessor.

getMaximumCoefficientOfVariation()

Accessor to the maximum coefficient of variation.

getMaximumOuterSampling()

Accessor to the maximum sample size.

getMaximumStandardDeviation()

Accessor to the maximum standard deviation.

getMaximumTimeDuration()

Accessor to the maximum duration.

getName()

Accessor to the object's name.

getOutputSample(*args)

Output sample accessor.

getQuantileLevel()

Accessor to the intermediate quantile level.

getResult()

Accessor to the intermediate quantile level.

getStepsNumber()

Subset steps number accessor.

getThresholdPerStep()

Threshold accessor.

hasName()

Test if the object is named.

run()

Launch simulation.

setBlockSize(blockSize)

Accessor to the block size.

setConvergenceStrategy(convergenceStrategy)

Accessor to the convergence strategy.

setKeepSample(keepSample)

Sample storage accessor.

setMaximumCoefficientOfVariation(...)

Accessor to the maximum coefficient of variation.

setMaximumOuterSampling(maximumOuterSampling)

Accessor to the maximum sample size.

setMaximumStandardDeviation(...)

Accessor to the maximum standard deviation.

setMaximumTimeDuration(maximumTimeDuration)

Accessor to the maximum duration.

setName(name)

Accessor to the object's name.

setProgressCallback(*args)

Set up a progress callback.

setQuantileLevel(quantileLevel)

Accessor to the intermediate quantile level.

setStopCallback(*args)

Set up a stop callback.

__init__(*args)
drawProbabilityConvergence(*args)

Draw the probability convergence at a given level.

Parameters:
levelfloat, optional

The probability convergence is drawn at this given confidence length level. By default level is 0.95.

Returns:
grapha Graph

probability convergence graph

getBlockSize()

Accessor to the block size.

Returns:
blockSizeint

Number of terms in the probability simulation estimator grouped together. It is set by default to 1.

getClassName()

Accessor to the object’s name.

Returns:
class_namestr

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

getConvergenceStrategy()

Accessor to the convergence strategy.

Returns:
storage_strategyHistoryStrategy

Storage strategy used to store the values of the probability estimator and its variance during the simulation algorithm.

getEvent()

Accessor to the event.

Returns:
eventRandomVector

Event we want to evaluate the probability.

getInputSample(*args)

Input sample accessor.

Parameters:
stepint

Iteration index

selectint, optional

Selection flag:

  • EVENT0 : points not realizing the event are selected

  • EVENT1 : points realizing the event are selected

  • BOTH : all points are selected (default)

Returns:
inputSampleSample

Input sample.

getMaximumCoefficientOfVariation()

Accessor to the maximum coefficient of variation.

Returns:
coefficientfloat

Maximum coefficient of variation of the simulated sample.

getMaximumOuterSampling()

Accessor to the maximum sample size.

Returns:
outerSamplingint

Maximum number of groups of terms in the probability simulation estimator.

getMaximumStandardDeviation()

Accessor to the maximum standard deviation.

Returns:
sigmafloat, \sigma > 0

Maximum standard deviation of the estimator.

getMaximumTimeDuration()

Accessor to the maximum duration.

Returns:
maximumTimeDurationfloat

Maximum optimization duration in seconds.

getName()

Accessor to the object’s name.

Returns:
namestr

The name of the object.

getOutputSample(*args)

Output sample accessor.

Parameters:
stepint

Iteration index

selectint, optional

Selection flag:

  • EVENT0 : points not realizing the event are selected

  • EVENT1 : points realizing the event are selected

  • BOTH : all points are selected (default)

Returns:
outputSampleSample

Output sample.

getQuantileLevel()

Accessor to the intermediate quantile level.

Returns:
quantileLevelfloat

Result of Cross Entropy algorithm.

getResult()

Accessor to the intermediate quantile level.

Returns:
CrossEntropyResultCrossEntropyResult

Contains all the results of the Cross Entropy algorithm simulation.

getStepsNumber()

Subset steps number accessor.

Returns:
nint

Number of subset steps, including the initial Monte Carlo sampling.

getThresholdPerStep()

Threshold accessor.

Returns:
thresholdPoint

Threshold values at each step.

hasName()

Test if the object is named.

Returns:
hasNamebool

True if the name is not empty.

run()

Launch simulation.

Notes

It launches the simulation and creates a SimulationResult, structure containing all the results obtained after simulation. It computes the probability of occurrence of the given event by computing the empirical mean of a sample of size at most outerSampling * blockSize, this sample being built by blocks of size blockSize. It allows one to use efficiently the distribution of the computation as well as it allows one to deal with a sample size > 2^{32} by a combination of blockSize and outerSampling.

setBlockSize(blockSize)

Accessor to the block size.

Parameters:
blockSizeint, blockSize \geq 1

Number of terms in the probability simulation estimator grouped together. It is set by default to 1.

Notes

For Monte Carlo, LHS and Importance Sampling methods, this allows one to save space while allowing multithreading, when available we recommend to use the number of available CPUs; for the Directional Sampling, we recommend to set it to 1.

setConvergenceStrategy(convergenceStrategy)

Accessor to the convergence strategy.

Parameters:
storage_strategyHistoryStrategy

Storage strategy used to store the values of the probability estimator and its variance during the simulation algorithm.

setKeepSample(keepSample)

Sample storage accessor.

Parameters:
keepsamplebool

Whether to keep the working samples at each iteration.

setMaximumCoefficientOfVariation(maximumCoefficientOfVariation)

Accessor to the maximum coefficient of variation.

Parameters:
coefficientfloat

Maximum coefficient of variation of the simulated sample.

setMaximumOuterSampling(maximumOuterSampling)

Accessor to the maximum sample size.

Parameters:
outerSamplingint

Maximum number of groups of terms in the probability simulation estimator.

setMaximumStandardDeviation(maximumStandardDeviation)

Accessor to the maximum standard deviation.

Parameters:
sigmafloat, \sigma > 0

Maximum standard deviation of the estimator.

setMaximumTimeDuration(maximumTimeDuration)

Accessor to the maximum duration.

Parameters:
maximumTimeDurationfloat

Maximum optimization duration in seconds.

setName(name)

Accessor to the object’s name.

Parameters:
namestr

The name of the object.

setProgressCallback(*args)

Set up a progress callback.

Can be used to programmatically report the progress of a simulation.

Parameters:
callbackcallable

Takes a float as argument as percentage of progress.

Examples

>>> import sys
>>> import openturns as ot
>>> experiment = ot.MonteCarloExperiment()
>>> X = ot.RandomVector(ot.Normal())
>>> Y = ot.CompositeRandomVector(ot.SymbolicFunction(['X'], ['1.1*X']), X)
>>> event = ot.ThresholdEvent(Y, ot.Less(), -2.0)
>>> algo = ot.ProbabilitySimulationAlgorithm(event, experiment)
>>> algo.setMaximumOuterSampling(100)
>>> algo.setMaximumCoefficientOfVariation(-1.0)
>>> def report_progress(progress):
...     sys.stderr.write('-- progress=' + str(progress) + '%\n')
>>> algo.setProgressCallback(report_progress)
>>> algo.run()
setQuantileLevel(quantileLevel)

Accessor to the intermediate quantile level.

Parameters:
quantileLevelfloat

Intermediate quantile level, must belong to the interval (0,1).

setStopCallback(*args)

Set up a stop callback.

Can be used to programmatically stop a simulation.

Parameters:
callbackcallable

Returns an int deciding whether to stop or continue.

Examples

Stop a Monte Carlo simulation algorithm using a time limit

>>> import openturns as ot
>>> experiment = ot.MonteCarloExperiment()
>>> X = ot.RandomVector(ot.Normal())
>>> Y = ot.CompositeRandomVector(ot.SymbolicFunction(['X'], ['1.1*X']), X)
>>> event = ot.ThresholdEvent(Y, ot.Less(), -2.0)
>>> algo = ot.ProbabilitySimulationAlgorithm(event, experiment)
>>> algo.setMaximumOuterSampling(10000000)
>>> algo.setMaximumCoefficientOfVariation(-1.0)
>>> algo.setMaximumTimeDuration(0.1)
>>> algo.run()

Examples using the class

Cross Entropy Importance Sampling

Cross Entropy Importance Sampling