import persalys
import os
anOTStudy = persalys.Study('anOTStudy')
# Model
X0 = persalys.Input('X0', 1, ot.Normal())
X1 = persalys.Input('X1', 2, ot.Normal())
Y0 = persalys.Output('Y0')
model = persalys.SymbolicPhysicalModel('aModelPhys', [X0, X1], [Y0],
['sin(X0)+8*X1'])
anOTStudy.add(model)
# Design of Experiment ##
aDesign = persalys.GridDesignOfExperiment('aDesign_1', model,
[[0.9, 1.1], [1.8, 2.2]])
anOTStudy.add(aDesign)
aDesign.run()
print('outs=', aDesign.getResult().getDesignOfExperiment().getOutputSample())
# Design of Experiment ##
filename = 'normal.csv'
ot.Normal(3).getSample(10).exportToCSVFile(filename)
aDesign2 = persalys.ImportedDesignOfExperiment('aDesign_2', model, filename,
[0, 2])
anOTStudy.add(aDesign2)
aDesign2.run()
print('outs=', aDesign2.getResult().getDesignOfExperiment().getOutputSample())
sample.exportToCSVFile(filename, ' ')
# Designs of Experiment ##
# fixed design ##
ot.RandomGenerator.SetSeed(0)
fixedDesign = persalys.FixedDesignOfExperiment('fixedDesign', symbolicModel)
inputSample = ot.LHSExperiment(ot.ComposedDistribution([ot.Uniform(0., 10.), ot.Uniform(0., 10.)]), 10).generate()
inputSample.stack(ot.Sample(10, [0.5]))
fixedDesign.setOriginalInputSample(inputSample)
fixedDesign.run()
myStudy.add(fixedDesign)
# grid ##
values = [[0.5+i*1.5 for i in range(7)], [0.5+i*1.5 for i in range(7)], [1]]
grid = persalys.GridDesignOfExperiment('grid', symbolicModel, values)
myStudy.add(grid)
# importDesign ##
importDesign = persalys.ImportedDesignOfExperiment('importDesign', symbolicModel, 'data.csv', [0, 2, 3])
importDesign.run()
myStudy.add(importDesign)
# onePointDesign ##
onePointDesign = persalys.GridDesignOfExperiment('onePointDesign', pythonModel)
myStudy.add(onePointDesign)
# twoPointsDesign ##
twoPointsDesign = persalys.GridDesignOfExperiment('twoPointsDesign', pythonModel, [[0.2], [1.2], [-0.2, 1.]])
myStudy.add(twoPointsDesign)
C = persalys.Input('C', 4000, dist_C, 'Valued thermal energy (W.h)')
Ep = persalys.Output('Ep', 'Primary energy savings (W.h)')
model = persalys.SymbolicPhysicalModel('myPhysicalModel', [Q, E, C], [Ep],
['1-(Q/((E/((1-0.05)*0.54))+(C/0.8)))'])
myStudy.add(model)
outputSample = [[0.060036508072], [0.0812679132055], [0.0684295269203],
[0.0892876773294], [0.0292238907867], [0.0511512752497],
[0.0378920382755], [0.0594339324804]]
# Design of Experiment - Parametric analysis ##
bounds = persalys.GridDesignOfExperiment.GetDefaultBounds(model)
values = [[bounds.getLowerBound()[i],
bounds.getUpperBound()[i]] for i in range(3)]
aDesign = persalys.GridDesignOfExperiment('aDesign_0', model, values)
myStudy.add(aDesign)
aDesign.run()
# Comparaison
openturns.testing.assert_almost_equal(
outputSample,
aDesign.getResult().getDesignOfExperiment().getOutputSample(), 1e-16)
# Taylor Expansions ##
taylorExpansionsMoments = persalys.TaylorExpansionMomentsAnalysis(
'myTaylorExpansionMoments', model)
myStudy.add(taylorExpansionsMoments)
taylorExpansionsMoments.run()
taylorExpansionsMomentsResult = taylorExpansionsMoments.getResult()
R = persalys.Input('R', 300., dist_R, 'Yield stress (Pa)')
F = persalys.Input('F', 75000., dist_F, 'Traction force (N)')
G = persalys.Output('G', 'Stress difference (Pa)')
code = 'from math import pi\n\ndef _exec(R, F):\n G = R-F/(pi*100.0)\n return G\n'
model = persalys.PythonPhysicalModel('myPhysicalModel', [R, F], [G], code)
myStudy.add(model)
f = model.getFunction()
print(f([300., 75000.]))
# Design of Experiment - Parametric analysis ##
bounds = persalys.GridDesignOfExperiment.GetDefaultBounds(model)
values = [[bounds.getLowerBound()[i],
bounds.getUpperBound()[i]] for i in range(2)]
aDesign = persalys.GridDesignOfExperiment('design', model, values)
myStudy.add(aDesign)
aDesign.run()
# Taylor Expansion ##
taylorExpansionsMoments = persalys.TaylorExpansionMomentsAnalysis(
'myTaylorExpansionMoments', model)
myStudy.add(taylorExpansionsMoments)
taylorExpansionsMoments.run()
taylorExpansionsMomentsResult = taylorExpansionsMoments.getResult()
# Monte Carlo ##
montecarlo = persalys.MonteCarloAnalysis('myMonteCarlo', model)
montecarlo.setMaximumCalls(1000)
myStudy.add(montecarlo)
montecarlo.run()
[ot.Normal(), ot.Gumbel(),
ot.Normal(), ot.Uniform()],
ot.ComposedCopula([ot.IndependentCopula(2),
ot.GumbelCopula()]))
sample = cDist.getSample(200)
sample.exportToCSVFile(filename, ' ')
model3 = persalys.DataModel('model3', 'data.csv', [0, 2, 3], [1],
['x_0', 'x_2', 'x_3'], ['x_1'])
myStudy.add(model3)
# Designs of Experiment ##
# design 2 ##
values = [[0.5 + i * 1.5 for i in range(7)], [0.5 + i * 1.5 for i in range(7)],
[1]]
design_2 = persalys.GridDesignOfExperiment('design_2', model1, values)
myStudy.add(design_2)
# design 4 ##
probaDesign = persalys.ProbabilisticDesignOfExperiment('probaDesign', model1,
100, "MONTE_CARLO")
probaDesign.run()
myStudy.add(probaDesign)
design_3 = persalys.ImportedDesignOfExperiment('design_3', model1, 'data.csv',
[0, 2, 3])
design_3.run()
myStudy.add(design_3)
# 1- meta model1 ##
dist_X0 = ot.Normal(0, 1)
X0 = persalys.Input('X0', 0, dist_X0, '')
dist_X1 = ot.Normal(0, 1)
X1 = persalys.Input('X1', 0, dist_X1, '')
X2 = persalys.Input('X2', 0, '')
X3 = persalys.Input('X3', 0, '')
Y0 = persalys.Output('Y0', '')
inputs = [X0, X1, X2, X3]
outputs = [Y0]
formulas = ['X0+X1+X2+X3']
SymbolicModel_0 = persalys.SymbolicPhysicalModel('SymbolicModel_0', inputs, outputs, formulas)
values = [[0],
[0],
[-0.1, 0.1],
[-0.1, 0.1]]
design_0 = persalys.GridDesignOfExperiment('design_0', SymbolicModel_0, values)
design_0.setBlockSize(1)
interestVariables = ['Y0']
design_0.setInterestVariables(interestVariables)
design_0.run()
metaModel_0 = persalys.FunctionalChaosAnalysis('metaModel_0', design_0)
metaModel_0.setChaosDegree(1)
metaModel_0.setSparseChaos(False)
metaModel_0.setAnalyticalValidation(True)
metaModel_0.setTestSampleValidation(False)
metaModel_0.setKFoldValidation(False)
metaModel_0.run()
print(metaModel_0.getResult())
# deterministic/chaos
X0 = persalys.Input('X0', 0, '')