1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
|
#! /usr/bin/env python
import openturns as ot
import openturns.experimental as otexp
import openturns.testing as ott
ot.TESTPREAMBLE()
f1 = ot.SymbolicFunction(
["a0", "a1"],
["-4 * exp((-25 / 8) * (a0^2 + a1^2)) + 7 * exp((-125 / 4) * (a0^2 + a1^2))"] * 2,
)
distribution = ot.JointDistribution([ot.Uniform(-1.0, 1.0)] * 2)
x0 = ot.LowDiscrepancyExperiment(ot.HaltonSequence(), distribution, 100).generate()
y0 = f1(x0)
# reference mean for each generated batch of input x
ref_mean = [
[-0.0271247, -0.0171091],
[-0.0763905, 0.0570268],
[0.0832077, -0.0235941],
[0.0488792, 0.158224],
[0.0511363, 0.113745],
]
algo = otexp.LOLAVoronoi(x0, y0, distribution)
newX = ot.Sample(0, x0.getDimension())
inc = 15
for i in range(5):
x = algo.generate(inc)
newX.add(x)
y = f1(x)
algo.update(x, y)
x_mean = x.computeMean()
print(f"iteration={i} x_mean={x_mean}")
lolaScore = algo.getLOLAScore()
voronoiScore = algo.getVoronoiScore()
assert len(lolaScore) == len(x0) + i * inc
assert len(lolaScore) == len(voronoiScore)
# FIXME: legacy KDTree is incorrect https://github.com/openturns/openturns/issues/2617
if ot.PlatformInfo.HasFeature("nanoflann"):
ott.assert_almost_equal(x_mean, ref_mean[i])
if False:
import openturns.viewer as otv
cloud1 = ot.Cloud(x0)
cloud1.setPointStyle("fcircle")
cloud1.setColor("blue")
cloud2 = ot.Cloud(newX)
cloud2.setPointStyle("fcircle")
cloud2.setColor("red")
graph = ot.Graph("LOLA-Voronoi", "x1", "x2", True)
graph.add(cloud1)
graph.add(cloud2)
otv.View(graph)
otv.View.ShowAll()
# linear model: the non linearity score should have no effect
ot.ResourceMap.SetAsString("LOLAVoronoi-NonLinearityAggregationMethod", "Maximum")
f2 = ot.SymbolicFunction(["x0", "x1"], ["2 * x0 + 3 * x1 + 8"])
y0 = f2(x0)
algo = otexp.LOLAVoronoi(x0, y0, distribution)
for i in range(10):
x = algo.generate(20)
y = f2(x)
algo.update(x, y)
x_mean = x.computeMean()
print(f"iteration={i} x_mean={x_mean}")
ott.assert_almost_equal(algo.getLOLAScore().asPoint().norm(), 0.0)
X = algo.getInputSample()
discrepancy = ot.SobolSequence.ComputeStarDiscrepancy(X)
assert discrepancy < 1.0, "wrong repartition"
# oscillating model on a grid
ot.ResourceMap.SetAsUnsignedInteger("LOLAVoronoi-MaximumCombinationsNumber", 3870)
f3 = ot.SymbolicFunction(["x0", "x1"], ["sin(10 * x0) + cos(10 * x1)"])
x0 = ot.Box([8, 8]).generate()
y0 = f3(x0)
algo = otexp.LOLAVoronoi(x0, y0, distribution)
for i in range(10):
x = algo.generate(20)
y = f3(x)
algo.update(x, y)
x_mean = x.computeMean()
print(f"iteration={i} x_mean={x_mean}")
X = algo.getInputSample()
discrepancy = ot.SobolSequence.ComputeStarDiscrepancy(X)
assert discrepancy < 1.0, "wrong repartition"
|
