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
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
|
#! /usr/bin/env python
import openturns as ot
import openturns.testing as ott
import time
print(
"Continuous univariate factories=",
ot.DistributionFactory.GetContinuousUniVariateFactories(),
)
print(
"Continuous multivariate factories=",
ot.DistributionFactory.GetContinuousMultiVariateFactories(),
)
print(
"Discrete univariate factories=",
ot.DistributionFactory.GetDiscreteUniVariateFactories(),
)
print(
"Discrete multivariate factories=",
ot.DistributionFactory.GetDiscreteMultiVariateFactories(),
)
print("Univariate factories=", ot.DistributionFactory.GetUniVariateFactories())
print("Multivariate factories=", ot.DistributionFactory.GetMultiVariateFactories())
factories = ot.DistributionFactory.GetContinuousUniVariateFactories()
factories.add(ot.DistributionFactory.GetDiscreteUniVariateFactories())
for factory in factories:
print(f"-- {factory}...")
dist = factory.build()
# check if raise on constant sample
if dist.isContinuous():
sample = ot.Sample(100, [1.0e5])
ok = False
try:
dist = factory.build(sample)
except Exception:
ok = True
print("constant:", "ok" if ok else "fail")
assert ok, str(factory) + " vs const"
# check if raises on sample with nan/inf
for weird in ["nan", "inf"]:
sample = dist.getSample(100)
sample[0, 0] = float(weird)
ok = False
try:
dist = factory.build(sample)
except Exception:
ok = True
print(weird + ":", "ok" if ok else "fail")
assert ok, str(factory) + " vs " + weird
factory = ot.DistributionFactory.GetByName("PoissonFactory")
print(factory)
assert factory.getImplementation().__class__.__name__ == "PoissonFactory", "wrong name"
# Now, test each factory for the partial estimation of any subset of its parameters
params = {
"FisherSnedecorFactory": 0.25,
"LogisticFactory": 0.025,
"ParetoFactory": 0.15,
"StudentFactory": 0.15,
"StudentFactory": 0.15,
"TrapezoidalFactory": 1.0,
"TriangularFactory": 0.025,
}
for factory in factories:
print(f"-- {factory}...")
try:
factoryName = factory.getImplementation().getClassName()
except Exception:
print("Skip", factory, "no getClassName()")
continue
# Skip HistogramFactory because it is nonparametric, so the rebuilt distribution has a number of parameters different from the reference distribution
# Skip some others because they are quite slow
distName = factoryName.replace("Factory", "")
if distName in [
"Rice",
"Student",
"Trapezoidal",
"TruncatedNormal",
"VonMises",
"Histogram",
"MeixnerDistribution",
"Pareto",
"Dirichlet",
"InverseNormal",
]:
continue
eps = params.get(factoryName, 0.05)
# Reference distribution
refDistribution = factory.build()
sample = refDistribution.getSample(10000)
# Get the ref parameter
refParameter = refDistribution.getParameter()
n = len(refParameter)
t0 = time.time()
for k in range(n):
# Generate all the combinations of known parameter
combinations = ot.Combinations(k, n).generate()
for combination in combinations:
indices = [int(x) for x in combination]
values = [refParameter[i] for i in indices]
factory.setKnownParameter(indices, values)
assert factory.getKnownParameterIndices() == indices
assert factory.getKnownParameterValues() == values
res = factory.build(sample)
resParameter = res.getParameter()
ott.assert_almost_equal(resParameter, refParameter, eps, eps)
t1 = time.time()
print(f"done in t={(t1 - t0):.3f} s")
|
