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
|
#! /usr/bin/env python
import openturns as ot
ot.TESTPREAMBLE()
ot.RandomGenerator.SetSeed(0)
# Create a collection of distribution
aCollection = [ot.Normal(0.0, 4), ot.Uniform(5.0, 7.0), ot.Triangular(7.0, 8.0, 9.0)]
# Instantiate one distribution object
distribution = ot.Mixture(aCollection)
print("mixture=", distribution)
classifier = ot.MixtureClassifier(distribution)
inS = ot.Sample([[2.0], [4.0], [6.0], [8.0]])
for i in range(inS.getSize()):
print("inP=", inS[i], " class=", classifier.classify(inS[i]))
print("classes=", classifier.classify(inS))
for i in range(inS.getSize()):
for j in range(len(aCollection)):
grade = classifier.grade(inS[i], j)
# TODO JM: remove the check after the use of infs has been thoroughly tested
if grade <= ot.SpecFunc.LowestScalar:
grade *= 2.0
print("inP=", inS[i], " grade|", j, "= %g" % grade)
for j in range(len(aCollection)):
grades = classifier.grade(inS, ot.Indices(inS.getSize(), j))
for num, grade in enumerate(grades):
# TODO JM: remove the check after the use of infs has been thoroughly tested
if grade <= ot.SpecFunc.LowestScalar:
grade *= 2.0
grades[num] = grade
print("grades|", j, "=", grades)
|
