9
9
# Instanciate one distribution object
10
10
distribution = Gumbel(2.0, -0.5)
11
print "Distribution " , distribution
11
print "Distribution " , distribution
13
13
# Is this distribution elliptical ?
14
print "Elliptical = ", distribution.isElliptical()
14
print "Elliptical = ", distribution.isElliptical()
16
16
# Is this distribution continuous ?
17
print "Continuous = ", distribution.isContinuous()
17
print "Continuous = ", distribution.isContinuous()
19
19
# Test for realization of distribution
20
20
oneRealization = distribution.getRealization()
21
21
print "oneRealization=", repr(oneRealization)
25
25
oneSample = distribution.getNumericalSample( size )
26
26
print "oneSample first=" , repr(oneSample[0]) , " last=" , repr(oneSample[1])
27
print "mean=" , repr(oneSample.computeMean())
28
print "covariance=" , oneSample.computeCovariance()
27
print "mean=" , repr(oneSample.computeMean())
28
print "covariance=" , repr(oneSample.computeCovariance())
31
31
point = NumericalPoint( distribution.getDimension(), 1.0 )
36
36
# derivative of PDF with regards its arguments
37
37
DDF = distribution.computeDDF( point )
38
38
print "ddf =" , repr(DDF)
39
# by the finite difference technique
39
# by the finite difference technique
40
40
print "ddf (FD)=" , repr(NumericalPoint(1, (distribution.computePDF( point + NumericalPoint(1, eps) ) - distribution.computePDF( point + NumericalPoint(1, -eps) )) / (2.0 * eps)))
43
43
PDF = distribution.computePDF( point )
44
print "pdf =%.6f" % PDF
44
print "pdf =%.6f" % PDF
45
45
# by the finite difference technique from CDF
46
46
print "pdf (FD)=%.6f" % ((distribution.computeCDF( point + NumericalPoint(1, eps) ) - distribution.computeCDF( point + NumericalPoint(1, -eps) )) / (2.0 * eps))
48
48
# derivative of the PDF with regards the parameters of the distribution
49
49
CDF = distribution.computeCDF( point )
50
print "cdf=%.6f" % CDF
50
print "cdf=%.6f" % CDF
51
51
PDFgr = distribution.computePDFGradient( point )
52
52
print "pdf gradient =" , repr(PDFgr)
53
53
# by the finite difference technique
55
55
PDFgrFD[0] = (Gumbel(distribution.getAlpha() + eps, distribution.getBeta()).computePDF(point) - Gumbel(distribution.getAlpha() - eps, distribution.getBeta()).computePDF(point)) / (2.0 * eps)
56
56
PDFgrFD[1] = (Gumbel(distribution.getAlpha(), distribution.getBeta() + eps).computePDF(point) - Gumbel(distribution.getAlpha(), distribution.getBeta() - eps).computePDF(point)) / (2.0 * eps)
57
57
print "pdf gradient (FD)=" , repr(PDFgrFD)
59
59
# derivative of the PDF with regards the parameters of the distribution
60
60
CDFgr = distribution.computeCDFGradient( point )
61
print "cdf gradient =" , repr(CDFgr)
61
print "cdf gradient =" , repr(CDFgr)
62
62
# by the finite difference technique
63
63
CDFgrFD = NumericalPoint(2)
64
64
CDFgrFD[0] = (Gumbel(distribution.getAlpha() + eps, distribution.getBeta()).computeCDF(point) - Gumbel(distribution.getAlpha() - eps, distribution.getBeta()).computeCDF(point)) / (2.0 * eps)
69
69
quantile = distribution.computeQuantile( 0.95 )
70
print "quantile=" , repr(quantile)
71
print "cdf(quantile)=%.6f" % distribution.computeCDF(quantile)
70
print "quantile=" , repr(quantile)
71
print "cdf(quantile)=%.6f" % distribution.computeCDF(quantile)
72
72
mean = distribution.getMean()
73
print "mean=" , repr(mean)
73
print "mean=" , repr(mean)
74
74
standardDeviation = distribution.getStandardDeviation()
75
print "standard deviation=" , repr(standardDeviation)
75
print "standard deviation=" , repr(standardDeviation)
76
76
skewness = distribution.getSkewness()
77
print "skewness=" , repr(skewness)
77
print "skewness=" , repr(skewness)
78
78
kurtosis = distribution.getKurtosis()
79
print "kurtosis=" , repr(kurtosis)
79
print "kurtosis=" , repr(kurtosis)
80
80
covariance = distribution.getCovariance()
81
print "covariance=" , covariance
81
print "covariance=" , repr(covariance)
82
82
parameters = distribution.getParametersCollection()
83
print "parameters=" , repr(parameters)
83
print "parameters=" , repr(parameters)
85
85
# Specific to this distribution
86
86
mu = distribution.getMu()
88
88
sigma = distribution.getSigma()
89
print "sigma=%.6f" % sigma
89
print "sigma=%.6f" % sigma
90
90
newDistribution = Gumbel(mu, sigma, 1)
91
91
print "alpha from (mu, sigma)=%.6f" % newDistribution.getAlpha()
92
92
print "beta from (mu, sigma)=%.6f" % newDistribution.getBeta()
96
print "t_Gumbel.py", sys.exc_type, sys.exc_value
96
print "t_Gumbel.py", sys.exc_type, sys.exc_value