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
116
117
118
119
120
121
122
123
124
125
126
127
|
[section:weibull Weibull Distribution]
``#include <boost/math/distributions/weibull.hpp>``
namespace boost{ namespace math{
template <class RealType = double,
class ``__Policy`` = ``__policy_class`` >
class weibull_distribution;
typedef weibull_distribution<> weibull;
template <class RealType, class ``__Policy``>
class weibull_distribution
{
public:
typedef RealType value_type;
typedef Policy policy_type;
// Construct:
weibull_distribution(RealType shape, RealType scale = 1)
// Accessors:
RealType shape()const;
RealType scale()const;
};
}} // namespaces
The [@http://en.wikipedia.org/wiki/Weibull_distribution Weibull distribution]
is a continuous distribution
with the
[@http://en.wikipedia.org/wiki/Probability_density_function probability density function]:
f(x; [alpha], [beta]) = ([alpha]\/[beta]) * (x \/ [beta])[super [alpha] - 1] * e[super -(x\/[beta])[super [alpha]]]
For shape parameter [alpha][space] > 0, and scale parameter [beta][space] > 0, and x > 0.
The Weibull distribution is often used in the field of failure analysis;
in particular it can mimic distributions where the failure rate varies over time.
If the failure rate is:
* constant over time, then [alpha][space] = 1, suggests that items are failing from random events.
* decreases over time, then [alpha][space] < 1, suggesting "infant mortality".
* increases over time, then [alpha][space] > 1, suggesting "wear out" - more likely to fail as time goes by.
The following graph illustrates how the PDF varies with the shape parameter [alpha]:
[$../graphs/weibull.png]
While this graph illustrates how the PDF varies with the scale parameter [beta]:
[$../graphs/weibull2.png]
[h4 Related distributions]
When [alpha][space] = 3, the
[@http://en.wikipedia.org/wiki/Weibull_distribution Weibull distribution] appears similar to the
[@http://en.wikipedia.org/wiki/Normal_distribution normal distribution].
When [alpha][space] = 1, the Weibull distribution reduces to the
[@http://en.wikipedia.org/wiki/Exponential_distribution exponential distribution].
[h4 Member Functions]
weibull_distribution(RealType shape, RealType scale = 1);
Constructs a [@http://en.wikipedia.org/wiki/Weibull_distribution
Weibull distribution] with shape /shape/ and scale /scale/.
Requires that the /shape/ and /scale/ parameters are both greater than zero,
otherwise calls __domain_error.
RealType shape()const;
Returns the /shape/ parameter of this distribution.
RealType scale()const;
Returns the /scale/ parameter of this distribution.
[h4 Non-member Accessors]
All the [link math_toolkit.dist.dist_ref.nmp usual non-member accessor functions] that are generic to all
distributions are supported: __usual_accessors.
The domain of the random variable is \[0, [infin]\].
[h4 Accuracy]
The Weibull distribution is implemented in terms of the
standard library `log` and `exp` functions plus __expm1 and __log1p
and as such should have very low error rates.
[h4 Implementation]
In the following table [alpha][space] is the shape parameter of the distribution,
[beta][space] is it's scale parameter, /x/ is the random variate, /p/ is the probability
and /q = 1-p/.
[table
[[Function][Implementation Notes]]
[[pdf][Using the relation: pdf = [alpha][beta][super -[alpha] ]x[super [alpha] - 1] e[super -(x/beta)[super alpha]] ]]
[[cdf][Using the relation: p = -__expm1(-(x\/[beta])[super [alpha]]) ]]
[[cdf complement][Using the relation: q = e[super -(x\/[beta])[super [alpha]]] ]]
[[quantile][Using the relation: x = [beta] * (-__log1p(-p))[super 1\/[alpha]] ]]
[[quantile from the complement][Using the relation: x = [beta] * (-log(q))[super 1\/[alpha]] ]]
[[mean][[beta] * [Gamma](1 + 1\/[alpha]) ]]
[[variance][[beta][super 2]([Gamma](1 + 2\/[alpha]) - [Gamma][super 2](1 + 1\/[alpha])) ]]
[[mode][[beta](([alpha] - 1) \/ [alpha])[super 1\/[alpha]] ]]
[[skewness][Refer to [@http://mathworld.wolfram.com/WeibullDistribution.html Weisstein, Eric W. "Weibull Distribution." From MathWorld--A Wolfram Web Resource.] ]]
[[kurtosis][Refer to [@http://mathworld.wolfram.com/WeibullDistribution.html Weisstein, Eric W. "Weibull Distribution." From MathWorld--A Wolfram Web Resource.] ]]
[[kurtosis excess][Refer to [@http://mathworld.wolfram.com/WeibullDistribution.html Weisstein, Eric W. "Weibull Distribution." From MathWorld--A Wolfram Web Resource.] ]]
]
[h4 References]
* [@http://en.wikipedia.org/wiki/Weibull_distribution ]
* [@http://mathworld.wolfram.com/WeibullDistribution.html Weisstein, Eric W. "Weibull Distribution." From MathWorld--A Wolfram Web Resource.]
* [@http://www.itl.nist.gov/div898/handbook/eda/section3/eda3668.htm Weibull in NIST Exploratory Data Analysis]
[endsect][/section:weibull Weibull]
[/
Copyright 2006 John Maddock and Paul A. Bristow.
Distributed under the Boost Software License, Version 1.0.
(See accompanying file LICENSE_1_0.txt or copy at
http://www.boost.org/LICENSE_1_0.txt).
]
|
