1 | #ifndef _theplu_yat_statistics_tscore_ |
---|
2 | #define _theplu_yat_statistics_tscore_ |
---|
3 | |
---|
4 | // $Id: tScore.h 683 2006-10-11 22:20:36Z jari $ |
---|
5 | |
---|
6 | /* |
---|
7 | Copyright (C) The authors contributing to this file. |
---|
8 | |
---|
9 | This file is part of the yat library, http://lev.thep.lu.se/trac/yat |
---|
10 | |
---|
11 | The yat library is free software; you can redistribute it and/or |
---|
12 | modify it under the terms of the GNU General Public License as |
---|
13 | published by the Free Software Foundation; either version 2 of the |
---|
14 | License, or (at your option) any later version. |
---|
15 | |
---|
16 | The yat library is distributed in the hope that it will be useful, |
---|
17 | but WITHOUT ANY WARRANTY; without even the implied warranty of |
---|
18 | MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU |
---|
19 | General Public License for more details. |
---|
20 | |
---|
21 | You should have received a copy of the GNU General Public License |
---|
22 | along with this program; if not, write to the Free Software |
---|
23 | Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA |
---|
24 | 02111-1307, USA. |
---|
25 | */ |
---|
26 | |
---|
27 | #include "Score.h" |
---|
28 | |
---|
29 | #include <gsl/gsl_cdf.h> |
---|
30 | |
---|
31 | namespace theplu { |
---|
32 | namespace yat { |
---|
33 | namespace utility { |
---|
34 | class vector; |
---|
35 | } |
---|
36 | namespace statistics { |
---|
37 | |
---|
38 | /// |
---|
39 | /// Class for Fisher's t-test. |
---|
40 | /// |
---|
41 | /// See <a href="http://en.wikipedia.org/wiki/Student's_t-test"> |
---|
42 | /// http://en.wikipedia.org/wiki/Student's_t-test</a> for more |
---|
43 | /// details on the t-test. |
---|
44 | /// |
---|
45 | class tScore : public Score |
---|
46 | { |
---|
47 | |
---|
48 | public: |
---|
49 | /// |
---|
50 | /// Default Constructor. |
---|
51 | /// |
---|
52 | tScore(bool absolute=true); |
---|
53 | |
---|
54 | |
---|
55 | /** |
---|
56 | Calculates the value of t-score, i.e. the ratio between |
---|
57 | difference in mean and standard deviation of this |
---|
58 | difference. \f$ t = \frac{ m_x - m_y } |
---|
59 | {s\sqrt{\frac{1}{n_x}+\frac{1}{n_y}}} \f$ where \f$ m \f$ is the |
---|
60 | mean, \f$ n \f$ is the number of data points and \f$ s^2 = |
---|
61 | \frac{ \sum_i (x_i-m_x)^2 + \sum_i (y_i-m_y)^2 }{ n_x + n_y - |
---|
62 | 2 } \f$ |
---|
63 | |
---|
64 | @return t-score. If absolute=true absolute value of t-score |
---|
65 | is returned |
---|
66 | */ |
---|
67 | double score(const classifier::Target& target, |
---|
68 | const utility::vector& value); |
---|
69 | |
---|
70 | /** |
---|
71 | Calculates the weighted t-score, i.e. the ratio between |
---|
72 | difference in mean and standard deviation of this |
---|
73 | difference. \f$ t = \frac{ m_x - m_y }{ |
---|
74 | s\sqrt{\frac{1}{n_x}+\frac{1}{n_y}}} \f$ where \f$ m \f$ is the |
---|
75 | weighted mean, n is the weighted version of number of data |
---|
76 | points \f$ \frac{\left(\sum w_i\right)^2}{\sum w_i^2} \f$, and |
---|
77 | \f$ s^2 \f$ is an estimation of the variance \f$ s^2 = \frac{ |
---|
78 | \sum_i w_i(x_i-m_x)^2 + \sum_i w_i(y_i-m_y)^2 }{ n_x + n_y - 2 |
---|
79 | } \f$. See AveragerWeighted for details. |
---|
80 | |
---|
81 | @return t-score. If absolute=true absolute value of t-score |
---|
82 | is returned |
---|
83 | */ |
---|
84 | double score(const classifier::Target& target, |
---|
85 | const classifier::DataLookupWeighted1D& value); |
---|
86 | |
---|
87 | /// |
---|
88 | /// Calculates the weighted t-score, i.e. the ratio between |
---|
89 | /// difference in mean and standard deviation of this |
---|
90 | /// difference. \f$ t = \frac{ m_x - m_y }{ |
---|
91 | /// \frac{s2}{n_x}+\frac{s2}{n_y}} \f$ where \f$ m \f$ is the |
---|
92 | /// weighted mean, n is the weighted version of number of data |
---|
93 | /// points and \f$ s2 \f$ is an estimation of the variance \f$ s^2 |
---|
94 | /// = \frac{ \sum_i w_i(x_i-m_x)^2 + \sum_i w_i(y_i-m_y)^2 }{ n_x |
---|
95 | /// + n_y - 2 } \f$. See AveragerWeighted for details. |
---|
96 | /// |
---|
97 | /// @return t-score if absolute=true absolute value of t-score |
---|
98 | /// is returned |
---|
99 | /// |
---|
100 | double score(const classifier::Target& target, |
---|
101 | const utility::vector& value, |
---|
102 | const utility::vector& weight); |
---|
103 | |
---|
104 | /// |
---|
105 | /// Calculates the p-value, i.e. the probability of observing a |
---|
106 | /// t-score equally or larger if the null hypothesis is true. If P |
---|
107 | /// is near zero, this casts doubt on this hypothesis. The null |
---|
108 | /// hypothesis is that the means of the two distributions are |
---|
109 | /// equal. Assumtions for this test is that the two distributions |
---|
110 | /// are normal distributions with equal variance. The latter |
---|
111 | /// assumtion is dropped in Welch's t-test. |
---|
112 | /// |
---|
113 | /// @return the one-sided p-value( if absolute=true is used |
---|
114 | /// the two-sided p-value) |
---|
115 | /// |
---|
116 | double p_value() const; |
---|
117 | |
---|
118 | private: |
---|
119 | double t_; |
---|
120 | double dof_; |
---|
121 | |
---|
122 | }; |
---|
123 | |
---|
124 | }}} // of namespace statistics, yat, and theplu |
---|
125 | |
---|
126 | #endif |
---|