1 | #ifndef _theplu_statistics_pearson_ |
---|
2 | #define _theplu_statistics_pearson_ |
---|
3 | |
---|
4 | // $Id: Pearson.h 623 2006-09-05 02:13:12Z peter $ |
---|
5 | |
---|
6 | #include <c++_tools/statistics/Score.h> |
---|
7 | |
---|
8 | namespace theplu { |
---|
9 | namespace utility { |
---|
10 | class vector; |
---|
11 | } |
---|
12 | namespace classifier { |
---|
13 | class VectorAbstract; |
---|
14 | } |
---|
15 | namespace statistics { |
---|
16 | |
---|
17 | /// |
---|
18 | /// Class for calculating Pearson correlation. |
---|
19 | /// |
---|
20 | |
---|
21 | class Pearson : public Score |
---|
22 | { |
---|
23 | |
---|
24 | public: |
---|
25 | /// |
---|
26 | /// Default Constructor. |
---|
27 | /// |
---|
28 | Pearson(bool absolute=true); |
---|
29 | |
---|
30 | /// |
---|
31 | /// Destructor |
---|
32 | /// |
---|
33 | virtual ~Pearson(void) {}; |
---|
34 | |
---|
35 | |
---|
36 | /// |
---|
37 | /// \f$ \frac{\vert \sum_i(x_i-\bar{x})(y_i-\bar{y})\vert |
---|
38 | /// }{\sqrt{\sum_i (x_i-\bar{x})^2\sum_i (x_i-\bar{x})^2}} \f$. |
---|
39 | /// @return Pearson correlation, if absolute=true absolute value |
---|
40 | /// of Pearson is used. |
---|
41 | /// |
---|
42 | double score(const classifier::Target& target, |
---|
43 | const utility::vector& value); |
---|
44 | |
---|
45 | /// |
---|
46 | /// \f$ \frac{\vert \sum_iw^2_i(x_i-\bar{x})(y_i-\bar{y})\vert } |
---|
47 | /// {\sqrt{\sum_iw^2_i(x_i-\bar{x})^2\sum_iw^2_i(y_i-\bar{y})^2}} |
---|
48 | /// \f$, where \f$ m_x = \frac{\sum w_ix_i}{\sum w_i} \f$ and \f$ |
---|
49 | /// m_x = \frac{\sum w_ix_i}{\sum w_i} \f$. This expression is |
---|
50 | /// chosen to get a correlation equal to unity when \a x and \a y |
---|
51 | /// are equal. @return absolute value of weighted version of |
---|
52 | /// Pearson correlation. |
---|
53 | /// |
---|
54 | double score(const classifier::Target& target, |
---|
55 | const classifier::DataLookupWeighted1D& value); |
---|
56 | |
---|
57 | /// |
---|
58 | /// \f$ \frac{\vert \sum_iw^2_i(x_i-\bar{x})(y_i-\bar{y})\vert } |
---|
59 | /// {\sqrt{\sum_iw^2_i(x_i-\bar{x})^2\sum_iw^2_i(y_i-\bar{y})^2}} |
---|
60 | /// \f$, where \f$ m_x = \frac{\sum w_ix_i}{\sum w_i} \f$ and \f$ |
---|
61 | /// m_x = \frac{\sum w_ix_i}{\sum w_i} \f$. This expression is |
---|
62 | /// chosen to get a correlation equal to unity when \a x and \a y |
---|
63 | /// are equal. @return absolute value of weighted version of |
---|
64 | /// Pearson correlation. |
---|
65 | /// |
---|
66 | double score(const classifier::Target& target, |
---|
67 | const utility::vector& value, |
---|
68 | const utility::vector& weight); |
---|
69 | |
---|
70 | /// |
---|
71 | /// The p-value is the probability of getting a correlation as |
---|
72 | /// large as the observed value by random chance, when the true |
---|
73 | /// correlation is zero (and the data is Gaussian). Note that this |
---|
74 | /// function can only be used together with the unweighted |
---|
75 | /// score. @return two-sided p-value |
---|
76 | /// |
---|
77 | double p_value() const; |
---|
78 | |
---|
79 | private: |
---|
80 | double r_; |
---|
81 | int nof_samples_; |
---|
82 | |
---|
83 | |
---|
84 | // void centralize(utility::vector&, const utility::vector&); |
---|
85 | }; |
---|
86 | |
---|
87 | }} // of namespace statistics and namespace theplu |
---|
88 | |
---|
89 | #endif |
---|
90 | |
---|