69 |
29 Apr 04 |
peter |
// $Id$ |
69 |
29 Apr 04 |
peter |
2 |
|
675 |
10 Oct 06 |
jari |
3 |
/* |
831 |
27 Mar 07 |
peter |
Copyright (C) 2004, 2005 Peter Johansson |
4359 |
23 Aug 23 |
peter |
Copyright (C) 2006 Jari Häkkinen, Peter Johansson |
4359 |
23 Aug 23 |
peter |
Copyright (C) 2007 Peter Johansson |
4359 |
23 Aug 23 |
peter |
Copyright (C) 2008 Jari Häkkinen, Peter Johansson |
4359 |
23 Aug 23 |
peter |
Copyright (C) 2011, 2012 Peter Johansson |
295 |
29 Apr 05 |
peter |
9 |
|
1437 |
25 Aug 08 |
peter |
This file is part of the yat library, http://dev.thep.lu.se/yat |
623 |
05 Sep 06 |
peter |
11 |
|
675 |
10 Oct 06 |
jari |
The yat library is free software; you can redistribute it and/or |
675 |
10 Oct 06 |
jari |
modify it under the terms of the GNU General Public License as |
1486 |
09 Sep 08 |
jari |
published by the Free Software Foundation; either version 3 of the |
675 |
10 Oct 06 |
jari |
License, or (at your option) any later version. |
675 |
10 Oct 06 |
jari |
16 |
|
675 |
10 Oct 06 |
jari |
The yat library is distributed in the hope that it will be useful, |
675 |
10 Oct 06 |
jari |
but WITHOUT ANY WARRANTY; without even the implied warranty of |
675 |
10 Oct 06 |
jari |
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU |
675 |
10 Oct 06 |
jari |
General Public License for more details. |
675 |
10 Oct 06 |
jari |
21 |
|
675 |
10 Oct 06 |
jari |
You should have received a copy of the GNU General Public License |
1487 |
10 Sep 08 |
jari |
along with yat. If not, see <http://www.gnu.org/licenses/>. |
675 |
10 Oct 06 |
jari |
24 |
*/ |
675 |
10 Oct 06 |
jari |
25 |
|
2881 |
18 Nov 12 |
peter |
26 |
#include <config.h> |
2881 |
18 Nov 12 |
peter |
27 |
|
2551 |
12 Aug 11 |
peter |
28 |
#include <iostream> |
2551 |
12 Aug 11 |
peter |
29 |
|
820 |
17 Mar 07 |
peter |
30 |
#include "AUC.h" |
675 |
10 Oct 06 |
jari |
31 |
#include "yat/classifier/DataLookupWeighted1D.h" |
747 |
11 Feb 07 |
peter |
32 |
#include "yat/classifier/Target.h" |
675 |
10 Oct 06 |
jari |
33 |
#include "yat/utility/stl_utility.h" |
1023 |
01 Feb 08 |
peter |
34 |
#include "yat/utility/VectorBase.h" |
675 |
10 Oct 06 |
jari |
35 |
|
779 |
05 Mar 07 |
peter |
36 |
#include <cassert> |
140 |
20 Aug 04 |
peter |
37 |
#include <cmath> |
112 |
07 Jul 04 |
peter |
38 |
#include <utility> |
779 |
05 Mar 07 |
peter |
39 |
#include <map> |
69 |
29 Apr 04 |
peter |
40 |
|
69 |
29 Apr 04 |
peter |
41 |
namespace theplu { |
680 |
11 Oct 06 |
jari |
42 |
namespace yat { |
4200 |
19 Aug 22 |
peter |
43 |
namespace statistics { |
69 |
29 Apr 04 |
peter |
44 |
|
820 |
17 Mar 07 |
peter |
45 |
AUC::AUC(bool absolute) |
779 |
05 Mar 07 |
peter |
46 |
: Score(absolute) |
779 |
05 Mar 07 |
peter |
47 |
{ |
779 |
05 Mar 07 |
peter |
48 |
} |
779 |
05 Mar 07 |
peter |
49 |
|
4200 |
19 Aug 22 |
peter |
50 |
double AUC::score(const classifier::Target& target, |
1023 |
01 Feb 08 |
peter |
51 |
const utility::VectorBase& value) const |
69 |
29 Apr 04 |
peter |
52 |
{ |
509 |
18 Feb 06 |
peter |
53 |
assert(target.size()==value.size()); |
779 |
05 Mar 07 |
peter |
// key data, pair<target, weight> |
779 |
05 Mar 07 |
peter |
55 |
std::multimap<double, std::pair<bool, double> > m; |
779 |
05 Mar 07 |
peter |
56 |
for (unsigned int i=0; i<target.size(); i++) |
4200 |
19 Aug 22 |
peter |
57 |
m.insert(std::make_pair(value(i), |
779 |
05 Mar 07 |
peter |
58 |
std::make_pair(target.binary(i),1.0))); |
4200 |
19 Aug 22 |
peter |
59 |
|
779 |
05 Mar 07 |
peter |
60 |
return score(m); |
779 |
05 Mar 07 |
peter |
61 |
} |
475 |
22 Dec 05 |
peter |
62 |
|
509 |
18 Feb 06 |
peter |
63 |
|
136 |
16 Aug 04 |
peter |
64 |
|
4200 |
19 Aug 22 |
peter |
65 |
double AUC::score(const classifier::Target& target, |
820 |
17 Mar 07 |
peter |
66 |
const classifier::DataLookupWeighted1D& value) const |
623 |
05 Sep 06 |
peter |
67 |
{ |
779 |
05 Mar 07 |
peter |
68 |
assert(target.size()==value.size()); |
779 |
05 Mar 07 |
peter |
// key data, pair<target, weight> |
779 |
05 Mar 07 |
peter |
70 |
std::multimap<double, std::pair<bool, double> > m; |
623 |
05 Sep 06 |
peter |
71 |
for (unsigned int i=0; i<target.size(); i++) |
623 |
05 Sep 06 |
peter |
72 |
if (value.weight(i)) |
4200 |
19 Aug 22 |
peter |
73 |
m.insert(std::make_pair(value.data(i), |
4200 |
19 Aug 22 |
peter |
74 |
std::make_pair(target.binary(i), |
779 |
05 Mar 07 |
peter |
75 |
value.weight(i)))); |
4200 |
19 Aug 22 |
peter |
76 |
|
779 |
05 Mar 07 |
peter |
77 |
return score(m); |
623 |
05 Sep 06 |
peter |
78 |
} |
623 |
05 Sep 06 |
peter |
79 |
|
623 |
05 Sep 06 |
peter |
80 |
|
4200 |
19 Aug 22 |
peter |
81 |
double AUC::score(const classifier::Target& target, |
4200 |
19 Aug 22 |
peter |
82 |
const utility::VectorBase& value, |
1023 |
01 Feb 08 |
peter |
83 |
const utility::VectorBase& weight) const |
103 |
15 Jun 04 |
peter |
84 |
{ |
779 |
05 Mar 07 |
peter |
85 |
assert(target.size()==value.size()); |
779 |
05 Mar 07 |
peter |
86 |
assert(target.size()==weight.size()); |
779 |
05 Mar 07 |
peter |
// key data, pair<target, weight> |
779 |
05 Mar 07 |
peter |
88 |
std::multimap<double, std::pair<bool, double> > m; |
475 |
22 Dec 05 |
peter |
89 |
for (unsigned int i=0; i<target.size(); i++) |
488 |
04 Jan 06 |
peter |
90 |
if (weight(i)) |
4200 |
19 Aug 22 |
peter |
91 |
m.insert(std::make_pair(value(i), |
779 |
05 Mar 07 |
peter |
92 |
std::make_pair(target.binary(i), weight(i)))); |
4200 |
19 Aug 22 |
peter |
93 |
|
779 |
05 Mar 07 |
peter |
94 |
return score(m); |
779 |
05 Mar 07 |
peter |
95 |
} |
488 |
04 Jan 06 |
peter |
96 |
|
475 |
22 Dec 05 |
peter |
97 |
|
820 |
17 Mar 07 |
peter |
98 |
double AUC::score(const MultiMap& m) const |
779 |
05 Mar 07 |
peter |
99 |
{ |
778 |
04 Mar 07 |
peter |
100 |
double area=0; |
779 |
05 Mar 07 |
peter |
101 |
double cumsum_pos_w=0; |
779 |
05 Mar 07 |
peter |
102 |
double cumsum_neg_w=0; |
779 |
05 Mar 07 |
peter |
103 |
typedef MultiMap::const_iterator iter; |
475 |
22 Dec 05 |
peter |
104 |
|
2551 |
12 Aug 11 |
peter |
105 |
iter first = m.begin(); |
2551 |
12 Aug 11 |
peter |
106 |
while (first!=m.end()) { |
2551 |
12 Aug 11 |
peter |
107 |
double local_cumsum_pos_w=0; |
2551 |
12 Aug 11 |
peter |
108 |
double local_cumsum_neg_w=0; |
2551 |
12 Aug 11 |
peter |
109 |
iter last = first; |
2551 |
12 Aug 11 |
peter |
110 |
while (last!=m.end() && first->first==last->first) { |
2551 |
12 Aug 11 |
peter |
111 |
if (last->second.first) |
2551 |
12 Aug 11 |
peter |
112 |
local_cumsum_pos_w += last->second.second; |
2551 |
12 Aug 11 |
peter |
113 |
else |
2551 |
12 Aug 11 |
peter |
114 |
local_cumsum_neg_w += last->second.second; |
2551 |
12 Aug 11 |
peter |
115 |
++last; |
4200 |
19 Aug 22 |
peter |
116 |
} |
2551 |
12 Aug 11 |
peter |
117 |
area += local_cumsum_pos_w * ( cumsum_neg_w + 0.5*local_cumsum_neg_w ); |
2551 |
12 Aug 11 |
peter |
118 |
cumsum_pos_w += local_cumsum_pos_w; |
2551 |
12 Aug 11 |
peter |
119 |
cumsum_neg_w += local_cumsum_neg_w; |
2551 |
12 Aug 11 |
peter |
120 |
first = last; |
998 |
13 Dec 07 |
peter |
121 |
} |
779 |
05 Mar 07 |
peter |
// max area is cumsum_neg_w * cumsum_pos_w |
779 |
05 Mar 07 |
peter |
123 |
area/=(cumsum_neg_w*cumsum_pos_w); |
4200 |
19 Aug 22 |
peter |
124 |
|
778 |
04 Mar 07 |
peter |
125 |
if (area<0.5 && absolute_) |
778 |
04 Mar 07 |
peter |
126 |
return 1.0-area; |
778 |
04 Mar 07 |
peter |
127 |
return area; |
779 |
05 Mar 07 |
peter |
128 |
} |
103 |
15 Jun 04 |
peter |
129 |
|
683 |
11 Oct 06 |
jari |
130 |
}}} // of namespace statistics, yat, and theplu |