-
-
Notifications
You must be signed in to change notification settings - Fork 1k
/
LibSVR_unittest.cc
142 lines (115 loc) · 3.88 KB
/
LibSVR_unittest.cc
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
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
/*
* This software is distributed under BSD 3-clause license (see LICENSE file).
*
* Authors: Heiko Strathmann, Evgeniy Andreev, Fernando Iglesias, Sanuj Sharma
*/
#include <shogun/kernel/GaussianKernel.h>
#include <shogun/labels/RegressionLabels.h>
#include <shogun/features/DenseFeatures.h>
#include <shogun/regression/svr/LibSVR.h>
#include <gtest/gtest.h>
using namespace shogun;
TEST(LibSVR,epsilon_svr_apply)
{
const float64_t rbf_width=1;
const float64_t svm_C=1;
const float64_t svm_eps=0.1;
/* create some easy regression data: 1d noisy sine wave */
index_t n=5;
SGMatrix<float64_t> feat_train(1, n);
SGMatrix<float64_t> feat_test(1, n);
SGVector<float64_t> lab_train(n);
/* a one dimensional quadratic function */
feat_train[0]=-2;
feat_train[1]=-1;
feat_train[2]=0;
feat_train[3]=1;
feat_train[4]=2;
lab_train[0]=4;
lab_train[1]=1;
lab_train[2]=0;
lab_train[3]=1;
lab_train[4]=4;
feat_test[0]=-2.2;
feat_test[1]=-1.1;
feat_test[2]=0.2;
feat_test[3]=1.3;
feat_test[4]=1.9;
/* shogun representation */
CRegressionLabels* labels_train=new CRegressionLabels(lab_train);
CDenseFeatures<float64_t>* features_train=new CDenseFeatures<float64_t>(
feat_train);
CDenseFeatures<float64_t>* features_test=new CDenseFeatures<float64_t>(
feat_test);
CGaussianKernel* kernel=new CGaussianKernel(rbf_width);
kernel->init(features_train, features_train);
LIBSVR_SOLVER_TYPE st=LIBSVR_EPSILON_SVR;
CLibSVR* svm=new CLibSVR(svm_C, svm_eps, kernel, labels_train, st);
svm->train();
/* predict */
CRegressionLabels* predicted_labels =
svm->apply(features_test)->as<CRegressionLabels>();
/* LibSVM regression comparison (with easy.py script) */
EXPECT_NEAR(predicted_labels->get_labels()[0], 2.44343, 1E-5);
EXPECT_NEAR(predicted_labels->get_labels()[1], 1.25466, 1E-5);
EXPECT_NEAR(predicted_labels->get_labels()[2], 0.313201, 1E-5);
EXPECT_NEAR(predicted_labels->get_labels()[3], 1.57767, 1E-5);
EXPECT_NEAR(predicted_labels->get_labels()[4], 2.34949, 1E-5);
EXPECT_NEAR(CMath::abs(svm->get_bias()), 1.60903, 1E-5);
EXPECT_EQ(svm->get_num_support_vectors(), 5);
/* clean up */
SG_UNREF(predicted_labels);
SG_UNREF(svm);
}
TEST(LibSVR,nu_svr_apply)
{
const float64_t rbf_width=1;
const float64_t svm_C=1;
const float64_t svm_nu=0.1;
/* create some easy regression data: 1d noisy sine wave */
index_t n=5;
SGMatrix<float64_t> feat_train(1, n);
SGMatrix<float64_t> feat_test(1, n);
SGVector<float64_t> lab_train(n);
/* a one dimensional quadratic function */
feat_train[0]=-2;
feat_train[1]=-1;
feat_train[2]=0;
feat_train[3]=1;
feat_train[4]=2;
lab_train[0]=4;
lab_train[1]=1;
lab_train[2]=0;
lab_train[3]=1;
lab_train[4]=4;
feat_test[0]=-2.2;
feat_test[1]=-1.1;
feat_test[2]=0.2;
feat_test[3]=1.3;
feat_test[4]=1.9;
/* shogun representation */
CRegressionLabels* labels_train=new CRegressionLabels(lab_train);
CDenseFeatures<float64_t>* features_train=new CDenseFeatures<float64_t>(
feat_train);
CDenseFeatures<float64_t>* features_test=new CDenseFeatures<float64_t>(
feat_test);
CGaussianKernel* kernel=new CGaussianKernel(rbf_width);
kernel->init(features_train, features_train);
LIBSVR_SOLVER_TYPE st=LIBSVR_NU_SVR;
CLibSVR* svm=new CLibSVR(svm_C, svm_nu, kernel, labels_train, st);
svm->train();
/* predict */
CRegressionLabels* predicted_labels =
svm->apply(features_test)->as<CRegressionLabels>();
/* LibSVM regression comparison (with easy.py script) */
EXPECT_NEAR(predicted_labels->get_labels()[0], 2.18062, 1E-5);
EXPECT_NEAR(predicted_labels->get_labels()[1], 2.04357, 1E-5);
EXPECT_NEAR(predicted_labels->get_labels()[2], 1.82819, 1E-5);
EXPECT_NEAR(predicted_labels->get_labels()[3], 2.09295, 1E-5);
EXPECT_NEAR(predicted_labels->get_labels()[4], 2.17949, 1E-5);
EXPECT_NEAR(CMath::abs(svm->get_bias()), 2.0625, 1E-5);
EXPECT_EQ(svm->get_num_support_vectors(), 3);
/* clean up */
SG_UNREF(predicted_labels);
SG_UNREF(svm);
}