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
143
144
145
146
147
|
import inspect
from ctypes import *
import distutils.sysconfig
import numpy as N
__all__ = [
'svm_node_dtype'
]
# load libsvm library from the shared library/DLL using the file name
# extension for Python extensions, because we are currently using
# distutils to build the shared library/DLL as a Python extension
so_ext = distutils.sysconfig.get_config_vars('SO')[0]
_libsvm = N.ctypeslib.load_library('libsvm_%s' % so_ext, __file__)
svm_node_dtype = \
N.dtype({'names' : ['index', 'value'],
'formats' : [N.intc, N.float64]},
align=True)
# svm types
C_SVC = 0
NU_SVC = 1
ONE_CLASS = 2
EPSILON_SVR = 3
NU_SVR = 4
# kernel types
LINEAR = 0
POLY = 1
RBF = 2
SIGMOID = 3
PRECOMPUTED = 4
class svm_node(Structure):
_fields_ = [
('index', c_int),
('value', c_double)
]
class svm_parameter(Structure):
_fields_ = [
('svm_type', c_int),
('kernel_type', c_int),
# degree in polynomial kernel function
('degree', c_int),
# gamma in kernel function
('gamma', c_double),
# coef0 in kernel function
('coef0', c_double),
# kernel evaluation cache memory size in MB
('cache_size', c_double),
# tolerance of termination criterion
('eps', c_double),
# parameter C of C-SVC, epsilon-SVR and nu-SVR
('C', c_double),
('nr_weight', c_int),
('weight_label', POINTER(c_int)),
# parameter weight of C-SVC and nu-SVC
# actual weight of class is this weight * C
('weight', POINTER(c_double)),
# parameter nu of nu-SVC, one-class SVM and nu-SVR
('nu', c_double),
# epsilon in loss function of epsilon-SVR
('p', c_double),
# whether to use the shrinking heuristics
('shrinking', c_int),
# whether to train a SVC or SVR model for probability estimates
('probability', c_int)
]
class svm_problem(Structure):
_fields_ = [
# length (number of elements)
('l', c_int),
# label for classification or response variable for regression
('y', POINTER(c_double)),
('x', POINTER(POINTER(svm_node)))
]
class svm_model(Structure):
_fields_ = [
# parameters used to train the model
('param', svm_parameter),
('nr_class', c_int),
# el
('l', c_int),
# support vectors (length el)
('SV', POINTER(POINTER(svm_node))),
# length nr_class-1, each length el
('sv_coef', POINTER(POINTER(c_double))),
# length nr_class*(nr_class-1)/2
('rho', POINTER(c_double)),
# length nr_class*(nr_class-1)/2 for classification
# length 1 for regression
('probA', POINTER(c_double)),
# length nr_class*(nr_class-1)/2 for classification
('probB', POINTER(c_double)),
# support vectors labels for classifcation (length nr_class)
('labels', POINTER(c_int)),
# length nr_class; nSV[0] + nSV[1] + ... + nSV[n-1] = el
('nSV', POINTER(c_int)),
# whether svm_destroy_model should free support vectors
('free_sv', c_int)
]
libsvm_api = {
'svm_check_parameter' :
(c_char_p, [POINTER(svm_problem), POINTER(svm_parameter)]),
'svm_train' :
(POINTER(svm_model), [POINTER(svm_problem), POINTER(svm_parameter)]),
'svm_check_probability_model' :
(None, [POINTER(svm_model)]),
'svm_predict' :
(c_double, [POINTER(svm_model), POINTER(svm_node)]),
'svm_predict_values' :
(None, [POINTER(svm_model), POINTER(svm_node), POINTER(c_double)]),
'svm_predict_probability' :
(c_double, [POINTER(svm_model), POINTER(svm_node), POINTER(c_double)]),
'svm_get_svr_probability' :
(c_double, [POINTER(svm_model)]),
'svm_cross_validation' :
(None,
[POINTER(svm_problem), POINTER(svm_parameter), c_int,
POINTER(c_double)]),
'svm_destroy_model' :
(None, [POINTER(svm_model)])
}
for f, (restype, argtypes) in libsvm_api.iteritems():
func = getattr(_libsvm, f)
func.restype = restype
func.argtypes = argtypes
inspect.currentframe().f_locals[f] = func
def create_svm_problem(data):
problem = svm_problem()
problem.l = len(data)
y = (c_double*problem.l)()
x = (POINTER(svm_node)*problem.l)()
for i, (yi, xi) in enumerate(data):
y[i] = yi
x[i] = xi.ctypes.data_as(POINTER(svm_node))
problem.x = x
problem.y = y
return problem
|
