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
|
from ctypes import POINTER, c_double, c_int
from dataset import LibSvmPrecomputedDataSet
from kernel import *
from predict import *
import libsvm
__all__ = [
'LibSvmModel'
]
class LibSvmModel:
def __init__(self, kernel,
tolerance=0.001, shrinking=True, cache_size=40):
"""
Parameters:
- `kernel`: XXX
- `tolerance`: tolerance of termination criterion
- `shrinking`: whether to use the shrinking heuristics
- `cache_size` kernel evaluation cache size (MB)
"""
self.kernel = kernel
self.tolerance = tolerance
self.shrinking = shrinking
self.cache_size = cache_size
# kernel parameters
param = libsvm.svm_parameter()
param.kernel_type = kernel.kernel_type
param.degree = getattr(kernel, 'degree', 0)
param.gamma = getattr(kernel, 'gamma', 0.0)
param.coef0 = getattr(kernel, 'coef0', 0.0)
# other parameters
param.eps = tolerance
param.shrinking = shrinking
param.cache_size = cache_size
# defaults for optional parameters
param.nr_weight = 0
param.weight = POINTER(c_double)()
param.weight_label = POINTER(c_int)()
param.probability = False
self.param = param
def fit(self, dataset, PredictorType=LibSvmPredictor):
if self.kernel.kernel_type == libsvm.PRECOMPUTED and \
not isinstance(dataset, LibSvmPrecomputedDataSet):
raise ValueError, 'kernel requires a precomputed dataset'
problem = dataset._create_svm_problem()
dataset._update_svm_parameter(self.param)
self._check_problem_param(problem, self.param)
model = libsvm.svm_train(problem, self.param)
return self.ResultsType(model, dataset, self.kernel, PredictorType)
def _check_problem_param(self, problem, param):
error_msg = libsvm.svm_check_parameter(problem, param)
if error_msg:
raise ValueError, error_msg
|
