-
Notifications
You must be signed in to change notification settings - Fork 0
/
spectral.py
133 lines (121 loc) · 5.32 KB
/
spectral.py
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
#!/usr/bin/env python
# -*- coding: UTF-8 -*-
import warnings
import itertools
from sklearn.cluster import SpectralClustering, spectral_clustering
# from sklearn.utils.validation import check_array
from sklearn.utils import check_random_state
from sklearn.metrics.pairwise import pairwise_kernels
import numpy as np
class ESpectralClustering(SpectralClustering):
def __init__(self, n_clusters=8, eigen_solver=None, random_state=None,
n_init=10, gamma=1., affinity='cosine', n_neighbors=10,
eigen_tol=0.0, assign_labels='kmeans', degree=3, coef0=1,
kernel_params=None, n_jobs=1, weight=np.array([]), merge='product'):
super(ESpectralClustering, self).__init__(n_clusters=n_clusters,
eigen_solver=eigen_solver,
random_state=random_state, n_init=n_init, gamma=gamma,
affinity=affinity,
n_neighbors=n_neighbors, eigen_tol=eigen_tol,
assign_labels=assign_labels, degree=degree,
coef0=coef0, kernel_params=kernel_params, n_jobs=n_jobs)
self.weight = weight
self.merge = merge
def calc_weight(self, w1, w2):
# self.merge = 'product' or 'min'
if self.merge == 'product':
weight = np.array([float(w1[i]) * float(w2[i])
for i in xrange(len(w1))])
return weight / sum(weight)
else:
raise
def fit(self, X, y=None):
'''
X is matrix((n_separate * n) * len(feature[0]))
w is matrix(n * n_separate)
'''
# X = check_array(X, accept_sparse=['csr', 'csc', 'coo'],
# dtype=np.float64)
if X.shape[0] == X.shape[1] and self.affinity != 'precomputed':
warnings.warn("The spectral clustering API has changed. ``fit``"
"set ``affinity=precomputed``.")
params = self.kernel_params
if params is None:
params = {}
if not callable(self.affinity):
params['gamma'] = self.gamma
params['degree'] = self.degree
params['coef0'] = self.coef0
# X = np.random.randint(1, 7, (10, 2))
# affity = 'cosine'
if len(self.weight) == 0:
self.affinity_matrix_ = pairwise_kernels(X,
metric=self.affinity, filter_params=True, **params)
else:
n_data = len(self.weight)
n_separate = len(self.weight[0])
print 'n_data: ', n_data
print 'se', n_separate
# zerosでよかった...
D = np.array([[[0.0000000 for ___ in range(len(X[0]))]
for __ in range(n_data)] for _ in range(n_separate)])
for i in xrange(n_data * n_separate):
D[i % n_separate, i / n_separate] = X[i]
mat_W = np.zeros([len(self.weight), len(self.weight),
len(self.weight[0])])
for i, j in itertools.product(range(len(self.weight)),
range(len(self.weight))):
print 'i:', self.weight[i]
print 'j:', self.weight[j]
mat_W[i, j] = self.calc_weight(self.weight[i], self.weight[j])
print 'W: ', mat_W
kernel_D = np.array([pairwise_kernels(mat, metric=self.affinity)
for mat in D])
sum_kernel_D = np.zeros((len(self.weight), len(self.weight)))
for i, j in itertools.product(range(len(self.weight)),
range(len(self.weight))):
temp = 0
for k in range(len(kernel_D)):
temp += kernel_D[k, i, j] * mat_W[i, j, k]
sum_kernel_D[i, j] = temp
self.affinity_matrix_ = sum_kernel_D
print 'result: ', self.affinity_matrix_
random_state = check_random_state(self.random_state)
print 'affinity mat: ', self.affinity_matrix_
self.labels_ = spectral_clustering(self.affinity_matrix_,
n_clusters=self.n_clusters,
eigen_solver=self.eigen_solver,
random_state=random_state,
n_init=self.n_init,
eigen_tol=self.eigen_tol,
assign_labels=self.assign_labels)
return self
if __name__ == '__main__':
np.random.seed(10)
data = np.random.randint(1, 7, (30, 2))
print('-' * 10 + 'sklearn' + '-' * 10)
I = SpectralClustering()
assign_labels = I.fit_predict(data)
print 'affinity', I.affinity
print(assign_labels)
np.random.seed(10)
data = np.random.randint(1, 7, (30, 2))
print('-' * 10 + 'sklearn' + '-' * 10)
I2 = SpectralClustering(affinity='cosine')
assign_labels = I2.fit_predict(data)
print 'affinity', I2.affinity
print(assign_labels)
print('-' * 10 + 'E' + '-' * 10)
np.random.seed(10)
E = ESpectralClustering()
e_assign_labels = E.fit_predict(data)
print 'affinity', E.affinity
print(e_assign_labels)
print('-' * 10 + 'E with weight' + '-' * 10)
np.random.seed(10)
weight = abs(np.random.randn(15, 2))
print data
E = ESpectralClustering(n_clusters=2, weight=weight)
e_assign_labels = E.fit_predict(data)
print 'affinity', E.affinity
print(e_assign_labels)