-
Notifications
You must be signed in to change notification settings - Fork 0
/
spin.py
173 lines (140 loc) · 4.33 KB
/
spin.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
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
# -*- coding: utf-8 -*-
import numpy as np
import matplotlib.pyplot as plt
#from scipy.spatial.distance import pdist
#from scipy.spatial.distance import squareform
#from scipy.stats import scoreatpercentile
def matrix_plot(dist_matrix, y=None, labels=False):
"""
Parameters
__________
dist_matrix: Distance matrix
y: set of labels
labels: boolean, if True show labels names
Returns
-------
plot the distance matrix dist_matrix
"""
fig = plt.figure(figsize=(5.1, 5))
ax1 = fig.add_axes([0, .05, 1, 1])
palette = plt.cm.jet
#palette.set_under('b')
#palette.set_over('r')
ax1.matshow(dist_matrix,
cmap=palette,
#vmin=scoreatpercentile(X,2),
#vmax=scoreatpercentile(X,98),
aspect='auto')
ax1.tick_params(length=0)
if labels:
ax1.xaxis.tick_top()
ax1.set_xticklabels(y, rotation='vertical')
ax1.set_yticklabels(y)
ax1.set_xticks(np.arange(len(dist_matrix)) + 0.55, minor=False)
ax1.set_yticks(np.arange(len(dist_matrix)) + 0.5, minor=False)
else:
ax1.set_xticks([])
ax1.set_yticks([])
if y is not None:
ax2 = fig.add_axes([0, 0, 1, .04])
cmap1 = plt.cm.get_cmap("RdYlBu", len(set(y)))
#yl = np.array([i[0] for i in enumerate(y)])
yl = np.array(y)
set_y = list(set(y))
for i in enumerate(set_y):
yl[yl == i[1]] = i[0]
yl = yl.astype(int)
sl = list(set(yl))
yl = yl.reshape(1, len(y))
im2 = ax2.matshow(yl,
cmap=cmap1,
aspect='auto')
ax2.set_xticks([])
ax2.set_yticks([])
ax2.tick_params(length=0)
cax = fig.add_axes([1.04, 0.0, 0.05, len(set_y) * 0.05])
cbar = fig.colorbar(im2, cax=cax, ticks=sl)
#cbar.ax.set_yticks(sl, minor=False)
cbar.ax.set_yticklabels(set_y)
def neighborhood_sort(dist_matrix, sigma, weight_matrix=None):
"""
Parameters
__________
dist_matrix: np.array([n,n])
sigma: float
weight_matrix: None or np.array([n,n])
Returns
-------
sorted_ind: np.array([n]) of integers
sort_score: np.array([n]) of floats
erg: float
"""
mat_size = len(dist_matrix)
if sigma < 2:
if sigma < 0:
print "Erro"
sigma = 1
if weight_matrix is None:
weight_matrix = make_weight_matrix(sigma=sigma, n=len(dist_matrix))
mismatch = np.dot(dist_matrix, weight_matrix)
#encontra os indices e os valores minimos por linha
idx_m = np.argmin(mismatch, axis=1)
val_m = mismatch[np.arange(mat_size), idx_m]
mx = max(val_m)
#print 0.1*np.sign((mat_size/2. - idx_m + 1.))*val_m/mx
sort_score = (idx_m + 1.
- 0.1 * np.sign((mat_size / 2. - idx_m + 1.)) * val_m / mx)
sorted_ind = np.argsort(sort_score)
#print np.argsort(idx_m)
#print sorted_ind
erg = np.trace(np.dot(dist_matrix[np.meshgrid(sorted_ind, sorted_ind)],
weight_matrix))
return sorted_ind, sort_score, mismatch, erg
def make_weight_matrix(sigma, n):
"""
Parameters
__________
sigma: float
n: integer
Returns
-------
weight_matrix: np.array([n.n])
"""
[i, j] = np.meshgrid(range(n), range(n))
w = np.exp(- (i - j) ** 2. / sigma / n)
for i in range(10):
w = w / np.reshape(np.sum(w, axis=1), (n, 1))
w = w / np.sum(w, axis=0)
w = 0.5 * (w + w.T)
return w
def spin(dist_matrix, y, shuffle=False):
"""
Parameters
__________
dist_matrix: np.array([n,n])
y: np.array([n])
shffle: boolean, if True distance_matrix will be shuffled
Returns
-------
dist_matrix: np.array([n,n])
y: np.array([n])
"""
si = range(len(dist_matrix))
if shuffle:
np.random.shuffle(si)
sigma = 2 ** 5.
for j in range(6):
w = make_weight_matrix(sigma, len(dist_matrix))
out = []
for i in range(20):
y = y[si]
dist_matrix = dist_matrix[np.meshgrid(si, si)]
si, se, mm, e = neighborhood_sort(dist_matrix, sigma, w)
if e in out:
break
else:
out = out + [e]
#print sigma, i, e
sigma = sigma / 2.0
#print e
return dist_matrix, y