-
Notifications
You must be signed in to change notification settings - Fork 89
/
kmeans.py
263 lines (245 loc) · 8.62 KB
/
kmeans.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
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
# k_means.py
# -*- coding: utf-8 -*-
# @Author: huzhu
# @Date: 2019-10-29 09:31:43
# @Last Modified by: huzhu
# @Last Modified time: 2019-11-14 21:14:20
import codecs
from numpy import *
import matplotlib.pyplot as plt
from mpl_toolkits.mplot3d import Axes3D
from sklearn.datasets import make_moons
import matplotlib.animation as animation
from sklearn.cluster import KMeans
def load_data(path):
"""
@brief Loads a data.
@param path The path
@return data set
"""
data_set = list()
with codecs.open(path) as f:
for line in f.readlines():
data = line.strip().split("\t")
flt_data = list(map(float, data))
data_set.append(flt_data)
return data_set
def dist_eucl(vecA, vecB):
"""
@brief the similarity function
@param vecA The vector a
@param vecB The vector b
@return the euclidean distance
"""
return sqrt(sum(power(vecA - vecB, 2)))
def rand_cent(data_mat, k):
"""
@brief select random centroid
@param data_mat The data matrix
@param k
@return centroids
"""
n = shape(data_mat)[1]
centroids = mat(zeros((k, n)))
for j in range(n):
minJ = min(data_mat[:,j])
rangeJ = float(max(data_mat[:,j]) - minJ)
centroids[:,j] = mat(minJ + rangeJ * random.rand(k,1))
return centroids
def kMeans(data_mat, k, dist = "dist_eucl", create_cent = "rand_cent"):
"""
@brief kMeans algorithm
@param data_mat The data matrix
@param k num of cluster
@param dist The distance funtion
@param create_cent The create centroid function
@return the cluster
"""
m = shape(data_mat)[0]
# 初始化点的簇
cluster_assment = mat(zeros((m, 2))) # 类别,距离
# 随机初始化聚类初始点
centroid = eval(create_cent)(data_mat, k)
cluster_changed = True
# 遍历每个点
while cluster_changed:
cluster_changed = False
for i in range(m):
min_index = -1
min_dist = inf
for j in range(k):
distance = eval(dist)(data_mat[i, :], centroid[j, :])
if distance < min_dist:
min_dist = distance
min_index = j
if cluster_assment[i, 0] != min_index:
cluster_changed = True
cluster_assment[i, :] = min_index, min_dist**2
# 计算簇中所有点的均值并重新将均值作为质心
for j in range(k):
per_data_set = data_mat[nonzero(cluster_assment[:,0].A == j)[0]]
centroid[j, :] = mean(per_data_set, axis = 0)
return centroid, cluster_assment
def plot_cluster(data_mat, cluster_assment, centroid):
"""
@brief plot cluster and centroid
@param data_mat The data matrix
@param cluster_assment The cluste assment
@param centroid The centroid
@return
"""
plt.figure(figsize=(15, 6), dpi=80)
plt.subplot(121)
plt.plot(data_mat[:, 0], data_mat[:, 1], 'o')
plt.title("source data", fontsize=15)
plt.subplot(122)
k = shape(centroid)[0]
colors = [plt.cm.get_cmap("Spectral")(each) for each in linspace(0, 1, k)]
for i, col in zip(range(k), colors):
per_data_set = data_mat[nonzero(cluster_assment[:,0].A == i)[0]]
plt.plot(per_data_set[:, 0], per_data_set[:, 1], 'o', markerfacecolor=tuple(col),
markeredgecolor='k', markersize=10)
for i in range(k):
plt.plot(centroid[:,0], centroid[:,1], '+', color = 'k', markersize=18)
plt.title("K-Means Cluster, k = 3", fontsize=15)
plt.show()
def plot_noncov():
"""
@brief 绘制非凸优化函数图像
@return { description_of_the_return_value }
"""
fig = plt.figure()
ax = fig.gca(projection='3d')
x1 = linspace(-2,2,100)
x2 = linspace(-2,2,100)
mu1 = array([1,1])
mu2 = array([-1,-1])
Z = zeros((len(x1), len(x2)))
for i in range(len(x1)):
for j in range(len(x2)):
itemx = x1[i]
itemy = x2[j]
z1 = dist_eucl(mu1, [itemx, itemy])
z2 = dist_eucl(mu2, [itemx, itemy])
Z[i,j] = min(z1,z2)
X1, X2 = meshgrid(x1, x2)
ax.plot_surface(X1, X2, Z, rstride=1, cstride=1, cmap='rainbow')
plt.show()
def test_diff_k():
plt.figure(figsize=(15, 4), dpi=80)
data_mat = mat(load_data("data/testSet2_kmeans.txt"))
centroid, cluster_assment = kMeans(data_mat, 2)
plt.subplot(131)
k = shape(centroid)[0]
colors = [plt.cm.Spectral(each) for each in linspace(0, 1, k)]
for i, col in zip(range(k), colors):
per_data_set = data_mat[nonzero(cluster_assment[:,0].A == i)[0]]
plt.plot(per_data_set[:, 0], per_data_set[:, 1], 'o', markerfacecolor=tuple(col),
markeredgecolor='k', markersize=10)
for i in range(k):
plt.plot(centroid[:,0], centroid[:,1], '+', color = 'k', markersize=18)
plt.title("K-Means Cluster, k = 2", fontsize=15)
centroid, cluster_assment = kMeans(data_mat, 3)
plt.subplot(132)
k = shape(centroid)[0]
colors = [plt.cm.Spectral(each) for each in linspace(0, 1, k)]
for i, col in zip(range(k), colors):
per_data_set = data_mat[nonzero(cluster_assment[:,0].A == i)[0]]
plt.plot(per_data_set[:, 0], per_data_set[:, 1], 'o', markerfacecolor=tuple(col),
markeredgecolor='k', markersize=10)
for i in range(k):
plt.plot(centroid[:,0], centroid[:,1], '+', color = 'k', markersize=18)
plt.title("K-Means Cluster, k = 3", fontsize=15)
centroid, cluster_assment = kMeans(data_mat, 4)
plt.subplot(133)
k = shape(centroid)[0]
colors = [plt.cm.Spectral(each) for each in linspace(0, 1, k)]
for i, col in zip(range(k), colors):
per_data_set = data_mat[nonzero(cluster_assment[:,0].A == i)[0]]
plt.plot(per_data_set[:, 0], per_data_set[:, 1], 'o', markerfacecolor=tuple(col),
markeredgecolor='k', markersize=10)
for i in range(k):
plt.plot(centroid[:,0], centroid[:,1], '+', color = 'k', markersize=18)
plt.title("K-Means Cluster, k = 4", fontsize=15)
plt.show()
def plot_fig(data_mat):
"""
@brief 绘制并保存gif图
@param data_mat The data matrix
@param k { parameter_description }
@return { description_of_the_return_value }
"""
centroid_list = list()
cluster_assment_list = list()
def sub_kMeans(data_mat, k, dist = "dist_eucl", create_cent = "rand_cent"):
m = shape(data_mat)[0]
# 初始化点的簇
cluster_assment = mat(zeros((m, 2))) # 类别,距离
# 随机初始化聚类初始点
centroid = eval(create_cent)(data_mat, k)
cluster_changed = True
# 遍历每个点
while cluster_changed:
centroid_list.append(array(centroid))
cluster_assment_list.append(array(cluster_assment))
cluster_changed = False
for i in range(m):
min_index = -1
min_dist = inf
for j in range(k):
distance = eval(dist)(data_mat[i, :], centroid[j, :])
if distance < min_dist:
min_dist = distance
min_index = j
if cluster_assment[i, 0] != min_index:
cluster_changed = True
cluster_assment[i, :] = min_index, min_dist**2
# 计算簇中所有点的均值并重新将均值作为质心
for j in range(k):
per_data_set = data_mat[nonzero(cluster_assment[:,0].A == j)[0]]
centroid[j, :] = mean(per_data_set, axis = 0)
return centroid_list,cluster_assment_list
centroid_list,cluster_assment_list = sub_kMeans(data_mat,4)
fig, ax = plt.subplots()
plt.scatter(data_mat[:, 0].flatten().A[0], data_mat[:, 1].flatten().A[0])
plt.title("K-Means Cluster Process", fontsize=15)
def update(i):
try:
ax.lines.pop()
except Exception:
pass
centroid = matrix(centroid_list[i])
cluster_assment = matrix(cluster_assment_list[i])
k = shape(centroid)[0]
colors = [plt.cm.Spectral(each) for each in linspace(0, 1, k)]
for i, col in zip(range(k), colors):
per_data_set = data_mat[nonzero(cluster_assment[:,0].A == i)[0]]
line, = plt.plot(per_data_set[:, 0], per_data_set[:, 1], 'o', markerfacecolor=tuple(col),markeredgecolor='k', markersize=10)
line, = plt.plot(centroid[:,0], centroid[:,1], '*', color = 'k', markersize=18)
return line,
anim = animation.FuncAnimation(fig, update, frames=len(centroid_list),interval=1000, repeat_delay=1000)
plt.show()
anim.save('test_animation.gif',writer='pillow')
def kmeans_lib():
data_mat = mat(load_data("data/testSet2_kmeans.txt"))
estimator = KMeans(n_clusters=3)#构造聚类器
estimator.fit(data_mat)#聚类
label_pred = estimator.labels_ #获取聚类标签
print(label_pred)
centroids = estimator.cluster_centers_ #获取聚类中心
inertia = estimator.inertia_ # 获取聚类准则的总和
plot_cluster(data_mat, mat(label_pred), centroids)
print(centroids)
print(inertia)
if __name__ == '__main__':
#data_mat = mat(load_data("data/testSet_kmeans.txt"))
data_mat = mat(load_data("data/testSet2_kmeans.txt"))
#data_mat,c = make_moons(n_samples=1000,noise=0.1)
centroid, cluster_assment = kMeans(data_mat, 3)
sse = sum(cluster_assment[:,1])
print("sse is ", sse)
plot_cluster(data_mat, cluster_assment, centroid)
#plot_fig(data_mat)
#plot_noncov()
#test_diff_k()
#kmeans_lib()