-
Notifications
You must be signed in to change notification settings - Fork 0
/
clustering.py
250 lines (213 loc) · 7.89 KB
/
clustering.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
__author__ = 'k148582'
import feature_extraction
import pymongo, pymongo_utill
import numpy as np
from time import time
import numpy as np
import matplotlib.pyplot as plt
from sklearn import metrics
from sklearn.cluster import KMeans, SpectralClustering
from sklearn.preprocessing import scale
from sklearn.metrics.pairwise import cosine_similarity, rbf_kernel, pairwise_kernels
from auto_spectral_clustering.autosp import predict_k
def bench_k_means(estimator, name, data):
t0 = time()
estimator.fit(data)
d_size = len(data)
print('% 9s %.2fs %i %.3f'
% (name, (time() - t0), estimator.inertia_,
metrics.silhouette_score(data, estimator.labels_,
metric='euclidean',
sample_size=d_size)))
def bench_spectral_clustering(name, data):
t0 = time()
print(data[0])
print(data[1])
data = scale(data)
d_size = len(data)
affinity_matrix = pairwise_kernels(data, data, metric='rbf')
print(type(affinity_matrix))
k = predict_k(affinity_matrix)
print(k)
sc = SpectralClustering(n_clusters=k,
affinity='precomputed',
assign_labels="kmeans").fit(affinity_matrix)
labels_pred = sc.labels_
print('% 9s %.2fs %.3f'
% (name, (time() - t0),
metrics.silhouette_score(data, labels_pred,
metric='cosine',)))
def KmeansForAgeEst(db, where, users, n_clusters):
X = []
map = []
cor_k = []
for at in where:
_users = [users[i] for i in at]
X.append(pymongo_utill.toTimeFreq(db, _users))
for i, x in enumerate(X):
km = KMeans(init='k-means++', n_clusters=n_clusters, n_init=10)
km.fit(x)
map = [i]*len(x)
cor_k += [tmp+(i*n_clusters) for tmp in km.predict(x)]
return cor_k, map
def KmeansForAgeEst2(db, where, users, n_clusters):
X = []
X_users = []
centers = []
est = []
est_v = []
for at in where:
_users = [users[i] for i in at]
X.append(pymongo_utill.toTimeFreq(db, _users))
X_users.append(_users)
for c, x in enumerate(X):
km = KMeans(init='k-means++', n_clusters=n_clusters, n_init=10)
km.fit(x)
centers.append(km.cluster_centers_)
max_0 = 0
max_1 = 0
est_0_v = ""
est_1_v = ""
for i, u in enumerate(x):
sim = pairwise_kernels(km.cluster_centers_, u, metric="cosine")
if max_0 < sim[0]:
est_0 = X_users[c][i]
max_0 = sim[0]
est_0_v = u
if max_1 < sim[1]:
est_1 = X_users[c][i]
max_1 = sim[1]
est_1_v = u
est.append((est_0, est_1))
est_v.append((est_0_v, est_1_v))
return centers
def reternOptimalK(X, init_k, last_k):
highest_score = 0
optimalK = 0
for k in range(init_k, last_k):
km = KMeans(n_clusters=k, init='k-means++', n_init=10).fit(X)
label_predic = km.labels_
score = metrics.silhouette_score(X, label_predic, metric='cosine')
if highest_score < score:
highest_score = score
optimalK = k
return optimalK
if __name__ == '__main__':
conn = pymongo_utill.getConnectionToMongoDB()
db = conn['TwitterInsert2']
feature_vecs, labels, screen_names = pymongo_utill.byTimeFreq(db=db, sample=100)
print(len(feature_vecs))
#screen_names, labels = pymongo_utill.loadUsers(db, sample=1254)
"""
where = []
for threshold in [0,1]:
where.append(np.argwhere(labels == threshold))
centers, est, est_v = KmeansForAgeEst2(db, where, screen_names, 2)
num_fig = 0
"""
"""
for i, center in enumerate(centers):
for ctr in center:
ti = range(24)
plt.figure(num_fig)
plt.ylim([0, 0.5])
plt.xlim([0, 1, 23])
plt.xlabel("fraction")
plt.ylabel("time-posted")
if i == 0:
plt.title("under_30")
elif i == 1:
plt.title("over_30")
plt.subplot(211)
plt.plot(ti, ctr[:24])
plt.subplot(212)
plt.plot(ti, ctr[24:])
num_fig += 1
"""
"""
for i, ctr in enumerate(centers):
x = range(24)
f, axarr = plt.subplots(1, 2, sharey=True)
axarr[0].set_xticks(range(0, 24))
axarr[0].plot(x, ctr[0][:24], label="WeekDay")
axarr[0].plot(x, ctr[0][24:], '--', label="WeekEnd")
axarr[0].set_title("center1")
axarr[0].legend(loc="upper left")
axarr[1].set_xticks(range(0, 24))
axarr[1].plot(x, ctr[1][:24], label="WeekDay")
axarr[1].plot(x, ctr[1][24::], '--', label="WeekEnd")
axarr[1].set_title("center2")
axarr[1].legend(loc="upper left")
if i/2 == 0:
f.text(0.5, 0.96, "Under 30", ha='center')
else:
f.text(0.5, 0.96, "Over 30", ha='center')
f.text(0.5, 0.04, "time-posted", ha='center')
f.text(0.04, 0.5, "fraction", va='center', rotation='vertical')
for i, mst in enumerate(est_v):
x = range(24)
f, axarr = plt.subplots(1, 2, sharey=True)
axarr[0].set_xticks(range(0, 24))
axarr[0].plot(x, mst[0][:24], label="WeekDay")
axarr[0].plot(x, mst[0][24:], '--', label="WeekEnd")
axarr[0].set_title("most similar to center1")
axarr[0].legend()
axarr[1].set_xticks(range(0, 24))
axarr[1].plot(x, mst[1][:24], label="WeekDay")
axarr[1].plot(x, mst[1][24:], label="WeekEnd")
axarr[1].set_title("most similar to center2")
axarr[1].legend()
if i/2 == 0:
f.text(0.5, 0.96, "Under 30", ha='center')
else:
f.text(0.5, 0.96, "Over 30", ha='center')
f.text(0.5, 0.04, "time-posted", ha='center')
f.text(0.04, 0.5, "fraction", va='center', rotation='vertical')
for i, m in enumerate(est):
if i == 0:
classname = "under30"
else:
classname = 'over30'
for k, us in enumerate(m):
print("Most Similar User to cluster center%s of %s:%s" % (k+1, classname, us))
plt.show()
"""
print(reternOptimalK(feature_vecs[:50], 2, 6))
#number_of_blobs = 3
#data, labels = datasets.make_blobs(n_samples=number_of_blobs*10, centers=number_of_blobs, )
#bench_spectral_clustering(name='kmeans',data=feature_vecs)
"""
age = {'A':(19,13)}
label, age_range = age.items()[0]
print(age_range)
label = ord(label)%65
upper_range, lower_range = age_range
users_list = [x['user'] for x in db['twitter_user'].find({"$and":[{"age":{"$lte":upper_range}},{"age":{"$gte":lower_range}}]}) \
if x['user'] in db.collection_names()]
print(len(users_list))
for user in users_list:
tweets = [post for post in db[user].find()]
users.append(tweets)
conn.disconnect()
vectorizer = feature_extraction.WordVectorizer()
usr_vec = vectorizer.fit_transform(users)
np.random.seed(42)
data = scale(usr_vec)
n_clusters = 3
sample_size = 200
print("n_clusters: %d" % n_clusters)
print(79 * '_')
print('% 9s' % 'init'
' time inertia homo compl v-meas ARI AMI silhouette')
bench_k_means(KMeans(init='k-means++', n_clusters=n_clusters, n_init=10),
name="k-means++", data=data)
bench_k_means(KMeans(init='random', n_clusters=n_clusters, n_init=10),
name="random", data=data)
# in this case the seeding of the centers is deterministic, hence we run the
# kmeans algorithm only once with n_init=1
pca = PCA(n_components=n_clusters).fit(data)
bench_k_means(KMeans(init=pca.components_, n_clusters=n_clusters, n_init=1),
name="PCA-based",
data=data)
print(79 * '_')
"""