-
Notifications
You must be signed in to change notification settings - Fork 1
/
Copy pathexp_clustering_agglomerative_print_titles.py
125 lines (83 loc) · 3.86 KB
/
exp_clustering_agglomerative_print_titles.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
from text.bag_of_words import bow_from_news
from exp_corpus_loader import load_cleaned_news
from exp_corpus_loader import get_corpus_bow
import os
import numpy as np
import pandas as pd
import pickle
from sklearn.cluster import AgglomerativeClustering
from sklearn.feature_selection import SelectKBest
from sklearn.feature_selection import chi2
from MulticoreTSNE import MulticoreTSNE as TSNE
def load(filename):
with open(filename, "rb") as fp:
return pickle.load(fp)
def save(obj, filename):
with open(filename, "wb") as fp:
pickle.dump(obj, fp, protocol=4)
news = load_cleaned_news()
corpus, labels = get_corpus_bow()
ground_truth_file = 'data/vectors_ground_truth.bin'
print("Computing ground truth...")
if os.path.isfile(ground_truth_file):
doc_bow_2d = load(ground_truth_file)
else:
doc_bow = bow_from_news(corpus,
filename=None,
normalize_words=False)
sel_doc_bow = SelectKBest(chi2, k=5000).fit_transform(doc_bow, labels)
tsne = TSNE(n_components=2, n_jobs=4, random_state=1)
doc_bow_2d = tsne.fit_transform(sel_doc_bow)
save(doc_bow_2d, ground_truth_file)
idx_filter = np.where(labels != 'Unclassified')
ids = np.array([n['id'] for n in news])
titles = np.array([n['title'] for n in news])
print("Computing ground truth... DONE!")
print("Computing agglomerative clustering Doc2Vec...")
agglomerative = AgglomerativeClustering(n_clusters=45,
affinity='precomputed',
linkage='average')
dist_file = 'data/dist_doc2vec.bin'
vectors_dist = load(dist_file)
pred_labels = agglomerative.fit_predict(vectors_dist)
pd.DataFrame({'id': ids[idx_filter],
'title': titles[idx_filter],
'cluster': pred_labels[idx_filter],
'label': labels[idx_filter]}).to_csv('data/agglomerative_doc2vec_groups.csv', index=False)
print("Computing agglomerative clustering Doc2Vec... DONE!")
print("Computing agglomerative clustering TF-IDF...")
agglomerative = AgglomerativeClustering(n_clusters=40,
affinity='precomputed',
linkage='average')
dist_file = 'data/dist_tfidf.bin'
vectors_dist = load(dist_file)
pred_labels = agglomerative.fit_predict(vectors_dist)
pd.DataFrame({'id': ids[idx_filter],
'title': titles[idx_filter],
'cluster': pred_labels[idx_filter],
'label': labels[idx_filter]}).to_csv('data/agglomerative_tfidf_groups.csv', index=False)
print("Computing agglomerative clustering TF-IDF... DONE!")
print("Computing agglomerative clustering BOW...")
agglomerative = AgglomerativeClustering(n_clusters=50,
affinity='precomputed',
linkage='average')
dist_file = 'data/dist_bow.bin'
vectors_dist = load(dist_file)
pred_labels = agglomerative.fit_predict(vectors_dist)
pd.DataFrame({'id': ids[idx_filter],
'title': titles[idx_filter],
'cluster': pred_labels[idx_filter],
'label': labels[idx_filter]}).to_csv('data/agglomerative_bow_groups.csv', index=False)
print("Computing agglomerative clustering BOW... DONE!")
print("Computing agglomerative clustering NEL...")
agglomerative = AgglomerativeClustering(n_clusters=85,
affinity='precomputed',
linkage='average')
dist_file = 'data/dist_nel.bin'
vectors_dist = load(dist_file)
pred_labels = agglomerative.fit_predict(vectors_dist)
pd.DataFrame({'id': ids[idx_filter],
'title': titles[idx_filter],
'cluster': pred_labels[idx_filter],
'label': labels[idx_filter]}).to_csv('data/agglomerative_nel_groups.csv', index=False)
print("Computing agglomerative clustering NEL... DONE!")