-
Notifications
You must be signed in to change notification settings - Fork 8
/
cluster.cc
150 lines (137 loc) · 4.49 KB
/
cluster.cc
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
//
// Cluster class
//
// Copyright(C) 2010 Mizuki Fujisawa <[email protected]>
//
// This program is free software; you can redistribute it and/or modify
// it under the terms of the GNU General Public License as published by
// the Free Software Foundation; version 2 of the License.
//
// This program is distributed in the hope that it will be useful,
// but WITHOUT ANY WARRANTY; without even the implied warranty of
// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
// GNU General Public License for more details.
//
// You should have received a copy of the GNU General Public License
// along with this program; if not, write to the Free Software
// Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA
//
#include <algorithm>
#include "cluster.h"
namespace bayon {
/**
* Get sorted documents in this clusters.
*/
void Cluster::sorted_documents(
std::vector<std::pair<Document *, double> > &pairs) {
Vector *centroid = centroid_vector();
for (size_t i = 0; i < documents_.size(); i++) {
double similarity = Vector::inner_product(*documents_[i]->feature(),
*centroid);
pairs.push_back(std::pair<Document *, double>(documents_[i], similarity));
}
std::sort(pairs.begin(), pairs.end(), greater_pair<Document *, double>);
}
/**
* Choose documents randomly.
*/
void Cluster::choose_randomly(size_t ndocs, std::vector<Document *> &docs) {
HashMap<size_t, bool>::type choosed;
size_t siz = size();
init_hash_map(siz, choosed, ndocs);
if (siz < ndocs) ndocs = siz;
size_t count = 0;
while (count < ndocs) {
size_t index = myrand(&seed_) % siz;
if (choosed.find(index) == choosed.end()) {
choosed.insert(std::pair<size_t, bool>(index, true));
docs.push_back(documents_[index]);
++count;
}
}
}
/**
* Choose documents smartly.
*/
void Cluster::choose_smartly(size_t ndocs, std::vector<Document *> &docs) {
HashMap<size_t, double>::type closest;
size_t siz = size();
init_hash_map(siz, closest, docs.size());
if (siz < ndocs) ndocs = siz;
size_t index, count = 0;
index = myrand(&seed_) % siz; // initial center
docs.push_back(documents_[index]);
++count;
double potential = 0.0;
for (size_t i = 0; i < documents_.size(); i++) {
double dist = 1.0 - Vector::inner_product(*documents_[i]->feature(),
*documents_[index]->feature());
potential += dist;
closest[i] = dist;
}
// choose each center
while (count < ndocs) {
double randval = (double)myrand(&seed_) / RAND_MAX * potential;
for (index = 0; index < documents_.size(); index++) {
double dist = closest[index];
if (randval <= dist) break;
randval -= dist;
}
if (index == documents_.size()) index--;
docs.push_back(documents_[index]);
++count;
double new_potential = 0.0;
for (size_t i = 0; i < documents_.size(); i++) {
double dist = 1.0 - Vector::inner_product(*documents_[i]->feature(),
*documents_[index]->feature());
double min = closest[i];
if (dist < min) {
closest[i] = dist;
min = dist;
}
new_potential += min;
}
potential = new_potential;
}
}
/**
* Set a gain when the cluster sectioned.
*/
void Cluster::set_sectioned_gain() {
double gain = 0.0;
if (sectioned_gain_ == 0 && sectioned_clusters_.size() > 1) {
for (size_t i = 0; i < sectioned_clusters_.size(); i++) {
gain += sectioned_clusters_[i]->composite_vector()->norm();
}
gain -= composite_.norm();
}
sectioned_gain_ = gain;
}
/**
* Section this cluster.
*/
void Cluster::section(size_t nclusters) {
if (size() < nclusters) return;
std::vector<Document *> centroids;
// choose_randomly(nclusters, centroids);
choose_smartly(nclusters, centroids);
for (size_t i = 0; i < centroids.size(); i++) {
Cluster *cluster = new Cluster();
cluster->set_seed(seed_);
sectioned_clusters_.push_back(cluster);
}
for (size_t i = 0; i < documents_.size(); i++) {
double max_similarity = -1.0;
size_t max_index = 0;
for (size_t j = 0; j < centroids.size(); j++) {
double similarity = Vector::inner_product(*documents_[i]->feature(),
*centroids[j]->feature());
if (max_similarity < similarity) {
max_similarity = similarity;
max_index = j;
}
}
sectioned_clusters_[max_index]->add_document(documents_[i]);
}
}
} /* namespace bayon */