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
|
/*
* cluster.cpp
*
*
* Created by Pat Schloss on 8/14/08.
* Copyright 2008 Patrick D. Schloss. All rights reserved.
*
*/
#include "cluster.hpp"
#include "rabundvector.hpp"
#include "listvector.hpp"
/***********************************************************************/
Cluster::Cluster(RAbundVector* rav, ListVector* lv, SparseDistanceMatrix* dm, float c, string f, float cs) :
rabund(rav), list(lv), dMatrix(dm), method(f), adjust(cs)
{
try {
mapWanted = false; //set to true by mgcluster to speed up overlap merge
//save so you can modify as it changes in average neighbor
cutoff = c;
m = MothurOut::getInstance();
}
catch(exception& e) {
m->errorOut(e, "Cluster", "Cluster");
exit(1);
}
}
/***********************************************************************/
void Cluster::clusterBins(){
try {
rabund->set(smallCol, rabund->get(smallRow)+rabund->get(smallCol));
rabund->set(smallRow, 0);
rabund->setLabel(toString(smallDist));
}
catch(exception& e) {
m->errorOut(e, "Cluster", "clusterBins");
exit(1);
}
}
/***********************************************************************/
void Cluster::clusterNames(){
try {
if (mapWanted) { updateMap(); }
list->set(smallCol, list->get(smallRow)+','+list->get(smallCol));
list->set(smallRow, "");
list->setLabel(toString(smallDist));
}
catch(exception& e) {
m->errorOut(e, "Cluster", "clusterNames");
exit(1);
}
}
/***********************************************************************/
void Cluster::update(double& cutOFF){
try {
smallCol = dMatrix->getSmallestCell(smallRow);
nColCells = dMatrix->seqVec[smallCol].size();
nRowCells = dMatrix->seqVec[smallRow].size();
vector<int> foundCol(nColCells, 0);
//cout << dMatrix->getNNodes() << " small cell: " << smallRow << '\t' << smallCol << endl;
int search;
bool changed;
for (int i=nRowCells-1;i>=0;i--) {
if (m->control_pressed) { break; }
//if you are not the smallCell
if (dMatrix->seqVec[smallRow][i].index != smallCol) {
search = dMatrix->seqVec[smallRow][i].index;
bool merged = false;
for (int j=0;j<nColCells;j++) {
if (dMatrix->seqVec[smallCol][j].index != smallRow) { //if you are not the smallest distance
if (dMatrix->seqVec[smallCol][j].index == search) {
foundCol[j] = 1;
merged = true;
changed = updateDistance(dMatrix->seqVec[smallCol][j], dMatrix->seqVec[smallRow][i]);
dMatrix->updateCellCompliment(smallCol, j);
break;
}else if (dMatrix->seqVec[smallCol][j].index < search) { //we don't have a distance for this cell
if (adjust != -1.0) { //adjust
merged = true;
PDistCell value(search, adjust); //create a distance for the missing value
int location = dMatrix->addCellSorted(smallCol, value);
changed = updateDistance(dMatrix->seqVec[smallCol][location], dMatrix->seqVec[smallRow][i]);
dMatrix->updateCellCompliment(smallCol, location);
nColCells++;
foundCol.push_back(0); //add a new found column
//adjust value
for (int k = foundCol.size()-1; k > location; k--) { foundCol[k] = foundCol[k-1]; }
foundCol[location] = 1;
}
j+=nColCells;
}
}
}
//if not merged it you need it for warning
if ((!merged) && (method == "average" || method == "weighted")) {
if (cutOFF > dMatrix->seqVec[smallRow][i].dist) {
cutOFF = dMatrix->seqVec[smallRow][i].dist;
//cout << "changing cutoff to " << cutOFF << endl;
}
}
dMatrix->rmCell(smallRow, i);
}
}
clusterBins();
clusterNames();
// Special handling for singlelinkage case, not sure whether this
// could be avoided
for (int i=nColCells-1;i>=0;i--) {
if (foundCol[i] == 0) {
if (adjust != -1.0) { //adjust
PDistCell value(smallCol, adjust); //create a distance for the missing value
changed = updateDistance(dMatrix->seqVec[smallCol][i], value);
dMatrix->updateCellCompliment(smallCol, i);
}else {
if (method == "average" || method == "weighted") {
if (dMatrix->seqVec[smallCol][i].index != smallRow) { //if you are not hte smallest distance
if (cutOFF > dMatrix->seqVec[smallCol][i].dist) {
cutOFF = dMatrix->seqVec[smallCol][i].dist;
}
}
}
}
dMatrix->rmCell(smallCol, i);
}
}
}
catch(exception& e) {
m->errorOut(e, "Cluster", "update");
exit(1);
}
}
/***********************************************************************/
void Cluster::setMapWanted(bool f) {
try {
mapWanted = f;
//initialize map
for (int k = 0; k < list->getNumBins(); k++) {
string names = list->get(k);
//parse bin
string individual = "";
int binNameslength = names.size();
for(int j=0;j<binNameslength;j++){
if(names[j] == ','){
seq2Bin[individual] = k;
individual = "";
}
else{ individual += names[j]; }
}
//get last name
seq2Bin[individual] = k;
}
}
catch(exception& e) {
m->errorOut(e, "Cluster", "setMapWanted");
exit(1);
}
}
/***********************************************************************/
void Cluster::updateMap() {
try {
//update location of seqs in smallRow since they move to smallCol now
string names = list->get(smallRow);
string individual = "";
int binNameslength = names.size();
for(int j=0;j<binNameslength;j++){
if(names[j] == ','){
seq2Bin[individual] = smallCol;
individual = "";
}
else{ individual += names[j]; }
}
//get last name
seq2Bin[individual] = smallCol;
}
catch(exception& e) {
m->errorOut(e, "Cluster", "updateMap");
exit(1);
}
}
/***********************************************************************/
|
