2 * treegroupscommand.cpp
5 * Created by Sarah Westcott on 4/8/09.
6 * Copyright 2009 Schloss Lab UMASS Amherst. All rights reserved.
10 #include "treegroupscommand.h"
11 #include "sharedjabund.h"
12 #include "sharedsorabund.h"
13 #include "sharedjclass.h"
14 #include "sharedsorclass.h"
15 #include "sharedjest.h"
16 #include "sharedsorest.h"
17 #include "sharedthetayc.h"
18 #include "sharedthetan.h"
19 #include "sharedmorisitahorn.h"
20 #include "sharedbraycurtis.h"
23 //**********************************************************************************************************************
25 TreeGroupCommand::TreeGroupCommand(){
27 globaldata = GlobalData::getInstance();
28 format = globaldata->getFormat();
29 validCalculator = new ValidCalculators();
30 util = new SharedUtil();
33 for (i=0; i<globaldata->Estimators.size(); i++) {
34 if (validCalculator->isValidCalculator("treegroup", globaldata->Estimators[i]) == true) {
35 if (globaldata->Estimators[i] == "jabund") {
36 treeCalculators.push_back(new JAbund());
37 }else if (globaldata->Estimators[i] == "sorabund") {
38 treeCalculators.push_back(new SorAbund());
39 }else if (globaldata->Estimators[i] == "jclass") {
40 treeCalculators.push_back(new Jclass());
41 }else if (globaldata->Estimators[i] == "sorclass") {
42 treeCalculators.push_back(new SorClass());
43 }else if (globaldata->Estimators[i] == "jest") {
44 treeCalculators.push_back(new Jest());
45 }else if (globaldata->Estimators[i] == "sorest") {
46 treeCalculators.push_back(new SorEst());
47 }else if (globaldata->Estimators[i] == "thetayc") {
48 treeCalculators.push_back(new ThetaYC());
49 }else if (globaldata->Estimators[i] == "thetan") {
50 treeCalculators.push_back(new ThetaN());
51 }else if (globaldata->Estimators[i] == "morisitahorn") {
52 treeCalculators.push_back(new MorHorn());
53 }else if (globaldata->Estimators[i] == "braycurtis") {
54 treeCalculators.push_back(new BrayCurtis());
59 //reset calc for next command
60 globaldata->setCalc("");
64 cout << "Standard Error: " << e.what() << " has occurred in the TreeGroupCommand class Function TreeGroupCommand. Please contact Pat Schloss at pschloss@microbio.umass.edu." << "\n";
68 cout << "An unknown error has occurred in the TreeGroupCommand class function TreeGroupCommand. Please contact Pat Schloss at pschloss@microbio.umass.edu." << "\n";
72 //**********************************************************************************************************************
74 TreeGroupCommand::~TreeGroupCommand(){
80 //**********************************************************************************************************************
82 int TreeGroupCommand::execute(){
86 vector<SharedRAbundVector*> subset;
88 //if the users entered no valid calculators don't execute command
89 if (treeCalculators.size() == 0) { return 0; }
92 read = new ReadOTUFile(globaldata->inputFileName);
93 read->read(&*globaldata);
95 input = globaldata->ginput;
96 lookup = input->getSharedRAbundVectors();
98 if (lookup.size() < 2) { cout << "You have not provided enough valid groups. I cannot run the command." << endl; }
100 numGroups = globaldata->Groups.size();
102 for (int i = 0; i < numGroups; i++) { groupNames += globaldata->Groups[i]; }
104 //clear globaldatas old tree names if any
105 globaldata->Treenames.clear();
107 //fills globaldatas tree names
108 globaldata->Treenames = globaldata->Groups;
110 //create treemap class from groupmap for tree class to use
111 tmap = new TreeMap();
112 tmap->makeSim(globaldata->gGroupmap);
113 globaldata->gTreemap = tmap;
115 while(lookup[0] != NULL){
117 if(globaldata->allLines == 1 || globaldata->lines.count(count) == 1 || globaldata->labels.count(lookup[0]->getLabel()) == 1){
119 cout << lookup[0]->getLabel() << '\t' << count << endl;
121 //for each calculator
122 for(int i = 0 ; i < treeCalculators.size(); i++) {
124 //initialize simMatrix
126 simMatrix.resize(numGroups);
127 for (int m = 0; m < simMatrix.size(); m++) {
128 for (int j = 0; j < simMatrix.size(); j++) {
129 simMatrix[m].push_back(0.0);
135 for (int g = 0; g < numGroups; g++) { index[g] = g; }
137 //create a new filename
138 outputFile = getRootName(globaldata->inputFileName) + treeCalculators[i]->getName() + "." + lookup[0]->getLabel() + ".tre";
140 for (int k = 0; k < lookup.size(); k++) {
141 for (int l = k; l < lookup.size(); l++) {
142 if (k != l) { //we dont need to similiarity of a groups to itself
143 //get estimated similarity between 2 groups
145 subset.clear(); //clear out old pair of sharedrabunds
146 //add new pair of sharedrabunds
147 subset.push_back(lookup[k]); subset.push_back(lookup[l]);
149 data = treeCalculators[i]->getValues(subset); //saves the calculator outputs
150 //save values in similarity matrix
151 simMatrix[k][l] = data[0];
152 simMatrix[l][k] = data[0];
157 //creates tree from similarity matrix and write out file
162 //get next line to process
163 lookup = input->getSharedRAbundVectors();
167 //reset groups parameter
168 globaldata->Groups.clear(); globaldata->setGroups("");
172 catch(exception& e) {
173 cout << "Standard Error: " << e.what() << " has occurred in the TreeGroupCommand class Function execute. Please contact Pat Schloss at pschloss@microbio.umass.edu." << "\n";
177 cout << "An unknown error has occurred in the TreeGroupCommand class function execute. Please contact Pat Schloss at pschloss@microbio.umass.edu." << "\n";
181 //**********************************************************************************************************************
183 void TreeGroupCommand::createTree(){
188 //do merges and create tree structure by setting parents and children
189 //there are numGroups - 1 merges to do
190 for (int i = 0; i < (numGroups - 1); i++) {
194 //find largest value in sims matrix by searching lower triangle
195 for (int j = 1; j < simMatrix.size(); j++) {
196 for (int k = 0; k < j; k++) {
197 if (simMatrix[j][k] > largest) { largest = simMatrix[j][k]; row = j; column = k; }
201 //set non-leaf node info and update leaves to know their parents
203 t->tree[numGroups + i].setChildren(index[row], index[column]);
206 t->tree[index[row]].setParent(numGroups + i);
207 t->tree[index[column]].setParent(numGroups + i);
209 //blength = distance / 2;
210 float blength = ((1.0 - largest) / 2);
213 t->tree[index[row]].setBranchLength(blength - t->tree[index[row]].getLengthToLeaves());
214 t->tree[index[column]].setBranchLength(blength - t->tree[index[column]].getLengthToLeaves());
216 //set your length to leaves to your childs length plus branchlength
217 t->tree[numGroups + i].setLengthToLeaves(t->tree[index[row]].getLengthToLeaves() + t->tree[index[row]].getBranchLength());
221 index[row] = numGroups+i;
222 index[column] = numGroups+i;
224 //zero out highest value that caused the merge.
225 simMatrix[row][column] = 0.0;
226 simMatrix[column][row] = 0.0;
228 //merge values in simsMatrix
229 for (int n = 0; n < simMatrix.size(); n++) {
230 //row becomes merge of 2 groups
231 simMatrix[row][n] = (simMatrix[row][n] + simMatrix[column][n]) / 2;
232 simMatrix[n][row] = simMatrix[row][n];
234 simMatrix[column][n] = 0.0;
235 simMatrix[n][column] = 0.0;
243 t->createNewickFile(outputFile);
249 catch(exception& e) {
250 cout << "Standard Error: " << e.what() << " has occurred in the TreeGroupCommand class Function createTree. Please contact Pat Schloss at pschloss@microbio.umass.edu." << "\n";
254 cout << "An unknown error has occurred in the TreeGroupCommand class function createTree. Please contact Pat Schloss at pschloss@microbio.umass.edu." << "\n";
258 /***********************************************************/
259 void TreeGroupCommand::printSims() {
261 cout << "simsMatrix" << endl;
262 for (int m = 0; m < simMatrix.size(); m++) {
263 for (int n = 0; n < simMatrix.size(); n++) {
264 cout << simMatrix[m][n] << '\t';
270 catch(exception& e) {
271 cout << "Standard Error: " << e.what() << " has occurred in the TreeGroupCommand class Function printSims. Please contact Pat Schloss at pschloss@microbio.umass.edu." << "\n";
275 cout << "An unknown error has occurred in the TreeGroupCommand class function printSims. Please contact Pat Schloss at pschloss@microbio.umass.edu." << "\n";
279 /***********************************************************/