2 * bootstrapsharedcommand.cpp
5 * Created by Sarah Westcott on 4/16/09.
6 * Copyright 2009 Schloss Lab UMASS Amherst. All rights reserved.
10 #include "bootstrapsharedcommand.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 BootSharedCommand::BootSharedCommand(){
27 globaldata = GlobalData::getInstance();
28 format = globaldata->getFormat();
29 convert(globaldata->getIters(), iters);
30 validCalculator = new ValidCalculators();
31 util = new SharedUtil();
35 for (i=0; i<globaldata->Estimators.size(); i++) {
36 if (validCalculator->isValidCalculator("boot", globaldata->Estimators[i]) == true) {
37 if (globaldata->Estimators[i] == "jabund") {
38 treeCalculators.push_back(new JAbund());
39 }else if (globaldata->Estimators[i] == "sorabund") {
40 treeCalculators.push_back(new SorAbund());
41 }else if (globaldata->Estimators[i] == "jclass") {
42 treeCalculators.push_back(new Jclass());
43 }else if (globaldata->Estimators[i] == "sorclass") {
44 treeCalculators.push_back(new SorClass());
45 }else if (globaldata->Estimators[i] == "jest") {
46 treeCalculators.push_back(new Jest());
47 }else if (globaldata->Estimators[i] == "sorest") {
48 treeCalculators.push_back(new SorEst());
49 }else if (globaldata->Estimators[i] == "thetayc") {
50 treeCalculators.push_back(new ThetaYC());
51 }else if (globaldata->Estimators[i] == "thetan") {
52 treeCalculators.push_back(new ThetaN());
53 }else if (globaldata->Estimators[i] == "morisitahorn") {
54 treeCalculators.push_back(new MorHorn());
55 }else if (globaldata->Estimators[i] == "braycurtis") {
56 treeCalculators.push_back(new BrayCurtis());
62 for (int i=0; i < treeCalculators.size(); i++) {
67 //reset calc for next command
68 globaldata->setCalc("");
72 cout << "Standard Error: " << e.what() << " has occurred in the BootSharedCommand class Function BootSharedCommand. Please contact Pat Schloss at pschloss@microbio.umass.edu." << "\n";
76 cout << "An unknown error has occurred in the BootSharedCommand class function BootSharedCommand. Please contact Pat Schloss at pschloss@microbio.umass.edu." << "\n";
80 //**********************************************************************************************************************
82 BootSharedCommand::~BootSharedCommand(){
88 //**********************************************************************************************************************
90 int BootSharedCommand::execute(){
94 vector<SharedRAbundVector*> subset;
96 //if the users entered no valid calculators don't execute command
97 if (treeCalculators.size() == 0) { return 0; }
100 read = new ReadOTUFile(globaldata->inputFileName);
101 read->read(&*globaldata);
102 input = globaldata->ginput;
103 order = input->getSharedOrderVector();
106 util->setGroups(globaldata->Groups, globaldata->gGroupmap->namesOfGroups, "treegroup");
107 numGroups = globaldata->Groups.size();
109 //clear globaldatas old tree names if any
110 globaldata->Treenames.clear();
112 //fills globaldatas tree names
113 globaldata->Treenames = globaldata->Groups;
115 //create treemap class from groupmap for tree class to use
116 tmap = new TreeMap();
117 tmap->makeSim(globaldata->gGroupmap);
118 globaldata->gTreemap = tmap;
120 while(order != NULL){
122 if(globaldata->allLines == 1 || globaldata->lines.count(count) == 1 || globaldata->labels.count(order->getLabel()) == 1){
124 cout << order->getLabel() << '\t' << count << endl;
126 //open an ostream for each calc to print to
127 for (int z = 0; z < treeCalculators.size(); z++) {
128 //create a new filename
129 outputFile = getRootName(globaldata->inputFileName) + treeCalculators[z]->getName() + ".boot" + order->getLabel() + ".tre";
130 openOutputFile(outputFile, *(out[z]));
133 //create a file for each calculator with the 1000 trees in it.
134 for (int p = 0; p < iters; p++) {
136 util->getSharedVectorswithReplacement(globaldata->Groups, lookup, order); //fills group vectors from order vector.
138 //for each calculator
139 for(int i = 0 ; i < treeCalculators.size(); i++) {
141 //initialize simMatrix
143 simMatrix.resize(numGroups);
144 for (int m = 0; m < simMatrix.size(); m++) {
145 for (int j = 0; j < simMatrix.size(); j++) {
146 simMatrix[m].push_back(0.0);
152 for (int g = 0; g < numGroups; g++) { index[g] = g; }
154 for (int k = 0; k < lookup.size(); k++) { // pass cdd each set of groups to commpare
155 for (int l = k; l < lookup.size(); l++) {
156 if (k != l) { //we dont need to similiarity of a groups to itself
157 subset.clear(); //clear out old pair of sharedrabunds
158 //add new pair of sharedrabunds
159 subset.push_back(lookup[k]); subset.push_back(lookup[l]);
161 //get estimated similarity between 2 groups
162 data = treeCalculators[i]->getValues(subset); //saves the calculator outputs
163 //save values in similarity matrix
164 simMatrix[k][l] = data[0];
165 simMatrix[l][k] = data[0];
170 //creates tree from similarity matrix and write out file
174 //close ostream for each calc
175 for (int z = 0; z < treeCalculators.size(); z++) { out[z]->close(); }
178 //get next line to process
179 order = input->getSharedOrderVector();
183 //reset groups parameter
184 globaldata->Groups.clear(); globaldata->setGroups("");
188 catch(exception& e) {
189 cout << "Standard Error: " << e.what() << " has occurred in the BootSharedCommand class Function execute. Please contact Pat Schloss at pschloss@microbio.umass.edu." << "\n";
193 cout << "An unknown error has occurred in the BootSharedCommand class function execute. Please contact Pat Schloss at pschloss@microbio.umass.edu." << "\n";
197 //**********************************************************************************************************************
199 void BootSharedCommand::createTree(ostream* out){
204 //do merges and create tree structure by setting parents and children
205 //there are numGroups - 1 merges to do
206 for (int i = 0; i < (numGroups - 1); i++) {
208 float largest = -1.0;
210 //find largest value in sims matrix by searching lower triangle
211 for (int j = 1; j < simMatrix.size(); j++) {
212 for (int k = 0; k < j; k++) {
213 if (simMatrix[j][k] > largest) { largest = simMatrix[j][k]; row = j; column = k; }
217 //set non-leaf node info and update leaves to know their parents
219 t->tree[numGroups + i].setChildren(index[row], index[column]);
222 t->tree[index[row]].setParent(numGroups + i);
223 t->tree[index[column]].setParent(numGroups + i);
225 //blength = distance / 2;
226 float blength = ((1.0 - largest) / 2);
229 t->tree[index[row]].setBranchLength(blength - t->tree[index[row]].getLengthToLeaves());
230 t->tree[index[column]].setBranchLength(blength - t->tree[index[column]].getLengthToLeaves());
232 //set your length to leaves to your childs length plus branchlength
233 t->tree[numGroups + i].setLengthToLeaves(t->tree[index[row]].getLengthToLeaves() + t->tree[index[row]].getBranchLength());
237 index[row] = numGroups+i;
238 index[column] = numGroups+i;
240 //zero out highest value that caused the merge.
241 simMatrix[row][column] = -1.0;
242 simMatrix[column][row] = -1.0;
244 //merge values in simsMatrix
245 for (int n = 0; n < simMatrix.size(); n++) {
246 //row becomes merge of 2 groups
247 simMatrix[row][n] = (simMatrix[row][n] + simMatrix[column][n]) / 2;
248 simMatrix[n][row] = simMatrix[row][n];
250 simMatrix[column][n] = -1.0;
251 simMatrix[n][column] = -1.0;
265 catch(exception& e) {
266 cout << "Standard Error: " << e.what() << " has occurred in the BootSharedCommand class Function createTree. Please contact Pat Schloss at pschloss@microbio.umass.edu." << "\n";
270 cout << "An unknown error has occurred in the BootSharedCommand class function createTree. Please contact Pat Schloss at pschloss@microbio.umass.edu." << "\n";
274 /***********************************************************/
275 void BootSharedCommand::printSims() {
277 cout << "simsMatrix" << endl;
278 for (int m = 0; m < simMatrix.size(); m++) {
279 for (int n = 0; n < simMatrix.size(); n++) {
280 cout << simMatrix[m][n] << '\t';
286 catch(exception& e) {
287 cout << "Standard Error: " << e.what() << " has occurred in the BootSharedCommand class Function printSims. Please contact Pat Schloss at pschloss@microbio.umass.edu." << "\n";
291 cout << "An unknown error has occurred in the BootSharedCommand class function printSims. Please contact Pat Schloss at pschloss@microbio.umass.edu." << "\n";
295 /***********************************************************/