X-Git-Url: https://git.donarmstrong.com/?a=blobdiff_plain;f=treegroupscommand.cpp;h=850df46929f8b4f7efa076a17b0052136f204b23;hb=0024441c1ff9a98467ba53cc7420a37c98f32e86;hp=6393684f2e79380788553e978504d0f0d306253f;hpb=154e089bcd37c2c2f773e53ffb88a20170b27037;p=mothur.git diff --git a/treegroupscommand.cpp b/treegroupscommand.cpp index 6393684..850df46 100644 --- a/treegroupscommand.cpp +++ b/treegroupscommand.cpp @@ -17,6 +17,7 @@ #include "sharedthetayc.h" #include "sharedthetan.h" #include "sharedmorisitahorn.h" +#include "sharedbraycurtis.h" //********************************************************************************************************************** @@ -26,29 +27,32 @@ TreeGroupCommand::TreeGroupCommand(){ globaldata = GlobalData::getInstance(); format = globaldata->getFormat(); validCalculator = new ValidCalculators(); - util = new SharedUtil(); - int i; - for (i=0; iEstimators.size(); i++) { - if (validCalculator->isValidCalculator("treegroup", globaldata->Estimators[i]) == true) { - if (globaldata->Estimators[i] == "jabund") { - treeCalculators.push_back(new JAbund()); - }else if (globaldata->Estimators[i] == "sorensonabund") { - treeCalculators.push_back(new SorAbund()); - }else if (globaldata->Estimators[i] == "jclass") { - treeCalculators.push_back(new Jclass()); - }else if (globaldata->Estimators[i] == "sorclass") { - treeCalculators.push_back(new SorClass()); - }else if (globaldata->Estimators[i] == "jest") { - treeCalculators.push_back(new Jest()); - }else if (globaldata->Estimators[i] == "sorest") { - treeCalculators.push_back(new SorEst()); - }else if (globaldata->Estimators[i] == "thetayc") { - treeCalculators.push_back(new ThetaYC()); - }else if (globaldata->Estimators[i] == "thetan") { - treeCalculators.push_back(new ThetaN()); - }else if (globaldata->Estimators[i] == "morisitahorn") { - treeCalculators.push_back(new MorHorn()); + if (format == "sharedfile") { + int i; + for (i=0; iEstimators.size(); i++) { + if (validCalculator->isValidCalculator("treegroup", globaldata->Estimators[i]) == true) { + if (globaldata->Estimators[i] == "jabund") { + treeCalculators.push_back(new JAbund()); + }else if (globaldata->Estimators[i] == "sorabund") { + treeCalculators.push_back(new SorAbund()); + }else if (globaldata->Estimators[i] == "jclass") { + treeCalculators.push_back(new Jclass()); + }else if (globaldata->Estimators[i] == "sorclass") { + treeCalculators.push_back(new SorClass()); + }else if (globaldata->Estimators[i] == "jest") { + treeCalculators.push_back(new Jest()); + }else if (globaldata->Estimators[i] == "sorest") { + treeCalculators.push_back(new SorEst()); + }else if (globaldata->Estimators[i] == "thetayc") { + treeCalculators.push_back(new ThetaYC()); + }else if (globaldata->Estimators[i] == "thetan") { + treeCalculators.push_back(new ThetaN()); + }else if (globaldata->Estimators[i] == "morisitahorn") { + treeCalculators.push_back(new MorHorn()); + }else if (globaldata->Estimators[i] == "braycurtis") { + treeCalculators.push_back(new BrayCurtis()); + } } } } @@ -70,111 +74,82 @@ TreeGroupCommand::TreeGroupCommand(){ TreeGroupCommand::~TreeGroupCommand(){ delete input; - delete read; - delete util; + if (format == "sharedfile") {delete read;} + else { delete readMatrix; delete matrix; delete list; } + delete tmap; + } //********************************************************************************************************************** int TreeGroupCommand::execute(){ try { - int count = 1; - EstOutput data; - - //if the users entered no valid calculators don't execute command - if (treeCalculators.size() == 0) { return 0; } - if (format == "sharedfile") { + //if the users entered no valid calculators don't execute command + if (treeCalculators.size() == 0) { cout << "You have given no valid calculators." << endl; return 0; } + + //you have groups read = new ReadOTUFile(globaldata->inputFileName); read->read(&*globaldata); input = globaldata->ginput; - order = input->getSharedOrderVector(); - }else { - //you are using a list and a groupfile - read = new ReadOTUFile(globaldata->inputFileName); - read->read(&*globaldata); - - input = globaldata->ginput; - SharedList = globaldata->gSharedList; - order = SharedList->getSharedOrderVector(); - } - - //set users groups - util->setGroups(globaldata->Groups, globaldata->gGroupmap->namesOfGroups, "treegroup"); - numGroups = globaldata->Groups.size(); - groupNames = ""; - for (int i = 0; i < numGroups; i++) { groupNames += globaldata->Groups[i]; } - - //clear globaldatas old tree names if any - globaldata->Treenames.clear(); - - //fills globaldatas tree names - globaldata->Treenames = globaldata->Groups; - - //create treemap class from groupmap for tree class to use - tmap = new TreeMap(); - tmap->makeSim(globaldata->gGroupmap); - globaldata->gTreemap = tmap; + lookup = input->getSharedRAbundVectors(); + lastLookup = lookup; - while(order != NULL){ + if (lookup.size() < 2) { cout << "You have not provided enough valid groups. I cannot run the command." << endl; return 0; } - if(globaldata->allLines == 1 || globaldata->lines.count(count) == 1 || globaldata->labels.count(order->getLabel()) == 1){ - - cout << order->getLabel() << '\t' << count << endl; - util->getSharedVectors(globaldata->Groups, lookup, order); //fills group vectors from order vector. - - //for each calculator - for(int i = 0 ; i < treeCalculators.size(); i++) { - - //initialize simMatrix - simMatrix.clear(); - simMatrix.resize(numGroups); - for (int m = 0; m < simMatrix.size(); m++) { - for (int j = 0; j < simMatrix.size(); j++) { - simMatrix[m].push_back(0.0); - } - } - - //initialize index - index.clear(); - for (int g = 0; g < numGroups; g++) { index[g] = g; } + //create tree file + makeSimsShared(); + }else{ + //read in dist file + filename = globaldata->inputFileName; - //create a new filename - outputFile = getRootName(globaldata->inputFileName) + treeCalculators[i]->getName() + "." + order->getLabel() + ".tre"; - - for (int k = 0; k < lookup.size(); k++) { - for (int l = k; l < lookup.size(); l++) { - if (k != l) { //we dont need to similiarity of a groups to itself - //get estimated similarity between 2 groups - data = treeCalculators[i]->getValues(lookup[k], lookup[l]); //saves the calculator outputs - //save values in similarity matrix - simMatrix[k][l] = data[0]; - simMatrix[l][k] = data[0]; - } - } - } - - //creates tree from similarity matrix and write out file - createTree(); - } + if (format == "column") { readMatrix = new ReadColumnMatrix(filename); } + else if (format == "phylip") { readMatrix = new ReadPhylipMatrix(filename); } + + if(globaldata->getPrecision() != ""){ + convert(globaldata->getPrecision(), precision); } - //get next line to process - if (format == "sharedfile") { - order = input->getSharedOrderVector(); - }else { - //you are using a list and a groupfile - SharedList = input->getSharedListVector(); //get new list vector to process - if (SharedList != NULL) { - order = SharedList->getSharedOrderVector(); //gets new order vector with group info. - }else { - break; - } + if(globaldata->getCutOff() != ""){ + convert(globaldata->getCutOff(), cutoff); + cutoff += (5 / (precision * 10.0)); } - count++; - } + readMatrix->setCutoff(cutoff); + + if(globaldata->getNameFile() != ""){ + nameMap = new NameAssignment(globaldata->getNameFile()); + nameMap->readMap(1,2); + } + else{ + nameMap = NULL; + } + + readMatrix->read(nameMap); + list = readMatrix->getListVector(); + matrix = readMatrix->getMatrix(); + + //make treemap + tmap = new TreeMap(); + tmap->makeSim(list); + globaldata->gTreemap = tmap; + + globaldata->Groups = tmap->namesOfGroups; + + //clear globaldatas old tree names if any + globaldata->Treenames.clear(); + //fills globaldatas tree names + globaldata->Treenames = globaldata->Groups; + + makeSimsDist(); + + //create a new filename + outputFile = getRootName(globaldata->inputFileName) + "tre"; + + createTree(); + } + //reset groups parameter globaldata->Groups.clear(); globaldata->setGroups(""); @@ -199,8 +174,8 @@ void TreeGroupCommand::createTree(){ //do merges and create tree structure by setting parents and children //there are numGroups - 1 merges to do for (int i = 0; i < (numGroups - 1); i++) { - - float largest = 0.0; + float largest = -1000.0; + int row, column; //find largest value in sims matrix by searching lower triangle for (int j = 1; j < simMatrix.size(); j++) { @@ -232,9 +207,9 @@ void TreeGroupCommand::createTree(){ index[row] = numGroups+i; index[column] = numGroups+i; - //zero out highest value that caused the merge. - simMatrix[row][column] = 0.0; - simMatrix[column][row] = 0.0; + //remove highest value that caused the merge. + simMatrix[row][column] = -1000.0; + simMatrix[column][row] = -1000.0; //merge values in simsMatrix for (int n = 0; n < simMatrix.size(); n++) { @@ -242,11 +217,15 @@ void TreeGroupCommand::createTree(){ simMatrix[row][n] = (simMatrix[row][n] + simMatrix[column][n]) / 2; simMatrix[n][row] = simMatrix[row][n]; //delete column - simMatrix[column][n] = 0.0; - simMatrix[n][column] = 0.0; + simMatrix[column][n] = -1000.0; + simMatrix[n][column] = -1000.0; } } - + + //adjust tree to make sure root to tip length is .5 + int root = t->findRoot(); + t->tree[root].setBranchLength((0.5 - t->tree[root].getLengthToLeaves())); + //assemble tree t->assembleTree(); @@ -267,14 +246,21 @@ void TreeGroupCommand::createTree(){ } } /***********************************************************/ -void TreeGroupCommand::printSims() { +void TreeGroupCommand::printSims(ostream& out) { try { - cout << "simsMatrix" << endl; + + //output column headers + //out << '\t'; + //for (int i = 0; i < lookup.size(); i++) { out << lookup[i]->getGroup() << '\t'; } + //out << endl; + + for (int m = 0; m < simMatrix.size(); m++) { + //out << lookup[m]->getGroup() << '\t'; for (int n = 0; n < simMatrix.size(); n++) { - cout << simMatrix[m][n] << '\t'; + out << simMatrix[m][n] << '\t'; } - cout << endl; + out << endl; } } @@ -288,5 +274,181 @@ void TreeGroupCommand::printSims() { } } /***********************************************************/ +void TreeGroupCommand::makeSimsDist() { + try { + numGroups = list->size(); + + //initialize index + index.clear(); + for (int g = 0; g < numGroups; g++) { index[g] = g; } + + //initialize simMatrix + simMatrix.clear(); + simMatrix.resize(numGroups); + for (int m = 0; m < simMatrix.size(); m++) { + for (int j = 0; j < simMatrix.size(); j++) { + simMatrix[m].push_back(0.0); + } + } + + //go through sparse matrix and fill sims + //go through each cell in the sparsematrix + for(MatData currentCell = matrix->begin(); currentCell != matrix->end(); currentCell++){ + //similairity = -(distance-1) + simMatrix[currentCell->row][currentCell->column] = -(currentCell->dist -1.0); + simMatrix[currentCell->column][currentCell->row] = -(currentCell->dist -1.0); + } + + } + catch(exception& e) { + cout << "Standard Error: " << e.what() << " has occurred in the TreeGroupCommand class Function makeSimsDist. Please contact Pat Schloss at pschloss@microbio.umass.edu." << "\n"; + exit(1); + } + catch(...) { + cout << "An unknown error has occurred in the TreeGroupCommand class function makeSimsDist. Please contact Pat Schloss at pschloss@microbio.umass.edu." << "\n"; + exit(1); + } +} + +/***********************************************************/ +void TreeGroupCommand::makeSimsShared() { + try { + int count = 1; + + //clear globaldatas old tree names if any + globaldata->Treenames.clear(); + + //fills globaldatas tree names + globaldata->Treenames = globaldata->Groups; + + //create treemap class from groupmap for tree class to use + tmap = new TreeMap(); + tmap->makeSim(globaldata->gGroupmap); + globaldata->gTreemap = tmap; + + set processedLabels; + set userLabels = globaldata->labels; + set userLines = globaldata->lines; + + //as long as you are not at the end of the file or done wih the lines you want + while((lookup[0] != NULL) && ((globaldata->allLines == 1) || (userLabels.size() != 0) || (userLines.size() != 0))) { + + if(globaldata->allLines == 1 || globaldata->lines.count(count) == 1 || globaldata->labels.count(lookup[0]->getLabel()) == 1){ + cout << lookup[0]->getLabel() << '\t' << count << endl; + process(lookup); + + processedLabels.insert(lookup[0]->getLabel()); + userLabels.erase(lookup[0]->getLabel()); + userLines.erase(count); + } + + if ((anyLabelsToProcess(lookup[0]->getLabel(), userLabels, "") == true) && (processedLabels.count(lastLookup[0]->getLabel()) != 1)) { + cout << lastLookup[0]->getLabel() << '\t' << count << endl; + process(lastLookup); + + processedLabels.insert(lastLookup[0]->getLabel()); + userLabels.erase(lastLookup[0]->getLabel()); + } + + //prevent memory leak + if (count != 1) { for (int i = 0; i < lastLookup.size(); i++) { delete lastLookup[i]; } } + lastLookup = lookup; + + //get next line to process + lookup = input->getSharedRAbundVectors(); + count++; + } + + //output error messages about any remaining user labels + set::iterator it; + bool needToRun = false; + for (it = userLabels.begin(); it != userLabels.end(); it++) { + cout << "Your file does not include the label "<< *it; + if (processedLabels.count(lastLookup[0]->getLabel()) != 1) { + cout << ". I will use " << lastLookup[0]->getLabel() << "." << endl; + needToRun = true; + }else { + cout << ". Please refer to " << lastLookup[0]->getLabel() << "." << endl; + } + } + + //run last line if you need to + if (needToRun == true) { + cout << lastLookup[0]->getLabel() << '\t' << count << endl; + process(lastLookup); + } + + for (int i = 0; i < lastLookup.size(); i++) { delete lastLookup[i]; } + for(int i = 0 ; i < treeCalculators.size(); i++) { delete treeCalculators[i]; } + } + catch(exception& e) { + cout << "Standard Error: " << e.what() << " has occurred in the TreeGroupCommand class Function makeSimsShared. Please contact Pat Schloss at pschloss@microbio.umass.edu." << "\n"; + exit(1); + } + catch(...) { + cout << "An unknown error has occurred in the TreeGroupCommand class function makeSimsShared. Please contact Pat Schloss at pschloss@microbio.umass.edu." << "\n"; + exit(1); + } +} + +/***********************************************************/ +void TreeGroupCommand::process(vector thisLookup) { + try{ + EstOutput data; + vector subset; + numGroups = globaldata->Groups.size(); + + //for each calculator + for(int i = 0 ; i < treeCalculators.size(); i++) { + //initialize simMatrix + simMatrix.clear(); + simMatrix.resize(numGroups); + for (int m = 0; m < simMatrix.size(); m++) { + for (int j = 0; j < simMatrix.size(); j++) { + simMatrix[m].push_back(0.0); + } + } + + //initialize index + index.clear(); + for (int g = 0; g < numGroups; g++) { index[g] = g; } + + //create a new filename + outputFile = getRootName(globaldata->inputFileName) + treeCalculators[i]->getName() + "." + thisLookup[0]->getLabel() + ".tre"; + + for (int k = 0; k < thisLookup.size(); k++) { + for (int l = k; l < thisLookup.size(); l++) { + if (k != l) { //we dont need to similiarity of a groups to itself + //get estimated similarity between 2 groups + + subset.clear(); //clear out old pair of sharedrabunds + //add new pair of sharedrabunds + subset.push_back(thisLookup[k]); subset.push_back(thisLookup[l]); + + data = treeCalculators[i]->getValues(subset); //saves the calculator outputs + //save values in similarity matrix + simMatrix[k][l] = data[0]; + simMatrix[l][k] = data[0]; + } + } + } + + //creates tree from similarity matrix and write out file + createTree(); + } + + } + catch(exception& e) { + cout << "Standard Error: " << e.what() << " has occurred in the TreeGroupCommand class Function process. Please contact Pat Schloss at pschloss@microbio.umass.edu." << "\n"; + exit(1); + } + catch(...) { + cout << "An unknown error has occurred in the TreeGroupCommand class function process. Please contact Pat Schloss at pschloss@microbio.umass.edu." << "\n"; + exit(1); + } +} +/***********************************************************/ + +