X-Git-Url: https://git.donarmstrong.com/?a=blobdiff_plain;f=bootstrapsharedcommand.cpp;h=12c0b5e647fefccded7b826e23efb7d019e1b703;hb=30b3ffcd1cfd08e7144ae721bb53e27eb3f7a5d1;hp=e88972cc19aae9c2e21efa665177925c3e86cf46;hpb=50ed3b6104d5821d6184f882e1e1423d47dcbf10;p=mothur.git diff --git a/bootstrapsharedcommand.cpp b/bootstrapsharedcommand.cpp index e88972c..12c0b5e 100644 --- a/bootstrapsharedcommand.cpp +++ b/bootstrapsharedcommand.cpp @@ -22,96 +22,193 @@ //********************************************************************************************************************** -BootSharedCommand::BootSharedCommand(){ +BootSharedCommand::BootSharedCommand(string option){ try { globaldata = GlobalData::getInstance(); - format = globaldata->getFormat(); - convert(globaldata->getIters(), iters); - validCalculator = new ValidCalculators(); - util = new SharedUtil(); + abort = false; + allLines = 1; + lines.clear(); + labels.clear(); + Groups.clear(); + Estimators.clear(); + //allow user to run help + if(option == "help") { help(); abort = true; } - int i; - for (i=0; iEstimators.size(); i++) { - if (validCalculator->isValidCalculator("boot", 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()); + else { + //valid paramters for this command + string Array[] = {"line","label","calc","groups","iters"}; + vector myArray (Array, Array+(sizeof(Array)/sizeof(string))); + + OptionParser parser(option); + map parameters = parser.getParameters(); + + ValidParameters validParameter; + + //check to make sure all parameters are valid for command + for (map::iterator it = parameters.begin(); it != parameters.end(); it++) { + if (validParameter.isValidParameter(it->first, myArray, it->second) != true) { abort = true; } + } + + //make sure the user has already run the read.otu command + if (globaldata->getSharedFile() == "") { + if (globaldata->getListFile() == "") { mothurOut("You must read a list and a group, or a shared before you can use the bootstrap.shared command."); mothurOutEndLine(); abort = true; } + else if (globaldata->getGroupFile() == "") { mothurOut("You must read a list and a group, or a shared before you can use the bootstrap.shared command."); mothurOutEndLine(); abort = true; } + } + + //check for optional parameter and set defaults + // ...at some point should added some additional type checking... + line = validParameter.validFile(parameters, "line", false); + if (line == "not found") { line = ""; } + else { + if(line != "all") { splitAtDash(line, lines); allLines = 0; } + else { allLines = 1; } + } + + label = validParameter.validFile(parameters, "label", false); + if (label == "not found") { label = ""; } + else { + if(label != "all") { splitAtDash(label, labels); allLines = 0; } + else { allLines = 1; } + } + + //make sure user did not use both the line and label parameters + if ((line != "") && (label != "")) { mothurOut("You cannot use both the line and label parameters at the same time. "); mothurOutEndLine(); abort = true; } + //if the user has not specified any line or labels use the ones from read.otu + else if((line == "") && (label == "")) { + allLines = globaldata->allLines; + labels = globaldata->labels; + lines = globaldata->lines; + } + + groups = validParameter.validFile(parameters, "groups", false); + if (groups == "not found") { groups = ""; } + else { + splitAtDash(groups, Groups); + globaldata->Groups = Groups; + } + + calc = validParameter.validFile(parameters, "calc", false); + if (calc == "not found") { calc = "jclass-thetayc"; } + else { + if (calc == "default") { calc = "jclass-thetayc"; } + } + splitAtDash(calc, Estimators); + + string temp; + temp = validParameter.validFile(parameters, "iters", false); if (temp == "not found") { temp = "1000"; } + convert(temp, iters); + + if (abort == false) { + + //used in tree constructor + globaldata->runParse = false; + + validCalculator = new ValidCalculators(); + + int i; + for (i=0; iisValidCalculator("boot", Estimators[i]) == true) { + if (Estimators[i] == "jabund") { + treeCalculators.push_back(new JAbund()); + }else if (Estimators[i] == "sorabund") { + treeCalculators.push_back(new SorAbund()); + }else if (Estimators[i] == "jclass") { + treeCalculators.push_back(new Jclass()); + }else if (Estimators[i] == "sorclass") { + treeCalculators.push_back(new SorClass()); + }else if (Estimators[i] == "jest") { + treeCalculators.push_back(new Jest()); + }else if (Estimators[i] == "sorest") { + treeCalculators.push_back(new SorEst()); + }else if (Estimators[i] == "thetayc") { + treeCalculators.push_back(new ThetaYC()); + }else if (Estimators[i] == "thetan") { + treeCalculators.push_back(new ThetaN()); + }else if (Estimators[i] == "morisitahorn") { + treeCalculators.push_back(new MorHorn()); + }else if (Estimators[i] == "braycurtis") { + treeCalculators.push_back(new BrayCurtis()); + } + } + } + + delete validCalculator; + + ofstream* tempo; + for (int i=0; i < treeCalculators.size(); i++) { + tempo = new ofstream; + out.push_back(tempo); } + + //make a vector of tree* for each calculator + trees.resize(treeCalculators.size()); } } - - ofstream* temp; - for (int i=0; i < treeCalculators.size(); i++) { - temp = new ofstream; - out.push_back(temp); - } - - //reset calc for next command - globaldata->setCalc(""); } catch(exception& e) { - cout << "Standard Error: " << e.what() << " has occurred in the BootSharedCommand class Function BootSharedCommand. Please contact Pat Schloss at pschloss@microbio.umass.edu." << "\n"; + errorOut(e, "BootSharedCommand", "BootSharedCommand"); exit(1); } - catch(...) { - cout << "An unknown error has occurred in the BootSharedCommand class function BootSharedCommand. Please contact Pat Schloss at pschloss@microbio.umass.edu." << "\n"; +} + +//********************************************************************************************************************** + +void BootSharedCommand::help(){ + try { + mothurOut("The bootstrap.shared command can only be executed after a successful read.otu command.\n"); + mothurOut("The bootstrap.shared command parameters are groups, calc, iters, line and label. You may not use line and label at the same time.\n"); + mothurOut("The groups parameter allows you to specify which of the groups in your groupfile you would like included used.\n"); + mothurOut("The group names are separated by dashes. The line and label allow you to select what distance levels you would like trees created for, and are also separated by dashes.\n"); + mothurOut("The bootstrap.shared command should be in the following format: bootstrap.shared(groups=yourGroups, calc=yourCalcs, line=yourLines, label=yourLabels, iters=yourIters).\n"); + mothurOut("Example bootstrap.shared(groups=A-B-C, line=1-3-5, calc=jabund-sorabund, iters=100).\n"); + mothurOut("The default value for groups is all the groups in your groupfile.\n"); + mothurOut("The default value for calc is jclass-thetayc. The default for iters is 1000.\n"); + } + catch(exception& e) { + errorOut(e, "BootSharedCommand", "help"); exit(1); - } + } } + //********************************************************************************************************************** BootSharedCommand::~BootSharedCommand(){ - delete input; - delete read; - delete util; + //made new in execute + if (abort == false) { + delete input; globaldata->ginput = NULL; + delete read; + delete util; + globaldata->gorder = NULL; + } } //********************************************************************************************************************** int BootSharedCommand::execute(){ try { - int count = 1; - EstOutput data; - vector subset; + if (abort == true) { return 0; } + + int count = 1; + util = new SharedUtil(); + + //read first line + read = new ReadOTUFile(globaldata->inputFileName); + read->read(&*globaldata); + input = globaldata->ginput; + order = input->getSharedOrderVector(); + string lastLabel = order->getLabel(); + //if the users entered no valid calculators don't execute command if (treeCalculators.size() == 0) { return 0; } - if (format == "sharedfile") { - 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(); - } - + //if the users enters label "0.06" and there is no "0.06" in their file use the next lowest label. + set processedLabels; + set userLabels = labels; + set userLines = lines; + //set users groups util->setGroups(globaldata->Groups, globaldata->gGroupmap->namesOfGroups, "treegroup"); numGroups = globaldata->Groups.size(); @@ -127,105 +224,84 @@ int BootSharedCommand::execute(){ tmap->makeSim(globaldata->gGroupmap); globaldata->gTreemap = tmap; - while(order != NULL){ + while((order != NULL) && ((allLines == 1) || (userLabels.size() != 0) || (userLines.size() != 0))) { - if(globaldata->allLines == 1 || globaldata->lines.count(count) == 1 || globaldata->labels.count(order->getLabel()) == 1){ - - cout << order->getLabel() << '\t' << count << endl; - - //open an ostream for each calc to print to - for (int z = 0; z < treeCalculators.size(); z++) { - //create a new filename - outputFile = getRootName(globaldata->inputFileName) + treeCalculators[z]->getName() + ".boot" + order->getLabel() + ".tre"; - openOutputFile(outputFile, *(out[z])); - } - - //create a file for each calculator with the 1000 trees in it. - for (int p = 0; p < iters; p++) { - - util->getSharedVectorswithReplacement(globaldata->Groups, lookup, order); //fills group vectors from order vector. + if(allLines == 1 || lines.count(count) == 1 || labels.count(order->getLabel()) == 1){ - //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); - } - } + mothurOut(order->getLabel()); mothurOutEndLine(); + process(order); - //initialize index - index.clear(); - for (int g = 0; g < numGroups; g++) { index[g] = g; } - - for (int k = 0; k < lookup.size(); k++) { // pass cdd each set of groups to commpare - for (int l = k; l < lookup.size(); l++) { - if (k != l) { //we dont need to similiarity of a groups to itself - subset.clear(); //clear out old pair of sharedrabunds - //add new pair of sharedrabunds - subset.push_back(lookup[k]); subset.push_back(lookup[l]); - - //get estimated similarity between 2 groups - data = treeCalculators[i]->getValues(subset); //saves the calculator outputs - //save values in similarity matrix - simMatrix[k][l] = data[0]; - simMatrix[l][k] = data[0]; - } - } - } + processedLabels.insert(order->getLabel()); + userLabels.erase(order->getLabel()); + userLines.erase(count); + } + + //you have a label the user want that is smaller than this line and the last line has not already been processed + if ((anyLabelsToProcess(order->getLabel(), userLabels, "") == true) && (processedLabels.count(lastLabel) != 1)) { - //creates tree from similarity matrix and write out file - createTree(out[i]); - } - } - //close ostream for each calc - for (int z = 0; z < treeCalculators.size(); z++) { out[z]->close(); } + delete order; + order = input->getSharedOrderVector(lastLabel); + mothurOut(order->getLabel()); mothurOutEndLine(); + process(order); + + processedLabels.insert(order->getLabel()); + userLabels.erase(order->getLabel()); } - + + + lastLabel = order->getLabel(); + //get next line to process - if (format == "sharedfile") { - order = input->getSharedOrderVector(); + delete order; + order = input->getSharedOrderVector(); + count++; + } + + //output error messages about any remaining user labels + set::iterator it; + bool needToRun = false; + for (it = userLabels.begin(); it != userLabels.end(); it++) { + mothurOut("Your file does not include the label " + *it); + if (processedLabels.count(lastLabel) != 1) { + mothurOut(". I will use " + lastLabel + "."); mothurOutEndLine(); + needToRun = true; }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; - } + mothurOut(". Please refer to " + lastLabel + "."); mothurOutEndLine(); } - count++; } + //run last line if you need to + if (needToRun == true) { + delete order; + order = input->getSharedOrderVector(lastLabel); + mothurOut(order->getLabel()); mothurOutEndLine(); + process(order); + delete order; + + } + + + //reset groups parameter - globaldata->Groups.clear(); globaldata->setGroups(""); + globaldata->Groups.clear(); return 0; } catch(exception& e) { - cout << "Standard Error: " << e.what() << " has occurred in the BootSharedCommand class Function execute. Please contact Pat Schloss at pschloss@microbio.umass.edu." << "\n"; + errorOut(e, "BootSharedCommand", "execute"); exit(1); } - catch(...) { - cout << "An unknown error has occurred in the BootSharedCommand class function execute. Please contact Pat Schloss at pschloss@microbio.umass.edu." << "\n"; - exit(1); - } } //********************************************************************************************************************** -void BootSharedCommand::createTree(ostream* out){ +void BootSharedCommand::createTree(ostream* out, Tree* t){ try { - //create tree - t = new Tree(); //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 = -1.0; + 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++) { @@ -258,8 +334,8 @@ void BootSharedCommand::createTree(ostream* out){ index[column] = numGroups+i; //zero out highest value that caused the merge. - simMatrix[row][column] = -1.0; - simMatrix[column][row] = -1.0; + simMatrix[row][column] = -1000.0; + simMatrix[column][row] = -1000.0; //merge values in simsMatrix for (int n = 0; n < simMatrix.size(); n++) { @@ -267,10 +343,14 @@ void BootSharedCommand::createTree(ostream* out){ simMatrix[row][n] = (simMatrix[row][n] + simMatrix[column][n]) / 2; simMatrix[n][row] = simMatrix[row][n]; //delete column - simMatrix[column][n] = -1.0; - simMatrix[n][column] = -1.0; + 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(); @@ -278,40 +358,130 @@ void BootSharedCommand::createTree(ostream* out){ //print newick file t->print(*out); - //delete tree - delete t; - } catch(exception& e) { - cout << "Standard Error: " << e.what() << " has occurred in the BootSharedCommand class Function createTree. Please contact Pat Schloss at pschloss@microbio.umass.edu." << "\n"; - exit(1); - } - catch(...) { - cout << "An unknown error has occurred in the BootSharedCommand class function createTree. Please contact Pat Schloss at pschloss@microbio.umass.edu." << "\n"; + errorOut(e, "BootSharedCommand", "createTree"); exit(1); } } /***********************************************************/ void BootSharedCommand::printSims() { try { - cout << "simsMatrix" << endl; + mothurOut("simsMatrix"); mothurOutEndLine(); for (int m = 0; m < simMatrix.size(); m++) { for (int n = 0; n < simMatrix.size(); n++) { - cout << simMatrix[m][n] << '\t'; + mothurOut(simMatrix[m][n]); mothurOut("\t"); } - cout << endl; + mothurOutEndLine(); } } catch(exception& e) { - cout << "Standard Error: " << e.what() << " has occurred in the BootSharedCommand class Function printSims. Please contact Pat Schloss at pschloss@microbio.umass.edu." << "\n"; + errorOut(e, "BootSharedCommand", "printSims"); exit(1); } - catch(...) { - cout << "An unknown error has occurred in the BootSharedCommand class function printSims. Please contact Pat Schloss at pschloss@microbio.umass.edu." << "\n"; +} +/***********************************************************/ +void BootSharedCommand::process(SharedOrderVector* order) { + try{ + EstOutput data; + vector subset; + + + //open an ostream for each calc to print to + for (int z = 0; z < treeCalculators.size(); z++) { + //create a new filename + outputFile = getRootName(globaldata->inputFileName) + treeCalculators[z]->getName() + ".boot" + order->getLabel() + ".tre"; + openOutputFile(outputFile, *(out[z])); + } + + mothurOut("Generating bootstrap trees..."); cout.flush(); + + //create a file for each calculator with the 1000 trees in it. + for (int p = 0; p < iters; p++) { + + util->getSharedVectorswithReplacement(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; } + + for (int k = 0; k < lookup.size(); k++) { // pass cdd each set of groups to commpare + for (int l = k; l < lookup.size(); l++) { + if (k != l) { //we dont need to similiarity of a groups to itself + subset.clear(); //clear out old pair of sharedrabunds + //add new pair of sharedrabunds + subset.push_back(lookup[k]); subset.push_back(lookup[l]); + + //get estimated similarity between 2 groups + data = treeCalculators[i]->getValues(subset); //saves the calculator outputs + //save values in similarity matrix + simMatrix[k][l] = data[0]; + simMatrix[l][k] = data[0]; + } + } + } + + tempTree = new Tree(); + + //creates tree from similarity matrix and write out file + createTree(out[i], tempTree); + + //save trees for consensus command. + trees[i].push_back(tempTree); + } + } + + mothurOut("\tDone."); mothurOutEndLine(); + //delete globaldata's tree + //for (int m = 0; m < globaldata->gTree.size(); m++) { delete globaldata->gTree[m]; } + //globaldata->gTree.clear(); + + + //create consensus trees for each bootstrapped tree set + for (int k = 0; k < trees.size(); k++) { + + mothurOut("Generating consensus tree for " + treeCalculators[k]->getName()); mothurOutEndLine(); + + //set global data to calc trees + globaldata->gTree = trees[k]; + + string filename = getRootName(globaldata->inputFileName) + treeCalculators[k]->getName() + ".boot" + order->getLabel(); + consensus = new ConcensusCommand(filename); + consensus->execute(); + delete consensus; + + //delete globaldata's tree + //for (int m = 0; m < globaldata->gTree.size(); m++) { delete globaldata->gTree[m]; } + //globaldata->gTree.clear(); + + } + + + + //close ostream for each calc + for (int z = 0; z < treeCalculators.size(); z++) { out[z]->close(); } + + } + catch(exception& e) { + errorOut(e, "BootSharedCommand", "process"); exit(1); - } + } } /***********************************************************/ +