X-Git-Url: https://git.donarmstrong.com/?a=blobdiff_plain;f=clustercommand.cpp;h=3f27560ae9bfbc69c5ac3d49875847a829f49bf3;hb=e4c80376cc4533f66c8dfc18f3e1a86a60ac17fe;hp=36fff2242ee65d3cdb1e72868d173c69ed5bc3ff;hpb=0470f6d037aacb3563c3f7010708120a4a67d4e6;p=mothur.git diff --git a/clustercommand.cpp b/clustercommand.cpp index 36fff22..3f27560 100644 --- a/clustercommand.cpp +++ b/clustercommand.cpp @@ -11,9 +11,10 @@ //********************************************************************************************************************** //This function checks to make sure the cluster command has no errors and then clusters based on the method chosen. -ClusterCommand::ClusterCommand(string option){ +ClusterCommand::ClusterCommand(string option) { try{ globaldata = GlobalData::getInstance(); + abort = false; //allow user to run help @@ -21,79 +22,93 @@ ClusterCommand::ClusterCommand(string option){ else { //valid paramters for this command - string Array[] = {"cutoff","precision","method"}; + string Array[] = {"cutoff","precision","method","showabund","timing","hard","outputdir","inputdir"}; vector myArray (Array, Array+(sizeof(Array)/sizeof(string))); - parser = new OptionParser(); - parser->parse(option, parameters); delete parser; + OptionParser parser(option); + map parameters = parser.getParameters(); - ValidParameters* validParameter = new ValidParameters(); + ValidParameters validParameter; //check to make sure all parameters are valid for command - for (it = parameters.begin(); it != parameters.end(); it++) { - if (validParameter->isValidParameter(it->first, myArray, it->second) != true) { abort = true; } + for (map::iterator it = parameters.begin(); it != parameters.end(); it++) { + if (validParameter.isValidParameter(it->first, myArray, it->second) != true) { + abort = true; + } } + //if the user changes the output directory command factory will send this info to us in the output parameter + outputDir = validParameter.validFile(parameters, "outputdir", false); if (outputDir == "not found"){ outputDir = ""; } + //error checking to make sure they read a distance file if ((globaldata->gSparseMatrix == NULL) || (globaldata->gListVector == NULL)) { - cout << "Before you use the cluster command, you first need to read in a distance matrix." << endl; abort = true; + m->mothurOut("Before you use the cluster command, you first need to read in a distance matrix."); m->mothurOutEndLine(); + abort = true; } //check for optional parameter and set defaults // ...at some point should added some additional type checking... //get user cutoff and precision or use defaults string temp; - temp = validParameter->validFile(parameters, "precision", false); if (temp == "not found") { temp = "100"; } + temp = validParameter.validFile(parameters, "precision", false); + if (temp == "not found") { temp = "100"; } //saves precision legnth for formatting below length = temp.length(); convert(temp, precision); - temp = validParameter->validFile(parameters, "cutoff", false); if (temp == "not found") { temp = "10"; } + temp = validParameter.validFile(parameters, "hard", false); if (temp == "not found") { temp = "F"; } + hard = isTrue(temp); + + temp = validParameter.validFile(parameters, "cutoff", false); + if (temp == "not found") { temp = "10"; } convert(temp, cutoff); - cutoff += (5 / (precision * 10.0)); + cutoff += (5 / (precision * 10.0)); - method = validParameter->validFile(parameters, "method", false); if (method == "not found") { method = "furthest"; } - - delete validParameter; + method = validParameter.validFile(parameters, "method", false); + if (method == "not found") { method = "furthest"; } - if ((method == "furthest") || (method == "nearest") || (method == "average")) { } - else {cout << "Not a valid clustering method. Valid clustering algorithms are furthest, nearest or average." << endl; abort = true; } + if ((method == "furthest") || (method == "nearest") || (method == "average") || (method == "weighted")) { } + else { m->mothurOut("Not a valid clustering method. Valid clustering algorithms are furthest, nearest, average, and weighted."); m->mothurOutEndLine(); abort = true; } + + showabund = validParameter.validFile(parameters, "showabund", false); + if (showabund == "not found") { showabund = "T"; } + timing = validParameter.validFile(parameters, "timing", false); + if (timing == "not found") { timing = "F"; } if (abort == false) { - //get matrix, list and rabund for execute - if(globaldata->gSparseMatrix != NULL) { matrix = new SparseMatrix(*globaldata->gSparseMatrix); } + + //get matrix, list and rabund for execute + if(globaldata->gSparseMatrix != NULL) { matrix = globaldata->gSparseMatrix; } if(globaldata->gListVector != NULL){ - list = new ListVector(*globaldata->gListVector); + list = globaldata->gListVector; rabund = new RAbundVector(list->getRAbundVector()); } //create cluster - if(method == "furthest") { cluster = new CompleteLinkage(rabund, list, matrix); tag = "fn"; } - else if(method == "nearest"){ cluster = new SingleLinkage(rabund, list, matrix); tag = "nn"; } - else if(method == "average"){ cluster = new AverageLinkage(rabund, list, matrix); tag = "an"; } - else { cout << "error - not recognized method" << endl; abort = true; } + if (method == "furthest") { cluster = new CompleteLinkage(rabund, list, matrix, cutoff, method); } + else if(method == "nearest"){ cluster = new SingleLinkage(rabund, list, matrix, cutoff, method); } + else if(method == "average"){ cluster = new AverageLinkage(rabund, list, matrix, cutoff, method); } + else if(method == "weighted"){ cluster = new WeightedLinkage(rabund, list, matrix, cutoff, method); } + tag = cluster->getTag(); - fileroot = getRootName(globaldata->inputFileName); + if (outputDir == "") { outputDir += hasPath(globaldata->inputFileName); } + fileroot = outputDir + getRootName(getSimpleName(globaldata->inputFileName)); openOutputFile(fileroot+ tag + ".sabund", sabundFile); openOutputFile(fileroot+ tag + ".rabund", rabundFile); openOutputFile(fileroot+ tag + ".list", listFile); - + outputNames.push_back(fileroot+ tag + ".sabund"); + outputNames.push_back(fileroot+ tag + ".rabund"); + outputNames.push_back(fileroot+ tag + ".list"); } - } - } catch(exception& e) { - cout << "Standard Error: " << e.what() << " has occurred in the ClusterCommand class Function ClusterCommand. Please contact Pat Schloss at pschloss@microbio.umass.edu." << "\n"; - exit(1); - } - catch(...) { - cout << "An unknown error has occurred in the ClusterCommand class function ClusterCommand. Please contact Pat Schloss at pschloss@microbio.umass.edu." << "\n"; + m->errorOut(e, "ClusterCommand", "ClusterCommand"); exit(1); } } @@ -102,29 +117,25 @@ ClusterCommand::ClusterCommand(string option){ void ClusterCommand::help(){ try { - cout << "The cluster command can only be executed after a successful read.dist command." << "\n"; - cout << "The cluster command parameter options are method, cuttoff and precision. No parameters are required." << "\n"; - cout << "The cluster command should be in the following format: " << "\n"; - cout << "cluster(method=yourMethod, cutoff=yourCutoff, precision=yourPrecision) " << "\n"; - cout << "The acceptable cluster methods are furthest, nearest and average. If no method is provided then furthest is assumed." << "\n" << "\n"; + m->mothurOut("The cluster command can only be executed after a successful read.dist command.\n"); + m->mothurOut("The cluster command parameter options are method, cuttoff, hard, precision, showabund and timing. No parameters are required.\n"); + m->mothurOut("The cluster command should be in the following format: \n"); + m->mothurOut("cluster(method=yourMethod, cutoff=yourCutoff, precision=yourPrecision) \n"); + m->mothurOut("The acceptable cluster methods are furthest, nearest and average. If no method is provided then furthest is assumed.\n\n"); } catch(exception& e) { - cout << "Standard Error: " << e.what() << " has occurred in the ClusterCommand class Function help. Please contact Pat Schloss at pschloss@microbio.umass.edu." << "\n"; + m->errorOut(e, "ClusterCommand", "help"); exit(1); } - catch(...) { - cout << "An unknown error has occurred in the ClusterCommand class function help. Please contact Pat Schloss at pschloss@microbio.umass.edu." << "\n"; - exit(1); - } } //********************************************************************************************************************** ClusterCommand::~ClusterCommand(){ - delete cluster; - delete matrix; - delete list; - delete rabund; + if (abort == false) { + delete cluster; + delete rabund; + } } //********************************************************************************************************************** @@ -134,19 +145,49 @@ int ClusterCommand::execute(){ if (abort == true) { return 0; } + time_t estart = time(NULL); + //int ndist = matrix->getNNodes(); float previousDist = 0.00000; float rndPreviousDist = 0.00000; oldRAbund = *rabund; oldList = *list; + + print_start = true; + start = time(NULL); + loops = 0; + double saveCutoff = cutoff; + + while (matrix->getSmallDist() < cutoff && matrix->getNNodes() > 0){ - float x; - x=0.1; - toString(x, 2); + if (m->control_pressed) { //clean up + delete globaldata->gSparseMatrix; globaldata->gSparseMatrix = NULL; + delete globaldata->gListVector; globaldata->gListVector = NULL; + if (globaldata->getFormat() == "phylip") { globaldata->setPhylipFile(""); } + else if (globaldata->getFormat() == "column") { globaldata->setColumnFile(""); } + sabundFile.close();rabundFile.close();listFile.close(); + for (int i = 0; i < outputNames.size(); i++) { remove(outputNames[i].c_str()); } + return 0; + } + + if (print_start && isTrue(timing)) { + m->mothurOut("Clustering (" + tag + ") dist " + toString(matrix->getSmallDist()) + "/" + + toString(roundDist(matrix->getSmallDist(), precision)) + + "\t(precision: " + toString(precision) + ", Nodes: " + toString(matrix->getNNodes()) + ")"); + cout.flush(); + print_start = false; + } + + loops++; + + cluster->update(cutoff); - while(matrix->getSmallDist() < cutoff && matrix->getNNodes() > 0){ - cluster->update(); float dist = matrix->getSmallDist(); - float rndDist = roundDist(dist, precision); + float rndDist; + if (hard) { + rndDist = ceilDist(dist, precision); + }else{ + rndDist = roundDist(dist, precision); + } if(previousDist <= 0.0000 && dist != previousDist){ printData("unique"); @@ -160,6 +201,13 @@ int ClusterCommand::execute(){ oldRAbund = *rabund; oldList = *list; } + + if (print_start && isTrue(timing)) { + m->mothurOut("Clustering (" + tag + ") for distance " + toString(previousDist) + "/" + toString(rndPreviousDist) + + "\t(precision: " + toString(precision) + ", Nodes: " + toString(matrix->getNNodes()) + ")"); + cout.flush(); + print_start = false; + } if(previousDist <= 0.0000){ printData("unique"); @@ -171,7 +219,7 @@ int ClusterCommand::execute(){ //delete globaldata's copy of the sparsematrix and listvector to free up memory delete globaldata->gSparseMatrix; globaldata->gSparseMatrix = NULL; delete globaldata->gListVector; globaldata->gListVector = NULL; - + //saves .list file so you can do the collect, rarefaction and summary commands without doing a read.list if (globaldata->getFormat() == "phylip") { globaldata->setPhylipFile(""); } else if (globaldata->getFormat() == "column") { globaldata->setColumnFile(""); } @@ -180,25 +228,52 @@ int ClusterCommand::execute(){ globaldata->setNameFile(""); globaldata->setFormat("list"); + sabundFile.close(); + rabundFile.close(); + listFile.close(); + + if (saveCutoff != cutoff) { + if (hard) { saveCutoff = ceilDist(saveCutoff, precision); } + else { saveCutoff = roundDist(saveCutoff, precision); } + + m->mothurOut("changed cutoff to " + toString(cutoff)); m->mothurOutEndLine(); + } + + m->mothurOutEndLine(); + m->mothurOut("Output File Names: "); m->mothurOutEndLine(); + for (int i = 0; i < outputNames.size(); i++) { m->mothurOut(outputNames[i]); m->mothurOutEndLine(); } + m->mothurOutEndLine(); + + + //if (isTrue(timing)) { + m->mothurOut("It took " + toString(time(NULL) - estart) + " seconds to cluster"); m->mothurOutEndLine(); + //} + + return 0; } catch(exception& e) { - cout << "Standard Error: " << e.what() << " has occurred in the ClusterCommand class Function execute. Please contact Pat Schloss at pschloss@microbio.umass.edu." << "\n"; - exit(1); - } - catch(...) { - cout << "An unknown error has occurred in the ClusterCommand class function execute. Please contact Pat Schloss at pschloss@microbio.umass.edu." << "\n"; + m->errorOut(e, "ClusterCommand", "execute"); exit(1); } - } //********************************************************************************************************************** void ClusterCommand::printData(string label){ try { + if (isTrue(timing)) { + m->mothurOut("\tTime: " + toString(time(NULL) - start) + "\tsecs for " + toString(oldRAbund.getNumBins()) + + "\tclusters. Updates: " + toString(loops)); m->mothurOutEndLine(); + } + print_start = true; + loops = 0; + start = time(NULL); + oldRAbund.setLabel(label); - oldRAbund.getSAbundVector().print(cout); + if (isTrue(showabund)) { + oldRAbund.getSAbundVector().print(cout); + } oldRAbund.print(rabundFile); oldRAbund.getSAbundVector().print(sabundFile); @@ -206,13 +281,10 @@ void ClusterCommand::printData(string label){ oldList.print(listFile); } catch(exception& e) { - cout << "Standard Error: " << e.what() << " has occurred in the ClusterCommand class Function printData. Please contact Pat Schloss at pschloss@microbio.umass.edu." << "\n"; - exit(1); - } - catch(...) { - cout << "An unknown error has occurred in the ClusterCommand class function printData. Please contact Pat Schloss at pschloss@microbio.umass.edu." << "\n"; + m->errorOut(e, "ClusterCommand", "printData"); exit(1); } + } //**********************************************************************************************************************