X-Git-Url: https://git.donarmstrong.com/?a=blobdiff_plain;f=clustercommand.cpp;h=8019def31cad9acc54c1308171bf796e949890a1;hb=078d7227da8dda9ae8620822fa32d51ec2706fc3;hp=391b20a9de857755854401f96869c774508bd0af;hpb=348de0f8b17d84ede77081dcf67bd6ef43496677;p=mothur.git diff --git a/clustercommand.cpp b/clustercommand.cpp index 391b20a..8019def 100644 --- a/clustercommand.cpp +++ b/clustercommand.cpp @@ -8,53 +8,85 @@ */ #include "clustercommand.h" +#include "readphylip.h" +#include "readcolumn.h" +#include "readmatrix.hpp" +#include "clusterdoturcommand.h" //********************************************************************************************************************** -vector ClusterCommand::getValidParameters(){ +vector ClusterCommand::setParameters(){ try { - string AlignArray[] = {"cutoff","precision","method","showabund","timing","hard","outputdir","inputdir"}; - vector myArray (AlignArray, AlignArray+(sizeof(AlignArray)/sizeof(string))); + CommandParameter pphylip("phylip", "InputTypes", "", "", "PhylipColumn", "PhylipColumn", "none",false,false); parameters.push_back(pphylip); + CommandParameter pname("name", "InputTypes", "", "", "none", "none", "ColumnName",false,false); parameters.push_back(pname); + CommandParameter pcolumn("column", "InputTypes", "", "", "PhylipColumn", "PhylipColumn", "ColumnName",false,false); parameters.push_back(pcolumn); + CommandParameter pcutoff("cutoff", "Number", "", "10", "", "", "",false,false); parameters.push_back(pcutoff); + CommandParameter pprecision("precision", "Number", "", "100", "", "", "",false,false); parameters.push_back(pprecision); + CommandParameter pmethod("method", "Multiple", "furthest-nearest-average-weighted", "average", "", "", "",false,false); parameters.push_back(pmethod); + CommandParameter pshowabund("showabund", "Boolean", "", "T", "", "", "",false,false); parameters.push_back(pshowabund); + CommandParameter ptiming("timing", "Boolean", "", "F", "", "", "",false,false); parameters.push_back(ptiming); + CommandParameter psim("sim", "Boolean", "", "F", "", "", "",false,false); parameters.push_back(psim); + CommandParameter phard("hard", "Boolean", "", "T", "", "", "",false,false); parameters.push_back(phard); + CommandParameter pinputdir("inputdir", "String", "", "", "", "", "",false,false); parameters.push_back(pinputdir); + CommandParameter poutputdir("outputdir", "String", "", "", "", "", "",false,false); parameters.push_back(poutputdir); + + vector myArray; + for (int i = 0; i < parameters.size(); i++) { myArray.push_back(parameters[i].name); } return myArray; } catch(exception& e) { - m->errorOut(e, "ClusterCommand", "getValidParameters"); + m->errorOut(e, "ClusterCommand", "setParameters"); exit(1); } } //********************************************************************************************************************** -ClusterCommand::ClusterCommand(){ +string ClusterCommand::getHelpString(){ try { - abort = true; calledHelp = true; - vector tempOutNames; - outputTypes["list"] = tempOutNames; - outputTypes["rabund"] = tempOutNames; - outputTypes["sabund"] = tempOutNames; + string helpString = ""; + helpString += "The cluster command parameter options are phylip, column, name, method, cuttoff, hard, precision, sim, showabund and timing. Phylip or column and name are required, unless you have a valid current file.\n"; + helpString += "The cluster command should be in the following format: \n"; + helpString += "cluster(method=yourMethod, cutoff=yourCutoff, precision=yourPrecision) \n"; + helpString += "The acceptable cluster methods are furthest, nearest, average and weighted. If no method is provided then average is assumed.\n"; + return helpString; } catch(exception& e) { - m->errorOut(e, "ClusterCommand", "ClusterCommand"); + m->errorOut(e, "ClusterCommand", "getHelpString"); exit(1); } } //********************************************************************************************************************** -vector ClusterCommand::getRequiredParameters(){ +string ClusterCommand::getOutputFileNameTag(string type, string inputName=""){ try { - vector myArray; - return myArray; + string outputFileName = ""; + map >::iterator it; + + //is this a type this command creates + it = outputTypes.find(type); + if (it == outputTypes.end()) { m->mothurOut("[ERROR]: this command doesn't create a " + type + " output file.\n"); } + else { + if (type == "list") { outputFileName = "list"; } + else if (type == "rabund") { outputFileName = "rabund"; } + else if (type == "sabund") { outputFileName = "sabund"; } + else { m->mothurOut("[ERROR]: No definition for type " + type + " output file tag.\n"); m->control_pressed = true; } + } + return outputFileName; } catch(exception& e) { - m->errorOut(e, "ClusterCommand", "getRequiredParameters"); + m->errorOut(e, "ClusterCommand", "getOutputFileNameTag"); exit(1); } } //********************************************************************************************************************** -vector ClusterCommand::getRequiredFiles(){ +ClusterCommand::ClusterCommand(){ try { - string Array[] = {"phylip","column","or"}; - vector myArray (Array, Array+(sizeof(Array)/sizeof(string))); - return myArray; + abort = true; calledHelp = true; + setParameters(); + vector tempOutNames; + outputTypes["list"] = tempOutNames; + outputTypes["rabund"] = tempOutNames; + outputTypes["sabund"] = tempOutNames; } catch(exception& e) { - m->errorOut(e, "ClusterCommand", "getRequiredFiles"); + m->errorOut(e, "ClusterCommand", "ClusterCommand"); exit(1); } } @@ -62,25 +94,23 @@ vector ClusterCommand::getRequiredFiles(){ //This function checks to make sure the cluster command has no errors and then clusters based on the method chosen. ClusterCommand::ClusterCommand(string option) { try{ - globaldata = GlobalData::getInstance(); - abort = false; calledHelp = false; //allow user to run help if(option == "help") { help(); abort = true; calledHelp = true; } + else if(option == "citation") { citation(); abort = true; calledHelp = true;} else { - //valid paramters for this command - string Array[] = {"cutoff","precision","method","showabund","timing","hard","outputdir","inputdir"}; - vector myArray (Array, Array+(sizeof(Array)/sizeof(string))); + vector myArray = setParameters(); OptionParser parser(option); map parameters = parser.getParameters(); + map::iterator it; ValidParameters validParameter; //check to make sure all parameters are valid for command - for (map::iterator it = parameters.begin(); it != parameters.end(); it++) { + for (it = parameters.begin(); it != parameters.end(); it++) { if (validParameter.isValidParameter(it->first, myArray, it->second) != true) { abort = true; } @@ -95,12 +125,78 @@ ClusterCommand::ClusterCommand(string option) { //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)) { - m->mothurOut("Before you use the cluster command, you first need to read in a distance matrix."); m->mothurOutEndLine(); - abort = true; - } - + string inputDir = validParameter.validFile(parameters, "inputdir", false); + if (inputDir == "not found"){ inputDir = ""; } + else { + string path; + it = parameters.find("phylip"); + //user has given a template file + if(it != parameters.end()){ + path = m->hasPath(it->second); + //if the user has not given a path then, add inputdir. else leave path alone. + if (path == "") { parameters["phylip"] = inputDir + it->second; } + } + + it = parameters.find("column"); + //user has given a template file + if(it != parameters.end()){ + path = m->hasPath(it->second); + //if the user has not given a path then, add inputdir. else leave path alone. + if (path == "") { parameters["column"] = inputDir + it->second; } + } + + it = parameters.find("name"); + //user has given a template file + if(it != parameters.end()){ + path = m->hasPath(it->second); + //if the user has not given a path then, add inputdir. else leave path alone. + if (path == "") { parameters["name"] = inputDir + it->second; } + } + } + + //check for required parameters + phylipfile = validParameter.validFile(parameters, "phylip", true); + if (phylipfile == "not open") { phylipfile = ""; abort = true; } + else if (phylipfile == "not found") { phylipfile = ""; } + else { distfile = phylipfile; format = "phylip"; m->setPhylipFile(phylipfile); } + + columnfile = validParameter.validFile(parameters, "column", true); + if (columnfile == "not open") { columnfile = ""; abort = true; } + else if (columnfile == "not found") { columnfile = ""; } + else { distfile = columnfile; format = "column"; m->setColumnFile(columnfile); } + + namefile = validParameter.validFile(parameters, "name", true); + if (namefile == "not open") { abort = true; } + else if (namefile == "not found") { namefile = ""; } + else { m->setNameFile(namefile); } + + if ((phylipfile == "") && (columnfile == "")) { + //is there are current file available for either of these? + //give priority to column, then phylip + columnfile = m->getColumnFile(); + if (columnfile != "") { distfile = columnfile; format = "column"; m->mothurOut("Using " + columnfile + " as input file for the column parameter."); m->mothurOutEndLine(); } + else { + phylipfile = m->getPhylipFile(); + if (phylipfile != "") { distfile = phylipfile; format = "phylip"; m->mothurOut("Using " + phylipfile + " as input file for the phylip parameter."); m->mothurOutEndLine(); } + else { + m->mothurOut("No valid current files. You must provide a phylip or column file before you can use the cluster command."); m->mothurOutEndLine(); + abort = true; + } + } + } + else if ((phylipfile != "") && (columnfile != "")) { m->mothurOut("When executing a cluster command you must enter ONLY ONE of the following: phylip or column."); m->mothurOutEndLine(); abort = true; } + + if (columnfile != "") { + if (namefile == "") { + namefile = m->getNameFile(); + if (namefile != "") { m->mothurOut("Using " + namefile + " as input file for the name parameter."); m->mothurOutEndLine(); } + else { + m->mothurOut("You need to provide a namefile if you are going to use the column format."); 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 @@ -109,18 +205,21 @@ ClusterCommand::ClusterCommand(string option) { if (temp == "not found") { temp = "100"; } //saves precision legnth for formatting below length = temp.length(); - convert(temp, precision); + m->mothurConvert(temp, precision); - temp = validParameter.validFile(parameters, "hard", false); if (temp == "not found") { temp = "F"; } + temp = validParameter.validFile(parameters, "hard", false); if (temp == "not found") { temp = "T"; } hard = m->isTrue(temp); + temp = validParameter.validFile(parameters, "sim", false); if (temp == "not found") { temp = "F"; } + sim = m->isTrue(temp); + temp = validParameter.validFile(parameters, "cutoff", false); if (temp == "not found") { temp = "10"; } - convert(temp, cutoff); + m->mothurConvert(temp, cutoff); cutoff += (5 / (precision * 10.0)); method = validParameter.validFile(parameters, "method", false); - if (method == "not found") { method = "furthest"; } + if (method == "not found") { method = "average"; } 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; } @@ -131,35 +230,6 @@ ClusterCommand::ClusterCommand(string option) { 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 = globaldata->gSparseMatrix; } - - if(globaldata->gListVector != NULL){ - list = globaldata->gListVector; - rabund = new RAbundVector(list->getRAbundVector()); - } - - //create cluster - 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(); - - if (outputDir == "") { outputDir += m->hasPath(globaldata->inputFileName); } - fileroot = outputDir + m->getRootName(m->getSimpleName(globaldata->inputFileName)); - - m->openOutputFile(fileroot+ tag + ".sabund", sabundFile); - m->openOutputFile(fileroot+ tag + ".rabund", rabundFile); - m->openOutputFile(fileroot+ tag + ".list", listFile); - - outputNames.push_back(fileroot+ tag + ".sabund"); outputTypes["sabund"].push_back(fileroot+ tag + ".sabund"); - outputNames.push_back(fileroot+ tag + ".rabund"); outputTypes["rabund"].push_back(fileroot+ tag + ".rabund"); - outputNames.push_back(fileroot+ tag + ".list"); outputTypes["list"].push_back(fileroot+ tag + ".list"); - } } } catch(exception& e) { @@ -167,32 +237,8 @@ ClusterCommand::ClusterCommand(string option) { exit(1); } } - -//********************************************************************************************************************** - -void ClusterCommand::help(){ - try { - 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) { - m->errorOut(e, "ClusterCommand", "help"); - exit(1); - } -} - //********************************************************************************************************************** - -ClusterCommand::~ClusterCommand(){ - if (abort == false) { - delete cluster; - delete rabund; - } -} - +ClusterCommand::~ClusterCommand(){} //********************************************************************************************************************** int ClusterCommand::execute(){ @@ -200,8 +246,83 @@ int ClusterCommand::execute(){ if (abort == true) { if (calledHelp) { return 0; } return 2; } + //phylip file given and cutoff not given - use cluster.classic because it uses less memory and is faster + if ((format == "phylip") && (cutoff > 10.0)) { + m->mothurOutEndLine(); m->mothurOut("You are using a phylip file and no cutoff. I will run cluster.classic to save memory and time."); m->mothurOutEndLine(); + + //run unique.seqs for deconvolute results + string inputString = "phylip=" + distfile; + if (namefile != "") { inputString += ", name=" + namefile; } + inputString += ", precision=" + toString(precision); + inputString += ", method=" + method; + if (hard) { inputString += ", hard=T"; } + else { inputString += ", hard=F"; } + if (sim) { inputString += ", sim=T"; } + else { inputString += ", sim=F"; } + + + m->mothurOutEndLine(); + m->mothurOut("/------------------------------------------------------------/"); m->mothurOutEndLine(); + m->mothurOut("Running command: cluster.classic(" + inputString + ")"); m->mothurOutEndLine(); + + Command* clusterClassicCommand = new ClusterDoturCommand(inputString); + clusterClassicCommand->execute(); + delete clusterClassicCommand; + + m->mothurOut("/------------------------------------------------------------/"); m->mothurOutEndLine(); + + return 0; + } + + ReadMatrix* read; + if (format == "column") { read = new ReadColumnMatrix(columnfile, sim); } //sim indicates whether its a similarity matrix + else if (format == "phylip") { read = new ReadPhylipMatrix(phylipfile, sim); } + + read->setCutoff(cutoff); + + NameAssignment* nameMap = NULL; + if(namefile != ""){ + nameMap = new NameAssignment(namefile); + nameMap->readMap(); + } + + read->read(nameMap); + list = read->getListVector(); + matrix = read->getMatrix(); + rabund = new RAbundVector(list->getRAbundVector()); + delete read; + + if (m->control_pressed) { //clean up + delete list; delete matrix; delete rabund; + sabundFile.close();rabundFile.close();listFile.close(); + for (int i = 0; i < outputNames.size(); i++) { m->mothurRemove(outputNames[i]); } outputTypes.clear(); + return 0; + } + + //create cluster + 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(); + + if (outputDir == "") { outputDir += m->hasPath(distfile); } + fileroot = outputDir + m->getRootName(m->getSimpleName(distfile)); + + string sabundFileName = fileroot+ tag + "." + getOutputFileNameTag("sabund"); + string rabundFileName = fileroot+ tag + "." + getOutputFileNameTag("rabund"); + string listFileName = fileroot+ tag + "." + getOutputFileNameTag("list"); + + m->openOutputFile(sabundFileName, sabundFile); + m->openOutputFile(rabundFileName, rabundFile); + m->openOutputFile(listFileName, listFile); + + outputNames.push_back(sabundFileName); outputTypes["sabund"].push_back(sabundFileName); + outputNames.push_back(rabundFileName); outputTypes["rabund"].push_back(rabundFileName); + outputNames.push_back(listFileName); outputTypes["list"].push_back(listFileName); + + time_t estart = time(NULL); - //int ndist = matrix->getNNodes(); float previousDist = 0.00000; float rndPreviousDist = 0.00000; oldRAbund = *rabund; @@ -215,12 +336,9 @@ int ClusterCommand::execute(){ while (matrix->getSmallDist() < cutoff && matrix->getNNodes() > 0){ 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(""); } + delete list; delete matrix; delete rabund; delete cluster; sabundFile.close();rabundFile.close();listFile.close(); - for (int i = 0; i < outputNames.size(); i++) { remove(outputNames[i].c_str()); } outputTypes.clear(); + for (int i = 0; i < outputNames.size(); i++) { m->mothurRemove(outputNames[i]); } outputTypes.clear(); return 0; } @@ -263,7 +381,7 @@ int ClusterCommand::execute(){ cout.flush(); print_start = false; } - + if(previousDist <= 0.0000){ printData("unique"); } @@ -271,17 +389,10 @@ int ClusterCommand::execute(){ printData(toString(rndPreviousDist, length-1)); } - //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(""); } - - globaldata->setListFile(fileroot+ tag + ".list"); - globaldata->setNameFile(""); - globaldata->setFormat("list"); + delete matrix; + delete list; + delete rabund; + delete cluster; sabundFile.close(); rabundFile.close();