X-Git-Url: https://git.donarmstrong.com/?a=blobdiff_plain;f=matrixoutputcommand.cpp;h=35a45532dc64d8d11ab415e777b3afe67ae94057;hb=19fcbbdba99658f5eca244803280f9ee7f9f6607;hp=baf449e34d8bac9e67ab424ba02539fb053033e6;hpb=c69e2e9749626cfbf1d6cb0125ae94f869e00b18;p=mothur.git diff --git a/matrixoutputcommand.cpp b/matrixoutputcommand.cpp index baf449e..35a4553 100644 --- a/matrixoutputcommand.cpp +++ b/matrixoutputcommand.cpp @@ -49,50 +49,61 @@ #include "memchord.h" #include "memeuclidean.h" #include "mempearson.h" + //********************************************************************************************************************** -vector MatrixOutputCommand::getValidParameters(){ +vector MatrixOutputCommand::setParameters(){ try { - string Array[] = {"label","calc","groups","outputdir","inputdir", "output"}; - vector myArray (Array, Array+(sizeof(Array)/sizeof(string))); + CommandParameter pshared("shared", "InputTypes", "", "", "none", "none", "none",false,true); parameters.push_back(pshared); + CommandParameter plabel("label", "String", "", "", "", "", "",false,false); parameters.push_back(plabel); + CommandParameter pgroups("groups", "String", "", "", "", "", "",false,false); parameters.push_back(pgroups); + CommandParameter pcalc("calc", "Multiple", "sharedsobs-sharedchao-sharedace-jabund-sorabund-jclass-sorclass-jest-sorest-thetayc-thetan-kstest-sharednseqs-ochiai-anderberg-kulczynski-kulczynskicody-lennon-morisitahorn-braycurtis-whittaker-odum-canberra-structeuclidean-structchord-hellinger-manhattan-structpearson-soergel-spearman-structkulczynski-speciesprofile-hamming-structchi2-gower-memchi2-memchord-memeuclidean-mempearson", "jclass-thetayc", "", "", "",true,false); parameters.push_back(pcalc); + CommandParameter poutput("output", "Multiple", "lt-square", "lt", "", "", "",false,false); parameters.push_back(poutput); + CommandParameter pprocessors("processors", "Number", "", "1", "", "", "",false,false); parameters.push_back(pprocessors); + 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, "MatrixOutputCommand", "getValidParameters"); - exit(1); - } -} -//********************************************************************************************************************** -MatrixOutputCommand::MatrixOutputCommand(){ - try { - abort = true; calledHelp = true; - vector tempOutNames; - outputTypes["phylip"] = tempOutNames; - } - catch(exception& e) { - m->errorOut(e, "MatrixOutputCommand", "MatrixOutputCommand"); + m->errorOut(e, "MatrixOutputCommand", "setParameters"); exit(1); } } //********************************************************************************************************************** -vector MatrixOutputCommand::getRequiredParameters(){ +string MatrixOutputCommand::getHelpString(){ try { - vector myArray; - return myArray; + string helpString = ""; + ValidCalculators validCalculator; + helpString += "The dist.shared command parameters are shared, groups, calc, output, processors and label. shared is a required, unless you have a valid current file.\n"; + helpString += "The groups parameter allows you to specify which of the groups in your groupfile you would like included used.\n"; + helpString += "The group names are separated by dashes. The label parameter allows you to select what distance levels you would like distance matrices created for, and is also separated by dashes.\n"; + helpString += "The dist.shared command should be in the following format: dist.shared(groups=yourGroups, calc=yourCalcs, label=yourLabels).\n"; + helpString += "The output parameter allows you to specify format of your distance matrix. Options are lt, and square. The default is lt.\n"; + helpString += "Example dist.shared(groups=A-B-C, calc=jabund-sorabund).\n"; + helpString += "The default value for groups is all the groups in your groupfile.\n"; + helpString += "The default value for calc is jclass and thetayc.\n"; + helpString += validCalculator.printCalc("matrix"); + helpString += "The dist.shared command outputs a .dist file for each calculator you specify at each distance you choose.\n"; + helpString += "Note: No spaces between parameter labels (i.e. groups), '=' and parameters (i.e.yourGroups).\n"; + return helpString; } catch(exception& e) { - m->errorOut(e, "MatrixOutputCommand", "getRequiredParameters"); + m->errorOut(e, "MatrixOutputCommand", "getHelpString"); exit(1); } } //********************************************************************************************************************** -vector MatrixOutputCommand::getRequiredFiles(){ +MatrixOutputCommand::MatrixOutputCommand(){ try { - string Array[] = {"shared"}; - vector myArray (Array, Array+(sizeof(Array)/sizeof(string))); - return myArray; + abort = true; calledHelp = true; + setParameters(); + vector tempOutNames; + outputTypes["phylip"] = tempOutNames; } catch(exception& e) { - m->errorOut(e, "MatrixOutputCommand", "getRequiredFiles"); + m->errorOut(e, "MatrixOutputCommand", "MatrixOutputCommand"); exit(1); } } @@ -100,28 +111,24 @@ vector MatrixOutputCommand::getRequiredFiles(){ MatrixOutputCommand::MatrixOutputCommand(string option) { try { - globaldata = GlobalData::getInstance(); abort = false; calledHelp = false; allLines = 1; - labels.clear(); - Groups.clear(); - Estimators.clear(); - + //allow user to run help - if(option == "help") { validCalculator = new ValidCalculators(); help(); abort = true; calledHelp = true; } + 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[] = {"label","calc","groups","outputdir","inputdir", "output"}; - 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; } } @@ -129,16 +136,33 @@ MatrixOutputCommand::MatrixOutputCommand(string option) { vector tempOutNames; outputTypes["phylip"] = tempOutNames; + //if the user changes the input directory command factory will send this info to us in the output parameter + string inputDir = validParameter.validFile(parameters, "inputdir", false); + if (inputDir == "not found"){ inputDir = ""; } + else { + string path; + it = parameters.find("shared"); + //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["shared"] = inputDir + it->second; } + } + } + + sharedfile = validParameter.validFile(parameters, "shared", true); + if (sharedfile == "not found") { + //if there is a current shared file, use it + sharedfile = m->getSharedFile(); + if (sharedfile != "") { m->mothurOut("Using " + sharedfile + " as input file for the shared parameter."); m->mothurOutEndLine(); } + else { m->mothurOut("You have no current sharedfile and the shared parameter is required."); m->mothurOutEndLine(); abort = true; } + }else if (sharedfile == "not open") { sharedfile = ""; abort = true; } + else { m->setSharedFile(sharedfile); } + //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 = ""; - outputDir += m->hasPath(globaldata->inputFileName); //if user entered a file with a path then preserve it - } - - //make sure the user has already run the read.otu command - if (globaldata->getSharedFile() == "") { - if (globaldata->getListFile() == "") { m->mothurOut("You must read a list and a group, or a shared before you can use the dist.shared command."); m->mothurOutEndLine(); abort = true; } - else if (globaldata->getGroupFile() == "") { m->mothurOut("You must read a list and a group, or a shared before you can use the dist.shared command."); m->mothurOutEndLine(); abort = true; } + outputDir += m->hasPath(sharedfile); //if user entered a file with a path then preserve it } //check for optional parameter and set defaults @@ -153,18 +177,16 @@ MatrixOutputCommand::MatrixOutputCommand(string option) { output = validParameter.validFile(parameters, "output", false); if(output == "not found"){ output = "lt"; } if ((output != "lt") && (output != "square")) { m->mothurOut(output + " is not a valid output form. Options are lt and square. I will use lt."); m->mothurOutEndLine(); output = "lt"; } - //if the user has not specified any labels use the ones from read.otu - if (label == "") { - allLines = globaldata->allLines; - labels = globaldata->labels; - } - groups = validParameter.validFile(parameters, "groups", false); if (groups == "not found") { groups = ""; } else { m->splitAtDash(groups, Groups); - globaldata->Groups = Groups; + m->Groups = Groups; } + + string temp = validParameter.validFile(parameters, "processors", false); if (temp == "not found"){ temp = m->getProcessors(); } + m->setProcessors(temp); + convert(temp, processors); calc = validParameter.validFile(parameters, "calc", false); if (calc == "not found") { calc = "jclass-thetayc"; } @@ -172,14 +194,19 @@ MatrixOutputCommand::MatrixOutputCommand(string option) { if (calc == "default") { calc = "jclass-thetayc"; } } m->splitAtDash(calc, Estimators); + if (m->inUsersGroups("citation", Estimators)) { + ValidCalculators validCalc; validCalc.printCitations(Estimators); + //remove citation from list of calcs + for (int i = 0; i < Estimators.size(); i++) { if (Estimators[i] == "citation") { Estimators.erase(Estimators.begin()+i); break; } } + } if (abort == false) { - validCalculator = new ValidCalculators(); + ValidCalculators validCalculator; int i; for (i=0; iisValidCalculator("matrix", Estimators[i]) == true) { + if (validCalculator.isValidCalculator("matrix", Estimators[i]) == true) { if (Estimators[i] == "sharedsobs") { matrixCalculators.push_back(new SharedSobsCS()); }else if (Estimators[i] == "sharedchao") { @@ -274,37 +301,7 @@ MatrixOutputCommand::MatrixOutputCommand(string option) { //********************************************************************************************************************** -void MatrixOutputCommand::help(){ - try { - m->mothurOut("The dist.shared command can only be executed after a successful read.otu command.\n"); - m->mothurOut("The dist.shared command parameters are groups, calc, output and label. No parameters are required.\n"); - m->mothurOut("The groups parameter allows you to specify which of the groups in your groupfile you would like included used.\n"); - m->mothurOut("The group names are separated by dashes. The label parameter allows you to select what distance levels you would like distance matrices created for, and is also separated by dashes.\n"); - m->mothurOut("The dist.shared command should be in the following format: dist.shared(groups=yourGroups, calc=yourCalcs, label=yourLabels).\n"); - m->mothurOut("The output parameter allows you to specify format of your distance matrix. Options are lt, and square. The default is lt.\n"); - m->mothurOut("Example dist.shared(groups=A-B-C, calc=jabund-sorabund).\n"); - m->mothurOut("The default value for groups is all the groups in your groupfile.\n"); - m->mothurOut("The default value for calc is jclass and thetayc.\n"); - validCalculator->printCalc("matrix", cout); - m->mothurOut("The dist.shared command outputs a .dist file for each calculator you specify at each distance you choose.\n"); - m->mothurOut("Note: No spaces between parameter labels (i.e. groups), '=' and parameters (i.e.yourGroups).\n\n"); - } - catch(exception& e) { - m->errorOut(e, "MatrixOutputCommand", "help"); - exit(1); - } -} - - -//********************************************************************************************************************** - -MatrixOutputCommand::~MatrixOutputCommand(){ - if (abort == false) { - delete input; globaldata->ginput = NULL; - delete read; - delete validCalculator; - } -} +MatrixOutputCommand::~MatrixOutputCommand(){} //********************************************************************************************************************** @@ -315,12 +312,8 @@ int MatrixOutputCommand::execute(){ //if the users entered no valid calculators don't execute command if (matrixCalculators.size() == 0) { m->mothurOut("No valid calculators."); m->mothurOutEndLine(); return 0; } - - //you have groups - read = new ReadOTUFile(globaldata->inputFileName); - read->read(&*globaldata); - input = globaldata->ginput; + input = new InputData(sharedfile, "sharedfile"); lookup = input->getSharedRAbundVectors(); string lastLabel = lookup[0]->getLabel(); @@ -328,16 +321,21 @@ int MatrixOutputCommand::execute(){ set processedLabels; set userLabels = labels; - if (lookup.size() < 2) { m->mothurOut("You have not provided enough valid groups. I cannot run the command."); m->mothurOutEndLine(); return 0;} + if (lookup.size() < 2) { m->mothurOut("You have not provided enough valid groups. I cannot run the command."); m->mothurOutEndLine(); delete input; for (int i = 0; i < lookup.size(); i++) { delete lookup[i]; } return 0;} numGroups = lookup.size(); + lines.resize(processors); + for (int i = 0; i < processors; i++) { + lines[i].start = int (sqrt(float(i)/float(processors)) * numGroups); + lines[i].end = int (sqrt(float(i+1)/float(processors)) * numGroups); + } - if (m->control_pressed) { delete read; delete input; globaldata->ginput = NULL; for (int i = 0; i < lookup.size(); i++) { delete lookup[i]; } globaldata->Groups.clear(); return 0; } + if (m->control_pressed) { delete input; for (int i = 0; i < lookup.size(); i++) { delete lookup[i]; } m->Groups.clear(); return 0; } //as long as you are not at the end of the file or done wih the lines you want while((lookup[0] != NULL) && ((allLines == 1) || (userLabels.size() != 0))) { - if (m->control_pressed) { outputTypes.clear(); delete read; delete input; globaldata->ginput = NULL; for (int i = 0; i < lookup.size(); i++) { delete lookup[i]; } for (int i = 0; i < outputNames.size(); i++) { remove(outputNames[i].c_str()); } globaldata->Groups.clear(); return 0; } + if (m->control_pressed) { outputTypes.clear(); delete input; for (int i = 0; i < lookup.size(); i++) { delete lookup[i]; } for (int i = 0; i < outputNames.size(); i++) { m->mothurRemove(outputNames[i]); } m->Groups.clear(); return 0; } if(allLines == 1 || labels.count(lookup[0]->getLabel()) == 1){ m->mothurOut(lookup[0]->getLabel()); m->mothurOutEndLine(); @@ -370,7 +368,7 @@ int MatrixOutputCommand::execute(){ lookup = input->getSharedRAbundVectors(); } - if (m->control_pressed) { outputTypes.clear(); delete read; delete input; globaldata->ginput = NULL; for (int i = 0; i < outputNames.size(); i++) { remove(outputNames[i].c_str()); } globaldata->Groups.clear(); return 0; } + if (m->control_pressed) { outputTypes.clear(); delete input; for (int i = 0; i < outputNames.size(); i++) { m->mothurRemove(outputNames[i]); } m->Groups.clear(); return 0; } //output error messages about any remaining user labels set::iterator it; @@ -385,7 +383,7 @@ int MatrixOutputCommand::execute(){ } } - if (m->control_pressed) { outputTypes.clear(); delete read; delete input; globaldata->ginput = NULL; for (int i = 0; i < outputNames.size(); i++) { remove(outputNames[i].c_str()); } globaldata->Groups.clear(); return 0; } + if (m->control_pressed) { outputTypes.clear(); delete input; for (int i = 0; i < outputNames.size(); i++) { m->mothurRemove(outputNames[i]); } m->Groups.clear(); return 0; } //run last label if you need to if (needToRun == true) { @@ -397,10 +395,17 @@ int MatrixOutputCommand::execute(){ for (int i = 0; i < lookup.size(); i++) { delete lookup[i]; } } - if (m->control_pressed) { outputTypes.clear(); delete read; delete input; globaldata->ginput = NULL; for (int i = 0; i < outputNames.size(); i++) { remove(outputNames[i].c_str()); } globaldata->Groups.clear(); return 0; } + if (m->control_pressed) { outputTypes.clear(); delete input; for (int i = 0; i < outputNames.size(); i++) { m->mothurRemove(outputNames[i]); } m->Groups.clear(); return 0; } //reset groups parameter - globaldata->Groups.clear(); + m->Groups.clear(); + + //set phylip file as new current phylipfile + string current = ""; + itTypes = outputTypes.find("phylip"); + if (itTypes != outputTypes.end()) { + if ((itTypes->second).size() != 0) { current = (itTypes->second)[0]; m->setPhylipFile(current); } + } m->mothurOutEndLine(); m->mothurOut("Output File Names: "); m->mothurOutEndLine(); @@ -416,7 +421,7 @@ int MatrixOutputCommand::execute(){ } } /***********************************************************/ -void MatrixOutputCommand::printSims(ostream& out) { +void MatrixOutputCommand::printSims(ostream& out, vector< vector >& simMatrix) { try { out.setf(ios::fixed, ios::floatfield); out.setf(ios::showpoint); @@ -450,63 +455,160 @@ void MatrixOutputCommand::printSims(ostream& out) { /***********************************************************/ int MatrixOutputCommand::process(vector thisLookup){ try { - - EstOutput data; - vector subset; - - //for each calculator - for(int i = 0 ; i < matrixCalculators.size(); i++) { + EstOutput data; + vector subset; + vector< vector > calcDists; calcDists.resize(matrixCalculators.size()); //one for each calc, this will be used to make .dist files + + #if defined (__APPLE__) || (__MACH__) || (linux) || (__linux) + if(processors == 1){ + driver(thisLookup, 0, numGroups, calcDists); + }else{ + int process = 1; + vector processIDS; + + //loop through and create all the processes you want + while (process != processors) { + int pid = fork(); + + if (pid > 0) { + processIDS.push_back(pid); + process++; + }else if (pid == 0){ + driver(thisLookup, lines[process].start, lines[process].end, calcDists); - //initialize simMatrix - simMatrix.clear(); - simMatrix.resize(numGroups); - for (int p = 0; p < simMatrix.size(); p++) { - for (int j = 0; j < simMatrix.size(); j++) { - simMatrix[p].push_back(0.0); + string tempdistFileName = m->getRootName(m->getSimpleName(sharedfile)) + toString(getpid()) + ".dist"; + ofstream outtemp; + m->openOutputFile(tempdistFileName, outtemp); + + for (int i = 0; i < calcDists.size(); i++) { + outtemp << calcDists[i].size() << endl; + + for (int j = 0; j < calcDists[i].size(); j++) { + outtemp << calcDists[i][j].seq1 << '\t' << calcDists[i][j].seq2 << '\t' << calcDists[i][j].dist << endl; } } + outtemp.close(); + + exit(0); + }else { + m->mothurOut("[ERROR]: unable to spawn the necessary processes."); m->mothurOutEndLine(); + for (int i = 0; i < processIDS.size(); i++) { kill (processIDS[i], SIGINT); } + exit(0); + } + } + + //parent do your part + driver(thisLookup, lines[0].start, lines[0].end, calcDists); + + //force parent to wait until all the processes are done + for (int i = 0; i < processIDS.size(); i++) { + int temp = processIDS[i]; + wait(&temp); + } + + for (int i = 0; i < processIDS.size(); i++) { + string tempdistFileName = m->getRootName(m->getSimpleName(sharedfile)) + toString(processIDS[i]) + ".dist"; + ifstream intemp; + m->openInputFile(tempdistFileName, intemp); + + for (int k = 0; k < calcDists.size(); k++) { + int size = 0; + intemp >> size; m->gobble(intemp); + + for (int j = 0; j < size; j++) { + int seq1 = 0; + int seq2 = 0; + float dist = 1.0; + + intemp >> seq1 >> seq2 >> dist; m->gobble(intemp); + + seqDist tempDist(seq1, seq2, dist); + calcDists[k].push_back(tempDist); + } + } + intemp.close(); + m->mothurRemove(tempdistFileName); + } + + } +#else + driver(thisLookup, 0, numGroups, calcDists); +#endif + + for (int i = 0; i < calcDists.size(); i++) { + if (m->control_pressed) { break; } + + //initialize matrix + vector< vector > matrix; //square matrix to represent the distance + matrix.resize(thisLookup.size()); + for (int k = 0; k < thisLookup.size(); k++) { matrix[k].resize(thisLookup.size(), 0.0); } - 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 - - if (m->control_pressed) { return 0; } - - subset.clear(); //clear out old pair of sharedrabunds - //add new pair of sharedrabunds - subset.push_back(thisLookup[k]); subset.push_back(thisLookup[l]); - - //if this calc needs all groups to calculate the pair load all groups - if (matrixCalculators[i]->getNeedsAll()) { - //load subset with rest of lookup for those calcs that need everyone to calc for a pair - for (int w = 0; w < thisLookup.size(); w++) { - if ((w != k) && (w != l)) { subset.push_back(thisLookup[w]); } - } - } - - data = matrixCalculators[i]->getValues(subset); //saves the calculator outputs - //save values in similarity matrix - simMatrix[k][l] = 1.0 - data[0]; //convert similiarity to distance - simMatrix[l][k] = 1.0 - data[0]; //convert similiarity to distance + for (int j = 0; j < calcDists[i].size(); j++) { + int row = calcDists[i][j].seq1; + int column = calcDists[i][j].seq2; + float dist = calcDists[i][j].dist; + + matrix[row][column] = dist; + matrix[column][row] = dist; + } + + string distFileName = outputDir + m->getRootName(m->getSimpleName(sharedfile)) + matrixCalculators[i]->getName() + "." + thisLookup[0]->getLabel() + "." + output + ".dist"; + outputNames.push_back(distFileName); outputTypes["phylip"].push_back(distFileName); + ofstream outDist; + m->openOutputFile(distFileName, outDist); + outDist.setf(ios::fixed, ios::floatfield); outDist.setf(ios::showpoint); + + printSims(outDist, matrix); + + outDist.close(); + } + + return 0; + } + catch(exception& e) { + m->errorOut(e, "MatrixOutputCommand", "process"); + exit(1); + } +} +/**************************************************************************************************/ +int MatrixOutputCommand::driver(vector thisLookup, int start, int end, vector< vector >& calcDists) { + try { + + vector subset; + for (int k = start; k < end; k++) { // pass cdd each set of groups to compare + + for (int l = 0; l < k; 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(thisLookup[k]); subset.push_back(thisLookup[l]); + + for(int i=0;igetNeedsAll()) { + //load subset with rest of lookup for those calcs that need everyone to calc for a pair + for (int w = 0; w < thisLookup.size(); w++) { + if ((w != k) && (w != l)) { subset.push_back(thisLookup[w]); } } } + + vector tempdata = matrixCalculators[i]->getValues(subset); //saves the calculator outputs + + if (m->control_pressed) { return 1; } + + seqDist temp(l, k, tempdata[0]); + calcDists[i].push_back(temp); } - - exportFileName = outputDir + m->getRootName(m->getSimpleName(globaldata->inputFileName)) + matrixCalculators[i]->getName() + "." + thisLookup[0]->getLabel() + "." + output + ".dist"; - m->openOutputFile(exportFileName, out); - outputNames.push_back(exportFileName); outputTypes["phylip"].push_back(exportFileName); - - printSims(out); - out.close(); - } - - return 0; + } + } + return 0; } catch(exception& e) { - m->errorOut(e, "MatrixOutputCommand", "process"); + m->errorOut(e, "MatrixOutputCommand", "driver"); exit(1); } }