X-Git-Url: https://git.donarmstrong.com/?a=blobdiff_plain;f=subsamplecommand.cpp;h=717b1d3231c20368a2d23e9b86d52a6707f03b12;hb=2c97dd48b8e27ee0a6a86c7a082f4c504c3357c6;hp=7a857e8df70d4ab99328e3988e1f23faf9837a49;hpb=3a13eff5c26d6fc156a299c9fa7f5497bded94a0;p=mothur.git diff --git a/subsamplecommand.cpp b/subsamplecommand.cpp index 7a857e8..717b1d3 100644 --- a/subsamplecommand.cpp +++ b/subsamplecommand.cpp @@ -9,24 +9,65 @@ #include "subsamplecommand.h" #include "sharedutilities.h" +#include "deconvolutecommand.h" +#include "subsample.h" //********************************************************************************************************************** -vector SubSampleCommand::getValidParameters(){ - try { - string Array[] = {"fasta", "group", "list","shared","rabund", "name","sabund","size","groups","label","outputdir","inputdir"}; - vector myArray (Array, Array+(sizeof(Array)/sizeof(string))); +vector SubSampleCommand::setParameters(){ + try { + CommandParameter pfasta("fasta", "InputTypes", "", "", "none", "FLSSR", "none",false,false); parameters.push_back(pfasta); + CommandParameter pname("name", "InputTypes", "", "", "none", "none", "none",false,false); parameters.push_back(pname); + CommandParameter pgroup("group", "InputTypes", "", "", "none", "none", "none",false,false); parameters.push_back(pgroup); + CommandParameter plist("list", "InputTypes", "", "", "none", "FLSSR", "none",false,false); parameters.push_back(plist); + CommandParameter pshared("shared", "InputTypes", "", "", "none", "FLSSR", "none",false,false); parameters.push_back(pshared); + CommandParameter prabund("rabund", "InputTypes", "", "", "none", "FLSSR", "none",false,false); parameters.push_back(prabund); + CommandParameter psabund("sabund", "InputTypes", "", "", "none", "FLSSR", "none",false,false); parameters.push_back(psabund); + CommandParameter plabel("label", "String", "", "", "", "", "",false,false); parameters.push_back(plabel); + CommandParameter pgroups("groups", "String", "", "", "", "", "",false,false); parameters.push_back(pgroups); + CommandParameter psize("size", "Number", "", "0", "", "", "",false,false); parameters.push_back(psize); + CommandParameter ppersample("persample", "Boolean", "", "F", "", "", "",false,false); parameters.push_back(ppersample); + 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, "SubSampleCommand", "getValidParameters"); + m->errorOut(e, "SubSampleCommand", "setParameters"); + exit(1); + } +} +//********************************************************************************************************************** +string SubSampleCommand::getHelpString(){ + try { + string helpString = ""; + helpString += "The sub.sample command is designed to be used as a way to normalize your data, or create a smaller set from your original set.\n"; + helpString += "The sub.sample command parameters are fasta, name, list, group, rabund, sabund, shared, groups, size, persample and label. You must provide a fasta, list, sabund, rabund or shared file as an input file.\n"; + helpString += "The namefile is only used with the fasta file, not with the listfile, because the list file should contain all sequences.\n"; + helpString += "The groups parameter allows you to specify which of the groups in your groupfile you would like included. The group names are separated by dashes.\n"; + helpString += "The label parameter allows you to select what distance levels you would like, and are also separated by dashes.\n"; + helpString += "The size parameter allows you indicate the size of your subsample.\n"; + helpString += "The persample parameter allows you indicate you want to select subsample of the same size from each of your groups, default=false. It is only used with the list and fasta files if a groupfile is given.\n"; + helpString += "persample=false will select a random set of sequences of the size you select, but the number of seqs from each group may differ.\n"; + helpString += "The size parameter is not set: with shared file size=number of seqs in smallest sample, with all other files if a groupfile is given and persample=true, then size=number of seqs in smallest sample, otherwise size=10% of number of seqs.\n"; + helpString += "The sub.sample command should be in the following format: sub.sample(list=yourListFile, group=yourGroupFile, groups=yourGroups, label=yourLabels).\n"; + helpString += "Example sub.sample(list=abrecovery.fn.list, group=abrecovery.groups, groups=B-C, size=20).\n"; + helpString += "The default value for groups is all the groups in your groupfile, and all labels in your inputfile will be used.\n"; + helpString += "The sub.sample command outputs a .subsample file.\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, "SubSampleCommand", "getHelpString"); exit(1); } } //********************************************************************************************************************** SubSampleCommand::SubSampleCommand(){ try { - abort = true; - //initialize outputTypes + abort = true; calledHelp = true; + setParameters(); vector tempOutNames; outputTypes["shared"] = tempOutNames; outputTypes["list"] = tempOutNames; @@ -42,43 +83,17 @@ SubSampleCommand::SubSampleCommand(){ } } //********************************************************************************************************************** -vector SubSampleCommand::getRequiredParameters(){ - try { - string Array[] = {"fasta","list","shared","rabund", "sabund","or"}; - vector myArray (Array, Array+(sizeof(Array)/sizeof(string))); - return myArray; - } - catch(exception& e) { - m->errorOut(e, "SubSampleCommand", "getRequiredParameters"); - exit(1); - } -} -//********************************************************************************************************************** -vector SubSampleCommand::getRequiredFiles(){ - try { - vector myArray; - return myArray; - } - catch(exception& e) { - m->errorOut(e, "SubSampleCommand", "getRequiredFiles"); - exit(1); - } -} -//********************************************************************************************************************** SubSampleCommand::SubSampleCommand(string option) { try { - globaldata = GlobalData::getInstance(); - abort = false; + abort = false; calledHelp = false; allLines = 1; - labels.clear(); //allow user to run help - if(option == "help") { help(); abort = 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[] = {"fasta", "group", "list","shared","rabund", "sabund","name","size","groups","label","outputdir","inputdir"}; - vector myArray (Array, Array+(sizeof(Array)/sizeof(string))); + vector myArray = setParameters(); OptionParser parser(option); map parameters = parser.getParameters(); @@ -169,32 +184,38 @@ SubSampleCommand::SubSampleCommand(string option) { //check for required parameters listfile = validParameter.validFile(parameters, "list", true); if (listfile == "not open") { listfile = ""; abort = true; } - else if (listfile == "not found") { listfile = ""; } + else if (listfile == "not found") { listfile = ""; } + else { m->setListFile(listfile); } sabundfile = validParameter.validFile(parameters, "sabund", true); if (sabundfile == "not open") { sabundfile = ""; abort = true; } else if (sabundfile == "not found") { sabundfile = ""; } + else { m->setSabundFile(sabundfile); } rabundfile = validParameter.validFile(parameters, "rabund", true); if (rabundfile == "not open") { rabundfile = ""; abort = true; } else if (rabundfile == "not found") { rabundfile = ""; } + else { m->setRabundFile(rabundfile); } fastafile = validParameter.validFile(parameters, "fasta", true); if (fastafile == "not open") { fastafile = ""; abort = true; } else if (fastafile == "not found") { fastafile = ""; } + else { m->setFastaFile(fastafile); } sharedfile = validParameter.validFile(parameters, "shared", true); if (sharedfile == "not open") { sharedfile = ""; abort = true; } else if (sharedfile == "not found") { sharedfile = ""; } + else { m->setSharedFile(sharedfile); } namefile = validParameter.validFile(parameters, "name", true); if (namefile == "not open") { namefile = ""; abort = true; } else if (namefile == "not found") { namefile = ""; } + else { m->setNameFile(namefile); } groupfile = validParameter.validFile(parameters, "group", true); if (groupfile == "not open") { groupfile = ""; abort = true; } else if (groupfile == "not found") { groupfile = ""; } - + else { m->setGroupFile(groupfile); } //check for optional parameter and set defaults // ...at some point should added some additional type checking... @@ -210,11 +231,16 @@ SubSampleCommand::SubSampleCommand(string option) { else { pickedGroups = true; m->splitAtDash(groups, Groups); - globaldata->Groups = Groups; + m->setGroups(Groups); } string temp = validParameter.validFile(parameters, "size", false); if (temp == "not found"){ temp = "0"; } - convert(temp, size); + m->mothurConvert(temp, size); + + temp = validParameter.validFile(parameters, "persample", false); if (temp == "not found"){ temp = "f"; } + persample = m->isTrue(temp); + + if (groupfile == "") { persample = false; } if ((namefile != "") && (fastafile == "")) { m->mothurOut("You may only use a namefile with a fastafile."); m->mothurOutEndLine(); abort = true; } @@ -230,6 +256,10 @@ SubSampleCommand::SubSampleCommand(string option) { if ((groupfile != "") && ((fastafile != "") && (listfile != ""))) { m->mothurOut("A new group file can only be made from the subsample of a listfile or fastafile, not both. Please correct."); m->mothurOutEndLine(); abort = true; } + if ((fastafile != "") && (namefile == "")) { + vector files; files.push_back(fastafile); + parser.getNameFile(files); + } } } @@ -238,57 +268,66 @@ SubSampleCommand::SubSampleCommand(string option) { exit(1); } } - -//********************************************************************************************************************** - -void SubSampleCommand::help(){ - try { - m->mothurOut("The sub.sample command is designed to be used as a way to normalize your data, or create a smaller set from your original set.\n"); - m->mothurOut("The sub.sample command parameters are fasta, name, list, group, rabund, sabund, shared, groups, size and label. You must provide a fasta, list, sabund, rabund or shared file as an input file.\n"); - m->mothurOut("The groups parameter allows you to specify which of the groups in your groupfile you would like included. The group names are separated by dashes.\n"); - m->mothurOut("The label parameter allows you to select what distance levels you would like, and are also separated by dashes.\n"); - m->mothurOut("The size parameter allows you indicate the size of your subsample.\n"); - m->mothurOut("The size parameter is not set: with shared file size=number of seqs in smallest sample, with all other files, 10% of number of seqs.\n"); - m->mothurOut("The sub.sample command should be in the following format: sub.sample(list=yourListFile, group=yourGroupFile, groups=yourGroups, label=yourLabels).\n"); - m->mothurOut("Example sub.sample(list=abrecovery.fn.list, group=abrecovery.groups, groups=B-C, size=20).\n"); - m->mothurOut("The default value for groups is all the groups in your groupfile, and all labels in your inputfile will be used.\n"); - m->mothurOut("The sub.sample command outputs a .subsample file.\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, "SubSampleCommand", "help"); - exit(1); - } -} - -//********************************************************************************************************************** - -SubSampleCommand::~SubSampleCommand(){} - //********************************************************************************************************************** int SubSampleCommand::execute(){ try { - if (abort == true) { return 0; } + if (abort == true) { if (calledHelp) { return 0; } return 2; } if (sharedfile != "") { getSubSampleShared(); } - if (m->control_pressed) { for (int i = 0; i < outputNames.size(); i++) { remove(outputNames[i].c_str()); return 0; } } + if (m->control_pressed) { for (int i = 0; i < outputNames.size(); i++) { m->mothurRemove(outputNames[i]); return 0; } } if (listfile != "") { getSubSampleList(); } - if (m->control_pressed) { for (int i = 0; i < outputNames.size(); i++) { remove(outputNames[i].c_str()); return 0; } } + if (m->control_pressed) { for (int i = 0; i < outputNames.size(); i++) { m->mothurRemove(outputNames[i]); return 0; } } if (rabundfile != "") { getSubSampleRabund(); } - if (m->control_pressed) { for (int i = 0; i < outputNames.size(); i++) { remove(outputNames[i].c_str()); return 0; } } + if (m->control_pressed) { for (int i = 0; i < outputNames.size(); i++) { m->mothurRemove(outputNames[i]); return 0; } } if (sabundfile != "") { getSubSampleSabund(); } - if (m->control_pressed) { for (int i = 0; i < outputNames.size(); i++) { remove(outputNames[i].c_str()); return 0; } } + if (m->control_pressed) { for (int i = 0; i < outputNames.size(); i++) { m->mothurRemove(outputNames[i]); return 0; } } if (fastafile != "") { getSubSampleFasta(); } - if (m->control_pressed) { for (int i = 0; i < outputNames.size(); i++) { remove(outputNames[i].c_str()); return 0; } } + if (m->control_pressed) { for (int i = 0; i < outputNames.size(); i++) { m->mothurRemove(outputNames[i]); return 0; } } - + //set fasta file as new current fastafile + string current = ""; + itTypes = outputTypes.find("fasta"); + if (itTypes != outputTypes.end()) { + if ((itTypes->second).size() != 0) { current = (itTypes->second)[0]; m->setFastaFile(current); } + } + + itTypes = outputTypes.find("name"); + if (itTypes != outputTypes.end()) { + if ((itTypes->second).size() != 0) { current = (itTypes->second)[0]; m->setNameFile(current); } + } + + itTypes = outputTypes.find("group"); + if (itTypes != outputTypes.end()) { + if ((itTypes->second).size() != 0) { current = (itTypes->second)[0]; m->setGroupFile(current); } + } + + itTypes = outputTypes.find("list"); + if (itTypes != outputTypes.end()) { + if ((itTypes->second).size() != 0) { current = (itTypes->second)[0]; m->setListFile(current); } + } + + itTypes = outputTypes.find("shared"); + if (itTypes != outputTypes.end()) { + if ((itTypes->second).size() != 0) { current = (itTypes->second)[0]; m->setSharedFile(current); } + } + + itTypes = outputTypes.find("rabund"); + if (itTypes != outputTypes.end()) { + if ((itTypes->second).size() != 0) { current = (itTypes->second)[0]; m->setRabundFile(current); } + } + + itTypes = outputTypes.find("sabund"); + if (itTypes != outputTypes.end()) { + if ((itTypes->second).size() != 0) { current = (itTypes->second)[0]; m->setSabundFile(current); } + } + + m->mothurOutEndLine(); m->mothurOut("Output File Names: "); m->mothurOutEndLine(); for (int i = 0; i < outputNames.size(); i++) { m->mothurOut(outputNames[i]); m->mothurOutEndLine(); } @@ -316,7 +355,8 @@ int SubSampleCommand::getSubSampleFasta() { //takes care of user setting groupNames that are invalid or setting groups=all SharedUtil* util = new SharedUtil(); - util->setGroups(Groups, groupMap->namesOfGroups); + vector namesGroups = groupMap->getNamesOfGroups(); + util->setGroups(Groups, namesGroups); delete util; //file mismatch quit @@ -330,80 +370,119 @@ int SubSampleCommand::getSubSampleFasta() { if (m->control_pressed) { return 0; } - string thisOutputDir = outputDir; - if (outputDir == "") { thisOutputDir += m->hasPath(fastafile); } - string outputFileName = thisOutputDir + m->getRootName(m->getSimpleName(fastafile)) + "subsample" + m->getExtension(fastafile); - - ofstream out; - m->openOutputFile(outputFileName, out); - outputTypes["fasta"].push_back(outputFileName); outputNames.push_back(outputFileName); //make sure that if your picked groups size is not too big - - if (pickedGroups) { - int total = 0; - for(int i = 0; i < Groups.size(); i++) { - total += groupMap->getNumSeqs(Groups[i]); + int thisSize = names.size(); + if (persample) { + if (size == 0) { //user has not set size, set size = smallest samples size + size = groupMap->getNumSeqs(Groups[0]); + for (int i = 1; i < Groups.size(); i++) { + int thisSize = groupMap->getNumSeqs(Groups[i]); + + if (thisSize < size) { size = thisSize; } + } + }else { //make sure size is not too large + vector newGroups; + for (int i = 0; i < Groups.size(); i++) { + int thisSize = groupMap->getNumSeqs(Groups[i]); + + if (thisSize >= size) { newGroups.push_back(Groups[i]); } + else { m->mothurOut("You have selected a size that is larger than " + Groups[i] + " number of sequences, removing " + Groups[i] + "."); m->mothurOutEndLine(); } + } + Groups = newGroups; + } + + m->mothurOut("Sampling " + toString(size) + " from each group."); m->mothurOutEndLine(); + }else { + if (pickedGroups) { + int total = 0; + for(int i = 0; i < Groups.size(); i++) { + total += groupMap->getNumSeqs(Groups[i]); + } + + if (size == 0) { //user has not set size, set size = 10% samples size + size = int (total * 0.10); + } + + if (total < size) { + if (size != 0) { + m->mothurOut("Your size is too large for the number of groups you selected. Adjusting to " + toString(int (total * 0.10)) + "."); m->mothurOutEndLine(); + } + size = int (total * 0.10); + } + + m->mothurOut("Sampling " + toString(size) + " from " + toString(total) + "."); m->mothurOutEndLine(); } if (size == 0) { //user has not set size, set size = 10% samples size - size = int (total * 0.10); + size = int (names.size() * 0.10); } - if (total < size) { - if (size != 0) { - m->mothurOut("Your size is too large for the number of groups you selected. Adjusting to " + toString(int (total * 0.10)) + "."); m->mothurOutEndLine(); - } - size = int (total * 0.10); + if (size > thisSize) { m->mothurOut("Your fasta file only contains " + toString(thisSize) + " sequences. Setting size to " + toString(thisSize) + "."); m->mothurOutEndLine(); + size = thisSize; } - m->mothurOut("Sampling " + toString(size) + " from " + toString(total) + "."); m->mothurOutEndLine(); - } - - if (size == 0) { //user has not set size, set size = 10% samples size - size = int (names.size() * 0.10); - } - - int thisSize = names.size(); - if (size > thisSize) { m->mothurOut("Your fasta file only contains " + toString(thisSize) + " sequences. Setting size to " + toString(thisSize) + "."); m->mothurOutEndLine(); - size = thisSize; + if (!pickedGroups) { m->mothurOut("Sampling " + toString(size) + " from " + toString(thisSize) + "."); m->mothurOutEndLine(); } + } - random_shuffle(names.begin(), names.end()); - if (!pickedGroups) { m->mothurOut("Sampling " + toString(size) + " from " + toString(thisSize) + "."); m->mothurOutEndLine(); } - - //randomly select a subset of those names to include in the subsample set subset; //dont want repeat sequence names added - for (int j = 0; j < size; j++) { + if (persample) { + //initialize counts + map groupCounts; + map::iterator itGroupCounts; + for (int i = 0; i < Groups.size(); i++) { groupCounts[Groups[i]] = 0; } - if (m->control_pressed) { return 0; } + for (int j = 0; j < names.size(); j++) { + + if (m->control_pressed) { return 0; } + + string group = groupMap->getGroup(names[j]); + if (group == "not found") { m->mothurOut("[ERROR]: " + names[j] + " is not in your groupfile. please correct."); m->mothurOutEndLine(); group = "NOTFOUND"; } + else{ + itGroupCounts = groupCounts.find(group); + if (itGroupCounts != groupCounts.end()) { + if (groupCounts[group] < size) { subset.insert(names[j]); groupCounts[group]++; } + } + } + } + }else { - //get random sequence to add, making sure we have not already added it - bool done = false; - int myrand; - while (!done) { - myrand = (int)((float)(rand()) / (RAND_MAX / (thisSize-1) + 1)); + //randomly select a subset of those names to include in the subsample + //since names was randomly shuffled just grab the next one + for (int j = 0; j < names.size(); j++) { + + if (m->control_pressed) { return 0; } - if (subset.count(names[myrand]) == 0) { + if (groupfile != "") { //if there is a groupfile given fill in group info + string group = groupMap->getGroup(names[j]); + if (group == "not found") { m->mothurOut("[ERROR]: " + names[j] + " is not in your groupfile. please correct."); m->mothurOutEndLine(); group = "NOTFOUND"; } - if (groupfile != "") { //if there is a groupfile given fill in group info - string group = groupMap->getGroup(names[myrand]); - if (group == "not found") { m->mothurOut("[ERROR]: " + names[myrand] + " is not in your groupfile. please correct."); m->mothurOutEndLine(); group = "NOTFOUND"; } - - if (pickedGroups) { //if hte user picked groups, we only want to keep the names of sequences from those groups - if (m->inUsersGroups(group, Groups)) { - subset.insert(names[myrand]); break; - } - }else{ - subset.insert(names[myrand]); break; + if (pickedGroups) { //if hte user picked groups, we only want to keep the names of sequences from those groups + if (m->inUsersGroups(group, Groups)) { + subset.insert(names[j]); } - }else{ //save everyone, group - subset.insert(names[myrand]); break; - } - } - } - } + }else{ + subset.insert(names[j]); + } + }else{ //save everyone, group + subset.insert(names[j]); + } + + //do we have enough?? + if (subset.size() == size) { break; } + } + } + + if (subset.size() == 0) { m->mothurOut("The size you selected is too large, skipping fasta file."); m->mothurOutEndLine(); return 0; } + + string thisOutputDir = outputDir; + if (outputDir == "") { thisOutputDir += m->hasPath(fastafile); } + string outputFileName = thisOutputDir + m->getRootName(m->getSimpleName(fastafile)) + "subsample" + m->getExtension(fastafile); + + ofstream out; + m->openOutputFile(outputFileName, out); //read through fasta file outputting only the names on the subsample list ifstream in; @@ -446,6 +525,34 @@ int SubSampleCommand::getSubSampleFasta() { m->mothurOut("[ERROR]: The subset selected contained " + toString(subset.size()) + " sequences, but I only found " + toString(count) + " of those in the fastafile."); m->mothurOutEndLine(); } + if (namefile != "") { + m->mothurOut("Deconvoluting subsampled fasta file... "); m->mothurOutEndLine(); + + //use unique.seqs to create new name and fastafile + string inputString = "fasta=" + outputFileName; + m->mothurOut("/******************************************/"); m->mothurOutEndLine(); + m->mothurOut("Running command: unique.seqs(" + inputString + ")"); m->mothurOutEndLine(); + m->mothurCalling = true; + + Command* uniqueCommand = new DeconvoluteCommand(inputString); + uniqueCommand->execute(); + + map > filenames = uniqueCommand->getOutputFiles(); + + delete uniqueCommand; + m->mothurCalling = false; + + outputTypes["name"].push_back(filenames["name"][0]); outputNames.push_back(filenames["name"][0]); + m->mothurRemove(outputFileName); + outputFileName = filenames["fasta"][0]; + + m->mothurOut("/******************************************/"); m->mothurOutEndLine(); + + m->mothurOut("Done."); m->mothurOutEndLine(); + } + + outputTypes["fasta"].push_back(outputFileName); outputNames.push_back(outputFileName); + //if a groupfile is provided read through the group file only outputting the names on the subsample list if (groupfile != "") { @@ -520,6 +627,8 @@ int SubSampleCommand::getNames() { } in.close(); + return 0; + } catch(exception& e) { m->errorOut(e, "SubSampleCommand", "getNames"); @@ -558,6 +667,8 @@ int SubSampleCommand::readNames() { } in.close(); + return 0; + } catch(exception& e) { m->errorOut(e, "SubSampleCommand", "readNames"); @@ -568,14 +679,6 @@ int SubSampleCommand::readNames() { int SubSampleCommand::getSubSampleShared() { try { - string thisOutputDir = outputDir; - if (outputDir == "") { thisOutputDir += m->hasPath(sharedfile); } - string outputFileName = thisOutputDir + m->getRootName(m->getSimpleName(sharedfile)) + "subsample" + m->getExtension(sharedfile); - - ofstream out; - m->openOutputFile(outputFileName, out); - outputTypes["shared"].push_back(outputFileName); outputNames.push_back(outputFileName); - InputData* input = new InputData(sharedfile, "sharedfile"); vector lookup = input->getSharedRAbundVectors(); string lastLabel = lookup[0]->getLabel(); @@ -591,19 +694,36 @@ int SubSampleCommand::getSubSampleShared() { if (thisSize < size) { size = thisSize; } } + }else { + m->clearGroups(); + Groups.clear(); + vector temp; + for (int i = 0; i < lookup.size(); i++) { + if (lookup[i]->getNumSeqs() < size) { + m->mothurOut(lookup[i]->getGroup() + " contains " + toString(lookup[i]->getNumSeqs()) + ". Eliminating."); m->mothurOutEndLine(); + delete lookup[i]; + }else { + Groups.push_back(lookup[i]->getGroup()); + temp.push_back(lookup[i]); + } + } + lookup = temp; + m->setGroups(Groups); } - m->mothurOut("Sampling " + toString(size) + " from " + toString(lookup[0]->getNumSeqs()) + "."); m->mothurOutEndLine(); + if (lookup.size() == 0) { m->mothurOut("The size you selected is too large, skipping shared file."); m->mothurOutEndLine(); delete input; return 0; } + + m->mothurOut("Sampling " + toString(size) + " from each group."); m->mothurOutEndLine(); //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) { delete input; for (int i = 0; i < lookup.size(); i++) { delete lookup[i]; lookup[i] = NULL; } out.close(); return 0; } + if (m->control_pressed) { delete input; for (int i = 0; i < lookup.size(); i++) { delete lookup[i]; lookup[i] = NULL; } return 0; } if(allLines == 1 || labels.count(lookup[0]->getLabel()) == 1){ m->mothurOut(lookup[0]->getLabel()); m->mothurOutEndLine(); - processShared(lookup, out); + processShared(lookup); processedLabels.insert(lookup[0]->getLabel()); userLabels.erase(lookup[0]->getLabel()); @@ -617,7 +737,7 @@ int SubSampleCommand::getSubSampleShared() { lookup = input->getSharedRAbundVectors(lastLabel); m->mothurOut(lookup[0]->getLabel()); m->mothurOutEndLine(); - processShared(lookup, out); + processShared(lookup); processedLabels.insert(lookup[0]->getLabel()); userLabels.erase(lookup[0]->getLabel()); @@ -635,7 +755,7 @@ int SubSampleCommand::getSubSampleShared() { } - if (m->control_pressed) { out.close(); return 0; } + if (m->control_pressed) { return 0; } //output error messages about any remaining user labels set::iterator it; @@ -657,13 +777,12 @@ int SubSampleCommand::getSubSampleShared() { m->mothurOut(lookup[0]->getLabel()); m->mothurOutEndLine(); - processShared(lookup, out); + processShared(lookup); for (int i = 0; i < lookup.size(); i++) { delete lookup[i]; } } delete input; - out.close(); return 0; @@ -674,58 +793,38 @@ int SubSampleCommand::getSubSampleShared() { } } //********************************************************************************************************************** -int SubSampleCommand::processShared(vector& thislookup, ofstream& out) { +int SubSampleCommand::processShared(vector& thislookup) { try { - int numBins = thislookup[0]->getNumBins(); - for (int i = 0; i < thislookup.size(); i++) { - int thisSize = thislookup[i]->getNumSeqs(); - - if (thisSize != size) { - - string thisgroup = thislookup[i]->getGroup(); - - OrderVector* order = new OrderVector(); - for(int p=0;pgetAbundance(p);j++){ - order->push_back(p); - } - } - random_shuffle(order->begin(), order->end()); - - SharedRAbundVector* temp = new SharedRAbundVector(numBins); - temp->setLabel(thislookup[i]->getLabel()); - temp->setGroup(thislookup[i]->getGroup()); - - delete thislookup[i]; - thislookup[i] = temp; - - - for (int j = 0; j < size; j++) { - - if (m->control_pressed) { delete order; return 0; } - - //get random number to sample from order between 0 and thisSize-1. - int myrand = (int)((float)(rand()) / (RAND_MAX / (thisSize-1) + 1)); - - int bin = order->get(myrand); - - int abund = thislookup[i]->getAbundance(bin); - thislookup[i]->set(bin, (abund+1), thisgroup); - } - delete order; - } - } + //save mothurOut's binLabels to restore for next label + vector saveBinLabels = m->currentBinLabels; - //subsampling may have created some otus with no sequences in them - eliminateZeroOTUS(thislookup); + string thisOutputDir = outputDir; + if (outputDir == "") { thisOutputDir += m->hasPath(sharedfile); } + string outputFileName = thisOutputDir + m->getRootName(m->getSimpleName(sharedfile)) + thislookup[0]->getLabel() + ".subsample" + m->getExtension(sharedfile); + + SubSample sample; + vector subsampledLabels = sample.getSample(thislookup, size); + + if (m->control_pressed) { return 0; } + + ofstream out; + m->openOutputFile(outputFileName, out); + outputTypes["shared"].push_back(outputFileName); outputNames.push_back(outputFileName); - if (m->control_pressed) { return 0; } + m->currentBinLabels = subsampledLabels; + + thislookup[0]->printHeaders(out); for (int i = 0; i < thislookup.size(); i++) { out << thislookup[i]->getLabel() << '\t' << thislookup[i]->getGroup() << '\t'; thislookup[i]->print(out); } + out.close(); + + + //save mothurOut's binLabels to restore for next label + m->currentBinLabels = saveBinLabels; return 0; @@ -764,7 +863,8 @@ int SubSampleCommand::getSubSampleList() { //takes care of user setting groupNames that are invalid or setting groups=all SharedUtil* util = new SharedUtil(); - util->setGroups(Groups, groupMap->namesOfGroups); + vector namesGroups = groupMap->getNamesOfGroups(); + util->setGroups(Groups, namesGroups); delete util; //create outputfiles @@ -789,37 +889,60 @@ int SubSampleCommand::getSubSampleList() { } //make sure that if your picked groups size is not too big - if (pickedGroups) { - int total = 0; - for(int i = 0; i < Groups.size(); i++) { - total += groupMap->getNumSeqs(Groups[i]); - } - - if (size == 0) { //user has not set size, set size = 10% samples size - size = int (total * 0.10); - } - - if (total < size) { - m->mothurOut("Your size is too large for the number of groups you selected. Adjusting to " + toString(int (total * 0.10)) + "."); m->mothurOutEndLine(); - size = int (total * 0.10); + if (persample) { + if (size == 0) { //user has not set size, set size = smallest samples size + size = groupMap->getNumSeqs(Groups[0]); + for (int i = 1; i < Groups.size(); i++) { + int thisSize = groupMap->getNumSeqs(Groups[i]); + + if (thisSize < size) { size = thisSize; } + } + }else { //make sure size is not too large + vector newGroups; + for (int i = 0; i < Groups.size(); i++) { + int thisSize = groupMap->getNumSeqs(Groups[i]); + + if (thisSize >= size) { newGroups.push_back(Groups[i]); } + else { m->mothurOut("You have selected a size that is larger than " + Groups[i] + " number of sequences, removing " + Groups[i] + "."); m->mothurOutEndLine(); } + } + Groups = newGroups; } - m->mothurOut("Sampling " + toString(size) + " from " + toString(total) + "."); m->mothurOutEndLine(); + m->mothurOut("Sampling " + toString(size) + " from each group."); m->mothurOutEndLine(); }else{ - - if (size == 0) { //user has not set size, set size = 10% samples size - size = int (list->getNumSeqs() * 0.10); - } - - int thisSize = list->getNumSeqs(); - if (size > thisSize) { m->mothurOut("Your list file only contains " + toString(thisSize) + " sequences. Setting size to " + toString(thisSize) + "."); m->mothurOutEndLine(); - size = thisSize; + if (pickedGroups) { + int total = 0; + for(int i = 0; i < Groups.size(); i++) { + total += groupMap->getNumSeqs(Groups[i]); + } + + if (size == 0) { //user has not set size, set size = 10% samples size + size = int (total * 0.10); + } + + if (total < size) { + m->mothurOut("Your size is too large for the number of groups you selected. Adjusting to " + toString(int (total * 0.10)) + "."); m->mothurOutEndLine(); + size = int (total * 0.10); + } + + m->mothurOut("Sampling " + toString(size) + " from " + toString(total) + "."); m->mothurOutEndLine(); + }else{ + + if (size == 0) { //user has not set size, set size = 10% samples size + size = int (list->getNumSeqs() * 0.10); + } + + int thisSize = list->getNumSeqs(); + if (size > thisSize) { m->mothurOut("Your list file only contains " + toString(thisSize) + " sequences. Setting size to " + toString(thisSize) + "."); m->mothurOutEndLine(); + size = thisSize; + } + + m->mothurOut("Sampling " + toString(size) + " from " + toString(list->getNumSeqs()) + "."); m->mothurOutEndLine(); } - - m->mothurOut("Sampling " + toString(size) + " from " + toString(list->getNumSeqs()) + "."); m->mothurOutEndLine(); } + //fill names for (int i = 0; i < list->getNumBins(); i++) { string binnames = list->get(i); @@ -867,22 +990,36 @@ int SubSampleCommand::getSubSampleList() { } random_shuffle(names.begin(), names.end()); - + //randomly select a subset of those names to include in the subsample set subset; //dont want repeat sequence names added - for (int j = 0; j < size; j++) { + if (persample) { + //initialize counts + map groupCounts; + map::iterator itGroupCounts; + for (int i = 0; i < Groups.size(); i++) { groupCounts[Groups[i]] = 0; } - if (m->control_pressed) { break; } - - //get random sequence to add, making sure we have not already added it - bool done = false; - int myrand; - while (!done) { - myrand = (int)((float)(rand()) / (RAND_MAX / (names.size()-1) + 1)); + for (int j = 0; j < names.size(); j++) { - if (subset.count(names[myrand]) == 0) { subset.insert(names[myrand]); break; } + if (m->control_pressed) { return 0; } + + string group = groupMap->getGroup(names[j]); + if (group == "not found") { m->mothurOut("[ERROR]: " + names[j] + " is not in your groupfile. please correct."); m->mothurOutEndLine(); group = "NOTFOUND"; } + else{ + itGroupCounts = groupCounts.find(group); + if (itGroupCounts != groupCounts.end()) { + if (groupCounts[group] < size) { subset.insert(names[j]); groupCounts[group]++; } + } + } } - } + }else{ + for (int j = 0; j < size; j++) { + + if (m->control_pressed) { break; } + + subset.insert(names[j]); + } + } if (groupfile != "") { //write out new groupfile @@ -1004,7 +1141,7 @@ int SubSampleCommand::processList(ListVector*& list, ofstream& out, set& individual += binnames[j]; } } - if (subset.count(individual) != 0) { newNames += individual; } + if (subset.count(individual) != 0) { newNames += individual + ","; } //if there are names in this bin add to new list @@ -1032,15 +1169,6 @@ int SubSampleCommand::processList(ListVector*& list, ofstream& out, set& //********************************************************************************************************************** int SubSampleCommand::getSubSampleRabund() { try { - - string thisOutputDir = outputDir; - if (outputDir == "") { thisOutputDir += m->hasPath(rabundfile); } - string outputFileName = thisOutputDir + m->getRootName(m->getSimpleName(rabundfile)) + "subsample" + m->getExtension(rabundfile); - - ofstream out; - m->openOutputFile(outputFileName, out); - outputTypes["rabund"].push_back(outputFileName); outputNames.push_back(outputFileName); - InputData* input = new InputData(rabundfile, "rabund"); RAbundVector* rabund = input->getRAbundVector(); string lastLabel = rabund->getLabel(); @@ -1051,10 +1179,18 @@ int SubSampleCommand::getSubSampleRabund() { if (size == 0) { //user has not set size, set size = 10% size = int((rabund->getNumSeqs()) * 0.10); - } + }else if (size > rabund->getNumSeqs()) { m->mothurOut("The size you selected is too large, skipping rabund file."); m->mothurOutEndLine(); delete input; delete rabund; return 0; } m->mothurOut("Sampling " + toString(size) + " from " + toString(rabund->getNumSeqs()) + "."); m->mothurOutEndLine(); + string thisOutputDir = outputDir; + if (outputDir == "") { thisOutputDir += m->hasPath(rabundfile); } + string outputFileName = thisOutputDir + m->getRootName(m->getSimpleName(rabundfile)) + "subsample" + m->getExtension(rabundfile); + + ofstream out; + m->openOutputFile(outputFileName, out); + outputTypes["rabund"].push_back(outputFileName); outputNames.push_back(outputFileName); + //as long as you are not at the end of the file or done wih the lines you want while((rabund != NULL) && ((allLines == 1) || (userLabels.size() != 0))) { if (m->control_pressed) { delete input; delete rabund; out.close(); return 0; } @@ -1162,10 +1298,7 @@ int SubSampleCommand::processRabund(RAbundVector*& rabund, ofstream& out) { if (m->control_pressed) { delete order; return 0; } - //get random number to sample from order between 0 and thisSize-1. - int myrand = (int)((float)(rand()) / (RAND_MAX / (thisSize-1) + 1)); - - int bin = order->get(myrand); + int bin = order->get(j); int abund = rabund->get(bin); rabund->set(bin, (abund+1)); @@ -1189,15 +1322,7 @@ int SubSampleCommand::processRabund(RAbundVector*& rabund, ofstream& out) { //********************************************************************************************************************** int SubSampleCommand::getSubSampleSabund() { try { - - string thisOutputDir = outputDir; - if (outputDir == "") { thisOutputDir += m->hasPath(sabundfile); } - string outputFileName = thisOutputDir + m->getRootName(m->getSimpleName(sabundfile)) + "subsample" + m->getExtension(sabundfile); - - ofstream out; - m->openOutputFile(outputFileName, out); - outputTypes["sabund"].push_back(outputFileName); outputNames.push_back(outputFileName); - + InputData* input = new InputData(sabundfile, "sabund"); SAbundVector* sabund = input->getSAbundVector(); string lastLabel = sabund->getLabel(); @@ -1208,10 +1333,20 @@ int SubSampleCommand::getSubSampleSabund() { if (size == 0) { //user has not set size, set size = 10% size = int((sabund->getNumSeqs()) * 0.10); - } + }else if (size > sabund->getNumSeqs()) { m->mothurOut("The size you selected is too large, skipping sabund file."); m->mothurOutEndLine(); delete input; delete sabund; return 0; } + m->mothurOut("Sampling " + toString(size) + " from " + toString(sabund->getNumSeqs()) + "."); m->mothurOutEndLine(); + string thisOutputDir = outputDir; + if (outputDir == "") { thisOutputDir += m->hasPath(sabundfile); } + string outputFileName = thisOutputDir + m->getRootName(m->getSimpleName(sabundfile)) + "subsample" + m->getExtension(sabundfile); + + ofstream out; + m->openOutputFile(outputFileName, out); + outputTypes["sabund"].push_back(outputFileName); outputNames.push_back(outputFileName); + + //as long as you are not at the end of the file or done wih the lines you want while((sabund != NULL) && ((allLines == 1) || (userLabels.size() != 0))) { if (m->control_pressed) { delete input; delete sabund; out.close(); return 0; } @@ -1322,10 +1457,7 @@ int SubSampleCommand::processSabund(SAbundVector*& sabund, ofstream& out) { if (m->control_pressed) { delete order; return 0; } - //get random number to sample from order between 0 and thisSize-1. - int myrand = (int)((float)(rand()) / (RAND_MAX / (thisSize-1) + 1)); - - int bin = order->get(myrand); + int bin = order->get(j); int abund = rabund->get(bin); rabund->set(bin, (abund+1)); @@ -1352,50 +1484,6 @@ int SubSampleCommand::processSabund(SAbundVector*& sabund, ofstream& out) { } } //********************************************************************************************************************** -int SubSampleCommand::eliminateZeroOTUS(vector& thislookup) { - try { - - vector newLookup; - for (int i = 0; i < thislookup.size(); i++) { - SharedRAbundVector* temp = new SharedRAbundVector(); - temp->setLabel(thislookup[i]->getLabel()); - temp->setGroup(thislookup[i]->getGroup()); - newLookup.push_back(temp); - } - - //for each bin - for (int i = 0; i < thislookup[0]->getNumBins(); i++) { - if (m->control_pressed) { for (int j = 0; j < newLookup.size(); j++) { delete newLookup[j]; } return 0; } - - //look at each sharedRabund and make sure they are not all zero - bool allZero = true; - for (int j = 0; j < thislookup.size(); j++) { - if (thislookup[j]->getAbundance(i) != 0) { allZero = false; break; } - } - - //if they are not all zero add this bin - if (!allZero) { - for (int j = 0; j < thislookup.size(); j++) { - newLookup[j]->push_back(thislookup[j]->getAbundance(i), thislookup[j]->getGroup()); - } - } - } - - for (int j = 0; j < thislookup.size(); j++) { delete thislookup[j]; } - thislookup.clear(); - - thislookup = newLookup; - - return 0; - - } - catch(exception& e) { - m->errorOut(e, "SubSampleCommand", "eliminateZeroOTUS"); - exit(1); - } -} - -//**********************************************************************************************************************