+int SubSampleCommand::getSubSampleList() {
+ try {
+
+ string thisOutputDir = outputDir;
+ if (outputDir == "") { thisOutputDir += m->hasPath(listfile); }
+ string outputFileName = thisOutputDir + m->getRootName(m->getSimpleName(listfile)) + "subsample" + m->getExtension(listfile);
+
+ ofstream out;
+ m->openOutputFile(outputFileName, out);
+ outputTypes["list"].push_back(outputFileName); outputNames.push_back(outputFileName);
+
+ InputData* input = new InputData(listfile, "list");
+ ListVector* list = input->getListVector();
+ string lastLabel = list->getLabel();
+
+ //if the users enters label "0.06" and there is no "0.06" in their file use the next lowest label.
+ set<string> processedLabels;
+ set<string> userLabels = labels;
+
+ ofstream outGroup;
+ GroupMap* groupMap;
+ if (groupfile != "") {
+
+ groupMap = new GroupMap(groupfile);
+ groupMap->readMap();
+
+ //takes care of user setting groupNames that are invalid or setting groups=all
+ SharedUtil* util = new SharedUtil();
+ util->setGroups(Groups, groupMap->namesOfGroups);
+ delete util;
+
+ //create outputfiles
+ string groupOutputDir = outputDir;
+ if (outputDir == "") { groupOutputDir += m->hasPath(groupfile); }
+ string groupOutputFileName = groupOutputDir + m->getRootName(m->getSimpleName(groupfile)) + "subsample" + m->getExtension(groupfile);
+
+ m->openOutputFile(groupOutputFileName, outGroup);
+ outputTypes["group"].push_back(groupOutputFileName); outputNames.push_back(groupOutputFileName);
+
+ //file mismatch quit
+ if (list->getNumSeqs() != groupMap->getNumSeqs()) {
+ m->mothurOut("[ERROR]: your list file contains " + toString(list->getNumSeqs()) + " sequences, and your groupfile contains " + toString(groupMap->getNumSeqs()) + ", please correct.");
+ m->mothurOutEndLine();
+ delete groupMap;
+ delete list;
+ delete input;
+ out.close();
+ outGroup.close();
+ return 0;
+ }
+ }
+
+ //make sure that if your picked groups size is not too big
+ 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
+ int smallestSize = groupMap->getNumSeqs(Groups[0]);
+ for (int i = 1; i < Groups.size(); i++) {
+ int thisSize = groupMap->getNumSeqs(Groups[i]);
+
+ if (thisSize < smallestSize) { smallestSize = thisSize; }
+ }
+ if (smallestSize < size) { size = smallestSize; m->mothurOut("You have selected a size that is larger than your smallest sample, using your samllest sample size, " + toString(smallestSize) + "."); m->mothurOutEndLine(); }
+ }
+
+ 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) {
+ 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();
+ }
+ }
+
+
+ //fill names
+ for (int i = 0; i < list->getNumBins(); i++) {
+ string binnames = list->get(i);
+
+ //parse names
+ string individual = "";
+ int length = binnames.length();
+ for(int j=0;j<length;j++){
+ if(binnames[j] == ','){
+
+ if (groupfile != "") { //if there is a groupfile given fill in group info
+ string group = groupMap->getGroup(individual);
+ if (group == "not found") { m->mothurOut("[ERROR]: " + individual + " 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)) {
+ names.push_back(individual);
+ }
+ }else{
+ names.push_back(individual);
+ }
+ }else{ //save everyone, group
+ names.push_back(individual);
+ }
+ individual = "";
+ }
+ else{
+ individual += binnames[j];
+ }
+ }
+ //save last name
+ if (groupfile != "") { //if there is a groupfile given fill in group info
+ string group = groupMap->getGroup(individual);
+ if (group == "not found") { m->mothurOut("[ERROR]: " + individual + " 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)) {
+ names.push_back(individual);
+ }
+ }else{
+ names.push_back(individual);
+ }
+ }else{ //save everyone, group
+ names.push_back(individual);
+ }
+ }
+
+ random_shuffle(names.begin(), names.end());
+
+ //randomly select a subset of those names to include in the subsample
+ set<string> subset; //dont want repeat sequence names added
+ if (persample) {
+ for (int i = 0; i < Groups.size(); i++) {
+
+ for (int j = 0; j < size; j++) {
+
+ 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)(names.size()) * (float)(rand()) / ((float)RAND_MAX+1.0));
+
+ if (subset.count(names[myrand]) == 0) { //you are not already added
+ if (groupMap->getGroup(names[myrand]) == Groups[i]) { subset.insert(names[myrand]); break; }
+ }
+ }
+ }
+ }
+ }else{
+ for (int j = 0; j < size; j++) {
+
+ 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)(names.size()) * (float)(rand()) / ((float)RAND_MAX+1.0));
+
+ if (subset.count(names[myrand]) == 0) { subset.insert(names[myrand]); break; }
+ }
+ }
+ }
+
+ if (groupfile != "") {
+ //write out new groupfile
+ for (set<string>::iterator it = subset.begin(); it != subset.end(); it++) {
+ string group = groupMap->getGroup(*it);
+ if (group == "not found") { group = "NOTFOUND"; }
+
+ outGroup << *it << '\t' << group << endl;
+ }
+ outGroup.close(); delete groupMap;
+ }
+
+
+ //as long as you are not at the end of the file or done wih the lines you want
+ while((list != NULL) && ((allLines == 1) || (userLabels.size() != 0))) {
+
+ if (m->control_pressed) { delete list; delete input; out.close(); return 0; }
+
+ if(allLines == 1 || labels.count(list->getLabel()) == 1){
+
+ m->mothurOut(list->getLabel()); m->mothurOutEndLine();
+
+ processList(list, out, subset);
+
+ processedLabels.insert(list->getLabel());
+ userLabels.erase(list->getLabel());
+ }
+
+ if ((m->anyLabelsToProcess(list->getLabel(), userLabels, "") == true) && (processedLabels.count(lastLabel) != 1)) {
+ string saveLabel = list->getLabel();
+
+ delete list;
+
+ list = input->getListVector(lastLabel);
+ m->mothurOut(list->getLabel()); m->mothurOutEndLine();
+
+ processList(list, out, subset);
+
+ processedLabels.insert(list->getLabel());
+ userLabels.erase(list->getLabel());
+
+ //restore real lastlabel to save below
+ list->setLabel(saveLabel);
+ }
+
+ lastLabel = list->getLabel();
+
+ delete list; list = NULL;
+
+ //get next line to process
+ list = input->getListVector();
+ }
+
+
+ if (m->control_pressed) { if (list != NULL) { delete list; } delete input; out.close(); return 0; }
+
+ //output error messages about any remaining user labels
+ set<string>::iterator it;
+ bool needToRun = false;
+ for (it = userLabels.begin(); it != userLabels.end(); it++) {
+ m->mothurOut("Your file does not include the label " + *it);
+ if (processedLabels.count(lastLabel) != 1) {
+ m->mothurOut(". I will use " + lastLabel + "."); m->mothurOutEndLine();
+ needToRun = true;
+ }else {
+ m->mothurOut(". Please refer to " + lastLabel + "."); m->mothurOutEndLine();
+ }
+ }
+
+ //run last label if you need to
+ if (needToRun == true) {
+ if (list != NULL) { delete list; }
+
+ list = input->getListVector(lastLabel);
+
+ m->mothurOut(list->getLabel()); m->mothurOutEndLine();
+
+ processList(list, out, subset);
+
+ delete list; list = NULL;
+ }
+
+ out.close();
+ if (list != NULL) { delete list; }
+ delete input;
+
+ return 0;
+
+ }
+ catch(exception& e) {
+ m->errorOut(e, "SubSampleCommand", "getSubSampleList");
+ exit(1);
+ }
+}
+//**********************************************************************************************************************
+int SubSampleCommand::processList(ListVector*& list, ofstream& out, set<string>& subset) {
+ try {
+
+ int numBins = list->getNumBins();
+
+ ListVector* temp = new ListVector();
+ temp->setLabel(list->getLabel());
+
+ for (int i = 0; i < numBins; i++) {
+
+ if (m->control_pressed) { break; }
+
+ string binnames = list->get(i);
+
+ //parse names
+ string individual = "";
+ string newNames = "";
+ int length = binnames.length();
+ for(int j=0;j<length;j++){
+ if(binnames[j] == ','){
+ if (subset.count(individual) != 0) { newNames += individual + ","; }
+ individual = "";
+ }else{
+ individual += binnames[j];
+ }
+ }
+ if (subset.count(individual) != 0) { newNames += individual; }
+
+
+ //if there are names in this bin add to new list
+ if (newNames != "") {
+ newNames = newNames.substr(0, newNames.length()-1); //rip off extra comma
+ temp->push_back(newNames);
+ }
+ }
+
+ delete list;
+ list = temp;
+
+ if (m->control_pressed) { return 0; }
+
+ list->print(out);
+
+ return 0;
+
+ }
+ catch(exception& e) {
+ m->errorOut(e, "SubSampleCommand", "processList");
+ exit(1);
+ }
+}
+//**********************************************************************************************************************
+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();
+
+ //if the users enters label "0.06" and there is no "0.06" in their file use the next lowest label.
+ set<string> processedLabels;
+ set<string> userLabels = labels;
+
+ if (size == 0) { //user has not set size, set size = 10%
+ size = int((rabund->getNumSeqs()) * 0.10);
+ }
+
+ m->mothurOut("Sampling " + toString(size) + " from " + toString(rabund->getNumSeqs()) + "."); m->mothurOutEndLine();
+
+ //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; }
+
+ if(allLines == 1 || labels.count(rabund->getLabel()) == 1){
+
+ m->mothurOut(rabund->getLabel()); m->mothurOutEndLine();
+
+ processRabund(rabund, out);
+
+ processedLabels.insert(rabund->getLabel());
+ userLabels.erase(rabund->getLabel());
+ }
+
+ if ((m->anyLabelsToProcess(rabund->getLabel(), userLabels, "") == true) && (processedLabels.count(lastLabel) != 1)) {
+ string saveLabel = rabund->getLabel();
+
+ delete rabund;
+
+ rabund = input->getRAbundVector(lastLabel);
+ m->mothurOut(rabund->getLabel()); m->mothurOutEndLine();
+
+ processRabund(rabund, out);
+
+ processedLabels.insert(rabund->getLabel());
+ userLabels.erase(rabund->getLabel());
+
+ //restore real lastlabel to save below
+ rabund->setLabel(saveLabel);
+ }
+
+ lastLabel = rabund->getLabel();
+
+ //prevent memory leak
+ delete rabund; rabund = NULL;
+
+ //get next line to process
+ rabund = input->getRAbundVector();
+ }
+
+
+ if (m->control_pressed) { out.close(); return 0; }
+
+ //output error messages about any remaining user labels
+ set<string>::iterator it;
+ bool needToRun = false;
+ for (it = userLabels.begin(); it != userLabels.end(); it++) {
+ m->mothurOut("Your file does not include the label " + *it);
+ if (processedLabels.count(lastLabel) != 1) {
+ m->mothurOut(". I will use " + lastLabel + "."); m->mothurOutEndLine();
+ needToRun = true;
+ }else {
+ m->mothurOut(". Please refer to " + lastLabel + "."); m->mothurOutEndLine();
+ }
+ }
+
+ //run last label if you need to
+ if (needToRun == true) {
+ if (rabund != NULL) { delete rabund; }
+
+ rabund = input->getRAbundVector(lastLabel);
+
+ m->mothurOut(rabund->getLabel()); m->mothurOutEndLine();
+
+ processRabund(rabund, out);
+
+ delete rabund;
+ }
+
+ delete input;
+ out.close();
+
+ return 0;
+
+ }
+ catch(exception& e) {
+ m->errorOut(e, "SubSampleCommand", "getSubSampleRabund");
+ exit(1);
+ }
+}
+//**********************************************************************************************************************
+int SubSampleCommand::processRabund(RAbundVector*& rabund, ofstream& out) {
+ try {
+
+ int numBins = rabund->getNumBins();
+ int thisSize = rabund->getNumSeqs();
+
+ if (thisSize != size) {
+
+ OrderVector* order = new OrderVector();
+ for(int p=0;p<numBins;p++){
+ for(int j=0;j<rabund->get(p);j++){
+ order->push_back(p);
+ }
+ }
+ random_shuffle(order->begin(), order->end());
+
+ RAbundVector* temp = new RAbundVector(numBins);
+ temp->setLabel(rabund->getLabel());
+
+ delete rabund;
+ rabund = 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)(thisSize) * (float)(rand()) / ((float)RAND_MAX+1.0));
+
+ int bin = order->get(myrand);
+
+ int abund = rabund->get(bin);
+ rabund->set(bin, (abund+1));
+ }
+
+ delete order;
+ }
+
+ if (m->control_pressed) { return 0; }
+
+ rabund->print(out);
+
+ return 0;
+
+ }
+ catch(exception& e) {
+ m->errorOut(e, "SubSampleCommand", "processRabund");
+ exit(1);
+ }
+}
+//**********************************************************************************************************************
+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();
+
+ //if the users enters label "0.06" and there is no "0.06" in their file use the next lowest label.
+ set<string> processedLabels;
+ set<string> userLabels = labels;
+
+ if (size == 0) { //user has not set size, set size = 10%
+ size = int((sabund->getNumSeqs()) * 0.10);
+ }
+
+ m->mothurOut("Sampling " + toString(size) + " from " + toString(sabund->getNumSeqs()) + "."); m->mothurOutEndLine();
+
+ //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; }
+
+ if(allLines == 1 || labels.count(sabund->getLabel()) == 1){
+
+ m->mothurOut(sabund->getLabel()); m->mothurOutEndLine();
+
+ processSabund(sabund, out);
+
+ processedLabels.insert(sabund->getLabel());
+ userLabels.erase(sabund->getLabel());
+ }
+
+ if ((m->anyLabelsToProcess(sabund->getLabel(), userLabels, "") == true) && (processedLabels.count(lastLabel) != 1)) {
+ string saveLabel = sabund->getLabel();
+
+ delete sabund;
+
+ sabund = input->getSAbundVector(lastLabel);
+ m->mothurOut(sabund->getLabel()); m->mothurOutEndLine();
+
+ processSabund(sabund, out);
+
+ processedLabels.insert(sabund->getLabel());
+ userLabels.erase(sabund->getLabel());
+
+ //restore real lastlabel to save below
+ sabund->setLabel(saveLabel);
+ }
+
+ lastLabel = sabund->getLabel();
+
+ //prevent memory leak
+ delete sabund; sabund = NULL;
+
+ //get next line to process
+ sabund = input->getSAbundVector();
+ }
+
+
+ if (m->control_pressed) { out.close(); return 0; }
+
+ //output error messages about any remaining user labels
+ set<string>::iterator it;
+ bool needToRun = false;
+ for (it = userLabels.begin(); it != userLabels.end(); it++) {
+ m->mothurOut("Your file does not include the label " + *it);
+ if (processedLabels.count(lastLabel) != 1) {
+ m->mothurOut(". I will use " + lastLabel + "."); m->mothurOutEndLine();
+ needToRun = true;
+ }else {
+ m->mothurOut(". Please refer to " + lastLabel + "."); m->mothurOutEndLine();
+ }
+ }
+
+ //run last label if you need to
+ if (needToRun == true) {
+ if (sabund != NULL) { delete sabund; }
+
+ sabund = input->getSAbundVector(lastLabel);
+
+ m->mothurOut(sabund->getLabel()); m->mothurOutEndLine();
+
+ processSabund(sabund, out);
+
+ delete sabund;
+ }
+
+ delete input;
+ out.close();
+
+ return 0;
+
+ }
+ catch(exception& e) {
+ m->errorOut(e, "SubSampleCommand", "getSubSampleSabund");
+ exit(1);
+ }
+}
+//**********************************************************************************************************************
+int SubSampleCommand::processSabund(SAbundVector*& sabund, ofstream& out) {
+ try {
+
+ RAbundVector* rabund = new RAbundVector();
+ *rabund = sabund->getRAbundVector();
+
+ int numBins = rabund->getNumBins();
+ int thisSize = rabund->getNumSeqs();
+
+ if (thisSize != size) {
+
+ OrderVector* order = new OrderVector();
+ for(int p=0;p<numBins;p++){
+ for(int j=0;j<rabund->get(p);j++){
+ order->push_back(p);
+ }
+ }
+ random_shuffle(order->begin(), order->end());
+
+ RAbundVector* temp = new RAbundVector(numBins);
+ temp->setLabel(rabund->getLabel());
+
+ delete rabund;
+ rabund = 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)(thisSize) * (float)(rand()) / ((float)RAND_MAX+1.0));
+
+ int bin = order->get(myrand);
+
+ int abund = rabund->get(bin);
+ rabund->set(bin, (abund+1));
+ }
+
+ delete order;
+ }
+
+ if (m->control_pressed) { return 0; }
+
+ delete sabund;
+ sabund = new SAbundVector();
+ *sabund = rabund->getSAbundVector();
+ delete rabund;
+
+ sabund->print(out);
+
+ return 0;
+
+ }
+ catch(exception& e) {
+ m->errorOut(e, "SubSampleCommand", "processSabund");
+ exit(1);
+ }
+}
+//**********************************************************************************************************************
+int SubSampleCommand::eliminateZeroOTUS(vector<SharedRAbundVector*>& thislookup) {
+ try {
+
+ vector<SharedRAbundVector*> 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);
+ }
+}
+
+//**********************************************************************************************************************
+