//**********************************************************************************************************************
int RemoveRareCommand::processList(){
try {
string thisOutputDir = outputDir;
if (outputDir == "") { thisOutputDir += m->hasPath(listfile); }
string outputFileName = thisOutputDir + m->getRootName(m->getSimpleName(listfile)) + "pick" + m->getExtension(listfile);
string outputGroupFileName = thisOutputDir + m->getRootName(m->getSimpleName(groupfile)) + "pick" + m->getExtension(groupfile);
ofstream out, outGroup;
m->openOutputFile(outputFileName, out);
bool wroteSomething = false;
//you must provide a label because the names in the listfile need to be consistent
string thisLabel = "";
if (allLines) { m->mothurOut("For the listfile you must select one label, using first label in your listfile."); m->mothurOutEndLine(); }
else if (labels.size() > 1) { m->mothurOut("For the listfile you must select one label, using " + (*labels.begin()) + "."); m->mothurOutEndLine(); thisLabel = *labels.begin(); }
else { thisLabel = *labels.begin(); }
InputData input(listfile, "list");
ListVector* list = input.getListVector();
//get first one or the one we want
if (thisLabel != "") {
//use smart distancing
set<string> userLabels; userLabels.insert(thisLabel);
set<string> processedLabels;
string lastLabel = list->getLabel();
while((list != NULL) && (userLabels.size() != 0)) {
if(userLabels.count(list->getLabel()) == 1){
processedLabels.insert(list->getLabel());
userLabels.erase(list->getLabel());
break;
}
if ((m->anyLabelsToProcess(list->getLabel(), userLabels, "") == true) && (processedLabels.count(lastLabel) != 1)) {
processedLabels.insert(list->getLabel());
userLabels.erase(list->getLabel());
delete list;
list = input.getListVector(lastLabel);
break;
}
lastLabel = list->getLabel();
delete list;
list = input.getListVector();
}
if (userLabels.size() != 0) {
m->mothurOut("Your file does not include the label " + thisLabel + ". I will use " + lastLabel + "."); m->mothurOutEndLine();
list = input.getListVector(lastLabel);
}
}
//if groupfile is given then use it
GroupMap* groupMap;
if (groupfile != "") {
groupMap = new GroupMap(groupfile); groupMap->readMap();
SharedUtil util;
vector<string> namesGroups = groupMap->getNamesOfGroups();
util.setGroups(Groups, namesGroups);
m->openOutputFile(outputGroupFileName, outGroup);
}
if (list != NULL) {
//make a new list vector
ListVector newList;
newList.setLabel(list->getLabel());
//for each bin
for (int i = 0; i < list->getNumBins(); i++) {
if (m->control_pressed) { if (groupfile != "") { delete groupMap; outGroup.close(); m->mothurRemove(outputGroupFileName); } out.close(); m->mothurRemove(outputFileName); return 0; }
//parse out names that are in accnos file
string binnames = list->get(i);
vector<string> names;
string saveBinNames = binnames;
m->splitAtComma(binnames, names);
vector<string> newGroupFile;
if (groupfile != "") {
vector<string> newNames;
saveBinNames = "";
for(int k = 0; k < names.size(); k++) {
string group = groupMap->getGroup(names[k]);
if (m->inUsersGroups(group, Groups)) {
newGroupFile.push_back(names[k] + "\t" + group);
newNames.push_back(names[k]);
saveBinNames += names[k] + ",";
}
}
names = newNames;
saveBinNames = saveBinNames.substr(0, saveBinNames.length()-1);
}
if (names.size() > nseqs) { //keep bin
newList.push_back(saveBinNames);
for(int k = 0; k < newGroupFile.size(); k++) { outGroup << newGroupFile[k] << endl; }
}
}
//print new listvector
if (newList.getNumBins() != 0) {
wroteSomething = true;
newList.print(out);
}
}
out.close();
if (groupfile != "") { outGroup.close(); outputTypes["group"].push_back(outputGroupFileName); outputNames.push_back(outputGroupFileName); }
if (wroteSomething == false) { m->mothurOut("Your file contains only rare sequences."); m->mothurOutEndLine(); }
outputTypes["list"].push_back(outputFileName); outputNames.push_back(outputFileName);
return 0;
}
catch(exception& e) {
m->errorOut(e, "RemoveRareCommand", "processList");
exit(1);
}
}
int CountTable::createTable(string namefile, string groupfile, bool createGroup) {
try {
if (namefile == "") { m->mothurOut("[ERROR]: namefile cannot be blank when creating a count table.\n"); m->control_pressed = true; }
GroupMap* groupMap;
int numGroups = 0;
groups.clear();
totalGroups.clear();
indexGroupMap.clear();
indexNameMap.clear();
counts.clear();
map<int, string> originalGroupIndexes;
if (groupfile != "") {
hasGroups = true;
groupMap = new GroupMap(groupfile); groupMap->readMap();
numGroups = groupMap->getNumGroups();
groups = groupMap->getNamesOfGroups();
totalGroups.resize(numGroups, 0);
}else if(createGroup) {
hasGroups = true;
numGroups = 1;
groups.push_back("Group1");
totalGroups.resize(numGroups, 0);
}
//sort groups to keep consistent with how we store the groups in groupmap
sort(groups.begin(), groups.end());
for (int i = 0; i < groups.size(); i++) { indexGroupMap[groups[i]] = i; }
m->setAllGroups(groups);
bool error = false;
string name;
uniques = 0;
total = 0;
//open input file
ifstream in;
m->openInputFile(namefile, in);
int total = 0;
while (!in.eof()) {
if (m->control_pressed) { break; }
string firstCol, secondCol;
in >> firstCol; m->gobble(in); in >> secondCol; m->gobble(in);
m->checkName(firstCol);
m->checkName(secondCol);
vector<string> names;
m->splitAtChar(secondCol, names, ',');
map<string, int> groupCounts;
int thisTotal = 0;
if (groupfile != "") {
//set to 0
for (int i = 0; i < groups.size(); i++) { groupCounts[groups[i]] = 0; }
//get counts for each of the users groups
for (int i = 0; i < names.size(); i++) {
string group = groupMap->getGroup(names[i]);
if (group == "not found") { m->mothurOut("[ERROR]: " + names[i] + " is not in your groupfile, please correct."); m->mothurOutEndLine(); error=true; }
else {
map<string, int>::iterator it = groupCounts.find(group);
//if not found, then this sequence is not from a group we care about
if (it != groupCounts.end()) {
it->second++;
thisTotal++;
}
}
}
}else if (createGroup) {
groupCounts["Group1"]=0;
for (int i = 0; i < names.size(); i++) {
string group = "Group1";
groupCounts["Group1"]++; thisTotal++;
}
}else { thisTotal = names.size(); }
//if group info, then read it
vector<int> thisGroupsCount; thisGroupsCount.resize(numGroups, 0);
for (int i = 0; i < numGroups; i++) {
thisGroupsCount[i] = groupCounts[groups[i]];
totalGroups[i] += thisGroupsCount[i];
}
map<string, int>::iterator it = indexNameMap.find(firstCol);
if (it == indexNameMap.end()) {
if (hasGroups) { counts.push_back(thisGroupsCount); }
indexNameMap[firstCol] = uniques;
totals.push_back(thisTotal);
total += thisTotal;
uniques++;
}else {
error = true;
m->mothurOut("[ERROR]: Your count table contains more than 1 sequence named " + firstCol + ", sequence names must be unique. Please correct."); m->mothurOutEndLine();
}
}
in.close();
if (error) { m->control_pressed = true; }
else { //check for zero groups
if (hasGroups) {
for (int i = 0; i < totalGroups.size(); i++) {
if (totalGroups[i] == 0) { m->mothurOut("\nRemoving group: " + groups[i] + " because all sequences have been removed.\n"); removeGroup(groups[i]); i--; }
}
}
}
if (groupfile != "") { delete groupMap; }
return 0;
}
catch(exception& e) {
m->errorOut(e, "CountTable", "createTable");
exit(1);
}
}
int ClassifySeqsCommand::execute(){
try {
if (abort) { if (calledHelp) { return 0; } return 2; }
string outputMethodTag = method;
if(method == "wang"){ classify = new Bayesian(taxonomyFileName, templateFileName, search, kmerSize, cutoff, iters, util.getRandomNumber(), flip, writeShortcuts, current->getVersion()); }
else if(method == "knn"){ classify = new Knn(taxonomyFileName, templateFileName, search, kmerSize, gapOpen, gapExtend, match, misMatch, numWanted, util.getRandomNumber(), current->getVersion()); }
else if(method == "zap"){
outputMethodTag = search + "_" + outputMethodTag;
if (search == "kmer") { classify = new KmerTree(templateFileName, taxonomyFileName, kmerSize, cutoff); }
else { classify = new AlignTree(templateFileName, taxonomyFileName, cutoff); }
}
else {
m->mothurOut(search + " is not a valid method option. I will run the command using wang.");
m->mothurOutEndLine();
classify = new Bayesian(taxonomyFileName, templateFileName, search, kmerSize, cutoff, iters, util.getRandomNumber(), flip, writeShortcuts, current->getVersion());
}
if (m->getControl_pressed()) { delete classify; return 0; }
m->mothurOut("Classifying sequences from " + fastafile + " ...\n" );
string baseTName = util.getSimpleName(taxonomyFileName);
//set rippedTaxName to
string RippedTaxName = "";
bool foundDot = false;
for (int i = baseTName.length()-1; i >= 0; i--) {
if (foundDot && (baseTName[i] != '.')) { RippedTaxName = baseTName[i] + RippedTaxName; }
else if (foundDot && (baseTName[i] == '.')) { break; }
else if (!foundDot && (baseTName[i] == '.')) { foundDot = true; }
}
if (outputDir == "") { outputDir += util.hasPath(fastafile); }
map<string, string> variables;
variables["[filename]"] = outputDir + util.getRootName(util.getSimpleName(fastafile));
variables["[tag]"] = RippedTaxName;
variables["[tag2]"] = outputMethodTag;
string newTaxonomyFile = getOutputFileName("taxonomy", variables);
string newaccnosFile = getOutputFileName("accnos", variables);
string tempTaxonomyFile = outputDir + util.getRootName(util.getSimpleName(fastafile)) + "taxonomy.temp";
string taxSummary = getOutputFileName("taxsummary", variables);
if ((method == "knn") && (search == "distance")) {
string DistName = getOutputFileName("matchdist", variables);
classify->setDistName(DistName); outputNames.push_back(DistName); outputTypes["matchdist"].push_back(DistName);
}
outputNames.push_back(newTaxonomyFile); outputTypes["taxonomy"].push_back(newTaxonomyFile);
outputNames.push_back(taxSummary); outputTypes["taxsummary"].push_back(taxSummary);
long start = time(NULL);
int numFastaSeqs = createProcesses(newTaxonomyFile, tempTaxonomyFile, newaccnosFile, fastafile);
if (!util.isBlank(newaccnosFile)) { m->mothurOut("\n[WARNING]: mothur reversed some your sequences for a better classification. If you would like to take a closer look, please check " + newaccnosFile + " for the list of the sequences.\n");
outputNames.push_back(newaccnosFile); outputTypes["accnos"].push_back(newaccnosFile);
}else { util.mothurRemove(newaccnosFile); }
m->mothurOut("\nIt took " + toString(time(NULL) - start) + " secs to classify " + toString(numFastaSeqs) + " sequences.\n\n");
start = time(NULL);
//read namefile
map<string, vector<string> > nameMap;
map<string, vector<string> >::iterator itNames;
if(namefile != "") {
m->mothurOut("Reading " + namefile + "..."); cout.flush();
nameMap.clear(); //remove old names
util.readNames(namefile, nameMap);
m->mothurOut(" Done.\n");
}
//output taxonomy with the unclassified bins added
ifstream inTax;
util.openInputFile(newTaxonomyFile, inTax);
ofstream outTax;
string unclass = newTaxonomyFile + ".unclass.temp";
util.openOutputFile(unclass, outTax);
//get maxLevel from phylotree so you know how many 'unclassified's to add
int maxLevel = classify->getMaxLevel();
//read taxfile - this reading and rewriting is done to preserve the confidence scores.
string name, taxon;
GroupMap* groupMap = NULL;
CountTable* ct = NULL;
PhyloSummary* taxaSum;
if (hasCount) {
ct = new CountTable();
ct->readTable(countfile, true, false);
taxaSum = new PhyloSummary(ct, relabund, printlevel);
}else {
if (groupfile != "") { groupMap = new GroupMap(groupfile); groupMap->readMap(); }
taxaSum = new PhyloSummary(groupMap, relabund, printlevel);
}
while (!inTax.eof()) {
if (m->getControl_pressed()) { outputTypes.clear(); if (ct != NULL) { delete ct; } if (groupMap != NULL) { delete groupMap; } delete taxaSum; for (int i = 0; i < outputNames.size(); i++) { util.mothurRemove(outputNames[i]); } delete classify; return 0; }
inTax >> name; util.gobble(inTax);
taxon = util.getline(inTax); util.gobble(inTax);
string newTax = util.addUnclassifieds(taxon, maxLevel, probs);
outTax << name << '\t' << newTax << endl;
if (namefile != "") {
itNames = nameMap.find(name);
if (itNames == nameMap.end()) {
m->mothurOut(name + " is not in your name file please correct.\n"); exit(1);
}else{
//add it as many times as there are identical seqs
for (int i = 0; i < itNames->second.size(); i++) { taxaSum->addSeqToTree(itNames->second[i], newTax); }
itNames->second.clear();
nameMap.erase(itNames->first);
}
}else { taxaSum->addSeqToTree(name, newTax); }
}
inTax.close();
outTax.close();
util.mothurRemove(newTaxonomyFile);
util.renameFile(unclass, newTaxonomyFile);
if (m->getControl_pressed()) { outputTypes.clear(); if (ct != NULL) { delete ct; } if (groupMap != NULL) { delete groupMap; } for (int i = 0; i < outputNames.size(); i++) { util.mothurRemove(outputNames[i]); } delete classify; return 0; }
//print summary file
ofstream outTaxTree;
util.openOutputFile(taxSummary, outTaxTree);
taxaSum->print(outTaxTree, output);
outTaxTree.close();
if (ct != NULL) { delete ct; }
if (groupMap != NULL) { delete groupMap; } delete taxaSum;
util.mothurRemove(tempTaxonomyFile);
delete classify;
m->mothurOut("\nIt took " + toString(time(NULL) - start) + " secs to create the summary file for " + toString(numFastaSeqs) + " sequences.\n\n");
m->mothurOut("\nOutput File Names: \n");
for (int i = 0; i < outputNames.size(); i++) { m->mothurOut(outputNames[i]); m->mothurOutEndLine(); }
m->mothurOutEndLine();
//set taxonomy file as new current taxonomyfile
string currentName = "";
itTypes = outputTypes.find("taxonomy");
if (itTypes != outputTypes.end()) { if ((itTypes->second).size() != 0) { currentName = (itTypes->second)[0]; current->setTaxonomyFile(currentName); } }
currentName = "";
itTypes = outputTypes.find("accnos");
if (itTypes != outputTypes.end()) { if ((itTypes->second).size() != 0) { currentName = (itTypes->second)[0]; current->setAccnosFile(currentName); } }
return 0;
}
catch(exception& e) {
m->errorOut(e, "ClassifySeqsCommand", "execute");
exit(1);
}
}
//**********************************************************************************************************************
int SharedCommand::createSharedFromListGroup() {
try {
GroupMap* groupMap = NULL;
CountTable* countTable = NULL;
pickedGroups = false;
if (groupfile != "") {
groupMap = new GroupMap(groupfile);
int groupError = groupMap->readMap();
if (groupError == 1) { delete groupMap; return 0; }
vector<string> allGroups = groupMap->getNamesOfGroups();
if (Groups.size() == 0) { Groups = allGroups; }
else { pickedGroups = true; }
}else{
countTable = new CountTable();
countTable->readTable(countfile, true, false);
vector<string> allGroups = countTable->getNamesOfGroups();
if (Groups.size() == 0) { Groups = allGroups; }
else { pickedGroups = true; }
}
int numGroups = Groups.size();
if (m->getControl_pressed()) { return 0; }
ofstream out;
string filename = "";
if (!pickedGroups) {
string filename = listfile;
if (outputDir == "") { outputDir += util.hasPath(filename); }
map<string, string> variables;
variables["[filename]"] = outputDir + util.getRootName(util.getSimpleName(filename));
filename = getOutputFileName("shared",variables);
outputNames.push_back(filename); outputTypes["shared"].push_back(filename);
util.openOutputFile(filename, out);
}
//set fileroot
fileroot = outputDir + util.getRootName(util.getSimpleName(listfile));
map<string, string> variables;
variables["[filename]"] = fileroot;
string errorOff = "no error";
InputData input(listfile, "shared", Groups);
SharedListVector* SharedList = input.getSharedListVector();
string lastLabel = SharedList->getLabel();
SharedRAbundVectors* lookup;
if (m->getControl_pressed()) {
delete SharedList; if (groupMap != NULL) { delete groupMap; } if (countTable != NULL) { delete countTable; }
out.close(); if (!pickedGroups) { util.mothurRemove(filename); }
return 0;
}
//sanity check
vector<string> namesSeqs;
int numGroupNames = 0;
if (current->getGroupMode() == "group") { namesSeqs = groupMap->getNamesSeqs(); numGroupNames = groupMap->getNumSeqs(); }
else { namesSeqs = countTable->getNamesOfSeqs(); numGroupNames = countTable->getNumUniqueSeqs(); }
int error = ListGroupSameSeqs(namesSeqs, SharedList);
if ((!pickedGroups) && (SharedList->getNumSeqs() != numGroupNames)) { //if the user has not specified any groups and their files don't match exit with error
m->mothurOut("Your group file contains " + toString(numGroupNames) + " sequences and list file contains " + toString(SharedList->getNumSeqs()) + " sequences. Please correct.\n"); m->setControl_pressed(true);
out.close(); if (!pickedGroups) { util.mothurRemove(filename); } //remove blank shared file you made
//delete memory
delete SharedList; if (groupMap != NULL) { delete groupMap; } if (countTable != NULL) { delete countTable; }
return 0;
}
if (error == 1) { m->setControl_pressed(true); }
//if user has specified groups make new groupfile for them
if ((pickedGroups) && (current->getGroupMode() == "group")) { //make new group file
string groups = "";
if (numGroups < 4) {
for (int i = 0; i < numGroups-1; i++) {
groups += Groups[i] + ".";
}
groups+=Groups[numGroups-1];
}else { groups = "merge"; }
map<string, string> variables;
variables["[filename]"] = outputDir + util.getRootName(util.getSimpleName(listfile));
variables["[group]"] = groups;
string newGroupFile = getOutputFileName("group",variables);
outputTypes["group"].push_back(newGroupFile);
outputNames.push_back(newGroupFile);
ofstream outGroups;
util.openOutputFile(newGroupFile, outGroups);
vector<string> names = groupMap->getNamesSeqs();
string groupName;
for (int i = 0; i < names.size(); i++) {
groupName = groupMap->getGroup(names[i]);
if (isValidGroup(groupName, Groups)) {
outGroups << names[i] << '\t' << groupName << endl;
}
}
outGroups.close();
}
//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;
bool printHeaders = true;
while((SharedList != NULL) && ((allLines == 1) || (userLabels.size() != 0))) {
if (m->getControl_pressed()) {
delete SharedList; if (groupMap != NULL) { delete groupMap; } if (countTable != NULL) { delete countTable; }
if (!pickedGroups) { out.close(); util.mothurRemove(filename); }
return 0;
}
if(allLines == 1 || labels.count(SharedList->getLabel()) == 1){
lookup = SharedList->getSharedRAbundVector();
m->mothurOut(lookup->getLabel()+"\n");
if (m->getControl_pressed()) {
delete SharedList; if (groupMap != NULL) { delete groupMap; } if (countTable != NULL) { delete countTable; }
delete lookup;
if (!pickedGroups) { out.close(); util.mothurRemove(filename); }
return 0;
}
//if picked groups must split the shared file by label
if (pickedGroups) {
string filename = listfile;
if (outputDir == "") { outputDir += util.hasPath(filename); }
map<string, string> variables;
variables["[filename]"] = outputDir + util.getRootName(util.getSimpleName(filename));
variables["[distance]"] = lookup->getLabel();
filename = getOutputFileName("shared",variables);
outputNames.push_back(filename); outputTypes["shared"].push_back(filename);
ofstream out2;
util.openOutputFile(filename, out2);
lookup->eliminateZeroOTUS();
printSharedData(lookup, out2, printHeaders);
out2.close();
}else {
printSharedData(lookup, out, printHeaders); //prints info to the .shared file
}
delete lookup;
processedLabels.insert(SharedList->getLabel());
userLabels.erase(SharedList->getLabel());
}
if ((util.anyLabelsToProcess(SharedList->getLabel(), userLabels, errorOff) ) && (processedLabels.count(lastLabel) != 1)) {
string saveLabel = SharedList->getLabel();
delete SharedList;
SharedList = input.getSharedListVector(lastLabel); //get new list vector to process
lookup = SharedList->getSharedRAbundVector();
m->mothurOut(lookup->getLabel()+"\n");
if (m->getControl_pressed()) {
delete SharedList; if (groupMap != NULL) { delete groupMap; } if (countTable != NULL) { delete countTable; }
delete lookup;
if (!pickedGroups) { out.close(); util.mothurRemove(filename); }
return 0;
}
//if picked groups must split the shared file by label
if (pickedGroups) {
string filename = listfile;
if (outputDir == "") { outputDir += util.hasPath(filename); }
map<string, string> variables;
variables["[filename]"] = outputDir + util.getRootName(util.getSimpleName(filename));
variables["[distance]"] = lookup->getLabel();
filename = getOutputFileName("shared",variables);
outputNames.push_back(filename); outputTypes["shared"].push_back(filename);
ofstream out2;
util.openOutputFile(filename, out2);
lookup->eliminateZeroOTUS();
printSharedData(lookup, out2, printHeaders);
out2.close();
}else {
printSharedData(lookup, out, printHeaders); //prints info to the .shared file
}
delete lookup;
processedLabels.insert(SharedList->getLabel());
userLabels.erase(SharedList->getLabel());
//restore real lastlabel to save below
SharedList->setLabel(saveLabel);
}
lastLabel = SharedList->getLabel();
delete SharedList;
SharedList = input.getSharedListVector(); //get new list vector to process
}
//output error messages about any remaining user labels
set<string>::iterator it;
bool needToRun = false;
for (it = userLabels.begin(); it != userLabels.end(); it++) {
if (processedLabels.count(lastLabel) != 1) {
needToRun = true;
}
}
//run last label if you need to
if (needToRun ) {
if (SharedList != NULL) { delete SharedList; }
SharedList = input.getSharedListVector(lastLabel); //get new list vector to process
lookup = SharedList->getSharedRAbundVector();
m->mothurOut(lookup->getLabel()+"\n");
if (m->getControl_pressed()) {
if (groupMap != NULL) { delete groupMap; } if (countTable != NULL) { delete countTable; }
if (!pickedGroups) { out.close(); util.mothurRemove(filename); }
return 0;
}
//if picked groups must split the shared file by label
if (pickedGroups) {
string filename = listfile;
if (outputDir == "") { outputDir += util.hasPath(filename); }
map<string, string> variables;
variables["[filename]"] = outputDir + util.getRootName(util.getSimpleName(filename));
variables["[distance]"] = lookup->getLabel();
filename = getOutputFileName("shared",variables);
outputNames.push_back(filename); outputTypes["shared"].push_back(filename);
ofstream out2;
util.openOutputFile(filename, out2);
lookup->eliminateZeroOTUS();
printSharedData(lookup, out2, printHeaders);
out2.close();
}else {
printSharedData(lookup, out, printHeaders); //prints info to the .shared file
}
delete lookup;
delete SharedList;
}
if (!pickedGroups) { out.close(); }
if (groupMap != NULL) { delete groupMap; } if (countTable != NULL) { delete countTable; }
if (m->getControl_pressed()) {
if (!pickedGroups) { util.mothurRemove(filename); }
return 0;
}
return 0;
}
catch(exception& e) {
m->errorOut(e, "SharedCommand", "createSharedFromListGroup");
exit(1);
}
}
int ClassifySeqsCommand::execute(){
try {
if (abort == true) { if (calledHelp) { return 0; } return 2; }
string outputMethodTag = method;
if(method == "wang"){ classify = new Bayesian(taxonomyFileName, templateFileName, search, kmerSize, cutoff, iters, rand(), flip, writeShortcuts); }
else if(method == "knn"){ classify = new Knn(taxonomyFileName, templateFileName, search, kmerSize, gapOpen, gapExtend, match, misMatch, numWanted, rand()); }
else if(method == "zap"){
outputMethodTag = search + "_" + outputMethodTag;
if (search == "kmer") { classify = new KmerTree(templateFileName, taxonomyFileName, kmerSize, cutoff); }
else { classify = new AlignTree(templateFileName, taxonomyFileName, cutoff); }
}
else {
m->mothurOut(search + " is not a valid method option. I will run the command using wang.");
m->mothurOutEndLine();
classify = new Bayesian(taxonomyFileName, templateFileName, search, kmerSize, cutoff, iters, rand(), flip, writeShortcuts);
}
if (m->control_pressed) { delete classify; return 0; }
for (int s = 0; s < fastaFileNames.size(); s++) {
m->mothurOut("Classifying sequences from " + fastaFileNames[s] + " ..." ); m->mothurOutEndLine();
string baseTName = m->getSimpleName(taxonomyFileName);
//set rippedTaxName to
string RippedTaxName = "";
bool foundDot = false;
for (int i = baseTName.length()-1; i >= 0; i--) {
if (foundDot && (baseTName[i] != '.')) { RippedTaxName = baseTName[i] + RippedTaxName; }
else if (foundDot && (baseTName[i] == '.')) { break; }
else if (!foundDot && (baseTName[i] == '.')) { foundDot = true; }
}
//if (RippedTaxName != "") { RippedTaxName += "."; }
if (outputDir == "") { outputDir += m->hasPath(fastaFileNames[s]); }
map<string, string> variables;
variables["[filename]"] = outputDir + m->getRootName(m->getSimpleName(fastaFileNames[s]));
variables["[tag]"] = RippedTaxName;
variables["[tag2]"] = outputMethodTag;
string newTaxonomyFile = getOutputFileName("taxonomy", variables);
string newaccnosFile = getOutputFileName("accnos", variables);
string tempTaxonomyFile = outputDir + m->getRootName(m->getSimpleName(fastaFileNames[s])) + "taxonomy.temp";
string taxSummary = getOutputFileName("taxsummary", variables);
if ((method == "knn") && (search == "distance")) {
string DistName = getOutputFileName("matchdist", variables);
classify->setDistName(DistName); outputNames.push_back(DistName); outputTypes["matchdist"].push_back(DistName);
}
outputNames.push_back(newTaxonomyFile); outputTypes["taxonomy"].push_back(newTaxonomyFile);
outputNames.push_back(taxSummary); outputTypes["taxsummary"].push_back(taxSummary);
int start = time(NULL);
int numFastaSeqs = 0;
for (int i = 0; i < lines.size(); i++) { delete lines[i]; } lines.clear();
vector<unsigned long long> positions;
#if defined (__APPLE__) || (__MACH__) || (linux) || (__linux) || (__linux__) || (__unix__) || (__unix)
positions = m->divideFile(fastaFileNames[s], processors);
for (int i = 0; i < (positions.size()-1); i++) { lines.push_back(new linePair(positions[i], positions[(i+1)])); }
#else
if (processors == 1) {
lines.push_back(new linePair(0, 1000));
}else {
positions = m->setFilePosFasta(fastaFileNames[s], numFastaSeqs);
if (numFastaSeqs < processors) { processors = numFastaSeqs; }
//figure out how many sequences you have to process
int numSeqsPerProcessor = numFastaSeqs / processors;
for (int i = 0; i < processors; i++) {
int startIndex = i * numSeqsPerProcessor;
if(i == (processors - 1)){ numSeqsPerProcessor = numFastaSeqs - i * numSeqsPerProcessor; }
lines.push_back(new linePair(positions[startIndex], numSeqsPerProcessor));
}
}
#endif
if(processors == 1){
numFastaSeqs = driver(lines[0], newTaxonomyFile, tempTaxonomyFile, newaccnosFile, fastaFileNames[s]);
}else{
numFastaSeqs = createProcesses(newTaxonomyFile, tempTaxonomyFile, newaccnosFile, fastaFileNames[s]);
}
if (!m->isBlank(newaccnosFile)) { m->mothurOutEndLine(); m->mothurOut("[WARNING]: mothur reversed some your sequences for a better classification. If you would like to take a closer look, please check " + newaccnosFile + " for the list of the sequences."); m->mothurOutEndLine();
outputNames.push_back(newaccnosFile); outputTypes["accnos"].push_back(newaccnosFile);
}else { m->mothurRemove(newaccnosFile); }
m->mothurOutEndLine();
m->mothurOut("It took " + toString(time(NULL) - start) + " secs to classify " + toString(numFastaSeqs) + " sequences."); m->mothurOutEndLine(); m->mothurOutEndLine();
start = time(NULL);
//read namefile
if(namefile != "") {
m->mothurOut("Reading " + namefileNames[s] + "..."); cout.flush();
nameMap.clear(); //remove old names
m->readNames(namefileNames[s], nameMap);
m->mothurOut(" Done."); m->mothurOutEndLine();
}
//output taxonomy with the unclassified bins added
ifstream inTax;
m->openInputFile(newTaxonomyFile, inTax);
ofstream outTax;
string unclass = newTaxonomyFile + ".unclass.temp";
m->openOutputFile(unclass, outTax);
//get maxLevel from phylotree so you know how many 'unclassified's to add
int maxLevel = classify->getMaxLevel();
//read taxfile - this reading and rewriting is done to preserve the confidence scores.
string name, taxon;
string group = "";
GroupMap* groupMap = NULL;
CountTable* ct = NULL;
PhyloSummary* taxaSum;
if (hasCount) {
ct = new CountTable();
ct->readTable(countfileNames[s], true, false);
taxaSum = new PhyloSummary(ct, relabund, printlevel);
}else {
if (groupfile != "") { group = groupfileNames[s]; groupMap = new GroupMap(group); groupMap->readMap(); }
taxaSum = new PhyloSummary(groupMap, relabund, printlevel);
}
while (!inTax.eof()) {
if (m->control_pressed) { outputTypes.clear(); if (ct != NULL) { delete ct; } if (groupMap != NULL) { delete groupMap; } delete taxaSum; for (int i = 0; i < outputNames.size(); i++) { m->mothurRemove(outputNames[i]); } delete classify; return 0; }
inTax >> name >> taxon; m->gobble(inTax);
string newTax = m->addUnclassifieds(taxon, maxLevel, probs);
outTax << name << '\t' << newTax << endl;
if (namefile != "") {
itNames = nameMap.find(name);
if (itNames == nameMap.end()) {
m->mothurOut(name + " is not in your name file please correct."); m->mothurOutEndLine(); exit(1);
}else{
for (int i = 0; i < itNames->second.size(); i++) {
taxaSum->addSeqToTree(itNames->second[i], newTax); //add it as many times as there are identical seqs
}
itNames->second.clear();
nameMap.erase(itNames->first);
}
}else {
taxaSum->addSeqToTree(name, newTax);
}
}
inTax.close();
outTax.close();
m->mothurRemove(newTaxonomyFile);
rename(unclass.c_str(), newTaxonomyFile.c_str());
if (m->control_pressed) { outputTypes.clear(); if (ct != NULL) { delete ct; } if (groupMap != NULL) { delete groupMap; } for (int i = 0; i < outputNames.size(); i++) { m->mothurRemove(outputNames[i]); } delete classify; return 0; }
//print summary file
ofstream outTaxTree;
m->openOutputFile(taxSummary, outTaxTree);
taxaSum->print(outTaxTree, output);
outTaxTree.close();
if (ct != NULL) { delete ct; }
if (groupMap != NULL) { delete groupMap; } delete taxaSum;
m->mothurRemove(tempTaxonomyFile);
m->mothurOutEndLine();
m->mothurOut("It took " + toString(time(NULL) - start) + " secs to create the summary file for " + toString(numFastaSeqs) + " sequences."); m->mothurOutEndLine(); m->mothurOutEndLine();
}
delete classify;
m->mothurOutEndLine();
m->mothurOut("Output File Names: "); m->mothurOutEndLine();
for (int i = 0; i < outputNames.size(); i++) { m->mothurOut(outputNames[i]); m->mothurOutEndLine(); }
m->mothurOutEndLine();
//set taxonomy file as new current taxonomyfile
string current = "";
itTypes = outputTypes.find("taxonomy");
if (itTypes != outputTypes.end()) {
if ((itTypes->second).size() != 0) { current = (itTypes->second)[0]; m->setTaxonomyFile(current); }
}
current = "";
itTypes = outputTypes.find("accnos");
if (itTypes != outputTypes.end()) {
if ((itTypes->second).size() != 0) { current = (itTypes->second)[0]; m->setAccnosFile(current); }
}
return 0;
}
catch(exception& e) {
m->errorOut(e, "ClassifySeqsCommand", "execute");
exit(1);
}
}