CommandParameter pcutoff("cutoff", "Number", "", "0.01", "", "", "",false,false); parameters.push_back(pcutoff);
CommandParameter pprocessors("processors", "Number", "", "1", "", "", "",false,false); parameters.push_back(pprocessors);
CommandParameter pmaxiter("maxiter", "Number", "", "1000", "", "", "",false,false); parameters.push_back(pmaxiter);
+ CommandParameter plarge("large", "Number", "", "-1", "", "", "",false,false); parameters.push_back(plarge);
CommandParameter psigma("sigma", "Number", "", "60", "", "", "",false,false); parameters.push_back(psigma);
CommandParameter pmindelta("mindelta", "Number", "", "0.000001", "", "", "",false,false); parameters.push_back(pmindelta);
CommandParameter porder("order", "String", "", "", "", "", "",false,false); parameters.push_back(porder);
}
}
//**********************************************************************************************************************
+string ShhherCommand::getOutputFileNameTag(string type, string inputName=""){
+ try {
+ string outputFileName = "";
+ map<string, vector<string> >::iterator it;
+
+ //is this a type this command creates
+ it = outputTypes.find(type);
+ if (it == outputTypes.end()) { m->mothurOut("[ERROR]: this command doesn't create a " + type + " output file.\n"); }
+ else {
+ if (type == "fasta") { outputFileName = "shhh.fasta"; }
+ else if (type == "name") { outputFileName = "shhh.names"; }
+ else if (type == "group") { outputFileName = "shhh.groups"; }
+ else if (type == "counts") { outputFileName = "shhh.counts"; }
+ else if (type == "qfile") { outputFileName = "shhh.qual"; }
+ else { m->mothurOut("[ERROR]: No definition for type " + type + " output file tag.\n"); m->control_pressed = true; }
+ }
+ return outputFileName;
+ }
+ catch(exception& e) {
+ m->errorOut(e, "ShhherCommand", "getOutputFileNameTag");
+ exit(1);
+ }
+}
+//**********************************************************************************************************************
ShhherCommand::ShhherCommand(){
try {
setParameters();
//initialize outputTypes
-// vector<string> tempOutNames;
-// outputTypes["pn.dist"] = tempOutNames;
+ vector<string> tempOutNames;
+ outputTypes["fasta"] = tempOutNames;
+ outputTypes["name"] = tempOutNames;
+ outputTypes["group"] = tempOutNames;
+ outputTypes["counts"] = tempOutNames;
+ outputTypes["qfile"] = tempOutNames;
}
catch(exception& e) {
}
//initialize outputTypes
- vector<string> tempOutNames;
-// outputTypes["pn.dist"] = tempOutNames;
- // outputTypes["fasta"] = tempOutNames;
+ vector<string> tempOutNames;
+ outputTypes["fasta"] = tempOutNames;
+ outputTypes["name"] = tempOutNames;
+ outputTypes["group"] = tempOutNames;
+ outputTypes["counts"] = tempOutNames;
+ outputTypes["qfile"] = tempOutNames;
+
//if the user changes the input directory command factory will send this info to us in the output parameter
string inputDir = validParameter.validFile(parameters, "inputdir", false);
if (path == "") { parameters["file"] = inputDir + it->second; }
}
}
-
-
+
+ //if the user changes the output directory command factory will send this info to us in the output parameter
+ outputDir = validParameter.validFile(parameters, "outputdir", false); if (outputDir == "not found"){ outputDir = ""; }
+
//check for required parameters
flowFileName = validParameter.validFile(parameters, "flow", true);
flowFilesFileName = validParameter.validFile(parameters, "file", true);
if (flowFileName == "not found" && flowFilesFileName == "not found") {
- m->mothurOut("values for either flow or file must be provided for the shhh.seqs command.");
+ m->mothurOut("values for either flow or file must be provided for the shhh.flows command.");
m->mothurOutEndLine();
abort = true;
}
}
else{
ofstream temp;
-
- //flow.files = 9 character offset
- compositeFASTAFileName = flowFilesFileName.substr(0, flowFilesFileName.length()-10) + "shhh.fasta";
+
+ string thisoutputDir = outputDir;
+ if (outputDir == "") { thisoutputDir = m->hasPath(flowFilesFileName); } //if user entered a file with a path then preserve it
+
+ //we want to rip off .files, and also .flow if its there
+ string fileroot = m->getRootName(m->getSimpleName(flowFilesFileName));
+ if (fileroot[fileroot.length()-1] == '.') { fileroot = fileroot.substr(0, fileroot.length()-1); } //rip off dot
+ string extension = m->getExtension(fileroot);
+ if (extension == ".flow") { fileroot = m->getRootName(fileroot); }
+ else { fileroot += "."; } //add back if needed
+
+ compositeFASTAFileName = thisoutputDir + fileroot + "shhh.fasta";
m->openOutputFile(compositeFASTAFileName, temp);
temp.close();
- compositeNamesFileName = flowFilesFileName.substr(0, flowFilesFileName.length()-10) + "shhh.names";
+ compositeNamesFileName = thisoutputDir + fileroot + "shhh.names";
m->openOutputFile(compositeNamesFileName, temp);
temp.close();
}
if (flowFileVector.size() == 0) { m->mothurOut("[ERROR]: no valid files."); m->mothurOutEndLine(); abort = true; }
}
else{
+ if (outputDir == "") { outputDir = m->hasPath(flowFileName); }
flowFileVector.push_back(flowFileName);
}
-
-
- //if the user changes the output directory command factory will send this info to us in the output parameter
- outputDir = validParameter.validFile(parameters, "outputdir", false); if (outputDir == "not found"){
- outputDir = "";
- outputDir += m->hasPath(flowFileName); //if user entered a file with a path then preserve it
- }
-
-
+
//check for optional parameter and set defaults
// ...at some point should added some additional type checking...
string temp;
temp = validParameter.validFile(parameters, "maxiter", false); if (temp == "not found"){ temp = "1000"; }
m->mothurConvert(temp, maxIters);
+
+ temp = validParameter.validFile(parameters, "large", false); if (temp == "not found"){ temp = "0"; }
+ m->mothurConvert(temp, largeSize);
+ if (largeSize != 0) { large = true; }
+ else { large = false; }
+ if (largeSize < 0) { m->mothurOut("The value of the large cannot be negative.\n"); }
+
+#ifdef USE_MPI
+ if (large) { m->mothurOut("The large parameter is not available with the MPI-Enabled version.\n"); large=false; }
+#endif
+
temp = validParameter.validFile(parameters, "sigma", false);if (temp == "not found") { temp = "60"; }
m->mothurConvert(temp, sigma);
duplicateNames[mapSeqToUnique[i]] += seqNameVector[i] + ',';
}
- string nameFileName = flowFileName.substr(0,flowFileName.find_last_of('.')) + ".shhh.names";
+ string nameFileName = outputDir + m->getRootName(m->getSimpleName(flowFileName) + getOutputFileNameTag("name");
ofstream nameFile;
m->openOutputFile(nameFileName, nameFile);
try{
if (numOTUs < processors) { processors = 1; }
+ if (m->debug) { m->mothurOut("[DEBUG]: numSeqs = " + toString(numSeqs) + " numOTUS = " + toString(numOTUs) + " about to alloc a dist vector with size = " + toString((numSeqs * numOTUs)) + ".\n"); }
+
dist.assign(numSeqs * numOTUs, 0);
change.assign(numOTUs, 1);
centroids.assign(numOTUs, -1);
nSeqsBreaks.assign(processors+1, 0);
nOTUsBreaks.assign(processors+1, 0);
+ if (m->debug) { m->mothurOut("[DEBUG]: made it through the memory allocation.\n"); }
+
nSeqsBreaks[0] = 0;
for(int i=0;i<processors;i++){
nSeqsBreaks[i+1] = nSeqsBreaks[i] + (int)((double) numSeqs / (double) processors);
try {
string thisOutputDir = outputDir;
if (outputDir == "") { thisOutputDir += m->hasPath(flowFileName); }
- string qualityFileName = thisOutputDir + m->getRootName(m->getSimpleName(flowFileName)) + "shhh.qual";
+ string qualityFileName = thisOutputDir + m->getRootName(m->getSimpleName(flowFileName)) + getOutputFileNameTag("qfile");
ofstream qualityFile;
m->openOutputFile(qualityFileName, qualityFile);
try {
string thisOutputDir = outputDir;
if (outputDir == "") { thisOutputDir += m->hasPath(flowFileName); }
- string fastaFileName = thisOutputDir + m->getRootName(m->getSimpleName(flowFileName)) + "shhh.fasta";
+ string fastaFileName = thisOutputDir + m->getRootName(m->getSimpleName(flowFileName)) + getOutputFileNameTag("fasta");
ofstream fastaFile;
m->openOutputFile(fastaFileName, fastaFile);
outputNames.push_back(fastaFileName);
if(compositeFASTAFileName != ""){
- m->appendFiles(fastaFileName, thisOutputDir + compositeFASTAFileName);
+ m->appendFiles(fastaFileName, compositeFASTAFileName);
}
}
catch(exception& e) {
try {
string thisOutputDir = outputDir;
if (outputDir == "") { thisOutputDir += m->hasPath(flowFileName); }
- string nameFileName = thisOutputDir + m->getRootName(m->getSimpleName(flowFileName)) + "shhh.names";
+ string nameFileName = thisOutputDir + m->getRootName(m->getSimpleName(flowFileName)) + getOutputFileNameTag("name");
ofstream nameFile;
m->openOutputFile(nameFileName, nameFile);
if(compositeNamesFileName != ""){
- m->appendFiles(nameFileName, thisOutputDir + uchimecompositeNamesFileName);
+ m->appendFiles(nameFileName, compositeNamesFileName);
}
}
catch(exception& e) {
try {
string thisOutputDir = outputDir;
if (outputDir == "") { thisOutputDir += m->hasPath(flowFileName); }
- string fileRoot = thisOutputDir + m->getRootName(m->getSimpleName(flowFileName));
- string groupFileName = fileRoot + "shhh.groups";
+ string fileRoot = m->getRootName(m->getSimpleName(flowFileName));
+ int pos = fileRoot.find_first_of('.');
+ string fileGroup = fileRoot;
+ if (pos != string::npos) { fileGroup = fileRoot.substr(pos+1, (fileRoot.length()-1-(pos+1))); }
+ string groupFileName = thisOutputDir + fileRoot + getOutputFileNameTag("group");
ofstream groupFile;
m->openOutputFile(groupFileName, groupFile);
for(int i=0;i<numSeqs;i++){
if (m->control_pressed) { break; }
- groupFile << seqNameVector[i] << '\t' << fileRoot << endl;
+ groupFile << seqNameVector[i] << '\t' << fileGroup << endl;
}
groupFile.close();
outputNames.push_back(groupFileName);
try {
string thisOutputDir = outputDir;
if (outputDir == "") { thisOutputDir += m->hasPath(flowFileName); }
- string otuCountsFileName = thisOutputDir + m->getRootName(m->getSimpleName(flowFileName)) + "shhh.counts";
+ string otuCountsFileName = thisOutputDir + m->getRootName(m->getSimpleName(flowFileName)) +getOutputFileNameTag("counts");
ofstream otuCountsFile;
m->openOutputFile(otuCountsFileName, otuCountsFile);
//divide the groups between the processors
vector<linePair> lines;
+ vector<int> numFilesToComplete;
int numFilesPerProcessor = filenames.size() / processors;
for (int i = 0; i < processors; i++) {
int startIndex = i * numFilesPerProcessor;
int endIndex = (i+1) * numFilesPerProcessor;
if(i == (processors - 1)){ endIndex = filenames.size(); }
lines.push_back(linePair(startIndex, endIndex));
+ numFilesToComplete.push_back((endIndex-startIndex));
}
#if defined (__APPLE__) || (__MACH__) || (linux) || (__linux) || (__linux__) || (__unix__) || (__unix)
process++;
}else if (pid == 0){
num = driver(filenames, compositeFASTAFileName + toString(getpid()) + ".temp", compositeNamesFileName + toString(getpid()) + ".temp", lines[process].start, lines[process].end);
+
+ //pass numSeqs to parent
+ ofstream out;
+ string tempFile = compositeFASTAFileName + toString(getpid()) + ".num.temp";
+ m->openOutputFile(tempFile, out);
+ out << num << endl;
+ out.close();
+
exit(0);
}else {
m->mothurOut("[ERROR]: unable to spawn the necessary processes."); m->mothurOutEndLine();
#endif
for (int i=0;i<processIDS.size();i++) {
+ ifstream in;
+ string tempFile = compositeFASTAFileName + toString(processIDS[i]) + ".num.temp";
+ m->openInputFile(tempFile, in);
+ if (!in.eof()) {
+ int tempNum = 0;
+ in >> tempNum;
+ if (tempNum != numFilesToComplete[i+1]) {
+ m->mothurOut("[ERROR]: main process expected " + toString(processIDS[i]) + " to complete " + toString(numFilesToComplete[i+1]) + " files, and it only reported completing " + toString(tempNum) + ". This will cause file mismatches. The flow files may be too large to process with multiple processors. \n");
+ }
+ }
+ in.close(); m->mothurRemove(tempFile);
+
if (compositeFASTAFileName != "") {
m->appendFiles((compositeFASTAFileName + toString(processIDS[i]) + ".temp"), compositeFASTAFileName);
m->appendFiles((compositeNamesFileName + toString(processIDS[i]) + ".temp"), compositeNamesFileName);
}
/**************************************************************************************************/
-int ShhherCommand::driver(vector<string> filenames, string thisCompositeFASTAFileName, string thisCompositeNamesFileName, int start, int end){
+vector<string> ShhherCommand::parseFlowFiles(string filename){
try {
+ vector<string> files;
+ int count = 0;
- for(int i=start;i<end;i++){
-
- if (m->control_pressed) { break; }
-
- string flowFileName = filenames[i];
-
- m->mothurOut("\n>>>>>\tProcessing " + flowFileName + " (file " + toString(i+1) + " of " + toString(filenames.size()) + ")\t<<<<<\n");
- m->mothurOut("Reading flowgrams...\n");
-
- vector<string> seqNameVector;
- vector<int> lengths;
- vector<short> flowDataIntI;
- vector<double> flowDataPrI;
- map<string, int> nameMap;
- vector<short> uniqueFlowgrams;
- vector<int> uniqueCount;
- vector<int> mapSeqToUnique;
- vector<int> mapUniqueToSeq;
- vector<int> uniqueLengths;
- int numFlowCells;
-
- int numSeqs = getFlowData(flowFileName, seqNameVector, lengths, flowDataIntI, nameMap, numFlowCells);
-
- if (m->control_pressed) { break; }
-
- m->mothurOut("Identifying unique flowgrams...\n");
- int numUniques = getUniques(numSeqs, numFlowCells, uniqueFlowgrams, uniqueCount, uniqueLengths, mapSeqToUnique, mapUniqueToSeq, lengths, flowDataPrI, flowDataIntI);
-
- if (m->control_pressed) { break; }
-
- m->mothurOut("Calculating distances between flowgrams...\n");
- string distFileName = flowFileName.substr(0,flowFileName.find_last_of('.')) + ".shhh.dist";
- unsigned long long begTime = time(NULL);
- double begClock = clock();
+ ifstream in;
+ m->openInputFile(filename, in);
- flowDistParentFork(numFlowCells, distFileName, numUniques, mapUniqueToSeq, mapSeqToUnique, lengths, flowDataPrI, flowDataIntI);
-
- m->mothurOutEndLine();
- m->mothurOut("Total time: " + toString(time(NULL) - begTime) + '\t' + toString((clock() - begClock)/CLOCKS_PER_SEC) + '\n');
-
+ int thisNumFLows = 0;
+ in >> thisNumFLows; m->gobble(in);
+
+ while (!in.eof()) {
+ if (m->control_pressed) { break; }
- string namesFileName = flowFileName.substr(0,flowFileName.find_last_of('.')) + ".shhh.names";
- createNamesFile(numSeqs, numUniques, namesFileName, seqNameVector, mapSeqToUnique, mapUniqueToSeq);
-
- if (m->control_pressed) { break; }
-
- m->mothurOut("\nClustering flowgrams...\n");
- string listFileName = flowFileName.substr(0,flowFileName.find_last_of('.')) + ".shhh.list";
- cluster(listFileName, distFileName, namesFileName);
-
- if (m->control_pressed) { break; }
+ ofstream out;
+ string outputFileName = filename + toString(count) + ".temp";
+ m->openOutputFile(outputFileName, out);
+ out << thisNumFLows << endl;
+ files.push_back(outputFileName);
+
+ int numLinesWrote = 0;
+ for (int i = 0; i < largeSize; i++) {
+ if (in.eof()) { break; }
+ string line = m->getline(in); m->gobble(in);
+ out << line << endl;
+ numLinesWrote++;
+ }
+ out.close();
- vector<int> otuData;
- vector<int> cumNumSeqs;
- vector<int> nSeqsPerOTU;
- vector<vector<int> > aaP; //tMaster->aanP: each row is a different otu / each col contains the sequence indices
- vector<vector<int> > aaI; //tMaster->aanI: that are in each otu - can't differentiate between aaP and aaI
- vector<int> seqNumber; //tMaster->anP: the sequence id number sorted by OTU
- vector<int> seqIndex; //tMaster->anI; the index that corresponds to seqNumber
+ if (numLinesWrote == 0) { m->mothurRemove(outputFileName); files.pop_back(); }
+ count++;
+ }
+ in.close();
+
+ if (m->control_pressed) { for (int i = 0; i < files.size(); i++) { m->mothurRemove(files[i]); } files.clear(); }
+
+ m->mothurOut("\nDivided " + filename + " into " + toString(files.size()) + " files.\n\n");
+
+ return files;
+ }
+ catch(exception& e) {
+ m->errorOut(e, "ShhherCommand", "parseFlowFiles");
+ exit(1);
+ }
+}
+/**************************************************************************************************/
-
- int numOTUs = getOTUData(numSeqs, listFileName, otuData, cumNumSeqs, nSeqsPerOTU, aaP, aaI, seqNumber, seqIndex, nameMap);
+int ShhherCommand::driver(vector<string> filenames, string thisCompositeFASTAFileName, string thisCompositeNamesFileName, int start, int end){
+ try {
+
+ int numCompleted = 0;
+
+ for(int i=start;i<end;i++){
if (m->control_pressed) { break; }
- m->mothurRemove(distFileName);
- m->mothurRemove(namesFileName);
- m->mothurRemove(listFileName);
-
- vector<double> dist; //adDist - distance of sequences to centroids
- vector<short> change; //did the centroid sequence change? 0 = no; 1 = yes
- vector<int> centroids; //the representative flowgram for each cluster m
- vector<double> weight;
- vector<double> singleTau; //tMaster->adTau: 1-D Tau vector (1xnumSeqs)
- vector<int> nSeqsBreaks;
- vector<int> nOTUsBreaks;
-
- dist.assign(numSeqs * numOTUs, 0);
- change.assign(numOTUs, 1);
- centroids.assign(numOTUs, -1);
- weight.assign(numOTUs, 0);
- singleTau.assign(numSeqs, 1.0);
+ vector<string> theseFlowFileNames; theseFlowFileNames.push_back(filenames[i]);
+ if (large) { theseFlowFileNames = parseFlowFiles(filenames[i]); }
- nSeqsBreaks.assign(2, 0);
- nOTUsBreaks.assign(2, 0);
+ if (m->control_pressed) { break; }
- nSeqsBreaks[0] = 0;
- nSeqsBreaks[1] = numSeqs;
- nOTUsBreaks[1] = numOTUs;
-
- if (m->control_pressed) { break; }
-
- double maxDelta = 0;
- int iter = 0;
-
- begClock = clock();
- begTime = time(NULL);
+ double begClock = clock();
+ unsigned long long begTime;
- m->mothurOut("\nDenoising flowgrams...\n");
- m->mothurOut("iter\tmaxDelta\tnLL\t\tcycletime\n");
-
- while((maxIters == 0 && maxDelta > minDelta) || iter < MIN_ITER || (maxDelta > minDelta && iter < maxIters)){
-
- if (m->control_pressed) { break; }
-
- double cycClock = clock();
- unsigned long long cycTime = time(NULL);
- fill(numOTUs, seqNumber, seqIndex, cumNumSeqs, nSeqsPerOTU, aaP, aaI);
-
- if (m->control_pressed) { break; }
+ for (int g = 0; g < theseFlowFileNames.size(); g++) {
- calcCentroidsDriver(numOTUs, cumNumSeqs, nSeqsPerOTU, seqIndex, change, centroids, singleTau, mapSeqToUnique, uniqueFlowgrams, flowDataIntI, lengths, numFlowCells, seqNumber);
-
- if (m->control_pressed) { break; }
+ string flowFileName = theseFlowFileNames[g];
+ m->mothurOut("\n>>>>>\tProcessing " + flowFileName + " (file " + toString(i+1) + " of " + toString(filenames.size()) + ")\t<<<<<\n");
+ m->mothurOut("Reading flowgrams...\n");
+
+ vector<string> seqNameVector;
+ vector<int> lengths;
+ vector<short> flowDataIntI;
+ vector<double> flowDataPrI;
+ map<string, int> nameMap;
+ vector<short> uniqueFlowgrams;
+ vector<int> uniqueCount;
+ vector<int> mapSeqToUnique;
+ vector<int> mapUniqueToSeq;
+ vector<int> uniqueLengths;
+ int numFlowCells;
- maxDelta = getNewWeights(numOTUs, cumNumSeqs, nSeqsPerOTU, singleTau, seqNumber, weight);
+ if (m->debug) { m->mothurOut("[DEBUG]: About to read flowgrams.\n"); }
+ int numSeqs = getFlowData(flowFileName, seqNameVector, lengths, flowDataIntI, nameMap, numFlowCells);
if (m->control_pressed) { break; }
- double nLL = getLikelihood(numSeqs, numOTUs, nSeqsPerOTU, seqNumber, cumNumSeqs, seqIndex, dist, weight);
+ m->mothurOut("Identifying unique flowgrams...\n");
+ int numUniques = getUniques(numSeqs, numFlowCells, uniqueFlowgrams, uniqueCount, uniqueLengths, mapSeqToUnique, mapUniqueToSeq, lengths, flowDataPrI, flowDataIntI);
if (m->control_pressed) { break; }
- checkCentroids(numOTUs, centroids, weight);
-
- if (m->control_pressed) { break; }
-
- calcNewDistances(numSeqs, numOTUs, nSeqsPerOTU, dist, weight, change, centroids, aaP, singleTau, aaI, seqNumber, seqIndex, uniqueFlowgrams, flowDataIntI, numFlowCells, lengths);
-
- if (m->control_pressed) { break; }
-
- iter++;
-
- m->mothurOut(toString(iter) + '\t' + toString(maxDelta) + '\t' + toString(nLL) + '\t' + toString(time(NULL) - cycTime) + '\t' + toString((clock() - cycClock)/(double)CLOCKS_PER_SEC) + '\n');
+ m->mothurOut("Calculating distances between flowgrams...\n");
+ string distFileName = flowFileName.substr(0,flowFileName.find_last_of('.')) + ".shhh.dist";
+ begTime = time(NULL);
+
- }
-
- if (m->control_pressed) { break; }
-
- m->mothurOut("\nFinalizing...\n");
- fill(numOTUs, seqNumber, seqIndex, cumNumSeqs, nSeqsPerOTU, aaP, aaI);
-
- if (m->control_pressed) { break; }
-
- setOTUs(numOTUs, numSeqs, seqNumber, seqIndex, cumNumSeqs, nSeqsPerOTU, otuData, singleTau, dist, aaP, aaI);
-
- if (m->control_pressed) { break; }
-
- vector<int> otuCounts(numOTUs, 0);
- for(int i=0;i<numSeqs;i++) { otuCounts[otuData[i]]++; }
-
- calcCentroidsDriver(numOTUs, cumNumSeqs, nSeqsPerOTU, seqIndex, change, centroids, singleTau, mapSeqToUnique, uniqueFlowgrams, flowDataIntI, lengths, numFlowCells, seqNumber);
-
- if (m->control_pressed) { break; }
+ flowDistParentFork(numFlowCells, distFileName, numUniques, mapUniqueToSeq, mapSeqToUnique, lengths, flowDataPrI, flowDataIntI);
+
+ m->mothurOutEndLine();
+ m->mothurOut("Total time: " + toString(time(NULL) - begTime) + '\t' + toString((clock() - begClock)/CLOCKS_PER_SEC) + '\n');
+
+
+ string namesFileName = flowFileName.substr(0,flowFileName.find_last_of('.')) + ".shhh.names";
+ createNamesFile(numSeqs, numUniques, namesFileName, seqNameVector, mapSeqToUnique, mapUniqueToSeq);
+
+ if (m->control_pressed) { break; }
+
+ m->mothurOut("\nClustering flowgrams...\n");
+ string listFileName = flowFileName.substr(0,flowFileName.find_last_of('.')) + ".shhh.list";
+ cluster(listFileName, distFileName, namesFileName);
+
+ if (m->control_pressed) { break; }
+
+ vector<int> otuData;
+ vector<int> cumNumSeqs;
+ vector<int> nSeqsPerOTU;
+ vector<vector<int> > aaP; //tMaster->aanP: each row is a different otu / each col contains the sequence indices
+ vector<vector<int> > aaI; //tMaster->aanI: that are in each otu - can't differentiate between aaP and aaI
+ vector<int> seqNumber; //tMaster->anP: the sequence id number sorted by OTU
+ vector<int> seqIndex; //tMaster->anI; the index that corresponds to seqNumber
+
+
+ int numOTUs = getOTUData(numSeqs, listFileName, otuData, cumNumSeqs, nSeqsPerOTU, aaP, aaI, seqNumber, seqIndex, nameMap);
+
+ if (m->control_pressed) { break; }
+
+ m->mothurRemove(distFileName);
+ m->mothurRemove(namesFileName);
+ m->mothurRemove(listFileName);
+
+ vector<double> dist; //adDist - distance of sequences to centroids
+ vector<short> change; //did the centroid sequence change? 0 = no; 1 = yes
+ vector<int> centroids; //the representative flowgram for each cluster m
+ vector<double> weight;
+ vector<double> singleTau; //tMaster->adTau: 1-D Tau vector (1xnumSeqs)
+ vector<int> nSeqsBreaks;
+ vector<int> nOTUsBreaks;
+
+ if (m->debug) { m->mothurOut("[DEBUG]: numSeqs = " + toString(numSeqs) + " numOTUS = " + toString(numOTUs) + " about to alloc a dist vector with size = " + toString((numSeqs * numOTUs)) + ".\n"); }
+
+ dist.assign(numSeqs * numOTUs, 0);
+ change.assign(numOTUs, 1);
+ centroids.assign(numOTUs, -1);
+ weight.assign(numOTUs, 0);
+ singleTau.assign(numSeqs, 1.0);
+
+ nSeqsBreaks.assign(2, 0);
+ nOTUsBreaks.assign(2, 0);
+
+ nSeqsBreaks[0] = 0;
+ nSeqsBreaks[1] = numSeqs;
+ nOTUsBreaks[1] = numOTUs;
+
+ if (m->debug) { m->mothurOut("[DEBUG]: done allocating memory, about to denoise.\n"); }
+
+ if (m->control_pressed) { break; }
+
+ double maxDelta = 0;
+ int iter = 0;
+
+ begClock = clock();
+ begTime = time(NULL);
+
+ m->mothurOut("\nDenoising flowgrams...\n");
+ m->mothurOut("iter\tmaxDelta\tnLL\t\tcycletime\n");
+
+ while((maxIters == 0 && maxDelta > minDelta) || iter < MIN_ITER || (maxDelta > minDelta && iter < maxIters)){
+
+ if (m->control_pressed) { break; }
+
+ double cycClock = clock();
+ unsigned long long cycTime = time(NULL);
+ fill(numOTUs, seqNumber, seqIndex, cumNumSeqs, nSeqsPerOTU, aaP, aaI);
+
+ if (m->control_pressed) { break; }
+
+ calcCentroidsDriver(numOTUs, cumNumSeqs, nSeqsPerOTU, seqIndex, change, centroids, singleTau, mapSeqToUnique, uniqueFlowgrams, flowDataIntI, lengths, numFlowCells, seqNumber);
+
+ if (m->control_pressed) { break; }
+
+ maxDelta = getNewWeights(numOTUs, cumNumSeqs, nSeqsPerOTU, singleTau, seqNumber, weight);
+
+ if (m->control_pressed) { break; }
+
+ double nLL = getLikelihood(numSeqs, numOTUs, nSeqsPerOTU, seqNumber, cumNumSeqs, seqIndex, dist, weight);
+
+ if (m->control_pressed) { break; }
+
+ checkCentroids(numOTUs, centroids, weight);
+
+ if (m->control_pressed) { break; }
+
+ calcNewDistances(numSeqs, numOTUs, nSeqsPerOTU, dist, weight, change, centroids, aaP, singleTau, aaI, seqNumber, seqIndex, uniqueFlowgrams, flowDataIntI, numFlowCells, lengths);
+
+ if (m->control_pressed) { break; }
+
+ iter++;
+
+ m->mothurOut(toString(iter) + '\t' + toString(maxDelta) + '\t' + toString(nLL) + '\t' + toString(time(NULL) - cycTime) + '\t' + toString((clock() - cycClock)/(double)CLOCKS_PER_SEC) + '\n');
+
+ }
+
+ if (m->control_pressed) { break; }
+
+ m->mothurOut("\nFinalizing...\n");
+ fill(numOTUs, seqNumber, seqIndex, cumNumSeqs, nSeqsPerOTU, aaP, aaI);
+
+ if (m->control_pressed) { break; }
+
+ setOTUs(numOTUs, numSeqs, seqNumber, seqIndex, cumNumSeqs, nSeqsPerOTU, otuData, singleTau, dist, aaP, aaI);
+
+ if (m->control_pressed) { break; }
+
+ vector<int> otuCounts(numOTUs, 0);
+ for(int i=0;i<numSeqs;i++) { otuCounts[otuData[i]]++; }
+
+ calcCentroidsDriver(numOTUs, cumNumSeqs, nSeqsPerOTU, seqIndex, change, centroids, singleTau, mapSeqToUnique, uniqueFlowgrams, flowDataIntI, lengths, numFlowCells, seqNumber);
+
+ if (m->control_pressed) { break; }
+
+ if ((large) && (g == 0)) { flowFileName = filenames[i]; theseFlowFileNames[0] = filenames[i]; }
+ string thisOutputDir = outputDir;
+ if (outputDir == "") { thisOutputDir = m->hasPath(flowFileName); }
+ string qualityFileName = thisOutputDir + m->getRootName(m->getSimpleName(flowFileName)) + getOutputFileNameTag("qfile");
+ string fastaFileName = thisOutputDir + m->getRootName(m->getSimpleName(flowFileName)) + getOutputFileNameTag("fasta");
+ string nameFileName = thisOutputDir + m->getRootName(m->getSimpleName(flowFileName)) + getOutputFileNameTag("name");
+ string otuCountsFileName = thisOutputDir + m->getRootName(m->getSimpleName(flowFileName)) + getOutputFileNameTag("counts");
+ string fileRoot = m->getRootName(m->getSimpleName(flowFileName));
+ int pos = fileRoot.find_first_of('.');
+ string fileGroup = fileRoot;
+ if (pos != string::npos) { fileGroup = fileRoot.substr(pos+1, (fileRoot.length()-1-(pos+1))); }
+ string groupFileName = thisOutputDir + fileRoot + getOutputFileNameTag("group");
+
+
+ writeQualities(numOTUs, numFlowCells, qualityFileName, otuCounts, nSeqsPerOTU, seqNumber, singleTau, flowDataIntI, uniqueFlowgrams, cumNumSeqs, mapUniqueToSeq, seqNameVector, centroids, aaI); if (m->control_pressed) { break; }
+ writeSequences(thisCompositeFASTAFileName, numOTUs, numFlowCells, fastaFileName, otuCounts, uniqueFlowgrams, seqNameVector, aaI, centroids);if (m->control_pressed) { break; }
+ writeNames(thisCompositeNamesFileName, numOTUs, nameFileName, otuCounts, seqNameVector, aaI, nSeqsPerOTU); if (m->control_pressed) { break; }
+ writeClusters(otuCountsFileName, numOTUs, numFlowCells,otuCounts, centroids, uniqueFlowgrams, seqNameVector, aaI, nSeqsPerOTU, lengths, flowDataIntI); if (m->control_pressed) { break; }
+ writeGroups(groupFileName, fileGroup, numSeqs, seqNameVector); if (m->control_pressed) { break; }
+
+ if (large) {
+ if (g > 0) {
+ m->appendFiles(qualityFileName, (thisOutputDir + m->getRootName(m->getSimpleName(theseFlowFileNames[0])) + getOutputFileNameTag("qfile")));
+ m->mothurRemove(qualityFileName);
+ m->appendFiles(fastaFileName, (thisOutputDir + m->getRootName(m->getSimpleName(theseFlowFileNames[0])) + getOutputFileNameTag("fasta")));
+ m->mothurRemove(fastaFileName);
+ m->appendFiles(nameFileName, (thisOutputDir + m->getRootName(m->getSimpleName(theseFlowFileNames[0])) + getOutputFileNameTag("name")));
+ m->mothurRemove(nameFileName);
+ m->appendFiles(otuCountsFileName, (thisOutputDir + m->getRootName(m->getSimpleName(theseFlowFileNames[0])) + getOutputFileNameTag("counts")));
+ m->mothurRemove(otuCountsFileName);
+ m->appendFiles(groupFileName, (thisOutputDir + m->getRootName(m->getSimpleName(theseFlowFileNames[0])) + getOutputFileNameTag("group")));
+ m->mothurRemove(groupFileName);
+ }
+ m->mothurRemove(theseFlowFileNames[g]);
+ }
+ }
- writeQualities(numOTUs, numFlowCells, flowFileName, otuCounts, nSeqsPerOTU, seqNumber, singleTau, flowDataIntI, uniqueFlowgrams, cumNumSeqs, mapUniqueToSeq, seqNameVector, centroids, aaI); if (m->control_pressed) { break; }
- writeSequences(thisCompositeFASTAFileName, numOTUs, numFlowCells, flowFileName, otuCounts, uniqueFlowgrams, seqNameVector, aaI, centroids);if (m->control_pressed) { break; }
- writeNames(thisCompositeNamesFileName, numOTUs, flowFileName, otuCounts, seqNameVector, aaI, nSeqsPerOTU); if (m->control_pressed) { break; }
- writeClusters(flowFileName, numOTUs, numFlowCells,otuCounts, centroids, uniqueFlowgrams, seqNameVector, aaI, nSeqsPerOTU, lengths, flowDataIntI); if (m->control_pressed) { break; }
- writeGroups(flowFileName, numSeqs, seqNameVector); if (m->control_pressed) { break; }
-
+ numCompleted++;
m->mothurOut("Total time to process " + flowFileName + ":\t" + toString(time(NULL) - begTime) + '\t' + toString((clock() - begClock)/(double)CLOCKS_PER_SEC) + '\n');
}
if (m->control_pressed) { for (int i = 0; i < outputNames.size(); i++) { m->mothurRemove(outputNames[i]); } return 0; }
- return 0;
+ return numCompleted;
}catch(exception& e) {
m->errorOut(e, "ShhherCommand", "driver");
thisNameMap.clear();
flowFile >> numFlowCells;
+ if (m->debug) { m->mothurOut("[DEBUG]: numFlowCells = " + toString(numFlowCells) + ".\n"); }
int index = 0;//pcluster
while(!flowFile.eof()){
if (m->control_pressed) { break; }
flowFile >> seqName >> currentNumFlowCells;
+
thisLengths.push_back(currentNumFlowCells);
thisSeqNameVector.push_back(seqName);
thisNameMap[seqName] = index++;//pcluster
-
+
+ if (m->debug) { m->mothurOut("[DEBUG]: seqName = " + seqName + " length = " + toString(currentNumFlowCells) + " index = " + toString(index) + "\n"); }
+
for(int i=0;i<numFlowCells;i++){
flowFile >> intensity;
if(intensity > 9.99) { intensity = 9.99; }
}
uniqueFlowDataIntI.resize(numFlowCells * numUniques);
uniqueLengths.resize(numUniques);
- uniqueFlowgrams.resize(numFlowCells * numUniques);
-
+
flowDataPrI.resize(numSeqs * numFlowCells, 0);
for(int i=0;i<flowDataPrI.size();i++) { if (m->control_pressed) { break; } flowDataPrI[i] = getProbIntensity(flowDataIntI[i]); }
listFile >> label >> numOTUs;
+ if (m->debug) { m->mothurOut("[DEBUG]: Getting OTU Data...\n"); }
+
otuData.assign(numSeqs, 0);
cumNumSeqs.assign(numOTUs, 0);
nSeqsPerOTU.assign(numOTUs, 0);
for(int i=0;i<numOTUs;i++){
if (m->control_pressed) { break; }
+ if (m->debug) { m->mothurOut("[DEBUG]: processing OTU " + toString(i) + ".\n"); }
listFile >> singleOTU;
}
/**************************************************************************************************/
-void ShhherCommand::writeQualities(int numOTUs, int numFlowCells, string filename, vector<int> otuCounts, vector<int>& nSeqsPerOTU, vector<int>& seqNumber,
+void ShhherCommand::writeQualities(int numOTUs, int numFlowCells, string qualityFileName, vector<int> otuCounts, vector<int>& nSeqsPerOTU, vector<int>& seqNumber,
vector<double>& singleTau, vector<short>& flowDataIntI, vector<short>& uniqueFlowgrams, vector<int>& cumNumSeqs,
vector<int>& mapUniqueToSeq, vector<string>& seqNameVector, vector<int>& centroids, vector<vector<int> >& aaI){
try {
- string thisOutputDir = outputDir;
- if (outputDir == "") { thisOutputDir += m->hasPath(filename); }
- string qualityFileName = thisOutputDir + m->getRootName(m->getSimpleName(filename)) + "shhh.qual";
ofstream qualityFile;
m->openOutputFile(qualityFileName, qualityFile);
/**************************************************************************************************/
-void ShhherCommand::writeSequences(string thisCompositeFASTAFileName, int numOTUs, int numFlowCells, string filename, vector<int> otuCounts, vector<short>& uniqueFlowgrams, vector<string>& seqNameVector, vector<vector<int> >& aaI, vector<int>& centroids){
+void ShhherCommand::writeSequences(string thisCompositeFASTAFileName, int numOTUs, int numFlowCells, string fastaFileName, vector<int> otuCounts, vector<short>& uniqueFlowgrams, vector<string>& seqNameVector, vector<vector<int> >& aaI, vector<int>& centroids){
try {
- string thisOutputDir = outputDir;
- if (outputDir == "") { thisOutputDir += m->hasPath(filename); }
- string fastaFileName = thisOutputDir + m->getRootName(m->getSimpleName(filename)) + "shhh.fasta";
+
ofstream fastaFile;
m->openOutputFile(fastaFileName, fastaFile);
}
}
- fastaFile << newSeq.substr(4) << endl;
+ if (newSeq.length() >= 4) { fastaFile << newSeq.substr(4) << endl; }
+ else { fastaFile << "NNNN" << endl; }
}
}
fastaFile.close();
/**************************************************************************************************/
-void ShhherCommand::writeNames(string thisCompositeNamesFileName, int numOTUs, string filename, vector<int> otuCounts, vector<string>& seqNameVector, vector<vector<int> >& aaI, vector<int>& nSeqsPerOTU){
+void ShhherCommand::writeNames(string thisCompositeNamesFileName, int numOTUs, string nameFileName, vector<int> otuCounts, vector<string>& seqNameVector, vector<vector<int> >& aaI, vector<int>& nSeqsPerOTU){
try {
- string thisOutputDir = outputDir;
- if (outputDir == "") { thisOutputDir += m->hasPath(filename); }
- string nameFileName = thisOutputDir + m->getRootName(m->getSimpleName(filename)) + "shhh.names";
+
ofstream nameFile;
m->openOutputFile(nameFileName, nameFile);
/**************************************************************************************************/
-void ShhherCommand::writeGroups(string filename, int numSeqs, vector<string>& seqNameVector){
+void ShhherCommand::writeGroups(string groupFileName, string fileRoot, int numSeqs, vector<string>& seqNameVector){
try {
- string thisOutputDir = outputDir;
- if (outputDir == "") { thisOutputDir += m->hasPath(filename); }
- string fileRoot = thisOutputDir + m->getRootName(m->getSimpleName(filename));
- string groupFileName = fileRoot + "shhh.groups";
- ofstream groupFile;
+ ofstream groupFile;
m->openOutputFile(groupFileName, groupFile);
for(int i=0;i<numSeqs;i++){
/**************************************************************************************************/
-void ShhherCommand::writeClusters(string filename, int numOTUs, int numFlowCells, vector<int> otuCounts, vector<int>& centroids, vector<short>& uniqueFlowgrams, vector<string>& seqNameVector, vector<vector<int> >& aaI, vector<int>& nSeqsPerOTU, vector<int>& lengths, vector<short>& flowDataIntI){
+void ShhherCommand::writeClusters(string otuCountsFileName, int numOTUs, int numFlowCells, vector<int> otuCounts, vector<int>& centroids, vector<short>& uniqueFlowgrams, vector<string>& seqNameVector, vector<vector<int> >& aaI, vector<int>& nSeqsPerOTU, vector<int>& lengths, vector<short>& flowDataIntI){
try {
- string thisOutputDir = outputDir;
- if (outputDir == "") { thisOutputDir += m->hasPath(filename); }
- string otuCountsFileName = thisOutputDir + m->getRootName(m->getSimpleName(filename)) + "shhh.counts";
ofstream otuCountsFile;
m->openOutputFile(otuCountsFileName, otuCountsFile);
//otuCountsFile << base;
}
}
- otuCountsFile << newSeq.substr(4) << endl;
+
+ if (newSeq.length() >= 4) { otuCountsFile << newSeq.substr(4) << endl; }
+ else { otuCountsFile << "NNNN" << endl; }
}
otuCountsFile << endl;
}