X-Git-Url: https://git.donarmstrong.com/?a=blobdiff_plain;f=chimeraperseuscommand.cpp;h=353b124f8d556760b57a8086573caefe65e575f8;hb=60928795782d8f8648ec373d6a96b53032a77429;hp=fd94fcab7e289590da462eef7f03fc7069ab551f;hpb=c48d91112209b841444923670dca5454da0e2a4d;p=mothur.git diff --git a/chimeraperseuscommand.cpp b/chimeraperseuscommand.cpp index fd94fca..353b124 100644 --- a/chimeraperseuscommand.cpp +++ b/chimeraperseuscommand.cpp @@ -516,7 +516,7 @@ int ChimeraPerseusCommand::execute(){ if (hasCount) { CountTable* ct = new CountTable(); - ct->readTable(nameFile); + ct->readTable(nameFile, true, false); if (ct->hasGroupInfo()) { cparser = new SequenceCountParser(fastaFileNames[s], *ct); @@ -534,7 +534,7 @@ int ChimeraPerseusCommand::execute(){ if(processors == 1) { numSeqs = driverGroups(outputFileName, accnosFileName, newCountFile, 0, groups.size(), groups); if (dups) { - CountTable c; c.readTable(nameFile); + CountTable c; c.readTable(nameFile, true, false); if (!m->isBlank(newCountFile)) { ifstream in2; m->openInputFile(newCountFile, in2); @@ -559,7 +559,7 @@ int ChimeraPerseusCommand::execute(){ numChimeras = deconvoluteResults(uniqueNames, outputFileName, accnosFileName); }else { set doNotRemove; - CountTable c; c.readTable(newCountFile); + CountTable c; c.readTable(newCountFile, true, true); vector namesInTable = c.getNamesOfSeqs(); for (int i = 0; i < namesInTable.size(); i++) { int temp = c.getNumSeqs(namesInTable[i]); @@ -1070,20 +1070,23 @@ int ChimeraPerseusCommand::createProcessesGroups(string outputFName, string accn int num = 0; CountTable newCount; - if (hasCount && dups) { newCount.readTable(name); } + if (hasCount && dups) { newCount.readTable(name, true, false); } //sanity check if (groups.size() < processors) { processors = groups.size(); } //divide the groups between the processors vector lines; - int numGroupsPerProcessor = groups.size() / processors; - for (int i = 0; i < processors; i++) { - int startIndex = i * numGroupsPerProcessor; - int endIndex = (i+1) * numGroupsPerProcessor; - if(i == (processors - 1)){ endIndex = groups.size(); } - lines.push_back(linePair(startIndex, endIndex)); - } + int remainingPairs = groups.size(); + int startIndex = 0; + for (int remainingProcessors = processors; remainingProcessors > 0; remainingProcessors--) { + int numPairs = remainingPairs; //case for last processor + if (remainingProcessors != 1) { numPairs = ceil(remainingPairs / remainingProcessors); } + lines.push_back(linePair(startIndex, (startIndex+numPairs))); //startIndex, endIndex + startIndex = startIndex + numPairs; + remainingPairs = remainingPairs - numPairs; + } + #if defined (__APPLE__) || (__MACH__) || (linux) || (__linux) || (__linux__) || (__unix__) || (__unix) @@ -1095,11 +1098,11 @@ int ChimeraPerseusCommand::createProcessesGroups(string outputFName, string accn processIDS.push_back(pid); //create map from line number to pid so you can append files in correct order later process++; }else if (pid == 0){ - num = driverGroups(outputFName + toString(getpid()) + ".temp", accnos + toString(getpid()) + ".temp", accnos + ".byCount." + toString(getpid()) + ".temp", lines[process].start, lines[process].end, groups); + num = driverGroups(outputFName + toString(m->mothurGetpid(process)) + ".temp", accnos + toString(m->mothurGetpid(process)) + ".temp", accnos + ".byCount." + toString(m->mothurGetpid(process)) + ".temp", lines[process].start, lines[process].end, groups); //pass numSeqs to parent ofstream out; - string tempFile = outputFName + toString(getpid()) + ".num.temp"; + string tempFile = outputFName + toString(m->mothurGetpid(process)) + ".num.temp"; m->openOutputFile(tempFile, out); out << num << endl; out.close();