X-Git-Url: https://git.donarmstrong.com/?a=blobdiff_plain;f=screenseqscommand.cpp;h=8a7a108eb322978377a9d09ffb880a2745c54caf;hb=191ae1be0679d5cf4eda950b3b1bf26fb7dd503d;hp=ab750321b8468670d626f5d5b724aba61a98c071;hpb=89d6711c2beed6ee75fb00e5e57f1a91564d3e89;p=mothur.git diff --git a/screenseqscommand.cpp b/screenseqscommand.cpp index ab75032..8a7a108 100644 --- a/screenseqscommand.cpp +++ b/screenseqscommand.cpp @@ -10,6 +10,59 @@ #include "screenseqscommand.h" #include "sequence.hpp" +//********************************************************************************************************************** +vector ScreenSeqsCommand::getValidParameters(){ + try { + string Array[] = {"fasta", "start", "end", "maxambig", "maxhomop","optimize","criteria", "minlength", "maxlength", + "name", "group", "alignreport","processors","outputdir","inputdir"}; + vector myArray (Array, Array+(sizeof(Array)/sizeof(string))); + return myArray; + } + catch(exception& e) { + m->errorOut(e, "ScreenSeqsCommand", "getValidParameters"); + exit(1); + } +} +//********************************************************************************************************************** +ScreenSeqsCommand::ScreenSeqsCommand(){ + try { + abort = true; + //initialize outputTypes + vector tempOutNames; + outputTypes["fasta"] = tempOutNames; + outputTypes["name"] = tempOutNames; + outputTypes["group"] = tempOutNames; + outputTypes["alignreport"] = tempOutNames; + outputTypes["accnos"] = tempOutNames; + } + catch(exception& e) { + m->errorOut(e, "ScreenSeqsCommand", "ScreenSeqsCommand"); + exit(1); + } +} +//********************************************************************************************************************** +vector ScreenSeqsCommand::getRequiredParameters(){ + try { + string Array[] = {"fasta"}; + vector myArray (Array, Array+(sizeof(Array)/sizeof(string))); + return myArray; + } + catch(exception& e) { + m->errorOut(e, "ScreenSeqsCommand", "getRequiredParameters"); + exit(1); + } +} +//********************************************************************************************************************** +vector ScreenSeqsCommand::getRequiredFiles(){ + try { + vector myArray; + return myArray; + } + catch(exception& e) { + m->errorOut(e, "ScreenSeqsCommand", "getRequiredFiles"); + exit(1); + } +} //*************************************************************************************************************** ScreenSeqsCommand::ScreenSeqsCommand(string option) { @@ -21,14 +74,14 @@ ScreenSeqsCommand::ScreenSeqsCommand(string option) { else { //valid paramters for this command - string AlignArray[] = {"fasta", "start", "end", "maxambig", "maxhomop", "minlength", "maxlength", + string AlignArray[] = {"fasta", "start", "end", "maxambig", "maxhomop","optimize","criteria", "minlength", "maxlength", "name", "group", "alignreport","processors","outputdir","inputdir"}; vector myArray (AlignArray, AlignArray+(sizeof(AlignArray)/sizeof(string))); OptionParser parser(option); map parameters = parser.getParameters(); - ValidParameters validParameter; + ValidParameters validParameter("screen.seqs"); map::iterator it; //check to make sure all parameters are valid for command @@ -36,6 +89,14 @@ ScreenSeqsCommand::ScreenSeqsCommand(string option) { if (validParameter.isValidParameter(it->first, myArray, it->second) != true) { abort = true; } } + //initialize outputTypes + vector tempOutNames; + outputTypes["fasta"] = tempOutNames; + outputTypes["name"] = tempOutNames; + outputTypes["group"] = tempOutNames; + outputTypes["alignreport"] = tempOutNames; + outputTypes["accnos"] = 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 (inputDir == "not found"){ inputDir = ""; } @@ -44,7 +105,7 @@ ScreenSeqsCommand::ScreenSeqsCommand(string option) { it = parameters.find("fasta"); //user has given a template file if(it != parameters.end()){ - path = hasPath(it->second); + path = m->hasPath(it->second); //if the user has not given a path then, add inputdir. else leave path alone. if (path == "") { parameters["fasta"] = inputDir + it->second; } } @@ -52,7 +113,7 @@ ScreenSeqsCommand::ScreenSeqsCommand(string option) { it = parameters.find("group"); //user has given a template file if(it != parameters.end()){ - path = hasPath(it->second); + path = m->hasPath(it->second); //if the user has not given a path then, add inputdir. else leave path alone. if (path == "") { parameters["group"] = inputDir + it->second; } } @@ -60,7 +121,7 @@ ScreenSeqsCommand::ScreenSeqsCommand(string option) { it = parameters.find("name"); //user has given a template file if(it != parameters.end()){ - path = hasPath(it->second); + path = m->hasPath(it->second); //if the user has not given a path then, add inputdir. else leave path alone. if (path == "") { parameters["name"] = inputDir + it->second; } } @@ -68,7 +129,7 @@ ScreenSeqsCommand::ScreenSeqsCommand(string option) { it = parameters.find("alignreport"); //user has given a template file if(it != parameters.end()){ - path = hasPath(it->second); + path = m->hasPath(it->second); //if the user has not given a path then, add inputdir. else leave path alone. if (path == "") { parameters["alignreport"] = inputDir + it->second; } } @@ -94,7 +155,7 @@ ScreenSeqsCommand::ScreenSeqsCommand(string option) { //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 += hasPath(fastafile); //if user entered a file with a path then preserve it + outputDir += m->hasPath(fastafile); //if user entered a file with a path then preserve it } //check for optional parameter and set defaults @@ -120,7 +181,25 @@ ScreenSeqsCommand::ScreenSeqsCommand(string option) { temp = validParameter.validFile(parameters, "processors", false); if (temp == "not found"){ temp = "1"; } convert(temp, processors); - + + temp = validParameter.validFile(parameters, "optimize", false); //optimizing trumps the optimized values original value + if (temp == "not found"){ temp = ""; } + else { m->splitAtDash(temp, optimize); } + + //check for invalid optimize options + set validOptimizers; + validOptimizers.insert("start"); validOptimizers.insert("end"); validOptimizers.insert("maxambig"); validOptimizers.insert("maxhomop"); validOptimizers.insert("minlength"); validOptimizers.insert("maxlength"); + + for (int i = 0; i < optimize.size(); i++) { + if (validOptimizers.count(optimize[i]) == 0) { + m->mothurOut(optimize[i] + " is not a valid optimizer. Valid options are start, end, maxambig, maxhomop, minlength and maxlength."); m->mothurOutEndLine(); + optimize.erase(optimize.begin()+i); + i--; + } + } + + temp = validParameter.validFile(parameters, "criteria", false); if (temp == "not found"){ temp = "90"; } + convert(temp, criteria); } } @@ -134,7 +213,7 @@ ScreenSeqsCommand::ScreenSeqsCommand(string option) { void ScreenSeqsCommand::help(){ try { m->mothurOut("The screen.seqs command reads a fastafile and creates .....\n"); - m->mothurOut("The screen.seqs command parameters are fasta, start, end, maxambig, maxhomop, minlength, maxlength, name, and group.\n"); + m->mothurOut("The screen.seqs command parameters are fasta, start, end, maxambig, maxhomop, minlength, maxlength, name, group, optimize, criteria and processors.\n"); m->mothurOut("The fasta parameter is required.\n"); m->mothurOut("The start parameter .... The default is -1.\n"); m->mothurOut("The end parameter .... The default is -1.\n"); @@ -142,6 +221,9 @@ void ScreenSeqsCommand::help(){ m->mothurOut("The maxhomop parameter .... The default is -1.\n"); m->mothurOut("The minlength parameter .... The default is -1.\n"); m->mothurOut("The maxlength parameter .... The default is -1.\n"); + m->mothurOut("The processors parameter allows you to specify the number of processors to use while running the command. The default is 1.\n"); + m->mothurOut("The optimize and criteria parameters allow you set the start, end, maxabig, maxhomop, minlength and maxlength parameters relative to your set of sequences .\n"); + m->mothurOut("For example optimize=start-end, criteria=90, would set the start and end values to the position 90% of your sequences started and ended.\n"); m->mothurOut("The name parameter allows you to provide a namesfile, and the group parameter allows you to provide a groupfile.\n"); m->mothurOut("The screen.seqs command should be in the following format: \n"); m->mothurOut("screen.seqs(fasta=yourFastaFile, name=youNameFile, group=yourGroupFIle, start=yourStart, end=yourEnd, maxambig=yourMaxambig, \n"); @@ -166,10 +248,19 @@ int ScreenSeqsCommand::execute(){ try{ if (abort == true) { return 0; } + + //if the user want to optimize we need to no the 90% mark + vector positions; + if (optimize.size() != 0) { getSummary(positions); } //get summary is paralellized so we need to divideFile, no need to do this step twice so I moved it here + else { + positions = m->divideFile(fastafile, processors); + for (int i = 0; i < (positions.size()-1); i++) { + lines.push_back(new linePair(positions[i], positions[(i+1)])); + } + } - string goodSeqFile = outputDir + getRootName(getSimpleName(fastafile)) + "good" + getExtension(fastafile); - string badSeqFile = outputDir + getRootName(getSimpleName(fastafile)) + "bad" + getExtension(fastafile); - string badAccnosFile = outputDir + getRootName(getSimpleName(fastafile)) + "bad.accnos"; + string goodSeqFile = outputDir + m->getRootName(m->getSimpleName(fastafile)) + "good" + m->getExtension(fastafile); + string badAccnosFile = outputDir + m->getRootName(m->getSimpleName(fastafile)) + "bad.accnos"; int numFastaSeqs = 0; set badSeqNames; @@ -178,7 +269,7 @@ int ScreenSeqsCommand::execute(){ #ifdef USE_MPI int pid, end, numSeqsPerProcessor; int tag = 2001; - vector MPIPos; + vector MPIPos; MPI_Status status; MPI_Comm_rank(MPI_COMM_WORLD, &pid); //find out who we are @@ -186,7 +277,6 @@ int ScreenSeqsCommand::execute(){ MPI_File inMPI; MPI_File outMPIGood; - MPI_File outMPIBad; MPI_File outMPIBadAccnos; int outMode=MPI_MODE_CREATE|MPI_MODE_WRONLY; @@ -195,9 +285,6 @@ int ScreenSeqsCommand::execute(){ char outGoodFilename[1024]; strcpy(outGoodFilename, goodSeqFile.c_str()); - char outBadFilename[1024]; - strcpy(outBadFilename, badSeqFile.c_str()); - char outBadAccnosFilename[1024]; strcpy(outBadAccnosFilename, badAccnosFile.c_str()); @@ -206,28 +293,29 @@ int ScreenSeqsCommand::execute(){ MPI_File_open(MPI_COMM_WORLD, inFileName, inMode, MPI_INFO_NULL, &inMPI); //comm, filename, mode, info, filepointer MPI_File_open(MPI_COMM_WORLD, outGoodFilename, outMode, MPI_INFO_NULL, &outMPIGood); - MPI_File_open(MPI_COMM_WORLD, outBadFilename, outMode, MPI_INFO_NULL, &outMPIBad); MPI_File_open(MPI_COMM_WORLD, outBadAccnosFilename, outMode, MPI_INFO_NULL, &outMPIBadAccnos); - if (m->control_pressed) { MPI_File_close(&inMPI); MPI_File_close(&outMPIGood); MPI_File_close(&outMPIBad); MPI_File_close(&outMPIBadAccnos); return 0; } + if (m->control_pressed) { MPI_File_close(&inMPI); MPI_File_close(&outMPIGood); MPI_File_close(&outMPIBadAccnos); return 0; } if (pid == 0) { //you are the root process - MPIPos = setFilePosFasta(fastafile, numFastaSeqs); //fills MPIPos, returns numSeqs + MPIPos = m->setFilePosFasta(fastafile, numFastaSeqs); //fills MPIPos, returns numSeqs //send file positions to all processes - MPI_Bcast(&numFastaSeqs, 1, MPI_INT, 0, MPI_COMM_WORLD); //send numSeqs - MPI_Bcast(&MPIPos[0], (numFastaSeqs+1), MPI_LONG, 0, MPI_COMM_WORLD); //send file pos + for(int i = 1; i < processors; i++) { + MPI_Send(&numFastaSeqs, 1, MPI_INT, i, tag, MPI_COMM_WORLD); + MPI_Send(&MPIPos[0], (numFastaSeqs+1), MPI_LONG, i, tag, MPI_COMM_WORLD); + } //figure out how many sequences you have to align numSeqsPerProcessor = numFastaSeqs / processors; int startIndex = pid * numSeqsPerProcessor; if(pid == (processors - 1)){ numSeqsPerProcessor = numFastaSeqs - pid * numSeqsPerProcessor; } - + // cout << pid << '\t' << numSeqsPerProcessor << '\t' << startIndex << endl; //align your part - driverMPI(startIndex, numSeqsPerProcessor, inMPI, outMPIGood, outMPIBad, outMPIBadAccnos, MPIPos, badSeqNames); - - if (m->control_pressed) { MPI_File_close(&inMPI); MPI_File_close(&outMPIGood); MPI_File_close(&outMPIBadAccnos); MPI_File_close(&outMPIBad); return 0; } + driverMPI(startIndex, numSeqsPerProcessor, inMPI, outMPIGood, outMPIBadAccnos, MPIPos, badSeqNames); + //cout << pid << " done" << endl; + if (m->control_pressed) { MPI_File_close(&inMPI); MPI_File_close(&outMPIGood); MPI_File_close(&outMPIBadAccnos); return 0; } for (int i = 1; i < processors; i++) { @@ -248,19 +336,19 @@ int ScreenSeqsCommand::execute(){ }*/ } }else{ //you are a child process - MPI_Bcast(&numFastaSeqs, 1, MPI_INT, 0, MPI_COMM_WORLD); //get numSeqs + MPI_Recv(&numFastaSeqs, 1, MPI_INT, 0, tag, MPI_COMM_WORLD, &status); MPIPos.resize(numFastaSeqs+1); - MPI_Bcast(&MPIPos[0], (numFastaSeqs+1), MPI_LONG, 0, MPI_COMM_WORLD); //get file positions - + MPI_Recv(&MPIPos[0], (numFastaSeqs+1), MPI_LONG, 0, tag, MPI_COMM_WORLD, &status); + //figure out how many sequences you have to align numSeqsPerProcessor = numFastaSeqs / processors; int startIndex = pid * numSeqsPerProcessor; if(pid == (processors - 1)){ numSeqsPerProcessor = numFastaSeqs - pid * numSeqsPerProcessor; } - + //cout << pid << '\t' << numSeqsPerProcessor << '\t' << startIndex << endl; //align your part - driverMPI(startIndex, numSeqsPerProcessor, inMPI, outMPIGood, outMPIBad, outMPIBadAccnos, MPIPos, badSeqNames); - - if (m->control_pressed) { MPI_File_close(&inMPI); MPI_File_close(&outMPIGood); MPI_File_close(&outMPIBad); MPI_File_close(&outMPIBadAccnos); return 0; } + driverMPI(startIndex, numSeqsPerProcessor, inMPI, outMPIGood, outMPIBadAccnos, MPIPos, badSeqNames); +//cout << pid << " done" << endl; + if (m->control_pressed) { MPI_File_close(&inMPI); MPI_File_close(&outMPIGood); MPI_File_close(&outMPIBadAccnos); return 0; } //send bad list int badSize = badSeqNames.size(); @@ -282,97 +370,54 @@ int ScreenSeqsCommand::execute(){ //close files MPI_File_close(&inMPI); MPI_File_close(&outMPIGood); - MPI_File_close(&outMPIBad); MPI_File_close(&outMPIBadAccnos); + MPI_Barrier(MPI_COMM_WORLD); //make everyone wait - just in case #else - + #if defined (__APPLE__) || (__MACH__) || (linux) || (__linux) if(processors == 1){ - ifstream inFASTA; - openInputFile(fastafile, inFASTA); - numFastaSeqs=count(istreambuf_iterator(inFASTA),istreambuf_iterator(), '>'); - inFASTA.close(); + numFastaSeqs = driver(lines[0], goodSeqFile, badAccnosFile, fastafile, badSeqNames); - lines.push_back(new linePair(0, numFastaSeqs)); - - driver(lines[0], goodSeqFile, badSeqFile, badAccnosFile, fastafile, badSeqNames); - - if (m->control_pressed) { remove(goodSeqFile.c_str()); remove(badSeqFile.c_str()); return 0; } + if (m->control_pressed) { remove(goodSeqFile.c_str()); return 0; } }else{ - vector positions; processIDS.resize(0); - ifstream inFASTA; - openInputFile(fastafile, inFASTA); - - string input; - while(!inFASTA.eof()){ - input = getline(inFASTA); - if (input.length() != 0) { - if(input[0] == '>'){ long int pos = inFASTA.tellg(); positions.push_back(pos - input.length() - 1); } - } - } - inFASTA.close(); - - numFastaSeqs = positions.size(); - - int numSeqsPerProcessor = numFastaSeqs / processors; - - for (int i = 0; i < processors; i++) { - long int startPos = positions[ i * numSeqsPerProcessor ]; - if(i == processors - 1){ - numSeqsPerProcessor = numFastaSeqs - i * numSeqsPerProcessor; - } - lines.push_back(new linePair(startPos, numSeqsPerProcessor)); - } - - createProcesses(goodSeqFile, badSeqFile, badAccnosFile, fastafile, badSeqNames); + numFastaSeqs = createProcesses(goodSeqFile, badAccnosFile, fastafile, badSeqNames); rename((goodSeqFile + toString(processIDS[0]) + ".temp").c_str(), goodSeqFile.c_str()); - rename((badSeqFile + toString(processIDS[0]) + ".temp").c_str(), badSeqFile.c_str()); rename((badAccnosFile + toString(processIDS[0]) + ".temp").c_str(), badAccnosFile.c_str()); //append alignment and report files for(int i=1;iappendFiles((goodSeqFile + toString(processIDS[i]) + ".temp"), goodSeqFile); remove((goodSeqFile + toString(processIDS[i]) + ".temp").c_str()); - - appendFiles((badSeqFile + toString(processIDS[i]) + ".temp"), badSeqFile); - remove((badSeqFile + toString(processIDS[i]) + ".temp").c_str()); - - appendFiles((badAccnosFile + toString(processIDS[i]) + ".temp"), badAccnosFile); + + m->appendFiles((badAccnosFile + toString(processIDS[i]) + ".temp"), badAccnosFile); remove((badAccnosFile + toString(processIDS[i]) + ".temp").c_str()); } - if (m->control_pressed) { remove(goodSeqFile.c_str()); remove(badSeqFile.c_str()); return 0; } + if (m->control_pressed) { remove(goodSeqFile.c_str()); return 0; } //read badSeqs in because root process doesnt know what other "bad" seqs the children found ifstream inBad; - int ableToOpen = openInputFile(badAccnosFile, inBad, "no error"); + int ableToOpen = m->openInputFile(badAccnosFile, inBad, "no error"); if (ableToOpen == 0) { badSeqNames.clear(); string tempName; while (!inBad.eof()) { - inBad >> tempName; gobble(inBad); + inBad >> tempName; m->gobble(inBad); badSeqNames.insert(tempName); } inBad.close(); } } #else - ifstream inFASTA; - openInputFile(fastafile, inFASTA); - numFastaSeqs=count(istreambuf_iterator(inFASTA),istreambuf_iterator(), '>'); - inFASTA.close(); - - lines.push_back(new linePair(0, numFastaSeqs)); - - driver(lines[0], goodSeqFile, badSeqFile, badAccnosFile, fastafile, badSeqNames); + numFastaSeqs = driver(lines[0], goodSeqFile, badAccnosFile, fastafile, badSeqNames); - if (m->control_pressed) { remove(goodSeqFile.c_str()); remove(badSeqFile.c_str()); return 0; } + if (m->control_pressed) { remove(goodSeqFile.c_str()); return 0; } #endif @@ -406,24 +451,24 @@ int ScreenSeqsCommand::execute(){ badSeqNames.clear(); string tempName; while (!iss.eof()) { - iss >> tempName; gobble(iss); + iss >> tempName; m->gobble(iss); badSeqNames.insert(tempName); } #endif if(namefile != "" && groupfile != "") { screenNameGroupFile(badSeqNames); - if (m->control_pressed) { remove(goodSeqFile.c_str()); remove(badSeqFile.c_str()); return 0; } + if (m->control_pressed) { remove(goodSeqFile.c_str()); return 0; } }else if(namefile != "") { screenNameGroupFile(badSeqNames); - if (m->control_pressed) { remove(goodSeqFile.c_str()); remove(badSeqFile.c_str()); return 0; } + if (m->control_pressed) { remove(goodSeqFile.c_str()); return 0; } }else if(groupfile != "") { screenGroupFile(badSeqNames); } // this screens just the group - if (m->control_pressed) { remove(goodSeqFile.c_str()); remove(badSeqFile.c_str()); return 0; } + if (m->control_pressed) { remove(goodSeqFile.c_str()); return 0; } if(alignreport != "") { screenAlignReport(badSeqNames); } - if (m->control_pressed) { remove(goodSeqFile.c_str()); remove(badSeqFile.c_str()); return 0; } + if (m->control_pressed) { remove(goodSeqFile.c_str()); return 0; } #ifdef USE_MPI } @@ -431,9 +476,8 @@ int ScreenSeqsCommand::execute(){ m->mothurOutEndLine(); m->mothurOut("Output File Names: "); m->mothurOutEndLine(); - m->mothurOut(goodSeqFile); m->mothurOutEndLine(); - m->mothurOut(badSeqFile); m->mothurOutEndLine(); - m->mothurOut(badAccnosFile); m->mothurOutEndLine(); + m->mothurOut(goodSeqFile); m->mothurOutEndLine(); outputTypes["fasta"].push_back(goodSeqFile); + m->mothurOut(badAccnosFile); m->mothurOutEndLine(); outputTypes["accnos"].push_back(badAccnosFile); for (int i = 0; i < outputNames.size(); i++) { m->mothurOut(outputNames[i]); m->mothurOutEndLine(); } m->mothurOutEndLine(); m->mothurOutEndLine(); @@ -454,28 +498,25 @@ int ScreenSeqsCommand::execute(){ int ScreenSeqsCommand::screenNameGroupFile(set badSeqNames){ try { ifstream inputNames; - openInputFile(namefile, inputNames); + m->openInputFile(namefile, inputNames); set badSeqGroups; string seqName, seqList, group; set::iterator it; - string goodNameFile = outputDir + getRootName(getSimpleName(namefile)) + "good" + getExtension(namefile); - string badNameFile = outputDir + getRootName(getSimpleName(namefile)) + "bad" + getExtension(namefile); - - outputNames.push_back(goodNameFile); outputNames.push_back(badNameFile); - - ofstream goodNameOut; openOutputFile(goodNameFile, goodNameOut); - ofstream badNameOut; openOutputFile(badNameFile, badNameOut); + string goodNameFile = outputDir + m->getRootName(m->getSimpleName(namefile)) + "good" + m->getExtension(namefile); + outputNames.push_back(goodNameFile); outputTypes["name"].push_back(goodNameFile); + ofstream goodNameOut; m->openOutputFile(goodNameFile, goodNameOut); + while(!inputNames.eof()){ - if (m->control_pressed) { goodNameOut.close(); badNameOut.close(); inputNames.close(); remove(goodNameFile.c_str()); remove(badNameFile.c_str()); return 0; } + if (m->control_pressed) { goodNameOut.close(); inputNames.close(); remove(goodNameFile.c_str()); return 0; } inputNames >> seqName >> seqList; it = badSeqNames.find(seqName); if(it != badSeqNames.end()){ badSeqNames.erase(it); - badNameOut << seqName << '\t' << seqList << endl; + if(namefile != ""){ int start = 0; for(int i=0;i badSeqNames){ else{ goodNameOut << seqName << '\t' << seqList << endl; } - gobble(inputNames); + m->gobble(inputNames); } inputNames.close(); goodNameOut.close(); - badNameOut.close(); - + //we were unable to remove some of the bad sequences if (badSeqNames.size() != 0) { for (it = badSeqNames.begin(); it != badSeqNames.end(); it++) { @@ -507,18 +547,15 @@ int ScreenSeqsCommand::screenNameGroupFile(set badSeqNames){ if(groupfile != ""){ ifstream inputGroups; - openInputFile(groupfile, inputGroups); + m->openInputFile(groupfile, inputGroups); - string goodGroupFile = outputDir + getRootName(getSimpleName(groupfile)) + "good" + getExtension(groupfile); - string badGroupFile = outputDir + getRootName(getSimpleName(groupfile)) + "bad" + getExtension(groupfile); + string goodGroupFile = outputDir + m->getRootName(m->getSimpleName(groupfile)) + "good" + m->getExtension(groupfile); + outputNames.push_back(goodGroupFile); outputTypes["group"].push_back(goodGroupFile); - outputNames.push_back(goodGroupFile); outputNames.push_back(badGroupFile); - - ofstream goodGroupOut; openOutputFile(goodGroupFile, goodGroupOut); - ofstream badGroupOut; openOutputFile(badGroupFile, badGroupOut); + ofstream goodGroupOut; m->openOutputFile(goodGroupFile, goodGroupOut); while(!inputGroups.eof()){ - if (m->control_pressed) { goodGroupOut.close(); badGroupOut.close(); inputGroups.close(); remove(goodNameFile.c_str()); remove(badNameFile.c_str()); remove(goodGroupFile.c_str()); remove(badGroupFile.c_str()); return 0; } + if (m->control_pressed) { goodGroupOut.close(); inputGroups.close(); remove(goodNameFile.c_str()); remove(goodGroupFile.c_str()); return 0; } inputGroups >> seqName >> group; @@ -526,21 +563,19 @@ int ScreenSeqsCommand::screenNameGroupFile(set badSeqNames){ if(it != badSeqGroups.end()){ badSeqGroups.erase(it); - badGroupOut << seqName << '\t' << group << endl; } else{ goodGroupOut << seqName << '\t' << group << endl; } - gobble(inputGroups); + m->gobble(inputGroups); } inputGroups.close(); goodGroupOut.close(); - badGroupOut.close(); //we were unable to remove some of the bad sequences if (badSeqGroups.size() != 0) { for (it = badSeqGroups.begin(); it != badSeqGroups.end(); it++) { - m->mothurOut("Your namefile does not include the sequence " + *it + " please correct."); + m->mothurOut("Your groupfile does not include the sequence " + *it + " please correct."); m->mothurOutEndLine(); } } @@ -554,41 +589,205 @@ int ScreenSeqsCommand::screenNameGroupFile(set badSeqNames){ exit(1); } } +//*************************************************************************************************************** +int ScreenSeqsCommand::getSummary(vector& positions){ + try { + + vector startPosition; + vector endPosition; + vector seqLength; + vector ambigBases; + vector longHomoPolymer; + + vector positions = m->divideFile(fastafile, processors); + + for (int i = 0; i < (positions.size()-1); i++) { + lines.push_back(new linePair(positions[i], positions[(i+1)])); + } + + int numSeqs = 0; + #if defined (__APPLE__) || (__MACH__) || (linux) || (__linux) + if(processors == 1){ + numSeqs = driverCreateSummary(startPosition, endPosition, seqLength, ambigBases, longHomoPolymer, fastafile, lines[0]); + }else{ + numSeqs = createProcessesCreateSummary(startPosition, endPosition, seqLength, ambigBases, longHomoPolymer, fastafile); + } + + if (m->control_pressed) { return 0; } + #else + numSeqs = driverCreateSummary(startPosition, endPosition, seqLength, ambigBases, longHomoPolymer, fastafile, lines[0]); + if (m->control_pressed) { return 0; } + #endif + + sort(startPosition.begin(), startPosition.end()); + sort(endPosition.begin(), endPosition.end()); + sort(seqLength.begin(), seqLength.end()); + sort(ambigBases.begin(), ambigBases.end()); + sort(longHomoPolymer.begin(), longHomoPolymer.end()); + + int criteriaPercentile = int(numSeqs * (criteria / (float) 100)); + + for (int i = 0; i < optimize.size(); i++) { + if (optimize[i] == "start") { startPos = startPosition[criteriaPercentile]; m->mothurOut("Optimizing start to " + toString(startPos) + "."); m->mothurOutEndLine(); } + else if (optimize[i] == "end") { int endcriteriaPercentile = int(numSeqs * ((100 - criteria) / (float) 100)); endPos = endPosition[endcriteriaPercentile]; m->mothurOut("Optimizing end to " + toString(endPos) + "."); m->mothurOutEndLine();} + else if (optimize[i] == "maxambig") { maxAmbig = ambigBases[criteriaPercentile]; m->mothurOut("Optimizing maxambig to " + toString(maxAmbig) + "."); m->mothurOutEndLine(); } + else if (optimize[i] == "maxhomop") { maxHomoP = longHomoPolymer[criteriaPercentile]; m->mothurOut("Optimizing maxhomop to " + toString(maxHomoP) + "."); m->mothurOutEndLine(); } + else if (optimize[i] == "minlength") { int mincriteriaPercentile = int(numSeqs * ((100 - criteria) / (float) 100)); minLength = seqLength[mincriteriaPercentile]; m->mothurOut("Optimizing minlength to " + toString(minLength) + "."); m->mothurOutEndLine(); } + else if (optimize[i] == "maxlength") { maxLength = seqLength[criteriaPercentile]; m->mothurOut("Optimizing maxlength to " + toString(maxLength) + "."); m->mothurOutEndLine(); } + } + + return 0; + } + catch(exception& e) { + m->errorOut(e, "ScreenSeqsCommand", "getSummary"); + exit(1); + } +} +/**************************************************************************************/ +int ScreenSeqsCommand::driverCreateSummary(vector& startPosition, vector& endPosition, vector& seqLength, vector& ambigBases, vector& longHomoPolymer, string filename, linePair* filePos) { + try { + + ifstream in; + m->openInputFile(filename, in); + + in.seekg(filePos->start); + + bool done = false; + int count = 0; + + while (!done) { + + if (m->control_pressed) { in.close(); return 1; } + + Sequence current(in); m->gobble(in); + + if (current.getName() != "") { + startPosition.push_back(current.getStartPos()); + endPosition.push_back(current.getEndPos()); + seqLength.push_back(current.getNumBases()); + ambigBases.push_back(current.getAmbigBases()); + longHomoPolymer.push_back(current.getLongHomoPolymer()); + count++; + } + + #if defined (__APPLE__) || (__MACH__) || (linux) || (__linux) + unsigned long int pos = in.tellg(); + if ((pos == -1) || (pos >= filePos->end)) { break; } + #else + if (in.eof()) { break; } + #endif + + } + + in.close(); + + return count; + } + catch(exception& e) { + m->errorOut(e, "ScreenSeqsCommand", "driverCreateSummary"); + exit(1); + } +} +/**************************************************************************************************/ +int ScreenSeqsCommand::createProcessesCreateSummary(vector& startPosition, vector& endPosition, vector& seqLength, vector& ambigBases, vector& longHomoPolymer, string filename) { + try { +#if defined (__APPLE__) || (__MACH__) || (linux) || (__linux) + int process = 1; + int num = 0; + processIDS.clear(); + + //loop through and create all the processes you want + while (process != processors) { + int pid = fork(); + + if (pid > 0) { + 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 = driverCreateSummary(startPosition, endPosition, seqLength, ambigBases, longHomoPolymer, fastafile, lines[process]); + + //pass numSeqs to parent + ofstream out; + string tempFile = fastafile + toString(getpid()) + ".num.temp"; + m->openOutputFile(tempFile, out); + + out << num << endl; + for (int k = 0; k < startPosition.size(); k++) { out << startPosition[k] << '\t'; } out << endl; + for (int k = 0; k < endPosition.size(); k++) { out << endPosition[k] << '\t'; } out << endl; + for (int k = 0; k < seqLength.size(); k++) { out << seqLength[k] << '\t'; } out << endl; + for (int k = 0; k < ambigBases.size(); k++) { out << ambigBases[k] << '\t'; } out << endl; + for (int k = 0; k < longHomoPolymer.size(); k++) { out << longHomoPolymer[k] << '\t'; } out << endl; + + out.close(); + + exit(0); + }else { m->mothurOut("unable to spawn the necessary processes."); m->mothurOutEndLine(); exit(0); } + } + + num = driverCreateSummary(startPosition, endPosition, seqLength, ambigBases, longHomoPolymer, fastafile, lines[0]); + + //force parent to wait until all the processes are done + for (int i=0;iopenInputFile(tempFilename, in); + + int temp, tempNum; + in >> tempNum; m->gobble(in); num += tempNum; + for (int k = 0; k < tempNum; k++) { in >> temp; startPosition.push_back(temp); } m->gobble(in); + for (int k = 0; k < tempNum; k++) { in >> temp; endPosition.push_back(temp); } m->gobble(in); + for (int k = 0; k < tempNum; k++) { in >> temp; seqLength.push_back(temp); } m->gobble(in); + for (int k = 0; k < tempNum; k++) { in >> temp; ambigBases.push_back(temp); } m->gobble(in); + for (int k = 0; k < tempNum; k++) { in >> temp; longHomoPolymer.push_back(temp); } m->gobble(in); + + in.close(); + remove(tempFilename.c_str()); + } + + return num; +#endif + } + catch(exception& e) { + m->errorOut(e, "ScreenSeqsCommand", "createProcessesCreateSummary"); + exit(1); + } +} //*************************************************************************************************************** int ScreenSeqsCommand::screenGroupFile(set badSeqNames){ try { ifstream inputGroups; - openInputFile(groupfile, inputGroups); + m->openInputFile(groupfile, inputGroups); string seqName, group; set::iterator it; - string goodGroupFile = outputDir + getRootName(getSimpleName(groupfile)) + "good" + getExtension(groupfile); - string badGroupFile = outputDir + getRootName(getSimpleName(groupfile)) + "bad" + getExtension(groupfile); - - outputNames.push_back(goodGroupFile); outputNames.push_back(badGroupFile); - - ofstream goodGroupOut; openOutputFile(goodGroupFile, goodGroupOut); - ofstream badGroupOut; openOutputFile(badGroupFile, badGroupOut); + string goodGroupFile = outputDir + m->getRootName(m->getSimpleName(groupfile)) + "good" + m->getExtension(groupfile); + outputNames.push_back(goodGroupFile); outputTypes["group"].push_back(goodGroupFile); + ofstream goodGroupOut; m->openOutputFile(goodGroupFile, goodGroupOut); while(!inputGroups.eof()){ - if (m->control_pressed) { goodGroupOut.close(); badGroupOut.close(); inputGroups.close(); remove(goodGroupFile.c_str()); remove(badGroupFile.c_str()); return 0; } + if (m->control_pressed) { goodGroupOut.close(); inputGroups.close(); remove(goodGroupFile.c_str()); return 0; } inputGroups >> seqName >> group; it = badSeqNames.find(seqName); if(it != badSeqNames.end()){ badSeqNames.erase(it); - badGroupOut << seqName << '\t' << group << endl; } else{ goodGroupOut << seqName << '\t' << group << endl; } - gobble(inputGroups); + m->gobble(inputGroups); } - if (m->control_pressed) { goodGroupOut.close(); badGroupOut.close(); inputGroups.close(); remove(goodGroupFile.c_str()); remove(badGroupFile.c_str()); return 0; } + if (m->control_pressed) { goodGroupOut.close(); inputGroups.close(); remove(goodGroupFile.c_str()); return 0; } //we were unable to remove some of the bad sequences if (badSeqNames.size() != 0) { @@ -600,10 +799,8 @@ int ScreenSeqsCommand::screenGroupFile(set badSeqNames){ inputGroups.close(); goodGroupOut.close(); - badGroupOut.close(); - if (m->control_pressed) { remove(goodGroupFile.c_str()); remove(badGroupFile.c_str()); } - + if (m->control_pressed) { remove(goodGroupFile.c_str()); } return 0; @@ -619,27 +816,22 @@ int ScreenSeqsCommand::screenGroupFile(set badSeqNames){ int ScreenSeqsCommand::screenAlignReport(set badSeqNames){ try { ifstream inputAlignReport; - openInputFile(alignreport, inputAlignReport); + m->openInputFile(alignreport, inputAlignReport); string seqName, group; set::iterator it; - string goodAlignReportFile = outputDir + getRootName(getSimpleName(alignreport)) + "good" + getExtension(alignreport); - string badAlignReportFile = outputDir + getRootName(getSimpleName(alignreport)) + "bad" + getExtension(alignreport); - - outputNames.push_back(goodAlignReportFile); outputNames.push_back(badAlignReportFile); - - ofstream goodAlignReportOut; openOutputFile(goodAlignReportFile, goodAlignReportOut); - ofstream badAlignReportOut; openOutputFile(badAlignReportFile, badAlignReportOut); + string goodAlignReportFile = outputDir + m->getRootName(m->getSimpleName(alignreport)) + "good" + m->getExtension(alignreport); + outputNames.push_back(goodAlignReportFile); outputTypes["alignreport"].push_back(goodAlignReportFile); + ofstream goodAlignReportOut; m->openOutputFile(goodAlignReportFile, goodAlignReportOut); while (!inputAlignReport.eof()) { // need to copy header char c = inputAlignReport.get(); goodAlignReportOut << c; - badAlignReportOut << c; if (c == 10 || c == 13){ break; } } while(!inputAlignReport.eof()){ - if (m->control_pressed) { goodAlignReportOut.close(); badAlignReportOut.close(); inputAlignReport.close(); remove(goodAlignReportFile.c_str()); remove(badAlignReportFile.c_str()); return 0; } + if (m->control_pressed) { goodAlignReportOut.close(); inputAlignReport.close(); remove(goodAlignReportFile.c_str()); return 0; } inputAlignReport >> seqName; it = badSeqNames.find(seqName); @@ -652,29 +844,27 @@ int ScreenSeqsCommand::screenAlignReport(set badSeqNames){ if(it != badSeqNames.end()){ badSeqNames.erase(it); - badAlignReportOut << seqName << '\t' << line; } else{ goodAlignReportOut << seqName << '\t' << line; } - gobble(inputAlignReport); + m->gobble(inputAlignReport); } - if (m->control_pressed) { goodAlignReportOut.close(); badAlignReportOut.close(); inputAlignReport.close(); remove(goodAlignReportFile.c_str()); remove(badAlignReportFile.c_str()); return 0; } + if (m->control_pressed) { goodAlignReportOut.close(); inputAlignReport.close(); remove(goodAlignReportFile.c_str()); return 0; } //we were unable to remove some of the bad sequences if (badSeqNames.size() != 0) { for (it = badSeqNames.begin(); it != badSeqNames.end(); it++) { - m->mothurOut("Your file does not include the sequence " + *it + " please correct."); + m->mothurOut("Your alignreport file does not include the sequence " + *it + " please correct."); m->mothurOutEndLine(); } } inputAlignReport.close(); goodAlignReportOut.close(); - badAlignReportOut.close(); - if (m->control_pressed) { remove(goodAlignReportFile.c_str()); remove(badAlignReportFile.c_str()); return 0; } + if (m->control_pressed) { remove(goodAlignReportFile.c_str()); return 0; } return 0; @@ -687,27 +877,27 @@ int ScreenSeqsCommand::screenAlignReport(set badSeqNames){ } //********************************************************************************************************************** -int ScreenSeqsCommand::driver(linePair* line, string goodFName, string badFName, string badAccnosFName, string filename, set& badSeqNames){ +int ScreenSeqsCommand::driver(linePair* filePos, string goodFName, string badAccnosFName, string filename, set& badSeqNames){ try { ofstream goodFile; - openOutputFile(goodFName, goodFile); - - ofstream badFile; - openOutputFile(badFName, badFile); + m->openOutputFile(goodFName, goodFile); ofstream badAccnosFile; - openOutputFile(badAccnosFName, badAccnosFile); + m->openOutputFile(badAccnosFName, badAccnosFile); ifstream inFASTA; - openInputFile(filename, inFASTA); + m->openInputFile(filename, inFASTA); + + inFASTA.seekg(filePos->start); - inFASTA.seekg(line->start); + bool done = false; + int count = 0; - for(int i=0;inumSeqs;i++){ - + while (!done) { + if (m->control_pressed) { return 0; } - Sequence currSeq(inFASTA); + Sequence currSeq(inFASTA); m->gobble(inFASTA); if (currSeq.getName() != "") { bool goodSeq = 1; // innocent until proven guilty if(goodSeq == 1 && startPos != -1 && startPos < currSeq.getStartPos()) { goodSeq = 0; } @@ -721,21 +911,31 @@ int ScreenSeqsCommand::driver(linePair* line, string goodFName, string badFName, currSeq.printSequence(goodFile); } else{ - currSeq.printSequence(badFile); badAccnosFile << currSeq.getName() << endl; badSeqNames.insert(currSeq.getName()); } + count++; } - gobble(inFASTA); + + #if defined (__APPLE__) || (__MACH__) || (linux) || (__linux) + unsigned long int pos = inFASTA.tellg(); + if ((pos == -1) || (pos >= filePos->end)) { break; } + #else + if (inFASTA.eof()) { break; } + #endif + + //report progress + if((count) % 100 == 0){ m->mothurOut("Processing sequence: " + toString(count)); m->mothurOutEndLine(); } } + //report progress + if((count) % 100 != 0){ m->mothurOut("Processing sequence: " + toString(count)); m->mothurOutEndLine(); } goodFile.close(); inFASTA.close(); - badFile.close(); badAccnosFile.close(); - return 1; + return count; } catch(exception& e) { m->errorOut(e, "ScreenSeqsCommand", "driver"); @@ -744,11 +944,10 @@ int ScreenSeqsCommand::driver(linePair* line, string goodFName, string badFName, } //********************************************************************************************************************** #ifdef USE_MPI -int ScreenSeqsCommand::driverMPI(int start, int num, MPI_File& inMPI, MPI_File& goodFile, MPI_File& badFile, MPI_File& badAccnosFile, vector& MPIPos, set& badSeqNames){ +int ScreenSeqsCommand::driverMPI(int start, int num, MPI_File& inMPI, MPI_File& goodFile, MPI_File& badAccnosFile, vector& MPIPos, set& badSeqNames){ try { string outputString = ""; MPI_Status statusGood; - MPI_Status statusBad; MPI_Status statusBadAccnos; MPI_Status status; int pid; @@ -794,15 +993,6 @@ int ScreenSeqsCommand::driverMPI(int start, int num, MPI_File& inMPI, MPI_File& delete buf2; } else{ - outputString = ">" + currSeq.getName() + "\n" + currSeq.getAligned() + "\n"; - - //print bad seq to fasta - length = outputString.length(); - char* buf2 = new char[length]; - memcpy(buf2, outputString.c_str(), length); - - MPI_File_write_shared(badFile, buf2, length, MPI_CHAR, &statusBad); - delete buf2; badSeqNames.insert(currSeq.getName()); @@ -829,11 +1019,11 @@ int ScreenSeqsCommand::driverMPI(int start, int num, MPI_File& inMPI, MPI_File& #endif /**************************************************************************************************/ -int ScreenSeqsCommand::createProcesses(string goodFileName, string badFileName, string badAccnos, string filename, set& badSeqNames) { +int ScreenSeqsCommand::createProcesses(string goodFileName, string badAccnos, string filename, set& badSeqNames) { try { #if defined (__APPLE__) || (__MACH__) || (linux) || (__linux) int process = 0; - int exitCommand = 1; + int num = 0; //loop through and create all the processes you want while (process != processors) { @@ -843,7 +1033,15 @@ int ScreenSeqsCommand::createProcesses(string goodFileName, string badFileName, 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){ - exitCommand = driver(lines[process], goodFileName + toString(getpid()) + ".temp", badFileName + toString(getpid()) + ".temp", badAccnos + toString(getpid()) + ".temp", filename, badSeqNames); + num = driver(lines[process], goodFileName + toString(getpid()) + ".temp", badAccnos + toString(getpid()) + ".temp", filename, badSeqNames); + + //pass numSeqs to parent + ofstream out; + string tempFile = filename + toString(getpid()) + ".num.temp"; + m->openOutputFile(tempFile, out); + out << num << endl; + out.close(); + exit(0); }else { m->mothurOut("unable to spawn the necessary processes."); m->mothurOutEndLine(); exit(0); } } @@ -854,7 +1052,15 @@ int ScreenSeqsCommand::createProcesses(string goodFileName, string badFileName, wait(&temp); } - return exitCommand; + for (int i = 0; i < processIDS.size(); i++) { + ifstream in; + string tempFile = filename + toString(processIDS[i]) + ".num.temp"; + m->openInputFile(tempFile, in); + if (!in.eof()) { int tempNum = 0; in >> tempNum; num += tempNum; } + in.close(); remove(tempFile.c_str()); + } + + return num; #endif } catch(exception& e) {