X-Git-Url: https://git.donarmstrong.com/?a=blobdiff_plain;f=distancecommand.cpp;h=67ad7beb77b69d72f11914579246eb35dfa707e4;hb=163b300cfd7d4ca4e70c454be20f07b1d8346650;hp=e2ca937ed971e87194fd3623c48aec343d2f3ca5;hpb=4f4070ac3fbb5da75f834b48062e9123364b006e;p=mothur.git diff --git a/distancecommand.cpp b/distancecommand.cpp index e2ca937..67ad7be 100644 --- a/distancecommand.cpp +++ b/distancecommand.cpp @@ -16,47 +16,84 @@ //********************************************************************************************************************** -DistanceCommand::DistanceCommand(){ +DistanceCommand::DistanceCommand(string option){ try { - globaldata = GlobalData::getInstance(); - validCalculator = new ValidCalculators(); - countends = globaldata->getCountEnds(); - convert(globaldata->getProcessors(), processors); - convert(globaldata->getCutOff(), cutoff); + abort = false; + Estimators.clear(); - //open file - string filename = globaldata->getFastaFile(); - openInputFile(filename, in); - + //allow user to run help + if(option == "help") { help(); abort = true; } - int i; - if (isTrue(countends) == true) { - for (i=0; iEstimators.size(); i++) { - if (validCalculator->isValidCalculator("distance", globaldata->Estimators[i]) == true) { - if (globaldata->Estimators[i] == "nogaps") { - distCalculator = new ignoreGaps(); - }else if (globaldata->Estimators[i] == "eachgap") { - distCalculator = new eachGapDist(); - }else if (globaldata->Estimators[i] == "onegap") { - distCalculator = new oneGapDist(); } - } + else { + //valid paramters for this command + string Array[] = {"fasta", "phylip", "calc", "countends", "cutoff", "processors"}; + vector myArray (Array, Array+(sizeof(Array)/sizeof(string))); + + OptionParser parser(option); + map parameters = parser.getParameters(); + + ValidParameters validParameter; + + //check to make sure all parameters are valid for command + for (map::iterator it2 = parameters.begin(); it2 != parameters.end(); it2++) { + if (validParameter.isValidParameter(it2->first, myArray, it2->second) != true) { abort = true; } + } + + //check for required parameters + fastafile = validParameter.validFile(parameters, "fasta", true); + if (fastafile == "not found") { cout << "fasta is a required parameter for the dist.seqs command." << endl; abort = true; } + else if (fastafile == "not open") { abort = true; } + else{ + ifstream inFASTA; + openInputFile(fastafile, inFASTA); + alignDB = SequenceDB(inFASTA); + inFASTA.close(); } - }else { - for (i=0; iEstimators.size(); i++) { - if (validCalculator->isValidCalculator("distance", globaldata->Estimators[i]) == true) { - if (globaldata->Estimators[i] == "nogaps") { - distCalculator = new ignoreGaps(); - }else if (globaldata->Estimators[i] == "eachgap") { - distCalculator = new eachGapIgnoreTermGapDist(); - }else if (globaldata->Estimators[i] == "onegap") { - distCalculator = new oneGapIgnoreTermGapDist(); + + //check for optional parameter and set defaults + // ...at some point should added some additional type checking... + calc = validParameter.validFile(parameters, "calc", false); + if (calc == "not found") { calc = "onegap"; } + else { + if (calc == "default") { calc = "onegap"; } + } + splitAtDash(calc, Estimators); + + string temp; + temp = validParameter.validFile(parameters, "countends", false); if(temp == "not found"){ temp = "T"; } + convert(temp, countends); + + temp = validParameter.validFile(parameters, "cutoff", false); if(temp == "not found"){ temp = "1.0"; } + convert(temp, cutoff); + + temp = validParameter.validFile(parameters, "processors", false); if(temp == "not found"){ temp = "1"; } + convert(temp, processors); + + phylip = validParameter.validFile(parameters, "phylip", false); if(phylip == "not found"){ phylip = "F"; } + + + ValidCalculators validCalculator; + + if (isTrue(countends) == true) { + for (int i=0; isetCalc(""); + } catch(exception& e) { cout << "Standard Error: " << e.what() << " has occurred in the DistanceCommand class Function DistanceCommand. Please contact Pat Schloss at pschloss@microbio.umass.edu." << "\n"; @@ -67,234 +104,100 @@ DistanceCommand::DistanceCommand(){ exit(1); } } + +//********************************************************************************************************************** + +DistanceCommand::~DistanceCommand(){ + + for(int i=0;i 0){ - driver(distCalculator, seqDB, 0, (numSeqs/sqrt(2)), distFile + "tempa", phylipFile + "tempPhylipA", cutoff); - appendFiles((distFile+"tempa"), distFile); - remove((distFile + "tempa").c_str()); - appendFiles((phylipFile + "tempPhylipA"), phylipFile); - remove((phylipFile + "tempPhylipA").c_str()); - } - else{ - driver(distCalculator, seqDB, (numSeqs/sqrt(2)), numSeqs, distFile + "tempb", phylipFile + "tempPhylipB", cutoff); - appendFiles((distFile+"tempb"), distFile); - remove((distFile + "tempb").c_str()); - appendFiles((phylipFile + "tempPhylipB"), phylipFile); - remove((phylipFile + "tempPhylipB").c_str()); - } - wait(NULL); - - } - else if(processors == 3){ - int pid1 = fork(); - if(pid1 > 0){ - int pid2 = fork(); - if(pid2 > 0){ - driver(distCalculator, seqDB, 0, sqrt(3) * numSeqs / 3, distFile + "tempa", phylipFile + "tempPhylipA", cutoff); - appendFiles(distFile+"tempa", distFile); - appendFiles((phylipFile + "tempPhylipA"), phylipFile); - remove((distFile + "tempa").c_str()); - remove((phylipFile + "tempPhylipA").c_str()); - } - else{ - driver(distCalculator, seqDB, sqrt(3) * numSeqs / 3, sqrt(6) * numSeqs / 3, distFile + "tempb", phylipFile + "tempPhylipB", cutoff); - appendFiles(distFile+"tempb", distFile); - appendFiles((phylipFile + "tempPhylipB"), phylipFile); - remove((distFile + "tempb").c_str()); - remove((phylipFile + "tempPhylipB").c_str()); - } - wait(NULL); - } - else{ - driver(distCalculator, seqDB, sqrt(6) * numSeqs / 3, numSeqs, distFile + "tempc", phylipFile + "tempPhylipC", cutoff); - appendFiles(distFile+"tempc", distFile); - appendFiles((phylipFile + "tempPhylipC"), phylipFile); - remove((distFile + "tempc").c_str()); - remove((phylipFile + "tempPhylipC").c_str()); - } - wait(NULL); - } - else if(processors == 4){ - int pid1 = fork(); - if(pid1 > 0){ - int pid2 = fork(); - if(pid2 > 0){ - driver(distCalculator, seqDB, 0, numSeqs / 2, distFile + "tempa", phylipFile + "tempPhylipA", cutoff); - appendFiles(distFile+"tempa", distFile); - appendFiles((phylipFile + "tempPhylipA"), phylipFile); - remove((distFile + "tempa").c_str()); - remove((phylipFile + "tempPhylipA").c_str()); - } - else{ - driver(distCalculator, seqDB, numSeqs / 2, (numSeqs/sqrt(2)), distFile + "tempb", phylipFile + "tempPhylipB", cutoff); - appendFiles(distFile+"tempb", distFile); - appendFiles((phylipFile + "tempPhylipB"), phylipFile); - remove((distFile + "tempb").c_str()); - remove((phylipFile + "tempPhylipB").c_str()); - } - wait(NULL); + for (int i = 0; i < processors; i++) { + lines.push_back(new linePair()); + lines[i]->start = int (sqrt(float(i)/float(processors)) * numSeqs); + lines[i]->end = int (sqrt(float(i+1)/float(processors)) * numSeqs); } - else{ - int pid3 = fork(); - if(pid3 > 0){ - driver(distCalculator, seqDB, (numSeqs/sqrt(2)), (sqrt(3) * numSeqs / 2), distFile + "tempc", phylipFile + "tempPhylipC", cutoff); - appendFiles(distFile+"tempc", distFile); - appendFiles((phylipFile + "tempPhylipC"), phylipFile); - remove((distFile + "tempc").c_str()); - remove((phylipFile + "tempPhylipC").c_str()); - - } - else{ - driver(distCalculator, seqDB, (sqrt(3) * numSeqs / 2), numSeqs, distFile + "tempd", phylipFile + "tempPhylipD", cutoff); - appendFiles(distFile+"tempd", distFile); - appendFiles((phylipFile + "tempPhylipD"), phylipFile); - remove((distFile + "tempd").c_str()); - remove((phylipFile + "tempPhylipD").c_str()); - } - wait(NULL); - } - wait(NULL); - } - wait(NULL); -#elif (linux) || (__linux) - if(processors == 1){ - driver(distCalculator, seqDB, 0, numSeqs, distFile, phylipFile + "tempPhylipA", cutoff); - appendFiles((phylipFile + "tempPhylipA"), phylipFile); - remove((phylipFile + "tempPhylipA").c_str()); - } - else if(processors == 2){ - - int pid = fork(); - if(pid > 0){ - driver(distCalculator, seqDB, 0, (numSeqs/sqrt(2)), distFile + "tempa", phylipFile + "tempPhylipA", cutoff); - appendFiles((distFile+"tempa"), distFile); - appendFiles((phylipFile + "tempPhylipA"), phylipFile); - remove((distFile + "tempa").c_str()); - remove((phylipFile + "tempPhylipA").c_str()); - } - else{ - driver(distCalculator, seqDB, (numSeqs/sqrt(2)), numSeqs, distFile + "tempb", phylipFile + "tempPhylipB", cutoff); - appendFiles((distFile+"tempb"), distFile); - appendFiles((phylipFile + "tempPhylipB"), phylipFile); - remove((distFile + "tempb").c_str()); - remove((phylipFile + "tempPhylipB").c_str()); - } - wait(); + createProcesses(outputFile); + + map::iterator it = processIDS.begin(); + rename((outputFile + toString(it->second) + ".temp").c_str(), outputFile.c_str()); + it++; - } - else if(processors == 3){ - int pid1 = fork(); - if(pid1 > 0){ - int pid2 = fork(); - if(pid2 > 0){ - driver(distCalculator, seqDB, 0, (numSeqs/sqrt(2)), distFile + "tempa", phylipFile + "tempPhylipA", cutoff); - appendFiles((distFile+"tempa"), distFile); - appendFiles((phylipFile + "tempPhylipA"), phylipFile); - remove((distFile + "tempa").c_str()); - remove((phylipFile + "tempPhylipA").c_str()); - - } - else{ - driver(distCalculator, seqDB, (numSeqs/sqrt(2)), numSeqs, distFile + "tempb", phylipFile + "tempPhylipB", cutoff); - appendFiles((distFile+"tempb"), distFile); - appendFiles((phylipFile + "tempPhylipB"), phylipFile); - remove((distFile + "tempb").c_str()); - remove((phylipFile + "tempPhylipB").c_str()); - } - wait(); - } - else{ - driver(distCalculator, seqDB, sqrt(6) * numSeqs / 3, numSeqs, distFile + "tempc", phylipFile + "tempPhylipC", cutoff); - appendFiles(distFile+"tempc", distFile); - appendFiles((phylipFile + "tempPhylipC"), phylipFile); - remove((distFile + "tempc").c_str()); - remove((phylipFile + "tempPhylipC").c_str()); - } - wait(); - } - else if(processors == 4){ - int pid1 = fork(); - if(pid1 > 0){ - int pid2 = fork(); - if(pid2 > 0){ - driver(distCalculator, seqDB, 0, (numSeqs/sqrt(2)), distFile + "tempa", phylipFile + "tempPhylipA", cutoff); - appendFiles((distFile+"tempa"), distFile); - appendFiles((phylipFile + "tempPhylipA"), phylipFile); - remove((distFile + "tempa").c_str()); - remove((phylipFile + "tempPhylipA").c_str()); - } - else{ - driver(distCalculator, seqDB, (numSeqs/sqrt(2)), numSeqs, distFile + "tempb", phylipFile + "tempPhylipB", cutoff); - appendFiles((distFile+"tempb"), distFile); - appendFiles((phylipFile + "tempPhylipB"), phylipFile); - remove((distFile + "tempb").c_str()); - remove((phylipFile + "tempPhylipB").c_str()); - } - wait(); - } - else{ - int pid3 = fork(); - if(pid3 > 0){ - driver(distCalculator, seqDB, sqrt(6) * numSeqs / 3, numSeqs, distFile + "tempc", phylipFile + "tempPhylipC", cutoff); - appendFiles(distFile+"tempc", distFile); - appendFiles((phylipFile + "tempPhylipC"), phylipFile); - remove((distFile + "tempc").c_str()); - remove((phylipFile + "tempPhylipC").c_str()); - } - else{ - driver(distCalculator, seqDB, (sqrt(3) * numSeqs / 2), numSeqs, distFile + "tempd", phylipFile + "tempPhylipD", cutoff); - appendFiles(distFile+"tempd", distFile); - appendFiles((phylipFile + "tempPhylipD"), phylipFile); - remove((distFile + "tempd").c_str()); - remove((phylipFile + "tempPhylipD").c_str()); - } - wait(); + //append and remove temp files + for (; it != processIDS.end(); it++) { + appendFiles((outputFile + toString(it->second) + ".temp"), outputFile); + remove((outputFile + toString(it->second) + ".temp").c_str()); } - wait(); } - wait(); - #else - driver(distCalculator, seqDB, 0, numSeqs, distFile, phylipFile + "tempPhylipA", cutoff); - appendFiles((phylipFile + "tempPhylipA"), phylipFile); - remove((phylipFile + "tempPhylipA").c_str()); + ifstream inFASTA; + driver(0, numSeqs, outputFile, cutoff); #endif delete distCalculator; @@ -311,36 +214,70 @@ int DistanceCommand::execute(){ exit(1); } } +/**************************************************************************************************/ +void DistanceCommand::createProcesses(string filename) { + try { +#if defined (__APPLE__) || (__MACH__) || (linux) || (__linux) + int process = 0; + processIDS.clear(); + + //loop through and create all the processes you want + while (process != processors) { + int pid = fork(); + + if (pid > 0) { + processIDS[lines[process]->end] = pid; //create map from line number to pid so you can append files in correct order later + process++; + }else if (pid == 0){ + driver(lines[process]->start, lines[process]->end, filename + toString(getpid()) + ".temp", cutoff); + exit(0); + }else { cout << "unable to spawn the necessary processes." << endl; exit(0); } + } + + //force parent to wait until all the processes are done + for (map::iterator it = processIDS.begin(); it != processIDS.end(); it++) { + int temp = it->second; + wait(&temp); + } +#endif + } + catch(exception& e) { + cout << "Standard Error: " << e.what() << " has occurred in the DistanceCommand class Function createProcesses. Please contact Pat Schloss at pschloss@microbio.umass.edu." << "\n"; + exit(1); + } + catch(...) { + cout << "An unknown error has occurred in the DistanceCommand class function createProcesses. Please contact Pat Schloss at pschloss@microbio.umass.edu." << "\n"; + exit(1); + } +} /**************************************************************************************************/ /////// need to fix to work with calcs and sequencedb -int DistanceCommand::driver(Dist* distCalculator, SequenceDB* align, int startLine, int endLine, string dFileName, string pFilename, float cutoff){ +int DistanceCommand::driver(int startLine, int endLine, string dFileName, float cutoff){ try { + int startTime = time(NULL); //column file - ofstream distFile(dFileName.c_str(), ios::trunc); - distFile.setf(ios::fixed, ios::showpoint); - distFile << setprecision(4); - - //column file - ofstream philFile(pFilename.c_str(), ios::trunc); - philFile.setf(ios::fixed, ios::showpoint); - philFile << setprecision(4); + ofstream outFile(dFileName.c_str(), ios::trunc); + outFile.setf(ios::fixed, ios::showpoint); + outFile << setprecision(4); + if(isTrue(phylip) && startLine == 0){ outFile << alignDB.getNumSeqs() << endl; } for(int i=startLine;icalcDist(*(align->get(i)), *(align->get(j))); + distCalculator->calcDist(alignDB.get(i), alignDB.get(j)); double dist = distCalculator->getDist(); if(dist <= cutoff){ - distFile << align->get(i)->getName() << ' ' << align->get(j)->getName() << ' ' << dist << endl; + if (!isTrue(phylip)) { outFile << alignDB.get(i).getName() << ' ' << alignDB.get(j).getName() << ' ' << dist << endl; } } - philFile << dist << '\t'; + if (isTrue(phylip)) { outFile << dist << '\t'; } + } - philFile << endl; + if (isTrue(phylip) == true) { outFile << endl; } if(i % 100 == 0){ cout << i << '\t' << time(NULL) - startTime << endl; @@ -349,8 +286,7 @@ int DistanceCommand::driver(Dist* distCalculator, SequenceDB* align, int startLi } cout << endLine-1 << '\t' << time(NULL) - startTime << endl; - //philFile.close(); - //distFile.close(); + outFile.close(); return 1; } @@ -370,21 +306,15 @@ void DistanceCommand::appendFiles(string temp, string filename) { try{ ofstream output; ifstream input; - + //open output file in append mode openOutputFileAppend(filename, output); - - //open temp file for reading openInputFile(temp, input); - string line; - //read input file and write to output file - while(input.eof() != true) { - getline(input, line); //getline removes the newline char - if (line != "") { - output << line << endl; // Appending back newline char - } - } + while(char c = input.get()){ + if(input.eof()) { break; } + else { output << c; } + } input.close(); output.close(); @@ -398,4 +328,4 @@ void DistanceCommand::appendFiles(string temp, string filename) { exit(1); } } -/**************************************************************************************************/ \ No newline at end of file +/**************************************************************************************************/