//**********************************************************************************************************************
-DistanceCommand::DistanceCommand(){
+DistanceCommand::DistanceCommand(string option){
try {
- globaldata = GlobalData::getInstance();
- validCalculator = new ValidCalculators();
- ends = globaldata->getEnds();
- convert(globaldata->getProcessors(), processors);
- convert(globaldata->getCutOff(), cutoff);
-
- int i;
- if (ends != "T") {
- for (i=0; i<globaldata->Estimators.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(); }
+ abort = false;
+ Estimators.clear();
+
+ //allow user to run help
+ if(option == "help") { help(); abort = true; }
+
+ else {
+ //valid paramters for this command
+ string Array[] = {"fasta", "output", "calc", "countends", "cutoff", "processors", "outputdir","inputdir"};
+ vector<string> myArray (Array, Array+(sizeof(Array)/sizeof(string)));
+
+ OptionParser parser(option);
+ map<string, string> parameters = parser.getParameters();
+
+ ValidParameters validParameter;
+ map<string, string>::iterator it2;
+
+ //check to make sure all parameters are valid for command
+ for (it2 = parameters.begin(); it2 != parameters.end(); it2++) {
+ if (validParameter.isValidParameter(it2->first, myArray, it2->second) != true) { abort = true; }
+ }
+
+ //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 = ""; }
+ else {
+ string path;
+ it2 = parameters.find("fasta");
+ //user has given a template file
+ if(it2 != parameters.end()){
+ path = hasPath(it2->second);
+ //if the user has not given a path then, add inputdir. else leave path alone.
+ if (path == "") { parameters["fasta"] = inputDir + it2->second; }
}
}
- }else {
- for (i=0; i<globaldata->Estimators.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 required parameters
+ fastafile = validParameter.validFile(parameters, "fasta", true);
+ if (fastafile == "not found") { mothurOut("fasta is a required parameter for the dist.seqs command."); mothurOutEndLine(); abort = true; }
+ else if (fastafile == "not open") { abort = true; }
+ else{
+ ifstream inFASTA;
+ openInputFile(fastafile, inFASTA);
+ alignDB = SequenceDB(inFASTA);
+ inFASTA.close();
+ }
+
+ //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
+ }
+
+ //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);
+
+ output = validParameter.validFile(parameters, "output", false); if(output == "not found"){ output = "column"; }
+
+ if ((output != "column") && (output != "lt") && (output != "square")) { mothurOut(output + " is not a valid output form. Options are column, lt and square. I will use column."); mothurOutEndLine(); output = "column"; }
+
+ ValidCalculators validCalculator;
+
+ if (isTrue(countends) == true) {
+ for (int i=0; i<Estimators.size(); i++) {
+ if (validCalculator.isValidCalculator("distance", Estimators[i]) == true) {
+ if (Estimators[i] == "nogaps") { distCalculator = new ignoreGaps(); }
+ else if (Estimators[i] == "eachgap") { distCalculator = new eachGapDist(); }
+ else if (Estimators[i] == "onegap") { distCalculator = new oneGapDist(); }
+ }
+ }
+ }else {
+ for (int i=0; i<Estimators.size(); i++) {
+ if (validCalculator.isValidCalculator("distance", Estimators[i]) == true) {
+ if (Estimators[i] == "nogaps") { distCalculator = new ignoreGaps(); }
+ else if (Estimators[i] == "eachgap"){ distCalculator = new eachGapIgnoreTermGapDist(); }
+ else if (Estimators[i] == "onegap") { distCalculator = new oneGapIgnoreTermGapDist(); }
}
}
}
+
}
- //reset calc for next command
- globaldata->setCalc("");
-
-
}
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";
+ errorOut(e, "DistanceCommand", "DistanceCommand");
exit(1);
}
- catch(...) {
- cout << "An unknown error has occurred in the DistanceCommand class function DistanceCommand. Please contact Pat Schloss at pschloss@microbio.umass.edu." << "\n";
+}
+
+//**********************************************************************************************************************
+
+DistanceCommand::~DistanceCommand(){
+
+ for(int i=0;i<lines.size();i++){
+ delete lines[i];
+ }
+
+}
+
+//**********************************************************************************************************************
+
+void DistanceCommand::help(){
+ try {
+ mothurOut("The dist.seqs command reads a file containing sequences and creates a distance file.\n");
+ mothurOut("The dist.seqs command parameters are fasta, calc, countends, output, cutoff and processors. \n");
+ mothurOut("The fasta parameter is required.\n");
+ mothurOut("The calc parameter allows you to specify the method of calculating the distances. Your options are: nogaps, onegap or eachgap. The default is onegap.\n");
+ mothurOut("The countends parameter allows you to specify whether to include terminal gaps in distance. Your options are: T or F. The default is T.\n");
+ mothurOut("The cutoff parameter allows you to specify maximum distance to keep. The default is 1.0.\n");
+ mothurOut("The output parameter allows you to specify format of your distance matrix. Options are column, lt, and square. The default is column.\n");
+ mothurOut("The processors parameter allows you to specify number of processors to use. The default is 1.\n");
+ mothurOut("The dist.seqs command should be in the following format: \n");
+ mothurOut("dist.seqs(fasta=yourFastaFile, calc=yourCalc, countends=yourEnds, cutoff= yourCutOff, processors=yourProcessors) \n");
+ mothurOut("Example dist.seqs(fasta=amazon.fasta, calc=eachgap, countends=F, cutoff= 2.0, processors=3).\n");
+ mothurOut("Note: No spaces between parameter labels (i.e. calc), '=' and parameters (i.e.yourCalc).\n\n");
+ }
+ catch(exception& e) {
+ errorOut(e, "DistanceCommand", "help");
exit(1);
- }
+ }
}
//**********************************************************************************************************************
int DistanceCommand::execute(){
try {
- //read file
- string filename = globaldata->inputFileName;
-
- if(globaldata->getFastaFile() != "") {
- readSeqs = new ReadFasta(filename); }
- else if(globaldata->getNexusFile() != "") {
- readSeqs = new ReadNexus(filename); }
- else if(globaldata->getClustalFile() != "") {
- readSeqs = new ReadClustal(filename); }
- else if(globaldata->getPhylipFile() != "") {
- readSeqs = new ReadPhylip(filename); }
-
- readSeqs->read();
- seqDB = readSeqs->getDB();
-
- int numSeqs = seqDB->getNumSeqs();
+ if (abort == true) { return 0; }
- string distFile = getRootName(globaldata->getFastaFile()) + "dist";
+ int numSeqs = alignDB.getNumSeqs();
+ cutoff += 0.005;
- remove(distFile.c_str());
+ string outputFile;
+ if (output == "lt") { //does the user want lower triangle phylip formatted file
+ outputFile = outputDir + getRootName(getSimpleName(fastafile)) + "phylip.dist";
+ remove(outputFile.c_str());
+
+ //output numSeqs to phylip formatted dist file
+ }else if (output == "column") { //user wants column format
+ outputFile = outputDir + getRootName(getSimpleName(fastafile)) + "dist";
+ remove(outputFile.c_str());
+ }else { //assume square
+ outputFile = outputDir + getRootName(getSimpleName(fastafile)) + "square.dist";
+ remove(outputFile.c_str());
+ }
+
+#if defined (__APPLE__) || (__MACH__) || (linux) || (__linux)
+ //if you don't need to fork anything
if(processors == 1){
- driver(distCalculator, seqDB, 0, numSeqs, distFile, cutoff);
- }
- else if(processors == 2){
-
- int pid = fork();
- if(pid > 0){
- driver(distCalculator, seqDB, 0, (numSeqs/sqrt(2)), distFile + "tempa", cutoff);
- appendFiles((distFile+"tempa"), distFile);
- remove((distFile + "tempa").c_str());
- }
- else{
- driver(distCalculator, seqDB, (numSeqs/sqrt(2)), numSeqs, distFile + "tempb", cutoff);
- appendFiles((distFile+"tempb"), distFile);
- remove((distFile + "tempb").c_str());
+ driver(0, numSeqs, outputFile, cutoff);
+ }else{ //you have multiple processors
+
+ 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);
}
- 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", cutoff);
- appendFiles(distFile+"tempa", distFile);
- remove((distFile + "tempa").c_str());
- }
- else{
- driver(distCalculator, seqDB, sqrt(3) * numSeqs / 3, sqrt(6) * numSeqs / 3, distFile + "tempb", cutoff);
- appendFiles(distFile+"tempb", distFile);
- remove((distFile + "tempb").c_str());
- }
- wait(NULL);
- }
- else{
- driver(distCalculator, seqDB, sqrt(6) * numSeqs / 3, numSeqs, distFile + "tempc", cutoff);
- appendFiles(distFile+"tempc", distFile);
- remove((distFile + "tempc").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", cutoff);
- //system(("cat " + distFile + "tempa" + " >> " + distFile).c_str());
- appendFiles(distFile+"tempa", distFile);
- //system(("rm " + distFile + "tempa").c_str());
- remove((distFile + "tempa").c_str());
- }
- else{
- driver(distCalculator, seqDB, numSeqs / 2, (numSeqs/sqrt(2)), distFile + "tempb", cutoff);
- //system(("cat " + distFile + "tempb" + " >> " + distFile).c_str());
- appendFiles(distFile+"tempb", distFile);
- //system(("rm " + distFile + "tempb").c_str());
- remove((distFile + "tempb").c_str());
- }
- wait(NULL);
- }
- else{
- int pid3 = fork();
- if(pid3 > 0){
- driver(distCalculator, seqDB, (numSeqs/sqrt(2)), (sqrt(3) * numSeqs / 2), distFile + "tempc", cutoff);
- //system(("cat " + distFile + "tempc" + " >> " + distFile).c_str());
- appendFiles(distFile+"tempc", distFile);
- //system(("rm " + distFile + "tempc").c_str());
- remove((distFile + "tempc").c_str());
- }
- else{
- driver(distCalculator, seqDB, (sqrt(3) * numSeqs / 2), numSeqs, distFile + "tempd", cutoff);
- //system(("cat " + distFile + "tempd" + " >> " + distFile).c_str());
- appendFiles(distFile+"tempd", distFile);
- //system(("rm " + distFile + "tempd").c_str());
- remove((distFile + "tempd").c_str());
- }
- wait(NULL);
+ createProcesses(outputFile);
+
+ map<int, int>::iterator it = processIDS.begin();
+ rename((outputFile + toString(it->second) + ".temp").c_str(), outputFile.c_str());
+ it++;
+
+ //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(NULL);
}
- wait(NULL);
-
+#else
+ ifstream inFASTA;
+ driver(0, numSeqs, outputFile, cutoff);
+#endif
+
+ if (output == "square") { convertMatrix(outputFile); }
+
delete distCalculator;
-
+
return 0;
-
+
}
catch(exception& e) {
- cout << "Standard Error: " << e.what() << " has occurred in the DistanceCommand class Function execute. Please contact Pat Schloss at pschloss@microbio.umass.edu." << "\n";
+ errorOut(e, "DistanceCommand", "execute");
exit(1);
}
- catch(...) {
- cout << "An unknown error has occurred in the DistanceCommand class function execute. Please contact Pat Schloss at pschloss@microbio.umass.edu." << "\n";
+}
+/**************************************************************************************************/
+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 { mothurOut("unable to spawn the necessary processes."); mothurOutEndLine(); exit(0); }
+ }
+
+ //force parent to wait until all the processes are done
+ for (map<int, int>::iterator it = processIDS.begin(); it != processIDS.end(); it++) {
+ int temp = it->second;
+ wait(&temp);
+ }
+#endif
+ }
+ catch(exception& e) {
+ errorOut(e, "DistanceCommand", "createProcesses");
exit(1);
- }
+ }
}
/**************************************************************************************************/
/////// need to fix to work with calcs and sequencedb
-int DistanceCommand::driver(Dist* distCalculator, SequenceDB* align, int startLine, int endLine, string dFileName, float cutoff){
+int DistanceCommand::driver(int startLine, int endLine, string dFileName, float cutoff){
try {
+
int startTime = time(NULL);
-
- ofstream distFile(dFileName.c_str(), ios::trunc);
- distFile.setf(ios::fixed, ios::showpoint);
- distFile << setprecision(4);
-
- for(int i=startLine;i<endLine;i++){
- for(int j=0;j<i;j++){
- distCalculator->calcDist(align->get(i), align->get(j));
+ //column file
+ ofstream outFile(dFileName.c_str(), ios::trunc);
+ outFile.setf(ios::fixed, ios::showpoint);
+ outFile << setprecision(4);
+
+ if((output == "lt") && startLine == 0){ outFile << alignDB.getNumSeqs() << endl; }
+
+ for(int i=startLine;i<endLine;i++){
+ if(output == "lt") {
+ string name = alignDB.get(i).getName();
+ if (name.length() < 10) { //pad with spaces to make compatible
+ while (name.length() < 10) { name += " "; }
+ }
+ outFile << name << '\t';
+ }
+ for(int j=0;j<i;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;
-//cout << align->get(i).getName() << ' ' << align->get(j).getName() << ' ' << dist << endl;
+ if (output == "column") { outFile << alignDB.get(i).getName() << ' ' << alignDB.get(j).getName() << ' ' << dist << endl; }
}
-
+ if (output == "lt") { outFile << dist << '\t'; }
+
+ if (output == "square") { //make a square column you can convert to square phylip
+ outFile << alignDB.get(i).getName() << '\t' << alignDB.get(j).getName() << '\t' << dist << endl;
+ outFile << alignDB.get(j).getName() << '\t' << alignDB.get(i).getName() << '\t' << dist << endl;
+ }
+
}
+
+ if (output == "lt") { outFile << endl; }
+
if(i % 100 == 0){
- cout << i << '\t' << time(NULL) - startTime << endl;
+ mothurOut(toString(i) + "\t" + toString(time(NULL) - startTime)); mothurOutEndLine();
}
-
+
}
- cout << endLine-1 << '\t' << time(NULL) - startTime << endl;
-
+ mothurOut(toString(endLine-1) + "\t" + toString(time(NULL) - startTime)); mothurOutEndLine();
+
+ outFile.close();
+
return 1;
}
catch(exception& e) {
- cout << "Standard Error: " << e.what() << " has occurred in the DistanceCommand class Function driver. Please contact Pat Schloss at pschloss@microbio.umass.edu." << "\n";
+ errorOut(e, "DistanceCommand", "driver");
exit(1);
}
- catch(...) {
- cout << "An unknown error has occurred in the DistanceCommand class function driver. Please contact Pat Schloss at pschloss@microbio.umass.edu." << "\n";
- exit(1);
- }
-
}
-
/**************************************************************************************************/
+void DistanceCommand::convertMatrix(string outputFile) {
+ try{
+
+ //sort file by first column so the distances for each row are together
+ string outfile = getRootName(outputFile) + "sorted.dist.temp";
+
+ //use the unix sort
+ #if defined (__APPLE__) || (__MACH__) || (linux) || (__linux)
+ string command = "sort -n " + outputFile + " -o " + outfile;
+ system(command.c_str());
+ #else //sort using windows sort
+ string command = "sort " + outputFile + " /O " + outfile;
+ system(command.c_str());
+ #endif
+
+
+ //output to new file distance for each row and save positions in file where new row begins
+ ifstream in;
+ openInputFile(outfile, in);
+
+ ofstream out;
+ openOutputFile(outputFile, out);
+
+ out.setf(ios::fixed, ios::floatfield); out.setf(ios::showpoint);
+
+ out << alignDB.getNumSeqs() << endl;
+
+ //get first currentRow
+ string first, currentRow, second;
+ float dist;
+ map<string, float> rowDists; //take advantage of the fact that maps are already sorted by key
+ map<string, float>::iterator it;
+
+ in >> first;
+ currentRow = first;
+
+ rowDists[first] = 0.00; //distance to yourself is 0.0
+
+ in.seekg(0);
+ //openInputFile(outfile, in);
+
+ while(!in.eof()) {
+ in >> first >> second >> dist; gobble(in);
+
+ if (first != currentRow) {
+ //print out last row
+ out << currentRow << '\t'; //print name
+
+ //print dists
+ for (it = rowDists.begin(); it != rowDists.end(); it++) {
+ out << it->second << '\t';
+ }
+ out << endl;
+
+ //start new row
+ currentRow = first;
+ rowDists.clear();
+ rowDists[first] = 0.00;
+ rowDists[second] = dist;
+ }else{
+ rowDists[second] = dist;
+ }
+ }
+ //print out last row
+ out << currentRow << '\t'; //print name
+
+ //print dists
+ for (it = rowDists.begin(); it != rowDists.end(); it++) {
+ out << it->second << '\t';
+ }
+ out << endl;
+
+ in.close();
+ out.close();
+
+ remove(outfile.c_str());
+
+ }
+ catch(exception& e) {
+ errorOut(e, "DistanceCommand", "convertMatrix");
+ exit(1);
+ }
+}
+/**************************************************************************************************
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();
}
catch(exception& e) {
- cout << "Standard Error: " << e.what() << " has occurred in the DistanceCommand class Function appendFiles. Please contact Pat Schloss at pschloss@microbio.umass.edu." << "\n";
+ errorOut(e, "DistanceCommand", "appendFiles");
exit(1);
}
- catch(...) {
- cout << "An unknown error has occurred in the DistanceCommand class function appendFiles. Please contact Pat Schloss at pschloss@microbio.umass.edu." << "\n";
- exit(1);
- }
}
+/**************************************************************************************************/
+
+