2 * pairwiseseqscommand.cpp
5 * Created by westcott on 10/20/10.
6 * Copyright 2010 Schloss Lab. All rights reserved.
10 #include "pairwiseseqscommand.h"
11 #include "sequence.hpp"
13 #include "gotohoverlap.hpp"
14 #include "needlemanoverlap.hpp"
15 #include "blastalign.hpp"
16 #include "noalign.hpp"
18 #include "ignoregaps.h"
19 #include "eachgapdist.h"
20 #include "eachgapignore.h"
21 #include "onegapdist.h"
22 #include "onegapignore.h"
25 //**********************************************************************************************************************
26 vector<string> PairwiseSeqsCommand::setParameters(){
28 CommandParameter pfasta("fasta", "InputTypes", "", "", "none", "none", "none",false,true); parameters.push_back(pfasta);
29 CommandParameter palign("align", "Multiple", "needleman-gotoh-blast-noalign", "needleman", "", "", "",false,false); parameters.push_back(palign);
30 CommandParameter pmatch("match", "Number", "", "1.0", "", "", "",false,false); parameters.push_back(pmatch);
31 CommandParameter pmismatch("mismatch", "Number", "", "-1.0", "", "", "",false,false); parameters.push_back(pmismatch);
32 CommandParameter pgapopen("gapopen", "Number", "", "-2.0", "", "", "",false,false); parameters.push_back(pgapopen);
33 CommandParameter pgapextend("gapextend", "Number", "", "-1.0", "", "", "",false,false); parameters.push_back(pgapextend);
34 CommandParameter pprocessors("processors", "Number", "", "1", "", "", "",false,false); parameters.push_back(pprocessors);
35 CommandParameter poutput("output", "Multiple", "column-lt-square", "column", "", "", "",false,false); parameters.push_back(poutput);
36 CommandParameter pcalc("calc", "Multiple", "nogaps-eachgap-onegap", "onegap", "", "", "",false,false); parameters.push_back(pcalc);
37 CommandParameter pcountends("countends", "Boolean", "", "T", "", "", "",false,false); parameters.push_back(pcountends);
38 CommandParameter pcompress("compress", "Boolean", "", "F", "", "", "",false,false); parameters.push_back(pcompress);
39 CommandParameter pcutoff("cutoff", "Number", "", "1.0", "", "", "",false,false); parameters.push_back(pcutoff);
40 CommandParameter pinputdir("inputdir", "String", "", "", "", "", "",false,false); parameters.push_back(pinputdir);
41 CommandParameter poutputdir("outputdir", "String", "", "", "", "", "",false,false); parameters.push_back(poutputdir);
43 vector<string> myArray;
44 for (int i = 0; i < parameters.size(); i++) { myArray.push_back(parameters[i].name); }
48 m->errorOut(e, "PairwiseSeqsCommand", "setParameters");
52 //**********************************************************************************************************************
53 string PairwiseSeqsCommand::getHelpString(){
55 string helpString = "";
56 helpString += "The pairwise.seqs command reads a fasta file and creates distance matrix.\n";
57 helpString += "The pairwise.seqs command parameters are fasta, align, match, mismatch, gapopen, gapextend, calc, output, cutoff and processors.\n";
58 helpString += "The fasta parameter is required. You may enter multiple fasta files by separating their names with dashes. ie. fasta=abrecovery.fasta-amzon.fasta \n";
59 helpString += "The align parameter allows you to specify the alignment method to use. Your options are: gotoh, needleman, blast and noalign. The default is needleman.\n";
60 helpString += "The match parameter allows you to specify the bonus for having the same base. The default is 1.0.\n";
61 helpString += "The mistmatch parameter allows you to specify the penalty for having different bases. The default is -1.0.\n";
62 helpString += "The gapopen parameter allows you to specify the penalty for opening a gap in an alignment. The default is -2.0.\n";
63 helpString += "The gapextend parameter allows you to specify the penalty for extending a gap in an alignment. The default is -1.0.\n";
64 helpString += "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";
65 helpString += "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";
66 helpString += "The cutoff parameter allows you to specify maximum distance to keep. The default is 1.0.\n";
67 helpString += "The output parameter allows you to specify format of your distance matrix. Options are column, lt, and square. The default is column.\n";
68 helpString += "The compress parameter allows you to indicate that you want the resulting distance file compressed. The default is false.\n";
69 helpString += "The pairwise.seqs command should be in the following format: \n";
70 helpString += "pairwise.seqs(fasta=yourfastaFile, align=yourAlignmentMethod) \n";
71 helpString += "Example pairwise.seqs(fasta=candidate.fasta, align=blast)\n";
72 helpString += "Note: No spaces between parameter labels (i.e. fasta), '=' and parameters (i.e.yourFastaFile).\n";
76 m->errorOut(e, "PairwiseSeqsCommand", "getHelpString");
81 //**********************************************************************************************************************
82 PairwiseSeqsCommand::PairwiseSeqsCommand(){
84 abort = true; calledHelp = true;
86 vector<string> tempOutNames;
87 outputTypes["phylip"] = tempOutNames;
88 outputTypes["column"] = tempOutNames;
91 m->errorOut(e, "PairwiseSeqsCommand", "PairwiseSeqsCommand");
95 //**********************************************************************************************************************
96 PairwiseSeqsCommand::PairwiseSeqsCommand(string option) {
98 abort = false; calledHelp = false;
100 //allow user to run help
101 if(option == "help") { help(); abort = true; calledHelp = true; }
102 else if(option == "citation") { citation(); abort = true; calledHelp = true;}
105 vector<string> myArray = setParameters();
107 OptionParser parser(option);
108 map<string, string> parameters = parser.getParameters();
110 ValidParameters validParameter("pairwise.seqs");
111 map<string, string>::iterator it;
113 //check to make sure all parameters are valid for command
114 for (it = parameters.begin(); it != parameters.end(); it++) {
115 if (validParameter.isValidParameter(it->first, myArray, it->second) != true) { abort = true; }
118 //initialize outputTypes
119 vector<string> tempOutNames;
120 outputTypes["phylip"] = tempOutNames;
121 outputTypes["column"] = tempOutNames;
123 //if the user changes the output directory command factory will send this info to us in the output parameter
124 outputDir = validParameter.validFile(parameters, "outputdir", false); if (outputDir == "not found"){ outputDir = ""; }
127 //if the user changes the input directory command factory will send this info to us in the output parameter
128 string inputDir = validParameter.validFile(parameters, "inputdir", false);
130 if (inputDir == "not found"){ inputDir = ""; }
132 fastaFileName = validParameter.validFile(parameters, "fasta", false);
133 if (fastaFileName == "not found") {
134 //if there is a current fasta file, use it
135 string filename = m->getFastaFile();
136 if (filename != "") { fastaFileNames.push_back(filename); m->mothurOut("Using " + filename + " as input file for the fasta parameter."); m->mothurOutEndLine(); }
137 else { m->mothurOut("You have no current fastafile and the fasta parameter is required."); m->mothurOutEndLine(); abort = true; }
140 m->splitAtDash(fastaFileName, fastaFileNames);
142 //go through files and make sure they are good, if not, then disregard them
143 for (int i = 0; i < fastaFileNames.size(); i++) {
146 if (fastaFileNames[i] == "current") {
147 fastaFileNames[i] = m->getFastaFile();
148 if (fastaFileNames[i] != "") { m->mothurOut("Using " + fastaFileNames[i] + " as input file for the fasta parameter where you had given current."); m->mothurOutEndLine(); }
150 m->mothurOut("You have no current fastafile, ignoring current."); m->mothurOutEndLine(); ignore=true;
151 //erase from file list
152 fastaFileNames.erase(fastaFileNames.begin()+i);
159 if (inputDir != "") {
160 string path = m->hasPath(fastaFileNames[i]);
161 //if the user has not given a path then, add inputdir. else leave path alone.
162 if (path == "") { fastaFileNames[i] = inputDir + fastaFileNames[i]; }
168 ableToOpen = m->openInputFile(fastaFileNames[i], in, "noerror");
170 //if you can't open it, try default location
171 if (ableToOpen == 1) {
172 if (m->getDefaultPath() != "") { //default path is set
173 string tryPath = m->getDefaultPath() + m->getSimpleName(fastaFileNames[i]);
174 m->mothurOut("Unable to open " + fastaFileNames[i] + ". Trying default " + tryPath); m->mothurOutEndLine();
176 ableToOpen = m->openInputFile(tryPath, in2, "noerror");
178 fastaFileNames[i] = tryPath;
182 //if you can't open it, try output location
183 if (ableToOpen == 1) {
184 if (m->getOutputDir() != "") { //default path is set
185 string tryPath = m->getOutputDir() + m->getSimpleName(fastaFileNames[i]);
186 m->mothurOut("Unable to open " + fastaFileNames[i] + ". Trying output directory " + tryPath); m->mothurOutEndLine();
188 ableToOpen = m->openInputFile(tryPath, in2, "noerror");
190 fastaFileNames[i] = tryPath;
196 if (ableToOpen == 1) {
197 m->mothurOut("Unable to open " + fastaFileNames[i] + ". It will be disregarded."); m->mothurOutEndLine();
198 //erase from file list
199 fastaFileNames.erase(fastaFileNames.begin()+i);
202 m->setFastaFile(fastaFileNames[i]);
207 //make sure there is at least one valid file left
208 if (fastaFileNames.size() == 0) { m->mothurOut("no valid files."); m->mothurOutEndLine(); abort = true; }
211 //check for optional parameter and set defaults
212 // ...at some point should added some additional type checking...
214 temp = validParameter.validFile(parameters, "match", false); if (temp == "not found"){ temp = "1.0"; }
215 m->mothurConvert(temp, match);
217 temp = validParameter.validFile(parameters, "mismatch", false); if (temp == "not found"){ temp = "-1.0"; }
218 m->mothurConvert(temp, misMatch);
220 temp = validParameter.validFile(parameters, "gapopen", false); if (temp == "not found"){ temp = "-2.0"; }
221 m->mothurConvert(temp, gapOpen);
223 temp = validParameter.validFile(parameters, "gapextend", false); if (temp == "not found"){ temp = "-1.0"; }
224 m->mothurConvert(temp, gapExtend);
226 temp = validParameter.validFile(parameters, "processors", false); if (temp == "not found"){ temp = m->getProcessors(); }
227 m->setProcessors(temp);
228 m->mothurConvert(temp, processors);
230 temp = validParameter.validFile(parameters, "cutoff", false); if(temp == "not found"){ temp = "1.0"; }
231 m->mothurConvert(temp, cutoff);
233 temp = validParameter.validFile(parameters, "countends", false); if(temp == "not found"){ temp = "T"; }
234 countends = m->isTrue(temp);
236 temp = validParameter.validFile(parameters, "compress", false); if(temp == "not found"){ temp = "F"; }
237 compress = m->isTrue(temp);
239 align = validParameter.validFile(parameters, "align", false); if (align == "not found"){ align = "needleman"; }
241 output = validParameter.validFile(parameters, "output", false); if(output == "not found"){ output = "column"; }
242 if ((output != "column") && (output != "lt") && (output != "square")) { m->mothurOut(output + " is not a valid output form. Options are column, lt and square. I will use column."); m->mothurOutEndLine(); output = "column"; }
244 calc = validParameter.validFile(parameters, "calc", false);
245 if (calc == "not found") { calc = "onegap"; }
247 if (calc == "default") { calc = "onegap"; }
249 m->splitAtDash(calc, Estimators);
251 ValidCalculators validCalculator;
253 for (int i=0; i<Estimators.size(); i++) {
254 if (validCalculator.isValidCalculator("distance", Estimators[i]) == true) {
255 if (Estimators[i] == "nogaps") { distCalculator = new ignoreGaps(); }
256 else if (Estimators[i] == "eachgap") { distCalculator = new eachGapDist(); }
257 else if (Estimators[i] == "onegap") { distCalculator = new oneGapDist(); }
261 for (int i=0; i<Estimators.size(); i++) {
262 if (validCalculator.isValidCalculator("distance", Estimators[i]) == true) {
263 if (Estimators[i] == "nogaps") { distCalculator = new ignoreGaps(); }
264 else if (Estimators[i] == "eachgap"){ distCalculator = new eachGapIgnoreTermGapDist(); }
265 else if (Estimators[i] == "onegap") { distCalculator = new oneGapIgnoreTermGapDist(); }
272 catch(exception& e) {
273 m->errorOut(e, "PairwiseSeqsCommand", "PairwiseSeqsCommand");
277 //**********************************************************************************************************************
279 int PairwiseSeqsCommand::execute(){
281 if (abort == true) { if (calledHelp) { return 0; } return 2; }
283 int longestBase = 2000; //will need to update this in driver if we find sequences with more bases. hardcoded so we don't have the pre-read user fasta file.
285 if(align == "gotoh") { alignment = new GotohOverlap(gapOpen, gapExtend, match, misMatch, longestBase); }
286 else if(align == "needleman") { alignment = new NeedlemanOverlap(gapOpen, match, misMatch, longestBase); }
287 else if(align == "blast") { alignment = new BlastAlignment(gapOpen, gapExtend, match, misMatch); }
288 else if(align == "noalign") { alignment = new NoAlign(); }
290 m->mothurOut(align + " is not a valid alignment option. I will run the command using needleman.");
291 m->mothurOutEndLine();
292 alignment = new NeedlemanOverlap(gapOpen, match, misMatch, longestBase);
297 for (int s = 0; s < fastaFileNames.size(); s++) {
298 if (m->control_pressed) { outputTypes.clear(); return 0; }
300 m->mothurOut("Processing sequences from " + fastaFileNames[s] + " ..." ); m->mothurOutEndLine();
302 if (outputDir == "") { outputDir += m->hasPath(fastaFileNames[s]); }
305 m->openInputFile(fastaFileNames[s], inFASTA);
306 alignDB = SequenceDB(inFASTA);
309 int numSeqs = alignDB.getNumSeqs();
310 int startTime = time(NULL);
311 string outputFile = "";
313 if (output == "lt") { //does the user want lower triangle phylip formatted file
314 outputFile = outputDir + m->getRootName(m->getSimpleName(fastaFileNames[s])) + "phylip.dist";
315 m->mothurRemove(outputFile); outputTypes["phylip"].push_back(outputFile);
316 }else if (output == "column") { //user wants column format
317 outputFile = outputDir + m->getRootName(m->getSimpleName(fastaFileNames[s])) + "dist";
318 outputTypes["column"].push_back(outputFile);
319 m->mothurRemove(outputFile);
320 }else { //assume square
321 outputFile = outputDir + m->getRootName(m->getSimpleName(fastaFileNames[s])) + "square.dist";
322 m->mothurRemove(outputFile);
323 outputTypes["phylip"].push_back(outputFile);
332 MPI_Comm_size(MPI_COMM_WORLD, &processors); //set processors to the number of mpi processes running
333 MPI_Comm_rank(MPI_COMM_WORLD, &pid); //find out who we are
335 //each process gets where it should start and stop in the file
336 if (output != "square") {
337 start = int (sqrt(float(pid)/float(processors)) * numSeqs);
338 end = int (sqrt(float(pid+1)/float(processors)) * numSeqs);
340 start = int ((float(pid)/float(processors)) * numSeqs);
341 end = int ((float(pid+1)/float(processors)) * numSeqs);
344 if (output == "column") {
346 int amode=MPI_MODE_CREATE|MPI_MODE_WRONLY;
349 strcpy(filename, outputFile.c_str());
351 MPI_File_open(MPI_COMM_WORLD, filename, amode, MPI_INFO_NULL, &outMPI);
353 if (pid == 0) { //you are the root process
358 driverMPI(start, end, outMPI, cutoff);
360 if (m->control_pressed) { outputTypes.clear(); MPI_File_close(&outMPI); m->mothurRemove(outputFile); delete distCalculator; return 0; }
363 for(int i = 1; i < processors; i++) {
364 if (m->control_pressed) { outputTypes.clear(); MPI_File_close(&outMPI); m->mothurRemove(outputFile); delete distCalculator; return 0; }
367 MPI_Recv(buf, 5, MPI_CHAR, i, tag, MPI_COMM_WORLD, &status);
369 }else { //you are a child process
371 driverMPI(start, end, outMPI, cutoff);
373 if (m->control_pressed) { outputTypes.clear(); MPI_File_close(&outMPI); m->mothurRemove(outputFile); delete distCalculator; return 0; }
377 //tell parent you are done.
378 MPI_Send(buf, 5, MPI_CHAR, 0, tag, MPI_COMM_WORLD);
381 MPI_File_close(&outMPI);
383 }else { //lower triangle format
384 if (pid == 0) { //you are the root process
388 unsigned long long mySize;
390 if (output != "square"){ driverMPI(start, end, outputFile, mySize); }
391 else { driverMPI(start, end, outputFile, mySize, output); }
393 if (m->control_pressed) { outputTypes.clear(); m->mothurRemove(outputFile); delete distCalculator; return 0; }
395 int amode=MPI_MODE_APPEND|MPI_MODE_WRONLY|MPI_MODE_CREATE; //
400 strcpy(filename, outputFile.c_str());
402 MPI_File_open(MPI_COMM_SELF, filename, amode, MPI_INFO_NULL, &outMPI);
405 for(int b = 1; b < processors; b++) {
406 unsigned long long fileSize;
408 if (m->control_pressed) { outputTypes.clear(); MPI_File_close(&outMPI); m->mothurRemove(outputFile); delete distCalculator; return 0; }
410 MPI_Recv(&fileSize, 1, MPI_LONG, b, tag, MPI_COMM_WORLD, &status);
412 string outTemp = outputFile + toString(b) + ".temp";
414 char* buf = new char[outTemp.length()];
415 memcpy(buf, outTemp.c_str(), outTemp.length());
417 MPI_File_open(MPI_COMM_SELF, buf, MPI_MODE_DELETE_ON_CLOSE|MPI_MODE_RDONLY, MPI_INFO_NULL, &inMPI);
421 while (count < fileSize) {
423 MPI_File_read(inMPI, buf2, 1, MPI_CHAR, &status);
424 MPI_File_write(outMPI, buf2, 1, MPI_CHAR, &status);
428 MPI_File_close(&inMPI); //deleted on close
431 MPI_File_close(&outMPI);
432 }else { //you are a child process
434 unsigned long long size;
435 if (output != "square"){ driverMPI(start, end, (outputFile + toString(pid) + ".temp"), size); }
436 else { driverMPI(start, end, (outputFile + toString(pid) + ".temp"), size, output); }
438 if (m->control_pressed) { delete distCalculator; return 0; }
440 //tell parent you are done.
441 MPI_Send(&size, 1, MPI_LONG, 0, tag, MPI_COMM_WORLD);
444 MPI_Barrier(MPI_COMM_WORLD); //make everyone wait - just in case
447 #if defined (__APPLE__) || (__MACH__) || (linux) || (__linux)
448 //if you don't need to fork anything
450 if (output != "square") { driver(0, numSeqs, outputFile, cutoff); }
451 else { driver(0, numSeqs, outputFile, "square"); }
452 }else{ //you have multiple processors
454 for (int i = 0; i < processors; i++) {
455 distlinePair tempLine;
456 lines.push_back(tempLine);
457 if (output != "square") {
458 lines[i].start = int (sqrt(float(i)/float(processors)) * numSeqs);
459 lines[i].end = int (sqrt(float(i+1)/float(processors)) * numSeqs);
461 lines[i].start = int ((float(i)/float(processors)) * numSeqs);
462 lines[i].end = int ((float(i+1)/float(processors)) * numSeqs);
466 createProcesses(outputFile);
470 if (output != "square") { driver(0, numSeqs, outputFile, cutoff); }
471 else { driver(0, numSeqs, outputFile, "square"); }
475 if (m->control_pressed) { outputTypes.clear(); delete distCalculator; m->mothurRemove(outputFile); return 0; }
478 MPI_Comm_rank(MPI_COMM_WORLD, &pid);
480 if (pid == 0) { //only one process should output to screen
484 fileHandle.open(outputFile.c_str());
486 m->gobble(fileHandle);
487 if (fileHandle.eof()) { m->mothurOut(outputFile + " is blank. This can result if there are no distances below your cutoff."); m->mothurOutEndLine(); }
491 m->mothurOut("Compressing..."); m->mothurOutEndLine();
492 m->mothurOut("(Replacing " + outputFile + " with " + outputFile + ".gz)"); m->mothurOutEndLine();
493 system(("gzip -v " + outputFile).c_str());
494 outputNames.push_back(outputFile + ".gz");
495 }else { outputNames.push_back(outputFile); }
501 m->mothurOut("It took " + toString(time(NULL) - startTime) + " to calculate the distances for " + toString(numSeqs) + " sequences."); m->mothurOutEndLine();
503 if (m->control_pressed) { outputTypes.clear(); delete distCalculator; m->mothurRemove(outputFile); return 0; }
506 delete distCalculator;
508 //set phylip file as new current phylipfile
510 itTypes = outputTypes.find("phylip");
511 if (itTypes != outputTypes.end()) {
512 if ((itTypes->second).size() != 0) { current = (itTypes->second)[0]; m->setPhylipFile(current); }
515 //set column file as new current columnfile
516 itTypes = outputTypes.find("column");
517 if (itTypes != outputTypes.end()) {
518 if ((itTypes->second).size() != 0) { current = (itTypes->second)[0]; m->setColumnFile(current); }
521 m->mothurOutEndLine();
522 m->mothurOut("Output File Name: "); m->mothurOutEndLine();
523 for (int i = 0; i < outputNames.size(); i++) { m->mothurOut(outputNames[i]); m->mothurOutEndLine(); }
524 m->mothurOutEndLine();
529 catch(exception& e) {
530 m->errorOut(e, "PairwiseSeqsCommand", "execute");
535 /**************************************************************************************************/
536 void PairwiseSeqsCommand::createProcesses(string filename) {
538 #if defined (__APPLE__) || (__MACH__) || (linux) || (__linux)
542 //loop through and create all the processes you want
543 while (process != processors) {
547 processIDS.push_back(pid);
550 if (output != "square") { driver(lines[process].start, lines[process].end, filename + toString(getpid()) + ".temp", cutoff); }
551 else { driver(lines[process].start, lines[process].end, filename + toString(getpid()) + ".temp", "square"); }
554 m->mothurOut("[ERROR]: unable to spawn the necessary processes."); m->mothurOutEndLine();
555 for (int i=0;i<processIDS.size();i++) { int temp = processIDS[i]; kill (temp, SIGINT); }
561 if (output != "square") { driver(lines[0].start, lines[0].end, filename, cutoff); }
562 else { driver(lines[0].start, lines[0].end, filename, "square"); }
565 //force parent to wait until all the processes are done
566 for (int i=0;i<processIDS.size();i++) {
567 int temp = processIDS[i];
571 //append and remove temp files
572 for (int i=0;i<processIDS.size();i++) {
573 m->appendFiles((filename + toString(processIDS[i]) + ".temp"), filename);
574 m->mothurRemove((filename + toString(processIDS[i]) + ".temp"));
578 catch(exception& e) {
579 m->errorOut(e, "PairwiseSeqsCommand", "createProcesses");
584 /**************************************************************************************************/
585 /////// need to fix to work with calcs and sequencedb
586 int PairwiseSeqsCommand::driver(int startLine, int endLine, string dFileName, float cutoff){
589 int startTime = time(NULL);
592 ofstream outFile(dFileName.c_str(), ios::trunc);
593 outFile.setf(ios::fixed, ios::showpoint);
594 outFile << setprecision(4);
596 if((output == "lt") && startLine == 0){ outFile << alignDB.getNumSeqs() << endl; }
598 for(int i=startLine;i<endLine;i++){
600 string name = alignDB.get(i).getName();
601 if (name.length() < 10) { //pad with spaces to make compatible
602 while (name.length() < 10) { name += " "; }
604 outFile << name << '\t';
607 for(int j=0;j<i;j++){
609 if (m->control_pressed) { outFile.close(); return 0; }
611 if (alignDB.get(i).getUnaligned().length() > alignment->getnRows()) {
612 alignment->resize(alignDB.get(i).getUnaligned().length()+1);
615 if (alignDB.get(j).getUnaligned().length() > alignment->getnRows()) {
616 alignment->resize(alignDB.get(j).getUnaligned().length()+1);
619 Sequence seqI(alignDB.get(i).getName(), alignDB.get(i).getAligned());
620 Sequence seqJ(alignDB.get(j).getName(), alignDB.get(j).getAligned());
622 alignment->align(seqI.getUnaligned(), seqJ.getUnaligned());
623 seqI.setAligned(alignment->getSeqAAln());
624 seqJ.setAligned(alignment->getSeqBAln());
627 distCalculator->calcDist(seqI, seqJ);
628 double dist = distCalculator->getDist();
631 if (output == "column") { outFile << alignDB.get(i).getName() << ' ' << alignDB.get(j).getName() << ' ' << dist << endl; }
633 if (output == "lt") { outFile << dist << '\t'; }
636 if (output == "lt") { outFile << endl; }
639 m->mothurOut(toString(i) + "\t" + toString(time(NULL) - startTime)); m->mothurOutEndLine();
643 m->mothurOut(toString(endLine-1) + "\t" + toString(time(NULL) - startTime)); m->mothurOutEndLine();
649 catch(exception& e) {
650 m->errorOut(e, "PairwiseSeqsCommand", "driver");
654 /**************************************************************************************************/
655 /////// need to fix to work with calcs and sequencedb
656 int PairwiseSeqsCommand::driver(int startLine, int endLine, string dFileName, string square){
659 int startTime = time(NULL);
662 ofstream outFile(dFileName.c_str(), ios::trunc);
663 outFile.setf(ios::fixed, ios::showpoint);
664 outFile << setprecision(4);
666 if(startLine == 0){ outFile << alignDB.getNumSeqs() << endl; }
668 for(int i=startLine;i<endLine;i++){
670 string name = alignDB.get(i).getName();
671 //pad with spaces to make compatible
672 if (name.length() < 10) { while (name.length() < 10) { name += " "; } }
674 outFile << name << '\t';
676 for(int j=0;j<alignDB.getNumSeqs();j++){
678 if (m->control_pressed) { outFile.close(); return 0; }
680 if (alignDB.get(i).getUnaligned().length() > alignment->getnRows()) {
681 alignment->resize(alignDB.get(i).getUnaligned().length()+1);
684 if (alignDB.get(j).getUnaligned().length() > alignment->getnRows()) {
685 alignment->resize(alignDB.get(j).getUnaligned().length()+1);
688 Sequence seqI(alignDB.get(i).getName(), alignDB.get(i).getAligned());
689 Sequence seqJ(alignDB.get(j).getName(), alignDB.get(j).getAligned());
691 alignment->align(seqI.getUnaligned(), seqJ.getUnaligned());
692 seqI.setAligned(alignment->getSeqAAln());
693 seqJ.setAligned(alignment->getSeqBAln());
695 distCalculator->calcDist(seqI, seqJ);
696 double dist = distCalculator->getDist();
698 outFile << dist << '\t';
704 m->mothurOut(toString(i) + "\t" + toString(time(NULL) - startTime)); m->mothurOutEndLine();
708 m->mothurOut(toString(endLine-1) + "\t" + toString(time(NULL) - startTime)); m->mothurOutEndLine();
714 catch(exception& e) {
715 m->errorOut(e, "PairwiseSeqsCommand", "driver");
720 /**************************************************************************************************/
721 /////// need to fix to work with calcs and sequencedb
722 int PairwiseSeqsCommand::driverMPI(int startLine, int endLine, MPI_File& outMPI, float cutoff){
725 int startTime = time(NULL);
727 string outputString = "";
729 for(int i=startLine;i<endLine;i++){
731 for(int j=0;j<i;j++){
733 if (m->control_pressed) { return 0; }
735 if (alignDB.get(i).getUnaligned().length() > alignment->getnRows()) {
736 alignment->resize(alignDB.get(i).getUnaligned().length()+1);
739 if (alignDB.get(j).getUnaligned().length() > alignment->getnRows()) {
740 alignment->resize(alignDB.get(j).getUnaligned().length()+1);
743 Sequence seqI(alignDB.get(i).getName(), alignDB.get(i).getAligned());
744 Sequence seqJ(alignDB.get(j).getName(), alignDB.get(j).getAligned());
746 alignment->align(seqI.getUnaligned(), seqJ.getUnaligned());
747 seqI.setAligned(alignment->getSeqAAln());
748 seqJ.setAligned(alignment->getSeqBAln());
750 distCalculator->calcDist(seqI, seqJ);
751 double dist = distCalculator->getDist();
754 outputString += (alignDB.get(i).getName() + ' ' + alignDB.get(j).getName() + ' ' + toString(dist) + '\n');
759 //m->mothurOut(toString(i) + "\t" + toString(time(NULL) - startTime)); m->mothurOutEndLine();
760 cout << i << '\t' << (time(NULL) - startTime) << endl;
764 //send results to parent
765 int length = outputString.length();
767 char* buf = new char[length];
768 memcpy(buf, outputString.c_str(), length);
770 MPI_File_write_shared(outMPI, buf, length, MPI_CHAR, &status);
778 catch(exception& e) {
779 m->errorOut(e, "PairwiseSeqsCommand", "driverMPI");
783 /**************************************************************************************************/
784 /////// need to fix to work with calcs and sequencedb
785 int PairwiseSeqsCommand::driverMPI(int startLine, int endLine, string file, unsigned long long& size){
790 int amode=MPI_MODE_CREATE|MPI_MODE_WRONLY;
793 strcpy(filename, file.c_str());
795 MPI_File_open(MPI_COMM_SELF, filename, amode, MPI_INFO_NULL, &outMPI);
799 string outputString = "";
802 if(startLine == 0){ outputString += toString(alignDB.getNumSeqs()) + "\n"; }
804 for(int i=startLine;i<endLine;i++){
806 string name = alignDB.get(i).getName();
807 if (name.length() < 10) { //pad with spaces to make compatible
808 while (name.length() < 10) { name += " "; }
810 outputString += name + "\t";
812 for(int j=0;j<i;j++){
814 if (m->control_pressed) { return 0; }
816 if (alignDB.get(i).getUnaligned().length() > alignment->getnRows()) {
817 alignment->resize(alignDB.get(i).getUnaligned().length()+1);
820 if (alignDB.get(j).getUnaligned().length() > alignment->getnRows()) {
821 alignment->resize(alignDB.get(j).getUnaligned().length()+1);
824 Sequence seqI(alignDB.get(i).getName(), alignDB.get(i).getAligned());
825 Sequence seqJ(alignDB.get(j).getName(), alignDB.get(j).getAligned());
827 alignment->align(seqI.getUnaligned(), seqJ.getUnaligned());
828 seqI.setAligned(alignment->getSeqAAln());
829 seqJ.setAligned(alignment->getSeqBAln());
831 distCalculator->calcDist(seqI, seqJ);
832 double dist = distCalculator->getDist();
834 outputString += toString(dist) + "\t";
837 outputString += "\n";
839 //send results to parent
840 int length = outputString.length();
841 char* buf = new char[length];
842 memcpy(buf, outputString.c_str(), length);
844 MPI_File_write(outMPI, buf, length, MPI_CHAR, &status);
845 size += outputString.length();
850 MPI_File_close(&outMPI);
854 catch(exception& e) {
855 m->errorOut(e, "PairwiseSeqsCommand", "driverMPI");
859 /**************************************************************************************************/
860 /////// need to fix to work with calcs and sequencedb
861 int PairwiseSeqsCommand::driverMPI(int startLine, int endLine, string file, unsigned long long& size, string square){
866 int amode=MPI_MODE_CREATE|MPI_MODE_WRONLY;
869 strcpy(filename, file.c_str());
871 MPI_File_open(MPI_COMM_SELF, filename, amode, MPI_INFO_NULL, &outMPI);
875 string outputString = "";
878 if(startLine == 0){ outputString += toString(alignDB.getNumSeqs()) + "\n"; }
880 for(int i=startLine;i<endLine;i++){
882 string name = alignDB.get(i).getName();
883 if (name.length() < 10) { //pad with spaces to make compatible
884 while (name.length() < 10) { name += " "; }
886 outputString += name + "\t";
888 for(int j=0;j<alignDB.getNumSeqs();j++){
890 if (m->control_pressed) { return 0; }
892 if (alignDB.get(i).getUnaligned().length() > alignment->getnRows()) {
893 alignment->resize(alignDB.get(i).getUnaligned().length()+1);
896 if (alignDB.get(j).getUnaligned().length() > alignment->getnRows()) {
897 alignment->resize(alignDB.get(j).getUnaligned().length()+1);
900 Sequence seqI(alignDB.get(i).getName(), alignDB.get(i).getAligned());
901 Sequence seqJ(alignDB.get(j).getName(), alignDB.get(j).getAligned());
903 alignment->align(seqI.getUnaligned(), seqJ.getUnaligned());
904 seqI.setAligned(alignment->getSeqAAln());
905 seqJ.setAligned(alignment->getSeqBAln());
907 distCalculator->calcDist(seqI, seqJ);
908 double dist = distCalculator->getDist();
910 outputString += toString(dist) + "\t";
913 outputString += "\n";
915 //send results to parent
916 int length = outputString.length();
917 char* buf = new char[length];
918 memcpy(buf, outputString.c_str(), length);
920 MPI_File_write(outMPI, buf, length, MPI_CHAR, &status);
921 size += outputString.length();
926 MPI_File_close(&outMPI);
930 catch(exception& e) {
931 m->errorOut(e, "PairwiseSeqsCommand", "driverMPI");
936 /**************************************************************************************************/