5 * Created by Sarah Westcott on 5/7/09.
6 * Copyright 2009 Schloss Lab UMASS Amherst. All rights reserved.
10 #include "distancecommand.h"
11 #include "ignoregaps.h"
12 #include "eachgapdist.h"
13 #include "eachgapignore.h"
14 #include "onegapdist.h"
15 #include "onegapignore.h"
17 //**********************************************************************************************************************
19 DistanceCommand::DistanceCommand(string option) {
24 //allow user to run help
25 if(option == "help") { help(); abort = true; }
28 //valid paramters for this command
29 string Array[] = {"fasta","oldfasta","column", "output", "calc", "countends", "cutoff", "processors", "outputdir","inputdir"};
30 vector<string> myArray (Array, Array+(sizeof(Array)/sizeof(string)));
32 OptionParser parser(option);
33 map<string, string> parameters = parser.getParameters();
35 ValidParameters validParameter("dist.seqs");
36 map<string, string>::iterator it2;
38 //check to make sure all parameters are valid for command
39 for (it2 = parameters.begin(); it2 != parameters.end(); it2++) {
40 if (validParameter.isValidParameter(it2->first, myArray, it2->second) != true) { abort = true; }
43 //if the user changes the input directory command factory will send this info to us in the output parameter
44 string inputDir = validParameter.validFile(parameters, "inputdir", false);
45 if (inputDir == "not found"){ inputDir = ""; }
48 it2 = parameters.find("fasta");
49 //user has given a template file
50 if(it2 != parameters.end()){
51 path = hasPath(it2->second);
52 //if the user has not given a path then, add inputdir. else leave path alone.
53 if (path == "") { parameters["fasta"] = inputDir + it2->second; }
56 it2 = parameters.find("oldfasta");
57 //user has given a template file
58 if(it2 != parameters.end()){
59 path = hasPath(it2->second);
60 //if the user has not given a path then, add inputdir. else leave path alone.
61 if (path == "") { parameters["oldfasta"] = inputDir + it2->second; }
64 it2 = parameters.find("column");
65 //user has given a template file
66 if(it2 != parameters.end()){
67 path = hasPath(it2->second);
68 //if the user has not given a path then, add inputdir. else leave path alone.
69 if (path == "") { parameters["column"] = inputDir + it2->second; }
73 //check for required parameters
74 fastafile = validParameter.validFile(parameters, "fasta", true);
75 if (fastafile == "not found") { m->mothurOut("fasta is a required parameter for the dist.seqs command."); m->mothurOutEndLine(); abort = true; }
76 else if (fastafile == "not open") { abort = true; }
79 openInputFile(fastafile, inFASTA);
80 alignDB = SequenceDB(inFASTA);
84 oldfastafile = validParameter.validFile(parameters, "oldfasta", true);
85 if (oldfastafile == "not found") { oldfastafile = ""; }
86 else if (oldfastafile == "not open") { abort = true; }
88 column = validParameter.validFile(parameters, "column", true);
89 if (column == "not found") { column = ""; }
90 else if (column == "not open") { abort = true; }
92 //if the user changes the output directory command factory will send this info to us in the output parameter
93 outputDir = validParameter.validFile(parameters, "outputdir", false); if (outputDir == "not found"){
95 outputDir += hasPath(fastafile); //if user entered a file with a path then preserve it
98 //check for optional parameter and set defaults
99 // ...at some point should added some additional type checking...
100 calc = validParameter.validFile(parameters, "calc", false);
101 if (calc == "not found") { calc = "onegap"; }
103 if (calc == "default") { calc = "onegap"; }
105 splitAtDash(calc, Estimators);
108 temp = validParameter.validFile(parameters, "countends", false); if(temp == "not found"){ temp = "T"; }
109 convert(temp, countends);
111 temp = validParameter.validFile(parameters, "cutoff", false); if(temp == "not found"){ temp = "1.0"; }
112 convert(temp, cutoff);
114 temp = validParameter.validFile(parameters, "processors", false); if(temp == "not found"){ temp = "1"; }
115 convert(temp, processors);
117 output = validParameter.validFile(parameters, "output", false); if(output == "not found"){ output = "column"; }
119 if (((column != "") && (oldfastafile == "")) || ((column == "") && (oldfastafile != ""))) { m->mothurOut("If you provide column or oldfasta, you must provide both."); m->mothurOutEndLine(); abort=true; }
121 if ((column != "") && (oldfastafile != "") && (output != "column")) { m->mothurOut("You have provided column and oldfasta, indicating you want to append distances to your column file. Your output must be in column format to do so."); m->mothurOutEndLine(); abort=true; }
123 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"; }
125 ValidCalculators validCalculator;
127 if (isTrue(countends) == true) {
128 for (int i=0; i<Estimators.size(); i++) {
129 if (validCalculator.isValidCalculator("distance", Estimators[i]) == true) {
130 if (Estimators[i] == "nogaps") { distCalculator = new ignoreGaps(); }
131 else if (Estimators[i] == "eachgap") { distCalculator = new eachGapDist(); }
132 else if (Estimators[i] == "onegap") { distCalculator = new oneGapDist(); }
136 for (int i=0; i<Estimators.size(); i++) {
137 if (validCalculator.isValidCalculator("distance", Estimators[i]) == true) {
138 if (Estimators[i] == "nogaps") { distCalculator = new ignoreGaps(); }
139 else if (Estimators[i] == "eachgap"){ distCalculator = new eachGapIgnoreTermGapDist(); }
140 else if (Estimators[i] == "onegap") { distCalculator = new oneGapIgnoreTermGapDist(); }
148 catch(exception& e) {
149 m->errorOut(e, "DistanceCommand", "DistanceCommand");
154 //**********************************************************************************************************************
156 DistanceCommand::~DistanceCommand(){
158 for(int i=0;i<lines.size();i++){
164 //**********************************************************************************************************************
166 void DistanceCommand::help(){
168 m->mothurOut("The dist.seqs command reads a file containing sequences and creates a distance file.\n");
169 m->mothurOut("The dist.seqs command parameters are fasta, oldfasta, column, calc, countends, output, cutoff and processors. \n");
170 m->mothurOut("The fasta parameter is required.\n");
171 m->mothurOut("The oldfasta and column parameters allow you to append the distances calculated to the column file.\n");
172 m->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");
173 m->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");
174 m->mothurOut("The cutoff parameter allows you to specify maximum distance to keep. The default is 1.0.\n");
175 m->mothurOut("The output parameter allows you to specify format of your distance matrix. Options are column, lt, and square. The default is column.\n");
176 m->mothurOut("The processors parameter allows you to specify number of processors to use. The default is 1.\n");
177 m->mothurOut("The dist.seqs command should be in the following format: \n");
178 m->mothurOut("dist.seqs(fasta=yourFastaFile, calc=yourCalc, countends=yourEnds, cutoff= yourCutOff, processors=yourProcessors) \n");
179 m->mothurOut("Example dist.seqs(fasta=amazon.fasta, calc=eachgap, countends=F, cutoff= 2.0, processors=3).\n");
180 m->mothurOut("Note: No spaces between parameter labels (i.e. calc), '=' and parameters (i.e.yourCalc).\n\n");
182 catch(exception& e) {
183 m->errorOut(e, "DistanceCommand", "help");
187 //**********************************************************************************************************************
189 int DistanceCommand::execute(){
192 if (abort == true) { return 0; }
194 int startTime = time(NULL);
196 //save number of new sequence
197 numNewFasta = alignDB.getNumSeqs();
199 //sanity check the oldfasta and column file as well as add oldfasta sequences to alignDB
200 if ((oldfastafile != "") && (column != "")) { if (!(sanityCheck())) { return 0; } }
202 if (m->control_pressed) { return 0; }
204 int numSeqs = alignDB.getNumSeqs();
209 if (output == "lt") { //does the user want lower triangle phylip formatted file
210 outputFile = outputDir + getRootName(getSimpleName(fastafile)) + "phylip.dist";
211 remove(outputFile.c_str());
213 //output numSeqs to phylip formatted dist file
214 }else if (output == "column") { //user wants column format
215 outputFile = outputDir + getRootName(getSimpleName(fastafile)) + "dist";
217 //so we don't accidentally overwrite
218 if (outputFile == column) {
219 string tempcolumn = column + ".old";
220 rename(column.c_str(), tempcolumn.c_str());
223 remove(outputFile.c_str());
224 }else { //assume square
225 outputFile = outputDir + getRootName(getSimpleName(fastafile)) + "square.dist";
226 remove(outputFile.c_str());
236 MPI_Comm_size(MPI_COMM_WORLD, &processors); //set processors to the number of mpi processes running
237 MPI_Comm_rank(MPI_COMM_WORLD, &pid); //find out who we are
239 //each process gets where it should start and stop in the file
240 start = int (sqrt(float(pid)/float(processors)) * numSeqs);
241 end = int (sqrt(float(pid+1)/float(processors)) * numSeqs);
243 if (output != "lt") {
245 int amode=MPI_MODE_CREATE|MPI_MODE_WRONLY;
247 //char* filename = new char[outputFile.length()];
248 //memcpy(filename, outputFile.c_str(), outputFile.length());
251 strcpy(filename, outputFile.c_str());
253 MPI_File_open(MPI_COMM_WORLD, filename, amode, MPI_INFO_NULL, &outMPI);
256 if (pid == 0) { //you are the root process
260 driverMPI(start, end, outMPI, cutoff);
262 if (m->control_pressed) { MPI_File_close(&outMPI); delete distCalculator; return 0; }
265 for(int i = 1; i < processors; i++) {
266 if (m->control_pressed) { MPI_File_close(&outMPI); delete distCalculator; return 0; }
269 MPI_Recv(buf, 4, MPI_CHAR, i, tag, MPI_COMM_WORLD, &status);
271 }else { //you are a child process
273 driverMPI(start, end, outMPI, cutoff);
275 if (m->control_pressed) { MPI_File_close(&outMPI); delete distCalculator; return 0; }
279 //tell parent you are done.
280 MPI_Send(buf, 4, MPI_CHAR, 0, tag, MPI_COMM_WORLD);
283 MPI_File_close(&outMPI);
285 }else { //lower triangle format
286 if (pid == 0) { //you are the root process
291 driverMPI(start, end, outputFile, mySize);
293 if (m->control_pressed) { delete distCalculator; return 0; }
295 int amode=MPI_MODE_APPEND|MPI_MODE_WRONLY|MPI_MODE_CREATE; //
299 //char* filename = new char[outputFile.length()];
300 //memcpy(filename, outputFile.c_str(), outputFile.length());
303 strcpy(filename, outputFile.c_str());
305 MPI_File_open(MPI_COMM_SELF, filename, amode, MPI_INFO_NULL, &outMPI);
309 for(int b = 1; b < processors; b++) {
312 if (m->control_pressed) { MPI_File_close(&outMPI); delete distCalculator; return 0; }
314 MPI_Recv(&fileSize, 1, MPI_LONG, b, tag, MPI_COMM_WORLD, &status);
316 string outTemp = outputFile + toString(b) + ".temp";
318 char* buf = new char[outTemp.length()];
319 memcpy(buf, outTemp.c_str(), outTemp.length());
321 MPI_File_open(MPI_COMM_SELF, buf, MPI_MODE_DELETE_ON_CLOSE|MPI_MODE_RDONLY, MPI_INFO_NULL, &inMPI);
325 while (count < fileSize) {
327 MPI_File_read(inMPI, buf2, 1, MPI_CHAR, &status);
328 MPI_File_write(outMPI, buf2, 1, MPI_CHAR, &status);
332 MPI_File_close(&inMPI); //deleted on close
335 MPI_File_close(&outMPI);
336 }else { //you are a child process
339 driverMPI(start, end, (outputFile + toString(pid) + ".temp"), size);
341 if (m->control_pressed) { delete distCalculator; return 0; }
343 //tell parent you are done.
344 MPI_Send(&size, 1, MPI_LONG, 0, tag, MPI_COMM_WORLD);
347 MPI_Barrier(MPI_COMM_WORLD); //make everyone wait - just in case
350 #if defined (__APPLE__) || (__MACH__) || (linux) || (__linux)
351 //if you don't need to fork anything
353 driver(0, numSeqs, outputFile, cutoff);
354 }else{ //you have multiple processors
356 for (int i = 0; i < processors; i++) {
357 lines.push_back(new linePair());
358 lines[i]->start = int (sqrt(float(i)/float(processors)) * numSeqs);
359 lines[i]->end = int (sqrt(float(i+1)/float(processors)) * numSeqs);
362 createProcesses(outputFile);
364 map<int, int>::iterator it = processIDS.begin();
365 rename((outputFile + toString(it->second) + ".temp").c_str(), outputFile.c_str());
368 //append and remove temp files
369 for (; it != processIDS.end(); it++) {
370 appendFiles((outputFile + toString(it->second) + ".temp"), outputFile);
371 remove((outputFile + toString(it->second) + ".temp").c_str());
376 driver(0, numSeqs, outputFile, cutoff);
380 if (m->control_pressed) { delete distCalculator; remove(outputFile.c_str()); return 0; }
383 MPI_Comm_rank(MPI_COMM_WORLD, &pid);
385 if (pid == 0) { //only one process should output to screen
388 if (output == "square") { convertMatrix(outputFile); }
391 fileHandle.open(outputFile.c_str());
394 if (fileHandle.eof()) { m->mothurOut(outputFile + " is blank. This can result if there are no distances below your cutoff."); m->mothurOutEndLine(); }
397 //append the old column file to the new one
398 if ((oldfastafile != "") && (column != "")) {
399 //we had to rename the column file so we didnt overwrite above, but we want to keep old name
400 if (outputFile == column) {
401 string tempcolumn = column + ".old";
402 appendFiles(tempcolumn, outputFile);
403 remove(tempcolumn.c_str());
405 appendFiles(outputFile, column);
406 remove(outputFile.c_str());
416 if (m->control_pressed) { delete distCalculator; remove(outputFile.c_str()); return 0; }
418 delete distCalculator;
420 m->mothurOutEndLine();
421 m->mothurOut("Output File Name: "); m->mothurOutEndLine();
422 m->mothurOut(outputFile); m->mothurOutEndLine();
423 m->mothurOutEndLine();
424 m->mothurOut("It took " + toString(time(NULL) - startTime) + " to calculate the distances for " + toString(numSeqs) + " sequences."); m->mothurOutEndLine();
428 catch(exception& e) {
429 m->errorOut(e, "DistanceCommand", "execute");
433 /**************************************************************************************************/
434 void DistanceCommand::createProcesses(string filename) {
436 #if defined (__APPLE__) || (__MACH__) || (linux) || (__linux)
440 //loop through and create all the processes you want
441 while (process != processors) {
445 processIDS[lines[process]->end] = pid; //create map from line number to pid so you can append files in correct order later
448 driver(lines[process]->start, lines[process]->end, filename + toString(getpid()) + ".temp", cutoff);
450 }else { m->mothurOut("unable to spawn the necessary processes."); m->mothurOutEndLine(); exit(0); }
453 //force parent to wait until all the processes are done
454 for (map<int, int>::iterator it = processIDS.begin(); it != processIDS.end(); it++) {
455 int temp = it->second;
460 catch(exception& e) {
461 m->errorOut(e, "DistanceCommand", "createProcesses");
466 /**************************************************************************************************/
467 /////// need to fix to work with calcs and sequencedb
468 int DistanceCommand::driver(int startLine, int endLine, string dFileName, float cutoff){
471 int startTime = time(NULL);
474 ofstream outFile(dFileName.c_str(), ios::trunc);
475 outFile.setf(ios::fixed, ios::showpoint);
476 outFile << setprecision(4);
478 if((output == "lt") && startLine == 0){ outFile << alignDB.getNumSeqs() << endl; }
480 for(int i=startLine;i<endLine;i++){
482 string name = alignDB.get(i).getName();
483 if (name.length() < 10) { //pad with spaces to make compatible
484 while (name.length() < 10) { name += " "; }
486 outFile << name << '\t';
488 for(int j=0;j<i;j++){
490 if (m->control_pressed) { outFile.close(); return 0; }
492 //if there was a column file given and we are appending, we don't want to calculate the distances that are already in the column file
493 //the alignDB contains the new sequences and then the old, so if i an oldsequence and j is an old sequence then break out of this loop
494 if ((i >= numNewFasta) && (j >= numNewFasta)) { break; }
496 distCalculator->calcDist(alignDB.get(i), alignDB.get(j));
497 double dist = distCalculator->getDist();
500 if (output == "column") { outFile << alignDB.get(i).getName() << ' ' << alignDB.get(j).getName() << ' ' << dist << endl; }
502 if (output == "lt") { outFile << dist << '\t'; }
504 if (output == "square") { //make a square column you can convert to square phylip
505 outFile << alignDB.get(i).getName() << '\t' << alignDB.get(j).getName() << '\t' << dist << endl;
506 outFile << alignDB.get(j).getName() << '\t' << alignDB.get(i).getName() << '\t' << dist << endl;
511 if (output == "lt") { outFile << endl; }
514 m->mothurOut(toString(i) + "\t" + toString(time(NULL) - startTime)); m->mothurOutEndLine();
518 m->mothurOut(toString(endLine-1) + "\t" + toString(time(NULL) - startTime)); m->mothurOutEndLine();
524 catch(exception& e) {
525 m->errorOut(e, "DistanceCommand", "driver");
530 /**************************************************************************************************/
531 /////// need to fix to work with calcs and sequencedb
532 int DistanceCommand::driverMPI(int startLine, int endLine, MPI_File& outMPI, float cutoff){
535 int startTime = time(NULL);
537 string outputString = "";
539 for(int i=startLine;i<endLine;i++){
541 for(int j=0;j<i;j++){
543 if (m->control_pressed) { return 0; }
545 //if there was a column file given and we are appending, we don't want to calculate the distances that are already in the column file
546 //the alignDB contains the new sequences and then the old, so if i an oldsequence and j is an old sequence then break out of this loop
547 if ((i >= numNewFasta) && (j >= numNewFasta)) { break; }
549 distCalculator->calcDist(alignDB.get(i), alignDB.get(j));
550 double dist = distCalculator->getDist();
553 if (output == "column") { outputString += (alignDB.get(i).getName() + ' ' + alignDB.get(j).getName() + ' ' + toString(dist) + '\n'); }
556 if (output == "square") { //make a square column you can convert to square phylip
557 outputString += (alignDB.get(i).getName() + ' ' + alignDB.get(j).getName() + ' ' + toString(dist) + '\n');
558 outputString += (alignDB.get(j).getName() + ' ' + alignDB.get(i).getName() + ' ' + toString(dist) + '\n');
563 //m->mothurOut(toString(i) + "\t" + toString(time(NULL) - startTime)); m->mothurOutEndLine();
564 cout << i << '\t' << (time(NULL) - startTime) << endl;
568 //send results to parent
569 int length = outputString.length();
571 char* buf = new char[length];
572 memcpy(buf, outputString.c_str(), length);
574 MPI_File_write_shared(outMPI, buf, length, MPI_CHAR, &status);
580 //m->mothurOut(toString(endLine-1) + "\t" + toString(time(NULL) - startTime)); m->mothurOutEndLine();
581 cout << (endLine-1) << '\t' << (time(NULL) - startTime) << endl;
584 catch(exception& e) {
585 m->errorOut(e, "DistanceCommand", "driverMPI");
589 /**************************************************************************************************/
590 /////// need to fix to work with calcs and sequencedb
591 int DistanceCommand::driverMPI(int startLine, int endLine, string file, long& size){
596 int amode=MPI_MODE_CREATE|MPI_MODE_WRONLY;
598 //char* filename = new char[file.length()];
599 //memcpy(filename, file.c_str(), file.length());
602 strcpy(filename, file.c_str());
604 MPI_File_open(MPI_COMM_SELF, filename, amode, MPI_INFO_NULL, &outMPI);
607 int startTime = time(NULL);
609 string outputString = "";
612 if((output == "lt") && startLine == 0){ outputString += toString(alignDB.getNumSeqs()) + "\n"; }
614 for(int i=startLine;i<endLine;i++){
616 string name = alignDB.get(i).getName();
617 if (name.length() < 10) { //pad with spaces to make compatible
618 while (name.length() < 10) { name += " "; }
620 outputString += name + "\t";
622 for(int j=0;j<i;j++){
624 if (m->control_pressed) { return 0; }
626 distCalculator->calcDist(alignDB.get(i), alignDB.get(j));
627 double dist = distCalculator->getDist();
629 if (output == "lt") { outputString += toString(dist) + "\t"; }
632 if (output == "lt") { outputString += "\n"; }
636 //m->mothurOut(toString(i) + "\t" + toString(time(NULL) - startTime)); m->mothurOutEndLine();
637 cout << i << '\t' << (time(NULL) - startTime) << endl;
641 //send results to parent
642 int length = outputString.length();
643 char* buf = new char[length];
644 memcpy(buf, outputString.c_str(), length);
646 MPI_File_write(outMPI, buf, length, MPI_CHAR, &status);
647 size += outputString.length();
652 //m->mothurOut(toString(endLine-1) + "\t" + toString(time(NULL) - startTime)); m->mothurOutEndLine();
653 cout << (endLine-1) << '\t' << (time(NULL) - startTime) << endl;
654 MPI_File_close(&outMPI);
658 catch(exception& e) {
659 m->errorOut(e, "DistanceCommand", "driverMPI");
664 /**************************************************************************************************/
665 int DistanceCommand::convertMatrix(string outputFile) {
668 //sort file by first column so the distances for each row are together
669 string outfile = getRootName(outputFile) + "sorted.dist.temp";
672 #if defined (__APPLE__) || (__MACH__) || (linux) || (__linux)
673 string command = "sort -n " + outputFile + " -o " + outfile;
674 system(command.c_str());
675 #else //sort using windows sort
676 string command = "sort " + outputFile + " /O " + outfile;
677 system(command.c_str());
681 //output to new file distance for each row and save positions in file where new row begins
683 openInputFile(outfile, in);
686 openOutputFile(outputFile, out);
688 out.setf(ios::fixed, ios::floatfield); out.setf(ios::showpoint);
690 out << alignDB.getNumSeqs() << endl;
692 //get first currentRow
693 string first, currentRow, second;
695 map<string, float> rowDists; //take advantage of the fact that maps are already sorted by key
696 map<string, float>::iterator it;
701 rowDists[first] = 0.00; //distance to yourself is 0.0
704 //openInputFile(outfile, in);
707 if (m->control_pressed) { in.close(); remove(outfile.c_str()); out.close(); return 0; }
709 in >> first >> second >> dist; gobble(in);
711 if (first != currentRow) {
713 out << currentRow << '\t'; //print name
716 for (it = rowDists.begin(); it != rowDists.end(); it++) {
717 out << it->second << '\t';
724 rowDists[first] = 0.00;
725 rowDists[second] = dist;
727 rowDists[second] = dist;
731 out << currentRow << '\t'; //print name
734 for (it = rowDists.begin(); it != rowDists.end(); it++) {
735 out << it->second << '\t';
742 remove(outfile.c_str());
747 catch(exception& e) {
748 m->errorOut(e, "DistanceCommand", "convertMatrix");
752 /**************************************************************************************************/
753 int DistanceCommand::convertToLowerTriangle(string outputFile) {
756 //sort file by first column so the distances for each row are together
757 string outfile = getRootName(outputFile) + "sorted.dist.temp";
760 #if defined (__APPLE__) || (__MACH__) || (linux) || (__linux)
761 string command = "sort -n " + outputFile + " -o " + outfile;
762 system(command.c_str());
763 #else //sort using windows sort
764 string command = "sort " + outputFile + " /O " + outfile;
765 system(command.c_str());
769 //output to new file distance for each row and save positions in file where new row begins
771 openInputFile(outfile, in);
774 openOutputFile(outputFile, out);
776 out.setf(ios::fixed, ios::floatfield); out.setf(ios::showpoint);
778 out << alignDB.getNumSeqs() << endl;
780 //get first currentRow
781 string first, currentRow, second;
785 map<string, float> rowDists; //take advantage of the fact that maps are already sorted by key
786 map<string, float>::iterator it;
791 rowDists[first] = 0.00; //distance to yourself is 0.0
794 //openInputFile(outfile, in);
797 if (m->control_pressed) { in.close(); remove(outfile.c_str()); out.close(); return 0; }
799 in >> first >> second >> dist; gobble(in);
801 if (first != currentRow) {
803 out << currentRow << '\t'; //print name
806 for (it = rowDists.begin(); it != rowDists.end(); it++) {
807 if (j >= i) { break; }
808 out << it->second << '\t';
816 rowDists[first] = 0.00;
817 rowDists[second] = dist;
821 rowDists[second] = dist;
825 out << currentRow << '\t'; //print name
828 for (it = rowDists.begin(); it != rowDists.end(); it++) {
829 out << it->second << '\t';
836 remove(outfile.c_str());
841 catch(exception& e) {
842 m->errorOut(e, "DistanceCommand", "convertToLowerTriangle");
846 /**************************************************************************************************/
847 //its okay if the column file does not contain all the names in the fasta file, since some distance may have been above a cutoff,
848 //but no sequences can be in the column file that are not in oldfasta. also, if a distance is above the cutoff given then remove it.
849 //also check to make sure the 2 files have the same alignment length.
850 bool DistanceCommand::sanityCheck() {
854 //make sure the 2 fasta files have the same alignment length
856 openInputFile(fastafile, in);
857 int fastaAlignLength = 0;
860 fastaAlignLength = tempIn.getAligned().length();
865 openInputFile(oldfastafile, in2);
866 int oldfastaAlignLength = 0;
868 Sequence tempIn2(in2);
869 oldfastaAlignLength = tempIn2.getAligned().length();
873 if (fastaAlignLength != oldfastaAlignLength) { m->mothurOut("fasta files do not have the same alignment length."); m->mothurOutEndLine(); return false; }
875 //read fasta file and save names as well as adding them to the alignDB
876 set<string> namesOldFasta;
879 openInputFile(oldfastafile, inFasta);
881 while (!inFasta.eof()) {
882 if (m->control_pressed) { inFasta.close(); return good; }
884 Sequence temp(inFasta);
886 if (temp.getName() != "") {
887 namesOldFasta.insert(temp.getName()); //save name
888 alignDB.push_back(temp); //add to DB
896 //read through the column file checking names and removing distances above the cutoff
898 openInputFile(column, inDist);
901 string outputFile = column + ".temp";
902 openOutputFile(outputFile, outDist);
906 while (!inDist.eof()) {
907 if (m->control_pressed) { inDist.close(); outDist.close(); remove(outputFile.c_str()); return good; }
909 inDist >> name1 >> name2 >> dist; gobble(inDist);
911 //both names are in fasta file and distance is below cutoff
912 if ((namesOldFasta.count(name1) == 0) || (namesOldFasta.count(name2) == 0)) { good = false; break; }
914 if (dist <= cutoff) {
915 outDist << name1 << '\t' << name2 << '\t' << dist << endl;
924 remove(column.c_str());
925 rename(outputFile.c_str(), column.c_str());
927 remove(outputFile.c_str()); //temp file is bad because file mismatch above
931 catch(exception& e) {
932 m->errorOut(e, "DistanceCommand", "appendFiles");
937 /**************************************************************************************************
938 void DistanceCommand::appendFiles(string temp, string filename) {
943 //open output file in append mode
944 openOutputFileAppend(filename, output);
945 openInputFile(temp, input);
947 while(char c = input.get()){
948 if(input.eof()) { break; }
949 else { output << c; }
955 catch(exception& e) {
956 m->errorOut(e, "DistanceCommand", "appendFiles");
960 /**************************************************************************************************/