2 * classifyseqscommand.cpp
5 * Created by westcott on 11/2/09.
6 * Copyright 2009 Schloss Lab. All rights reserved.
10 #include "classifyseqscommand.h"
14 //**********************************************************************************************************************
15 vector<string> ClassifySeqsCommand::setParameters(){
17 CommandParameter ptaxonomy("taxonomy", "InputTypes", "", "", "none", "none", "none",false,true); parameters.push_back(ptaxonomy);
18 CommandParameter ptemplate("reference", "InputTypes", "", "", "none", "none", "none",false,true); parameters.push_back(ptemplate);
19 CommandParameter pfasta("fasta", "InputTypes", "", "", "none", "none", "none",false,true); parameters.push_back(pfasta);
20 CommandParameter pname("name", "InputTypes", "", "", "NameCount", "none", "none",false,false); parameters.push_back(pname);
21 CommandParameter pcount("count", "InputTypes", "", "", "NameCount-CountGroup", "none", "none",false,false); parameters.push_back(pcount);
22 CommandParameter pgroup("group", "InputTypes", "", "", "CountGroup", "none", "none",false,false); parameters.push_back(pgroup);
24 CommandParameter psearch("search", "Multiple", "kmer-blast-suffix-distance", "kmer", "", "", "",false,false); parameters.push_back(psearch);
25 CommandParameter pksize("ksize", "Number", "", "8", "", "", "",false,false); parameters.push_back(pksize);
26 CommandParameter pmethod("method", "Multiple", "bayesian-knn", "bayesian", "", "", "",false,false); parameters.push_back(pmethod);
27 CommandParameter pprocessors("processors", "Number", "", "1", "", "", "",false,false); parameters.push_back(pprocessors);
28 CommandParameter pmatch("match", "Number", "", "1.0", "", "", "",false,false); parameters.push_back(pmatch);
29 CommandParameter pmismatch("mismatch", "Number", "", "-1.0", "", "", "",false,false); parameters.push_back(pmismatch);
30 CommandParameter pgapopen("gapopen", "Number", "", "-2.0", "", "", "",false,false); parameters.push_back(pgapopen);
31 CommandParameter pgapextend("gapextend", "Number", "", "-1.0", "", "", "",false,false); parameters.push_back(pgapextend);
32 //CommandParameter pflip("flip", "Boolean", "", "F", "", "", "",false,false); parameters.push_back(pflip);
33 CommandParameter pcutoff("cutoff", "Number", "", "0", "", "", "",false,true); parameters.push_back(pcutoff);
34 CommandParameter pprobs("probs", "Boolean", "", "T", "", "", "",false,false); parameters.push_back(pprobs);
35 CommandParameter piters("iters", "Number", "", "100", "", "", "",false,true); parameters.push_back(piters);
36 CommandParameter psave("save", "Boolean", "", "F", "", "", "",false,false); parameters.push_back(psave);
37 CommandParameter pnumwanted("numwanted", "Number", "", "10", "", "", "",false,true); parameters.push_back(pnumwanted);
38 CommandParameter pinputdir("inputdir", "String", "", "", "", "", "",false,false); parameters.push_back(pinputdir);
39 CommandParameter poutputdir("outputdir", "String", "", "", "", "", "",false,false); parameters.push_back(poutputdir);
41 vector<string> myArray;
42 for (int i = 0; i < parameters.size(); i++) { myArray.push_back(parameters[i].name); }
46 m->errorOut(e, "ClassifySeqsCommand", "setParameters");
50 //**********************************************************************************************************************
51 string ClassifySeqsCommand::getHelpString(){
53 string helpString = "";
54 helpString += "The classify.seqs command reads a fasta file containing sequences and creates a .taxonomy file and a .tax.summary file.\n";
55 helpString += "The classify.seqs command parameters are reference, fasta, name, group, count, search, ksize, method, taxonomy, processors, match, mismatch, gapopen, gapextend, numwanted and probs.\n";
56 helpString += "The reference, fasta and taxonomy parameters are required. You may enter multiple fasta files by separating their names with dashes. ie. fasta=abrecovery.fasta-amzon.fasta \n";
57 helpString += "The search parameter allows you to specify the method to find most similar template. Your options are: suffix, kmer, blast and distance. The default is kmer.\n";
58 helpString += "The name parameter allows you add a names file with your fasta file, if you enter multiple fasta files, you must enter matching names files for them.\n";
59 helpString += "The group parameter allows you add a group file so you can have the summary totals broken up by group.\n";
60 helpString += "The count parameter allows you add a count file so you can have the summary totals broken up by group.\n";
61 helpString += "The method parameter allows you to specify classification method to use. Your options are: bayesian and knn. The default is bayesian.\n";
62 helpString += "The ksize parameter allows you to specify the kmer size for finding most similar template to candidate. The default is 8.\n";
63 helpString += "The processors parameter allows you to specify the number of processors to use. The default is 1.\n";
65 helpString += "When using MPI, the processors parameter is set to the number of MPI processes running. \n";
67 helpString += "If the save parameter is set to true the reference sequences will be saved in memory, to clear them later you can use the clear.memory command. Default=f.";
68 helpString += "The match parameter allows you to specify the bonus for having the same base. The default is 1.0.\n";
69 helpString += "The mistmatch parameter allows you to specify the penalty for having different bases. The default is -1.0.\n";
70 helpString += "The gapopen parameter allows you to specify the penalty for opening a gap in an alignment. The default is -2.0.\n";
71 helpString += "The gapextend parameter allows you to specify the penalty for extending a gap in an alignment. The default is -1.0.\n";
72 helpString += "The numwanted parameter allows you to specify the number of sequence matches you want with the knn method. The default is 10.\n";
73 helpString += "The cutoff parameter allows you to specify a bootstrap confidence threshold for your taxonomy. The default is 0.\n";
74 helpString += "The probs parameter shuts off the bootstrapping results for the bayesian method. The default is true, meaning you want the bootstrapping to be shown.\n";
75 helpString += "The iters parameter allows you to specify how many iterations to do when calculating the bootstrap confidence score for your taxonomy with the bayesian method. The default is 100.\n";
76 //helpString += "The flip parameter allows you shut off mothur's The default is T.\n";
77 helpString += "The classify.seqs command should be in the following format: \n";
78 helpString += "classify.seqs(reference=yourTemplateFile, fasta=yourFastaFile, method=yourClassificationMethod, search=yourSearchmethod, ksize=yourKmerSize, taxonomy=yourTaxonomyFile, processors=yourProcessors) \n";
79 helpString += "Example classify.seqs(fasta=amazon.fasta, reference=core.filtered, method=knn, search=gotoh, ksize=8, processors=2)\n";
80 helpString += "The .taxonomy file consists of 2 columns: 1 = your sequence name, 2 = the taxonomy for your sequence. \n";
81 helpString += "The .tax.summary is a summary of the different taxonomies represented in your fasta file. \n";
82 helpString += "Note: No spaces between parameter labels (i.e. fasta), '=' and parameters (i.e.yourFastaFile).\n";
86 m->errorOut(e, "ClassifySeqsCommand", "getHelpString");
90 //**********************************************************************************************************************
91 string ClassifySeqsCommand::getOutputFileNameTag(string type, string inputName=""){
93 string outputFileName = "";
94 map<string, vector<string> >::iterator it;
96 //is this a type this command creates
97 it = outputTypes.find(type);
98 if (it == outputTypes.end()) { m->mothurOut("[ERROR]: this command doesn't create a " + type + " output file.\n"); }
100 if (type == "taxonomy") { outputFileName = "taxonomy"; }
101 else if (type == "accnos") { outputFileName = "flip.accnos"; }
102 else if (type == "taxsummary") { outputFileName = "tax.summary"; }
103 else if (type == "matchdist") { outputFileName = "match.dist"; }
104 else { m->mothurOut("[ERROR]: No definition for type " + type + " output file tag.\n"); m->control_pressed = true; }
106 return outputFileName;
108 catch(exception& e) {
109 m->errorOut(e, "ClassifySeqsCommand", "getOutputFileNameTag");
113 //**********************************************************************************************************************
114 ClassifySeqsCommand::ClassifySeqsCommand(){
116 abort = true; calledHelp = true;
118 vector<string> tempOutNames;
119 outputTypes["taxonomy"] = tempOutNames;
120 outputTypes["accnos"] = tempOutNames;
121 outputTypes["taxsummary"] = tempOutNames;
122 outputTypes["matchdist"] = tempOutNames;
124 catch(exception& e) {
125 m->errorOut(e, "ClassifySeqsCommand", "ClassifySeqsCommand");
129 //**********************************************************************************************************************
130 ClassifySeqsCommand::ClassifySeqsCommand(string option) {
132 abort = false; calledHelp = false;
133 rdb = ReferenceDB::getInstance(); hasName = false; hasCount=false;
135 //allow user to run help
136 if(option == "help") { help(); abort = true; calledHelp = true; }
137 else if(option == "citation") { citation(); abort = true; calledHelp = true;}
140 vector<string> myArray = setParameters();
142 OptionParser parser(option);
143 map<string, string> parameters = parser.getParameters();
145 ValidParameters validParameter("classify.seqs");
146 map<string, string>::iterator it;
148 //check to make sure all parameters are valid for command
149 for (it = parameters.begin(); it != parameters.end(); it++) {
150 if (validParameter.isValidParameter(it->first, myArray, it->second) != true) { abort = true; }
153 //initialize outputTypes
154 vector<string> tempOutNames;
155 outputTypes["taxonomy"] = tempOutNames;
156 outputTypes["taxsummary"] = tempOutNames;
157 outputTypes["matchdist"] = tempOutNames;
158 outputTypes["accnos"] = tempOutNames;
160 //if the user changes the output directory command factory will send this info to us in the output parameter
161 outputDir = validParameter.validFile(parameters, "outputdir", false); if (outputDir == "not found"){ outputDir = ""; }
163 //if the user changes the input directory command factory will send this info to us in the output parameter
164 string inputDir = validParameter.validFile(parameters, "inputdir", false);
165 if (inputDir == "not found"){ inputDir = ""; }
168 it = parameters.find("reference");
169 //user has given a template file
170 if(it != parameters.end()){
171 path = m->hasPath(it->second);
172 //if the user has not given a path then, add inputdir. else leave path alone.
173 if (path == "") { parameters["reference"] = inputDir + it->second; }
176 it = parameters.find("taxonomy");
177 //user has given a template file
178 if(it != parameters.end()){
179 path = m->hasPath(it->second);
180 //if the user has not given a path then, add inputdir. else leave path alone.
181 if (path == "") { parameters["taxonomy"] = inputDir + it->second; }
184 it = parameters.find("group");
185 //user has given a template file
186 if(it != parameters.end()){
187 path = m->hasPath(it->second);
188 //if the user has not given a path then, add inputdir. else leave path alone.
189 if (path == "") { parameters["group"] = inputDir + it->second; }
192 it = parameters.find("count");
193 //user has given a template file
194 if(it != parameters.end()){
195 path = m->hasPath(it->second);
196 //if the user has not given a path then, add inputdir. else leave path alone.
197 if (path == "") { parameters["count"] = inputDir + it->second; }
201 fastaFileName = validParameter.validFile(parameters, "fasta", false);
202 if (fastaFileName == "not found") {
203 //if there is a current fasta file, use it
204 string filename = m->getFastaFile();
205 if (filename != "") { fastaFileNames.push_back(filename); m->mothurOut("Using " + filename + " as input file for the fasta parameter."); m->mothurOutEndLine(); }
206 else { m->mothurOut("You have no current fastafile and the fasta parameter is required."); m->mothurOutEndLine(); abort = true; }
209 m->splitAtDash(fastaFileName, fastaFileNames);
211 //go through files and make sure they are good, if not, then disregard them
212 for (int i = 0; i < fastaFileNames.size(); i++) {
215 if (fastaFileNames[i] == "current") {
216 fastaFileNames[i] = m->getFastaFile();
217 if (fastaFileNames[i] != "") { m->mothurOut("Using " + fastaFileNames[i] + " as input file for the fasta parameter where you had given current."); m->mothurOutEndLine(); }
219 m->mothurOut("You have no current fastafile, ignoring current."); m->mothurOutEndLine(); ignore=true;
220 //erase from file list
221 fastaFileNames.erase(fastaFileNames.begin()+i);
228 if (inputDir != "") {
229 string path = m->hasPath(fastaFileNames[i]);
230 //if the user has not given a path then, add inputdir. else leave path alone.
231 if (path == "") { fastaFileNames[i] = inputDir + fastaFileNames[i]; }
237 ableToOpen = m->openInputFile(fastaFileNames[i], in, "noerror");
239 //if you can't open it, try default location
240 if (ableToOpen == 1) {
241 if (m->getDefaultPath() != "") { //default path is set
242 string tryPath = m->getDefaultPath() + m->getSimpleName(fastaFileNames[i]);
243 m->mothurOut("Unable to open " + fastaFileNames[i] + ". Trying default " + tryPath); m->mothurOutEndLine();
245 ableToOpen = m->openInputFile(tryPath, in2, "noerror");
247 fastaFileNames[i] = tryPath;
251 if (ableToOpen == 1) {
252 if (m->getOutputDir() != "") { //default path is set
253 string tryPath = m->getOutputDir() + m->getSimpleName(fastaFileNames[i]);
254 m->mothurOut("Unable to open " + fastaFileNames[i] + ". Trying output directory " + tryPath); m->mothurOutEndLine();
256 ableToOpen = m->openInputFile(tryPath, in2, "noerror");
258 fastaFileNames[i] = tryPath;
264 if (ableToOpen == 1) {
265 m->mothurOut("Unable to open " + fastaFileNames[i] + ". It will be disregarded."); m->mothurOutEndLine();
266 //erase from file list
267 fastaFileNames.erase(fastaFileNames.begin()+i);
270 m->setFastaFile(fastaFileNames[i]);
276 //make sure there is at least one valid file left
277 if (fastaFileNames.size() == 0) { m->mothurOut("no valid files."); m->mothurOutEndLine(); abort = true; }
280 namefile = validParameter.validFile(parameters, "name", false);
281 if (namefile == "not found") { namefile = ""; }
284 m->splitAtDash(namefile, namefileNames);
286 //go through files and make sure they are good, if not, then disregard them
287 for (int i = 0; i < namefileNames.size(); i++) {
289 if (namefileNames[i] == "current") {
290 namefileNames[i] = m->getNameFile();
291 if (namefileNames[i] != "") { m->mothurOut("Using " + namefileNames[i] + " as input file for the name parameter where you had given current."); m->mothurOutEndLine(); }
293 m->mothurOut("You have no current namefile, ignoring current."); m->mothurOutEndLine(); ignore=true;
294 //erase from file list
295 namefileNames.erase(namefileNames.begin()+i);
302 if (inputDir != "") {
303 string path = m->hasPath(namefileNames[i]);
304 //if the user has not given a path then, add inputdir. else leave path alone.
305 if (path == "") { namefileNames[i] = inputDir + namefileNames[i]; }
310 ableToOpen = m->openInputFile(namefileNames[i], in, "noerror");
312 //if you can't open it, try default location
313 if (ableToOpen == 1) {
314 if (m->getDefaultPath() != "") { //default path is set
315 string tryPath = m->getDefaultPath() + m->getSimpleName(namefileNames[i]);
316 m->mothurOut("Unable to open " + namefileNames[i] + ". Trying default " + tryPath); m->mothurOutEndLine();
318 ableToOpen = m->openInputFile(tryPath, in2, "noerror");
320 namefileNames[i] = tryPath;
324 if (ableToOpen == 1) {
325 if (m->getOutputDir() != "") { //default path is set
326 string tryPath = m->getOutputDir() + m->getSimpleName(namefileNames[i]);
327 m->mothurOut("Unable to open " + namefileNames[i] + ". Trying output directory " + tryPath); m->mothurOutEndLine();
329 ableToOpen = m->openInputFile(tryPath, in2, "noerror");
331 namefileNames[i] = tryPath;
336 if (ableToOpen == 1) {
337 m->mothurOut("Unable to open " + namefileNames[i] + ". It will be disregarded."); m->mothurOutEndLine(); abort = true;
338 //erase from file list
339 namefileNames.erase(namefileNames.begin()+i);
342 m->setNameFile(namefileNames[i]);
348 if (namefileNames.size() != 0) { hasName = true; }
350 if (namefile != "") {
351 if (namefileNames.size() != fastaFileNames.size()) { abort = true; m->mothurOut("If you provide a name file, you must have one for each fasta file."); m->mothurOutEndLine(); }
354 //check for required parameters
355 countfile = validParameter.validFile(parameters, "count", false);
356 if (countfile == "not found") {
359 m->splitAtDash(countfile, countfileNames);
361 //go through files and make sure they are good, if not, then disregard them
362 for (int i = 0; i < countfileNames.size(); i++) {
365 if (countfileNames[i] == "current") {
366 countfileNames[i] = m->getCountTableFile();
367 if (countfileNames[i] != "") { m->mothurOut("Using " + countfileNames[i] + " as input file for the count parameter where you had given current."); m->mothurOutEndLine(); }
369 m->mothurOut("You have no current count file, ignoring current."); m->mothurOutEndLine(); ignore=true;
370 //erase from file list
371 countfileNames.erase(countfileNames.begin()+i);
378 if (inputDir != "") {
379 string path = m->hasPath(countfileNames[i]);
380 //if the user has not given a path then, add inputdir. else leave path alone.
381 if (path == "") { countfileNames[i] = inputDir + countfileNames[i]; }
387 ableToOpen = m->openInputFile(countfileNames[i], in, "noerror");
389 //if you can't open it, try default location
390 if (ableToOpen == 1) {
391 if (m->getDefaultPath() != "") { //default path is set
392 string tryPath = m->getDefaultPath() + m->getSimpleName(countfileNames[i]);
393 m->mothurOut("Unable to open " + countfileNames[i] + ". Trying default " + tryPath); m->mothurOutEndLine();
395 ableToOpen = m->openInputFile(tryPath, in2, "noerror");
397 countfileNames[i] = tryPath;
401 if (ableToOpen == 1) {
402 if (m->getOutputDir() != "") { //default path is set
403 string tryPath = m->getOutputDir() + m->getSimpleName(countfileNames[i]);
404 m->mothurOut("Unable to open " + countfileNames[i] + ". Trying output directory " + tryPath); m->mothurOutEndLine();
406 ableToOpen = m->openInputFile(tryPath, in2, "noerror");
408 countfileNames[i] = tryPath;
414 if (ableToOpen == 1) {
415 m->mothurOut("Unable to open " + countfileNames[i] + ". It will be disregarded."); m->mothurOutEndLine();
416 //erase from file list
417 countfileNames.erase(countfileNames.begin()+i);
420 m->setCountTableFile(countfileNames[i]);
426 if (countfileNames.size() != 0) { hasCount = true; if (countfileNames.size() != fastaFileNames.size()) {m->mothurOut("If you provide a count file, you must have one for each fasta file."); m->mothurOutEndLine(); } }
428 //make sure there is at least one valid file left
429 if (hasName && hasCount) { m->mothurOut("[ERROR]: You must enter ONLY ONE of the following: count or name."); m->mothurOutEndLine(); abort = true; }
431 groupfile = validParameter.validFile(parameters, "group", false);
432 if (groupfile == "not found") { groupfile = ""; }
434 m->splitAtDash(groupfile, groupfileNames);
436 //go through files and make sure they are good, if not, then disregard them
437 for (int i = 0; i < groupfileNames.size(); i++) {
438 if (inputDir != "") {
439 string path = m->hasPath(groupfileNames[i]);
440 //if the user has not given a path then, add inputdir. else leave path alone.
441 if (path == "") { groupfileNames[i] = inputDir + groupfileNames[i]; }
446 ableToOpen = m->openInputFile(groupfileNames[i], in, "noerror");
448 //if you can't open it, try default location
449 if (ableToOpen == 1) {
450 if (m->getDefaultPath() != "") { //default path is set
451 string tryPath = m->getDefaultPath() + m->getSimpleName(groupfileNames[i]);
452 m->mothurOut("Unable to open " + groupfileNames[i] + ". Trying default " + tryPath); m->mothurOutEndLine();
454 ableToOpen = m->openInputFile(tryPath, in2, "noerror");
456 groupfileNames[i] = tryPath;
460 if (ableToOpen == 1) {
461 if (m->getOutputDir() != "") { //default path is set
462 string tryPath = m->getOutputDir() + m->getSimpleName(groupfileNames[i]);
463 m->mothurOut("Unable to open " + groupfileNames[i] + ". Trying output directory " + tryPath); m->mothurOutEndLine();
465 ableToOpen = m->openInputFile(tryPath, in2, "noerror");
467 groupfileNames[i] = tryPath;
473 if (ableToOpen == 1) {
474 m->mothurOut("Unable to open " + groupfileNames[i] + ". It will be disregarded."); m->mothurOutEndLine(); groupfileNames[i] = "";
475 //erase from file list
476 groupfileNames.erase(groupfileNames.begin()+i);
479 m->setGroupFile(groupfileNames[i]);
484 if (groupfile != "") {
485 if (groupfileNames.size() != fastaFileNames.size()) { abort = true; m->mothurOut("If you provide a group file, you must have one for each fasta file."); m->mothurOutEndLine(); }
486 if (hasCount) { m->mothurOut("[ERROR]: You must enter ONLY ONE of the following: count or group."); m->mothurOutEndLine(); abort = true; }
488 for (int i = 0; i < fastaFileNames.size(); i++) { groupfileNames.push_back(""); }
491 //check for optional parameter and set defaults
492 // ...at some point should added some additional type checking...
494 temp = validParameter.validFile(parameters, "ksize", false); if (temp == "not found"){ temp = "8"; }
495 m->mothurConvert(temp, kmerSize);
497 temp = validParameter.validFile(parameters, "processors", false); if (temp == "not found"){ temp = m->getProcessors(); }
498 m->setProcessors(temp);
499 m->mothurConvert(temp, processors);
501 temp = validParameter.validFile(parameters, "save", false); if (temp == "not found"){ temp = "f"; }
502 save = m->isTrue(temp);
504 if (save) { //clear out old references
508 //this has to go after save so that if the user sets save=t and provides no reference we abort
509 templateFileName = validParameter.validFile(parameters, "reference", true);
510 if (templateFileName == "not found") {
511 //check for saved reference sequences
512 if (rdb->referenceSeqs.size() != 0) {
513 templateFileName = "saved";
515 m->mothurOut("[ERROR]: You don't have any saved reference sequences and the reference parameter is a required for the classify.seqs command.");
516 m->mothurOutEndLine();
519 }else if (templateFileName == "not open") { abort = true; }
520 else { if (save) { rdb->setSavedReference(templateFileName); } }
522 //this has to go after save so that if the user sets save=t and provides no reference we abort
523 taxonomyFileName = validParameter.validFile(parameters, "taxonomy", true);
524 if (taxonomyFileName == "not found") {
525 //check for saved reference sequences
526 if (rdb->wordGenusProb.size() != 0) {
527 taxonomyFileName = "saved";
529 m->mothurOut("[ERROR]: You don't have any saved taxonomy information and the taxonomy parameter is a required for the classify.seqs command.");
530 m->mothurOutEndLine();
533 }else if (taxonomyFileName == "not open") { abort = true; }
534 else { if (save) { rdb->setSavedTaxonomy(taxonomyFileName); } }
536 search = validParameter.validFile(parameters, "search", false); if (search == "not found"){ search = "kmer"; }
538 method = validParameter.validFile(parameters, "method", false); if (method == "not found"){ method = "bayesian"; }
540 temp = validParameter.validFile(parameters, "match", false); if (temp == "not found"){ temp = "1.0"; }
541 m->mothurConvert(temp, match);
543 temp = validParameter.validFile(parameters, "mismatch", false); if (temp == "not found"){ temp = "-1.0"; }
544 m->mothurConvert(temp, misMatch);
546 temp = validParameter.validFile(parameters, "gapopen", false); if (temp == "not found"){ temp = "-2.0"; }
547 m->mothurConvert(temp, gapOpen);
549 temp = validParameter.validFile(parameters, "gapextend", false); if (temp == "not found"){ temp = "-1.0"; }
550 m->mothurConvert(temp, gapExtend);
552 temp = validParameter.validFile(parameters, "numwanted", false); if (temp == "not found"){ temp = "10"; }
553 m->mothurConvert(temp, numWanted);
555 temp = validParameter.validFile(parameters, "cutoff", false); if (temp == "not found"){ temp = "0"; }
556 m->mothurConvert(temp, cutoff);
558 temp = validParameter.validFile(parameters, "probs", false); if (temp == "not found"){ temp = "true"; }
559 probs = m->isTrue(temp);
561 //temp = validParameter.validFile(parameters, "flip", false); if (temp == "not found"){ temp = "T"; }
562 //flip = m->isTrue(temp);
565 temp = validParameter.validFile(parameters, "iters", false); if (temp == "not found") { temp = "100"; }
566 m->mothurConvert(temp, iters);
569 if ((method == "bayesian") && (search != "kmer")) {
570 m->mothurOut("The bayesian method requires the kmer search." + search + "will be disregarded." ); m->mothurOutEndLine();
576 if (namefileNames.size() == 0){
577 if (fastaFileNames.size() != 0) {
578 vector<string> files; files.push_back(fastaFileNames[fastaFileNames.size()-1]);
579 parser.getNameFile(files);
586 catch(exception& e) {
587 m->errorOut(e, "ClassifySeqsCommand", "ClassifySeqsCommand");
592 //**********************************************************************************************************************
593 ClassifySeqsCommand::~ClassifySeqsCommand(){
594 if (abort == false) {
595 for (int i = 0; i < lines.size(); i++) { delete lines[i]; } lines.clear();
598 //**********************************************************************************************************************
600 int ClassifySeqsCommand::execute(){
602 if (abort == true) { if (calledHelp) { return 0; } return 2; }
604 if(method == "bayesian"){ classify = new Bayesian(taxonomyFileName, templateFileName, search, kmerSize, cutoff, iters, rand(), flip); }
605 else if(method == "knn"){ classify = new Knn(taxonomyFileName, templateFileName, search, kmerSize, gapOpen, gapExtend, match, misMatch, numWanted, rand()); }
607 m->mothurOut(search + " is not a valid method option. I will run the command using bayesian.");
608 m->mothurOutEndLine();
609 classify = new Bayesian(taxonomyFileName, templateFileName, search, kmerSize, cutoff, iters, rand(), flip);
612 if (m->control_pressed) { delete classify; return 0; }
614 for (int s = 0; s < fastaFileNames.size(); s++) {
616 m->mothurOut("Classifying sequences from " + fastaFileNames[s] + " ..." ); m->mothurOutEndLine();
618 string baseTName = taxonomyFileName;
619 if (taxonomyFileName == "saved") {baseTName = rdb->getSavedTaxonomy(); }
621 //set rippedTaxName to
622 string RippedTaxName = "";
623 bool foundDot = false;
624 for (int i = baseTName.length()-1; i >= 0; i--) {
625 if (foundDot && (baseTName[i] != '.')) { RippedTaxName = baseTName[i] + RippedTaxName; }
626 else if (foundDot && (baseTName[i] == '.')) { break; }
627 else if (!foundDot && (baseTName[i] == '.')) { foundDot = true; }
629 if (RippedTaxName != "") { RippedTaxName += "."; }
631 if (outputDir == "") { outputDir += m->hasPath(fastaFileNames[s]); }
632 string newTaxonomyFile = outputDir + m->getRootName(m->getSimpleName(fastaFileNames[s])) + RippedTaxName + getOutputFileNameTag("taxonomy");
633 string newaccnosFile = outputDir + m->getRootName(m->getSimpleName(fastaFileNames[s])) + RippedTaxName + getOutputFileNameTag("accnos");
634 string tempTaxonomyFile = outputDir + m->getRootName(m->getSimpleName(fastaFileNames[s])) + "taxonomy.temp";
635 string taxSummary = outputDir + m->getRootName(m->getSimpleName(fastaFileNames[s])) + RippedTaxName + getOutputFileNameTag("taxsummary");
637 if ((method == "knn") && (search == "distance")) {
638 string DistName = outputDir + m->getRootName(m->getSimpleName(fastaFileNames[s])) + getOutputFileNameTag("matchdist");
639 classify->setDistName(DistName); outputNames.push_back(DistName); outputTypes["matchdist"].push_back(DistName);
642 outputNames.push_back(newTaxonomyFile); outputTypes["taxonomy"].push_back(newTaxonomyFile);
643 outputNames.push_back(newaccnosFile); outputTypes["accnos"].push_back(newaccnosFile);
644 outputNames.push_back(taxSummary); outputTypes["taxsummary"].push_back(taxSummary);
646 int start = time(NULL);
647 int numFastaSeqs = 0;
648 for (int i = 0; i < lines.size(); i++) { delete lines[i]; } lines.clear();
651 int pid, numSeqsPerProcessor;
653 vector<unsigned long long> MPIPos;
656 MPI_Comm_rank(MPI_COMM_WORLD, &pid); //find out who we are
657 MPI_Comm_size(MPI_COMM_WORLD, &processors);
660 MPI_File outMPINewTax;
661 MPI_File outMPITempTax;
664 int outMode=MPI_MODE_CREATE|MPI_MODE_WRONLY;
665 int inMode=MPI_MODE_RDONLY;
667 char outNewTax[1024];
668 strcpy(outNewTax, newTaxonomyFile.c_str());
670 char outTempTax[1024];
671 strcpy(outTempTax, tempTaxonomyFile.c_str());
674 strcpy(outAcc, newaccnosFile.c_str());
676 char inFileName[1024];
677 strcpy(inFileName, fastaFileNames[s].c_str());
679 MPI_File_open(MPI_COMM_WORLD, inFileName, inMode, MPI_INFO_NULL, &inMPI); //comm, filename, mode, info, filepointer
680 MPI_File_open(MPI_COMM_WORLD, outNewTax, outMode, MPI_INFO_NULL, &outMPINewTax);
681 MPI_File_open(MPI_COMM_WORLD, outTempTax, outMode, MPI_INFO_NULL, &outMPITempTax);
682 MPI_File_open(MPI_COMM_WORLD, outAcc, outMode, MPI_INFO_NULL, &outMPIAcc);
684 if (m->control_pressed) { outputTypes.clear(); MPI_File_close(&inMPI); MPI_File_close(&outMPINewTax); MPI_File_close(&outMPIAcc); MPI_File_close(&outMPITempTax); delete classify; return 0; }
686 if (pid == 0) { //you are the root process
688 MPIPos = m->setFilePosFasta(fastaFileNames[s], numFastaSeqs); //fills MPIPos, returns numSeqs
690 //send file positions to all processes
691 for(int i = 1; i < processors; i++) {
692 MPI_Send(&numFastaSeqs, 1, MPI_INT, i, tag, MPI_COMM_WORLD);
693 MPI_Send(&MPIPos[0], (numFastaSeqs+1), MPI_LONG, i, tag, MPI_COMM_WORLD);
696 //figure out how many sequences you have to align
697 numSeqsPerProcessor = numFastaSeqs / processors;
698 int startIndex = pid * numSeqsPerProcessor;
699 if(pid == (processors - 1)){ numSeqsPerProcessor = numFastaSeqs - pid * numSeqsPerProcessor; }
703 driverMPI(startIndex, numSeqsPerProcessor, inMPI, outMPINewTax, outMPITempTax, outMPIAcc, MPIPos);
705 if (m->control_pressed) { outputTypes.clear(); MPI_File_close(&inMPI); MPI_File_close(&outMPINewTax); MPI_File_close(&outMPIAcc); MPI_File_close(&outMPITempTax); for (int i = 0; i < outputNames.size(); i++) { m->mothurRemove(outputNames[i]); } delete classify; return 0; }
707 for (int i = 1; i < processors; i++) {
709 MPI_Recv(&done, 1, MPI_INT, i, tag, MPI_COMM_WORLD, &status);
711 }else{ //you are a child process
712 MPI_Recv(&numFastaSeqs, 1, MPI_INT, 0, tag, MPI_COMM_WORLD, &status);
713 MPIPos.resize(numFastaSeqs+1);
714 MPI_Recv(&MPIPos[0], (numFastaSeqs+1), MPI_LONG, 0, tag, MPI_COMM_WORLD, &status);
716 //figure out how many sequences you have to align
717 numSeqsPerProcessor = numFastaSeqs / processors;
718 int startIndex = pid * numSeqsPerProcessor;
719 if(pid == (processors - 1)){ numSeqsPerProcessor = numFastaSeqs - pid * numSeqsPerProcessor; }
723 driverMPI(startIndex, numSeqsPerProcessor, inMPI, outMPINewTax, outMPITempTax, outMPIAcc, MPIPos);
725 if (m->control_pressed) { outputTypes.clear(); MPI_File_close(&inMPI); MPI_File_close(&outMPINewTax); MPI_File_close(&outMPIAcc); MPI_File_close(&outMPITempTax); delete classify; return 0; }
728 MPI_Send(&done, 1, MPI_INT, 0, tag, MPI_COMM_WORLD);
732 MPI_File_close(&inMPI);
733 MPI_File_close(&outMPINewTax);
734 MPI_File_close(&outMPITempTax);
735 MPI_File_close(&outMPIAcc);
736 MPI_Barrier(MPI_COMM_WORLD); //make everyone wait - just in case
740 vector<unsigned long long> positions;
741 #if defined (__APPLE__) || (__MACH__) || (linux) || (__linux) || (__linux__) || (__unix__) || (__unix)
742 positions = m->divideFile(fastaFileNames[s], processors);
743 for (int i = 0; i < (positions.size()-1); i++) { lines.push_back(new linePair(positions[i], positions[(i+1)])); }
745 if (processors == 1) {
746 lines.push_back(new linePair(0, 1000));
748 positions = m->setFilePosFasta(fastaFileNames[s], numFastaSeqs);
749 if (positions.size() < processors) { processors = positions.size(); }
751 //figure out how many sequences you have to process
752 int numSeqsPerProcessor = numFastaSeqs / processors;
753 for (int i = 0; i < processors; i++) {
754 int startIndex = i * numSeqsPerProcessor;
755 if(i == (processors - 1)){ numSeqsPerProcessor = numFastaSeqs - i * numSeqsPerProcessor; }
756 lines.push_back(new linePair(positions[startIndex], numSeqsPerProcessor));
761 numFastaSeqs = driver(lines[0], newTaxonomyFile, tempTaxonomyFile, newaccnosFile, fastaFileNames[s]);
763 numFastaSeqs = createProcesses(newTaxonomyFile, tempTaxonomyFile, newaccnosFile, fastaFileNames[s]);
767 if (!m->isBlank(newaccnosFile)) { m->mothurOutEndLine(); m->mothurOut("[WARNING]: mothur suspects some of your sequences may be reversed, please check " + newaccnosFile + " for the list of the sequences."); m->mothurOutEndLine(); }
769 m->mothurOutEndLine();
770 m->mothurOut("It took " + toString(time(NULL) - start) + " secs to classify " + toString(numFastaSeqs) + " sequences."); m->mothurOutEndLine(); m->mothurOutEndLine();
776 if (pid == 0) { //this part does not need to be paralellized
778 if(namefile != "") { m->mothurOut("Reading " + namefileNames[s] + "..."); cout.flush(); MPIReadNamesFile(namefileNames[s]); m->mothurOut(" Done."); m->mothurOutEndLine(); }
783 m->mothurOut("Reading " + namefileNames[s] + "..."); cout.flush();
784 nameMap.clear(); //remove old names
785 m->readNames(namefileNames[s], nameMap);
786 m->mothurOut(" Done."); m->mothurOutEndLine();
791 GroupMap* groupMap = NULL;
792 CountTable* ct = NULL;
793 PhyloSummary* taxaSum;
795 ct = new CountTable();
796 ct->readTable(countfileNames[s]);
797 taxaSum = new PhyloSummary(baseTName, ct);
798 taxaSum->summarize(tempTaxonomyFile);
800 if (groupfile != "") { group = groupfileNames[s]; groupMap = new GroupMap(group); groupMap->readMap(); }
802 taxaSum = new PhyloSummary(baseTName, groupMap);
804 if (m->control_pressed) { outputTypes.clear(); if (ct != NULL) { delete ct; } if (groupMap != NULL) { delete groupMap; } delete taxaSum; for (int i = 0; i < outputNames.size(); i++) { m->mothurRemove(outputNames[i]); } delete classify; return 0; }
806 if (namefile == "") { taxaSum->summarize(tempTaxonomyFile); }
809 m->openInputFile(tempTaxonomyFile, in);
811 //read in users taxonomy file and add sequences to tree
815 if (m->control_pressed) { outputTypes.clear(); if (ct != NULL) { delete ct; } if (groupMap != NULL) { delete groupMap; } delete taxaSum; for (int i = 0; i < outputNames.size(); i++) { m->mothurRemove(outputNames[i]); } delete classify; return 0; }
817 in >> name >> taxon; m->gobble(in);
819 itNames = nameMap.find(name);
821 if (itNames == nameMap.end()) {
822 m->mothurOut(name + " is not in your name file please correct."); m->mothurOutEndLine(); exit(1);
824 for (int i = 0; i < itNames->second.size(); i++) {
825 taxaSum->addSeqToTree(itNames->second[i], taxon); //add it as many times as there are identical seqs
827 itNames->second.clear();
828 nameMap.erase(itNames->first);
834 m->mothurRemove(tempTaxonomyFile);
836 if (m->control_pressed) { outputTypes.clear(); if (ct != NULL) { delete ct; } if (groupMap != NULL) { delete groupMap; } for (int i = 0; i < outputNames.size(); i++) { m->mothurRemove(outputNames[i]); } delete classify; return 0; }
840 m->openOutputFile(taxSummary, outTaxTree);
841 taxaSum->print(outTaxTree);
844 //output taxonomy with the unclassified bins added
846 m->openInputFile(newTaxonomyFile, inTax);
849 string unclass = newTaxonomyFile + ".unclass.temp";
850 m->openOutputFile(unclass, outTax);
852 //get maxLevel from phylotree so you know how many 'unclassified's to add
853 int maxLevel = taxaSum->getMaxLevel();
855 //read taxfile - this reading and rewriting is done to preserve the confidence scores.
857 while (!inTax.eof()) {
858 if (m->control_pressed) { outputTypes.clear(); if (ct != NULL) { delete ct; } if (groupMap != NULL) { delete groupMap; } delete taxaSum; for (int i = 0; i < outputNames.size(); i++) { m->mothurRemove(outputNames[i]); } m->mothurRemove(unclass); delete classify; return 0; }
860 inTax >> name >> taxon; m->gobble(inTax);
862 string newTax = addUnclassifieds(taxon, maxLevel);
864 outTax << name << '\t' << newTax << endl;
869 if (ct != NULL) { delete ct; }
870 if (groupMap != NULL) { delete groupMap; } delete taxaSum;
871 m->mothurRemove(newTaxonomyFile);
872 rename(unclass.c_str(), newTaxonomyFile.c_str());
874 m->mothurOutEndLine();
875 m->mothurOut("It took " + toString(time(NULL) - start) + " secs to create the summary file for " + toString(numFastaSeqs) + " sequences."); m->mothurOutEndLine(); m->mothurOutEndLine();
881 m->mothurOutEndLine();
882 m->mothurOut("Output File Names: "); m->mothurOutEndLine();
883 for (int i = 0; i < outputNames.size(); i++) { m->mothurOut(outputNames[i]); m->mothurOutEndLine(); }
884 m->mothurOutEndLine();
887 //set taxonomy file as new current taxonomyfile
889 itTypes = outputTypes.find("taxonomy");
890 if (itTypes != outputTypes.end()) {
891 if ((itTypes->second).size() != 0) { current = (itTypes->second)[0]; m->setTaxonomyFile(current); }
895 itTypes = outputTypes.find("accnos");
896 if (itTypes != outputTypes.end()) {
897 if ((itTypes->second).size() != 0) { current = (itTypes->second)[0]; m->setAccnosFile(current); }
904 catch(exception& e) {
905 m->errorOut(e, "ClassifySeqsCommand", "execute");
910 /**************************************************************************************************/
911 string ClassifySeqsCommand::addUnclassifieds(string tax, int maxlevel) {
913 string newTax, taxon;
916 //keep what you have counting the levels
917 while (tax.find_first_of(';') != -1) {
919 taxon = tax.substr(0,tax.find_first_of(';'))+';';
920 tax = tax.substr(tax.find_first_of(';')+1, tax.length());
925 //add "unclassified" until you reach maxLevel
926 while (level < maxlevel) {
927 newTax += "unclassified;";
933 catch(exception& e) {
934 m->errorOut(e, "ClassifySeqsCommand", "addUnclassifieds");
939 /**************************************************************************************************/
941 int ClassifySeqsCommand::createProcesses(string taxFileName, string tempTaxFile, string accnos, string filename) {
947 #if defined (__APPLE__) || (__MACH__) || (linux) || (__linux) || (__linux__) || (__unix__) || (__unix)
950 //loop through and create all the processes you want
951 while (process != processors) {
955 processIDS.push_back(pid); //create map from line number to pid so you can append files in correct order later
958 num = driver(lines[process], taxFileName + toString(getpid()) + ".temp", tempTaxFile + toString(getpid()) + ".temp", accnos + toString(getpid()) + ".temp", filename);
960 //pass numSeqs to parent
962 string tempFile = filename + toString(getpid()) + ".num.temp";
963 m->openOutputFile(tempFile, out);
969 m->mothurOut("[ERROR]: unable to spawn the necessary processes."); m->mothurOutEndLine();
970 for (int i = 0; i < processIDS.size(); i++) { kill (processIDS[i], SIGINT); }
975 //parent does its part
976 num = driver(lines[0], taxFileName, tempTaxFile, accnos, filename);
978 //force parent to wait until all the processes are done
979 for (int i=0;i<processIDS.size();i++) {
980 int temp = processIDS[i];
984 for (int i = 0; i < processIDS.size(); i++) {
986 string tempFile = filename + toString(processIDS[i]) + ".num.temp";
987 m->openInputFile(tempFile, in);
988 if (!in.eof()) { int tempNum = 0; in >> tempNum; num += tempNum; }
989 in.close(); m->mothurRemove(m->getFullPathName(tempFile));
992 //////////////////////////////////////////////////////////////////////////////////////////////////////
993 //Windows version shared memory, so be careful when passing variables through the alignData struct.
994 //Above fork() will clone, so memory is separate, but that's not the case with windows,
995 //////////////////////////////////////////////////////////////////////////////////////////////////////
997 vector<classifyData*> pDataArray;
998 DWORD dwThreadIdArray[processors-1];
999 HANDLE hThreadArray[processors-1];
1001 //Create processor worker threads.
1002 for( int i=0; i<processors-1; i++ ){
1003 // Allocate memory for thread data.
1004 string extension = "";
1005 if (i != 0) { extension = toString(i) + ".temp"; processIDS.push_back(i); }
1007 classifyData* tempclass = new classifyData((accnos + extension), probs, method, templateFileName, taxonomyFileName, (taxFileName + extension), (tempTaxFile + extension), filename, search, kmerSize, iters, numWanted, m, lines[i]->start, lines[i]->end, match, misMatch, gapOpen, gapExtend, cutoff, i, flip);
1008 pDataArray.push_back(tempclass);
1010 //MySeqSumThreadFunction is in header. It must be global or static to work with the threads.
1011 //default security attributes, thread function name, argument to thread function, use default creation flags, returns the thread identifier
1012 hThreadArray[i] = CreateThread(NULL, 0, MyClassThreadFunction, pDataArray[i], 0, &dwThreadIdArray[i]);
1016 //parent does its part
1017 num = driver(lines[processors-1], taxFileName + toString(processors-1) + ".temp", tempTaxFile + toString(processors-1) + ".temp", accnos + toString(processors-1) + ".temp", filename);
1018 processIDS.push_back((processors-1));
1020 //Wait until all threads have terminated.
1021 WaitForMultipleObjects(processors-1, hThreadArray, TRUE, INFINITE);
1023 //Close all thread handles and free memory allocations.
1024 for(int i=0; i < pDataArray.size(); i++){
1025 num += pDataArray[i]->count;
1026 CloseHandle(hThreadArray[i]);
1027 delete pDataArray[i];
1031 vector<string> nonBlankAccnosFiles;
1032 if (!(m->isBlank(accnos))) { nonBlankAccnosFiles.push_back(accnos); }
1033 else { m->mothurRemove(accnos); } //remove so other files can be renamed to it
1035 for(int i=0;i<processIDS.size();i++){
1036 m->appendFiles((taxFileName + toString(processIDS[i]) + ".temp"), taxFileName);
1037 m->appendFiles((tempTaxFile + toString(processIDS[i]) + ".temp"), tempTaxFile);
1038 if (!(m->isBlank(accnos + toString(processIDS[i]) + ".temp"))) {
1039 nonBlankAccnosFiles.push_back(accnos + toString(processIDS[i]) + ".temp");
1040 }else { m->mothurRemove((accnos + toString(processIDS[i]) + ".temp")); }
1042 m->mothurRemove((m->getFullPathName(taxFileName) + toString(processIDS[i]) + ".temp"));
1043 m->mothurRemove((m->getFullPathName(tempTaxFile) + toString(processIDS[i]) + ".temp"));
1046 //append accnos files
1047 if (nonBlankAccnosFiles.size() != 0) {
1048 rename(nonBlankAccnosFiles[0].c_str(), accnos.c_str());
1050 for (int h=1; h < nonBlankAccnosFiles.size(); h++) {
1051 m->appendFiles(nonBlankAccnosFiles[h], accnos);
1052 m->mothurRemove(nonBlankAccnosFiles[h]);
1054 }else { //recreate the accnosfile if needed
1056 m->openOutputFile(accnos, out);
1063 catch(exception& e) {
1064 m->errorOut(e, "ClassifySeqsCommand", "createProcesses");
1068 //**********************************************************************************************************************
1070 int ClassifySeqsCommand::driver(linePair* filePos, string taxFName, string tempTFName, string accnos, string filename){
1073 m->openOutputFile(taxFName, outTax);
1075 ofstream outTaxSimple;
1076 m->openOutputFile(tempTFName, outTaxSimple);
1079 m->openOutputFile(accnos, outAcc);
1082 m->openInputFile(filename, inFASTA);
1086 inFASTA.seekg(filePos->start);
1092 if (m->control_pressed) {
1095 outTaxSimple.close();
1096 outAcc.close(); return 0; }
1098 Sequence* candidateSeq = new Sequence(inFASTA); m->gobble(inFASTA);
1100 if (candidateSeq->getName() != "") {
1102 taxonomy = classify->getTaxonomy(candidateSeq);
1104 if (m->control_pressed) { delete candidateSeq; return 0; }
1106 if (taxonomy == "unknown;") { m->mothurOut("[WARNING]: " + candidateSeq->getName() + " could not be classified. You can use the remove.lineage command with taxon=unknown; to remove such sequences."); m->mothurOutEndLine(); }
1108 //output confidence scores or not
1110 outTax << candidateSeq->getName() << '\t' << taxonomy << endl;
1112 outTax << candidateSeq->getName() << '\t' << classify->getSimpleTax() << endl;
1115 if (classify->getFlipped()) { outAcc << candidateSeq->getName() << endl; }
1117 outTaxSimple << candidateSeq->getName() << '\t' << classify->getSimpleTax() << endl;
1121 delete candidateSeq;
1123 #if defined (__APPLE__) || (__MACH__) || (linux) || (__linux) || (__linux__) || (__unix__) || (__unix)
1124 unsigned long long pos = inFASTA.tellg();
1125 if ((pos == -1) || (pos >= filePos->end)) { break; }
1127 if (inFASTA.eof()) { break; }
1131 if((count) % 100 == 0){ m->mothurOut("Processing sequence: " + toString(count)); m->mothurOutEndLine(); }
1135 if((count) % 100 != 0){ m->mothurOut("Processing sequence: " + toString(count)); m->mothurOutEndLine(); }
1139 outTaxSimple.close();
1144 catch(exception& e) {
1145 m->errorOut(e, "ClassifySeqsCommand", "driver");
1149 //**********************************************************************************************************************
1151 int ClassifySeqsCommand::driverMPI(int start, int num, MPI_File& inMPI, MPI_File& newFile, MPI_File& tempFile, MPI_File& accFile, vector<unsigned long long>& MPIPos){
1153 MPI_Status statusNew;
1154 MPI_Status statusTemp;
1155 MPI_Status statusAcc;
1159 MPI_Comm_rank(MPI_COMM_WORLD, &pid); //find out who we are
1162 string outputString;
1164 for(int i=0;i<num;i++){
1166 if (m->control_pressed) { return 0; }
1168 //read next sequence
1169 int length = MPIPos[start+i+1] - MPIPos[start+i];
1170 char* buf4 = new char[length];
1171 MPI_File_read_at(inMPI, MPIPos[start+i], buf4, length, MPI_CHAR, &status);
1173 string tempBuf = buf4;
1174 if (tempBuf.length() > length) { tempBuf = tempBuf.substr(0, length); }
1175 istringstream iss (tempBuf,istringstream::in);
1178 Sequence* candidateSeq = new Sequence(iss);
1180 if (candidateSeq->getName() != "") {
1181 taxonomy = classify->getTaxonomy(candidateSeq);
1183 if (taxonomy == "unknown;") { m->mothurOut("[WARNING]: " + candidateSeq->getName() + " could not be classified. You can use the remove.lineage command with taxon=unknown; to remove such sequences."); m->mothurOutEndLine(); }
1185 //output confidence scores or not
1187 outputString = candidateSeq->getName() + "\t" + taxonomy + "\n";
1189 outputString = candidateSeq->getName() + "\t" + classify->getSimpleTax() + "\n";
1192 int length = outputString.length();
1193 char* buf2 = new char[length];
1194 memcpy(buf2, outputString.c_str(), length);
1196 MPI_File_write_shared(newFile, buf2, length, MPI_CHAR, &statusNew);
1199 outputString = candidateSeq->getName() + "\t" + classify->getSimpleTax() + "\n";
1200 length = outputString.length();
1201 char* buf = new char[length];
1202 memcpy(buf, outputString.c_str(), length);
1204 MPI_File_write_shared(tempFile, buf, length, MPI_CHAR, &statusTemp);
1207 if (classify->getFlipped()) {
1208 outputString = candidateSeq->getName() + "\n";
1209 length = outputString.length();
1210 char* buf3 = new char[length];
1211 memcpy(buf3, outputString.c_str(), length);
1213 MPI_File_write_shared(accFile, buf3, length, MPI_CHAR, &statusAcc);
1218 delete candidateSeq;
1220 if((i+1) % 100 == 0){ cout << "Classifying sequence " << (i+1) << endl; }
1223 if(num % 100 != 0){ cout << "Classifying sequence " << (num) << endl; }
1228 catch(exception& e) {
1229 m->errorOut(e, "ClassifySeqsCommand", "driverMPI");
1234 //**********************************************************************************************************************
1235 int ClassifySeqsCommand::MPIReadNamesFile(string nameFilename){
1238 nameMap.clear(); //remove old names
1244 //char* inFileName = new char[nameFilename.length()];
1245 //memcpy(inFileName, nameFilename.c_str(), nameFilename.length());
1247 char inFileName[1024];
1248 strcpy(inFileName, nameFilename.c_str());
1250 MPI_File_open(MPI_COMM_WORLD, inFileName, MPI_MODE_RDONLY, MPI_INFO_NULL, &inMPI);
1251 MPI_File_get_size(inMPI, &size);
1252 //delete inFileName;
1254 char* buffer = new char[size];
1255 MPI_File_read(inMPI, buffer, size, MPI_CHAR, &status);
1257 string tempBuf = buffer;
1258 if (tempBuf.length() > size) { tempBuf = tempBuf.substr(0, size); }
1259 istringstream iss (tempBuf,istringstream::in);
1262 string firstCol, secondCol;
1264 iss >> firstCol >> secondCol; m->gobble(iss);
1266 vector<string> temp;
1267 m->splitAtComma(secondCol, temp);
1269 nameMap[firstCol] = temp;
1272 MPI_File_close(&inMPI);
1276 catch(exception& e) {
1277 m->errorOut(e, "ClassifySeqsCommand", "MPIReadNamesFile");
1282 /**************************************************************************************************/