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,true); parameters.push_back(ptaxonomy);
18 CommandParameter ptemplate("reference", "InputTypes", "", "", "none", "none", "none","",false,true,true); parameters.push_back(ptemplate);
19 CommandParameter pfasta("fasta", "InputTypes", "", "", "none", "none", "none","taxonomy",false,true,true); parameters.push_back(pfasta);
20 CommandParameter pname("name", "InputTypes", "", "", "NameCount", "none", "none","",false,false,true); parameters.push_back(pname);
21 CommandParameter pcount("count", "InputTypes", "", "", "NameCount-CountGroup", "none", "none","",false,false,true); parameters.push_back(pcount);
22 CommandParameter pgroup("group", "InputTypes", "", "", "CountGroup", "none", "none","",false,false,true); parameters.push_back(pgroup);
24 CommandParameter psearch("search", "Multiple", "kmer-blast-suffix-distance-align", "kmer", "", "", "","",false,false); parameters.push_back(psearch);
25 CommandParameter pksize("ksize", "Number", "", "8", "", "", "","",false,false); parameters.push_back(pksize);
26 CommandParameter pmethod("method", "Multiple", "wang-knn-zap", "wang", "", "", "","",false,false); parameters.push_back(pmethod);
27 CommandParameter pprocessors("processors", "Number", "", "1", "", "", "","",false,false,true); 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 pcutoff("cutoff", "Number", "", "0", "", "", "","",false,true); parameters.push_back(pcutoff);
33 CommandParameter pprobs("probs", "Boolean", "", "T", "", "", "","",false,false); parameters.push_back(pprobs);
34 CommandParameter piters("iters", "Number", "", "100", "", "", "","",false,true); parameters.push_back(piters);
35 CommandParameter psave("save", "Boolean", "", "F", "", "", "","",false,false); parameters.push_back(psave);
36 CommandParameter pshortcuts("shortcuts", "Boolean", "", "T", "", "", "","",false,false); parameters.push_back(pshortcuts);
37 CommandParameter prelabund("relabund", "Boolean", "", "F", "", "", "","",false,false); parameters.push_back(prelabund);
38 CommandParameter pnumwanted("numwanted", "Number", "", "10", "", "", "","",false,true); parameters.push_back(pnumwanted);
39 CommandParameter pinputdir("inputdir", "String", "", "", "", "", "","",false,false); parameters.push_back(pinputdir);
40 CommandParameter poutputdir("outputdir", "String", "", "", "", "", "","",false,false); parameters.push_back(poutputdir);
42 vector<string> myArray;
43 for (int i = 0; i < parameters.size(); i++) { myArray.push_back(parameters[i].name); }
47 m->errorOut(e, "ClassifySeqsCommand", "setParameters");
51 //**********************************************************************************************************************
52 string ClassifySeqsCommand::getHelpString(){
54 string helpString = "";
55 helpString += "The classify.seqs command reads a fasta file containing sequences and creates a .taxonomy file and a .tax.summary file.\n";
56 helpString += "The classify.seqs command parameters are reference, fasta, name, group, count, search, ksize, method, taxonomy, processors, match, mismatch, gapopen, gapextend, numwanted, relabund and probs.\n";
57 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";
58 helpString += "The search parameter allows you to specify the method to find most similar template. Your options are: suffix, kmer, blast, align and distance. The default is kmer.\n";
59 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";
60 helpString += "The group parameter allows you add a group file so you can have the summary totals broken up by group.\n";
61 helpString += "The count parameter allows you add a count file so you can have the summary totals broken up by group.\n";
62 helpString += "The method parameter allows you to specify classification method to use. Your options are: wang, knn and zap. The default is wang.\n";
63 helpString += "The ksize parameter allows you to specify the kmer size for finding most similar template to candidate. The default is 8.\n";
64 helpString += "The processors parameter allows you to specify the number of processors to use. The default is 1.\n";
66 helpString += "When using MPI, the processors parameter is set to the number of MPI processes running. \n";
68 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.";
69 helpString += "The match parameter allows you to specify the bonus for having the same base. The default is 1.0.\n";
70 helpString += "The mistmatch parameter allows you to specify the penalty for having different bases. The default is -1.0.\n";
71 helpString += "The gapopen parameter allows you to specify the penalty for opening a gap in an alignment. The default is -2.0.\n";
72 helpString += "The gapextend parameter allows you to specify the penalty for extending a gap in an alignment. The default is -1.0.\n";
73 helpString += "The numwanted parameter allows you to specify the number of sequence matches you want with the knn method. The default is 10.\n";
74 helpString += "The cutoff parameter allows you to specify a bootstrap confidence threshold for your taxonomy. The default is 0.\n";
75 helpString += "The probs parameter shuts off the bootstrapping results for the wang and zap method. The default is true, meaning you want the bootstrapping to be shown.\n";
76 helpString += "The relabund parameter allows you to indicate you want the summary file values to be relative abundances rather than raw abundances. Default=F. \n";
77 helpString += "The iters parameter allows you to specify how many iterations to do when calculating the bootstrap confidence score for your taxonomy with the wang method. The default is 100.\n";
78 //helpString += "The flip parameter allows you shut off mothur's The default is T.\n";
79 helpString += "The classify.seqs command should be in the following format: \n";
80 helpString += "classify.seqs(reference=yourTemplateFile, fasta=yourFastaFile, method=yourClassificationMethod, search=yourSearchmethod, ksize=yourKmerSize, taxonomy=yourTaxonomyFile, processors=yourProcessors) \n";
81 helpString += "Example classify.seqs(fasta=amazon.fasta, reference=core.filtered, method=knn, search=gotoh, ksize=8, processors=2)\n";
82 helpString += "The .taxonomy file consists of 2 columns: 1 = your sequence name, 2 = the taxonomy for your sequence. \n";
83 helpString += "The .tax.summary is a summary of the different taxonomies represented in your fasta file. \n";
84 helpString += "Note: No spaces between parameter labels (i.e. fasta), '=' and parameters (i.e.yourFastaFile).\n";
88 m->errorOut(e, "ClassifySeqsCommand", "getHelpString");
92 //**********************************************************************************************************************
93 string ClassifySeqsCommand::getOutputPattern(string type) {
97 if (type == "taxonomy") { pattern = "[filename],[tag],[tag2],taxonomy"; }
98 else if (type == "taxsummary") { pattern = "[filename],[tag],[tag2],tax.summary"; }
99 else if (type == "accnos") { pattern = "[filename],[tag],[tag2],flip.accnos"; }
100 else if (type == "matchdist") { pattern = "[filename],[tag],[tag2],match.dist"; }
101 else { m->mothurOut("[ERROR]: No definition for type " + type + " output pattern.\n"); m->control_pressed = true; }
105 catch(exception& e) {
106 m->errorOut(e, "ClassifySeqsCommand", "getOutputPattern");
110 //**********************************************************************************************************************
111 ClassifySeqsCommand::ClassifySeqsCommand(){
113 abort = true; calledHelp = true;
115 vector<string> tempOutNames;
116 outputTypes["taxonomy"] = tempOutNames;
117 outputTypes["accnos"] = tempOutNames;
118 outputTypes["taxsummary"] = tempOutNames;
119 outputTypes["matchdist"] = tempOutNames;
121 catch(exception& e) {
122 m->errorOut(e, "ClassifySeqsCommand", "ClassifySeqsCommand");
126 //**********************************************************************************************************************
127 ClassifySeqsCommand::ClassifySeqsCommand(string option) {
129 abort = false; calledHelp = false;
130 rdb = ReferenceDB::getInstance(); hasName = false; hasCount=false;
132 //allow user to run help
133 if(option == "help") { help(); abort = true; calledHelp = true; }
134 else if(option == "citation") { citation(); abort = true; calledHelp = true;}
137 vector<string> myArray = setParameters();
139 OptionParser parser(option);
140 map<string, string> parameters = parser.getParameters();
142 ValidParameters validParameter("classify.seqs");
143 map<string, string>::iterator it;
145 //check to make sure all parameters are valid for command
146 for (it = parameters.begin(); it != parameters.end(); it++) {
147 if (validParameter.isValidParameter(it->first, myArray, it->second) != true) { abort = true; }
150 //initialize outputTypes
151 vector<string> tempOutNames;
152 outputTypes["taxonomy"] = tempOutNames;
153 outputTypes["taxsummary"] = tempOutNames;
154 outputTypes["matchdist"] = tempOutNames;
155 outputTypes["accnos"] = tempOutNames;
157 //if the user changes the output directory command factory will send this info to us in the output parameter
158 outputDir = validParameter.validFile(parameters, "outputdir", false); if (outputDir == "not found"){ outputDir = ""; }
160 //if the user changes the input directory command factory will send this info to us in the output parameter
161 string inputDir = validParameter.validFile(parameters, "inputdir", false);
162 if (inputDir == "not found"){ inputDir = ""; }
165 it = parameters.find("reference");
166 //user has given a template file
167 if(it != parameters.end()){
168 path = m->hasPath(it->second);
169 //if the user has not given a path then, add inputdir. else leave path alone.
170 if (path == "") { parameters["reference"] = inputDir + it->second; }
173 it = parameters.find("taxonomy");
174 //user has given a template file
175 if(it != parameters.end()){
176 path = m->hasPath(it->second);
177 //if the user has not given a path then, add inputdir. else leave path alone.
178 if (path == "") { parameters["taxonomy"] = inputDir + it->second; }
182 fastaFileName = validParameter.validFile(parameters, "fasta", false);
183 if (fastaFileName == "not found") {
184 //if there is a current fasta file, use it
185 string filename = m->getFastaFile();
186 if (filename != "") { fastaFileNames.push_back(filename); m->mothurOut("Using " + filename + " as input file for the fasta parameter."); m->mothurOutEndLine(); }
187 else { m->mothurOut("You have no current fastafile and the fasta parameter is required."); m->mothurOutEndLine(); abort = true; }
190 m->splitAtDash(fastaFileName, fastaFileNames);
192 //go through files and make sure they are good, if not, then disregard them
193 for (int i = 0; i < fastaFileNames.size(); i++) {
196 if (fastaFileNames[i] == "current") {
197 fastaFileNames[i] = m->getFastaFile();
198 if (fastaFileNames[i] != "") { m->mothurOut("Using " + fastaFileNames[i] + " as input file for the fasta parameter where you had given current."); m->mothurOutEndLine(); }
200 m->mothurOut("You have no current fastafile, ignoring current."); m->mothurOutEndLine(); ignore=true;
201 //erase from file list
202 fastaFileNames.erase(fastaFileNames.begin()+i);
209 if (inputDir != "") {
210 string path = m->hasPath(fastaFileNames[i]);
211 //if the user has not given a path then, add inputdir. else leave path alone.
212 if (path == "") { fastaFileNames[i] = inputDir + fastaFileNames[i]; }
218 ableToOpen = m->openInputFile(fastaFileNames[i], in, "noerror");
220 //if you can't open it, try default location
221 if (ableToOpen == 1) {
222 if (m->getDefaultPath() != "") { //default path is set
223 string tryPath = m->getDefaultPath() + m->getSimpleName(fastaFileNames[i]);
224 m->mothurOut("Unable to open " + fastaFileNames[i] + ". Trying default " + tryPath); m->mothurOutEndLine();
226 ableToOpen = m->openInputFile(tryPath, in2, "noerror");
228 fastaFileNames[i] = tryPath;
232 if (ableToOpen == 1) {
233 if (m->getOutputDir() != "") { //default path is set
234 string tryPath = m->getOutputDir() + m->getSimpleName(fastaFileNames[i]);
235 m->mothurOut("Unable to open " + fastaFileNames[i] + ". Trying output directory " + tryPath); m->mothurOutEndLine();
237 ableToOpen = m->openInputFile(tryPath, in2, "noerror");
239 fastaFileNames[i] = tryPath;
245 if (ableToOpen == 1) {
246 m->mothurOut("Unable to open " + fastaFileNames[i] + ". It will be disregarded."); m->mothurOutEndLine();
247 //erase from file list
248 fastaFileNames.erase(fastaFileNames.begin()+i);
251 m->setFastaFile(fastaFileNames[i]);
257 //make sure there is at least one valid file left
258 if (fastaFileNames.size() == 0) { m->mothurOut("no valid files."); m->mothurOutEndLine(); abort = true; }
261 namefile = validParameter.validFile(parameters, "name", false);
262 if (namefile == "not found") { namefile = ""; }
264 m->splitAtDash(namefile, namefileNames);
266 //go through files and make sure they are good, if not, then disregard them
267 for (int i = 0; i < namefileNames.size(); i++) {
269 if (namefileNames[i] == "current") {
270 namefileNames[i] = m->getNameFile();
271 if (namefileNames[i] != "") { m->mothurOut("Using " + namefileNames[i] + " as input file for the name parameter where you had given current."); m->mothurOutEndLine(); }
273 m->mothurOut("You have no current namefile, ignoring current."); m->mothurOutEndLine(); ignore=true;
274 //erase from file list
275 namefileNames.erase(namefileNames.begin()+i);
282 if (inputDir != "") {
283 string path = m->hasPath(namefileNames[i]);
284 //if the user has not given a path then, add inputdir. else leave path alone.
285 if (path == "") { namefileNames[i] = inputDir + namefileNames[i]; }
290 ableToOpen = m->openInputFile(namefileNames[i], in, "noerror");
292 //if you can't open it, try default location
293 if (ableToOpen == 1) {
294 if (m->getDefaultPath() != "") { //default path is set
295 string tryPath = m->getDefaultPath() + m->getSimpleName(namefileNames[i]);
296 m->mothurOut("Unable to open " + namefileNames[i] + ". Trying default " + tryPath); m->mothurOutEndLine();
298 ableToOpen = m->openInputFile(tryPath, in2, "noerror");
300 namefileNames[i] = tryPath;
304 if (ableToOpen == 1) {
305 if (m->getOutputDir() != "") { //default path is set
306 string tryPath = m->getOutputDir() + m->getSimpleName(namefileNames[i]);
307 m->mothurOut("Unable to open " + namefileNames[i] + ". Trying output directory " + tryPath); m->mothurOutEndLine();
309 ableToOpen = m->openInputFile(tryPath, in2, "noerror");
311 namefileNames[i] = tryPath;
316 if (ableToOpen == 1) {
317 m->mothurOut("Unable to open " + namefileNames[i] + ". It will be disregarded."); m->mothurOutEndLine(); abort = true;
318 //erase from file list
319 namefileNames.erase(namefileNames.begin()+i);
322 m->setNameFile(namefileNames[i]);
328 if (namefileNames.size() != 0) { hasName = true; }
330 if (namefile != "") {
331 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(); }
334 //check for required parameters
335 countfile = validParameter.validFile(parameters, "count", false);
336 if (countfile == "not found") {
339 m->splitAtDash(countfile, countfileNames);
341 //go through files and make sure they are good, if not, then disregard them
342 for (int i = 0; i < countfileNames.size(); i++) {
345 if (countfileNames[i] == "current") {
346 countfileNames[i] = m->getCountTableFile();
347 if (countfileNames[i] != "") { m->mothurOut("Using " + countfileNames[i] + " as input file for the count parameter where you had given current."); m->mothurOutEndLine(); }
349 m->mothurOut("You have no current count file, ignoring current."); m->mothurOutEndLine(); ignore=true;
350 //erase from file list
351 countfileNames.erase(countfileNames.begin()+i);
358 if (inputDir != "") {
359 string path = m->hasPath(countfileNames[i]);
360 //if the user has not given a path then, add inputdir. else leave path alone.
361 if (path == "") { countfileNames[i] = inputDir + countfileNames[i]; }
367 ableToOpen = m->openInputFile(countfileNames[i], in, "noerror");
369 //if you can't open it, try default location
370 if (ableToOpen == 1) {
371 if (m->getDefaultPath() != "") { //default path is set
372 string tryPath = m->getDefaultPath() + m->getSimpleName(countfileNames[i]);
373 m->mothurOut("Unable to open " + countfileNames[i] + ". Trying default " + tryPath); m->mothurOutEndLine();
375 ableToOpen = m->openInputFile(tryPath, in2, "noerror");
377 countfileNames[i] = tryPath;
381 if (ableToOpen == 1) {
382 if (m->getOutputDir() != "") { //default path is set
383 string tryPath = m->getOutputDir() + m->getSimpleName(countfileNames[i]);
384 m->mothurOut("Unable to open " + countfileNames[i] + ". Trying output directory " + tryPath); m->mothurOutEndLine();
386 ableToOpen = m->openInputFile(tryPath, in2, "noerror");
388 countfileNames[i] = tryPath;
394 if (ableToOpen == 1) {
395 m->mothurOut("Unable to open " + countfileNames[i] + ". It will be disregarded."); m->mothurOutEndLine();
396 //erase from file list
397 countfileNames.erase(countfileNames.begin()+i);
400 m->setCountTableFile(countfileNames[i]);
406 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(); } }
408 //make sure there is at least one valid file left
409 if (hasName && hasCount) { m->mothurOut("[ERROR]: You must enter ONLY ONE of the following: count or name."); m->mothurOutEndLine(); abort = true; }
411 groupfile = validParameter.validFile(parameters, "group", false);
412 if (groupfile == "not found") { groupfile = ""; }
414 m->splitAtDash(groupfile, groupfileNames);
416 //go through files and make sure they are good, if not, then disregard them
417 for (int i = 0; i < groupfileNames.size(); i++) {
420 if (groupfileNames[i] == "current") {
421 groupfileNames[i] = m->getGroupFile();
422 if (groupfileNames[i] != "") { m->mothurOut("Using " + groupfileNames[i] + " as input file for the group parameter where you had given current."); m->mothurOutEndLine(); }
424 m->mothurOut("You have no current group file, ignoring current."); m->mothurOutEndLine(); ignore=true;
425 //erase from file list
426 groupfileNames.erase(groupfileNames.begin()+i);
433 if (inputDir != "") {
434 string path = m->hasPath(groupfileNames[i]);
435 cout << path << '\t' << inputDir << endl;
436 //if the user has not given a path then, add inputdir. else leave path alone.
437 if (path == "") { groupfileNames[i] = inputDir + groupfileNames[i]; }
443 ableToOpen = m->openInputFile(groupfileNames[i], in, "noerror");
445 //if you can't open it, try default location
446 if (ableToOpen == 1) {
447 if (m->getDefaultPath() != "") { //default path is set
448 string tryPath = m->getDefaultPath() + m->getSimpleName(groupfileNames[i]);
449 m->mothurOut("Unable to open " + groupfileNames[i] + ". Trying default " + tryPath); m->mothurOutEndLine();
451 ableToOpen = m->openInputFile(tryPath, in2, "noerror");
453 groupfileNames[i] = tryPath;
457 if (ableToOpen == 1) {
458 if (m->getOutputDir() != "") { //default path is set
459 string tryPath = m->getOutputDir() + m->getSimpleName(groupfileNames[i]);
460 m->mothurOut("Unable to open " + groupfileNames[i] + ". Trying output directory " + tryPath); m->mothurOutEndLine();
462 ableToOpen = m->openInputFile(tryPath, in2, "noerror");
464 groupfileNames[i] = tryPath;
470 if (ableToOpen == 1) {
471 m->mothurOut("Unable to open " + groupfileNames[i] + ". It will be disregarded."); m->mothurOutEndLine();
472 //erase from file list
473 groupfileNames.erase(groupfileNames.begin()+i);
476 m->setGroupFile(groupfileNames[i]);
483 if (groupfile != "") {
484 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(); }
485 if (hasCount) { m->mothurOut("[ERROR]: You must enter ONLY ONE of the following: count or group."); m->mothurOutEndLine(); abort = true; }
487 for (int i = 0; i < fastaFileNames.size(); i++) { groupfileNames.push_back(""); }
490 //check for optional parameter and set defaults
491 // ...at some point should added some additional type checking...
493 temp = validParameter.validFile(parameters, "processors", false); if (temp == "not found"){ temp = m->getProcessors(); }
494 m->setProcessors(temp);
495 m->mothurConvert(temp, processors);
497 temp = validParameter.validFile(parameters, "save", false); if (temp == "not found"){ temp = "f"; }
498 save = m->isTrue(temp);
500 if (save) { //clear out old references
504 //this has to go after save so that if the user sets save=t and provides no reference we abort
505 templateFileName = validParameter.validFile(parameters, "reference", true);
506 if (templateFileName == "not found") {
507 //check for saved reference sequences
508 if (rdb->referenceSeqs.size() != 0) {
509 templateFileName = "saved";
511 m->mothurOut("[ERROR]: You don't have any saved reference sequences and the reference parameter is a required for the classify.seqs command.");
512 m->mothurOutEndLine();
515 }else if (templateFileName == "not open") { abort = true; }
516 else { if (save) { rdb->setSavedReference(templateFileName); } }
518 //this has to go after save so that if the user sets save=t and provides no reference we abort
519 taxonomyFileName = validParameter.validFile(parameters, "taxonomy", true);
520 if (taxonomyFileName == "not found") {
521 //check for saved reference sequences
522 if (rdb->wordGenusProb.size() != 0) {
523 taxonomyFileName = "saved";
525 m->mothurOut("[ERROR]: You don't have any saved taxonomy information and the taxonomy parameter is a required for the classify.seqs command.");
526 m->mothurOutEndLine();
529 }else if (taxonomyFileName == "not open") { abort = true; }
530 else { if (save) { rdb->setSavedTaxonomy(taxonomyFileName); } }
532 search = validParameter.validFile(parameters, "search", false); if (search == "not found"){ search = "kmer"; }
534 method = validParameter.validFile(parameters, "method", false); if (method == "not found"){ method = "wang"; }
536 temp = validParameter.validFile(parameters, "ksize", false); if (temp == "not found"){
538 if (method == "zap") { temp = "7"; }
540 m->mothurConvert(temp, kmerSize);
542 temp = validParameter.validFile(parameters, "match", false); if (temp == "not found"){ temp = "1.0"; }
543 m->mothurConvert(temp, match);
545 temp = validParameter.validFile(parameters, "mismatch", false); if (temp == "not found"){ temp = "-1.0"; }
546 m->mothurConvert(temp, misMatch);
548 temp = validParameter.validFile(parameters, "gapopen", false); if (temp == "not found"){ temp = "-2.0"; }
549 m->mothurConvert(temp, gapOpen);
551 temp = validParameter.validFile(parameters, "gapextend", false); if (temp == "not found"){ temp = "-1.0"; }
552 m->mothurConvert(temp, gapExtend);
554 temp = validParameter.validFile(parameters, "numwanted", false); if (temp == "not found"){ temp = "10"; }
555 m->mothurConvert(temp, numWanted);
557 temp = validParameter.validFile(parameters, "cutoff", false); if (temp == "not found"){ temp = "0"; }
558 m->mothurConvert(temp, cutoff);
560 temp = validParameter.validFile(parameters, "probs", false); if (temp == "not found"){ temp = "true"; }
561 probs = m->isTrue(temp);
563 temp = validParameter.validFile(parameters, "relabund", false); if (temp == "not found"){ temp = "false"; }
564 relabund = m->isTrue(temp);
566 temp = validParameter.validFile(parameters, "shortcuts", false); if (temp == "not found"){ temp = "true"; }
567 writeShortcuts = m->isTrue(temp);
569 //temp = validParameter.validFile(parameters, "flip", false); if (temp == "not found"){ temp = "T"; }
570 //flip = m->isTrue(temp);
573 temp = validParameter.validFile(parameters, "iters", false); if (temp == "not found") { temp = "100"; }
574 m->mothurConvert(temp, iters);
577 if ((method == "wang") && (search != "kmer")) {
578 m->mothurOut("The wang method requires the kmer search. " + search + " will be disregarded, and kmer will be used." ); m->mothurOutEndLine();
582 if ((method == "zap") && ((search != "kmer") && (search != "align"))) {
583 m->mothurOut("The zap method requires the kmer or align search. " + search + " will be disregarded, and kmer will be used." ); m->mothurOutEndLine();
589 if (namefileNames.size() == 0){
590 if (fastaFileNames.size() != 0) {
591 vector<string> files; files.push_back(fastaFileNames[fastaFileNames.size()-1]);
592 parser.getNameFile(files);
599 catch(exception& e) {
600 m->errorOut(e, "ClassifySeqsCommand", "ClassifySeqsCommand");
605 //**********************************************************************************************************************
606 ClassifySeqsCommand::~ClassifySeqsCommand(){
607 if (abort == false) {
608 for (int i = 0; i < lines.size(); i++) { delete lines[i]; } lines.clear();
611 //**********************************************************************************************************************
613 int ClassifySeqsCommand::execute(){
615 if (abort == true) { if (calledHelp) { return 0; } return 2; }
617 string outputMethodTag = method;
618 if(method == "wang"){ classify = new Bayesian(taxonomyFileName, templateFileName, search, kmerSize, cutoff, iters, rand(), flip, writeShortcuts); }
619 else if(method == "knn"){ classify = new Knn(taxonomyFileName, templateFileName, search, kmerSize, gapOpen, gapExtend, match, misMatch, numWanted, rand()); }
620 else if(method == "zap"){
621 outputMethodTag = search + "_" + outputMethodTag;
622 if (search == "kmer") { classify = new KmerTree(templateFileName, taxonomyFileName, kmerSize, cutoff); }
623 else { classify = new AlignTree(templateFileName, taxonomyFileName, cutoff); }
626 m->mothurOut(search + " is not a valid method option. I will run the command using wang.");
627 m->mothurOutEndLine();
628 classify = new Bayesian(taxonomyFileName, templateFileName, search, kmerSize, cutoff, iters, rand(), flip, writeShortcuts);
631 if (m->control_pressed) { delete classify; return 0; }
633 for (int s = 0; s < fastaFileNames.size(); s++) {
635 m->mothurOut("Classifying sequences from " + fastaFileNames[s] + " ..." ); m->mothurOutEndLine();
637 string baseTName = m->getSimpleName(taxonomyFileName);
638 if (taxonomyFileName == "saved") { baseTName = rdb->getSavedTaxonomy(); }
640 //set rippedTaxName to
641 string RippedTaxName = "";
642 bool foundDot = false;
643 for (int i = baseTName.length()-1; i >= 0; i--) {
644 if (foundDot && (baseTName[i] != '.')) { RippedTaxName = baseTName[i] + RippedTaxName; }
645 else if (foundDot && (baseTName[i] == '.')) { break; }
646 else if (!foundDot && (baseTName[i] == '.')) { foundDot = true; }
648 //if (RippedTaxName != "") { RippedTaxName += "."; }
650 if (outputDir == "") { outputDir += m->hasPath(fastaFileNames[s]); }
651 map<string, string> variables;
652 variables["[filename]"] = outputDir + m->getRootName(m->getSimpleName(fastaFileNames[s]));
653 variables["[tag]"] = RippedTaxName;
654 variables["[tag2]"] = outputMethodTag;
655 string newTaxonomyFile = getOutputFileName("taxonomy", variables);
656 string newaccnosFile = getOutputFileName("accnos", variables);
657 string tempTaxonomyFile = outputDir + m->getRootName(m->getSimpleName(fastaFileNames[s])) + "taxonomy.temp";
658 string taxSummary = getOutputFileName("taxsummary", variables);
660 if ((method == "knn") && (search == "distance")) {
661 string DistName = getOutputFileName("matchdist", variables);
662 classify->setDistName(DistName); outputNames.push_back(DistName); outputTypes["matchdist"].push_back(DistName);
665 outputNames.push_back(newTaxonomyFile); outputTypes["taxonomy"].push_back(newTaxonomyFile);
666 outputNames.push_back(taxSummary); outputTypes["taxsummary"].push_back(taxSummary);
668 int start = time(NULL);
669 int numFastaSeqs = 0;
670 for (int i = 0; i < lines.size(); i++) { delete lines[i]; } lines.clear();
673 int pid, numSeqsPerProcessor;
675 vector<unsigned long long> MPIPos;
678 MPI_Comm_rank(MPI_COMM_WORLD, &pid); //find out who we are
679 MPI_Comm_size(MPI_COMM_WORLD, &processors);
682 MPI_File outMPINewTax;
683 MPI_File outMPITempTax;
686 int outMode=MPI_MODE_CREATE|MPI_MODE_WRONLY;
687 int inMode=MPI_MODE_RDONLY;
689 char outNewTax[1024];
690 strcpy(outNewTax, newTaxonomyFile.c_str());
692 char outTempTax[1024];
693 strcpy(outTempTax, tempTaxonomyFile.c_str());
696 strcpy(outAcc, newaccnosFile.c_str());
698 char inFileName[1024];
699 strcpy(inFileName, fastaFileNames[s].c_str());
701 MPI_File_open(MPI_COMM_WORLD, inFileName, inMode, MPI_INFO_NULL, &inMPI); //comm, filename, mode, info, filepointer
702 MPI_File_open(MPI_COMM_WORLD, outNewTax, outMode, MPI_INFO_NULL, &outMPINewTax);
703 MPI_File_open(MPI_COMM_WORLD, outTempTax, outMode, MPI_INFO_NULL, &outMPITempTax);
704 MPI_File_open(MPI_COMM_WORLD, outAcc, outMode, MPI_INFO_NULL, &outMPIAcc);
706 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; }
708 if (pid == 0) { //you are the root process
710 MPIPos = m->setFilePosFasta(fastaFileNames[s], numFastaSeqs); //fills MPIPos, returns numSeqs
712 //send file positions to all processes
713 for(int i = 1; i < processors; i++) {
714 MPI_Send(&numFastaSeqs, 1, MPI_INT, i, tag, MPI_COMM_WORLD);
715 MPI_Send(&MPIPos[0], (numFastaSeqs+1), MPI_LONG, i, tag, MPI_COMM_WORLD);
718 //figure out how many sequences you have to align
719 numSeqsPerProcessor = numFastaSeqs / processors;
720 int startIndex = pid * numSeqsPerProcessor;
721 if(pid == (processors - 1)){ numSeqsPerProcessor = numFastaSeqs - pid * numSeqsPerProcessor; }
725 driverMPI(startIndex, numSeqsPerProcessor, inMPI, outMPINewTax, outMPITempTax, outMPIAcc, MPIPos);
727 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; }
729 for (int i = 1; i < processors; i++) {
731 MPI_Recv(&done, 1, MPI_INT, i, tag, MPI_COMM_WORLD, &status);
733 }else{ //you are a child process
734 MPI_Recv(&numFastaSeqs, 1, MPI_INT, 0, tag, MPI_COMM_WORLD, &status);
735 MPIPos.resize(numFastaSeqs+1);
736 MPI_Recv(&MPIPos[0], (numFastaSeqs+1), MPI_LONG, 0, tag, MPI_COMM_WORLD, &status);
738 //figure out how many sequences you have to align
739 numSeqsPerProcessor = numFastaSeqs / processors;
740 int startIndex = pid * numSeqsPerProcessor;
741 if(pid == (processors - 1)){ numSeqsPerProcessor = numFastaSeqs - pid * numSeqsPerProcessor; }
745 driverMPI(startIndex, numSeqsPerProcessor, inMPI, outMPINewTax, outMPITempTax, outMPIAcc, MPIPos);
747 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; }
750 MPI_Send(&done, 1, MPI_INT, 0, tag, MPI_COMM_WORLD);
754 MPI_File_close(&inMPI);
755 MPI_File_close(&outMPINewTax);
756 MPI_File_close(&outMPITempTax);
757 MPI_File_close(&outMPIAcc);
758 MPI_Barrier(MPI_COMM_WORLD); //make everyone wait - just in case
762 vector<unsigned long long> positions;
763 #if defined (__APPLE__) || (__MACH__) || (linux) || (__linux) || (__linux__) || (__unix__) || (__unix)
764 positions = m->divideFile(fastaFileNames[s], processors);
765 for (int i = 0; i < (positions.size()-1); i++) { lines.push_back(new linePair(positions[i], positions[(i+1)])); }
767 if (processors == 1) {
768 lines.push_back(new linePair(0, 1000));
770 positions = m->setFilePosFasta(fastaFileNames[s], numFastaSeqs);
771 if (positions.size() < processors) { processors = positions.size(); }
773 //figure out how many sequences you have to process
774 int numSeqsPerProcessor = numFastaSeqs / processors;
775 for (int i = 0; i < processors; i++) {
776 int startIndex = i * numSeqsPerProcessor;
777 if(i == (processors - 1)){ numSeqsPerProcessor = numFastaSeqs - i * numSeqsPerProcessor; }
778 lines.push_back(new linePair(positions[startIndex], numSeqsPerProcessor));
783 numFastaSeqs = driver(lines[0], newTaxonomyFile, tempTaxonomyFile, newaccnosFile, fastaFileNames[s]);
785 numFastaSeqs = createProcesses(newTaxonomyFile, tempTaxonomyFile, newaccnosFile, fastaFileNames[s]);
789 if (!m->isBlank(newaccnosFile)) { m->mothurOutEndLine(); m->mothurOut("[WARNING]: mothur reversed some your sequences for a better classification. If you would like to take a closer look, please check " + newaccnosFile + " for the list of the sequences."); m->mothurOutEndLine();
790 outputNames.push_back(newaccnosFile); outputTypes["accnos"].push_back(newaccnosFile);
791 }else { m->mothurRemove(newaccnosFile); }
793 m->mothurOutEndLine();
794 m->mothurOut("It took " + toString(time(NULL) - start) + " secs to classify " + toString(numFastaSeqs) + " sequences."); m->mothurOutEndLine(); m->mothurOutEndLine();
800 if (pid == 0) { //this part does not need to be paralellized
802 if(namefile != "") { m->mothurOut("Reading " + namefileNames[s] + "..."); cout.flush(); MPIReadNamesFile(namefileNames[s]); m->mothurOut(" Done."); m->mothurOutEndLine(); }
807 m->mothurOut("Reading " + namefileNames[s] + "..."); cout.flush();
808 nameMap.clear(); //remove old names
809 m->readNames(namefileNames[s], nameMap);
810 m->mothurOut(" Done."); m->mothurOutEndLine();
815 GroupMap* groupMap = NULL;
816 CountTable* ct = NULL;
817 PhyloSummary* taxaSum;
819 ct = new CountTable();
820 ct->readTable(countfileNames[s], true, false);
821 taxaSum = new PhyloSummary(taxonomyFileName, ct, relabund);
822 taxaSum->summarize(tempTaxonomyFile);
824 if (groupfile != "") { group = groupfileNames[s]; groupMap = new GroupMap(group); groupMap->readMap(); }
826 taxaSum = new PhyloSummary(taxonomyFileName, groupMap, relabund);
828 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; }
830 if (namefile == "") { taxaSum->summarize(tempTaxonomyFile); }
833 m->openInputFile(tempTaxonomyFile, in);
835 //read in users taxonomy file and add sequences to tree
839 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; }
841 in >> name >> taxon; m->gobble(in);
843 itNames = nameMap.find(name);
845 if (itNames == nameMap.end()) {
846 m->mothurOut(name + " is not in your name file please correct."); m->mothurOutEndLine(); exit(1);
848 for (int i = 0; i < itNames->second.size(); i++) {
849 taxaSum->addSeqToTree(itNames->second[i], taxon); //add it as many times as there are identical seqs
851 itNames->second.clear();
852 nameMap.erase(itNames->first);
858 m->mothurRemove(tempTaxonomyFile);
860 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; }
864 m->openOutputFile(taxSummary, outTaxTree);
865 taxaSum->print(outTaxTree);
868 //output taxonomy with the unclassified bins added
870 m->openInputFile(newTaxonomyFile, inTax);
873 string unclass = newTaxonomyFile + ".unclass.temp";
874 m->openOutputFile(unclass, outTax);
876 //get maxLevel from phylotree so you know how many 'unclassified's to add
877 int maxLevel = taxaSum->getMaxLevel();
879 //read taxfile - this reading and rewriting is done to preserve the confidence scores.
881 while (!inTax.eof()) {
882 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; }
884 inTax >> name >> taxon; m->gobble(inTax);
886 string newTax = addUnclassifieds(taxon, maxLevel);
888 outTax << name << '\t' << newTax << endl;
893 if (ct != NULL) { delete ct; }
894 if (groupMap != NULL) { delete groupMap; } delete taxaSum;
895 m->mothurRemove(newTaxonomyFile);
896 rename(unclass.c_str(), newTaxonomyFile.c_str());
898 m->mothurOutEndLine();
899 m->mothurOut("It took " + toString(time(NULL) - start) + " secs to create the summary file for " + toString(numFastaSeqs) + " sequences."); m->mothurOutEndLine(); m->mothurOutEndLine();
907 m->mothurOutEndLine();
908 m->mothurOut("Output File Names: "); m->mothurOutEndLine();
909 for (int i = 0; i < outputNames.size(); i++) { m->mothurOut(outputNames[i]); m->mothurOutEndLine(); }
910 m->mothurOutEndLine();
912 //set taxonomy file as new current taxonomyfile
914 itTypes = outputTypes.find("taxonomy");
915 if (itTypes != outputTypes.end()) {
916 if ((itTypes->second).size() != 0) { current = (itTypes->second)[0]; m->setTaxonomyFile(current); }
920 itTypes = outputTypes.find("accnos");
921 if (itTypes != outputTypes.end()) {
922 if ((itTypes->second).size() != 0) { current = (itTypes->second)[0]; m->setAccnosFile(current); }
929 catch(exception& e) {
930 m->errorOut(e, "ClassifySeqsCommand", "execute");
935 /**************************************************************************************************/
936 string ClassifySeqsCommand::addUnclassifieds(string tax, int maxlevel) {
938 string newTax, taxon;
941 //keep what you have counting the levels
942 while (tax.find_first_of(';') != -1) {
944 taxon = tax.substr(0,tax.find_first_of(';'))+';';
945 tax = tax.substr(tax.find_first_of(';')+1, tax.length());
950 //add "unclassified" until you reach maxLevel
951 while (level < maxlevel) {
952 newTax += "unclassified;";
958 catch(exception& e) {
959 m->errorOut(e, "ClassifySeqsCommand", "addUnclassifieds");
964 /**************************************************************************************************/
966 int ClassifySeqsCommand::createProcesses(string taxFileName, string tempTaxFile, string accnos, string filename) {
972 #if defined (__APPLE__) || (__MACH__) || (linux) || (__linux) || (__linux__) || (__unix__) || (__unix)
975 //loop through and create all the processes you want
976 while (process != processors) {
980 processIDS.push_back(pid); //create map from line number to pid so you can append files in correct order later
983 num = driver(lines[process], taxFileName + m->mothurGetpid(process) + ".temp", tempTaxFile + m->mothurGetpid(process) + ".temp", accnos + m->mothurGetpid(process) + ".temp", filename);
985 //pass numSeqs to parent
987 string tempFile = filename + m->mothurGetpid(process) + ".num.temp";
988 m->openOutputFile(tempFile, out);
994 m->mothurOut("[ERROR]: unable to spawn the necessary processes."); m->mothurOutEndLine();
995 for (int i = 0; i < processIDS.size(); i++) { kill (processIDS[i], SIGINT); }
1000 //parent does its part
1001 num = driver(lines[0], taxFileName, tempTaxFile, accnos, filename);
1003 //force parent to wait until all the processes are done
1004 for (int i=0;i<processIDS.size();i++) {
1005 int temp = processIDS[i];
1009 for (int i = 0; i < processIDS.size(); i++) {
1011 string tempFile = filename + toString(processIDS[i]) + ".num.temp";
1012 m->openInputFile(tempFile, in);
1013 if (!in.eof()) { int tempNum = 0; in >> tempNum; num += tempNum; }
1014 in.close(); m->mothurRemove(m->getFullPathName(tempFile));
1017 //////////////////////////////////////////////////////////////////////////////////////////////////////
1018 //Windows version shared memory, so be careful when passing variables through the alignData struct.
1019 //Above fork() will clone, so memory is separate, but that's not the case with windows,
1020 //////////////////////////////////////////////////////////////////////////////////////////////////////
1022 vector<classifyData*> pDataArray;
1023 DWORD dwThreadIdArray[processors-1];
1024 HANDLE hThreadArray[processors-1];
1026 //Create processor worker threads.
1027 for( int i=0; i<processors-1; i++ ){
1028 // Allocate memory for thread data.
1029 string extension = "";
1030 if (i != 0) { extension = toString(i) + ".temp"; processIDS.push_back(i); }
1032 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, writeShortcuts);
1033 pDataArray.push_back(tempclass);
1035 //MySeqSumThreadFunction is in header. It must be global or static to work with the threads.
1036 //default security attributes, thread function name, argument to thread function, use default creation flags, returns the thread identifier
1037 hThreadArray[i] = CreateThread(NULL, 0, MyClassThreadFunction, pDataArray[i], 0, &dwThreadIdArray[i]);
1041 //parent does its part
1042 num = driver(lines[processors-1], taxFileName + toString(processors-1) + ".temp", tempTaxFile + toString(processors-1) + ".temp", accnos + toString(processors-1) + ".temp", filename);
1043 processIDS.push_back((processors-1));
1045 //Wait until all threads have terminated.
1046 WaitForMultipleObjects(processors-1, hThreadArray, TRUE, INFINITE);
1048 //Close all thread handles and free memory allocations.
1049 for(int i=0; i < pDataArray.size(); i++){
1050 num += pDataArray[i]->count;
1051 if (pDataArray[i]->count != pDataArray[i]->end) {
1052 m->mothurOut("[ERROR]: process " + toString(i) + " only processed " + toString(pDataArray[i]->count) + " of " + toString(pDataArray[i]->end) + " sequences assigned to it, quitting. \n"); m->control_pressed = true;
1054 CloseHandle(hThreadArray[i]);
1055 delete pDataArray[i];
1059 vector<string> nonBlankAccnosFiles;
1060 if (!(m->isBlank(accnos))) { nonBlankAccnosFiles.push_back(accnos); }
1061 else { m->mothurRemove(accnos); } //remove so other files can be renamed to it
1063 for(int i=0;i<processIDS.size();i++){
1064 m->appendFiles((taxFileName + toString(processIDS[i]) + ".temp"), taxFileName);
1065 m->appendFiles((tempTaxFile + toString(processIDS[i]) + ".temp"), tempTaxFile);
1066 if (!(m->isBlank(accnos + toString(processIDS[i]) + ".temp"))) {
1067 nonBlankAccnosFiles.push_back(accnos + toString(processIDS[i]) + ".temp");
1068 }else { m->mothurRemove((accnos + toString(processIDS[i]) + ".temp")); }
1070 m->mothurRemove((m->getFullPathName(taxFileName) + toString(processIDS[i]) + ".temp"));
1071 m->mothurRemove((m->getFullPathName(tempTaxFile) + toString(processIDS[i]) + ".temp"));
1074 //append accnos files
1075 if (nonBlankAccnosFiles.size() != 0) {
1076 rename(nonBlankAccnosFiles[0].c_str(), accnos.c_str());
1078 for (int h=1; h < nonBlankAccnosFiles.size(); h++) {
1079 m->appendFiles(nonBlankAccnosFiles[h], accnos);
1080 m->mothurRemove(nonBlankAccnosFiles[h]);
1082 }else { //recreate the accnosfile if needed
1084 m->openOutputFile(accnos, out);
1091 catch(exception& e) {
1092 m->errorOut(e, "ClassifySeqsCommand", "createProcesses");
1096 //**********************************************************************************************************************
1098 int ClassifySeqsCommand::driver(linePair* filePos, string taxFName, string tempTFName, string accnos, string filename){
1101 m->openOutputFile(taxFName, outTax);
1103 ofstream outTaxSimple;
1104 m->openOutputFile(tempTFName, outTaxSimple);
1107 m->openOutputFile(accnos, outAcc);
1110 m->openInputFile(filename, inFASTA);
1114 inFASTA.seekg(filePos->start);
1120 if (m->control_pressed) {
1123 outTaxSimple.close();
1124 outAcc.close(); return 0; }
1126 Sequence* candidateSeq = new Sequence(inFASTA); m->gobble(inFASTA);
1128 if (candidateSeq->getName() != "") {
1130 taxonomy = classify->getTaxonomy(candidateSeq);
1132 if (m->control_pressed) { delete candidateSeq; return 0; }
1134 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(); }
1136 //output confidence scores or not
1138 outTax << candidateSeq->getName() << '\t' << taxonomy << endl;
1140 outTax << candidateSeq->getName() << '\t' << classify->getSimpleTax() << endl;
1143 if (classify->getFlipped()) { outAcc << candidateSeq->getName() << endl; }
1145 outTaxSimple << candidateSeq->getName() << '\t' << classify->getSimpleTax() << endl;
1149 delete candidateSeq;
1151 #if defined (__APPLE__) || (__MACH__) || (linux) || (__linux) || (__linux__) || (__unix__) || (__unix)
1152 unsigned long long pos = inFASTA.tellg();
1153 if ((pos == -1) || (pos >= filePos->end)) { break; }
1155 if (inFASTA.eof()) { break; }
1159 if((count) % 100 == 0){ m->mothurOutJustToScreen("Processing sequence: " + toString(count) +"\n"); }
1163 if((count) % 100 != 0){ m->mothurOutJustToScreen("Processing sequence: " + toString(count)+"\n"); }
1167 outTaxSimple.close();
1172 catch(exception& e) {
1173 m->errorOut(e, "ClassifySeqsCommand", "driver");
1177 //**********************************************************************************************************************
1179 int ClassifySeqsCommand::driverMPI(int start, int num, MPI_File& inMPI, MPI_File& newFile, MPI_File& tempFile, MPI_File& accFile, vector<unsigned long long>& MPIPos){
1181 MPI_Status statusNew;
1182 MPI_Status statusTemp;
1183 MPI_Status statusAcc;
1187 MPI_Comm_rank(MPI_COMM_WORLD, &pid); //find out who we are
1190 string outputString;
1192 for(int i=0;i<num;i++){
1194 if (m->control_pressed) { return 0; }
1196 //read next sequence
1197 int length = MPIPos[start+i+1] - MPIPos[start+i];
1198 char* buf4 = new char[length];
1199 MPI_File_read_at(inMPI, MPIPos[start+i], buf4, length, MPI_CHAR, &status);
1201 string tempBuf = buf4;
1202 if (tempBuf.length() > length) { tempBuf = tempBuf.substr(0, length); }
1203 istringstream iss (tempBuf,istringstream::in);
1206 Sequence* candidateSeq = new Sequence(iss);
1208 if (candidateSeq->getName() != "") {
1209 taxonomy = classify->getTaxonomy(candidateSeq);
1211 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(); }
1213 //output confidence scores or not
1215 outputString = candidateSeq->getName() + "\t" + taxonomy + "\n";
1217 outputString = candidateSeq->getName() + "\t" + classify->getSimpleTax() + "\n";
1220 int length = outputString.length();
1221 char* buf2 = new char[length];
1222 memcpy(buf2, outputString.c_str(), length);
1224 MPI_File_write_shared(newFile, buf2, length, MPI_CHAR, &statusNew);
1227 outputString = candidateSeq->getName() + "\t" + classify->getSimpleTax() + "\n";
1228 length = outputString.length();
1229 char* buf = new char[length];
1230 memcpy(buf, outputString.c_str(), length);
1232 MPI_File_write_shared(tempFile, buf, length, MPI_CHAR, &statusTemp);
1235 if (classify->getFlipped()) {
1236 outputString = candidateSeq->getName() + "\n";
1237 length = outputString.length();
1238 char* buf3 = new char[length];
1239 memcpy(buf3, outputString.c_str(), length);
1241 MPI_File_write_shared(accFile, buf3, length, MPI_CHAR, &statusAcc);
1246 delete candidateSeq;
1248 if((i+1) % 100 == 0){ cout << "Classifying sequence " << (i+1) << endl; }
1251 if(num % 100 != 0){ cout << "Classifying sequence " << (num) << endl; }
1256 catch(exception& e) {
1257 m->errorOut(e, "ClassifySeqsCommand", "driverMPI");
1262 //**********************************************************************************************************************
1263 int ClassifySeqsCommand::MPIReadNamesFile(string nameFilename){
1266 nameMap.clear(); //remove old names
1272 //char* inFileName = new char[nameFilename.length()];
1273 //memcpy(inFileName, nameFilename.c_str(), nameFilename.length());
1275 char inFileName[1024];
1276 strcpy(inFileName, nameFilename.c_str());
1278 MPI_File_open(MPI_COMM_WORLD, inFileName, MPI_MODE_RDONLY, MPI_INFO_NULL, &inMPI);
1279 MPI_File_get_size(inMPI, &size);
1280 //delete inFileName;
1282 char* buffer = new char[size];
1283 MPI_File_read(inMPI, buffer, size, MPI_CHAR, &status);
1285 string tempBuf = buffer;
1286 if (tempBuf.length() > size) { tempBuf = tempBuf.substr(0, size); }
1287 istringstream iss (tempBuf,istringstream::in);
1290 string firstCol, secondCol;
1292 iss >> firstCol >> secondCol; m->gobble(iss);
1294 vector<string> temp;
1295 m->splitAtComma(secondCol, temp);
1297 nameMap[firstCol] = temp;
1300 MPI_File_close(&inMPI);
1304 catch(exception& e) {
1305 m->errorOut(e, "ClassifySeqsCommand", "MPIReadNamesFile");
1310 /**************************************************************************************************/