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 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, align 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: wang, knn and zap. The default is wang.\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 wang and zap 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 wang 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::getOutputPattern(string type) {
95 if (type == "taxonomy") { pattern = "[filename],[tag],[tag2],taxonomy"; }
96 else if (type == "taxsummary") { pattern = "[filename],[tag],[tag2],tax.summary"; }
97 else if (type == "accnos") { pattern = "[filename],[tag],[tag2],flip.accnos"; }
98 else if (type == "matchdist") { pattern = "[filename],[tag],[tag2],match.dist"; }
99 else { m->mothurOut("[ERROR]: No definition for type " + type + " output pattern.\n"); m->control_pressed = true; }
103 catch(exception& e) {
104 m->errorOut(e, "ClassifySeqsCommand", "getOutputPattern");
108 //**********************************************************************************************************************
109 ClassifySeqsCommand::ClassifySeqsCommand(){
111 abort = true; calledHelp = true;
113 vector<string> tempOutNames;
114 outputTypes["taxonomy"] = tempOutNames;
115 outputTypes["accnos"] = tempOutNames;
116 outputTypes["taxsummary"] = tempOutNames;
117 outputTypes["matchdist"] = tempOutNames;
119 catch(exception& e) {
120 m->errorOut(e, "ClassifySeqsCommand", "ClassifySeqsCommand");
124 //**********************************************************************************************************************
125 ClassifySeqsCommand::ClassifySeqsCommand(string option) {
127 abort = false; calledHelp = false;
128 rdb = ReferenceDB::getInstance(); hasName = false; hasCount=false;
130 //allow user to run help
131 if(option == "help") { help(); abort = true; calledHelp = true; }
132 else if(option == "citation") { citation(); abort = true; calledHelp = true;}
135 vector<string> myArray = setParameters();
137 OptionParser parser(option);
138 map<string, string> parameters = parser.getParameters();
140 ValidParameters validParameter("classify.seqs");
141 map<string, string>::iterator it;
143 //check to make sure all parameters are valid for command
144 for (it = parameters.begin(); it != parameters.end(); it++) {
145 if (validParameter.isValidParameter(it->first, myArray, it->second) != true) { abort = true; }
148 //initialize outputTypes
149 vector<string> tempOutNames;
150 outputTypes["taxonomy"] = tempOutNames;
151 outputTypes["taxsummary"] = tempOutNames;
152 outputTypes["matchdist"] = tempOutNames;
153 outputTypes["accnos"] = tempOutNames;
155 //if the user changes the output directory command factory will send this info to us in the output parameter
156 outputDir = validParameter.validFile(parameters, "outputdir", false); if (outputDir == "not found"){ outputDir = ""; }
158 //if the user changes the input directory command factory will send this info to us in the output parameter
159 string inputDir = validParameter.validFile(parameters, "inputdir", false);
160 if (inputDir == "not found"){ inputDir = ""; }
163 it = parameters.find("reference");
164 //user has given a template file
165 if(it != parameters.end()){
166 path = m->hasPath(it->second);
167 //if the user has not given a path then, add inputdir. else leave path alone.
168 if (path == "") { parameters["reference"] = inputDir + it->second; }
171 it = parameters.find("taxonomy");
172 //user has given a template file
173 if(it != parameters.end()){
174 path = m->hasPath(it->second);
175 //if the user has not given a path then, add inputdir. else leave path alone.
176 if (path == "") { parameters["taxonomy"] = inputDir + it->second; }
180 fastaFileName = validParameter.validFile(parameters, "fasta", false);
181 if (fastaFileName == "not found") {
182 //if there is a current fasta file, use it
183 string filename = m->getFastaFile();
184 if (filename != "") { fastaFileNames.push_back(filename); m->mothurOut("Using " + filename + " as input file for the fasta parameter."); m->mothurOutEndLine(); }
185 else { m->mothurOut("You have no current fastafile and the fasta parameter is required."); m->mothurOutEndLine(); abort = true; }
188 m->splitAtDash(fastaFileName, fastaFileNames);
190 //go through files and make sure they are good, if not, then disregard them
191 for (int i = 0; i < fastaFileNames.size(); i++) {
194 if (fastaFileNames[i] == "current") {
195 fastaFileNames[i] = m->getFastaFile();
196 if (fastaFileNames[i] != "") { m->mothurOut("Using " + fastaFileNames[i] + " as input file for the fasta parameter where you had given current."); m->mothurOutEndLine(); }
198 m->mothurOut("You have no current fastafile, ignoring current."); m->mothurOutEndLine(); ignore=true;
199 //erase from file list
200 fastaFileNames.erase(fastaFileNames.begin()+i);
207 if (inputDir != "") {
208 string path = m->hasPath(fastaFileNames[i]);
209 //if the user has not given a path then, add inputdir. else leave path alone.
210 if (path == "") { fastaFileNames[i] = inputDir + fastaFileNames[i]; }
216 ableToOpen = m->openInputFile(fastaFileNames[i], in, "noerror");
218 //if you can't open it, try default location
219 if (ableToOpen == 1) {
220 if (m->getDefaultPath() != "") { //default path is set
221 string tryPath = m->getDefaultPath() + m->getSimpleName(fastaFileNames[i]);
222 m->mothurOut("Unable to open " + fastaFileNames[i] + ". Trying default " + tryPath); m->mothurOutEndLine();
224 ableToOpen = m->openInputFile(tryPath, in2, "noerror");
226 fastaFileNames[i] = tryPath;
230 if (ableToOpen == 1) {
231 if (m->getOutputDir() != "") { //default path is set
232 string tryPath = m->getOutputDir() + m->getSimpleName(fastaFileNames[i]);
233 m->mothurOut("Unable to open " + fastaFileNames[i] + ". Trying output directory " + tryPath); m->mothurOutEndLine();
235 ableToOpen = m->openInputFile(tryPath, in2, "noerror");
237 fastaFileNames[i] = tryPath;
243 if (ableToOpen == 1) {
244 m->mothurOut("Unable to open " + fastaFileNames[i] + ". It will be disregarded."); m->mothurOutEndLine();
245 //erase from file list
246 fastaFileNames.erase(fastaFileNames.begin()+i);
249 m->setFastaFile(fastaFileNames[i]);
255 //make sure there is at least one valid file left
256 if (fastaFileNames.size() == 0) { m->mothurOut("no valid files."); m->mothurOutEndLine(); abort = true; }
259 namefile = validParameter.validFile(parameters, "name", false);
260 if (namefile == "not found") { namefile = ""; }
262 m->splitAtDash(namefile, namefileNames);
264 //go through files and make sure they are good, if not, then disregard them
265 for (int i = 0; i < namefileNames.size(); i++) {
267 if (namefileNames[i] == "current") {
268 namefileNames[i] = m->getNameFile();
269 if (namefileNames[i] != "") { m->mothurOut("Using " + namefileNames[i] + " as input file for the name parameter where you had given current."); m->mothurOutEndLine(); }
271 m->mothurOut("You have no current namefile, ignoring current."); m->mothurOutEndLine(); ignore=true;
272 //erase from file list
273 namefileNames.erase(namefileNames.begin()+i);
280 if (inputDir != "") {
281 string path = m->hasPath(namefileNames[i]);
282 //if the user has not given a path then, add inputdir. else leave path alone.
283 if (path == "") { namefileNames[i] = inputDir + namefileNames[i]; }
288 ableToOpen = m->openInputFile(namefileNames[i], in, "noerror");
290 //if you can't open it, try default location
291 if (ableToOpen == 1) {
292 if (m->getDefaultPath() != "") { //default path is set
293 string tryPath = m->getDefaultPath() + m->getSimpleName(namefileNames[i]);
294 m->mothurOut("Unable to open " + namefileNames[i] + ". Trying default " + tryPath); m->mothurOutEndLine();
296 ableToOpen = m->openInputFile(tryPath, in2, "noerror");
298 namefileNames[i] = tryPath;
302 if (ableToOpen == 1) {
303 if (m->getOutputDir() != "") { //default path is set
304 string tryPath = m->getOutputDir() + m->getSimpleName(namefileNames[i]);
305 m->mothurOut("Unable to open " + namefileNames[i] + ". Trying output directory " + tryPath); m->mothurOutEndLine();
307 ableToOpen = m->openInputFile(tryPath, in2, "noerror");
309 namefileNames[i] = tryPath;
314 if (ableToOpen == 1) {
315 m->mothurOut("Unable to open " + namefileNames[i] + ". It will be disregarded."); m->mothurOutEndLine(); abort = true;
316 //erase from file list
317 namefileNames.erase(namefileNames.begin()+i);
320 m->setNameFile(namefileNames[i]);
326 if (namefileNames.size() != 0) { hasName = true; }
328 if (namefile != "") {
329 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(); }
332 //check for required parameters
333 countfile = validParameter.validFile(parameters, "count", false);
334 if (countfile == "not found") {
337 m->splitAtDash(countfile, countfileNames);
339 //go through files and make sure they are good, if not, then disregard them
340 for (int i = 0; i < countfileNames.size(); i++) {
343 if (countfileNames[i] == "current") {
344 countfileNames[i] = m->getCountTableFile();
345 if (countfileNames[i] != "") { m->mothurOut("Using " + countfileNames[i] + " as input file for the count parameter where you had given current."); m->mothurOutEndLine(); }
347 m->mothurOut("You have no current count file, ignoring current."); m->mothurOutEndLine(); ignore=true;
348 //erase from file list
349 countfileNames.erase(countfileNames.begin()+i);
356 if (inputDir != "") {
357 string path = m->hasPath(countfileNames[i]);
358 //if the user has not given a path then, add inputdir. else leave path alone.
359 if (path == "") { countfileNames[i] = inputDir + countfileNames[i]; }
365 ableToOpen = m->openInputFile(countfileNames[i], in, "noerror");
367 //if you can't open it, try default location
368 if (ableToOpen == 1) {
369 if (m->getDefaultPath() != "") { //default path is set
370 string tryPath = m->getDefaultPath() + m->getSimpleName(countfileNames[i]);
371 m->mothurOut("Unable to open " + countfileNames[i] + ". Trying default " + tryPath); m->mothurOutEndLine();
373 ableToOpen = m->openInputFile(tryPath, in2, "noerror");
375 countfileNames[i] = tryPath;
379 if (ableToOpen == 1) {
380 if (m->getOutputDir() != "") { //default path is set
381 string tryPath = m->getOutputDir() + m->getSimpleName(countfileNames[i]);
382 m->mothurOut("Unable to open " + countfileNames[i] + ". Trying output directory " + tryPath); m->mothurOutEndLine();
384 ableToOpen = m->openInputFile(tryPath, in2, "noerror");
386 countfileNames[i] = tryPath;
392 if (ableToOpen == 1) {
393 m->mothurOut("Unable to open " + countfileNames[i] + ". It will be disregarded."); m->mothurOutEndLine();
394 //erase from file list
395 countfileNames.erase(countfileNames.begin()+i);
398 m->setCountTableFile(countfileNames[i]);
404 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(); } }
406 //make sure there is at least one valid file left
407 if (hasName && hasCount) { m->mothurOut("[ERROR]: You must enter ONLY ONE of the following: count or name."); m->mothurOutEndLine(); abort = true; }
409 groupfile = validParameter.validFile(parameters, "group", false);
410 if (groupfile == "not found") { groupfile = ""; }
412 m->splitAtDash(groupfile, groupfileNames);
414 //go through files and make sure they are good, if not, then disregard them
415 for (int i = 0; i < groupfileNames.size(); i++) {
418 if (groupfileNames[i] == "current") {
419 groupfileNames[i] = m->getGroupFile();
420 if (groupfileNames[i] != "") { m->mothurOut("Using " + groupfileNames[i] + " as input file for the group parameter where you had given current."); m->mothurOutEndLine(); }
422 m->mothurOut("You have no current group file, ignoring current."); m->mothurOutEndLine(); ignore=true;
423 //erase from file list
424 groupfileNames.erase(groupfileNames.begin()+i);
431 if (inputDir != "") {
432 string path = m->hasPath(groupfileNames[i]);
433 cout << path << '\t' << inputDir << endl;
434 //if the user has not given a path then, add inputdir. else leave path alone.
435 if (path == "") { groupfileNames[i] = inputDir + groupfileNames[i]; }
441 ableToOpen = m->openInputFile(groupfileNames[i], in, "noerror");
443 //if you can't open it, try default location
444 if (ableToOpen == 1) {
445 if (m->getDefaultPath() != "") { //default path is set
446 string tryPath = m->getDefaultPath() + m->getSimpleName(groupfileNames[i]);
447 m->mothurOut("Unable to open " + groupfileNames[i] + ". Trying default " + tryPath); m->mothurOutEndLine();
449 ableToOpen = m->openInputFile(tryPath, in2, "noerror");
451 groupfileNames[i] = tryPath;
455 if (ableToOpen == 1) {
456 if (m->getOutputDir() != "") { //default path is set
457 string tryPath = m->getOutputDir() + m->getSimpleName(groupfileNames[i]);
458 m->mothurOut("Unable to open " + groupfileNames[i] + ". Trying output directory " + tryPath); m->mothurOutEndLine();
460 ableToOpen = m->openInputFile(tryPath, in2, "noerror");
462 groupfileNames[i] = tryPath;
468 if (ableToOpen == 1) {
469 m->mothurOut("Unable to open " + groupfileNames[i] + ". It will be disregarded."); m->mothurOutEndLine();
470 //erase from file list
471 groupfileNames.erase(groupfileNames.begin()+i);
474 m->setGroupFile(groupfileNames[i]);
481 if (groupfile != "") {
482 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(); }
483 if (hasCount) { m->mothurOut("[ERROR]: You must enter ONLY ONE of the following: count or group."); m->mothurOutEndLine(); abort = true; }
485 for (int i = 0; i < fastaFileNames.size(); i++) { groupfileNames.push_back(""); }
488 //check for optional parameter and set defaults
489 // ...at some point should added some additional type checking...
491 temp = validParameter.validFile(parameters, "processors", false); if (temp == "not found"){ temp = m->getProcessors(); }
492 m->setProcessors(temp);
493 m->mothurConvert(temp, processors);
495 temp = validParameter.validFile(parameters, "save", false); if (temp == "not found"){ temp = "f"; }
496 save = m->isTrue(temp);
498 if (save) { //clear out old references
502 //this has to go after save so that if the user sets save=t and provides no reference we abort
503 templateFileName = validParameter.validFile(parameters, "reference", true);
504 if (templateFileName == "not found") {
505 //check for saved reference sequences
506 if (rdb->referenceSeqs.size() != 0) {
507 templateFileName = "saved";
509 m->mothurOut("[ERROR]: You don't have any saved reference sequences and the reference parameter is a required for the classify.seqs command.");
510 m->mothurOutEndLine();
513 }else if (templateFileName == "not open") { abort = true; }
514 else { if (save) { rdb->setSavedReference(templateFileName); } }
516 //this has to go after save so that if the user sets save=t and provides no reference we abort
517 taxonomyFileName = validParameter.validFile(parameters, "taxonomy", true);
518 if (taxonomyFileName == "not found") {
519 //check for saved reference sequences
520 if (rdb->wordGenusProb.size() != 0) {
521 taxonomyFileName = "saved";
523 m->mothurOut("[ERROR]: You don't have any saved taxonomy information and the taxonomy parameter is a required for the classify.seqs command.");
524 m->mothurOutEndLine();
527 }else if (taxonomyFileName == "not open") { abort = true; }
528 else { if (save) { rdb->setSavedTaxonomy(taxonomyFileName); } }
530 search = validParameter.validFile(parameters, "search", false); if (search == "not found"){ search = "kmer"; }
532 method = validParameter.validFile(parameters, "method", false); if (method == "not found"){ method = "wang"; }
534 temp = validParameter.validFile(parameters, "ksize", false); if (temp == "not found"){
536 if (method == "zap") { temp = "7"; }
538 m->mothurConvert(temp, kmerSize);
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, "shortcuts", false); if (temp == "not found"){ temp = "true"; }
562 writeShortcuts = m->isTrue(temp);
564 //temp = validParameter.validFile(parameters, "flip", false); if (temp == "not found"){ temp = "T"; }
565 //flip = m->isTrue(temp);
568 temp = validParameter.validFile(parameters, "iters", false); if (temp == "not found") { temp = "100"; }
569 m->mothurConvert(temp, iters);
572 if ((method == "wang") && (search != "kmer")) {
573 m->mothurOut("The wang method requires the kmer search. " + search + " will be disregarded, and kmer will be used." ); m->mothurOutEndLine();
577 if ((method == "zap") && ((search != "kmer") && (search != "align"))) {
578 m->mothurOut("The zap method requires the kmer or align search. " + search + " will be disregarded, and kmer will be used." ); m->mothurOutEndLine();
584 if (namefileNames.size() == 0){
585 if (fastaFileNames.size() != 0) {
586 vector<string> files; files.push_back(fastaFileNames[fastaFileNames.size()-1]);
587 parser.getNameFile(files);
594 catch(exception& e) {
595 m->errorOut(e, "ClassifySeqsCommand", "ClassifySeqsCommand");
600 //**********************************************************************************************************************
601 ClassifySeqsCommand::~ClassifySeqsCommand(){
602 if (abort == false) {
603 for (int i = 0; i < lines.size(); i++) { delete lines[i]; } lines.clear();
606 //**********************************************************************************************************************
608 int ClassifySeqsCommand::execute(){
610 if (abort == true) { if (calledHelp) { return 0; } return 2; }
612 string outputMethodTag = method;
613 if(method == "wang"){ classify = new Bayesian(taxonomyFileName, templateFileName, search, kmerSize, cutoff, iters, rand(), flip, writeShortcuts); }
614 else if(method == "knn"){ classify = new Knn(taxonomyFileName, templateFileName, search, kmerSize, gapOpen, gapExtend, match, misMatch, numWanted, rand()); }
615 else if(method == "zap"){
616 outputMethodTag = search + "_" + outputMethodTag;
617 if (search == "kmer") { classify = new KmerTree(templateFileName, taxonomyFileName, kmerSize, cutoff); }
618 else { classify = new AlignTree(templateFileName, taxonomyFileName, cutoff); }
621 m->mothurOut(search + " is not a valid method option. I will run the command using wang.");
622 m->mothurOutEndLine();
623 classify = new Bayesian(taxonomyFileName, templateFileName, search, kmerSize, cutoff, iters, rand(), flip, writeShortcuts);
626 if (m->control_pressed) { delete classify; return 0; }
628 for (int s = 0; s < fastaFileNames.size(); s++) {
630 m->mothurOut("Classifying sequences from " + fastaFileNames[s] + " ..." ); m->mothurOutEndLine();
632 string baseTName = m->getSimpleName(taxonomyFileName);
633 if (taxonomyFileName == "saved") { baseTName = rdb->getSavedTaxonomy(); }
635 //set rippedTaxName to
636 string RippedTaxName = "";
637 bool foundDot = false;
638 for (int i = baseTName.length()-1; i >= 0; i--) {
639 if (foundDot && (baseTName[i] != '.')) { RippedTaxName = baseTName[i] + RippedTaxName; }
640 else if (foundDot && (baseTName[i] == '.')) { break; }
641 else if (!foundDot && (baseTName[i] == '.')) { foundDot = true; }
643 //if (RippedTaxName != "") { RippedTaxName += "."; }
645 if (outputDir == "") { outputDir += m->hasPath(fastaFileNames[s]); }
646 map<string, string> variables;
647 variables["[filename]"] = outputDir + m->getRootName(m->getSimpleName(fastaFileNames[s]));
648 variables["[tag]"] = RippedTaxName;
649 variables["[tag2]"] = outputMethodTag;
650 string newTaxonomyFile = getOutputFileName("taxonomy", variables);
651 string newaccnosFile = getOutputFileName("accnos", variables);
652 string tempTaxonomyFile = outputDir + m->getRootName(m->getSimpleName(fastaFileNames[s])) + "taxonomy.temp";
653 string taxSummary = getOutputFileName("taxsummary", variables);
655 if ((method == "knn") && (search == "distance")) {
656 string DistName = getOutputFileName("matchdist", variables);
657 classify->setDistName(DistName); outputNames.push_back(DistName); outputTypes["matchdist"].push_back(DistName);
660 outputNames.push_back(newTaxonomyFile); outputTypes["taxonomy"].push_back(newTaxonomyFile);
661 outputNames.push_back(taxSummary); outputTypes["taxsummary"].push_back(taxSummary);
663 int start = time(NULL);
664 int numFastaSeqs = 0;
665 for (int i = 0; i < lines.size(); i++) { delete lines[i]; } lines.clear();
668 int pid, numSeqsPerProcessor;
670 vector<unsigned long long> MPIPos;
673 MPI_Comm_rank(MPI_COMM_WORLD, &pid); //find out who we are
674 MPI_Comm_size(MPI_COMM_WORLD, &processors);
677 MPI_File outMPINewTax;
678 MPI_File outMPITempTax;
681 int outMode=MPI_MODE_CREATE|MPI_MODE_WRONLY;
682 int inMode=MPI_MODE_RDONLY;
684 char outNewTax[1024];
685 strcpy(outNewTax, newTaxonomyFile.c_str());
687 char outTempTax[1024];
688 strcpy(outTempTax, tempTaxonomyFile.c_str());
691 strcpy(outAcc, newaccnosFile.c_str());
693 char inFileName[1024];
694 strcpy(inFileName, fastaFileNames[s].c_str());
696 MPI_File_open(MPI_COMM_WORLD, inFileName, inMode, MPI_INFO_NULL, &inMPI); //comm, filename, mode, info, filepointer
697 MPI_File_open(MPI_COMM_WORLD, outNewTax, outMode, MPI_INFO_NULL, &outMPINewTax);
698 MPI_File_open(MPI_COMM_WORLD, outTempTax, outMode, MPI_INFO_NULL, &outMPITempTax);
699 MPI_File_open(MPI_COMM_WORLD, outAcc, outMode, MPI_INFO_NULL, &outMPIAcc);
701 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; }
703 if (pid == 0) { //you are the root process
705 MPIPos = m->setFilePosFasta(fastaFileNames[s], numFastaSeqs); //fills MPIPos, returns numSeqs
707 //send file positions to all processes
708 for(int i = 1; i < processors; i++) {
709 MPI_Send(&numFastaSeqs, 1, MPI_INT, i, tag, MPI_COMM_WORLD);
710 MPI_Send(&MPIPos[0], (numFastaSeqs+1), MPI_LONG, i, tag, MPI_COMM_WORLD);
713 //figure out how many sequences you have to align
714 numSeqsPerProcessor = numFastaSeqs / processors;
715 int startIndex = pid * numSeqsPerProcessor;
716 if(pid == (processors - 1)){ numSeqsPerProcessor = numFastaSeqs - pid * numSeqsPerProcessor; }
720 driverMPI(startIndex, numSeqsPerProcessor, inMPI, outMPINewTax, outMPITempTax, outMPIAcc, MPIPos);
722 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; }
724 for (int i = 1; i < processors; i++) {
726 MPI_Recv(&done, 1, MPI_INT, i, tag, MPI_COMM_WORLD, &status);
728 }else{ //you are a child process
729 MPI_Recv(&numFastaSeqs, 1, MPI_INT, 0, tag, MPI_COMM_WORLD, &status);
730 MPIPos.resize(numFastaSeqs+1);
731 MPI_Recv(&MPIPos[0], (numFastaSeqs+1), MPI_LONG, 0, tag, MPI_COMM_WORLD, &status);
733 //figure out how many sequences you have to align
734 numSeqsPerProcessor = numFastaSeqs / processors;
735 int startIndex = pid * numSeqsPerProcessor;
736 if(pid == (processors - 1)){ numSeqsPerProcessor = numFastaSeqs - pid * numSeqsPerProcessor; }
740 driverMPI(startIndex, numSeqsPerProcessor, inMPI, outMPINewTax, outMPITempTax, outMPIAcc, MPIPos);
742 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; }
745 MPI_Send(&done, 1, MPI_INT, 0, tag, MPI_COMM_WORLD);
749 MPI_File_close(&inMPI);
750 MPI_File_close(&outMPINewTax);
751 MPI_File_close(&outMPITempTax);
752 MPI_File_close(&outMPIAcc);
753 MPI_Barrier(MPI_COMM_WORLD); //make everyone wait - just in case
757 vector<unsigned long long> positions;
758 #if defined (__APPLE__) || (__MACH__) || (linux) || (__linux) || (__linux__) || (__unix__) || (__unix)
759 positions = m->divideFile(fastaFileNames[s], processors);
760 for (int i = 0; i < (positions.size()-1); i++) { lines.push_back(new linePair(positions[i], positions[(i+1)])); }
762 if (processors == 1) {
763 lines.push_back(new linePair(0, 1000));
765 positions = m->setFilePosFasta(fastaFileNames[s], numFastaSeqs);
766 if (positions.size() < processors) { processors = positions.size(); }
768 //figure out how many sequences you have to process
769 int numSeqsPerProcessor = numFastaSeqs / processors;
770 for (int i = 0; i < processors; i++) {
771 int startIndex = i * numSeqsPerProcessor;
772 if(i == (processors - 1)){ numSeqsPerProcessor = numFastaSeqs - i * numSeqsPerProcessor; }
773 lines.push_back(new linePair(positions[startIndex], numSeqsPerProcessor));
778 numFastaSeqs = driver(lines[0], newTaxonomyFile, tempTaxonomyFile, newaccnosFile, fastaFileNames[s]);
780 numFastaSeqs = createProcesses(newTaxonomyFile, tempTaxonomyFile, newaccnosFile, fastaFileNames[s]);
784 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();
785 outputNames.push_back(newaccnosFile); outputTypes["accnos"].push_back(newaccnosFile);
786 }else { m->mothurRemove(newaccnosFile); }
788 m->mothurOutEndLine();
789 m->mothurOut("It took " + toString(time(NULL) - start) + " secs to classify " + toString(numFastaSeqs) + " sequences."); m->mothurOutEndLine(); m->mothurOutEndLine();
795 if (pid == 0) { //this part does not need to be paralellized
797 if(namefile != "") { m->mothurOut("Reading " + namefileNames[s] + "..."); cout.flush(); MPIReadNamesFile(namefileNames[s]); m->mothurOut(" Done."); m->mothurOutEndLine(); }
802 m->mothurOut("Reading " + namefileNames[s] + "..."); cout.flush();
803 nameMap.clear(); //remove old names
804 m->readNames(namefileNames[s], nameMap);
805 m->mothurOut(" Done."); m->mothurOutEndLine();
810 GroupMap* groupMap = NULL;
811 CountTable* ct = NULL;
812 PhyloSummary* taxaSum;
814 ct = new CountTable();
815 ct->readTable(countfileNames[s], true, false);
816 taxaSum = new PhyloSummary(taxonomyFileName, ct);
817 taxaSum->summarize(tempTaxonomyFile);
819 if (groupfile != "") { group = groupfileNames[s]; groupMap = new GroupMap(group); groupMap->readMap(); }
821 taxaSum = new PhyloSummary(taxonomyFileName, groupMap);
823 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; }
825 if (namefile == "") { taxaSum->summarize(tempTaxonomyFile); }
828 m->openInputFile(tempTaxonomyFile, in);
830 //read in users taxonomy file and add sequences to tree
834 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; }
836 in >> name >> taxon; m->gobble(in);
838 itNames = nameMap.find(name);
840 if (itNames == nameMap.end()) {
841 m->mothurOut(name + " is not in your name file please correct."); m->mothurOutEndLine(); exit(1);
843 for (int i = 0; i < itNames->second.size(); i++) {
844 taxaSum->addSeqToTree(itNames->second[i], taxon); //add it as many times as there are identical seqs
846 itNames->second.clear();
847 nameMap.erase(itNames->first);
853 m->mothurRemove(tempTaxonomyFile);
855 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; }
859 m->openOutputFile(taxSummary, outTaxTree);
860 taxaSum->print(outTaxTree);
863 //output taxonomy with the unclassified bins added
865 m->openInputFile(newTaxonomyFile, inTax);
868 string unclass = newTaxonomyFile + ".unclass.temp";
869 m->openOutputFile(unclass, outTax);
871 //get maxLevel from phylotree so you know how many 'unclassified's to add
872 int maxLevel = taxaSum->getMaxLevel();
874 //read taxfile - this reading and rewriting is done to preserve the confidence scores.
876 while (!inTax.eof()) {
877 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; }
879 inTax >> name >> taxon; m->gobble(inTax);
881 string newTax = addUnclassifieds(taxon, maxLevel);
883 outTax << name << '\t' << newTax << endl;
888 if (ct != NULL) { delete ct; }
889 if (groupMap != NULL) { delete groupMap; } delete taxaSum;
890 m->mothurRemove(newTaxonomyFile);
891 rename(unclass.c_str(), newTaxonomyFile.c_str());
893 m->mothurOutEndLine();
894 m->mothurOut("It took " + toString(time(NULL) - start) + " secs to create the summary file for " + toString(numFastaSeqs) + " sequences."); m->mothurOutEndLine(); m->mothurOutEndLine();
902 m->mothurOutEndLine();
903 m->mothurOut("Output File Names: "); m->mothurOutEndLine();
904 for (int i = 0; i < outputNames.size(); i++) { m->mothurOut(outputNames[i]); m->mothurOutEndLine(); }
905 m->mothurOutEndLine();
907 //set taxonomy file as new current taxonomyfile
909 itTypes = outputTypes.find("taxonomy");
910 if (itTypes != outputTypes.end()) {
911 if ((itTypes->second).size() != 0) { current = (itTypes->second)[0]; m->setTaxonomyFile(current); }
915 itTypes = outputTypes.find("accnos");
916 if (itTypes != outputTypes.end()) {
917 if ((itTypes->second).size() != 0) { current = (itTypes->second)[0]; m->setAccnosFile(current); }
924 catch(exception& e) {
925 m->errorOut(e, "ClassifySeqsCommand", "execute");
930 /**************************************************************************************************/
931 string ClassifySeqsCommand::addUnclassifieds(string tax, int maxlevel) {
933 string newTax, taxon;
936 //keep what you have counting the levels
937 while (tax.find_first_of(';') != -1) {
939 taxon = tax.substr(0,tax.find_first_of(';'))+';';
940 tax = tax.substr(tax.find_first_of(';')+1, tax.length());
945 //add "unclassified" until you reach maxLevel
946 while (level < maxlevel) {
947 newTax += "unclassified;";
953 catch(exception& e) {
954 m->errorOut(e, "ClassifySeqsCommand", "addUnclassifieds");
959 /**************************************************************************************************/
961 int ClassifySeqsCommand::createProcesses(string taxFileName, string tempTaxFile, string accnos, string filename) {
967 #if defined (__APPLE__) || (__MACH__) || (linux) || (__linux) || (__linux__) || (__unix__) || (__unix)
970 //loop through and create all the processes you want
971 while (process != processors) {
975 processIDS.push_back(pid); //create map from line number to pid so you can append files in correct order later
978 num = driver(lines[process], taxFileName + toString(getpid()) + ".temp", tempTaxFile + toString(getpid()) + ".temp", accnos + toString(getpid()) + ".temp", filename);
980 //pass numSeqs to parent
982 string tempFile = filename + toString(getpid()) + ".num.temp";
983 m->openOutputFile(tempFile, out);
989 m->mothurOut("[ERROR]: unable to spawn the necessary processes."); m->mothurOutEndLine();
990 for (int i = 0; i < processIDS.size(); i++) { kill (processIDS[i], SIGINT); }
995 //parent does its part
996 num = driver(lines[0], taxFileName, tempTaxFile, accnos, filename);
998 //force parent to wait until all the processes are done
999 for (int i=0;i<processIDS.size();i++) {
1000 int temp = processIDS[i];
1004 for (int i = 0; i < processIDS.size(); i++) {
1006 string tempFile = filename + toString(processIDS[i]) + ".num.temp";
1007 m->openInputFile(tempFile, in);
1008 if (!in.eof()) { int tempNum = 0; in >> tempNum; num += tempNum; }
1009 in.close(); m->mothurRemove(m->getFullPathName(tempFile));
1012 //////////////////////////////////////////////////////////////////////////////////////////////////////
1013 //Windows version shared memory, so be careful when passing variables through the alignData struct.
1014 //Above fork() will clone, so memory is separate, but that's not the case with windows,
1015 //////////////////////////////////////////////////////////////////////////////////////////////////////
1017 vector<classifyData*> pDataArray;
1018 DWORD dwThreadIdArray[processors-1];
1019 HANDLE hThreadArray[processors-1];
1021 //Create processor worker threads.
1022 for( int i=0; i<processors-1; i++ ){
1023 // Allocate memory for thread data.
1024 string extension = "";
1025 if (i != 0) { extension = toString(i) + ".temp"; processIDS.push_back(i); }
1027 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);
1028 pDataArray.push_back(tempclass);
1030 //MySeqSumThreadFunction is in header. It must be global or static to work with the threads.
1031 //default security attributes, thread function name, argument to thread function, use default creation flags, returns the thread identifier
1032 hThreadArray[i] = CreateThread(NULL, 0, MyClassThreadFunction, pDataArray[i], 0, &dwThreadIdArray[i]);
1036 //parent does its part
1037 num = driver(lines[processors-1], taxFileName + toString(processors-1) + ".temp", tempTaxFile + toString(processors-1) + ".temp", accnos + toString(processors-1) + ".temp", filename);
1038 processIDS.push_back((processors-1));
1040 //Wait until all threads have terminated.
1041 WaitForMultipleObjects(processors-1, hThreadArray, TRUE, INFINITE);
1043 //Close all thread handles and free memory allocations.
1044 for(int i=0; i < pDataArray.size(); i++){
1045 num += pDataArray[i]->count;
1046 if (pDataArray[i]->count != pDataArray[i]->end) {
1047 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;
1049 CloseHandle(hThreadArray[i]);
1050 delete pDataArray[i];
1054 vector<string> nonBlankAccnosFiles;
1055 if (!(m->isBlank(accnos))) { nonBlankAccnosFiles.push_back(accnos); }
1056 else { m->mothurRemove(accnos); } //remove so other files can be renamed to it
1058 for(int i=0;i<processIDS.size();i++){
1059 m->appendFiles((taxFileName + toString(processIDS[i]) + ".temp"), taxFileName);
1060 m->appendFiles((tempTaxFile + toString(processIDS[i]) + ".temp"), tempTaxFile);
1061 if (!(m->isBlank(accnos + toString(processIDS[i]) + ".temp"))) {
1062 nonBlankAccnosFiles.push_back(accnos + toString(processIDS[i]) + ".temp");
1063 }else { m->mothurRemove((accnos + toString(processIDS[i]) + ".temp")); }
1065 m->mothurRemove((m->getFullPathName(taxFileName) + toString(processIDS[i]) + ".temp"));
1066 m->mothurRemove((m->getFullPathName(tempTaxFile) + toString(processIDS[i]) + ".temp"));
1069 //append accnos files
1070 if (nonBlankAccnosFiles.size() != 0) {
1071 rename(nonBlankAccnosFiles[0].c_str(), accnos.c_str());
1073 for (int h=1; h < nonBlankAccnosFiles.size(); h++) {
1074 m->appendFiles(nonBlankAccnosFiles[h], accnos);
1075 m->mothurRemove(nonBlankAccnosFiles[h]);
1077 }else { //recreate the accnosfile if needed
1079 m->openOutputFile(accnos, out);
1086 catch(exception& e) {
1087 m->errorOut(e, "ClassifySeqsCommand", "createProcesses");
1091 //**********************************************************************************************************************
1093 int ClassifySeqsCommand::driver(linePair* filePos, string taxFName, string tempTFName, string accnos, string filename){
1096 m->openOutputFile(taxFName, outTax);
1098 ofstream outTaxSimple;
1099 m->openOutputFile(tempTFName, outTaxSimple);
1102 m->openOutputFile(accnos, outAcc);
1105 m->openInputFile(filename, inFASTA);
1109 inFASTA.seekg(filePos->start);
1115 if (m->control_pressed) {
1118 outTaxSimple.close();
1119 outAcc.close(); return 0; }
1121 Sequence* candidateSeq = new Sequence(inFASTA); m->gobble(inFASTA);
1123 if (candidateSeq->getName() != "") {
1125 taxonomy = classify->getTaxonomy(candidateSeq);
1127 if (m->control_pressed) { delete candidateSeq; return 0; }
1129 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(); }
1131 //output confidence scores or not
1133 outTax << candidateSeq->getName() << '\t' << taxonomy << endl;
1135 outTax << candidateSeq->getName() << '\t' << classify->getSimpleTax() << endl;
1138 if (classify->getFlipped()) { outAcc << candidateSeq->getName() << endl; }
1140 outTaxSimple << candidateSeq->getName() << '\t' << classify->getSimpleTax() << endl;
1144 delete candidateSeq;
1146 #if defined (__APPLE__) || (__MACH__) || (linux) || (__linux) || (__linux__) || (__unix__) || (__unix)
1147 unsigned long long pos = inFASTA.tellg();
1148 if ((pos == -1) || (pos >= filePos->end)) { break; }
1150 if (inFASTA.eof()) { break; }
1154 if((count) % 100 == 0){ m->mothurOutJustToScreen("Processing sequence: " + toString(count) +"\n"); }
1158 if((count) % 100 != 0){ m->mothurOutJustToScreen("Processing sequence: " + toString(count)+"\n"); }
1162 outTaxSimple.close();
1167 catch(exception& e) {
1168 m->errorOut(e, "ClassifySeqsCommand", "driver");
1172 //**********************************************************************************************************************
1174 int ClassifySeqsCommand::driverMPI(int start, int num, MPI_File& inMPI, MPI_File& newFile, MPI_File& tempFile, MPI_File& accFile, vector<unsigned long long>& MPIPos){
1176 MPI_Status statusNew;
1177 MPI_Status statusTemp;
1178 MPI_Status statusAcc;
1182 MPI_Comm_rank(MPI_COMM_WORLD, &pid); //find out who we are
1185 string outputString;
1187 for(int i=0;i<num;i++){
1189 if (m->control_pressed) { return 0; }
1191 //read next sequence
1192 int length = MPIPos[start+i+1] - MPIPos[start+i];
1193 char* buf4 = new char[length];
1194 MPI_File_read_at(inMPI, MPIPos[start+i], buf4, length, MPI_CHAR, &status);
1196 string tempBuf = buf4;
1197 if (tempBuf.length() > length) { tempBuf = tempBuf.substr(0, length); }
1198 istringstream iss (tempBuf,istringstream::in);
1201 Sequence* candidateSeq = new Sequence(iss);
1203 if (candidateSeq->getName() != "") {
1204 taxonomy = classify->getTaxonomy(candidateSeq);
1206 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(); }
1208 //output confidence scores or not
1210 outputString = candidateSeq->getName() + "\t" + taxonomy + "\n";
1212 outputString = candidateSeq->getName() + "\t" + classify->getSimpleTax() + "\n";
1215 int length = outputString.length();
1216 char* buf2 = new char[length];
1217 memcpy(buf2, outputString.c_str(), length);
1219 MPI_File_write_shared(newFile, buf2, length, MPI_CHAR, &statusNew);
1222 outputString = candidateSeq->getName() + "\t" + classify->getSimpleTax() + "\n";
1223 length = outputString.length();
1224 char* buf = new char[length];
1225 memcpy(buf, outputString.c_str(), length);
1227 MPI_File_write_shared(tempFile, buf, length, MPI_CHAR, &statusTemp);
1230 if (classify->getFlipped()) {
1231 outputString = candidateSeq->getName() + "\n";
1232 length = outputString.length();
1233 char* buf3 = new char[length];
1234 memcpy(buf3, outputString.c_str(), length);
1236 MPI_File_write_shared(accFile, buf3, length, MPI_CHAR, &statusAcc);
1241 delete candidateSeq;
1243 if((i+1) % 100 == 0){ cout << "Classifying sequence " << (i+1) << endl; }
1246 if(num % 100 != 0){ cout << "Classifying sequence " << (num) << endl; }
1251 catch(exception& e) {
1252 m->errorOut(e, "ClassifySeqsCommand", "driverMPI");
1257 //**********************************************************************************************************************
1258 int ClassifySeqsCommand::MPIReadNamesFile(string nameFilename){
1261 nameMap.clear(); //remove old names
1267 //char* inFileName = new char[nameFilename.length()];
1268 //memcpy(inFileName, nameFilename.c_str(), nameFilename.length());
1270 char inFileName[1024];
1271 strcpy(inFileName, nameFilename.c_str());
1273 MPI_File_open(MPI_COMM_WORLD, inFileName, MPI_MODE_RDONLY, MPI_INFO_NULL, &inMPI);
1274 MPI_File_get_size(inMPI, &size);
1275 //delete inFileName;
1277 char* buffer = new char[size];
1278 MPI_File_read(inMPI, buffer, size, MPI_CHAR, &status);
1280 string tempBuf = buffer;
1281 if (tempBuf.length() > size) { tempBuf = tempBuf.substr(0, size); }
1282 istringstream iss (tempBuf,istringstream::in);
1285 string firstCol, secondCol;
1287 iss >> firstCol >> secondCol; m->gobble(iss);
1289 vector<string> temp;
1290 m->splitAtComma(secondCol, temp);
1292 nameMap[firstCol] = temp;
1295 MPI_File_close(&inMPI);
1299 catch(exception& e) {
1300 m->errorOut(e, "ClassifySeqsCommand", "MPIReadNamesFile");
1305 /**************************************************************************************************/