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 = ""; }
263 m->splitAtDash(namefile, namefileNames);
265 //go through files and make sure they are good, if not, then disregard them
266 for (int i = 0; i < namefileNames.size(); i++) {
268 if (namefileNames[i] == "current") {
269 namefileNames[i] = m->getNameFile();
270 if (namefileNames[i] != "") { m->mothurOut("Using " + namefileNames[i] + " as input file for the name parameter where you had given current."); m->mothurOutEndLine(); }
272 m->mothurOut("You have no current namefile, ignoring current."); m->mothurOutEndLine(); ignore=true;
273 //erase from file list
274 namefileNames.erase(namefileNames.begin()+i);
281 if (inputDir != "") {
282 string path = m->hasPath(namefileNames[i]);
283 //if the user has not given a path then, add inputdir. else leave path alone.
284 if (path == "") { namefileNames[i] = inputDir + namefileNames[i]; }
289 ableToOpen = m->openInputFile(namefileNames[i], in, "noerror");
291 //if you can't open it, try default location
292 if (ableToOpen == 1) {
293 if (m->getDefaultPath() != "") { //default path is set
294 string tryPath = m->getDefaultPath() + m->getSimpleName(namefileNames[i]);
295 m->mothurOut("Unable to open " + namefileNames[i] + ". Trying default " + tryPath); m->mothurOutEndLine();
297 ableToOpen = m->openInputFile(tryPath, in2, "noerror");
299 namefileNames[i] = tryPath;
303 if (ableToOpen == 1) {
304 if (m->getOutputDir() != "") { //default path is set
305 string tryPath = m->getOutputDir() + m->getSimpleName(namefileNames[i]);
306 m->mothurOut("Unable to open " + namefileNames[i] + ". Trying output directory " + tryPath); m->mothurOutEndLine();
308 ableToOpen = m->openInputFile(tryPath, in2, "noerror");
310 namefileNames[i] = tryPath;
315 if (ableToOpen == 1) {
316 m->mothurOut("Unable to open " + namefileNames[i] + ". It will be disregarded."); m->mothurOutEndLine(); abort = true;
317 //erase from file list
318 namefileNames.erase(namefileNames.begin()+i);
321 m->setNameFile(namefileNames[i]);
327 if (namefileNames.size() != 0) { hasName = true; }
329 if (namefile != "") {
330 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(); }
333 //check for required parameters
334 countfile = validParameter.validFile(parameters, "count", false);
335 if (countfile == "not found") {
338 m->splitAtDash(countfile, countfileNames);
340 //go through files and make sure they are good, if not, then disregard them
341 for (int i = 0; i < countfileNames.size(); i++) {
344 if (countfileNames[i] == "current") {
345 countfileNames[i] = m->getCountTableFile();
346 if (countfileNames[i] != "") { m->mothurOut("Using " + countfileNames[i] + " as input file for the count parameter where you had given current."); m->mothurOutEndLine(); }
348 m->mothurOut("You have no current count file, ignoring current."); m->mothurOutEndLine(); ignore=true;
349 //erase from file list
350 countfileNames.erase(countfileNames.begin()+i);
357 if (inputDir != "") {
358 string path = m->hasPath(countfileNames[i]);
359 //if the user has not given a path then, add inputdir. else leave path alone.
360 if (path == "") { countfileNames[i] = inputDir + countfileNames[i]; }
366 ableToOpen = m->openInputFile(countfileNames[i], in, "noerror");
368 //if you can't open it, try default location
369 if (ableToOpen == 1) {
370 if (m->getDefaultPath() != "") { //default path is set
371 string tryPath = m->getDefaultPath() + m->getSimpleName(countfileNames[i]);
372 m->mothurOut("Unable to open " + countfileNames[i] + ". Trying default " + tryPath); m->mothurOutEndLine();
374 ableToOpen = m->openInputFile(tryPath, in2, "noerror");
376 countfileNames[i] = tryPath;
380 if (ableToOpen == 1) {
381 if (m->getOutputDir() != "") { //default path is set
382 string tryPath = m->getOutputDir() + m->getSimpleName(countfileNames[i]);
383 m->mothurOut("Unable to open " + countfileNames[i] + ". Trying output directory " + tryPath); m->mothurOutEndLine();
385 ableToOpen = m->openInputFile(tryPath, in2, "noerror");
387 countfileNames[i] = tryPath;
393 if (ableToOpen == 1) {
394 m->mothurOut("Unable to open " + countfileNames[i] + ". It will be disregarded."); m->mothurOutEndLine();
395 //erase from file list
396 countfileNames.erase(countfileNames.begin()+i);
399 m->setCountTableFile(countfileNames[i]);
405 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(); } }
407 //make sure there is at least one valid file left
408 if (hasName && hasCount) { m->mothurOut("[ERROR]: You must enter ONLY ONE of the following: count or name."); m->mothurOutEndLine(); abort = true; }
410 groupfile = validParameter.validFile(parameters, "group", false);
411 if (groupfile == "not found") { groupfile = ""; }
413 m->splitAtDash(groupfile, groupfileNames);
415 //go through files and make sure they are good, if not, then disregard them
416 for (int i = 0; i < groupfileNames.size(); i++) {
419 if (groupfileNames[i] == "current") {
420 groupfileNames[i] = m->getGroupFile();
421 if (groupfileNames[i] != "") { m->mothurOut("Using " + groupfileNames[i] + " as input file for the group parameter where you had given current."); m->mothurOutEndLine(); }
423 m->mothurOut("You have no current group file, ignoring current."); m->mothurOutEndLine(); ignore=true;
424 //erase from file list
425 groupfileNames.erase(groupfileNames.begin()+i);
432 if (inputDir != "") {
433 string path = m->hasPath(groupfileNames[i]);
434 cout << path << '\t' << inputDir << endl;
435 //if the user has not given a path then, add inputdir. else leave path alone.
436 if (path == "") { groupfileNames[i] = inputDir + groupfileNames[i]; }
442 ableToOpen = m->openInputFile(groupfileNames[i], in, "noerror");
444 //if you can't open it, try default location
445 if (ableToOpen == 1) {
446 if (m->getDefaultPath() != "") { //default path is set
447 string tryPath = m->getDefaultPath() + m->getSimpleName(groupfileNames[i]);
448 m->mothurOut("Unable to open " + groupfileNames[i] + ". Trying default " + tryPath); m->mothurOutEndLine();
450 ableToOpen = m->openInputFile(tryPath, in2, "noerror");
452 groupfileNames[i] = tryPath;
456 if (ableToOpen == 1) {
457 if (m->getOutputDir() != "") { //default path is set
458 string tryPath = m->getOutputDir() + m->getSimpleName(groupfileNames[i]);
459 m->mothurOut("Unable to open " + groupfileNames[i] + ". Trying output directory " + tryPath); m->mothurOutEndLine();
461 ableToOpen = m->openInputFile(tryPath, in2, "noerror");
463 groupfileNames[i] = tryPath;
469 if (ableToOpen == 1) {
470 m->mothurOut("Unable to open " + groupfileNames[i] + ". It will be disregarded."); m->mothurOutEndLine();
471 //erase from file list
472 groupfileNames.erase(groupfileNames.begin()+i);
475 m->setGroupFile(groupfileNames[i]);
482 if (groupfile != "") {
483 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(); }
484 if (hasCount) { m->mothurOut("[ERROR]: You must enter ONLY ONE of the following: count or group."); m->mothurOutEndLine(); abort = true; }
486 for (int i = 0; i < fastaFileNames.size(); i++) { groupfileNames.push_back(""); }
489 //check for optional parameter and set defaults
490 // ...at some point should added some additional type checking...
492 temp = validParameter.validFile(parameters, "processors", false); if (temp == "not found"){ temp = m->getProcessors(); }
493 m->setProcessors(temp);
494 m->mothurConvert(temp, processors);
496 temp = validParameter.validFile(parameters, "save", false); if (temp == "not found"){ temp = "f"; }
497 save = m->isTrue(temp);
499 if (save) { //clear out old references
503 //this has to go after save so that if the user sets save=t and provides no reference we abort
504 templateFileName = validParameter.validFile(parameters, "reference", true);
505 if (templateFileName == "not found") {
506 //check for saved reference sequences
507 if (rdb->referenceSeqs.size() != 0) {
508 templateFileName = "saved";
510 m->mothurOut("[ERROR]: You don't have any saved reference sequences and the reference parameter is a required for the classify.seqs command.");
511 m->mothurOutEndLine();
514 }else if (templateFileName == "not open") { abort = true; }
515 else { if (save) { rdb->setSavedReference(templateFileName); } }
517 //this has to go after save so that if the user sets save=t and provides no reference we abort
518 taxonomyFileName = validParameter.validFile(parameters, "taxonomy", true);
519 if (taxonomyFileName == "not found") {
520 //check for saved reference sequences
521 if (rdb->wordGenusProb.size() != 0) {
522 taxonomyFileName = "saved";
524 m->mothurOut("[ERROR]: You don't have any saved taxonomy information and the taxonomy parameter is a required for the classify.seqs command.");
525 m->mothurOutEndLine();
528 }else if (taxonomyFileName == "not open") { abort = true; }
529 else { if (save) { rdb->setSavedTaxonomy(taxonomyFileName); } }
531 search = validParameter.validFile(parameters, "search", false); if (search == "not found"){ search = "kmer"; }
533 method = validParameter.validFile(parameters, "method", false); if (method == "not found"){ method = "wang"; }
535 temp = validParameter.validFile(parameters, "ksize", false); if (temp == "not found"){
537 if (method == "zap") { temp = "7"; }
539 m->mothurConvert(temp, kmerSize);
541 temp = validParameter.validFile(parameters, "match", false); if (temp == "not found"){ temp = "1.0"; }
542 m->mothurConvert(temp, match);
544 temp = validParameter.validFile(parameters, "mismatch", false); if (temp == "not found"){ temp = "-1.0"; }
545 m->mothurConvert(temp, misMatch);
547 temp = validParameter.validFile(parameters, "gapopen", false); if (temp == "not found"){ temp = "-2.0"; }
548 m->mothurConvert(temp, gapOpen);
550 temp = validParameter.validFile(parameters, "gapextend", false); if (temp == "not found"){ temp = "-1.0"; }
551 m->mothurConvert(temp, gapExtend);
553 temp = validParameter.validFile(parameters, "numwanted", false); if (temp == "not found"){ temp = "10"; }
554 m->mothurConvert(temp, numWanted);
556 temp = validParameter.validFile(parameters, "cutoff", false); if (temp == "not found"){ temp = "0"; }
557 m->mothurConvert(temp, cutoff);
559 temp = validParameter.validFile(parameters, "probs", false); if (temp == "not found"){ temp = "true"; }
560 probs = m->isTrue(temp);
562 temp = validParameter.validFile(parameters, "shortcuts", false); if (temp == "not found"){ temp = "true"; }
563 writeShortcuts = m->isTrue(temp);
565 //temp = validParameter.validFile(parameters, "flip", false); if (temp == "not found"){ temp = "T"; }
566 //flip = m->isTrue(temp);
569 temp = validParameter.validFile(parameters, "iters", false); if (temp == "not found") { temp = "100"; }
570 m->mothurConvert(temp, iters);
573 if ((method == "wang") && (search != "kmer")) {
574 m->mothurOut("The wang method requires the kmer search. " + search + " will be disregarded, and kmer will be used." ); m->mothurOutEndLine();
578 if ((method == "zap") && ((search != "kmer") && (search != "align"))) {
579 m->mothurOut("The zap method requires the kmer or align search. " + search + " will be disregarded, and kmer will be used." ); m->mothurOutEndLine();
585 if (namefileNames.size() == 0){
586 if (fastaFileNames.size() != 0) {
587 vector<string> files; files.push_back(fastaFileNames[fastaFileNames.size()-1]);
588 parser.getNameFile(files);
595 catch(exception& e) {
596 m->errorOut(e, "ClassifySeqsCommand", "ClassifySeqsCommand");
601 //**********************************************************************************************************************
602 ClassifySeqsCommand::~ClassifySeqsCommand(){
603 if (abort == false) {
604 for (int i = 0; i < lines.size(); i++) { delete lines[i]; } lines.clear();
607 //**********************************************************************************************************************
609 int ClassifySeqsCommand::execute(){
611 if (abort == true) { if (calledHelp) { return 0; } return 2; }
613 string outputMethodTag = method;
614 if(method == "wang"){ classify = new Bayesian(taxonomyFileName, templateFileName, search, kmerSize, cutoff, iters, rand(), flip, writeShortcuts); }
615 else if(method == "knn"){ classify = new Knn(taxonomyFileName, templateFileName, search, kmerSize, gapOpen, gapExtend, match, misMatch, numWanted, rand()); }
616 else if(method == "zap"){
617 outputMethodTag = search + "_" + outputMethodTag;
618 if (search == "kmer") { classify = new KmerTree(templateFileName, taxonomyFileName, kmerSize, cutoff); }
619 else { classify = new AlignTree(templateFileName, taxonomyFileName, cutoff); }
622 m->mothurOut(search + " is not a valid method option. I will run the command using wang.");
623 m->mothurOutEndLine();
624 classify = new Bayesian(taxonomyFileName, templateFileName, search, kmerSize, cutoff, iters, rand(), flip, writeShortcuts);
627 if (m->control_pressed) { delete classify; return 0; }
629 for (int s = 0; s < fastaFileNames.size(); s++) {
631 m->mothurOut("Classifying sequences from " + fastaFileNames[s] + " ..." ); m->mothurOutEndLine();
633 string baseTName = m->getSimpleName(taxonomyFileName);
634 if (taxonomyFileName == "saved") { baseTName = rdb->getSavedTaxonomy(); }
636 //set rippedTaxName to
637 string RippedTaxName = "";
638 bool foundDot = false;
639 for (int i = baseTName.length()-1; i >= 0; i--) {
640 if (foundDot && (baseTName[i] != '.')) { RippedTaxName = baseTName[i] + RippedTaxName; }
641 else if (foundDot && (baseTName[i] == '.')) { break; }
642 else if (!foundDot && (baseTName[i] == '.')) { foundDot = true; }
644 //if (RippedTaxName != "") { RippedTaxName += "."; }
646 if (outputDir == "") { outputDir += m->hasPath(fastaFileNames[s]); }
647 map<string, string> variables;
648 variables["[filename]"] = outputDir + m->getRootName(m->getSimpleName(fastaFileNames[s]));
649 variables["[tag]"] = RippedTaxName;
650 variables["[tag2]"] = outputMethodTag;
651 string newTaxonomyFile = getOutputFileName("taxonomy", variables);
652 string newaccnosFile = getOutputFileName("accnos", variables);
653 string tempTaxonomyFile = outputDir + m->getRootName(m->getSimpleName(fastaFileNames[s])) + "taxonomy.temp";
654 string taxSummary = getOutputFileName("taxsummary", variables);
656 if ((method == "knn") && (search == "distance")) {
657 string DistName = getOutputFileName("matchdist", variables);
658 classify->setDistName(DistName); outputNames.push_back(DistName); outputTypes["matchdist"].push_back(DistName);
661 outputNames.push_back(newTaxonomyFile); outputTypes["taxonomy"].push_back(newTaxonomyFile);
662 outputNames.push_back(taxSummary); outputTypes["taxsummary"].push_back(taxSummary);
664 int start = time(NULL);
665 int numFastaSeqs = 0;
666 for (int i = 0; i < lines.size(); i++) { delete lines[i]; } lines.clear();
669 int pid, numSeqsPerProcessor;
671 vector<unsigned long long> MPIPos;
674 MPI_Comm_rank(MPI_COMM_WORLD, &pid); //find out who we are
675 MPI_Comm_size(MPI_COMM_WORLD, &processors);
678 MPI_File outMPINewTax;
679 MPI_File outMPITempTax;
682 int outMode=MPI_MODE_CREATE|MPI_MODE_WRONLY;
683 int inMode=MPI_MODE_RDONLY;
685 char outNewTax[1024];
686 strcpy(outNewTax, newTaxonomyFile.c_str());
688 char outTempTax[1024];
689 strcpy(outTempTax, tempTaxonomyFile.c_str());
692 strcpy(outAcc, newaccnosFile.c_str());
694 char inFileName[1024];
695 strcpy(inFileName, fastaFileNames[s].c_str());
697 MPI_File_open(MPI_COMM_WORLD, inFileName, inMode, MPI_INFO_NULL, &inMPI); //comm, filename, mode, info, filepointer
698 MPI_File_open(MPI_COMM_WORLD, outNewTax, outMode, MPI_INFO_NULL, &outMPINewTax);
699 MPI_File_open(MPI_COMM_WORLD, outTempTax, outMode, MPI_INFO_NULL, &outMPITempTax);
700 MPI_File_open(MPI_COMM_WORLD, outAcc, outMode, MPI_INFO_NULL, &outMPIAcc);
702 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; }
704 if (pid == 0) { //you are the root process
706 MPIPos = m->setFilePosFasta(fastaFileNames[s], numFastaSeqs); //fills MPIPos, returns numSeqs
708 //send file positions to all processes
709 for(int i = 1; i < processors; i++) {
710 MPI_Send(&numFastaSeqs, 1, MPI_INT, i, tag, MPI_COMM_WORLD);
711 MPI_Send(&MPIPos[0], (numFastaSeqs+1), MPI_LONG, i, tag, MPI_COMM_WORLD);
714 //figure out how many sequences you have to align
715 numSeqsPerProcessor = numFastaSeqs / processors;
716 int startIndex = pid * numSeqsPerProcessor;
717 if(pid == (processors - 1)){ numSeqsPerProcessor = numFastaSeqs - pid * numSeqsPerProcessor; }
721 driverMPI(startIndex, numSeqsPerProcessor, inMPI, outMPINewTax, outMPITempTax, outMPIAcc, MPIPos);
723 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; }
725 for (int i = 1; i < processors; i++) {
727 MPI_Recv(&done, 1, MPI_INT, i, tag, MPI_COMM_WORLD, &status);
729 }else{ //you are a child process
730 MPI_Recv(&numFastaSeqs, 1, MPI_INT, 0, tag, MPI_COMM_WORLD, &status);
731 MPIPos.resize(numFastaSeqs+1);
732 MPI_Recv(&MPIPos[0], (numFastaSeqs+1), MPI_LONG, 0, tag, MPI_COMM_WORLD, &status);
734 //figure out how many sequences you have to align
735 numSeqsPerProcessor = numFastaSeqs / processors;
736 int startIndex = pid * numSeqsPerProcessor;
737 if(pid == (processors - 1)){ numSeqsPerProcessor = numFastaSeqs - pid * numSeqsPerProcessor; }
741 driverMPI(startIndex, numSeqsPerProcessor, inMPI, outMPINewTax, outMPITempTax, outMPIAcc, MPIPos);
743 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; }
746 MPI_Send(&done, 1, MPI_INT, 0, tag, MPI_COMM_WORLD);
750 MPI_File_close(&inMPI);
751 MPI_File_close(&outMPINewTax);
752 MPI_File_close(&outMPITempTax);
753 MPI_File_close(&outMPIAcc);
754 MPI_Barrier(MPI_COMM_WORLD); //make everyone wait - just in case
758 vector<unsigned long long> positions;
759 #if defined (__APPLE__) || (__MACH__) || (linux) || (__linux) || (__linux__) || (__unix__) || (__unix)
760 positions = m->divideFile(fastaFileNames[s], processors);
761 for (int i = 0; i < (positions.size()-1); i++) { lines.push_back(new linePair(positions[i], positions[(i+1)])); }
763 if (processors == 1) {
764 lines.push_back(new linePair(0, 1000));
766 positions = m->setFilePosFasta(fastaFileNames[s], numFastaSeqs);
767 if (positions.size() < processors) { processors = positions.size(); }
769 //figure out how many sequences you have to process
770 int numSeqsPerProcessor = numFastaSeqs / processors;
771 for (int i = 0; i < processors; i++) {
772 int startIndex = i * numSeqsPerProcessor;
773 if(i == (processors - 1)){ numSeqsPerProcessor = numFastaSeqs - i * numSeqsPerProcessor; }
774 lines.push_back(new linePair(positions[startIndex], numSeqsPerProcessor));
779 numFastaSeqs = driver(lines[0], newTaxonomyFile, tempTaxonomyFile, newaccnosFile, fastaFileNames[s]);
781 numFastaSeqs = createProcesses(newTaxonomyFile, tempTaxonomyFile, newaccnosFile, fastaFileNames[s]);
785 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();
786 outputNames.push_back(newaccnosFile); outputTypes["accnos"].push_back(newaccnosFile);
787 }else { m->mothurRemove(newaccnosFile); }
789 m->mothurOutEndLine();
790 m->mothurOut("It took " + toString(time(NULL) - start) + " secs to classify " + toString(numFastaSeqs) + " sequences."); m->mothurOutEndLine(); m->mothurOutEndLine();
796 if (pid == 0) { //this part does not need to be paralellized
798 if(namefile != "") { m->mothurOut("Reading " + namefileNames[s] + "..."); cout.flush(); MPIReadNamesFile(namefileNames[s]); m->mothurOut(" Done."); m->mothurOutEndLine(); }
803 m->mothurOut("Reading " + namefileNames[s] + "..."); cout.flush();
804 nameMap.clear(); //remove old names
805 m->readNames(namefileNames[s], nameMap);
806 m->mothurOut(" Done."); m->mothurOutEndLine();
811 GroupMap* groupMap = NULL;
812 CountTable* ct = NULL;
813 PhyloSummary* taxaSum;
815 ct = new CountTable();
816 ct->readTable(countfileNames[s]);
817 taxaSum = new PhyloSummary(taxonomyFileName, ct);
818 taxaSum->summarize(tempTaxonomyFile);
820 if (groupfile != "") { group = groupfileNames[s]; groupMap = new GroupMap(group); groupMap->readMap(); }
822 taxaSum = new PhyloSummary(taxonomyFileName, groupMap);
824 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; }
826 if (namefile == "") { taxaSum->summarize(tempTaxonomyFile); }
829 m->openInputFile(tempTaxonomyFile, in);
831 //read in users taxonomy file and add sequences to tree
835 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; }
837 in >> name >> taxon; m->gobble(in);
839 itNames = nameMap.find(name);
841 if (itNames == nameMap.end()) {
842 m->mothurOut(name + " is not in your name file please correct."); m->mothurOutEndLine(); exit(1);
844 for (int i = 0; i < itNames->second.size(); i++) {
845 taxaSum->addSeqToTree(itNames->second[i], taxon); //add it as many times as there are identical seqs
847 itNames->second.clear();
848 nameMap.erase(itNames->first);
854 m->mothurRemove(tempTaxonomyFile);
856 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; }
860 m->openOutputFile(taxSummary, outTaxTree);
861 taxaSum->print(outTaxTree);
864 //output taxonomy with the unclassified bins added
866 m->openInputFile(newTaxonomyFile, inTax);
869 string unclass = newTaxonomyFile + ".unclass.temp";
870 m->openOutputFile(unclass, outTax);
872 //get maxLevel from phylotree so you know how many 'unclassified's to add
873 int maxLevel = taxaSum->getMaxLevel();
875 //read taxfile - this reading and rewriting is done to preserve the confidence scores.
877 while (!inTax.eof()) {
878 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; }
880 inTax >> name >> taxon; m->gobble(inTax);
882 string newTax = addUnclassifieds(taxon, maxLevel);
884 outTax << name << '\t' << newTax << endl;
889 if (ct != NULL) { delete ct; }
890 if (groupMap != NULL) { delete groupMap; } delete taxaSum;
891 m->mothurRemove(newTaxonomyFile);
892 rename(unclass.c_str(), newTaxonomyFile.c_str());
894 m->mothurOutEndLine();
895 m->mothurOut("It took " + toString(time(NULL) - start) + " secs to create the summary file for " + toString(numFastaSeqs) + " sequences."); m->mothurOutEndLine(); m->mothurOutEndLine();
903 m->mothurOutEndLine();
904 m->mothurOut("Output File Names: "); m->mothurOutEndLine();
905 for (int i = 0; i < outputNames.size(); i++) { m->mothurOut(outputNames[i]); m->mothurOutEndLine(); }
906 m->mothurOutEndLine();
908 //set taxonomy file as new current taxonomyfile
910 itTypes = outputTypes.find("taxonomy");
911 if (itTypes != outputTypes.end()) {
912 if ((itTypes->second).size() != 0) { current = (itTypes->second)[0]; m->setTaxonomyFile(current); }
916 itTypes = outputTypes.find("accnos");
917 if (itTypes != outputTypes.end()) {
918 if ((itTypes->second).size() != 0) { current = (itTypes->second)[0]; m->setAccnosFile(current); }
925 catch(exception& e) {
926 m->errorOut(e, "ClassifySeqsCommand", "execute");
931 /**************************************************************************************************/
932 string ClassifySeqsCommand::addUnclassifieds(string tax, int maxlevel) {
934 string newTax, taxon;
937 //keep what you have counting the levels
938 while (tax.find_first_of(';') != -1) {
940 taxon = tax.substr(0,tax.find_first_of(';'))+';';
941 tax = tax.substr(tax.find_first_of(';')+1, tax.length());
946 //add "unclassified" until you reach maxLevel
947 while (level < maxlevel) {
948 newTax += "unclassified;";
954 catch(exception& e) {
955 m->errorOut(e, "ClassifySeqsCommand", "addUnclassifieds");
960 /**************************************************************************************************/
962 int ClassifySeqsCommand::createProcesses(string taxFileName, string tempTaxFile, string accnos, string filename) {
968 #if defined (__APPLE__) || (__MACH__) || (linux) || (__linux) || (__linux__) || (__unix__) || (__unix)
971 //loop through and create all the processes you want
972 while (process != processors) {
976 processIDS.push_back(pid); //create map from line number to pid so you can append files in correct order later
979 num = driver(lines[process], taxFileName + toString(getpid()) + ".temp", tempTaxFile + toString(getpid()) + ".temp", accnos + toString(getpid()) + ".temp", filename);
981 //pass numSeqs to parent
983 string tempFile = filename + toString(getpid()) + ".num.temp";
984 m->openOutputFile(tempFile, out);
990 m->mothurOut("[ERROR]: unable to spawn the necessary processes."); m->mothurOutEndLine();
991 for (int i = 0; i < processIDS.size(); i++) { kill (processIDS[i], SIGINT); }
996 //parent does its part
997 num = driver(lines[0], taxFileName, tempTaxFile, accnos, filename);
999 //force parent to wait until all the processes are done
1000 for (int i=0;i<processIDS.size();i++) {
1001 int temp = processIDS[i];
1005 for (int i = 0; i < processIDS.size(); i++) {
1007 string tempFile = filename + toString(processIDS[i]) + ".num.temp";
1008 m->openInputFile(tempFile, in);
1009 if (!in.eof()) { int tempNum = 0; in >> tempNum; num += tempNum; }
1010 in.close(); m->mothurRemove(m->getFullPathName(tempFile));
1013 //////////////////////////////////////////////////////////////////////////////////////////////////////
1014 //Windows version shared memory, so be careful when passing variables through the alignData struct.
1015 //Above fork() will clone, so memory is separate, but that's not the case with windows,
1016 //////////////////////////////////////////////////////////////////////////////////////////////////////
1018 vector<classifyData*> pDataArray;
1019 DWORD dwThreadIdArray[processors-1];
1020 HANDLE hThreadArray[processors-1];
1022 //Create processor worker threads.
1023 for( int i=0; i<processors-1; i++ ){
1024 // Allocate memory for thread data.
1025 string extension = "";
1026 if (i != 0) { extension = toString(i) + ".temp"; processIDS.push_back(i); }
1028 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);
1029 pDataArray.push_back(tempclass);
1031 //MySeqSumThreadFunction is in header. It must be global or static to work with the threads.
1032 //default security attributes, thread function name, argument to thread function, use default creation flags, returns the thread identifier
1033 hThreadArray[i] = CreateThread(NULL, 0, MyClassThreadFunction, pDataArray[i], 0, &dwThreadIdArray[i]);
1037 //parent does its part
1038 num = driver(lines[processors-1], taxFileName + toString(processors-1) + ".temp", tempTaxFile + toString(processors-1) + ".temp", accnos + toString(processors-1) + ".temp", filename);
1039 processIDS.push_back((processors-1));
1041 //Wait until all threads have terminated.
1042 WaitForMultipleObjects(processors-1, hThreadArray, TRUE, INFINITE);
1044 //Close all thread handles and free memory allocations.
1045 for(int i=0; i < pDataArray.size(); i++){
1046 num += pDataArray[i]->count;
1047 if (pDataArray[i]->count != pDataArray[i]->end) {
1048 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;
1050 CloseHandle(hThreadArray[i]);
1051 delete pDataArray[i];
1055 vector<string> nonBlankAccnosFiles;
1056 if (!(m->isBlank(accnos))) { nonBlankAccnosFiles.push_back(accnos); }
1057 else { m->mothurRemove(accnos); } //remove so other files can be renamed to it
1059 for(int i=0;i<processIDS.size();i++){
1060 m->appendFiles((taxFileName + toString(processIDS[i]) + ".temp"), taxFileName);
1061 m->appendFiles((tempTaxFile + toString(processIDS[i]) + ".temp"), tempTaxFile);
1062 if (!(m->isBlank(accnos + toString(processIDS[i]) + ".temp"))) {
1063 nonBlankAccnosFiles.push_back(accnos + toString(processIDS[i]) + ".temp");
1064 }else { m->mothurRemove((accnos + toString(processIDS[i]) + ".temp")); }
1066 m->mothurRemove((m->getFullPathName(taxFileName) + toString(processIDS[i]) + ".temp"));
1067 m->mothurRemove((m->getFullPathName(tempTaxFile) + toString(processIDS[i]) + ".temp"));
1070 //append accnos files
1071 if (nonBlankAccnosFiles.size() != 0) {
1072 rename(nonBlankAccnosFiles[0].c_str(), accnos.c_str());
1074 for (int h=1; h < nonBlankAccnosFiles.size(); h++) {
1075 m->appendFiles(nonBlankAccnosFiles[h], accnos);
1076 m->mothurRemove(nonBlankAccnosFiles[h]);
1078 }else { //recreate the accnosfile if needed
1080 m->openOutputFile(accnos, out);
1087 catch(exception& e) {
1088 m->errorOut(e, "ClassifySeqsCommand", "createProcesses");
1092 //**********************************************************************************************************************
1094 int ClassifySeqsCommand::driver(linePair* filePos, string taxFName, string tempTFName, string accnos, string filename){
1097 m->openOutputFile(taxFName, outTax);
1099 ofstream outTaxSimple;
1100 m->openOutputFile(tempTFName, outTaxSimple);
1103 m->openOutputFile(accnos, outAcc);
1106 m->openInputFile(filename, inFASTA);
1110 inFASTA.seekg(filePos->start);
1116 if (m->control_pressed) {
1119 outTaxSimple.close();
1120 outAcc.close(); return 0; }
1122 Sequence* candidateSeq = new Sequence(inFASTA); m->gobble(inFASTA);
1124 if (candidateSeq->getName() != "") {
1126 taxonomy = classify->getTaxonomy(candidateSeq);
1128 if (m->control_pressed) { delete candidateSeq; return 0; }
1130 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(); }
1132 //output confidence scores or not
1134 outTax << candidateSeq->getName() << '\t' << taxonomy << endl;
1136 outTax << candidateSeq->getName() << '\t' << classify->getSimpleTax() << endl;
1139 if (classify->getFlipped()) { outAcc << candidateSeq->getName() << endl; }
1141 outTaxSimple << candidateSeq->getName() << '\t' << classify->getSimpleTax() << endl;
1145 delete candidateSeq;
1147 #if defined (__APPLE__) || (__MACH__) || (linux) || (__linux) || (__linux__) || (__unix__) || (__unix)
1148 unsigned long long pos = inFASTA.tellg();
1149 if ((pos == -1) || (pos >= filePos->end)) { break; }
1151 if (inFASTA.eof()) { break; }
1155 if((count) % 100 == 0){ m->mothurOut("Processing sequence: " + toString(count)); m->mothurOutEndLine(); }
1159 if((count) % 100 != 0){ m->mothurOut("Processing sequence: " + toString(count)); m->mothurOutEndLine(); }
1163 outTaxSimple.close();
1168 catch(exception& e) {
1169 m->errorOut(e, "ClassifySeqsCommand", "driver");
1173 //**********************************************************************************************************************
1175 int ClassifySeqsCommand::driverMPI(int start, int num, MPI_File& inMPI, MPI_File& newFile, MPI_File& tempFile, MPI_File& accFile, vector<unsigned long long>& MPIPos){
1177 MPI_Status statusNew;
1178 MPI_Status statusTemp;
1179 MPI_Status statusAcc;
1183 MPI_Comm_rank(MPI_COMM_WORLD, &pid); //find out who we are
1186 string outputString;
1188 for(int i=0;i<num;i++){
1190 if (m->control_pressed) { return 0; }
1192 //read next sequence
1193 int length = MPIPos[start+i+1] - MPIPos[start+i];
1194 char* buf4 = new char[length];
1195 MPI_File_read_at(inMPI, MPIPos[start+i], buf4, length, MPI_CHAR, &status);
1197 string tempBuf = buf4;
1198 if (tempBuf.length() > length) { tempBuf = tempBuf.substr(0, length); }
1199 istringstream iss (tempBuf,istringstream::in);
1202 Sequence* candidateSeq = new Sequence(iss);
1204 if (candidateSeq->getName() != "") {
1205 taxonomy = classify->getTaxonomy(candidateSeq);
1207 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(); }
1209 //output confidence scores or not
1211 outputString = candidateSeq->getName() + "\t" + taxonomy + "\n";
1213 outputString = candidateSeq->getName() + "\t" + classify->getSimpleTax() + "\n";
1216 int length = outputString.length();
1217 char* buf2 = new char[length];
1218 memcpy(buf2, outputString.c_str(), length);
1220 MPI_File_write_shared(newFile, buf2, length, MPI_CHAR, &statusNew);
1223 outputString = candidateSeq->getName() + "\t" + classify->getSimpleTax() + "\n";
1224 length = outputString.length();
1225 char* buf = new char[length];
1226 memcpy(buf, outputString.c_str(), length);
1228 MPI_File_write_shared(tempFile, buf, length, MPI_CHAR, &statusTemp);
1231 if (classify->getFlipped()) {
1232 outputString = candidateSeq->getName() + "\n";
1233 length = outputString.length();
1234 char* buf3 = new char[length];
1235 memcpy(buf3, outputString.c_str(), length);
1237 MPI_File_write_shared(accFile, buf3, length, MPI_CHAR, &statusAcc);
1242 delete candidateSeq;
1244 if((i+1) % 100 == 0){ cout << "Classifying sequence " << (i+1) << endl; }
1247 if(num % 100 != 0){ cout << "Classifying sequence " << (num) << endl; }
1252 catch(exception& e) {
1253 m->errorOut(e, "ClassifySeqsCommand", "driverMPI");
1258 //**********************************************************************************************************************
1259 int ClassifySeqsCommand::MPIReadNamesFile(string nameFilename){
1262 nameMap.clear(); //remove old names
1268 //char* inFileName = new char[nameFilename.length()];
1269 //memcpy(inFileName, nameFilename.c_str(), nameFilename.length());
1271 char inFileName[1024];
1272 strcpy(inFileName, nameFilename.c_str());
1274 MPI_File_open(MPI_COMM_WORLD, inFileName, MPI_MODE_RDONLY, MPI_INFO_NULL, &inMPI);
1275 MPI_File_get_size(inMPI, &size);
1276 //delete inFileName;
1278 char* buffer = new char[size];
1279 MPI_File_read(inMPI, buffer, size, MPI_CHAR, &status);
1281 string tempBuf = buffer;
1282 if (tempBuf.length() > size) { tempBuf = tempBuf.substr(0, size); }
1283 istringstream iss (tempBuf,istringstream::in);
1286 string firstCol, secondCol;
1288 iss >> firstCol >> secondCol; m->gobble(iss);
1290 vector<string> temp;
1291 m->splitAtComma(secondCol, temp);
1293 nameMap[firstCol] = temp;
1296 MPI_File_close(&inMPI);
1300 catch(exception& e) {
1301 m->errorOut(e, "ClassifySeqsCommand", "MPIReadNamesFile");
1306 /**************************************************************************************************/