]> git.donarmstrong.com Git - mothur.git/blob - treegroupscommand.cpp
Merge remote-tracking branch 'mothur/master'
[mothur.git] / treegroupscommand.cpp
1 /*
2  *  treegroupscommand.cpp
3  *  Mothur
4  *
5  *  Created by Sarah Westcott on 4/8/09.
6  *  Copyright 2009 Schloss Lab UMASS Amherst. All rights reserved.
7  *
8  */
9
10 #include "treegroupscommand.h"
11 #include "subsample.h"
12 #include "consensus.h"
13
14 //**********************************************************************************************************************
15 vector<string> TreeGroupCommand::setParameters(){       
16         try {
17                 CommandParameter pshared("shared", "InputTypes", "", "", "PhylipColumnShared", "PhylipColumnShared", "none",false,false); parameters.push_back(pshared);
18                 CommandParameter pphylip("phylip", "InputTypes", "", "", "PhylipColumnShared", "PhylipColumnShared", "none",false,false); parameters.push_back(pphylip);
19                 CommandParameter pname("name", "InputTypes", "", "", "none", "none", "ColumnName",false,false); parameters.push_back(pname);
20                 CommandParameter pcolumn("column", "InputTypes", "", "", "PhylipColumnShared", "PhylipColumnShared", "ColumnName",false,false); parameters.push_back(pcolumn);  
21         CommandParameter piters("iters", "Number", "", "1000", "", "", "",false,false); parameters.push_back(piters);
22         CommandParameter psubsample("subsample", "String", "", "", "", "", "",false,false); parameters.push_back(psubsample);
23         CommandParameter pcutoff("cutoff", "Number", "", "10", "", "", "",false,false); parameters.push_back(pcutoff);
24                 CommandParameter pprecision("precision", "Number", "", "100", "", "", "",false,false); parameters.push_back(pprecision);                
25                 CommandParameter plabel("label", "String", "", "", "", "", "",false,false); parameters.push_back(plabel);
26                 CommandParameter pgroups("groups", "String", "", "", "", "", "",false,false); parameters.push_back(pgroups);
27                 CommandParameter pcalc("calc", "Multiple", "sharedsobs-sharedchao-sharedace-jabund-sorabund-jclass-sorclass-jest-sorest-thetayc-thetan-kstest-sharednseqs-ochiai-anderberg-kulczynski-kulczynskicody-lennon-morisitahorn-braycurtis-whittaker-odum-canberra-structeuclidean-structchord-hellinger-manhattan-structpearson-soergel-spearman-structkulczynski-speciesprofile-hamming-structchi2-gower-memchi2-memchord-memeuclidean-mempearson", "jclass-thetayc", "", "", "",true,false); parameters.push_back(pcalc);
28                 
29         CommandParameter pprocessors("processors", "Number", "", "1", "", "", "",false,false); parameters.push_back(pprocessors);
30 //CommandParameter poutput("output", "Multiple", "lt-square", "lt", "", "", "",false,false); parameters.push_back(poutput);
31                 CommandParameter pinputdir("inputdir", "String", "", "", "", "", "",false,false); parameters.push_back(pinputdir);
32                 CommandParameter poutputdir("outputdir", "String", "", "", "", "", "",false,false); parameters.push_back(poutputdir);
33                 
34                 vector<string> myArray;
35                 for (int i = 0; i < parameters.size(); i++) {   myArray.push_back(parameters[i].name);          }
36                 return myArray;
37         }
38         catch(exception& e) {
39                 m->errorOut(e, "TreeGroupCommand", "setParameters");
40                 exit(1);
41         }
42 }
43 //**********************************************************************************************************************
44 string TreeGroupCommand::getHelpString(){       
45         try {
46                 string helpString = "";
47                 ValidCalculators validCalculator;
48                 helpString += "The tree.shared command creates a .tre to represent the similiarity between groups or sequences.\n";
49                 helpString += "The tree.shared command parameters are shared, groups, calc, phylip, column, name, cutoff, precision, processors, subsample, iters and label.\n";
50                 helpString += "The groups parameter allows you to specify which of the groups in your groupfile you would like included used.\n";
51                 helpString += "The group names are separated by dashes. The label allow you to select what distance levels you would like trees created for, and are also separated by dashes.\n";
52                 helpString += "The phylip or column parameter are required if you do not provide a sharedfile, and only one may be used.  If you use a column file the name filename is required. \n";
53                 helpString += "If you do not provide a cutoff value 10.00 is assumed. If you do not provide a precision value then 100 is assumed.\n";
54                 helpString += "The tree.shared command should be in the following format: tree.shared(groups=yourGroups, calc=yourCalcs, label=yourLabels).\n";
55         helpString += "The iters parameter allows you to choose the number of times you would like to run the subsample.\n";
56         helpString += "The subsample parameter allows you to enter the size pergroup of the sample or you can set subsample=T and mothur will use the size of your smallest group. The subsample parameter may only be used with a shared file.\n";
57                 helpString += "Example tree.shared(groups=A-B-C, calc=jabund-sorabund).\n";
58                 helpString += "The default value for groups is all the groups in your groupfile.\n";
59                 helpString += "The default value for calc is jclass-thetayc.\n";
60                 helpString += "The tree.shared command outputs a .tre file for each calculator you specify at each distance you choose.\n";
61                 helpString += validCalculator.printCalc("treegroup");
62                 helpString += "Or the tree.shared command can be in the following format: tree.shared(phylip=yourPhylipFile).\n";
63                 helpString += "Example tree.shared(phylip=abrecovery.dist).\n";
64                 helpString += "Note: No spaces between parameter labels (i.e. groups), '=' and parameters (i.e.yourGroups).\n";
65                 return helpString;
66         }
67         catch(exception& e) {
68                 m->errorOut(e, "TreeGroupCommand", "getHelpString");
69                 exit(1);
70         }
71 }
72 //**********************************************************************************************************************
73 string TreeGroupCommand::getOutputFileNameTag(string type, string inputName=""){        
74         try {
75         string outputFileName = "";
76                 map<string, vector<string> >::iterator it;
77         
78         //is this a type this command creates
79         it = outputTypes.find(type);
80         if (it == outputTypes.end()) {  m->mothurOut("[ERROR]: this command doesn't create a " + type + " output file.\n"); }
81         else {
82             if (type == "tree")            {   outputFileName =  "tre";   }
83             else { m->mothurOut("[ERROR]: No definition for type " + type + " output file tag.\n"); m->control_pressed = true;  }
84         }
85         return outputFileName;
86         }
87         catch(exception& e) {
88                 m->errorOut(e, "TreeGroupCommand", "getOutputFileNameTag");
89                 exit(1);
90         }
91 }
92 //**********************************************************************************************************************
93 TreeGroupCommand::TreeGroupCommand(){   
94         try {
95                 abort = true; calledHelp = true;
96                 setParameters();
97                 //initialize outputTypes
98                 vector<string> tempOutNames;
99                 outputTypes["tree"] = tempOutNames;
100         }
101         catch(exception& e) {
102                 m->errorOut(e, "TreeGroupCommand", "TreeGroupCommand");
103                 exit(1);
104         }
105 }
106 //**********************************************************************************************************************
107
108 TreeGroupCommand::TreeGroupCommand(string option)  {
109         try {
110                 abort = false; calledHelp = false;   
111                 allLines = 1;
112                 
113                 //allow user to run help
114                 if(option == "help") { help(); abort = true; calledHelp = true; }
115                 else if(option == "citation") { citation(); abort = true; calledHelp = true;}
116                 
117                 else {
118                         vector<string> myArray = setParameters();
119                         
120                         OptionParser parser(option);
121                         map<string, string> parameters = parser. getParameters();
122                         
123                         ValidParameters validParameter;
124                         map<string, string>::iterator it;
125                 
126                         //check to make sure all parameters are valid for command
127                         for (it = parameters.begin(); it != parameters.end(); it++) { 
128                                 if (validParameter.isValidParameter(it->first, myArray, it->second) != true) {  abort = true;  }
129                         }
130                         
131                         //initialize outputTypes
132                         vector<string> tempOutNames;
133                         outputTypes["tree"] = tempOutNames;
134                         
135                         //if the user changes the input directory command factory will send this info to us in the output parameter 
136                         string inputDir = validParameter.validFile(parameters, "inputdir", false);              
137                         if (inputDir == "not found"){   inputDir = "";          }
138                         else {
139                                 string path;
140                                 it = parameters.find("phylip");
141                                 //user has given a template file
142                                 if(it != parameters.end()){ 
143                                         path = m->hasPath(it->second);
144                                         //if the user has not given a path then, add inputdir. else leave path alone.
145                                         if (path == "") {       parameters["phylip"] = inputDir + it->second;           }
146                                 }
147                                 
148                                 it = parameters.find("column");
149                                 //user has given a template file
150                                 if(it != parameters.end()){ 
151                                         path = m->hasPath(it->second);
152                                         //if the user has not given a path then, add inputdir. else leave path alone.
153                                         if (path == "") {       parameters["column"] = inputDir + it->second;           }
154                                 }
155                                 
156                                 it = parameters.find("name");
157                                 //user has given a template file
158                                 if(it != parameters.end()){ 
159                                         path = m->hasPath(it->second);
160                                         //if the user has not given a path then, add inputdir. else leave path alone.
161                                         if (path == "") {       parameters["name"] = inputDir + it->second;             }
162                                 }
163                         }
164                         
165                         //check for required parameters
166                         phylipfile = validParameter.validFile(parameters, "phylip", true);
167                         if (phylipfile == "not open") { phylipfile = ""; abort = true; }
168                         else if (phylipfile == "not found") { phylipfile = ""; }        
169                         else {  inputfile = phylipfile;  format = "phylip"; m->setPhylipFile(phylipfile);       }
170                         
171                         columnfile = validParameter.validFile(parameters, "column", true);
172                         if (columnfile == "not open") { columnfile = ""; abort = true; }        
173                         else if (columnfile == "not found") { columnfile = ""; }
174                         else {  inputfile = columnfile; format = "column";      m->setColumnFile(columnfile); }
175                         
176                         sharedfile = validParameter.validFile(parameters, "shared", true);
177                         if (sharedfile == "not open") { sharedfile = ""; abort = true; }        
178                         else if (sharedfile == "not found") { sharedfile = ""; }
179                         else {  inputfile = sharedfile; format = "sharedfile";  m->setSharedFile(sharedfile); }
180                         
181                         namefile = validParameter.validFile(parameters, "name", true);
182                         if (namefile == "not open") { abort = true; }   
183                         else if (namefile == "not found") { namefile = ""; }
184                         else { m->setNameFile(namefile); }
185                         
186                         if ((phylipfile == "") && (columnfile == "") && (sharedfile == "")) { 
187                                 //is there are current file available for either of these?
188                                 //give priority to shared, then column, then phylip
189                                 sharedfile = m->getSharedFile(); 
190                                 if (sharedfile != "") {  inputfile = sharedfile; format = "sharedfile"; m->mothurOut("Using " + sharedfile + " as input file for the shared parameter."); m->mothurOutEndLine(); }
191                                 else { 
192                                         columnfile = m->getColumnFile(); 
193                                         if (columnfile != "") { inputfile = columnfile; format = "column";  m->mothurOut("Using " + columnfile + " as input file for the column parameter."); m->mothurOutEndLine(); }
194                                         else { 
195                                                 phylipfile = m->getPhylipFile(); 
196                                                 if (phylipfile != "") { inputfile = phylipfile;  format = "phylip";  m->mothurOut("Using " + phylipfile + " as input file for the phylip parameter."); m->mothurOutEndLine(); }
197                                                 else { 
198                                                         m->mothurOut("No valid current files. You must provide a shared, phylip or column file."); m->mothurOutEndLine(); 
199                                                         abort = true;
200                                                 }
201                                         }
202                                 }
203                         }
204                         else if ((phylipfile != "") && (columnfile != "")) { m->mothurOut("When running the tree.shared command with a distance file you may not use both the column and the phylip parameters."); m->mothurOutEndLine(); abort = true; }
205                         
206                         if (columnfile != "") {
207                                 if (namefile == "") { 
208                                         namefile = m->getNameFile(); 
209                                         if (namefile != "") {  m->mothurOut("Using " + namefile + " as input file for the name parameter."); m->mothurOutEndLine(); }
210                                         else { 
211                                                 m->mothurOut("You need to provide a namefile if you are going to use the column format."); m->mothurOutEndLine(); 
212                                                 abort = true; 
213                                         }       
214                                 }
215                         }
216                         
217                         //check for optional parameter and set defaults
218                         // ...at some point should added some additional type checking...
219                         label = validParameter.validFile(parameters, "label", false);                   
220                         if (label == "not found") { label = ""; }
221                         else { 
222                                 if(label != "all") {  m->splitAtDash(label, labels);  allLines = 0;  }
223                                 else { allLines = 1;  }
224                         }
225                         
226                         groups = validParameter.validFile(parameters, "groups", false);                 
227                         if (groups == "not found") { groups = ""; }
228                         else { 
229                                 m->splitAtDash(groups, Groups);
230                                 m->setGroups(Groups);
231                         }
232                                 
233                         calc = validParameter.validFile(parameters, "calc", false);                     
234                         if (calc == "not found") { calc = "jclass-thetayc";  }
235                         else { 
236                                  if (calc == "default")  {  calc = "jclass-thetayc";  }
237                         }
238                         m->splitAtDash(calc, Estimators);
239                         if (m->inUsersGroups("citation", Estimators)) { 
240                                 ValidCalculators validCalc; validCalc.printCitations(Estimators); 
241                                 //remove citation from list of calcs
242                                 for (int i = 0; i < Estimators.size(); i++) { if (Estimators[i] == "citation") {  Estimators.erase(Estimators.begin()+i); break; } }
243                         }
244
245                         string temp;
246                         temp = validParameter.validFile(parameters, "precision", false);                        if (temp == "not found") { temp = "100"; }
247                         m->mothurConvert(temp, precision); 
248                         
249                         temp = validParameter.validFile(parameters, "cutoff", false);                   if (temp == "not found") { temp = "10"; }
250                         m->mothurConvert(temp, cutoff); 
251                         cutoff += (5 / (precision * 10.0));
252                         
253             temp = validParameter.validFile(parameters, "processors", false);   if (temp == "not found"){       temp = m->getProcessors();      }
254                         m->setProcessors(temp);
255                         m->mothurConvert(temp, processors); 
256             
257             temp = validParameter.validFile(parameters, "iters", false);                        if (temp == "not found") { temp = "1000"; }
258                         m->mothurConvert(temp, iters); 
259             
260             temp = validParameter.validFile(parameters, "subsample", false);            if (temp == "not found") { temp = "F"; }
261                         if (m->isNumeric1(temp)) { m->mothurConvert(temp, subsampleSize); subsample = true; }
262             else {  
263                 if (m->isTrue(temp)) { subsample = true; subsampleSize = -1; }  //we will set it to smallest group later 
264                 else { subsample = false; }
265             }
266             
267             if (subsample == false) { iters = 1; }
268             
269             if (subsample && (format != "sharedfile")) { m->mothurOut("[ERROR]: the subsample parameter can only be used with a shared file.\n"); abort=true; }
270             
271                         //if the user changes the output directory command factory will send this info to us in the output parameter 
272                         outputDir = validParameter.validFile(parameters, "outputdir", false);           if (outputDir == "not found"){  
273                                 outputDir = ""; 
274                                 outputDir += m->hasPath(inputfile); //if user entered a file with a path then preserve it       
275                         }
276                 }
277
278         }
279         catch(exception& e) {
280                 m->errorOut(e, "TreeGroupCommand", "TreeGroupCommand");
281                 exit(1);
282         }
283 }
284 //**********************************************************************************************************************
285
286 TreeGroupCommand::~TreeGroupCommand(){
287         if (abort == false) {
288                 if (format == "sharedfile") {  delete input; }
289                 else { delete list; }
290                 delete tmap;  
291         }
292         
293 }
294
295 //**********************************************************************************************************************
296
297 int TreeGroupCommand::execute(){
298         try {
299         
300                 if (abort == true) { if (calledHelp) { return 0; }  return 2;   }
301                 
302                 if (format == "sharedfile") {
303                         
304                         ValidCalculators validCalculator;
305                 
306                         for (int i=0; i<Estimators.size(); i++) {
307                                 if (validCalculator.isValidCalculator("treegroup", Estimators[i]) == true) { 
308                                         if (Estimators[i] == "sharedsobs") { 
309                                                 treeCalculators.push_back(new SharedSobsCS());
310                                         }else if (Estimators[i] == "sharedchao") { 
311                                                 treeCalculators.push_back(new SharedChao1());
312                                         }else if (Estimators[i] == "sharedace") { 
313                                                 treeCalculators.push_back(new SharedAce());
314                                         }else if (Estimators[i] == "jabund") {  
315                                                 treeCalculators.push_back(new JAbund());
316                                         }else if (Estimators[i] == "sorabund") { 
317                                                 treeCalculators.push_back(new SorAbund());
318                                         }else if (Estimators[i] == "jclass") { 
319                                                 treeCalculators.push_back(new Jclass());
320                                         }else if (Estimators[i] == "sorclass") { 
321                                                 treeCalculators.push_back(new SorClass());
322                                         }else if (Estimators[i] == "jest") { 
323                                                 treeCalculators.push_back(new Jest());
324                                         }else if (Estimators[i] == "sorest") { 
325                                                 treeCalculators.push_back(new SorEst());
326                                         }else if (Estimators[i] == "thetayc") { 
327                                                 treeCalculators.push_back(new ThetaYC());
328                                         }else if (Estimators[i] == "thetan") { 
329                                                 treeCalculators.push_back(new ThetaN());
330                                         }else if (Estimators[i] == "kstest") { 
331                                                 treeCalculators.push_back(new KSTest());
332                                         }else if (Estimators[i] == "sharednseqs") { 
333                                                 treeCalculators.push_back(new SharedNSeqs());
334                                         }else if (Estimators[i] == "ochiai") { 
335                                                 treeCalculators.push_back(new Ochiai());
336                                         }else if (Estimators[i] == "anderberg") { 
337                                                 treeCalculators.push_back(new Anderberg());
338                                         }else if (Estimators[i] == "kulczynski") { 
339                                                 treeCalculators.push_back(new Kulczynski());
340                                         }else if (Estimators[i] == "kulczynskicody") { 
341                                                 treeCalculators.push_back(new KulczynskiCody());
342                                         }else if (Estimators[i] == "lennon") { 
343                                                 treeCalculators.push_back(new Lennon());
344                                         }else if (Estimators[i] == "morisitahorn") { 
345                                                 treeCalculators.push_back(new MorHorn());
346                                         }else if (Estimators[i] == "braycurtis") { 
347                                                 treeCalculators.push_back(new BrayCurtis());
348                                         }else if (Estimators[i] == "whittaker") { 
349                                                 treeCalculators.push_back(new Whittaker());
350                                         }else if (Estimators[i] == "odum") { 
351                                                 treeCalculators.push_back(new Odum());
352                                         }else if (Estimators[i] == "canberra") { 
353                                                 treeCalculators.push_back(new Canberra());
354                                         }else if (Estimators[i] == "structeuclidean") { 
355                                                 treeCalculators.push_back(new StructEuclidean());
356                                         }else if (Estimators[i] == "structchord") { 
357                                                 treeCalculators.push_back(new StructChord());
358                                         }else if (Estimators[i] == "hellinger") { 
359                                                 treeCalculators.push_back(new Hellinger());
360                                         }else if (Estimators[i] == "manhattan") { 
361                                                 treeCalculators.push_back(new Manhattan());
362                                         }else if (Estimators[i] == "structpearson") { 
363                                                 treeCalculators.push_back(new StructPearson());
364                                         }else if (Estimators[i] == "soergel") { 
365                                                 treeCalculators.push_back(new Soergel());
366                                         }else if (Estimators[i] == "spearman") { 
367                                                 treeCalculators.push_back(new Spearman());
368                                         }else if (Estimators[i] == "structkulczynski") { 
369                                                 treeCalculators.push_back(new StructKulczynski());
370                                         }else if (Estimators[i] == "speciesprofile") { 
371                                                 treeCalculators.push_back(new SpeciesProfile());
372                                         }else if (Estimators[i] == "hamming") { 
373                                                 treeCalculators.push_back(new Hamming());
374                                         }else if (Estimators[i] == "structchi2") { 
375                                                 treeCalculators.push_back(new StructChi2());
376                                         }else if (Estimators[i] == "gower") { 
377                                                 treeCalculators.push_back(new Gower());
378                                         }else if (Estimators[i] == "memchi2") { 
379                                                 treeCalculators.push_back(new MemChi2());
380                                         }else if (Estimators[i] == "memchord") { 
381                                                 treeCalculators.push_back(new MemChord());
382                                         }else if (Estimators[i] == "memeuclidean") { 
383                                                 treeCalculators.push_back(new MemEuclidean());
384                                         }else if (Estimators[i] == "mempearson") { 
385                                                 treeCalculators.push_back(new MemPearson());
386                                         }
387                                 }
388                         }
389                         
390                         //if the users entered no valid calculators don't execute command
391                         if (treeCalculators.size() == 0) { m->mothurOut("You have given no valid calculators."); m->mothurOutEndLine(); return 0; }
392                         
393                         input = new InputData(sharedfile, "sharedfile");
394                         lookup = input->getSharedRAbundVectors();
395                         lastLabel = lookup[0]->getLabel();
396                         
397                         if (lookup.size() < 2) { m->mothurOut("You have not provided enough valid groups.  I cannot run the command."); m->mothurOutEndLine(); return 0; }
398                         
399                         //used in tree constructor 
400                         m->runParse = false;
401                         
402                         //create treemap class from groupmap for tree class to use
403                         tmap = new TreeMap();
404                         tmap->makeSim(m->getAllGroups());
405                         
406                         //clear globaldatas old tree names if any
407                         m->Treenames.clear();
408                         
409                         //fills globaldatas tree names
410                         m->Treenames = m->getGroups();
411                 
412                         if (m->control_pressed) { return 0; }
413                         
414                         //create tree file
415                         makeSimsShared();
416                         
417                         if (m->control_pressed) { for (int i = 0; i < outputNames.size(); i++) {        m->mothurRemove(outputNames[i]);  } return 0; }
418                 }else{
419                         //read in dist file
420                         filename = inputfile;
421             
422             ReadMatrix* readMatrix;
423                         if (format == "column") { readMatrix = new ReadColumnMatrix(filename); }        
424                         else if (format == "phylip") { readMatrix = new ReadPhylipMatrix(filename); }
425                                 
426                         readMatrix->setCutoff(cutoff);
427         
428                         if(namefile != ""){     
429                                 nameMap = new NameAssignment(namefile);
430                                 nameMap->readMap();
431                         }
432                         else{
433                                 nameMap = NULL;
434                         }
435         
436                         readMatrix->read(nameMap);
437                         list = readMatrix->getListVector();
438                         SparseDistanceMatrix* dMatrix = readMatrix->getDMatrix();
439
440                         //make treemap
441                         tmap = new TreeMap();
442                         
443                         if (m->control_pressed) { return 0; }
444                         
445                         tmap->makeSim(list);
446                         
447                         vector<string> namesGroups = tmap->getNamesOfGroups();
448                         m->setGroups(namesGroups);
449                 
450                         //clear globaldatas old tree names if any
451                         m->Treenames.clear();
452                 
453                         //fills globaldatas tree names
454                         m->Treenames = m->getGroups();
455                         
456                         //used in tree constructor 
457                         m->runParse = false;
458                         
459                         if (m->control_pressed) { return 0; }
460                         
461                         vector< vector<double> > matrix = makeSimsDist(dMatrix);
462             delete readMatrix;
463             delete dMatrix;
464                         
465                         if (m->control_pressed) { return 0; }
466
467                         //create a new filename
468                         string outputFile = outputDir + m->getRootName(m->getSimpleName(inputfile)) + getOutputFileNameTag("tree");     
469                         outputNames.push_back(outputFile); outputTypes["tree"].push_back(outputFile);
470                                 
471                         Tree* newTree = createTree(matrix);
472             
473             if (newTree != NULL) {  writeTree(outputFile, newTree); delete newTree; }
474                         
475                         if (m->control_pressed) { return 0; }
476
477                         m->mothurOut("Tree complete. "); m->mothurOutEndLine();
478                         
479                 }
480                                 
481                 //reset groups parameter
482                 m->clearGroups(); 
483                 
484                 //set tree file as new current treefile
485                 string current = "";
486                 itTypes = outputTypes.find("tree");
487                 if (itTypes != outputTypes.end()) {
488                         if ((itTypes->second).size() != 0) { current = (itTypes->second)[0]; m->setTreeFile(current); }
489                 }
490                 
491                 m->mothurOutEndLine();
492                 m->mothurOut("Output File Names: "); m->mothurOutEndLine();
493                 for (int i = 0; i < outputNames.size(); i++) {  m->mothurOut(outputNames[i]); m->mothurOutEndLine();    }
494                 m->mothurOutEndLine();
495
496                 return 0;
497         }
498         catch(exception& e) {
499                 m->errorOut(e, "TreeGroupCommand", "execute");
500                 exit(1);
501         }
502 }
503 //**********************************************************************************************************************
504
505 Tree* TreeGroupCommand::createTree(vector< vector<double> >& simMatrix){
506         try {
507                 //create tree
508                 t = new Tree(tmap, simMatrix);
509         
510         if (m->control_pressed) { delete t; t = NULL; return t; }
511                 
512         //assemble tree
513         map<string, string> empty;
514                 t->assembleTree(empty);
515
516                 return t;
517         }
518         catch(exception& e) {
519                 m->errorOut(e, "TreeGroupCommand", "createTree");
520                 exit(1);
521         }
522 }
523 /***********************************************************/
524 int TreeGroupCommand::writeTree(string out, Tree* T) {
525         try {
526                 
527         //print newick file
528                 t->createNewickFile(out);
529                 
530         if (m->control_pressed) { m->mothurRemove(out); outputNames.pop_back(); return 1; }
531         
532         return 0;
533         
534         }
535         catch(exception& e) {
536                 m->errorOut(e, "TreeGroupCommand", "printSims");
537                 exit(1);
538         }
539 }
540
541 /***********************************************************/
542 void TreeGroupCommand::printSims(ostream& out, vector< vector<double> >& simMatrix) {
543         try {
544                 
545         for (int m = 0; m < simMatrix.size(); m++)      {
546                         //out << lookup[m]->getGroup() << '\t';
547                         for (int n = 0; n < simMatrix.size(); n++)      {
548                                 out << simMatrix[m][n] << '\t'; 
549                         }
550                         out << endl;
551                 }
552
553         }
554         catch(exception& e) {
555                 m->errorOut(e, "TreeGroupCommand", "printSims");
556                 exit(1);
557         }
558 }
559 /***********************************************************/
560 vector< vector<double> > TreeGroupCommand::makeSimsDist(SparseDistanceMatrix* matrix) {
561         try {
562                 numGroups = list->size();
563                 
564                 //initialize simMatrix
565                 vector< vector<double> > simMatrix;
566                 simMatrix.resize(numGroups);
567                 for (int k = 0; k < simMatrix.size(); k++)      {
568                         for (int j = 0; j < simMatrix.size(); j++)      {
569                                 simMatrix[k].push_back(0.0);
570                         }
571                 }
572                 
573                 //go through sparse matrix and fill sims
574                 //go through each cell in the sparsematrix
575         for (int i = 0; i < matrix->seqVec.size(); i++) {
576             for (int j = 0; j < matrix->seqVec[i].size(); j++) {
577                 
578                 //already checked everyone else in row
579                 if (i < matrix->seqVec[i][j].index) {   
580                     simMatrix[i][matrix->seqVec[i][j].index] = -(matrix->seqVec[i][j].dist -1.0);       
581                     simMatrix[matrix->seqVec[i][j].index][i] = -(matrix->seqVec[i][j].dist -1.0);       
582                         
583                     if (m->control_pressed) { return simMatrix; }
584                 }
585             }
586                 }
587
588                 return simMatrix;
589         }
590         catch(exception& e) {
591                 m->errorOut(e, "TreeGroupCommand", "makeSimsDist");
592                 exit(1);
593         }
594 }
595
596 /***********************************************************/
597 int TreeGroupCommand::makeSimsShared() {
598         try {
599         
600         if (subsample) { 
601             if (subsampleSize == -1) { //user has not set size, set size = smallest samples size
602                 subsampleSize = lookup[0]->getNumSeqs();
603                 for (int i = 1; i < lookup.size(); i++) {
604                     int thisSize = lookup[i]->getNumSeqs();
605                     
606                     if (thisSize < subsampleSize) {     subsampleSize = thisSize;       }
607                 }
608             }else {
609                 m->clearGroups();
610                 Groups.clear();
611                 vector<SharedRAbundVector*> temp;
612                 for (int i = 0; i < lookup.size(); i++) {
613                     if (lookup[i]->getNumSeqs() < subsampleSize) { 
614                         m->mothurOut(lookup[i]->getGroup() + " contains " + toString(lookup[i]->getNumSeqs()) + ". Eliminating."); m->mothurOutEndLine();
615                         delete lookup[i];
616                     }else { 
617                         Groups.push_back(lookup[i]->getGroup()); 
618                         temp.push_back(lookup[i]);
619                     }
620                 } 
621                 lookup = temp;
622                 m->setGroups(Groups);
623             }
624             
625             if (lookup.size() < 2) { m->mothurOut("You have not provided enough valid groups.  I cannot run the command."); m->mothurOutEndLine(); m->control_pressed = true; return 0; }
626         }
627         
628         numGroups = lookup.size();
629                 lines.resize(processors);
630                 for (int i = 0; i < processors; i++) {
631                         lines[i].start = int (sqrt(float(i)/float(processors)) * numGroups);
632                         lines[i].end = int (sqrt(float(i+1)/float(processors)) * numGroups);
633                 }       
634         
635                 set<string> processedLabels;
636                 set<string> userLabels = labels;
637                 
638                 //as long as you are not at the end of the file or done wih the lines you want
639                 while((lookup[0] != NULL) && ((allLines == 1) || (userLabels.size() != 0))) {
640                         if (m->control_pressed) { for (int i = 0; i < lookup.size(); i++) {  delete lookup[i];  } for(int i = 0 ; i < treeCalculators.size(); i++) {  delete treeCalculators[i]; } return 1; }
641                 
642                         if(allLines == 1 || labels.count(lookup[0]->getLabel()) == 1){                  
643                                 m->mothurOut(lookup[0]->getLabel()); m->mothurOutEndLine();
644                                 process(lookup);
645                                 
646                                 processedLabels.insert(lookup[0]->getLabel());
647                                 userLabels.erase(lookup[0]->getLabel());
648                         }
649                         
650                         if ((m->anyLabelsToProcess(lookup[0]->getLabel(), userLabels, "") == true) && (processedLabels.count(lastLabel) != 1)) {
651                                 string saveLabel = lookup[0]->getLabel();
652                         
653                                 for (int i = 0; i < lookup.size(); i++) {  delete lookup[i];  } 
654                                 lookup = input->getSharedRAbundVectors(lastLabel);
655
656                                 m->mothurOut(lookup[0]->getLabel()); m->mothurOutEndLine();
657                                 process(lookup);
658                                         
659                                 processedLabels.insert(lookup[0]->getLabel());
660                                 userLabels.erase(lookup[0]->getLabel());
661                                 
662                                 //restore real lastlabel to save below
663                                 lookup[0]->setLabel(saveLabel);
664                         }
665
666                         lastLabel = lookup[0]->getLabel();                      
667                         
668                         //get next line to process
669                         for (int i = 0; i < lookup.size(); i++) {  delete lookup[i];  } 
670                         lookup = input->getSharedRAbundVectors();
671                 }
672                 
673                 if (m->control_pressed) { for (int i = 0; i < lookup.size(); i++) {  delete lookup[i];  } for(int i = 0 ; i < treeCalculators.size(); i++) {  delete treeCalculators[i]; } return 1; }
674
675                 //output error messages about any remaining user labels
676                 set<string>::iterator it;
677                 bool needToRun = false;
678                 for (it = userLabels.begin(); it != userLabels.end(); it++) {  
679                         m->mothurOut("Your file does not include the label " + *it); 
680                         if (processedLabels.count(lastLabel) != 1) {
681                                 m->mothurOut(". I will use " + lastLabel + "."); m->mothurOutEndLine();
682                                 needToRun = true;
683                         }else {
684                                 m->mothurOut(". Please refer to " + lastLabel + "."); m->mothurOutEndLine();
685                         }
686                 }
687                 
688                 //run last label if you need to
689                 if (needToRun == true)  {
690                         for (int i = 0; i < lookup.size(); i++) {  if (lookup[i] != NULL) {             delete lookup[i]; }             } 
691                         lookup = input->getSharedRAbundVectors(lastLabel);
692
693                         m->mothurOut(lookup[0]->getLabel()); m->mothurOutEndLine();
694                         process(lookup);
695                         for (int i = 0; i < lookup.size(); i++) {  delete lookup[i];  }         
696                 }
697                 
698                 for(int i = 0 ; i < treeCalculators.size(); i++) {  delete treeCalculators[i]; }
699                 
700                 return 0;
701         }
702         catch(exception& e) {
703                 m->errorOut(e, "TreeGroupCommand", "makeSimsShared");
704                 exit(1);
705         }
706 }
707
708 /***********************************************************/
709 int TreeGroupCommand::process(vector<SharedRAbundVector*> thisLookup) {
710         try{
711                 vector< vector< vector<seqDist> > > calcDistsTotals;  //each iter, one for each calc, then each groupCombos dists. this will be used to make .dist files
712         vector< vector<seqDist>  > calcDists; calcDists.resize(treeCalculators.size());                 
713         
714         for (int thisIter = 0; thisIter < iters; thisIter++) {
715             
716             vector<SharedRAbundVector*> thisItersLookup = thisLookup;
717             
718             if (subsample) {
719                 SubSample sample;
720                 vector<string> tempLabels; //dont need since we arent printing the sampled sharedRabunds
721                 
722                 //make copy of lookup so we don't get access violations
723                 vector<SharedRAbundVector*> newLookup;
724                 for (int k = 0; k < thisItersLookup.size(); k++) {
725                     SharedRAbundVector* temp = new SharedRAbundVector();
726                     temp->setLabel(thisItersLookup[k]->getLabel());
727                     temp->setGroup(thisItersLookup[k]->getGroup());
728                     newLookup.push_back(temp);
729                 }
730                 
731                 //for each bin
732                 for (int k = 0; k < thisItersLookup[0]->getNumBins(); k++) {
733                     if (m->control_pressed) { for (int j = 0; j < newLookup.size(); j++) {  delete newLookup[j];  } return 0; }
734                     for (int j = 0; j < thisItersLookup.size(); j++) { newLookup[j]->push_back(thisItersLookup[j]->getAbundance(k), thisItersLookup[j]->getGroup()); }
735                 }
736                 
737                 tempLabels = sample.getSample(newLookup, subsampleSize);
738                 thisItersLookup = newLookup;
739             }
740             
741             if(processors == 1){
742                 driver(thisItersLookup, 0, numGroups, calcDists);
743             }else{
744                 int process = 1;
745                 vector<int> processIDS;
746                 
747 #if defined (__APPLE__) || (__MACH__) || (linux) || (__linux) || (__linux__) || (__unix__) || (__unix)
748                 //loop through and create all the processes you want
749                 while (process != processors) {
750                     int pid = fork();
751                     
752                     if (pid > 0) {
753                         processIDS.push_back(pid); 
754                         process++;
755                     }else if (pid == 0){
756                         
757                         driver(thisItersLookup, lines[process].start, lines[process].end, calcDists);   
758                         
759                         string tempdistFileName = m->getRootName(m->getSimpleName(sharedfile)) + toString(getpid()) + ".dist";
760                         ofstream outtemp;
761                         m->openOutputFile(tempdistFileName, outtemp);
762                         
763                         for (int i = 0; i < calcDists.size(); i++) {
764                             outtemp << calcDists[i].size() << endl;
765                             
766                             for (int j = 0; j < calcDists[i].size(); j++) {
767                                 outtemp << calcDists[i][j].seq1 << '\t' << calcDists[i][j].seq2 << '\t' << calcDists[i][j].dist << endl;
768                             }
769                         }
770                         outtemp.close();
771                         
772                         exit(0);
773                     }else { 
774                         m->mothurOut("[ERROR]: unable to spawn the necessary processes."); m->mothurOutEndLine(); 
775                         for (int i = 0; i < processIDS.size(); i++) { kill (processIDS[i], SIGINT); }
776                         exit(0);
777                     }
778                 }
779                 
780                 //parent do your part
781                 driver(thisItersLookup, lines[0].start, lines[0].end, calcDists);   
782                 
783                 //force parent to wait until all the processes are done
784                 for (int i = 0; i < processIDS.size(); i++) {
785                     int temp = processIDS[i];
786                     wait(&temp);
787                 }
788                 
789                 for (int i = 0; i < processIDS.size(); i++) {
790                     string tempdistFileName = m->getRootName(m->getSimpleName(sharedfile)) + toString(processIDS[i]) +  ".dist";
791                     ifstream intemp;
792                     m->openInputFile(tempdistFileName, intemp);
793                     
794                     for (int k = 0; k < calcDists.size(); k++) {
795                         int size = 0;
796                         intemp >> size; m->gobble(intemp);
797                         
798                         for (int j = 0; j < size; j++) {
799                             int seq1 = 0;
800                             int seq2 = 0;
801                             float dist = 1.0;
802                             
803                             intemp >> seq1 >> seq2 >> dist;   m->gobble(intemp);
804                             
805                             seqDist tempDist(seq1, seq2, dist);
806                             calcDists[k].push_back(tempDist);
807                         }
808                     }
809                     intemp.close();
810                     m->mothurRemove(tempdistFileName);
811                 }
812 #else
813                 //////////////////////////////////////////////////////////////////////////////////////////////////////
814                 //Windows version shared memory, so be careful when passing variables through the treeSharedData struct. 
815                 //Above fork() will clone, so memory is separate, but that's not the case with windows, 
816                 //Taking advantage of shared memory to pass results vectors.
817                 //////////////////////////////////////////////////////////////////////////////////////////////////////
818                 
819                 vector<treeSharedData*> pDataArray; 
820                 DWORD   dwThreadIdArray[processors-1];
821                 HANDLE  hThreadArray[processors-1]; 
822                 
823                 //Create processor worker threads.
824                 for( int i=1; i<processors; i++ ){
825                     
826                     //make copy of lookup so we don't get access violations
827                     vector<SharedRAbundVector*> newLookup;
828                     for (int k = 0; k < thisItersLookup.size(); k++) {
829                         SharedRAbundVector* temp = new SharedRAbundVector();
830                         temp->setLabel(thisItersLookup[k]->getLabel());
831                         temp->setGroup(thisItersLookup[k]->getGroup());
832                         newLookup.push_back(temp);
833                     }
834                     
835                     //for each bin
836                     for (int k = 0; k < thisItersLookup[0]->getNumBins(); k++) {
837                         if (m->control_pressed) { for (int j = 0; j < newLookup.size(); j++) {  delete newLookup[j];  } return 0; }
838                         for (int j = 0; j < thisItersLookup.size(); j++) { newLookup[j]->push_back(thisItersLookup[j]->getAbundance(k), thisItersLookup[j]->getGroup()); }
839                     }
840                     
841                     // Allocate memory for thread data.
842                     treeSharedData* tempSum = new treeSharedData(m, lines[i].start, lines[i].end, Estimators, newLookup);
843                     pDataArray.push_back(tempSum);
844                     processIDS.push_back(i);
845                     
846                     hThreadArray[i-1] = CreateThread(NULL, 0, MyTreeSharedThreadFunction, pDataArray[i-1], 0, &dwThreadIdArray[i-1]);   
847                 }
848                 
849                 //parent do your part
850                 driver(thisItersLookup, lines[0].start, lines[0].end, calcDists);   
851                 
852                 //Wait until all threads have terminated.
853                 WaitForMultipleObjects(processors-1, hThreadArray, TRUE, INFINITE);
854                 
855                 //Close all thread handles and free memory allocations.
856                 for(int i=0; i < pDataArray.size(); i++){
857                     for (int j = 0; j < pDataArray[i]->thisLookup.size(); j++) {  delete pDataArray[i]->thisLookup[j];  } 
858                     
859                     for (int k = 0; k < calcDists.size(); k++) {
860                         int size = pDataArray[i]->calcDists[k].size();
861                         for (int j = 0; j < size; j++) {    calcDists[k].push_back(pDataArray[i]->calcDists[k][j]);    }
862                     }
863                     
864                     CloseHandle(hThreadArray[i]);
865                     delete pDataArray[i];
866                 }
867                 
868 #endif
869             }
870             
871             calcDistsTotals.push_back(calcDists);
872             
873             if (subsample) {  
874                 
875                 //clean up memory
876                 for (int i = 0; i < thisItersLookup.size(); i++) { delete thisItersLookup[i]; }
877                 thisItersLookup.clear();
878                 for (int i = 0; i < calcDists.size(); i++) {  calcDists[i].clear(); }
879             }
880                 }
881                 
882         if (iters != 1) {
883             //we need to find the average distance and standard deviation for each groups distance
884             
885             vector< vector<seqDist>  > calcAverages; calcAverages.resize(treeCalculators.size()); 
886             for (int i = 0; i < calcAverages.size(); i++) {  //initialize sums to zero.
887                 calcAverages[i].resize(calcDistsTotals[0][i].size());
888                 
889                 for (int j = 0; j < calcAverages[i].size(); j++) {
890                     calcAverages[i][j].seq1 = calcDists[i][j].seq1;
891                     calcAverages[i][j].seq2 = calcDists[i][j].seq2;
892                     calcAverages[i][j].dist = 0.0;
893                 }
894             }
895             
896             for (int thisIter = 0; thisIter < iters; thisIter++) { //sum all groups dists for each calculator
897                 for (int i = 0; i < calcAverages.size(); i++) {  //initialize sums to zero.
898                     for (int j = 0; j < calcAverages[i].size(); j++) {
899                         calcAverages[i][j].dist += calcDistsTotals[thisIter][i][j].dist;
900                     }
901                 }
902             }
903             
904             for (int i = 0; i < calcAverages.size(); i++) {  //finds average.
905                 for (int j = 0; j < calcAverages[i].size(); j++) {
906                     calcAverages[i][j].dist /= (float) iters;
907                 }
908             }
909             
910             //create average tree for each calc
911             for (int i = 0; i < calcDists.size(); i++) {
912                 vector< vector<double> > matrix; //square matrix to represent the distance
913                 matrix.resize(thisLookup.size());
914                 for (int k = 0; k < thisLookup.size(); k++) {  matrix[k].resize(thisLookup.size(), 0.0); }
915                 
916                 for (int j = 0; j < calcAverages[i].size(); j++) {
917                     int row = calcAverages[i][j].seq1;
918                     int column = calcAverages[i][j].seq2;
919                     float dist = calcAverages[i][j].dist;
920                     
921                     matrix[row][column] = dist;
922                     matrix[column][row] = dist;
923                 }
924                 
925                 //create a new filename
926                 string outputFile = outputDir + m->getRootName(m->getSimpleName(inputfile)) + treeCalculators[i]->getName() + "." + thisLookup[0]->getLabel() + ".ave." + getOutputFileNameTag("tree");                         
927                 outputNames.push_back(outputFile); outputTypes["tree"].push_back(outputFile); 
928                 
929                 //creates tree from similarity matrix and write out file
930                 Tree* newTree = createTree(matrix);
931                 if (newTree != NULL) { writeTree(outputFile, newTree); }                
932             }
933             
934             //create all trees for each calc and find their consensus tree
935             for (int i = 0; i < calcDists.size(); i++) {
936                 if (m->control_pressed) { break; }
937                 
938                 //create a new filename
939                 string outputFile = outputDir + m->getRootName(m->getSimpleName(inputfile)) + treeCalculators[i]->getName() + "." + thisLookup[0]->getLabel() + ".all." + getOutputFileNameTag("tree");                         
940                 outputNames.push_back(outputFile); outputTypes["tree"].push_back(outputFile); 
941                 
942                 ofstream outAll;
943                 m->openOutputFile(outputFile, outAll);
944                 
945                 vector<Tree*> trees; 
946                 for (int myIter = 0; myIter < iters; myIter++) {
947                     
948                     if(m->control_pressed) { break; }
949                     
950                     //initialize matrix
951                     vector< vector<double> > matrix; //square matrix to represent the distance
952                     matrix.resize(thisLookup.size());
953                     for (int k = 0; k < thisLookup.size(); k++) {  matrix[k].resize(thisLookup.size(), 0.0); }
954                     
955                     for (int j = 0; j < calcDistsTotals[myIter][i].size(); j++) {
956                         int row = calcDistsTotals[myIter][i][j].seq1;
957                         int column = calcDistsTotals[myIter][i][j].seq2;
958                         double dist = calcDistsTotals[myIter][i][j].dist;
959                         
960                         matrix[row][column] = dist;
961                         matrix[column][row] = dist;
962                     }
963                     
964                     //creates tree from similarity matrix and write out file
965                     Tree* newTree = createTree(matrix);
966                     if (newTree != NULL) { 
967                         newTree->print(outAll);
968                         trees.push_back(newTree);
969                     }
970                 }
971                 outAll.close();
972                 if (m->control_pressed) { for (int k = 0; k < trees.size(); k++) { delete trees[k]; } }
973                 
974                 Consensus consensus;
975                 //clear old tree names if any
976                 m->Treenames.clear(); m->Treenames = m->getGroups(); //may have changed if subsample eliminated groups
977                 Tree* conTree = consensus.getTree(trees);
978                 
979                 //create a new filename
980                 string conFile = outputDir + m->getRootName(m->getSimpleName(inputfile)) + treeCalculators[i]->getName() + "." + thisLookup[0]->getLabel() + ".cons." + getOutputFileNameTag("tree");                           
981                 outputNames.push_back(conFile); outputTypes["tree"].push_back(conFile); 
982                 ofstream outTree;
983                 m->openOutputFile(conFile, outTree);
984                 
985                 if (conTree != NULL) { conTree->print(outTree, "boot"); delete conTree; }
986             }
987
988         }else {
989             
990             for (int i = 0; i < calcDists.size(); i++) {
991                 if (m->control_pressed) { break; }
992                 
993                 //initialize matrix
994                 vector< vector<double> > matrix; //square matrix to represent the distance
995                 matrix.resize(thisLookup.size());
996                 for (int k = 0; k < thisLookup.size(); k++) {  matrix[k].resize(thisLookup.size(), 0.0); }
997                 
998                 for (int j = 0; j < calcDists[i].size(); j++) {
999                     int row = calcDists[i][j].seq1;
1000                     int column = calcDists[i][j].seq2;
1001                     double dist = calcDists[i][j].dist;
1002                     
1003                     matrix[row][column] = dist;
1004                     matrix[column][row] = dist;
1005                 }
1006                 
1007                 //create a new filename
1008                 string outputFile = outputDir + m->getRootName(m->getSimpleName(inputfile)) + treeCalculators[i]->getName() + "." + thisLookup[0]->getLabel() + "." + getOutputFileNameTag("tree");                             
1009                 outputNames.push_back(outputFile); outputTypes["tree"].push_back(outputFile); 
1010                 
1011                 //creates tree from similarity matrix and write out file
1012                 Tree* newTree = createTree(matrix);
1013                 if (newTree != NULL) { writeTree(outputFile, newTree); delete newTree; }
1014             }
1015         }
1016                 
1017                 return 0;
1018         }
1019         catch(exception& e) {
1020                 m->errorOut(e, "TreeGroupCommand", "process");
1021                 exit(1);
1022         }
1023 }
1024 /**************************************************************************************************/
1025 int TreeGroupCommand::driver(vector<SharedRAbundVector*> thisLookup, int start, int end, vector< vector<seqDist> >& calcDists) { 
1026         try {
1027                 vector<SharedRAbundVector*> subset;
1028                 for (int k = start; k < end; k++) { // pass cdd each set of groups to compare
1029                         
1030                         for (int l = 0; l < k; l++) {
1031                                 
1032                                 if (k != l) { //we dont need to similiarity of a groups to itself
1033                                         subset.clear(); //clear out old pair of sharedrabunds
1034                                         //add new pair of sharedrabunds
1035                                         subset.push_back(thisLookup[k]); subset.push_back(thisLookup[l]); 
1036                                         
1037                                         for(int i=0;i<treeCalculators.size();i++) {
1038                                                 
1039                                                 //if this calc needs all groups to calculate the pair load all groups
1040                                                 if (treeCalculators[i]->getNeedsAll()) { 
1041                                                         //load subset with rest of lookup for those calcs that need everyone to calc for a pair
1042                                                         for (int w = 0; w < thisLookup.size(); w++) {
1043                                                                 if ((w != k) && (w != l)) { subset.push_back(thisLookup[w]); }
1044                                                         }
1045                                                 }
1046                                                 
1047                                                 vector<double> tempdata = treeCalculators[i]->getValues(subset); //saves the calculator outputs
1048                                                 
1049                                                 if (m->control_pressed) { return 1; }
1050                                                 
1051                                                 seqDist temp(l, k, -(tempdata[0]-1.0));
1052                                                 calcDists[i].push_back(temp);
1053                                         }
1054                                 }
1055                         }
1056                 }
1057                 
1058                 return 0;
1059         }
1060         catch(exception& e) {
1061                 m->errorOut(e, "TreeGroupCommand", "driver");
1062                 exit(1);
1063         }
1064 }
1065 /***********************************************************/
1066
1067         
1068