unifracweightedcommand.cpp

   1 /*
   2  *  unifracweightedcommand.cpp
   3  *  Mothur
   4  *
   5  *  Created by Sarah Westcott on 2/9/09.
   6  *  Copyright 2009 Schloss Lab UMASS Amherst. All rights reserved.
   7  *
   8  */
   9
  10 #include "unifracweightedcommand.h"
  11
  12 //**********************************************************************************************************************
  13 vector<string> UnifracWeightedCommand::getValidParameters(){
  14         try {
  15                 string Array[] =  {"groups","iters","distance","random","processors","outputdir","inputdir"};
  16                 vector<string> myArray (Array, Array+(sizeof(Array)/sizeof(string)));
  17                 return myArray;
  18         }
  19         catch(exception& e) {
  20                 m->errorOut(e, "UnifracWeightedCommand", "getValidParameters");
  21                 exit(1);
  22         }
  23 }
  24 //**********************************************************************************************************************
  25 UnifracWeightedCommand::UnifracWeightedCommand(){
  26         try {
  27                 //initialize outputTypes
  28                 vector<string> tempOutNames;
  29                 outputTypes["weighted"] = tempOutNames;
  30                 outputTypes["wsummary"] = tempOutNames;
  31                 outputTypes["phylip"] = tempOutNames;
  32         }
  33         catch(exception& e) {
  34                 m->errorOut(e, "UnifracWeightedCommand", "UnifracWeightedCommand");
  35                 exit(1);
  36         }
  37 }
  38 //**********************************************************************************************************************
  39 vector<string> UnifracWeightedCommand::getRequiredParameters(){
  40         try {
  41                 vector<string> myArray;
  42                 return myArray;
  43         }
  44         catch(exception& e) {
  45                 m->errorOut(e, "UnifracWeightedCommand", "getRequiredParameters");
  46                 exit(1);
  47         }
  48 }
  49 //**********************************************************************************************************************
  50 vector<string> UnifracWeightedCommand::getRequiredFiles(){
  51         try {
  52                 string Array[] =  {"tree","group"};
  53                 vector<string> myArray (Array, Array+(sizeof(Array)/sizeof(string)));
  54
  55                 return myArray;
  56         }
  57         catch(exception& e) {
  58                 m->errorOut(e, "UnifracWeightedCommand", "getRequiredFiles");
  59                 exit(1);
  60         }
  61 }
  62 /***********************************************************/
  63 UnifracWeightedCommand::UnifracWeightedCommand(string option) {
  64         try {
  65                 globaldata = GlobalData::getInstance();
  66                 abort = false;
  67                 Groups.clear();
  68
  69                 //allow user to run help
  70                 if(option == "help") { help(); abort = true; }
  71
  72                 else {
  73                         //valid paramters for this command
  74                         string Array[] =  {"groups","iters","distance","random","processors","outputdir","inputdir"};
  75                         vector<string> myArray (Array, Array+(sizeof(Array)/sizeof(string)));
  76
  77                         OptionParser parser(option);
  78                         map<string,string> parameters=parser.getParameters();
  79
  80                         ValidParameters validParameter;
  81
  82                         //check to make sure all parameters are valid for command
  83                         for (map<string,string>::iterator it = parameters.begin(); it != parameters.end(); it++) {
  84                                 if (validParameter.isValidParameter(it->first, myArray, it->second) != true) {  abort = true;  }
  85                         }
  86
  87                         //initialize outputTypes
  88                         vector<string> tempOutNames;
  89                         outputTypes["weighted"] = tempOutNames;
  90                         outputTypes["wsummary"] = tempOutNames;
  91                         outputTypes["phylip"] = tempOutNames;
  92
  93                         if (globaldata->gTree.size() == 0) {//no trees were read
  94                                 m->mothurOut("You must execute the read.tree command, before you may execute the unifrac.weighted command."); m->mothurOutEndLine(); abort = true;  }
  95
  96                         //if the user changes the output directory command factory will send this info to us in the output parameter
  97                         outputDir = validParameter.validFile(parameters, "outputdir", false);           if (outputDir == "not found"){
  98                                 outputDir = "";
  99                                 outputDir += m->hasPath(globaldata->inputFileName); //if user entered a file with a path then preserve it
 100                         }
 101
 102                         //check for optional parameter and set defaults
 103                         // ...at some point should added some additional type checking...
 104                         groups = validParameter.validFile(parameters, "groups", false);
 105                         if (groups == "not found") { groups = ""; }
 106                         else {
 107                                 m->splitAtDash(groups, Groups);
 108                                 globaldata->Groups = Groups;
 109                         }
 110
 111                         itersString = validParameter.validFile(parameters, "iters", false);                     if (itersString == "not found") { itersString = "1000"; }
 112                         convert(itersString, iters);
 113
 114                         string temp = validParameter.validFile(parameters, "distance", false);                  if (temp == "not found") { temp = "false"; }
 115                         phylip = m->isTrue(temp);
 116
 117                         temp = validParameter.validFile(parameters, "random", false);                                   if (temp == "not found") { temp = "F"; }
 118                         random = m->isTrue(temp);
 119
 120                         temp = validParameter.validFile(parameters, "processors", false);       if (temp == "not found"){       temp = "1";                             }
 121                         convert(temp, processors);
 122
 123                         if (!random) {  iters = 0;  } //turn off random calcs
 124
 125
 126                         if (abort == false) {
 127                                 T = globaldata->gTree;
 128                                 tmap = globaldata->gTreemap;
 129                                 sumFile = outputDir + m->getSimpleName(globaldata->getTreeFile()) + ".wsummary";
 130                                 m->openOutputFile(sumFile, outSum);
 131                                 outputNames.push_back(sumFile);  outputTypes["wsummary"].push_back(sumFile);
 132
 133                                 util = new SharedUtil();
 134                                 string s; //to make work with setgroups
 135                                 util->setGroups(globaldata->Groups, tmap->namesOfGroups, s, numGroups, "weighted");     //sets the groups the user wants to analyze
 136                                 util->getCombos(groupComb, globaldata->Groups, numComp);
 137
 138                                 weighted = new Weighted(tmap);
 139
 140                         }
 141                 }
 142
 143
 144         }
 145         catch(exception& e) {
 146                 m->errorOut(e, "UnifracWeightedCommand", "UnifracWeightedCommand");
 147                 exit(1);
 148         }
 149 }
 150 //**********************************************************************************************************************
 151
 152 void UnifracWeightedCommand::help(){
 153         try {
 154                 m->mothurOut("The unifrac.weighted command can only be executed after a successful read.tree command.\n");
 155                 m->mothurOut("The unifrac.weighted command parameters are groups, iters, distance, processors and random.  No parameters are required.\n");
 156                 m->mothurOut("The groups parameter allows you to specify which of the groups in your groupfile you would like analyzed.  You must enter at least 2 valid groups.\n");
 157                 m->mothurOut("The group names are separated by dashes.  The iters parameter allows you to specify how many random trees you would like compared to your tree.\n");
 158                 m->mothurOut("The distance parameter allows you to create a distance file from the results. The default is false.\n");
 159                 m->mothurOut("The random parameter allows you to shut off the comparison to random trees. The default is false, meaning don't compare your trees with randomly generated trees.\n");
 160                 m->mothurOut("The processors parameter allows you to specify the number of processors to use. The default is 1.\n");
 161                 m->mothurOut("The unifrac.weighted command should be in the following format: unifrac.weighted(groups=yourGroups, iters=yourIters).\n");
 162                 m->mothurOut("Example unifrac.weighted(groups=A-B-C, iters=500).\n");
 163                 m->mothurOut("The default value for groups is all the groups in your groupfile, and iters is 1000.\n");
 164                 m->mothurOut("The unifrac.weighted command output two files: .weighted and .wsummary their descriptions are in the manual.\n");
 165                 m->mothurOut("Note: No spaces between parameter labels (i.e. groups), '=' and parameters (i.e.yourGroups).\n\n");
 166         }
 167         catch(exception& e) {
 168                 m->errorOut(e, "UnifracWeightedCommand", "help");
 169                 exit(1);
 170         }
 171 }
 172
 173 /***********************************************************/
 174 int UnifracWeightedCommand::execute() {
 175         try {
 176
 177                 if (abort == true) { return 0; }
 178
 179                 int start = time(NULL);
 180
 181                 //get weighted for users tree
 182                 userData.resize(numComp,0);  //data[0] = weightedscore AB, data[1] = weightedscore AC...
 183                 randomData.resize(numComp,0); //data[0] = weightedscore AB, data[1] = weightedscore AC...
 184
 185                 //get weighted scores for users trees
 186                 for (int i = 0; i < T.size(); i++) {
 187
 188                         if (m->control_pressed) { outSum.close(); for (int i = 0; i < outputNames.size(); i++) {        remove(outputNames[i].c_str());  } return 0; }
 189
 190                         counter = 0;
 191                         rScores.resize(numComp);  //data[0] = weightedscore AB, data[1] = weightedscore AC...
 192                         uScores.resize(numComp);  //data[0] = weightedscore AB, data[1] = weightedscore AC...
 193
 194                         if (random) {
 195                                 output = new ColumnFile(outputDir + m->getSimpleName(globaldata->getTreeFile())  + toString(i+1) + ".weighted", itersString);
 196                                 outputNames.push_back(outputDir + m->getSimpleName(globaldata->getTreeFile())  + toString(i+1) + ".weighted");
 197                                 outputTypes["weighted"].push_back(outputDir + m->getSimpleName(globaldata->getTreeFile())  + toString(i+1) + ".weighted");
 198                         }
 199
 200                         userData = weighted->getValues(T[i], processors, outputDir);  //userData[0] = weightedscore
 201
 202                         if (m->control_pressed) { if (random) { delete output; } outSum.close(); for (int i = 0; i < outputNames.size(); i++) { remove(outputNames[i].c_str());  } return 0; }
 203
 204                         //save users score
 205                         for (int s=0; s<numComp; s++) {
 206                                 //add users score to vector of user scores
 207                                 uScores[s].push_back(userData[s]);
 208
 209                                 //save users tree score for summary file
 210                                 utreeScores.push_back(userData[s]);
 211                         }
 212
 213                         if (random) {
 214                                 vector<double> sums = weighted->getBranchLengthSums(T[i]);
 215
 216                                 //calculate number of comparisons i.e. with groups A,B,C = AB, AC, BC = 3;
 217                                 vector< vector<string> > namesOfGroupCombos;
 218                                 for (int a=0; a<numGroups; a++) {
 219                                         for (int l = 0; l < a; l++) {
 220                                                 vector<string> groups; groups.push_back(globaldata->Groups[a]); groups.push_back(globaldata->Groups[l]);
 221                                                 namesOfGroupCombos.push_back(groups);
 222                                         }
 223                                 }
 224
 225                                 lines.clear();
 226
 227                                 #if defined (__APPLE__) || (__MACH__) || (linux) || (__linux)
 228                                         if(processors != 1){
 229                                                 int numPairs = namesOfGroupCombos.size();
 230                                                 int numPairsPerProcessor = numPairs / processors;
 231
 232                                                 for (int i = 0; i < processors; i++) {
 233                                                         int startPos = i * numPairsPerProcessor;
 234                                                         if(i == processors - 1){
 235                                                                 numPairsPerProcessor = numPairs - i * numPairsPerProcessor;
 236                                                         }
 237                                                         lines.push_back(linePair(startPos, numPairsPerProcessor));
 238                                                 }
 239                                         }
 240                                 #endif
 241
 242
 243                                 //get scores for random trees
 244                                 for (int j = 0; j < iters; j++) {
 245
 246                                         #if defined (__APPLE__) || (__MACH__) || (linux) || (__linux)
 247                                                 if(processors == 1){
 248                                                         driver(T[i],  namesOfGroupCombos, 0, namesOfGroupCombos.size(), sums, rScores);
 249                                                 }else{
 250                                                         createProcesses(T[i],  namesOfGroupCombos, sums, rScores);
 251                                                 }
 252                                         #else
 253                                                 driver(T[i], namesOfGroupCombos, 0, namesOfGroupCombos.size(), sums, rScores);
 254                                         #endif
 255
 256                                         if (m->control_pressed) { delete output; outSum.close(); for (int i = 0; i < outputNames.size(); i++) { remove(outputNames[i].c_str());  } return 0; }
 257
 258                                         //report progress
 259                                         m->mothurOut("Iter: " + toString(j+1)); m->mothurOutEndLine();
 260                                 }
 261                                 lines.clear();
 262
 263                                 //find the signifigance of the score for summary file
 264                                 for (int f = 0; f < numComp; f++) {
 265                                         //sort random scores
 266                                         sort(rScores[f].begin(), rScores[f].end());
 267
 268                                         //the index of the score higher than yours is returned
 269                                         //so if you have 1000 random trees the index returned is 100
 270                                         //then there are 900 trees with a score greater then you.
 271                                         //giving you a signifigance of 0.900
 272                                         int index = findIndex(userData[f], f);    if (index == -1) { m->mothurOut("error in UnifracWeightedCommand"); m->mothurOutEndLine(); exit(1); } //error code
 273
 274                                         //the signifigance is the number of trees with the users score or higher
 275                                         WScoreSig.push_back((iters-index)/(float)iters);
 276                                 }
 277
 278                                 //out << "Tree# " << i << endl;
 279                                 calculateFreqsCumuls();
 280                                 printWeightedFile();
 281
 282                                 delete output;
 283
 284                         }
 285
 286                         //clear data
 287                         rScores.clear();
 288                         uScores.clear();
 289                         validScores.clear();
 290                 }
 291
 292
 293                 if (m->control_pressed) { outSum.close(); for (int i = 0; i < outputNames.size(); i++) {        remove(outputNames[i].c_str());  } return 0;  }
 294
 295                 printWSummaryFile();
 296
 297                 if (phylip) {   createPhylipFile();             }
 298
 299                 //clear out users groups
 300                 globaldata->Groups.clear();
 301
 302
 303                 if (m->control_pressed) {
 304                         for (int i = 0; i < outputNames.size(); i++) {  remove(outputNames[i].c_str());  }
 305                         return 0;
 306                 }
 307
 308                 m->mothurOut("It took " + toString(time(NULL) - start) + " secs to run unifrac.weighted."); m->mothurOutEndLine();
 309
 310                 m->mothurOutEndLine();
 311                 m->mothurOut("Output File Names: "); m->mothurOutEndLine();
 312                 for (int i = 0; i < outputNames.size(); i++) {  m->mothurOut(outputNames[i]); m->mothurOutEndLine();    }
 313                 m->mothurOutEndLine();
 314
 315                 return 0;
 316
 317         }
 318         catch(exception& e) {
 319                 m->errorOut(e, "UnifracWeightedCommand", "execute");
 320                 exit(1);
 321         }
 322 }
 323 /**************************************************************************************************/
 324
 325 int UnifracWeightedCommand::createProcesses(Tree* t, vector< vector<string> > namesOfGroupCombos, vector<double>& sums, vector< vector<double> >& scores) {
 326         try {
 327 #if defined (__APPLE__) || (__MACH__) || (linux) || (__linux)
 328                 int process = 1;
 329                 int num = 0;
 330                 vector<int> processIDS;
 331
 332                 EstOutput results;
 333
 334                 //loop through and create all the processes you want
 335                 while (process != processors) {
 336                         int pid = fork();
 337
 338                         if (pid > 0) {
 339                                 processIDS.push_back(pid);  //create map from line number to pid so you can append files in correct order later
 340                                 process++;
 341                         }else if (pid == 0){
 342                                 driver(t, namesOfGroupCombos, lines[process].start, lines[process].num, sums, scores);
 343
 344                                 //pass numSeqs to parent
 345                                 ofstream out;
 346                                 string tempFile = outputDir + toString(getpid()) + ".weightedcommand.results.temp";
 347                                 m->openOutputFile(tempFile, out);
 348                                 for (int i = lines[process].start; i < (lines[process].start + lines[process].num); i++) { out << scores[i][(scores[i].size()-1)] << '\t';  } out << endl;
 349                                 out.close();
 350
 351                                 exit(0);
 352                         }else { m->mothurOut("unable to spawn the necessary processes."); m->mothurOutEndLine(); exit(0); }
 353                 }
 354
 355                 driver(t, namesOfGroupCombos, lines[0].start, lines[0].num, sums, scores);
 356
 357                 //force parent to wait until all the processes are done
 358                 for (int i=0;i<(processors-1);i++) {
 359                         int temp = processIDS[i];
 360                         wait(&temp);
 361                 }
 362
 363                 //get data created by processes
 364                 for (int i=0;i<(processors-1);i++) {
 365
 366                         ifstream in;
 367                         string s = outputDir + toString(processIDS[i]) + ".weightedcommand.results.temp";
 368                         m->openInputFile(s, in);
 369
 370                         double tempScore;
 371                         for (int j = lines[(i+1)].start; j < (lines[(i+1)].start + lines[(i+1)].num); j++) { in >> tempScore; scores[j].push_back(tempScore); }
 372                         in.close();
 373                         remove(s.c_str());
 374                 }
 375
 376                 return 0;
 377 #endif
 378         }
 379         catch(exception& e) {
 380                 m->errorOut(e, "UnifracWeightedCommand", "createProcesses");
 381                 exit(1);
 382         }
 383 }
 384
 385 /**************************************************************************************************/
 386 int UnifracWeightedCommand::driver(Tree* t, vector< vector<string> > namesOfGroupCombos, int start, int num, vector<double>& sums, vector< vector<double> >& scores) {
 387  try {
 388                 Tree* randT = new Tree();
 389
 390                 for (int h = start; h < (start+num); h++) {
 391
 392                         if (m->control_pressed) { return 0; }
 393
 394                         //initialize weighted score
 395                         string groupA = namesOfGroupCombos[h][0];
 396                         string groupB = namesOfGroupCombos[h][1];
 397
 398                         //copy T[i]'s info.
 399                         randT->getCopy(t);
 400
 401                         //create a random tree with same topology as T[i], but different labels
 402                         randT->assembleRandomUnifracTree(groupA, groupB);
 403
 404                         if (m->control_pressed) { delete randT;  return 0;  }
 405
 406                         //get wscore of random tree
 407                         EstOutput randomData = weighted->getValues(randT, groupA, groupB, sums);
 408
 409                         if (m->control_pressed) { delete randT;  return 0;  }
 410
 411                         //save scores
 412                         scores[h].push_back(randomData[0]);
 413                 }
 414
 415                 delete randT;
 416
 417                 return 0;
 418
 419         }
 420         catch(exception& e) {
 421                 m->errorOut(e, "UnifracWeightedCommand", "driver");
 422                 exit(1);
 423         }
 424 }
 425 /***********************************************************/
 426 void UnifracWeightedCommand::printWeightedFile() {
 427         try {
 428                 vector<double> data;
 429                 vector<string> tags;
 430                 tags.push_back("Score"); tags.push_back("RandFreq"); tags.push_back("RandCumul");
 431
 432                 for(int a = 0; a < numComp; a++) {
 433                         output->initFile(groupComb[a], tags);
 434                         //print each line
 435                         for (map<float,float>::iterator it = validScores.begin(); it != validScores.end(); it++) {
 436                                 data.push_back(it->first);  data.push_back(rScoreFreq[a][it->first]); data.push_back(rCumul[a][it->first]);
 437                                 output->output(data);
 438                                 data.clear();
 439                         }
 440                         output->resetFile();
 441                 }
 442         }
 443         catch(exception& e) {
 444                 m->errorOut(e, "UnifracWeightedCommand", "printWeightedFile");
 445                 exit(1);
 446         }
 447 }
 448
 449
 450 /***********************************************************/
 451 void UnifracWeightedCommand::printWSummaryFile() {
 452         try {
 453                 //column headers
 454                 outSum << "Tree#" << '\t' << "Groups" << '\t' << "WScore" << '\t' << "WSig" <<  endl;
 455                 m->mothurOut("Tree#\tGroups\tWScore\tWSig"); m->mothurOutEndLine();
 456
 457                 //format output
 458                 outSum.setf(ios::fixed, ios::floatfield); outSum.setf(ios::showpoint);
 459
 460                 //print each line
 461                 int count = 0;
 462                 for (int i = 0; i < T.size(); i++) {
 463                         for (int j = 0; j < numComp; j++) {
 464                                 if (random) {
 465                                         if (WScoreSig[count] > (1/(float)iters)) {
 466                                                 outSum << setprecision(6) << i+1 << '\t' << groupComb[j] << '\t' << utreeScores[count] << '\t' << setprecision(itersString.length()) << WScoreSig[count] << endl;
 467                                                 cout << setprecision(6) << i+1 << '\t' << groupComb[j] << '\t' << utreeScores[count] << '\t' << setprecision(itersString.length()) << WScoreSig[count] << endl;
 468                                                 m->mothurOutJustToLog(toString(i+1) +"\t" + groupComb[j] +"\t" + toString(utreeScores[count]) +"\t" +  toString(WScoreSig[count]) + "\n");
 469                                         }else{
 470                                                 outSum << setprecision(6) << i+1 << '\t' << groupComb[j] << '\t' << utreeScores[count] << '\t' << setprecision(itersString.length()) << "<" << (1/float(iters)) << endl;
 471                                                 cout << setprecision(6) << i+1 << '\t' << groupComb[j] << '\t' << utreeScores[count] << '\t' << setprecision(itersString.length()) << "<" << (1/float(iters)) << endl;
 472                                                 m->mothurOutJustToLog(toString(i+1) +"\t" + groupComb[j] +"\t" + toString(utreeScores[count]) +"\t<" +  toString((1/float(iters))) + "\n");
 473                                         }
 474                                 }else{
 475                                         outSum << setprecision(6) << i+1 << '\t' << groupComb[j] << '\t' << utreeScores[count] << '\t' << "0.00" << endl;
 476                                         cout << setprecision(6) << i+1 << '\t' << groupComb[j] << '\t' << utreeScores[count] << '\t' << "0.00" << endl;
 477                                         m->mothurOutJustToLog(toString(i+1) +"\t" + groupComb[j] +"\t" + toString(utreeScores[count]) +"\t0.00\n");
 478                                 }
 479                                 count++;
 480                         }
 481                 }
 482                 outSum.close();
 483         }
 484         catch(exception& e) {
 485                 m->errorOut(e, "UnifracWeightedCommand", "printWSummaryFile");
 486                 exit(1);
 487         }
 488 }
 489 /***********************************************************/
 490 void UnifracWeightedCommand::createPhylipFile() {
 491         try {
 492                 int count = 0;
 493                 //for each tree
 494                 for (int i = 0; i < T.size(); i++) {
 495
 496                         string phylipFileName = outputDir + m->getSimpleName(globaldata->getTreeFile())  + toString(i+1) + ".weighted.dist";
 497                         outputNames.push_back(phylipFileName);
 498                         outputTypes["phylip"].push_back(phylipFileName);
 499                         ofstream out;
 500                         m->openOutputFile(phylipFileName, out);
 501
 502                         //output numSeqs
 503                         out << globaldata->Groups.size() << endl;
 504
 505                         //make matrix with scores in it
 506                         vector< vector<float> > dists;  dists.resize(globaldata->Groups.size());
 507                         for (int i = 0; i < globaldata->Groups.size(); i++) {
 508                                 dists[i].resize(globaldata->Groups.size(), 0.0);
 509                         }
 510
 511                         //flip it so you can print it
 512                         for (int r=0; r<globaldata->Groups.size(); r++) {
 513                                 for (int l = r+1; l < globaldata->Groups.size(); l++) {
 514                                         dists[r][l] = utreeScores[count];
 515                                         dists[l][r] = utreeScores[count];
 516                                         count++;
 517                                 }
 518                         }
 519
 520                         //output to file
 521                         for (int r=0; r<globaldata->Groups.size(); r++) {
 522                                 //output name
 523                                 string name = globaldata->Groups[r];
 524                                 if (name.length() < 10) { //pad with spaces to make compatible
 525                                         while (name.length() < 10) {  name += " ";  }
 526                                 }
 527                                 out << name << '\t';
 528
 529                                 //output distances
 530                                 for (int l = 0; l < r; l++) {   out  << dists[r][l] << '\t';  }
 531                                 out << endl;
 532                         }
 533                         out.close();
 534                 }
 535         }
 536         catch(exception& e) {
 537                 m->errorOut(e, "UnifracWeightedCommand", "createPhylipFile");
 538                 exit(1);
 539         }
 540 }
 541 /***********************************************************/
 542 int UnifracWeightedCommand::findIndex(float score, int index) {
 543         try{
 544                 for (int i = 0; i < rScores[index].size(); i++) {
 545                         if (rScores[index][i] >= score) {       return i;       }
 546                 }
 547                 return rScores[index].size();
 548         }
 549         catch(exception& e) {
 550                 m->errorOut(e, "UnifracWeightedCommand", "findIndex");
 551                 exit(1);
 552         }
 553 }
 554
 555 /***********************************************************/
 556
 557 void UnifracWeightedCommand::calculateFreqsCumuls() {
 558         try {
 559                 //clear out old tree values
 560                 rScoreFreq.clear();
 561                 rScoreFreq.resize(numComp);
 562                 rCumul.clear();
 563                 rCumul.resize(numComp);
 564                 validScores.clear();
 565
 566                 //calculate frequency
 567                 for (int f = 0; f < numComp; f++) {
 568                         for (int i = 0; i < rScores[f].size(); i++) { //looks like 0,0,1,1,1,2,4,7...  you want to make a map that say rScoreFreq[0] = 2, rScoreFreq[1] = 3...
 569                                 validScores[rScores[f][i]] = rScores[f][i];
 570                                 map<float,float>::iterator it = rScoreFreq[f].find(rScores[f][i]);
 571                                 if (it != rScoreFreq[f].end()) {
 572                                         rScoreFreq[f][rScores[f][i]]++;
 573                                 }else{
 574                                         rScoreFreq[f][rScores[f][i]] = 1;
 575                                 }
 576                         }
 577                 }
 578
 579                 //calculate rcumul
 580                 for(int a = 0; a < numComp; a++) {
 581                         float rcumul = 1.0000;
 582                         //this loop fills the cumulative maps and put 0.0000 in the score freq map to make it easier to print.
 583                         for (map<float,float>::iterator it = validScores.begin(); it != validScores.end(); it++) {
 584                                 //make rscoreFreq map and rCumul
 585                                 map<float,float>::iterator it2 = rScoreFreq[a].find(it->first);
 586                                 rCumul[a][it->first] = rcumul;
 587                                 //get percentage of random trees with that info
 588                                 if (it2 != rScoreFreq[a].end()) {  rScoreFreq[a][it->first] /= iters; rcumul-= it2->second;  }
 589                                 else { rScoreFreq[a][it->first] = 0.0000; } //no random trees with that score
 590                         }
 591                 }
 592
 593         }
 594         catch(exception& e) {
 595                 m->errorOut(e, "UnifracWeightedCommand", "calculateFreqsCums");
 596                 exit(1);
 597         }
 598 }
 599
 600 /***********************************************************/
 601
 602
 603
 604
 605