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                 abort = true;
  28                 //initialize outputTypes
  29                 vector<string> tempOutNames;
  30                 outputTypes["weighted"] = tempOutNames;
  31                 outputTypes["wsummary"] = tempOutNames;
  32                 outputTypes["phylip"] = tempOutNames;
  33         }
  34         catch(exception& e) {
  35                 m->errorOut(e, "UnifracWeightedCommand", "UnifracWeightedCommand");
  36                 exit(1);
  37         }
  38 }
  39 //**********************************************************************************************************************
  40 vector<string> UnifracWeightedCommand::getRequiredParameters(){
  41         try {
  42                 vector<string> myArray;
  43                 return myArray;
  44         }
  45         catch(exception& e) {
  46                 m->errorOut(e, "UnifracWeightedCommand", "getRequiredParameters");
  47                 exit(1);
  48         }
  49 }
  50 //**********************************************************************************************************************
  51 vector<string> UnifracWeightedCommand::getRequiredFiles(){
  52         try {
  53                 string Array[] =  {"tree","group"};
  54                 vector<string> myArray (Array, Array+(sizeof(Array)/sizeof(string)));
  55
  56                 return myArray;
  57         }
  58         catch(exception& e) {
  59                 m->errorOut(e, "UnifracWeightedCommand", "getRequiredFiles");
  60                 exit(1);
  61         }
  62 }
  63 /***********************************************************/
  64 UnifracWeightedCommand::UnifracWeightedCommand(string option) {
  65         try {
  66                 globaldata = GlobalData::getInstance();
  67                 abort = false;
  68                 Groups.clear();
  69
  70                 //allow user to run help
  71                 if(option == "help") { help(); abort = true; }
  72
  73                 else {
  74                         //valid paramters for this command
  75                         string Array[] =  {"groups","iters","distance","random","processors","outputdir","inputdir"};
  76                         vector<string> myArray (Array, Array+(sizeof(Array)/sizeof(string)));
  77
  78                         OptionParser parser(option);
  79                         map<string,string> parameters=parser.getParameters();
  80
  81                         ValidParameters validParameter;
  82
  83                         //check to make sure all parameters are valid for command
  84                         for (map<string,string>::iterator it = parameters.begin(); it != parameters.end(); it++) {
  85                                 if (validParameter.isValidParameter(it->first, myArray, it->second) != true) {  abort = true;  }
  86                         }
  87
  88                         //initialize outputTypes
  89                         vector<string> tempOutNames;
  90                         outputTypes["weighted"] = tempOutNames;
  91                         outputTypes["wsummary"] = tempOutNames;
  92                         outputTypes["phylip"] = tempOutNames;
  93
  94                         if (globaldata->gTree.size() == 0) {//no trees were read
  95                                 m->mothurOut("You must execute the read.tree command, before you may execute the unifrac.weighted command."); m->mothurOutEndLine(); abort = true;  }
  96
  97                         //if the user changes the output directory command factory will send this info to us in the output parameter
  98                         outputDir = validParameter.validFile(parameters, "outputdir", false);           if (outputDir == "not found"){
  99                                 outputDir = "";
 100                                 outputDir += m->hasPath(globaldata->inputFileName); //if user entered a file with a path then preserve it
 101                         }
 102
 103                         //check for optional parameter and set defaults
 104                         // ...at some point should added some additional type checking...
 105                         groups = validParameter.validFile(parameters, "groups", false);
 106                         if (groups == "not found") { groups = ""; }
 107                         else {
 108                                 m->splitAtDash(groups, Groups);
 109                                 globaldata->Groups = Groups;
 110                         }
 111
 112                         itersString = validParameter.validFile(parameters, "iters", false);                     if (itersString == "not found") { itersString = "1000"; }
 113                         convert(itersString, iters);
 114
 115                         string temp = validParameter.validFile(parameters, "distance", false);                  if (temp == "not found") { temp = "false"; }
 116                         phylip = m->isTrue(temp);
 117
 118                         temp = validParameter.validFile(parameters, "random", false);                                   if (temp == "not found") { temp = "F"; }
 119                         random = m->isTrue(temp);
 120
 121                         temp = validParameter.validFile(parameters, "processors", false);       if (temp == "not found"){       temp = "1";                             }
 122                         convert(temp, processors);
 123
 124                         if (!random) {  iters = 0;  } //turn off random calcs
 125
 126
 127                         if (abort == false) {
 128                                 T = globaldata->gTree;
 129                                 tmap = globaldata->gTreemap;
 130                                 sumFile = outputDir + m->getSimpleName(globaldata->getTreeFile()) + ".wsummary";
 131                                 m->openOutputFile(sumFile, outSum);
 132                                 outputNames.push_back(sumFile);  outputTypes["wsummary"].push_back(sumFile);
 133
 134                                 util = new SharedUtil();
 135                                 string s; //to make work with setgroups
 136                                 util->setGroups(globaldata->Groups, tmap->namesOfGroups, s, numGroups, "weighted");     //sets the groups the user wants to analyze
 137                                 util->getCombos(groupComb, globaldata->Groups, numComp);
 138
 139                                 weighted = new Weighted(tmap);
 140
 141                         }
 142                 }
 143
 144
 145         }
 146         catch(exception& e) {
 147                 m->errorOut(e, "UnifracWeightedCommand", "UnifracWeightedCommand");
 148                 exit(1);
 149         }
 150 }
 151 //**********************************************************************************************************************
 152
 153 void UnifracWeightedCommand::help(){
 154         try {
 155                 m->mothurOut("The unifrac.weighted command can only be executed after a successful read.tree command.\n");
 156                 m->mothurOut("The unifrac.weighted command parameters are groups, iters, distance, processors and random.  No parameters are required.\n");
 157                 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");
 158                 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");
 159                 m->mothurOut("The distance parameter allows you to create a distance file from the results. The default is false.\n");
 160                 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");
 161                 m->mothurOut("The processors parameter allows you to specify the number of processors to use. The default is 1.\n");
 162                 m->mothurOut("The unifrac.weighted command should be in the following format: unifrac.weighted(groups=yourGroups, iters=yourIters).\n");
 163                 m->mothurOut("Example unifrac.weighted(groups=A-B-C, iters=500).\n");
 164                 m->mothurOut("The default value for groups is all the groups in your groupfile, and iters is 1000.\n");
 165                 m->mothurOut("The unifrac.weighted command output two files: .weighted and .wsummary their descriptions are in the manual.\n");
 166                 m->mothurOut("Note: No spaces between parameter labels (i.e. groups), '=' and parameters (i.e.yourGroups).\n\n");
 167         }
 168         catch(exception& e) {
 169                 m->errorOut(e, "UnifracWeightedCommand", "help");
 170                 exit(1);
 171         }
 172 }
 173
 174 /***********************************************************/
 175 int UnifracWeightedCommand::execute() {
 176         try {
 177
 178                 if (abort == true) { return 0; }
 179
 180                 int start = time(NULL);
 181
 182                 //get weighted for users tree
 183                 userData.resize(numComp,0);  //data[0] = weightedscore AB, data[1] = weightedscore AC...
 184                 randomData.resize(numComp,0); //data[0] = weightedscore AB, data[1] = weightedscore AC...
 185
 186                 //get weighted scores for users trees
 187                 for (int i = 0; i < T.size(); i++) {
 188
 189                         if (m->control_pressed) { outSum.close(); for (int i = 0; i < outputNames.size(); i++) {        remove(outputNames[i].c_str());  } return 0; }
 190
 191                         counter = 0;
 192                         rScores.resize(numComp);  //data[0] = weightedscore AB, data[1] = weightedscore AC...
 193                         uScores.resize(numComp);  //data[0] = weightedscore AB, data[1] = weightedscore AC...
 194
 195                         if (random) {
 196                                 output = new ColumnFile(outputDir + m->getSimpleName(globaldata->getTreeFile())  + toString(i+1) + ".weighted", itersString);
 197                                 outputNames.push_back(outputDir + m->getSimpleName(globaldata->getTreeFile())  + toString(i+1) + ".weighted");
 198                                 outputTypes["weighted"].push_back(outputDir + m->getSimpleName(globaldata->getTreeFile())  + toString(i+1) + ".weighted");
 199                         }
 200
 201                         userData = weighted->getValues(T[i], processors, outputDir);  //userData[0] = weightedscore
 202
 203                         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; }
 204
 205                         //save users score
 206                         for (int s=0; s<numComp; s++) {
 207                                 //add users score to vector of user scores
 208                                 uScores[s].push_back(userData[s]);
 209
 210                                 //save users tree score for summary file
 211                                 utreeScores.push_back(userData[s]);
 212                         }
 213
 214                         if (random) {
 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(),  rScores);
 249                                                 }else{
 250                                                         createProcesses(T[i],  namesOfGroupCombos, rScores);
 251                                                 }
 252                                         #else
 253                                                 driver(T[i], namesOfGroupCombos, 0, namesOfGroupCombos.size(), 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< 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, 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 {
 353                                 m->mothurOut("[ERROR]: unable to spawn the necessary processes."); m->mothurOutEndLine();
 354                                 for (int i = 0; i < processIDS.size(); i++) { kill (processIDS[i], SIGINT); }
 355                                 exit(0);
 356                         }
 357                 }
 358
 359                 driver(t, namesOfGroupCombos, lines[0].start, lines[0].num, scores);
 360
 361                 //force parent to wait until all the processes are done
 362                 for (int i=0;i<(processors-1);i++) {
 363                         int temp = processIDS[i];
 364                         wait(&temp);
 365                 }
 366
 367                 //get data created by processes
 368                 for (int i=0;i<(processors-1);i++) {
 369
 370                         ifstream in;
 371                         string s = outputDir + toString(processIDS[i]) + ".weightedcommand.results.temp";
 372                         m->openInputFile(s, in);
 373
 374                         double tempScore;
 375                         for (int j = lines[(i+1)].start; j < (lines[(i+1)].start + lines[(i+1)].num); j++) { in >> tempScore; scores[j].push_back(tempScore); }
 376                         in.close();
 377                         remove(s.c_str());
 378                 }
 379
 380                 return 0;
 381 #endif
 382         }
 383         catch(exception& e) {
 384                 m->errorOut(e, "UnifracWeightedCommand", "createProcesses");
 385                 exit(1);
 386         }
 387 }
 388
 389 /**************************************************************************************************/
 390 int UnifracWeightedCommand::driver(Tree* t, vector< vector<string> > namesOfGroupCombos, int start, int num, vector< vector<double> >& scores) {
 391  try {
 392                 Tree* randT = new Tree();
 393
 394                 for (int h = start; h < (start+num); h++) {
 395
 396                         if (m->control_pressed) { return 0; }
 397
 398                         //initialize weighted score
 399                         string groupA = namesOfGroupCombos[h][0];
 400                         string groupB = namesOfGroupCombos[h][1];
 401
 402                         //copy T[i]'s info.
 403                         randT->getCopy(t);
 404
 405                         //create a random tree with same topology as T[i], but different labels
 406                         randT->assembleRandomUnifracTree(groupA, groupB);
 407
 408                         if (m->control_pressed) { delete randT;  return 0;  }
 409
 410                         //get wscore of random tree
 411                         EstOutput randomData = weighted->getValues(randT, groupA, groupB);
 412
 413                         if (m->control_pressed) { delete randT;  return 0;  }
 414
 415                         //save scores
 416                         scores[h].push_back(randomData[0]);
 417                 }
 418
 419                 delete randT;
 420
 421                 return 0;
 422
 423         }
 424         catch(exception& e) {
 425                 m->errorOut(e, "UnifracWeightedCommand", "driver");
 426                 exit(1);
 427         }
 428 }
 429 /***********************************************************/
 430 void UnifracWeightedCommand::printWeightedFile() {
 431         try {
 432                 vector<double> data;
 433                 vector<string> tags;
 434                 tags.push_back("Score"); tags.push_back("RandFreq"); tags.push_back("RandCumul");
 435
 436                 for(int a = 0; a < numComp; a++) {
 437                         output->initFile(groupComb[a], tags);
 438                         //print each line
 439                         for (map<float,float>::iterator it = validScores.begin(); it != validScores.end(); it++) {
 440                                 data.push_back(it->first);  data.push_back(rScoreFreq[a][it->first]); data.push_back(rCumul[a][it->first]);
 441                                 output->output(data);
 442                                 data.clear();
 443                         }
 444                         output->resetFile();
 445                 }
 446         }
 447         catch(exception& e) {
 448                 m->errorOut(e, "UnifracWeightedCommand", "printWeightedFile");
 449                 exit(1);
 450         }
 451 }
 452
 453
 454 /***********************************************************/
 455 void UnifracWeightedCommand::printWSummaryFile() {
 456         try {
 457                 //column headers
 458                 outSum << "Tree#" << '\t' << "Groups" << '\t' << "WScore" << '\t' << "WSig" <<  endl;
 459                 m->mothurOut("Tree#\tGroups\tWScore\tWSig"); m->mothurOutEndLine();
 460
 461                 //format output
 462                 outSum.setf(ios::fixed, ios::floatfield); outSum.setf(ios::showpoint);
 463
 464                 //print each line
 465                 int count = 0;
 466                 for (int i = 0; i < T.size(); i++) {
 467                         for (int j = 0; j < numComp; j++) {
 468                                 if (random) {
 469                                         if (WScoreSig[count] > (1/(float)iters)) {
 470                                                 outSum << setprecision(6) << i+1 << '\t' << groupComb[j] << '\t' << utreeScores[count] << '\t' << setprecision(itersString.length()) << WScoreSig[count] << endl;
 471                                                 cout << setprecision(6) << i+1 << '\t' << groupComb[j] << '\t' << utreeScores[count] << '\t' << setprecision(itersString.length()) << WScoreSig[count] << endl;
 472                                                 m->mothurOutJustToLog(toString(i+1) +"\t" + groupComb[j] +"\t" + toString(utreeScores[count]) +"\t" +  toString(WScoreSig[count]) + "\n");
 473                                         }else{
 474                                                 outSum << setprecision(6) << i+1 << '\t' << groupComb[j] << '\t' << utreeScores[count] << '\t' << setprecision(itersString.length()) << "<" << (1/float(iters)) << endl;
 475                                                 cout << setprecision(6) << i+1 << '\t' << groupComb[j] << '\t' << utreeScores[count] << '\t' << setprecision(itersString.length()) << "<" << (1/float(iters)) << endl;
 476                                                 m->mothurOutJustToLog(toString(i+1) +"\t" + groupComb[j] +"\t" + toString(utreeScores[count]) +"\t<" +  toString((1/float(iters))) + "\n");
 477                                         }
 478                                 }else{
 479                                         outSum << setprecision(6) << i+1 << '\t' << groupComb[j] << '\t' << utreeScores[count] << '\t' << "0.00" << endl;
 480                                         cout << setprecision(6) << i+1 << '\t' << groupComb[j] << '\t' << utreeScores[count] << '\t' << "0.00" << endl;
 481                                         m->mothurOutJustToLog(toString(i+1) +"\t" + groupComb[j] +"\t" + toString(utreeScores[count]) +"\t0.00\n");
 482                                 }
 483                                 count++;
 484                         }
 485                 }
 486                 outSum.close();
 487         }
 488         catch(exception& e) {
 489                 m->errorOut(e, "UnifracWeightedCommand", "printWSummaryFile");
 490                 exit(1);
 491         }
 492 }
 493 /***********************************************************/
 494 void UnifracWeightedCommand::createPhylipFile() {
 495         try {
 496                 int count = 0;
 497                 //for each tree
 498                 for (int i = 0; i < T.size(); i++) {
 499
 500                         string phylipFileName = outputDir + m->getSimpleName(globaldata->getTreeFile())  + toString(i+1) + ".weighted.dist";
 501                         outputNames.push_back(phylipFileName);
 502                         outputTypes["phylip"].push_back(phylipFileName);
 503                         ofstream out;
 504                         m->openOutputFile(phylipFileName, out);
 505
 506                         //output numSeqs
 507                         out << globaldata->Groups.size() << endl;
 508
 509                         //make matrix with scores in it
 510                         vector< vector<float> > dists;  dists.resize(globaldata->Groups.size());
 511                         for (int i = 0; i < globaldata->Groups.size(); i++) {
 512                                 dists[i].resize(globaldata->Groups.size(), 0.0);
 513                         }
 514
 515                         //flip it so you can print it
 516                         for (int r=0; r<globaldata->Groups.size(); r++) {
 517                                 for (int l = r+1; l < globaldata->Groups.size(); l++) {
 518                                         dists[r][l] = utreeScores[count];
 519                                         dists[l][r] = utreeScores[count];
 520                                         count++;
 521                                 }
 522                         }
 523
 524                         //output to file
 525                         for (int r=0; r<globaldata->Groups.size(); r++) {
 526                                 //output name
 527                                 string name = globaldata->Groups[r];
 528                                 if (name.length() < 10) { //pad with spaces to make compatible
 529                                         while (name.length() < 10) {  name += " ";  }
 530                                 }
 531                                 out << name << '\t';
 532
 533                                 //output distances
 534                                 for (int l = 0; l < r; l++) {   out  << dists[r][l] << '\t';  }
 535                                 out << endl;
 536                         }
 537                         out.close();
 538                 }
 539         }
 540         catch(exception& e) {
 541                 m->errorOut(e, "UnifracWeightedCommand", "createPhylipFile");
 542                 exit(1);
 543         }
 544 }
 545 /***********************************************************/
 546 int UnifracWeightedCommand::findIndex(float score, int index) {
 547         try{
 548                 for (int i = 0; i < rScores[index].size(); i++) {
 549                         if (rScores[index][i] >= score) {       return i;       }
 550                 }
 551                 return rScores[index].size();
 552         }
 553         catch(exception& e) {
 554                 m->errorOut(e, "UnifracWeightedCommand", "findIndex");
 555                 exit(1);
 556         }
 557 }
 558
 559 /***********************************************************/
 560
 561 void UnifracWeightedCommand::calculateFreqsCumuls() {
 562         try {
 563                 //clear out old tree values
 564                 rScoreFreq.clear();
 565                 rScoreFreq.resize(numComp);
 566                 rCumul.clear();
 567                 rCumul.resize(numComp);
 568                 validScores.clear();
 569
 570                 //calculate frequency
 571                 for (int f = 0; f < numComp; f++) {
 572                         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...
 573                                 validScores[rScores[f][i]] = rScores[f][i];
 574                                 map<float,float>::iterator it = rScoreFreq[f].find(rScores[f][i]);
 575                                 if (it != rScoreFreq[f].end()) {
 576                                         rScoreFreq[f][rScores[f][i]]++;
 577                                 }else{
 578                                         rScoreFreq[f][rScores[f][i]] = 1;
 579                                 }
 580                         }
 581                 }
 582
 583                 //calculate rcumul
 584                 for(int a = 0; a < numComp; a++) {
 585                         float rcumul = 1.0000;
 586                         //this loop fills the cumulative maps and put 0.0000 in the score freq map to make it easier to print.
 587                         for (map<float,float>::iterator it = validScores.begin(); it != validScores.end(); it++) {
 588                                 //make rscoreFreq map and rCumul
 589                                 map<float,float>::iterator it2 = rScoreFreq[a].find(it->first);
 590                                 rCumul[a][it->first] = rcumul;
 591                                 //get percentage of random trees with that info
 592                                 if (it2 != rScoreFreq[a].end()) {  rScoreFreq[a][it->first] /= iters; rcumul-= it2->second;  }
 593                                 else { rScoreFreq[a][it->first] = 0.0000; } //no random trees with that score
 594                         }
 595                 }
 596
 597         }
 598         catch(exception& e) {
 599                 m->errorOut(e, "UnifracWeightedCommand", "calculateFreqsCums");
 600                 exit(1);
 601         }
 602 }
 603
 604 /***********************************************************/
 605
 606
 607
 608
 609