formatcolumn.cpp

   1 /*
   2  *  formatcolumn.cpp
   3  *  Mothur
   4  *
   5  *  Created by westcott on 1/13/10.
   6  *  Copyright 2010 Schloss Lab. All rights reserved.
   7  *
   8  */
   9
  10 #include "formatcolumn.h"
  11 #include "progress.hpp"
  12
  13 /***********************************************************************/
  14 FormatColumnMatrix::FormatColumnMatrix(string df) : filename(df){
  15         openInputFile(filename, fileHandle);
  16 }
  17 /***********************************************************************/
  18
  19 void FormatColumnMatrix::read(NameAssignment* nameMap){
  20         try {
  21
  22                 string firstName, secondName;
  23                 float distance;
  24                 int nseqs = nameMap->size();
  25
  26                 list = new ListVector(nameMap->getListVector());
  27
  28                 Progress* reading = new Progress("Formatting matrix:     ", nseqs * nseqs);
  29
  30                 int lt = 1;
  31                 int refRow = 0; //we'll keep track of one cell - Cell(refRow,refCol) - and see if it's transpose
  32                 int refCol = 0; //shows up later - Cell(refCol,refRow).  If it does, then its a square matrix
  33
  34                 //need to see if this is a square or a triangular matrix...
  35
  36                 ofstream out;
  37                 string tempOutFile = filename + ".temp";
  38                 openOutputFile(tempOutFile, out);
  39
  40                 while(fileHandle && lt == 1){  //let's assume it's a triangular matrix...
  41
  42                         fileHandle >> firstName >> secondName >> distance;      // get the row and column names and distance
  43
  44                         map<string,int>::iterator itA = nameMap->find(firstName);
  45                         map<string,int>::iterator itB = nameMap->find(secondName);
  46                         if(itA == nameMap->end()){      cerr << "AAError: Sequence '" << firstName << "' was not found in the names file, please correct\n"; exit(1);   }
  47                         if(itB == nameMap->end()){      cerr << "ABError: Sequence '" << secondName << "' was not found in the names file, please correct\n"; exit(1);  }
  48
  49                         if (distance == -1) { distance = 1000000; }
  50
  51                         if((distance < cutoff) && (itA != itB)){
  52                                 if(refRow == refCol){           // in other words, if we haven't loaded refRow and refCol...
  53                                         refRow = itA->second;
  54                                         refCol = itB->second;
  55
  56                                         //making it square
  57                                         out << itA->second << '\t' << itB->second << '\t' << distance << endl;
  58                                         out << itB->second << '\t' << itA->second << '\t' << distance << endl;
  59                                 }
  60                                 else if(refRow == itA->second && refCol == itB->second){        lt = 0;         } //you are square
  61                                 else if(refRow == itB->second && refCol == itA->second){        lt = 0;         }  //you are square
  62                                 else{   //making it square
  63                                         out << itA->second << '\t' << itB->second << '\t' << distance << endl;
  64                                         out << itB->second << '\t' << itA->second << '\t' << distance << endl;
  65                                 }
  66
  67                                 reading->update(itA->second * nseqs / 2);
  68                         }
  69                         gobble(fileHandle);
  70                 }
  71                 out.close();
  72                 fileHandle.close();
  73
  74                 string squareFile;
  75                 if(lt == 0){  // oops, it was square
  76                         squareFile = filename;
  77                 }else{ squareFile = tempOutFile; }
  78
  79                 //sort file by first column so the distances for each row are together
  80                 string outfile = getRootName(squareFile) + "sorted.dist.temp";
  81
  82                 //use the unix sort
  83                 #if defined (__APPLE__) || (__MACH__) || (linux) || (__linux)
  84                         string command = "sort -n " + squareFile + " -o " + outfile;
  85                         system(command.c_str());
  86                 #else //sort using windows sort
  87                         string command = "sort " + squareFile + " /O " + outfile;
  88                         system(command.c_str());
  89                 #endif
  90
  91
  92                 //output to new file distance for each row and save positions in file where new row begins
  93                 ifstream in;
  94                 openInputFile(outfile, in);
  95
  96                 distFile = outfile + ".rowFormatted";
  97                 openOutputFile(distFile, out);
  98
  99                 rowPos.resize(nseqs, -1);
 100                 int currentRow;
 101                 int first, second;
 102                 float dist;
 103                 map<int, float> rowMap;
 104                 map<int, float>::iterator itRow;
 105
 106                 //get first currentRow
 107                 in >> first;
 108                 currentRow = first;
 109
 110                 string firstString = toString(first);
 111                 for(int k = 0; k < firstString.length(); k++)  {   in.putback(firstString[k]);  }
 112
 113                 while(!in.eof()) {
 114                         in >> first >> second >> dist; gobble(in);
 115
 116                         if (first != currentRow) {
 117                                 //save position in file of each new row
 118                                 rowPos[currentRow] = out.tellp();
 119
 120                                 out << currentRow << '\t' << rowMap.size() << '\t';
 121
 122                                 for (itRow = rowMap.begin(); itRow != rowMap.end(); itRow++) {
 123                                         out << itRow->first << '\t' << itRow->second << '\t';
 124                                 }
 125                                 out << endl;
 126
 127                                 currentRow = first;
 128                                 rowMap.clear();
 129
 130                                 //save row you just read
 131                                 if (dist < cutoff) {
 132                                         rowMap[second] = dist;
 133                                 }
 134                         }else{
 135                                 if (dist < cutoff) {
 136                                         rowMap[second] = dist;
 137                                 }
 138                         }
 139                 }
 140
 141                 //print last Row
 142                 //save position in file of each new row
 143                 rowPos[currentRow] = out.tellp();
 144
 145                 out << currentRow << '\t' << rowMap.size() << '\t';
 146
 147                 for (itRow = rowMap.begin(); itRow != rowMap.end(); itRow++) {
 148                         out << itRow->first << '\t' << itRow->second << '\t';
 149                 }
 150                 out << endl;
 151
 152
 153                 in.close();
 154                 out.close();
 155
 156
 157                 remove(tempOutFile.c_str());
 158                 remove(outfile.c_str());
 159
 160                 reading->finish();
 161                 list->setLabel("0");
 162
 163         }
 164         catch(exception& e) {
 165                 m->errorOut(e, "FormatColumnMatrix", "read");
 166                 exit(1);
 167         }
 168 }
 169
 170 /***********************************************************************/
 171 FormatColumnMatrix::~FormatColumnMatrix(){}
 172 /***********************************************************************/
 173
 174
 175