From 626e77d477ecca8b03c942bdc66f8bb8c413b356 Mon Sep 17 00:00:00 2001 From: westcott Date: Tue, 3 Mar 2009 16:41:02 +0000 Subject: [PATCH] unifrac.unweighted command changes --- unifracunweightedcommand.cpp | 146 ++++++++++++++++++++++++----------- unifracunweightedcommand.h | 14 ++-- unweighted.cpp | 65 ++++++++-------- 3 files changed, 142 insertions(+), 83 deletions(-) diff --git a/unifracunweightedcommand.cpp b/unifracunweightedcommand.cpp index f1326a8..5a7cc36 100644 --- a/unifracunweightedcommand.cpp +++ b/unifracunweightedcommand.cpp @@ -16,15 +16,8 @@ UnifracUnweightedCommand::UnifracUnweightedCommand() { T = globaldata->gTree; tmap = globaldata->gTreemap; - unweightedFile = globaldata->getTreeFile() + ".unweighted"; - openOutputFile(unweightedFile, out); - //column headers - out << "Comb" << '\t' << "Score" << '\t' << "UserFreq" << '\t' << "UserCumul" << '\t' << "RandFreq" << '\t' << "RandCumul" << endl; - sumFile = globaldata->getTreeFile() + ".uwsummary"; openOutputFile(sumFile, outSum); - //column headers - outSum << "Tree#" << '\t' << "Comb" << '\t' << "UWScore" << '\t' << '\t' << "UWSig" << endl; setGroups(); //sets users groups to analyze convert(globaldata->getIters(), iters); //how many random trees to generate @@ -47,32 +40,27 @@ int UnifracUnweightedCommand::execute() { userData.resize(numComp,0); //data[0] = unweightedscore randomData.resize(numComp,0); //data[0] = unweightedscore //create new tree with same num nodes and leaves as users - randT = new Tree(); //get pscores for users trees for (int i = 0; i < T.size(); i++) { + unweightedFile = globaldata->getTreeFile() + toString(i+1) + ".unweighted"; + unweightedFileout = globaldata->getTreeFile() + "temp." + toString(i+1) + ".unweighted"; + //column headers + outSum << "Tree# " << i+1 << endl; + outSum << "Comb" << '\t' << "UWScore" << '\t' << '\t' << "UWSig" << endl; + //get unweighted for users tree rscoreFreq.resize(numComp); - uscoreFreq.resize(numComp); rCumul.resize(numComp); - uCumul.resize(numComp); validScores.resize(numComp); utreeScores.resize(numComp); UWScoreSig.resize(numComp); cout << "Processing tree " << i+1 << endl; - outSum << "Tree#" << i+1 << endl; - out << "Tree#" << i+1 << endl; userData = unweighted->getValues(T[i]); //userData[0] = unweightedscore //output scores for each combination for(int k = 0; k < numComp; k++) { - //update uscoreFreq - it = uscoreFreq[k].find(userData[k]); - if (it == uscoreFreq[k].end()) {//new score - uscoreFreq[k][userData[k]] = 1; - }else{ uscoreFreq[k][userData[k]]++; } - //add users score to valid scores validScores[k][userData[k]] = userData[k]; @@ -80,16 +68,13 @@ int UnifracUnweightedCommand::execute() { utreeScores[k].push_back(userData[k]); } - //copy T[i]'s info. - randT->getCopy(T[i]); - //get unweighted scores for random trees for (int j = 0; j < iters; j++) { //we need a different getValues because when we swap the labels we only want to swap those in each parwise comparison - randomData = unweighted->getValues(randT, "", ""); + randomData = unweighted->getValues(T[i], "", ""); for(int k = 0; k < numComp; k++) { -cout << "iter " << j << " comp " << k << " = " << randomData[k] << endl; +//cout << "iter " << j << " comp " << k << " = " << randomData[k] << endl; //add trees unweighted score to map of scores it2 = rscoreFreq[k].find(randomData[k]); if (it2 != rscoreFreq[k].end()) {//already have that score @@ -104,17 +89,9 @@ cout << "iter " << j << " comp " << k << " = " << randomData[k] << endl; } for(int a = 0; a < numComp; a++) { - float ucumul = 1.0000; float rcumul = 1.0000; //this loop fills the cumulative maps and put 0.0000 in the score freq map to make it easier to print. for (it = validScores[a].begin(); it != validScores[a].end(); it++) { - it2 = uscoreFreq[a].find(it->first); - //make uCumul map - uCumul[a][it->first] = ucumul; - //user data has that score - if (it2 != uscoreFreq[a].end()) { uscoreFreq[a][it->first] /= T.size(); ucumul-= it2->second; } - else { uscoreFreq[a][it->first] = 0.0000; } //no user trees with that score - //make rscoreFreq map and rCumul it2 = rscoreFreq[a].find(it->first); rCumul[a][it->first] = rcumul; @@ -128,18 +105,15 @@ cout << "iter " << j << " comp " << k << " = " << randomData[k] << endl; printUnweightedFile(); printUWSummaryFile(); - rscoreFreq.clear(); - uscoreFreq.clear(); + rscoreFreq.clear(); rCumul.clear(); - uCumul.clear(); validScores.clear(); utreeScores.clear(); UWScoreSig.clear(); } //reset groups parameter globaldata->Groups.clear(); globaldata->setGroups(""); - - delete randT; + outSum.close(); return 0; @@ -156,16 +130,22 @@ cout << "iter " << j << " comp " << k << " = " << randomData[k] << endl; /***********************************************************/ void UnifracUnweightedCommand::printUnweightedFile() { try { - //format output - out.setf(ios::fixed, ios::floatfield); out.setf(ios::showpoint); + vector data; for(int a = 0; a < numComp; a++) { + initFile(groupComb[a]); //print each line for (it = validScores[a].begin(); it != validScores[a].end(); it++) { - out << setprecision(6) << groupComb[a] << '\t' << it->first << '\t' << '\t' << uscoreFreq[a][it->first] << '\t' << uCumul[a][it->first] << '\t' << rscoreFreq[a][it->first] << '\t' << rCumul[a][it->first] << endl; + data.push_back(it->first); data.push_back(rscoreFreq[a][it->first]); data.push_back(rCumul[a][it->first]); + output(data); + data.clear(); } + resetFile(); } + out.close(); + inFile.close(); + remove(unweightedFileout.c_str()); } catch(exception& e) { @@ -186,14 +166,11 @@ void UnifracUnweightedCommand::printUWSummaryFile() { outSum.setf(ios::fixed, ios::floatfield); outSum.setf(ios::showpoint); //print each line - for (int i = 0; i< T.size(); i++) { - for(int a = 0; a < numComp; a++) { - outSum << setprecision(6) << i+1 << '\t' << groupComb[a] << '\t' << '\t' << utreeScores[a][i] << '\t' << UWScoreSig[a][i] << endl; - cout << setprecision(6) << i+1 << '\t' << groupComb[a] << '\t' << '\t' << utreeScores[a][i] << '\t' << UWScoreSig[a][i] << endl; - } + for(int a = 0; a < numComp; a++) { + outSum << setprecision(6) << groupComb[a] << '\t' << '\t' << utreeScores[a][0] << '\t' << UWScoreSig[a][0] << endl; + cout << setprecision(6) << groupComb[a] << '\t' << '\t' << utreeScores[a][0] << '\t' << UWScoreSig[a][0] << endl; } - outSum.close(); } catch(exception& e) { cout << "Standard Error: " << e.what() << " has occurred in the UnifracUnweightedCommand class Function printUWSummaryFile. Please contact Pat Schloss at pschloss@microbio.umass.edu." << "\n"; @@ -278,3 +255,82 @@ void UnifracUnweightedCommand::setGroups() { } /*****************************************************************/ +void UnifracUnweightedCommand::initFile(string label){ + try { + if(counter != 0){ + openOutputFile(unweightedFileout, out); + openInputFile(unweightedFile, inFile); + + string inputBuffer; + getline(inFile, inputBuffer); + + out << inputBuffer << '\t' << label + "Score" << '\t' << label + "RandFreq" << '\t' << label + "RandCumul" << endl; + }else{ + openOutputFile(unweightedFileout, out); + out << label + "Score" << '\t' << label + "RandFreq" << '\t' << label + "RandCumul" << endl; + } + + out.setf(ios::fixed, ios::floatfield); + out.setf(ios::showpoint); + } + catch(exception& e) { + cout << "Standard Error: " << e.what() << " has occurred in the UnifracUnweightedCommand class Function initFile. Please contact Pat Schloss at pschloss@microbio.umass.edu." << "\n"; + exit(1); + } + catch(...) { + cout << "An unknown error has occurred in the UnifracUnweightedCommand class function initFile. Please contact Pat Schloss at pschloss@microbio.umass.edu." << "\n"; + exit(1); + } +} + +/***********************************************************************/ + +void UnifracUnweightedCommand::output(vector data){ + try { + if(counter != 0){ + string inputBuffer; + getline(inFile, inputBuffer); + + out << inputBuffer << setprecision(6) << '\t' << data[0] << '\t' << data[1] << '\t' << data[2] << endl; + } + else{ + out << setprecision(6) << data[0] << '\t' << data[1] << '\t' << data[2] << endl; + } + + } + catch(exception& e) { + cout << "Standard Error: " << e.what() << " has occurred in the UnifracUnweightedCommand class Function output. Please contact Pat Schloss at pschloss@microbio.umass.edu." << "\n"; + exit(1); + } + catch(...) { + cout << "An unknown error has occurred in the UnifracUnweightedCommand class function output. Please contact Pat Schloss at pschloss@microbio.umass.edu." << "\n"; + exit(1); + } +}; + +/***********************************************************************/ + +void UnifracUnweightedCommand::resetFile(){ + try { + if(counter != 0){ + out.close(); + inFile.close(); + } + else{ + out.close(); + } + counter = 1; + + remove(unweightedFile.c_str()); + rename(unweightedFileout.c_str(), unweightedFile.c_str()); + } + catch(exception& e) { + cout << "Standard Error: " << e.what() << " has occurred in the UnifracUnweightedCommand class Function resetFile. Please contact Pat Schloss at pschloss@microbio.umass.edu." << "\n"; + exit(1); + } + catch(...) { + cout << "An unknown error has occurred in the UnifracUnweightedCommand class function resetFile. Please contact Pat Schloss at pschloss@microbio.umass.edu." << "\n"; + exit(1); + } +} + diff --git a/unifracunweightedcommand.h b/unifracunweightedcommand.h index e026bc2..b572e46 100644 --- a/unifracunweightedcommand.h +++ b/unifracunweightedcommand.h @@ -28,29 +28,31 @@ class UnifracUnweightedCommand : public Command { private: GlobalData* globaldata; vector T; //user trees - Tree* randT; //random tree TreeMap* tmap; Unweighted* unweighted; - string sumFile, unweightedFile; + string sumFile, unweightedFile, unweightedFileout; vector groupComb; // AB. AC, BC... - int iters, numGroups, numComp; + int iters, numGroups, numComp, counter; EstOutput userData; //unweighted score info for user tree EstOutput randomData; //unweighted score info for random trees vector< vector > utreeScores; //scores for users trees for each comb. vector< vector > UWScoreSig; //tree score signifigance when compared to random trees - percentage of random trees with that score or higher. vector< map > validScores; //map contains scores from both user and random vector< map > rscoreFreq; //map -vector entry for each combination. - vector< map > uscoreFreq; //map -vector entry for each combination. vector< map > rCumul; //map -vector entry for each combination. - vector< map > uCumul; //map -vector entry for each combination. map::iterator it; map::iterator it2; map::iterator it; ofstream outSum, out; + ifstream inFile; void printUWSummaryFile(); void printUnweightedFile(); - void setGroups(); + void setGroups(); + void initFile(string); + void output(vector); + void resetFile(); + }; diff --git a/unweighted.cpp b/unweighted.cpp index 16a2203..47ddd6b 100644 --- a/unweighted.cpp +++ b/unweighted.cpp @@ -205,7 +205,7 @@ EstOutput Unweighted::getValues(Tree* t, string groupA, string groupB) { //if the users enters no groups then give them the score of all groups int numGroups = globaldata->Groups.size(); -cout << "NumGroups = " << numGroups << endl; + //calculate number of comparsions int numComp = 0; for (int r=0; rgetNumLeaves();igetNumNodes();i++){ + for(int i=copyTree->getNumLeaves();igetNumNodes();i++){ - int lc = t->tree[i].getLChild(); //lc = vector index of left child - int rc = t->tree[i].getRChild(); //rc = vector index of right child + int lc = copyTree->tree[i].getLChild(); //lc = vector index of left child + int rc = copyTree->tree[i].getRChild(); //rc = vector index of right child /**********************************************************************/ //This section adds in all lengths that are non leaf - copyIpcount = t->tree[i].pcount; + copyIpcount = copyTree->tree[i].pcount; for (it = copyIpcount.begin(); it != copyIpcount.end(); it++) { if (inUsersGroups(it->first, groups) != true) { copyIpcount.erase(it->first); } } @@ -251,30 +252,30 @@ cout << "NumGroups = " << numGroups << endl; //if i's children are from the same group then i's pcount size will be 1 //if copyIpcount.size() = 0 they are from a branch that is entirely from a group the user doesn't want if (copyIpcount.size() == 0) { } - else if ((t->tree[i].getBranchLength() != -1) && (copyIpcount.size() == 1)) { UniqueBL += t->tree[i].getBranchLength(); } + else if ((copyTree->tree[i].getBranchLength() != -1) && (copyIpcount.size() == 1)) { UniqueBL += copyTree->tree[i].getBranchLength(); } //add i's BL to total if it is from the groups the user wants - if ((t->tree[i].getBranchLength() != -1) && (copyIpcount.size() != 0)) { - totalBL += t->tree[i].getBranchLength(); + if ((copyTree->tree[i].getBranchLength() != -1) && (copyIpcount.size() != 0)) { + totalBL += copyTree->tree[i].getBranchLength(); } /**********************************************************************/ //This section adds in all lengths that are leaf //if i's chidren are leaves - if (t->tree[rc].getRChild() == -1) { + if (copyTree->tree[rc].getRChild() == -1) { //if rc is a valid group and rc has a BL - if ((inUsersGroups(t->tree[rc].getGroup(), groups) == true) && (t->tree[rc].getBranchLength() != -1)) { - UniqueBL += t->tree[rc].getBranchLength(); - totalBL += t->tree[rc].getBranchLength(); + if ((inUsersGroups(copyTree->tree[rc].getGroup(), groups) == true) && (copyTree->tree[rc].getBranchLength() != -1)) { + UniqueBL += copyTree->tree[rc].getBranchLength(); + totalBL += copyTree->tree[rc].getBranchLength(); } } - if (t->tree[lc].getLChild() == -1) { + if (copyTree->tree[lc].getLChild() == -1) { //if lc is a valid group and lc has a BL - if ((inUsersGroups(t->tree[lc].getGroup(), groups) == true) && (t->tree[lc].getBranchLength() != -1)) { - UniqueBL += t->tree[lc].getBranchLength(); - totalBL += t->tree[lc].getBranchLength(); + if ((inUsersGroups(copyTree->tree[lc].getGroup(), groups) == true) && (copyTree->tree[lc].getBranchLength() != -1)) { + UniqueBL += copyTree->tree[lc].getBranchLength(); + totalBL += copyTree->tree[lc].getBranchLength(); } } @@ -315,15 +316,15 @@ cout << "NumGroups = " << numGroups << endl; //swap labels in all the groups you want to compare copyTree->assembleRandomUnifracTree(groups); - for(int i=t->getNumLeaves();igetNumNodes();i++){ + for(int i=copyTree->getNumLeaves();igetNumNodes();i++){ - int lc = t->tree[i].getLChild(); //lc = vector index of left child - int rc = t->tree[i].getRChild(); //rc = vector index of right child + int lc = copyTree->tree[i].getLChild(); //lc = vector index of left child + int rc = copyTree->tree[i].getRChild(); //rc = vector index of right child /**********************************************************************/ //This section adds in all lengths that are non leaf - copyIpcount = t->tree[i].pcount; + copyIpcount = copyTree->tree[i].pcount; for (it = copyIpcount.begin(); it != copyIpcount.end(); it++) { if (inUsersGroups(it->first, groups) != true) { copyIpcount.erase(it->first); } } @@ -331,30 +332,30 @@ cout << "NumGroups = " << numGroups << endl; //if i's children are from the same group then i's pcount size will be 1 //if copyIpcount.size() = 0 they are from a branch that is entirely from a group the user doesn't want if (copyIpcount.size() == 0) { } - else if ((t->tree[i].getBranchLength() != -1) && (copyIpcount.size() == 1)) { UniqueBL += t->tree[i].getBranchLength(); } + else if ((copyTree->tree[i].getBranchLength() != -1) && (copyIpcount.size() == 1)) { UniqueBL += copyTree->tree[i].getBranchLength(); } //add i's BL to total if it is from the groups the user wants - if ((t->tree[i].getBranchLength() != -1) && (copyIpcount.size() != 0)) { - totalBL += t->tree[i].getBranchLength(); + if ((copyTree->tree[i].getBranchLength() != -1) && (copyIpcount.size() != 0)) { + totalBL += copyTree->tree[i].getBranchLength(); } /**********************************************************************/ //This section adds in all lengths that are leaf //if i's chidren are leaves - if (t->tree[rc].getRChild() == -1) { + if (copyTree->tree[rc].getRChild() == -1) { //if rc is a valid group and rc has a BL - if ((inUsersGroups(t->tree[rc].getGroup(), groups) == true) && (t->tree[rc].getBranchLength() != -1)) { - UniqueBL += t->tree[rc].getBranchLength(); - totalBL += t->tree[rc].getBranchLength(); + if ((inUsersGroups(copyTree->tree[rc].getGroup(), groups) == true) && (copyTree->tree[rc].getBranchLength() != -1)) { + UniqueBL += copyTree->tree[rc].getBranchLength(); + totalBL += copyTree->tree[rc].getBranchLength(); } } - if (t->tree[lc].getLChild() == -1) { + if (copyTree->tree[lc].getLChild() == -1) { //if lc is a valid group and lc has a BL - if ((inUsersGroups(t->tree[lc].getGroup(), groups) == true) && (t->tree[lc].getBranchLength() != -1)) { - UniqueBL += t->tree[lc].getBranchLength(); - totalBL += t->tree[lc].getBranchLength(); + if ((inUsersGroups(copyTree->tree[lc].getGroup(), groups) == true) && (copyTree->tree[lc].getBranchLength() != -1)) { + UniqueBL += copyTree->tree[lc].getBranchLength(); + totalBL += copyTree->tree[lc].getBranchLength(); } } -- 2.39.2