X-Git-Url: https://git.donarmstrong.com/?a=blobdiff_plain;f=tree.cpp;h=d9d71aea2dd1859d6e198302325fb51d2fdf9367;hb=d6c0a11d1cecfac18b323285e7ffadb7f58e848f;hp=f8bb76fd664383f148c73eca62aa5cc2291202c8;hpb=9b884118e97081d743c7b001216567f9b9131258;p=mothur.git diff --git a/tree.cpp b/tree.cpp index f8bb76f..d9d71ae 100644 --- a/tree.cpp +++ b/tree.cpp @@ -10,12 +10,28 @@ #include "tree.h" /*****************************************************************/ -Tree::Tree(string g) { +Tree::Tree(int num, TreeMap* t) : tmap(t) { try { - globaldata = GlobalData::getInstance(); m = MothurOut::getInstance(); - parseTreeFile(); globaldata->runParse = false; + numLeaves = num; + numNodes = 2*numLeaves - 1; + + tree.resize(numNodes); + } + catch(exception& e) { + m->errorOut(e, "Tree", "Tree - numNodes"); + exit(1); + } +} +/*****************************************************************/ +Tree::Tree(string g) { //do not use tree generated by this its just to extract the treenames, its a chicken before the egg thing that needs to be revisited. + try { + m = MothurOut::getInstance(); + + tmap = NULL; + + parseTreeFile(); m->runParse = false; } catch(exception& e) { m->errorOut(e, "Tree", "Tree - just parse"); @@ -23,31 +39,31 @@ Tree::Tree(string g) { } } /*****************************************************************/ -Tree::Tree() { +Tree::Tree(TreeMap* t) : tmap(t) { try { - globaldata = GlobalData::getInstance(); m = MothurOut::getInstance(); - if (globaldata->runParse == true) { parseTreeFile(); globaldata->runParse = false; } + if (m->runParse == true) { parseTreeFile(); m->runParse = false; } //for(int i = 0; i < globaldata->Treenames.size(); i++) { cout << i << '\t' << globaldata->Treenames[i] << endl; } - numLeaves = globaldata->Treenames.size(); + numLeaves = m->Treenames.size(); numNodes = 2*numLeaves - 1; tree.resize(numNodes); - + //initialize groupNodeInfo - for (int i = 0; i < globaldata->gTreemap->namesOfGroups.size(); i++) { - groupNodeInfo[globaldata->gTreemap->namesOfGroups[i]].resize(0); + for (int i = 0; i < (tmap->getNamesOfGroups()).size(); i++) { + groupNodeInfo[(tmap->getNamesOfGroups())[i]].resize(0); } - + //initialize tree with correct number of nodes, name and group info. for (int i = 0; i < numNodes; i++) { //initialize leaf nodes if (i <= (numLeaves-1)) { - tree[i].setName(globaldata->Treenames[i]); + tree[i].setName(m->Treenames[i]); //save group info - string group = globaldata->gTreemap->getGroup(globaldata->Treenames[i]); + string group = tmap->getGroup(m->Treenames[i]); + vector tempGroups; tempGroups.push_back(group); tree[i].setGroup(tempGroups); groupNodeInfo[group].push_back(i); @@ -57,7 +73,7 @@ Tree::Tree() { tree[i].pGroups[group] = 1; //Treemap knows name, group and index to speed up search - globaldata->gTreemap->setIndex(globaldata->Treenames[i], i); + tmap->setIndex(m->Treenames[i], i); //intialize non leaf nodes }else if (i > (numLeaves-1)) { @@ -66,6 +82,7 @@ Tree::Tree() { tree[i].setGroup(tempGroups); } } + } catch(exception& e) { m->errorOut(e, "Tree", "Tree"); @@ -99,19 +116,19 @@ void Tree::addNamesToCounts() { string name = tree[i].getName(); - map::iterator itNames = globaldata->names.find(name); + map::iterator itNames = m->names.find(name); - if (itNames == globaldata->names.end()) { m->mothurOut(name + " is not in your name file, please correct."); m->mothurOutEndLine(); exit(1); } + if (itNames == m->names.end()) { m->mothurOut(name + " is not in your name file, please correct."); m->mothurOutEndLine(); exit(1); } else { vector dupNames; - m->splitAtComma(globaldata->names[name], dupNames); + m->splitAtComma(m->names[name], dupNames); map::iterator itCounts; int maxPars = 1; set groupsAddedForThisNode; for (int j = 0; j < dupNames.size(); j++) { - string group = globaldata->gTreemap->getGroup(dupNames[j]); + string group = tmap->getGroup(dupNames[j]); if (dupNames[j] != name) {//you already added yourself in the constructor @@ -129,7 +146,7 @@ void Tree::addNamesToCounts() { itCounts = tree[i].pGroups.find(group); if (itCounts == tree[i].pGroups.end()) { //new group, add it tree[i].pGroups[group] = 1; - }else { + }else{ tree[i].pGroups[group]++; } @@ -178,7 +195,7 @@ int Tree::getIndex(string searchName) { try { //Treemap knows name, group and index to speed up search // getIndex function will return the vector index or -1 if seq is not found. - int index = globaldata->gTreemap->getIndex(searchName); + int index = tmap->getIndex(searchName); return index; } @@ -192,7 +209,7 @@ int Tree::getIndex(string searchName) { void Tree::setIndex(string searchName, int index) { try { //set index in treemap - globaldata->gTreemap->setIndex(searchName, index); + tmap->setIndex(searchName, index); } catch(exception& e) { m->errorOut(e, "Tree", "setIndex"); @@ -205,7 +222,7 @@ int Tree::assembleTree() { //float A = clock(); //if user has given a names file we want to include that info in the pgroups and pcount info. - if(globaldata->names.size() != 0) { addNamesToCounts(); } + if(m->names.size() != 0) { addNamesToCounts(); } //build the pGroups in non leaf nodes to be used in the parsimony calcs. for (int i = numLeaves; i < numNodes; i++) { @@ -243,7 +260,190 @@ int Tree::assembleTree(string n) { exit(1); } } +/*****************************************************************/ +void Tree::getSubTree(Tree* copy, vector Groups) { + try { + + //we want to select some of the leaf nodes to create the output tree + //go through the input Tree starting at parents of leaves + for (int i = 0; i < numNodes; i++) { + + //initialize leaf nodes + if (i <= (numLeaves-1)) { + tree[i].setName(Groups[i]); + + //save group info + string group = tmap->getGroup(Groups[i]); + vector tempGroups; tempGroups.push_back(group); + tree[i].setGroup(tempGroups); + groupNodeInfo[group].push_back(i); + + //set pcount and pGroup for groupname to 1. + tree[i].pcount[group] = 1; + tree[i].pGroups[group] = 1; + + //Treemap knows name, group and index to speed up search + tmap->setIndex(Groups[i], i); + + //intialize non leaf nodes + }else if (i > (numLeaves-1)) { + tree[i].setName(""); + vector tempGroups; + tree[i].setGroup(tempGroups); + } + } + + set removedLeaves; + for (int i = 0; i < copy->getNumLeaves(); i++) { + + if (removedLeaves.count(i) == 0) { + + //am I in the group + int parent = copy->tree[i].getParent(); + + if (parent != -1) { + + if (m->inUsersGroups(copy->tree[i].getName(), Groups)) { + //find my siblings name + int parentRC = copy->tree[parent].getRChild(); + int parentLC = copy->tree[parent].getLChild(); + + //if I am the right child, then my sib is the left child + int sibIndex = parentRC; + if (parentRC == i) { sibIndex = parentLC; } + + string sibsName = copy->tree[sibIndex].getName(); + + //if yes, is my sibling + if ((m->inUsersGroups(sibsName, Groups)) || (sibsName == "")) { + //we both are okay no trimming required + }else{ + //i am, my sib is not, so remove sib by setting my parent to my grandparent + int grandparent = copy->tree[parent].getParent(); + int grandparentLC = copy->tree[grandparent].getLChild(); + int grandparentRC = copy->tree[grandparent].getRChild(); + + //whichever of my granparents children was my parent now equals me + if (grandparentLC == parent) { grandparentLC = i; } + else { grandparentRC = i; } + + copy->tree[i].setParent(grandparent); + copy->tree[i].setBranchLength((copy->tree[i].getBranchLength()+copy->tree[parent].getBranchLength())); + if (grandparent != -1) { + copy->tree[grandparent].setChildren(grandparentLC, grandparentRC); + } + removedLeaves.insert(sibIndex); + } + }else{ + //find my siblings name + int parentRC = copy->tree[parent].getRChild(); + int parentLC = copy->tree[parent].getLChild(); + + //if I am the right child, then my sib is the left child + int sibIndex = parentRC; + if (parentRC == i) { sibIndex = parentLC; } + + string sibsName = copy->tree[sibIndex].getName(); + + //if no is my sibling + if ((m->inUsersGroups(sibsName, Groups)) || (sibsName == "")) { + //i am not, but my sib is + int grandparent = copy->tree[parent].getParent(); + int grandparentLC = copy->tree[grandparent].getLChild(); + int grandparentRC = copy->tree[grandparent].getRChild(); + + //whichever of my granparents children was my parent now equals my sib + if (grandparentLC == parent) { grandparentLC = sibIndex; } + else { grandparentRC = sibIndex; } + + copy->tree[sibIndex].setParent(grandparent); + copy->tree[sibIndex].setBranchLength((copy->tree[sibIndex].getBranchLength()+copy->tree[parent].getBranchLength())); + if (grandparent != -1) { + copy->tree[grandparent].setChildren(grandparentLC, grandparentRC); + } + removedLeaves.insert(i); + }else{ + //neither of us are, so we want to eliminate ourselves and our parent + //so set our parents sib to our great-grandparent + int parent = copy->tree[i].getParent(); + int grandparent = copy->tree[parent].getParent(); + int parentsSibIndex; + if (grandparent != -1) { + int greatgrandparent = copy->tree[grandparent].getParent(); + int greatgrandparentLC, greatgrandparentRC; + if (greatgrandparent != -1) { + greatgrandparentLC = copy->tree[greatgrandparent].getLChild(); + greatgrandparentRC = copy->tree[greatgrandparent].getRChild(); + } + + int grandparentLC = copy->tree[grandparent].getLChild(); + int grandparentRC = copy->tree[grandparent].getRChild(); + + parentsSibIndex = grandparentLC; + if (grandparentLC == parent) { parentsSibIndex = grandparentRC; } + + //whichever of my greatgrandparents children was my grandparent + if (greatgrandparentLC == grandparent) { greatgrandparentLC = parentsSibIndex; } + else { greatgrandparentRC = parentsSibIndex; } + + copy->tree[parentsSibIndex].setParent(greatgrandparent); + copy->tree[parentsSibIndex].setBranchLength((copy->tree[parentsSibIndex].getBranchLength()+copy->tree[grandparent].getBranchLength())); + if (greatgrandparent != -1) { + copy->tree[greatgrandparent].setChildren(greatgrandparentLC, greatgrandparentRC); + } + }else{ + copy->tree[parent].setParent(-1); + //cout << "issues with making subtree" << endl; + } + removedLeaves.insert(sibIndex); + removedLeaves.insert(i); + } + } + } + } + } + + int root = 0; + for (int i = 0; i < copy->getNumNodes(); i++) { + //you found the root + if (copy->tree[i].getParent() == -1) { root = i; break; } + } + + int nextSpot = numLeaves; + populateNewTree(copy->tree, root, nextSpot); + } + catch(exception& e) { + m->errorOut(e, "Tree", "getCopy"); + exit(1); + } +} +/*****************************************************************/ +int Tree::populateNewTree(vector& oldtree, int node, int& index) { + try { + + if (oldtree[node].getLChild() != -1) { + int rc = populateNewTree(oldtree, oldtree[node].getLChild(), index); + int lc = populateNewTree(oldtree, oldtree[node].getRChild(), index); + tree[index].setChildren(lc, rc); + tree[rc].setParent(index); + tree[lc].setParent(index); + + tree[index].setBranchLength(oldtree[node].getBranchLength()); + tree[rc].setBranchLength(oldtree[oldtree[node].getLChild()].getBranchLength()); + tree[lc].setBranchLength(oldtree[oldtree[node].getRChild()].getBranchLength()); + + return (index++); + }else { //you are a leaf + int indexInNewTree = tmap->getIndex(oldtree[node].getName()); + return indexInNewTree; + } + } + catch(exception& e) { + m->errorOut(e, "Tree", "populateNewTree"); + exit(1); + } +} /*****************************************************************/ void Tree::getCopy(Tree* copy) { try { @@ -432,8 +632,8 @@ void Tree::randomLabels(vector g) { try { //initialize groupNodeInfo - for (int i = 0; i < globaldata->gTreemap->namesOfGroups.size(); i++) { - groupNodeInfo[globaldata->gTreemap->namesOfGroups[i]].resize(0); + for (int i = 0; i < (tmap->getNamesOfGroups()).size(); i++) { + groupNodeInfo[(tmap->getNamesOfGroups())[i]].resize(0); } for(int i = 0; i < numLeaves; i++){ @@ -506,7 +706,7 @@ void Tree::randomLabels(string groupA, string groupB) { exit(1); } } -/**************************************************************************************************/ +**************************************************************************************************/ void Tree::randomBlengths() { try { for(int i=numNodes-1;i>=0;i--){ @@ -592,24 +792,23 @@ void Tree::print(ostream& out) { } } /*****************************************************************/ -void Tree::printForBoot(ostream& out) { +void Tree::print(ostream& out, string mode) { try { int root = findRoot(); - printBranch(root, out, "boot"); + printBranch(root, out, mode); out << ";" << endl; } catch(exception& e) { - m->errorOut(e, "Tree", "printForBoot"); + m->errorOut(e, "Tree", "print"); exit(1); } } - /*****************************************************************/ // This prints out the tree in Newick form. void Tree::createNewickFile(string f) { try { int root = findRoot(); - //filename = m->getRootName(globaldata->getTreeFile()) + "newick"; + filename = f; m->openOutputFile(filename, out); @@ -644,12 +843,11 @@ int Tree::findRoot() { exit(1); } } - /*****************************************************************/ void Tree::printBranch(int node, ostream& out, string mode) { - try { - - // you are not a leaf +try { + +// you are not a leaf if (tree[node].getLChild() != -1) { out << "("; printBranch(tree[node].getLChild(), out, mode); @@ -666,21 +864,154 @@ void Tree::printBranch(int node, ostream& out, string mode) { if (tree[node].getLabel() != -1) { out << tree[node].getLabel(); } + }else if (mode == "both") { + if (tree[node].getLabel() != -1) { + out << tree[node].getLabel(); + } + //if there is a branch length then print it + if (tree[node].getBranchLength() != -1) { + out << ":" << tree[node].getBranchLength(); + } + }else if (mode == "deunique") { + //if there is a branch length then print it + if (tree[node].getBranchLength() != -1) { + out << ":" << tree[node].getBranchLength(); + } } }else { //you are a leaf - string leafGroup = globaldata->gTreemap->getGroup(tree[node].getName()); + string leafGroup = tmap->getGroup(tree[node].getName()); - out << leafGroup; if (mode == "branch") { + out << leafGroup; //if there is a branch length then print it if (tree[node].getBranchLength() != -1) { out << ":" << tree[node].getBranchLength(); } }else if (mode == "boot") { + out << leafGroup; //if there is a label then print it if (tree[node].getLabel() != -1) { out << tree[node].getLabel(); } + }else if (mode == "both") { + out << tree[node].getName(); + if (tree[node].getLabel() != -1) { + out << tree[node].getLabel(); + } + //if there is a branch length then print it + if (tree[node].getBranchLength() != -1) { + out << ":" << tree[node].getBranchLength(); + } + }else if (mode == "deunique") { + map::iterator itNames = m->names.find(tree[node].getName()); + + string outputString = ""; + if (itNames != m->names.end()) { + + vector dupNames; + m->splitAtComma((itNames->second), dupNames); + + if (dupNames.size() == 1) { + outputString += tree[node].getName(); + if (tree[node].getBranchLength() != -1) { + outputString += ":" + toString(tree[node].getBranchLength()); + } + }else { + outputString += "("; + + for (int u = 0; u < dupNames.size()-1; u++) { + outputString += dupNames[u]; + + if (tree[node].getBranchLength() != -1) { + outputString += ":" + toString(0.0); + } + outputString += ","; + } + + outputString += dupNames[dupNames.size()-1]; + if (tree[node].getBranchLength() != -1) { + outputString += ":" + toString(0.0); + } + + outputString += ")"; + if (tree[node].getBranchLength() != -1) { + outputString += ":" + toString(tree[node].getBranchLength()); + } + } + }else { + outputString = tree[node].getName(); + //if there is a branch length then print it + if (tree[node].getBranchLength() != -1) { + outputString += ":" + toString(tree[node].getBranchLength()); + } + + m->mothurOut("[ERROR]: " + tree[node].getName() + " is not in your namefile, please correct."); m->mothurOutEndLine(); + } + + out << outputString; + } + } + + } + catch(exception& e) { + m->errorOut(e, "Tree", "printBranch"); + exit(1); + } +} +/*****************************************************************/ +void Tree::printBranch(int node, ostream& out, string mode, vector& theseNodes) { + try { + + // you are not a leaf + if (theseNodes[node].getLChild() != -1) { + out << "("; + printBranch(theseNodes[node].getLChild(), out, mode); + out << ","; + printBranch(theseNodes[node].getRChild(), out, mode); + out << ")"; + if (mode == "branch") { + //if there is a branch length then print it + if (theseNodes[node].getBranchLength() != -1) { + out << ":" << theseNodes[node].getBranchLength(); + } + }else if (mode == "boot") { + //if there is a label then print it + if (theseNodes[node].getLabel() != -1) { + out << theseNodes[node].getLabel(); + } + }else if (mode == "both") { + if (theseNodes[node].getLabel() != -1) { + out << theseNodes[node].getLabel(); + } + //if there is a branch length then print it + if (theseNodes[node].getBranchLength() != -1) { + out << ":" << theseNodes[node].getBranchLength(); + } + } + }else { //you are a leaf + string leafGroup = tmap->getGroup(theseNodes[node].getName()); + + if (mode == "branch") { + out << leafGroup; + //if there is a branch length then print it + if (theseNodes[node].getBranchLength() != -1) { + out << ":" << theseNodes[node].getBranchLength(); + } + }else if (mode == "boot") { + out << leafGroup; + //if there is a label then print it + if (theseNodes[node].getLabel() != -1) { + out << theseNodes[node].getLabel(); + } + }else if (mode == "both") { + out << theseNodes[node].getName(); + if (theseNodes[node].getLabel() != -1) { + out << theseNodes[node].getLabel(); + } + //if there is a branch length then print it + if (theseNodes[node].getBranchLength() != -1) { + out << ":" << theseNodes[node].getBranchLength(); + } } } @@ -690,7 +1021,6 @@ void Tree::printBranch(int node, ostream& out, string mode) { exit(1); } } - /*****************************************************************/ void Tree::printTree() { @@ -708,7 +1038,7 @@ void Tree::parseTreeFile() { //only takes names from the first tree and assumes that all trees use the same names. try { - string filename = globaldata->getTreeFile(); + string filename = m->getTreeFile(); ifstream filehandle; m->openInputFile(filename, filehandle); int c, comment; @@ -774,9 +1104,9 @@ void Tree::parseTreeFile() { //c = , until done with translation then c = ; h = name.substr(name.length()-1, name.length()); name.erase(name.end()-1); //erase the comma - globaldata->Treenames.push_back(number); + m->Treenames.push_back(number); } - if(number == ";") { globaldata->Treenames.pop_back(); } //in case ';' from translation is on next line instead of next to last name + if(number == ";") { m->Treenames.pop_back(); } //in case ';' from translation is on next line instead of next to last name } } filehandle.close(); @@ -823,7 +1153,7 @@ int Tree::readTreeString(ifstream& filehandle) { } //cout << "name = " << name << endl; - globaldata->Treenames.push_back(name); + m->Treenames.push_back(name); filehandle.putback(c); //k = c; //cout << " after putback" << k << endl;