#include "tree.h"
-
+/*****************************************************************/
+Tree::Tree(int num) {
+ try {
+ globaldata = GlobalData::getInstance();
+ m = MothurOut::getInstance();
+
+ numLeaves = num;
+ numNodes = 2*numLeaves - 1;
+
+ tree.resize(numNodes);
+ }
+ catch(exception& e) {
+ m->errorOut(e, "Tree", "Tree - numNodes");
+ exit(1);
+ }
+}
+/*****************************************************************/
+Tree::Tree(string g) {
+ try {
+ globaldata = GlobalData::getInstance();
+ m = MothurOut::getInstance();
+
+ parseTreeFile(); globaldata->runParse = false;
+ }
+ catch(exception& e) {
+ m->errorOut(e, "Tree", "Tree - just parse");
+ exit(1);
+ }
+}
/*****************************************************************/
Tree::Tree() {
try {
globaldata = GlobalData::getInstance();
+ m = MothurOut::getInstance();
if (globaldata->runParse == true) { parseTreeFile(); globaldata->runParse = false; }
//for(int i = 0; i < globaldata->Treenames.size(); i++) { cout << i << '\t' << globaldata->Treenames[i] << endl; }
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);
+ }
//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].setGroup(globaldata->gTreemap->getGroup(globaldata->Treenames[i]));
+
+ //save group info
+ string group = globaldata->gTreemap->getGroup(globaldata->Treenames[i]);
+ vector<string> 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[globaldata->gTreemap->getGroup(globaldata->Treenames[i])] = 1;
- tree[i].pGroups[globaldata->gTreemap->getGroup(globaldata->Treenames[i])] = 1;
+ tree[i].pcount[group] = 1;
+ tree[i].pGroups[group] = 1;
+
//Treemap knows name, group and index to speed up search
globaldata->gTreemap->setIndex(globaldata->Treenames[i], i);
//intialize non leaf nodes
}else if (i > (numLeaves-1)) {
tree[i].setName("");
- tree[i].setGroup("");
+ vector<string> tempGroups;
+ tree[i].setGroup(tempGroups);
}
}
}
catch(exception& e) {
- errorOut(e, "Tree", "Tree");
+ m->errorOut(e, "Tree", "Tree");
exit(1);
}
}
//go through each leaf and update its pcounts and pgroups
+
+ //float A = clock();
+
for (int i = 0; i < numLeaves; i++) {
+
string name = tree[i].getName();
-
+
map<string, string>::iterator itNames = globaldata->names.find(name);
-
- if (itNames == globaldata->names.end()) { mothurOut(name + " is not in your name file, please correct."); mothurOutEndLine(); exit(1); }
+
+ if (itNames == globaldata->names.end()) { m->mothurOut(name + " is not in your name file, please correct."); m->mothurOutEndLine(); exit(1); }
else {
vector<string> dupNames;
- splitAtComma(globaldata->names[name], dupNames);
+ m->splitAtComma(globaldata->names[name], dupNames);
map<string, int>::iterator itCounts;
int maxPars = 1;
+ set<string> groupsAddedForThisNode;
for (int j = 0; j < dupNames.size(); j++) {
-
+
+ string group = globaldata->gTreemap->getGroup(dupNames[j]);
+
if (dupNames[j] != name) {//you already added yourself in the constructor
- string group = globaldata->gTreemap->getGroup(dupNames[j]);
+
+ if (groupsAddedForThisNode.count(group) == 0) { groupNodeInfo[group].push_back(i); groupsAddedForThisNode.insert(group); } //if you have not already added this node for this group, then add it
//update pcounts
itCounts = tree[i].pcount.find(group);
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]++;
}
if(tree[i].pGroups[group] > maxPars){
maxPars = tree[i].pGroups[group];
}
- }//end if
+ }else { groupsAddedForThisNode.insert(group); } //add it so you don't add it to groupNodeInfo again
}//end for
if (maxPars > 1) { //then we have some more dominant groups
}
}//end if
- }//end else
- }//end for
+ //update groups to reflect all the groups this node represents
+ vector<string> nodeGroups;
+ map<string, int>::iterator itGroups;
+ for (itGroups = tree[i].pcount.begin(); itGroups != tree[i].pcount.end(); itGroups++) {
+ nodeGroups.push_back(itGroups->first);
+ }
+ tree[i].setGroup(nodeGroups);
+ }//end else
+ }//end for
+
+ //float B = clock();
+ //cout << "addNamesToCounts\t" << (B - A) / CLOCKS_PER_SEC << endl;
+
}
catch(exception& e) {
- errorOut(e, "Tree", "addNamesToCounts");
+ m->errorOut(e, "Tree", "addNamesToCounts");
exit(1);
}
}
}
catch(exception& e) {
- errorOut(e, "Tree", "getIndex");
+ m->errorOut(e, "Tree", "getIndex");
exit(1);
}
}
globaldata->gTreemap->setIndex(searchName, index);
}
catch(exception& e) {
- errorOut(e, "Tree", "setIndex");
+ m->errorOut(e, "Tree", "setIndex");
exit(1);
}
}
/*****************************************************************/
-void Tree::assembleTree() {
+int Tree::assembleTree() {
try {
-
+ //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(); }
//build the pGroups in non leaf nodes to be used in the parsimony calcs.
for (int i = numLeaves; i < numNodes; i++) {
+ if (m->control_pressed) { return 1; }
+
+ tree[i].pGroups = (mergeGroups(i));
+ tree[i].pcount = (mergeGcounts(i));
+ }
+ //float B = clock();
+ //cout << "assembleTree\t" << (B-A) / CLOCKS_PER_SEC << endl;
+ return 0;
+ }
+ catch(exception& e) {
+ m->errorOut(e, "Tree", "assembleTree");
+ exit(1);
+ }
+}
+/*****************************************************************/
+int Tree::assembleTree(string n) {
+ try {
+
+ //build the pGroups in non leaf nodes to be used in the parsimony calcs.
+ for (int i = numLeaves; i < numNodes; i++) {
+ if (m->control_pressed) { return 1; }
+
tree[i].pGroups = (mergeGroups(i));
tree[i].pcount = (mergeGcounts(i));
}
+ //float B = clock();
+ //cout << "assembleTree\t" << (B-A) / CLOCKS_PER_SEC << endl;
+ return 0;
+ }
+ catch(exception& e) {
+ m->errorOut(e, "Tree", "assembleTree");
+ exit(1);
+ }
+}
+/*****************************************************************/
+void Tree::getSubTree(Tree* copy, vector<string> 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 = globaldata->gTreemap->getGroup(Groups[i]);
+ vector<string> 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
+ globaldata->gTreemap->setIndex(Groups[i], i);
+
+ //intialize non leaf nodes
+ }else if (i > (numLeaves-1)) {
+ tree[i].setName("");
+ vector<string> tempGroups;
+ tree[i].setGroup(tempGroups);
+ }
+ }
+
+ set<int> 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()));
+ 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()));
+ 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 = copy->tree[greatgrandparent].getLChild();
+ int 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()));
+ copy->tree[greatgrandparent].setChildren(greatgrandparentLC, greatgrandparentRC);
+ }else{
+ copy->tree[parent].setChildren(-1, -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) {
- errorOut(e, "Tree", "assembleTree");
+ m->errorOut(e, "Tree", "getCopy");
+ exit(1);
+ }
+}
+/*****************************************************************/
+int Tree::populateNewTree(vector<Node>& 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 = globaldata->gTreemap->getIndex(oldtree[node].getName());
+ return indexInNewTree;
+ }
+ }
+ catch(exception& e) {
+ m->errorOut(e, "Tree", "populateNewTree");
exit(1);
}
}
tree[i].pcount = copy->tree[i].pcount;
}
+ groupNodeInfo = copy->groupNodeInfo;
+
}
catch(exception& e) {
- errorOut(e, "Tree", "getCopy");
+ m->errorOut(e, "Tree", "getCopy");
exit(1);
}
}
return parsimony;
}
catch(exception& e) {
- errorOut(e, "Tree", "mergeGroups");
+ m->errorOut(e, "Tree", "mergeGroups");
exit(1);
}
}
//loop through nodes groups removing the ones the user doesn't want
for(it=tree[lc].pGroups.begin();it!=tree[lc].pGroups.end();){
- if (inUsersGroups(it->first, g) != true) {
+ if (m->inUsersGroups(it->first, g) != true) {
tree[lc].pGroups.erase(it++);
}else { it++; }
}
//loop through nodes groups removing the ones the user doesn't want
for(it=tree[rc].pGroups.begin();it!=tree[rc].pGroups.end();){
- if (inUsersGroups(it->first, g) != true) {
+ if (m->inUsersGroups(it->first, g) != true) {
tree[rc].pGroups.erase(it++);
}else { it++; }
}
return parsimony;
}
catch(exception& e) {
- errorOut(e, "Tree", "mergeUserGroups");
+ m->errorOut(e, "Tree", "mergeUserGroups");
exit(1);
}
}
return sum;
}
catch(exception& e) {
- errorOut(e, "Tree", "mergeGcounts");
+ m->errorOut(e, "Tree", "mergeGcounts");
exit(1);
}
}
void Tree::randomLabels(vector<string> 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 < numLeaves; i++){
int z;
//if either of the leaf nodes you are about to switch are not in the users groups then you don't want to switch them.
bool treez, treei;
- treez = inUsersGroups(tree[z].getGroup(), g);
- treei = inUsersGroups(tree[i].getGroup(), g);
+ treez = m->inUsersGroups(tree[z].getGroup(), g);
+ treei = m->inUsersGroups(tree[i].getGroup(), g);
if ((treez == true) && (treei == true)) {
//switches node i and node z's info.
tree[z].pGroups = (tree[i].pGroups);
tree[i].pGroups = (lib_hold);
- string zgroup = tree[z].getGroup();
+ vector<string> zgroup = tree[z].getGroup();
tree[z].setGroup(tree[i].getGroup());
tree[i].setGroup(zgroup);
tree[z].pcount = (tree[i].pcount);
tree[i].pcount = (gcount_hold);
}
+
+ for (int k = 0; k < (tree[i].getGroup()).size(); k++) { groupNodeInfo[(tree[i].getGroup())[k]].push_back(i); }
+ for (int k = 0; k < (tree[z].getGroup()).size(); k++) { groupNodeInfo[(tree[z].getGroup())[k]].push_back(z); }
}
}
catch(exception& e) {
- errorOut(e, "Tree", "randomLabels");
+ m->errorOut(e, "Tree", "randomLabels");
exit(1);
}
}
-/**************************************************************************************************/
+/**************************************************************************************************
void Tree::randomLabels(string groupA, string groupB) {
try {
}
}
catch(exception& e) {
- errorOut(e, "Tree", "randomLabels");
+ m->errorOut(e, "Tree", "randomLabels");
exit(1);
}
}
}
}
catch(exception& e) {
- errorOut(e, "Tree", "randomBlengths");
+ m->errorOut(e, "Tree", "randomBlengths");
exit(1);
}
}
/*************************************************************************************************/
void Tree::assembleRandomUnifracTree(vector<string> g) {
randomLabels(g);
- assembleTree();
+ assembleTree("noNameCounts");
}
/*************************************************************************************************/
void Tree::assembleRandomUnifracTree(string groupA, string groupB) {
- randomLabels(groupA, groupB);
- assembleTree();
+
+ vector<string> temp; temp.push_back(groupA); temp.push_back(groupB);
+ randomLabels(temp);
+ assembleTree("noNameCounts");
}
/*************************************************************************************************/
}
}
catch(exception& e) {
- errorOut(e, "Tree", "randomTopology");
+ m->errorOut(e, "Tree", "randomTopology");
exit(1);
}
}
out << ";" << endl;
}
catch(exception& e) {
- errorOut(e, "Tree", "print");
+ m->errorOut(e, "Tree", "print");
exit(1);
}
}
/*****************************************************************/
-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) {
- 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 = getRootName(globaldata->getTreeFile()) + "newick";
+ //filename = m->getRootName(globaldata->getTreeFile()) + "newick";
filename = f;
- openOutputFile(filename, out);
+ m->openOutputFile(filename, out);
printBranch(root, out, "branch");
out.close();
}
catch(exception& e) {
- errorOut(e, "Tree", "createNewickFile");
+ m->errorOut(e, "Tree", "createNewickFile");
exit(1);
}
}
return -1;
}
catch(exception& e) {
- errorOut(e, "Tree", "findRoot");
+ m->errorOut(e, "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);
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 { //you are a leaf
- out << tree[node].getGroup();
+ string leafGroup = globaldata->gTreemap->getGroup(tree[node].getName());
+
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();
+ }
}
}
}
catch(exception& e) {
- errorOut(e, "Tree", "printBranch");
+ m->errorOut(e, "Tree", "printBranch");
+ exit(1);
+ }
+}
+/*****************************************************************/
+void Tree::printBranch(int node, ostream& out, string mode, vector<Node>& 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 = globaldata->gTreemap->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();
+ }
+ }
+ }
+
+ }
+ catch(exception& e) {
+ m->errorOut(e, "Tree", "printBranch");
exit(1);
}
}
-
/*****************************************************************/
void Tree::printTree() {
try {
string filename = globaldata->getTreeFile();
ifstream filehandle;
- openInputFile(filename, filehandle);
+ m->openInputFile(filename, filehandle);
int c, comment;
comment = 0;
int done = 1;
}
}
filehandle.close();
+
+ //for (int i = 0; i < globaldata->Treenames.size(); i++) {
+//cout << globaldata->Treenames[i] << endl; }
+//cout << globaldata->Treenames.size() << endl;
}
catch(exception& e) {
- errorOut(e, "Tree", "parseTreeFile");
+ m->errorOut(e, "Tree", "parseTreeFile");
exit(1);
}
}
return 0;
}
catch(exception& e) {
- errorOut(e, "Tree", "readTreeString");
+ m->errorOut(e, "Tree", "readTreeString");
exit(1);
}
}