}
/**************************************************************************************************/
-PhyloTree::PhyloTree(ifstream& in){
+PhyloTree::PhyloTree(ifstream& in, string filename){
try {
m = MothurOut::getInstance();
calcTotals = false;
+ numNodes = 0;
+ numSeqs = 0;
- in >> numNodes; gobble(in);
-
- tree.resize(numNodes);
-
- for (int i = 0; i < tree.size(); i++) {
- in >> tree[i].name >> tree[i].level >> tree[i].parent; gobble(in);
-
- }
-
- //read genus nodes
- int numGenus = 0;
- in >> numGenus; gobble(in);
-
- int gnode, gsize;
- totals.clear();
- for (int i = 0; i < numGenus; i++) {
- in >> gnode >> gsize; gobble(in);
+ #ifdef USE_MPI
+ MPI_File inMPI;
+ MPI_Offset size;
+ MPI_Status status;
+
+ char inFileName[1024];
+ strcpy(inFileName, filename.c_str());
+
+ MPI_File_open(MPI_COMM_WORLD, inFileName, MPI_MODE_RDONLY, MPI_INFO_NULL, &inMPI);
+ MPI_File_get_size(inMPI, &size);
- uniqueTaxonomies[gnode] = gnode;
- totals.push_back(gsize);
- }
+ char* buffer = new char[size];
+ MPI_File_read(inMPI, buffer, size, MPI_CHAR, &status);
+
+ string tempBuf = buffer;
+ if (tempBuf.length() > size) { tempBuf = tempBuf.substr(0, size); }
+ istringstream iss (tempBuf,istringstream::in);
+ delete buffer;
+
+ iss >> numNodes; gobble(iss);
+
+ tree.resize(numNodes);
+
+ for (int i = 0; i < tree.size(); i++) {
+ iss >> tree[i].name >> tree[i].level >> tree[i].parent; gobble(iss);
+ }
+
+ //read genus nodes
+ int numGenus = 0;
+ iss >> numGenus; gobble(iss);
+
+ int gnode, gsize;
+ totals.clear();
+ for (int i = 0; i < numGenus; i++) {
+ iss >> gnode >> gsize; gobble(iss);
+
+ uniqueTaxonomies[gnode] = gnode;
+ totals.push_back(gsize);
+ }
+
+ MPI_File_close(&inMPI);
+
+ #else
+ in >> numNodes; gobble(in);
- in.close();
+ tree.resize(numNodes);
+
+ for (int i = 0; i < tree.size(); i++) {
+ in >> tree[i].name >> tree[i].level >> tree[i].parent; gobble(in);
+ }
+
+ //read genus nodes
+ int numGenus = 0;
+ in >> numGenus; gobble(in);
+
+ int gnode, gsize;
+ totals.clear();
+ for (int i = 0; i < numGenus; i++) {
+ in >> gnode >> gsize; gobble(in);
+
+ uniqueTaxonomies[gnode] = gnode;
+ totals.push_back(gsize);
+ }
+
+ in.close();
+
+ #endif
}
catch(exception& e) {
tree[0].heirarchyID = "0";
maxLevel = 0;
calcTotals = true;
+ string name, tax;
+
- ifstream in;
- openInputFile(tfile, in);
+ #ifdef USE_MPI
+ int pid, num, processors;
+ vector<long> positions;
+
+ MPI_Status status;
+ MPI_File inMPI;
+ MPI_Comm_rank(MPI_COMM_WORLD, &pid); //find out who we are
+ MPI_Comm_size(MPI_COMM_WORLD, &processors);
+
+ char inFileName[1024];
+ strcpy(inFileName, tfile.c_str());
+
+ MPI_File_open(MPI_COMM_WORLD, inFileName, MPI_MODE_RDONLY, MPI_INFO_NULL, &inMPI); //comm, filename, mode, info, filepointer
+
+ if (pid == 0) {
+ positions = setFilePosEachLine(tfile, num);
+
+ //send file positions to all processes
+ for(int i = 1; i < processors; i++) {
+ MPI_Send(&num, 1, MPI_INT, i, 2001, MPI_COMM_WORLD);
+ MPI_Send(&positions[0], (num+1), MPI_LONG, i, 2001, MPI_COMM_WORLD);
+ }
+ }else{
+ MPI_Recv(&num, 1, MPI_INT, 0, 2001, MPI_COMM_WORLD, &status);
+ positions.resize(num+1);
+ MPI_Recv(&positions[0], (num+1), MPI_LONG, 0, 2001, MPI_COMM_WORLD, &status);
+ }
- //read in users taxonomy file and add sequences to tree
- string name, tax;
- while(!in.eof()){
- in >> name >> tax; gobble(in);
+ //read file
+ for(int i=0;i<num;i++){
+ //read next sequence
+ int length = positions[i+1] - positions[i];
+ char* buf4 = new char[length];
+
+ MPI_File_read_at(inMPI, positions[i], buf4, length, MPI_CHAR, &status);
+
+ string tempBuf = buf4;
+ if (tempBuf.length() > length) { tempBuf = tempBuf.substr(0, length); }
+ delete buf4;
+
+ istringstream iss (tempBuf,istringstream::in);
+ iss >> name >> tax;
+ addSeqToTree(name, tax);
+ }
- addSeqToTree(name, tax);
- }
- in.close();
-
+ MPI_File_close(&inMPI);
+ MPI_Barrier(MPI_COMM_WORLD); //make everyone wait - just in case
+
+ #else
+ ifstream in;
+ openInputFile(tfile, in);
+
+ //read in users taxonomy file and add sequences to tree
+ while(!in.eof()){
+ in >> name >> tax; gobble(in);
+
+ addSeqToTree(name, tax);
+ }
+ in.close();
+ #endif
+
assignHeirarchyIDs(0);
+
+ //create file for summary if needed
+ setUp(tfile);
}
catch(exception& e) {
m->errorOut(e, "PhyloTree", "PhyloTree");
int PhyloTree::addSeqToTree(string seqName, string seqTaxonomy){
try {
+
numSeqs++;
map<string, int>::iterator childPointer;
//use print to reassign the taxa id
taxon = getNextTaxon(seqTaxonomy);
+ if (taxon == "") { m->mothurOut(seqName + " has an error in the taxonomy. This may be due to a ;;"); m->mothurOutEndLine(); if (currentNode != 0) { uniqueTaxonomies[currentNode] = currentNode; } break; }
+
childPointer = tree[currentNode].children.find(taxon);
if(childPointer != tree[currentNode].children.end()){ //if the node already exists, move on
// tree[currentNode].childNumber = numChildren;
// tree[currentNode].heirarchyID = heirarchyID + '.' + toString(tree[currentNode].childNumber);
}
-
+
if (seqTaxonomy == "") { uniqueTaxonomies[currentNode] = currentNode; }
}
}
}
/**************************************************************************************************/
-void PhyloTree::binUnclassified(){
+void PhyloTree::setUp(string tfile){
+ try{
+ string taxFileNameTest = tfile.substr(0,tfile.find_last_of(".")+1) + "tree.sum";
+
+ #ifdef USE_MPI
+ int pid;
+ MPI_Comm_rank(MPI_COMM_WORLD, &pid); //find out who we are
+
+ if (pid == 0) { binUnclassified(taxFileNameTest); }
+
+ #else
+ //create file needed for summary if it doesn't exist
+ ifstream FileTest(taxFileNameTest.c_str());
+
+ if (!FileTest) {
+ binUnclassified(taxFileNameTest);
+ }
+ #endif
+ }
+ catch(exception& e) {
+ m->errorOut(e, "PhyloTree", "setUp");
+ exit(1);
+ }
+}
+/**************************************************************************************************/
+void PhyloTree::binUnclassified(string file){
try {
+
+ ofstream out;
+ openOutputFile(file, out);
+
map<string, int>::iterator itBin;
map<string, int>::iterator childPointer;
- //go through the seqs and if a sequence finest taxon is not the same level as the most finely defined taxon then classify it as unclassified where necessary
- for (itBin = name2Taxonomy.begin(); itBin != name2Taxonomy.end(); itBin++) {
+ vector<TaxNode> copy = tree;
+
+ //fill out tree
+ fillOutTree(0, copy);
+
+ //get leaf nodes that may need externsion
+ for (int i = 0; i < copy.size(); i++) {
+
+ if (copy[i].children.size() == 0) {
+ leafNodes[i] = i;
+ }
+ }
+
+ int copyNodes = copy.size();
- int level = tree[itBin->second].level;
- int currentNode = itBin->second;
+ //go through the seqs and if a sequence finest taxon is not the same level as the most finely defined taxon then classify it as unclassified where necessary
+ map<int, int>::iterator itLeaf;
+ for (itLeaf = leafNodes.begin(); itLeaf != leafNodes.end(); itLeaf++) {
- //this sequence is unclassified at some levels
- while(level != maxLevel){
+ if (m->control_pressed) { out.close(); break; }
+ int level = copy[itLeaf->second].level;
+ int currentNode = itLeaf->second;
+
+ //this sequence is unclassified at some levels
+ while(level <= maxLevel){
+
level++;
string taxon = "unclassified";
//does the parent have a child names 'unclassified'?
- childPointer = tree[currentNode].children.find(taxon);
+ childPointer = copy[currentNode].children.find(taxon);
- if(childPointer != tree[currentNode].children.end()){ //if the node already exists, move on
+ if(childPointer != copy[currentNode].children.end()){ //if the node already exists, move on
currentNode = childPointer->second; //currentNode becomes 'unclassified'
- tree[currentNode].accessions.push_back(itBin->first); //add this seq
- name2Taxonomy[itBin->first] = currentNode;
}
else{ //otherwise, create it
- tree.push_back(TaxNode(taxon));
- numNodes++;
- tree[currentNode].children[taxon] = numNodes-1;
- tree[numNodes-1].parent = currentNode;
+ copy.push_back(TaxNode(taxon));
+ copyNodes++;
+ copy[currentNode].children[taxon] = copyNodes-1;
+ copy[copyNodes-1].parent = currentNode;
+ copy[copyNodes-1].level = copy[currentNode].level + 1;
- currentNode = tree[currentNode].children[taxon];
- tree[currentNode].accessions.push_back(itBin->first);
- name2Taxonomy[itBin->first] = currentNode;
+ currentNode = copy[currentNode].children[taxon];
}
-
- if (level == maxLevel) { uniqueTaxonomies[currentNode] = currentNode; }
}
}
- //clear HeirarchyIDs and reset them
- for (int i = 1; i < tree.size(); i++) {
- tree[i].heirarchyID = "";
+ if (!m->control_pressed) {
+ //print copy tree
+ print(out, copy);
}
- assignHeirarchyIDs(0);
-
+
}
catch(exception& e) {
m->errorOut(e, "PhyloTree", "binUnclassified");
}
}
/**************************************************************************************************/
+void PhyloTree::fillOutTree(int index, vector<TaxNode>& copy) {
+ try {
+ map<string,int>::iterator it;
+
+ it = copy[index].children.find("unclassified");
+ if (it == copy[index].children.end()) { //no unclassified at this level
+ string taxon = "unclassified";
+ copy.push_back(TaxNode(taxon));
+ copy[index].children[taxon] = copy.size()-1;
+ copy[copy.size()-1].parent = index;
+ copy[copy.size()-1].level = copy[index].level + 1;
+ }
+
+ if (tree[index].level <= maxLevel) {
+ for(it=tree[index].children.begin();it!=tree[index].children.end();it++){ //check your children
+ fillOutTree(it->second, copy);
+ }
+ }
+
+ }
+ catch(exception& e) {
+ m->errorOut(e, "PhyloTree", "fillOutTree");
+ exit(1);
+ }
+}
+/**************************************************************************************************/
string PhyloTree::getFullTaxonomy(string seqName) {
try {
string tax = "";
}
/**************************************************************************************************/
-void PhyloTree::print(ofstream& out){
+void PhyloTree::print(ofstream& out, vector<TaxNode>& copy){
try {
- out << tree[0].level << '\t'<< tree[0].heirarchyID << '\t' << tree[0].name << '\t' << tree[0].children.size() << '\t' << tree[0].accessions.size() << endl;
- print(0, out);
+ out << copy.size() << endl;
+
+ for (int i = 0; i < copy.size(); i++) {
+
+ out << copy[i].level << '\t'<< copy[i].name << '\t' << copy[i].children.size() << '\t';
+
+ map<string,int>::iterator it;
+ for(it=copy[i].children.begin();it!=copy[i].children.end();it++){
+ out << it->first << '\t' << it->second << '\t';
+ }
+ out << endl;
+ }
+
+ out.close();
}
catch(exception& e) {
m->errorOut(e, "PhyloTree", "print");
exit(1);
}
}
-
/**************************************************************************************************/
+void PhyloTree::printTreeNodes(string treefilename) {
+ try {
+
+ #ifdef USE_MPI
+ int pid;
+ MPI_Comm_rank(MPI_COMM_WORLD, &pid); //find out who we are
-void PhyloTree::print(int i, ofstream& out){
+ if (pid == 0) {
+
+ #endif
+
+ ofstream outTree;
+ openOutputFile(treefilename, outTree);
+
+ //print treenodes
+ outTree << tree.size() << endl;
+ for (int i = 0; i < tree.size(); i++) {
+ outTree << tree[i].name << '\t' << tree[i].level << '\t' << tree[i].parent << endl;
+ }
+
+ //print genus nodes
+ outTree << endl << uniqueTaxonomies.size() << endl;
+ map<int, int>::iterator it2;
+ for (it2=uniqueTaxonomies.begin(); it2!=uniqueTaxonomies.end(); it2++) { outTree << it2->first << '\t' << tree[it2->first].accessions.size() << endl; }
+ outTree << endl;
+
+ outTree.close();
+
+ #ifdef USE_MPI
+ }
+ #endif
+
+
+ }
+ catch(exception& e) {
+ m->errorOut(e, "PhyloTree", "printTreeNodes");
+ exit(1);
+ }
+}
+/**************************************************************************************************/
+TaxNode PhyloTree::get(int i ){
try {
- map<string,int>::iterator it;
- for(it=tree[i].children.begin();it!=tree[i].children.end();it++){
- out <<tree[it->second].level << '\t' << tree[it->second].heirarchyID << '\t' << tree[it->second].name << '\t' << tree[it->second].children.size() << '\t' << tree[it->second].accessions.size() << endl;
- print(it->second, out);
- }
+ if (i < tree.size()) { return tree[i]; }
+ else { cout << i << '\t' << tree.size() << endl ; m->mothurOut("Mismatch with taxonomy and template files. Cannot continue."); m->mothurOutEndLine(); exit(1); }
}
catch(exception& e) {
- m->errorOut(e, "PhyloTree", "print");
+ m->errorOut(e, "PhyloTree", "get");
exit(1);
}
}
/**************************************************************************************************/
-void PhyloTree::printTreeNodes(string treefilename) {
+TaxNode PhyloTree::get(string seqName){
try {
- ofstream outTree;
- openOutputFile(treefilename, outTree);
-
- //print treenodes
- outTree << tree.size() << endl;
- for (int i = 0; i < tree.size(); i++) {
- outTree << tree[i].name << '\t' << tree[i].level << '\t' << tree[i].parent << endl;
- }
+ map<string, int>::iterator itFind = name2Taxonomy.find(seqName);
+
+ if (itFind != name2Taxonomy.end()) { return tree[name2Taxonomy[seqName]]; }
+ else { m->mothurOut("Cannot find " + seqName + ". Mismatch with taxonomy and template files. Cannot continue."); m->mothurOutEndLine(); exit(1);}
+ }
+ catch(exception& e) {
+ m->errorOut(e, "PhyloTree", "get");
+ exit(1);
+ }
+}
+/**************************************************************************************************/
+string PhyloTree::getName(int i ){
+ try {
+ if (i < tree.size()) { return tree[i].name; }
+ else { m->mothurOut("Mismatch with taxonomy and template files. Cannot continue."); m->mothurOutEndLine(); exit(1); }
+ }
+ catch(exception& e) {
+ m->errorOut(e, "PhyloTree", "get");
+ exit(1);
+ }
+}
+/**************************************************************************************************/
+int PhyloTree::getIndex(string seqName){
+ try {
+ map<string, int>::iterator itFind = name2Taxonomy.find(seqName);
+
+ if (itFind != name2Taxonomy.end()) { return name2Taxonomy[seqName]; }
+ else { m->mothurOut("Cannot find " + seqName + ". Mismatch with taxonomy and template files. Cannot continue."); m->mothurOutEndLine(); exit(1);}
+ }
+ catch(exception& e) {
+ m->errorOut(e, "PhyloTree", "get");
+ exit(1);
+ }
+}
+/**************************************************************************************************/
+bool PhyloTree::ErrorCheck(vector<string> templateFileNames){
+ try {
+
+ bool okay = true;
- //print genus nodes
- outTree << endl << uniqueTaxonomies.size() << endl;
- map<int, int>::iterator it2;
- for (it2=uniqueTaxonomies.begin(); it2!=uniqueTaxonomies.end(); it2++) { outTree << it2->first << '\t' << tree[it2->first].accessions.size() << endl; }
- outTree << endl;
+ map<string, int>::iterator itFind;
+ map<string, int> taxonomyFileNames = name2Taxonomy;
+ for (int i = 0; i < templateFileNames.size(); i++) {
+ itFind = taxonomyFileNames.find(templateFileNames[i]);
+
+ if (itFind != taxonomyFileNames.end()) { //found it so erase it
+ taxonomyFileNames.erase(itFind);
+ }else {
+ m->mothurOut(templateFileNames[i] + " is in your template file and is not in your taxonomy file. Please correct."); m->mothurOutEndLine();
+ okay = false;
+ }
+
+ templateFileNames.erase(templateFileNames.begin()+i);
+ i--;
+ }
- outTree.close();
+ if (taxonomyFileNames.size() > 0) { //there are names in tax file that are not in template
+ okay = false;
+
+ for (itFind = taxonomyFileNames.begin(); itFind != taxonomyFileNames.end(); itFind++) {
+ m->mothurOut(itFind->first + " is in your taxonomy file and is not in your template file. Please correct."); m->mothurOutEndLine();
+ }
+ }
+ return okay;
}
catch(exception& e) {
- m->errorOut(e, "PhyloTree", "printTreeNodes");
+ m->errorOut(e, "PhyloTree", "ErrorCheck");
exit(1);
}
}
/**************************************************************************************************/
+