try {
m = MothurOut::getInstance();
calcTotals = false;
+ numNodes = 0;
+ numSeqs = 0;
#ifdef USE_MPI
MPI_File inMPI;
#ifdef USE_MPI
- int pid, num;
+ 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());
positions = setFilePosEachLine(tfile, num);
//send file positions to all processes
- MPI_Bcast(&num, 1, MPI_INT, 0, MPI_COMM_WORLD); //send numSeqs
- MPI_Bcast(&positions[0], (num+1), MPI_LONG, 0, MPI_COMM_WORLD); //send file pos
+ 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_Bcast(&num, 1, MPI_INT, 0, MPI_COMM_WORLD); //get numSeqs
- positions.resize(num);
- MPI_Bcast(&positions[0], (num+1), MPI_LONG, 0, MPI_COMM_WORLD); //get file positions
+ 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 file
}
MPI_File_close(&inMPI);
+ MPI_Barrier(MPI_COMM_WORLD); //make everyone wait - just in case
#else
ifstream in;
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; }
}
map<string, int>::iterator childPointer;
vector<TaxNode> copy = tree;
- int copyNodes = numNodes;
+
+ //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();
//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++) {
+ map<int, int>::iterator itLeaf;
+ for (itLeaf = leafNodes.begin(); itLeaf != leafNodes.end(); itLeaf++) {
if (m->control_pressed) { out.close(); break; }
- int level = copy[itBin->second].level;
- int currentNode = itBin->second;
+ int level = copy[itLeaf->second].level;
+ int currentNode = itLeaf->second;
//this sequence is unclassified at some levels
- while(level != maxLevel){
+ while(level <= maxLevel){
level++;
if(childPointer != copy[currentNode].children.end()){ //if the node already exists, move on
currentNode = childPointer->second; //currentNode becomes 'unclassified'
- copy[currentNode].accessions.push_back(itBin->first); //add this seq
}
else{ //otherwise, create it
copy.push_back(TaxNode(taxon));
copy[copyNodes-1].level = copy[currentNode].level + 1;
currentNode = copy[currentNode].children[taxon];
- copy[currentNode].accessions.push_back(itBin->first);
}
}
}
}
}
/**************************************************************************************************/
+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 = "";
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
}
}
/**************************************************************************************************/
+TaxNode PhyloTree::get(int i ){
+ try {
+ 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", "get");
+ exit(1);
+ }
+}
+/**************************************************************************************************/
+TaxNode PhyloTree::get(string seqName){
+ try {
+ 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;
+
+ 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 != name2Taxonomy.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--;
+ }
+
+ 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", "ErrorCheck");
+ exit(1);
+ }
+}
+/**************************************************************************************************/
+
projects / mothur.git / blobdiff