X-Git-Url: https://git.donarmstrong.com/?a=blobdiff_plain;f=classify.cpp;h=8aa3cdb381ed7a389667ce61962d47cefac15ddd;hb=e8e13c129ba8184ec5932a090773f353f3ae3406;hp=7770999e779917996a916e87836a03c3a2a8f26b;hpb=65b6a38d00b3a72021611211e7c25392022c69ed;p=mothur.git

diff --git a/classify.cpp b/classify.cpp
index 7770999..8aa3cdb 100644
--- a/classify.cpp
+++ b/classify.cpp
@@ -47,7 +47,7 @@ void Classify::generateDatabaseAndNames(string tfile, string tempFile, string me
 				}
 			}
 			else if(method == "suffix")		{	database = new SuffixDB(numSeqs);								}
-			else if(method == "blast")		{	database = new BlastDB(tempFile.substr(0,tempFile.find_last_of(".")+1), gapOpen, gapExtend, match, misMatch);	}
+			else if(method == "blast")		{	database = new BlastDB(tempFile.substr(0,tempFile.find_last_of(".")+1), gapOpen, gapExtend, match, misMatch, "", threadID);	}
 			else if(method == "distance")	{	database = new DistanceDB();	}
 			else {
 				m->mothurOut(method + " is not a valid search option. I will run the command using kmer, ksize=8.");
@@ -61,7 +61,8 @@ void Classify::generateDatabaseAndNames(string tfile, string tempFile, string me
 					names.push_back(temp.getName());
 					database->addSequence(temp);	
 				}
-				database->generateDB();
+				if ((method == "kmer") && (!shortcuts)) {;} //don't print
+                else {database->generateDB(); }
 			}else if ((method == "kmer") && (!needToGenerate)) {	
 				ifstream kmerFileTest(kmerDBName.c_str());
 				database->readKmerDB(kmerFileTest);	
@@ -89,7 +90,7 @@ void Classify::generateDatabaseAndNames(string tfile, string tempFile, string me
 			m->mothurOut("Generating search database...    "); cout.flush();
 	#ifdef USE_MPI	
 				int pid, processors;
-				vector<unsigned long int> positions;
+				vector<unsigned long long> positions;
 				int tag = 2001;
 			
 				MPI_Status status; 
@@ -123,7 +124,7 @@ void Classify::generateDatabaseAndNames(string tfile, string tempFile, string me
 				//create database
 				if(method == "kmer")			{	database = new KmerDB(tempFile, kmerSize);			}
 				else if(method == "suffix")		{	database = new SuffixDB(numSeqs);								}
-				else if(method == "blast")		{	database = new BlastDB(tempFile.substr(0,tempFile.find_last_of(".")+1), gapOpen, gapExtend, match, misMatch);	}
+				else if(method == "blast")		{	database = new BlastDB(tempFile.substr(0,tempFile.find_last_of(".")+1), gapOpen, gapExtend, match, misMatch, "", pid);	}
 				else if(method == "distance")	{	database = new DistanceDB();	}
 				else {
 					m->mothurOut(method + " is not a valid search option. I will run the command using kmer, ksize=8."); m->mothurOutEndLine();
@@ -176,7 +177,7 @@ void Classify::generateDatabaseAndNames(string tfile, string tempFile, string me
 				}
 			}
 			else if(method == "suffix")		{	database = new SuffixDB(numSeqs);								}
-			else if(method == "blast")		{	database = new BlastDB(tempFile.substr(0,tempFile.find_last_of(".")+1), gapOpen, gapExtend, match, misMatch);	}
+			else if(method == "blast")		{	database = new BlastDB(tempFile.substr(0,tempFile.find_last_of(".")+1), gapOpen, gapExtend, match, misMatch, "", threadID);	}
 			else if(method == "distance")	{	database = new DistanceDB();	}
 			else {
 				m->mothurOut(method + " is not a valid search option. I will run the command using kmer, ksize=8.");
@@ -200,7 +201,8 @@ void Classify::generateDatabaseAndNames(string tfile, string tempFile, string me
 				}
 				fastaFile.close();
 
-				database->generateDB();
+                if ((method == "kmer") && (!shortcuts)) {;} //don't print
+                else {database->generateDB(); } 
 				
 			}else if ((method == "kmer") && (!needToGenerate)) {	
 				ifstream kmerFileTest(kmerDBName.c_str());
@@ -238,7 +240,7 @@ void Classify::generateDatabaseAndNames(string tfile, string tempFile, string me
 	}
 }
 /**************************************************************************************************/
-Classify::Classify() {		m = MothurOut::getInstance();	database = NULL;	}
+Classify::Classify() {		m = MothurOut::getInstance();	database = NULL;	flipped=false; }
 /**************************************************************************************************/
 
 int Classify::readTaxonomy(string file) {
@@ -249,19 +251,17 @@ int Classify::readTaxonomy(string file) {
 		
 		m->mothurOutEndLine();
 		m->mothurOut("Reading in the " + file + " taxonomy...\t");	cout.flush();
-
+        if (m->debug) { m->mothurOut("[DEBUG]: Taxonomies read in...\n"); }
+        
 #ifdef USE_MPI	
 		int pid, num, processors;
-		vector<unsigned long int> positions;
+		vector<unsigned long long> positions;
 		int tag = 2001;
 		
 		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 = new char[file.length()];
-		//memcpy(inFileName, file.c_str(), file.length());
 		
 		char inFileName[1024];
 		strcpy(inFileName, file.c_str());
@@ -296,30 +296,21 @@ int Classify::readTaxonomy(string file) {
 			delete buf4;
 
 			istringstream iss (tempBuf,istringstream::in);
-			iss >> name >> taxInfo;
+			iss >> name; m->gobble(iss);
+            iss >> taxInfo;
+            if (m->debug) { m->mothurOut("[DEBUG]: name = " + name + " tax = " + taxInfo + "\n"); }
 			taxonomy[name] = taxInfo;
 			phyloTree->addSeqToTree(name, taxInfo);
 		}
 		
 		MPI_File_close(&inMPI);
 		MPI_Barrier(MPI_COMM_WORLD); //make everyone wait - just in case
-#else				
-		ifstream inTax;
-		m->openInputFile(file, inTax);
-	
-		//read template seqs and save
-		while (!inTax.eof()) {
-			inTax >> name >> taxInfo;
-			
-			taxonomy[name] = taxInfo;
-			
-			phyloTree->addSeqToTree(name, taxInfo);
-		
-			m->gobble(inTax);
-		}
-		inTax.close();
+#else	
+        
+        taxonomy.clear(); 
+        m->readTax(file, taxonomy);
+        for (map<string, string>::iterator itTax = taxonomy.begin(); itTax != taxonomy.end(); itTax++) {  phyloTree->addSeqToTree(itTax->first, itTax->second);  }
 #endif	
-	
 		phyloTree->assignHeirarchyIDs(0);
 		
 		phyloTree->setUp(file);
@@ -363,3 +354,37 @@ vector<string> Classify::parseTax(string tax) {
 }
 /**************************************************************************************************/
 
+double Classify::getLogExpSum(vector<double> probabilities, int& maxIndex){
+	try {
+        //	http://jblevins.org/notes/log-sum-exp
+        
+        double maxProb = probabilities[0];
+        maxIndex = 0;
+        
+        int numProbs = (int)probabilities.size();
+        
+        for(int i=1;i<numProbs;i++){
+            if(probabilities[i] >= maxProb){
+                maxProb = probabilities[i];
+                maxIndex = i;
+            }
+        }
+        
+        double probSum = 0.0000;
+        
+        for(int i=0;i<numProbs;i++){
+            probSum += exp(probabilities[i] - maxProb);		
+        }
+        
+        probSum = log(probSum) + maxProb;
+        
+        return probSum;
+	}
+	catch(exception& e) {
+		m->errorOut(e, "Classify", "getLogExpSum");
+		exit(1);
+	}
+}
+
+/**************************************************************************************************/
+