added cutoff change to mgcluster

[mothur.git] / mgclustercommand.cpp

diff --git a/mgclustercommand.cpp b/mgclustercommand.cpp
index 10932bc065ca69f6546503e3c0d402571d27cdc2..b982b8d1deb3cbc51716cf74339b8dcaea5d136d 100644 (file)
--- a/mgclustercommand.cpp
+++ b/mgclustercommand.cpp
@@ -18,8 +18,8 @@ vector<string> MGClusterCommand::setParameters(){
                CommandParameter ppenalty("penalty", "Number", "", "0.10", "", "", "",false,false); parameters.push_back(ppenalty);
                CommandParameter pcutoff("cutoff", "Number", "", "0.70", "", "", "",false,false); parameters.push_back(pcutoff);
                CommandParameter pprecision("precision", "Number", "", "100", "", "", "",false,false); parameters.push_back(pprecision);
-               CommandParameter pmethod("method", "Multiple", "furthest-nearest-average", "furthest", "", "", "",false,false); parameters.push_back(pmethod);
-               CommandParameter phard("hard", "Boolean", "", "F", "", "", "",false,false); parameters.push_back(phard);
+               CommandParameter pmethod("method", "Multiple", "furthest-nearest-average", "average", "", "", "",false,false); parameters.push_back(pmethod);
+               CommandParameter phard("hard", "Boolean", "", "T", "", "", "",false,false); parameters.push_back(phard);
                CommandParameter pmin("min", "Boolean", "", "T", "", "", "",false,false); parameters.push_back(pmin);
                CommandParameter pmerge("merge", "Boolean", "", "T", "", "", "",false,false); parameters.push_back(pmerge);
                CommandParameter phcluster("hcluster", "Boolean", "", "F", "", "", "",false,false); parameters.push_back(phcluster);
@@ -44,7 +44,7 @@ string MGClusterCommand::getHelpString(){
                helpString += "This command outputs a .list, .rabund and .sabund file that can be used with mothur other commands to estimate richness.\n";
                helpString += "The cutoff parameter is used to specify the maximum distance you would like to cluster to. The default is 0.70.\n";
                helpString += "The precision parameter's default value is 100. \n";
-               helpString += "The acceptable mgcluster methods are furthest, nearest and average.  If no method is provided then furthest is assumed.\n";      
+               helpString += "The acceptable mgcluster methods are furthest, nearest and average.  If no method is provided then average is assumed.\n";       
                helpString += "The min parameter allows you to specify is you want the minimum or maximum blast score ratio used in calculating the distance. The default is true, meaning you want the minimum.\n";
                helpString += "The length parameter is used to specify the minimum overlap required.  The default is 5.\n";
                helpString += "The penalty parameter is used to adjust the error rate.  The default is 0.10.\n";
@@ -82,6 +82,7 @@ MGClusterCommand::MGClusterCommand(string option) {
                
                //allow user to run help
                if(option == "help") { help(); abort = true; calledHelp = true; }
+               else if(option == "citation") { citation(); abort = true; calledHelp = true;}
                
                else {
                        vector<string> myArray = setParameters();
@@ -140,6 +141,7 @@ MGClusterCommand::MGClusterCommand(string option) {
                        namefile = validParameter.validFile(parameters, "name", true);
                        if (namefile == "not open") { abort = true; }   
                        else if (namefile == "not found") { namefile = ""; }
+                       else { m->setNameFile(namefile); }
                        
                        if ((blastfile == "")) { m->mothurOut("When executing a mgcluster command you must provide a blastfile."); m->mothurOutEndLine(); abort = true; }
                        
@@ -154,7 +156,7 @@ MGClusterCommand::MGClusterCommand(string option) {
                        cutoff += (5 / (precision * 10.0));
                        
                        method = validParameter.validFile(parameters, "method", false);
-                       if (method == "not found") { method = "furthest"; }
+                       if (method == "not found") { method = "average"; }
                        
                        if ((method == "furthest") || (method == "nearest") || (method == "average")) { }
                        else { m->mothurOut("Not a valid clustering method.  Valid clustering algorithms are furthest, nearest or average."); m->mothurOutEndLine(); abort = true; }
@@ -174,7 +176,7 @@ MGClusterCommand::MGClusterCommand(string option) {
                        temp = validParameter.validFile(parameters, "hcluster", false);                 if (temp == "not found") { temp = "false"; }
                        hclusterWanted = m->isTrue(temp); 
                        
-                       temp = validParameter.validFile(parameters, "hard", false);                     if (temp == "not found") { temp = "F"; }
+                       temp = validParameter.validFile(parameters, "hard", false);                     if (temp == "not found") { temp = "T"; }
                        hard = m->isTrue(temp);
                }
 
@@ -233,6 +235,8 @@ int MGClusterCommand::execute(){
                        return 0; 
                }
                
+               double saveCutoff = cutoff;
+               
                if (!hclusterWanted) {
                        //get distmatrix and overlap
                        SparseMatrix* distMatrix = read->getDistMatrix();
@@ -371,6 +375,11 @@ int MGClusterCommand::execute(){
                
                                seqs = hcluster->getSeqs();
                                
+                               //to account for cutoff change in average neighbor
+                               if (seqs.size() != 0) {
+                                       if (seqs[0].dist > cutoff) { break; }
+                               }
+                               
                                if (m->control_pressed) { 
                                        delete nameMap;  delete list; delete rabund; delete hcluster;
                                        listFile.close(); rabundFile.close(); sabundFile.close(); remove((fileroot+ tag + ".list").c_str()); remove((fileroot+ tag + ".rabund").c_str()); remove((fileroot+ tag + ".sabund").c_str());
@@ -384,7 +393,7 @@ int MGClusterCommand::execute(){
                                        
                                        if (seqs[i].seq1 != seqs[i].seq2) {
                
-                                               hcluster->update(seqs[i].seq1, seqs[i].seq2, seqs[i].dist);
+                                               cutoff = hcluster->update(seqs[i].seq1, seqs[i].seq2, seqs[i].dist);
                                                
                                                if (m->control_pressed) { 
                                                        delete nameMap;  delete list; delete rabund; delete hcluster;
@@ -489,6 +498,13 @@ int MGClusterCommand::execute(){
                m->mothurOut(fileroot+ tag + ".sabund"); m->mothurOutEndLine(); outputNames.push_back(fileroot+ tag + ".sabund"); outputTypes["sabund"].push_back(fileroot+ tag + ".sabund");
                m->mothurOutEndLine();
                
+               if (saveCutoff != cutoff) { 
+                       if (hard)       {  saveCutoff = m->ceilDist(saveCutoff, precision);     }
+                       else            {       saveCutoff = m->roundDist(saveCutoff, precision);  }
+                       
+                       m->mothurOut("changed cutoff to " + toString(cutoff)); m->mothurOutEndLine(); 
+               }
+               
                //set list file as new current listfile
                string current = "";
                itTypes = outputTypes.find("list");