//**********************************************************************************************************************
//This function checks to make sure the cluster command has no errors and then clusters based on the method chosen.
-ClusterCommand::ClusterCommand(string option){
+ClusterCommand::ClusterCommand(string option) {
try{
globaldata = GlobalData::getInstance();
+
abort = false;
//allow user to run help
else {
//valid paramters for this command
- string Array[] = {"cutoff","precision","method","showabund","timing","outputdir","inputdir"};
+ string Array[] = {"cutoff","precision","method","showabund","timing","hard","outputdir","inputdir"};
vector<string> myArray (Array, Array+(sizeof(Array)/sizeof(string)));
OptionParser parser(option);
//error checking to make sure they read a distance file
if ((globaldata->gSparseMatrix == NULL) || (globaldata->gListVector == NULL)) {
- mothurOut("Before you use the cluster command, you first need to read in a distance matrix."); mothurOutEndLine();
+ m->mothurOut("Before you use the cluster command, you first need to read in a distance matrix."); m->mothurOutEndLine();
abort = true;
}
length = temp.length();
convert(temp, precision);
+ temp = validParameter.validFile(parameters, "hard", false); if (temp == "not found") { temp = "F"; }
+ hard = isTrue(temp);
+
temp = validParameter.validFile(parameters, "cutoff", false);
if (temp == "not found") { temp = "10"; }
convert(temp, cutoff);
- cutoff += (5 / (precision * 10.0));
+ if (!hard) { cutoff += (5 / (precision * 10.0)); }
method = validParameter.validFile(parameters, "method", false);
if (method == "not found") { method = "furthest"; }
if ((method == "furthest") || (method == "nearest") || (method == "average")) { }
- else { mothurOut("Not a valid clustering method. Valid clustering algorithms are furthest, nearest or average."); mothurOutEndLine(); abort = true; }
+ else { m->mothurOut("Not a valid clustering method. Valid clustering algorithms are furthest, nearest or average."); m->mothurOutEndLine(); abort = true; }
showabund = validParameter.validFile(parameters, "showabund", false);
if (showabund == "not found") { showabund = "T"; }
}
//create cluster
- if (method == "furthest") { cluster = new CompleteLinkage(rabund, list, matrix, cutoff); }
- else if(method == "nearest"){ cluster = new SingleLinkage(rabund, list, matrix, cutoff); }
- else if(method == "average"){ cluster = new AverageLinkage(rabund, list, matrix, cutoff); }
+ if (method == "furthest") { cluster = new CompleteLinkage(rabund, list, matrix, cutoff, method); }
+ else if(method == "nearest"){ cluster = new SingleLinkage(rabund, list, matrix, cutoff, method); }
+ else if(method == "average"){ cluster = new AverageLinkage(rabund, list, matrix, cutoff, method); }
tag = cluster->getTag();
if (outputDir == "") { outputDir += hasPath(globaldata->inputFileName); }
openOutputFile(fileroot+ tag + ".sabund", sabundFile);
openOutputFile(fileroot+ tag + ".rabund", rabundFile);
openOutputFile(fileroot+ tag + ".list", listFile);
+
+ outputNames.push_back(fileroot+ tag + ".sabund");
+ outputNames.push_back(fileroot+ tag + ".rabund");
+ outputNames.push_back(fileroot+ tag + ".list");
}
}
}
catch(exception& e) {
- errorOut(e, "ClusterCommand", "ClusterCommand");
+ m->errorOut(e, "ClusterCommand", "ClusterCommand");
exit(1);
}
}
void ClusterCommand::help(){
try {
- mothurOut("The cluster command can only be executed after a successful read.dist command.\n");
- mothurOut("The cluster command parameter options are method, cuttoff, precision, showabund and timing. No parameters are required.\n");
- mothurOut("The cluster command should be in the following format: \n");
- mothurOut("cluster(method=yourMethod, cutoff=yourCutoff, precision=yourPrecision) \n");
- mothurOut("The acceptable cluster methods are furthest, nearest and average. If no method is provided then furthest is assumed.\n\n");
+ m->mothurOut("The cluster command can only be executed after a successful read.dist command.\n");
+ m->mothurOut("The cluster command parameter options are method, cuttoff, hard, precision, showabund and timing. No parameters are required.\n");
+ m->mothurOut("The cluster command should be in the following format: \n");
+ m->mothurOut("cluster(method=yourMethod, cutoff=yourCutoff, precision=yourPrecision) \n");
+ m->mothurOut("The acceptable cluster methods are furthest, nearest and average. If no method is provided then furthest is assumed.\n\n");
}
catch(exception& e) {
- errorOut(e, "ClusterCommand", "help");
+ m->errorOut(e, "ClusterCommand", "help");
exit(1);
}
}
print_start = true;
start = time(NULL);
loops = 0;
+ double saveCutoff = cutoff;
while (matrix->getSmallDist() < cutoff && matrix->getNNodes() > 0){
+
+ if (m->control_pressed) { //clean up
+ delete globaldata->gSparseMatrix; globaldata->gSparseMatrix = NULL;
+ delete globaldata->gListVector; globaldata->gListVector = NULL;
+ if (globaldata->getFormat() == "phylip") { globaldata->setPhylipFile(""); }
+ else if (globaldata->getFormat() == "column") { globaldata->setColumnFile(""); }
+ sabundFile.close();rabundFile.close();listFile.close();
+ for (int i = 0; i < outputNames.size(); i++) { remove(outputNames[i].c_str()); }
+ return 0;
+ }
+
if (print_start && isTrue(timing)) {
- mothurOut("Clustering (" + tag + ") dist " + toString(matrix->getSmallDist()) + "/"
+ m->mothurOut("Clustering (" + tag + ") dist " + toString(matrix->getSmallDist()) + "/"
+ toString(roundDist(matrix->getSmallDist(), precision))
+ "\t(precision: " + toString(precision) + ", Nodes: " + toString(matrix->getNNodes()) + ")");
cout.flush();
loops++;
cluster->update(cutoff);
+
float dist = matrix->getSmallDist();
float rndDist = roundDist(dist, precision);
}
if (print_start && isTrue(timing)) {
- mothurOut("Clustering (" + tag + ") for distance " + toString(previousDist) + "/" + toString(rndPreviousDist)
+ m->mothurOut("Clustering (" + tag + ") for distance " + toString(previousDist) + "/" + toString(rndPreviousDist)
+ "\t(precision: " + toString(precision) + ", Nodes: " + toString(matrix->getNNodes()) + ")");
cout.flush();
print_start = false;
//delete globaldata's copy of the sparsematrix and listvector to free up memory
delete globaldata->gSparseMatrix; globaldata->gSparseMatrix = NULL;
delete globaldata->gListVector; globaldata->gListVector = NULL;
-
+
//saves .list file so you can do the collect, rarefaction and summary commands without doing a read.list
if (globaldata->getFormat() == "phylip") { globaldata->setPhylipFile(""); }
else if (globaldata->getFormat() == "column") { globaldata->setColumnFile(""); }
sabundFile.close();
rabundFile.close();
listFile.close();
+
+ if (saveCutoff != cutoff) { m->mothurOut("changed cutoff to " + toString(cutoff)); m->mothurOutEndLine(); }
+
+ m->mothurOutEndLine();
+ m->mothurOut("Output File Names: "); m->mothurOutEndLine();
+ for (int i = 0; i < outputNames.size(); i++) { m->mothurOut(outputNames[i]); m->mothurOutEndLine(); }
+ m->mothurOutEndLine();
+
+
//if (isTrue(timing)) {
- mothurOut("It took " + toString(time(NULL) - estart) + " seconds to cluster"); mothurOutEndLine();
+ m->mothurOut("It took " + toString(time(NULL) - estart) + " seconds to cluster"); m->mothurOutEndLine();
//}
+
+
return 0;
}
catch(exception& e) {
- errorOut(e, "ClusterCommand", "execute");
+ m->errorOut(e, "ClusterCommand", "execute");
exit(1);
}
}
void ClusterCommand::printData(string label){
try {
if (isTrue(timing)) {
- mothurOut("\tTime: " + toString(time(NULL) - start) + "\tsecs for " + toString(oldRAbund.getNumBins())
- + "\tclusters. Updates: " + toString(loops)); mothurOutEndLine();
+ m->mothurOut("\tTime: " + toString(time(NULL) - start) + "\tsecs for " + toString(oldRAbund.getNumBins())
+ + "\tclusters. Updates: " + toString(loops)); m->mothurOutEndLine();
}
print_start = true;
loops = 0;
oldList.print(listFile);
}
catch(exception& e) {
- errorOut(e, "ClusterCommand", "printData");
+ m->errorOut(e, "ClusterCommand", "printData");
exit(1);
}