int main()
{
// detect nuclear particle type
ReadMatrix dataReader;
string filename = "nuclear";
string delimiter = ", ";
MatrixXd data= dataReader.read(filename, delimiter);
LINE;
LDA lda(false);
lda.train(data);
LINE;
cout << "Prediction of the first five labels: " << endl;
MatrixXd X_ = data.block(0, 0, 5, data.cols() - 1);
cout << lda.predict(X_).transpose() << endl;
LINE; LINE;
}
int main()
{
ReadMatrix dataReader;
string filename = "dataset";
string delimiter = ", ";
MatrixXd data = dataReader.read(filename, delimiter);
LINE;
LAR lar;
lar.train(data);
//predict on first 20 X;
LINE;
cout << "Prediction of y on first 20 x: " << endl << endl;
cout << lar.predict(data.block(0, 0, 20, data.cols() - 1)).transpose() << endl<<endl;
cout << (data.block(0, data.cols() - 1, 20, 1)).transpose() << endl;
LINE;
}
int main()
{
// detect nuclear particle type
ReadMatrix dataReader;
string filename = "eigenFace";
string delimiter = ", ";
MatrixXd data = dataReader.read(filename, delimiter);
LINE;
KM km;
km.train(data);
LINE;
cout << "Clusters label of original data: " << endl<<endl;
MatrixXd clusterLabel = km.clusterLabel(data);
cout << clusterLabel.transpose() << endl << endl;
LINE;
}
int ClusterCommand::execute(){
try {
if (abort == true) { if (calledHelp) { return 0; } return 2; }
//phylip file given and cutoff not given - use cluster.classic because it uses less memory and is faster
if ((format == "phylip") && (cutoff > 10.0)) {
m->mothurOutEndLine(); m->mothurOut("You are using a phylip file and no cutoff. I will run cluster.classic to save memory and time."); m->mothurOutEndLine();
//run unique.seqs for deconvolute results
string inputString = "phylip=" + distfile;
if (namefile != "") { inputString += ", name=" + namefile; }
inputString += ", precision=" + toString(precision);
inputString += ", method=" + method;
if (hard) { inputString += ", hard=T"; }
else { inputString += ", hard=F"; }
if (sim) { inputString += ", sim=T"; }
else { inputString += ", sim=F"; }
m->mothurOutEndLine();
m->mothurOut("/------------------------------------------------------------/"); m->mothurOutEndLine();
m->mothurOut("Running command: cluster.classic(" + inputString + ")"); m->mothurOutEndLine();
Command* clusterClassicCommand = new ClusterDoturCommand(inputString);
clusterClassicCommand->execute();
delete clusterClassicCommand;
m->mothurOut("/------------------------------------------------------------/"); m->mothurOutEndLine();
return 0;
}
ReadMatrix* read;
if (format == "column") { read = new ReadColumnMatrix(columnfile, sim); } //sim indicates whether its a similarity matrix
else if (format == "phylip") { read = new ReadPhylipMatrix(phylipfile, sim); }
read->setCutoff(cutoff);
NameAssignment* nameMap = NULL;
if(namefile != ""){
nameMap = new NameAssignment(namefile);
nameMap->readMap();
}
read->read(nameMap);
list = read->getListVector();
matrix = read->getMatrix();
rabund = new RAbundVector(list->getRAbundVector());
delete read;
if (m->control_pressed) { //clean up
delete list; delete matrix; delete rabund;
sabundFile.close();rabundFile.close();listFile.close();
for (int i = 0; i < outputNames.size(); i++) { m->mothurRemove(outputNames[i]); } outputTypes.clear();
return 0;
}
//create cluster
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); }
else if(method == "weighted"){ cluster = new WeightedLinkage(rabund, list, matrix, cutoff, method); }
tag = cluster->getTag();
if (outputDir == "") { outputDir += m->hasPath(distfile); }
fileroot = outputDir + m->getRootName(m->getSimpleName(distfile));
m->openOutputFile(fileroot+ tag + ".sabund", sabundFile);
m->openOutputFile(fileroot+ tag + ".rabund", rabundFile);
m->openOutputFile(fileroot+ tag + ".list", listFile);
outputNames.push_back(fileroot+ tag + ".sabund"); outputTypes["sabund"].push_back(fileroot+ tag + ".sabund");
outputNames.push_back(fileroot+ tag + ".rabund"); outputTypes["rabund"].push_back(fileroot+ tag + ".rabund");
outputNames.push_back(fileroot+ tag + ".list"); outputTypes["list"].push_back(fileroot+ tag + ".list");
time_t estart = time(NULL);
float previousDist = 0.00000;
float rndPreviousDist = 0.00000;
oldRAbund = *rabund;
oldList = *list;
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 list; delete matrix; delete rabund; delete cluster;
sabundFile.close();rabundFile.close();listFile.close();
for (int i = 0; i < outputNames.size(); i++) { m->mothurRemove(outputNames[i]); } outputTypes.clear();
return 0;
}
if (print_start && m->isTrue(timing)) {
m->mothurOut("Clustering (" + tag + ") dist " + toString(matrix->getSmallDist()) + "/"
+ toString(m->roundDist(matrix->getSmallDist(), precision))
+ "\t(precision: " + toString(precision) + ", Nodes: " + toString(matrix->getNNodes()) + ")");
cout.flush();
print_start = false;
}
loops++;
cluster->update(cutoff);
float dist = matrix->getSmallDist();
float rndDist;
if (hard) {
rndDist = m->ceilDist(dist, precision);
}else{
rndDist = m->roundDist(dist, precision);
}
if(previousDist <= 0.0000 && dist != previousDist){
printData("unique");
}
else if(rndDist != rndPreviousDist){
printData(toString(rndPreviousDist, length-1));
}
previousDist = dist;
rndPreviousDist = rndDist;
oldRAbund = *rabund;
oldList = *list;
}
if (print_start && m->isTrue(timing)) {
m->mothurOut("Clustering (" + tag + ") for distance " + toString(previousDist) + "/" + toString(rndPreviousDist)
+ "\t(precision: " + toString(precision) + ", Nodes: " + toString(matrix->getNNodes()) + ")");
cout.flush();
print_start = false;
}
if(previousDist <= 0.0000){
printData("unique");
}
else if(rndPreviousDist<cutoff){
printData(toString(rndPreviousDist, length-1));
}
delete matrix;
delete list;
delete rabund;
delete cluster;
sabundFile.close();
rabundFile.close();
listFile.close();
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");
if (itTypes != outputTypes.end()) {
if ((itTypes->second).size() != 0) { current = (itTypes->second)[0]; m->setListFile(current); }
}
//set rabund file as new current rabundfile
itTypes = outputTypes.find("rabund");
if (itTypes != outputTypes.end()) {
if ((itTypes->second).size() != 0) { current = (itTypes->second)[0]; m->setRabundFile(current); }
}
//set sabund file as new current sabundfile
itTypes = outputTypes.find("sabund");
if (itTypes != outputTypes.end()) {
if ((itTypes->second).size() != 0) { current = (itTypes->second)[0]; m->setSabundFile(current); }
}
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 (m->isTrue(timing)) {
m->mothurOut("It took " + toString(time(NULL) - estart) + " seconds to cluster"); m->mothurOutEndLine();
//}
return 0;
}
catch(exception& e) {
m->errorOut(e, "ClusterCommand", "execute");
exit(1);
}
}
int TreeGroupCommand::execute(){
try {
if (abort == true) { if (calledHelp) { return 0; } return 2; }
if (format == "sharedfile") {
ValidCalculators validCalculator;
for (int i=0; i<Estimators.size(); i++) {
if (validCalculator.isValidCalculator("treegroup", Estimators[i]) == true) {
if (Estimators[i] == "sharedsobs") {
treeCalculators.push_back(new SharedSobsCS());
}else if (Estimators[i] == "sharedchao") {
treeCalculators.push_back(new SharedChao1());
}else if (Estimators[i] == "sharedace") {
treeCalculators.push_back(new SharedAce());
}else if (Estimators[i] == "jabund") {
treeCalculators.push_back(new JAbund());
}else if (Estimators[i] == "sorabund") {
treeCalculators.push_back(new SorAbund());
}else if (Estimators[i] == "jclass") {
treeCalculators.push_back(new Jclass());
}else if (Estimators[i] == "sorclass") {
treeCalculators.push_back(new SorClass());
}else if (Estimators[i] == "jest") {
treeCalculators.push_back(new Jest());
}else if (Estimators[i] == "sorest") {
treeCalculators.push_back(new SorEst());
}else if (Estimators[i] == "thetayc") {
treeCalculators.push_back(new ThetaYC());
}else if (Estimators[i] == "thetan") {
treeCalculators.push_back(new ThetaN());
}else if (Estimators[i] == "kstest") {
treeCalculators.push_back(new KSTest());
}else if (Estimators[i] == "sharednseqs") {
treeCalculators.push_back(new SharedNSeqs());
}else if (Estimators[i] == "ochiai") {
treeCalculators.push_back(new Ochiai());
}else if (Estimators[i] == "anderberg") {
treeCalculators.push_back(new Anderberg());
}else if (Estimators[i] == "kulczynski") {
treeCalculators.push_back(new Kulczynski());
}else if (Estimators[i] == "kulczynskicody") {
treeCalculators.push_back(new KulczynskiCody());
}else if (Estimators[i] == "lennon") {
treeCalculators.push_back(new Lennon());
}else if (Estimators[i] == "morisitahorn") {
treeCalculators.push_back(new MorHorn());
}else if (Estimators[i] == "braycurtis") {
treeCalculators.push_back(new BrayCurtis());
}else if (Estimators[i] == "whittaker") {
treeCalculators.push_back(new Whittaker());
}else if (Estimators[i] == "odum") {
treeCalculators.push_back(new Odum());
}else if (Estimators[i] == "canberra") {
treeCalculators.push_back(new Canberra());
}else if (Estimators[i] == "structeuclidean") {
treeCalculators.push_back(new StructEuclidean());
}else if (Estimators[i] == "structchord") {
treeCalculators.push_back(new StructChord());
}else if (Estimators[i] == "hellinger") {
treeCalculators.push_back(new Hellinger());
}else if (Estimators[i] == "manhattan") {
treeCalculators.push_back(new Manhattan());
}else if (Estimators[i] == "structpearson") {
treeCalculators.push_back(new StructPearson());
}else if (Estimators[i] == "soergel") {
treeCalculators.push_back(new Soergel());
}else if (Estimators[i] == "spearman") {
treeCalculators.push_back(new Spearman());
}else if (Estimators[i] == "structkulczynski") {
treeCalculators.push_back(new StructKulczynski());
}else if (Estimators[i] == "speciesprofile") {
treeCalculators.push_back(new SpeciesProfile());
}else if (Estimators[i] == "hamming") {
treeCalculators.push_back(new Hamming());
}else if (Estimators[i] == "structchi2") {
treeCalculators.push_back(new StructChi2());
}else if (Estimators[i] == "gower") {
treeCalculators.push_back(new Gower());
}else if (Estimators[i] == "memchi2") {
treeCalculators.push_back(new MemChi2());
}else if (Estimators[i] == "memchord") {
treeCalculators.push_back(new MemChord());
}else if (Estimators[i] == "memeuclidean") {
treeCalculators.push_back(new MemEuclidean());
}else if (Estimators[i] == "mempearson") {
treeCalculators.push_back(new MemPearson());
}else if (Estimators[i] == "jsd") {
treeCalculators.push_back(new JSD());
}else if (Estimators[i] == "rjsd") {
treeCalculators.push_back(new RJSD());
}
}
}
//if the users entered no valid calculators don't execute command
if (treeCalculators.size() == 0) { m->mothurOut("You have given no valid calculators."); m->mothurOutEndLine(); return 0; }
input = new InputData(sharedfile, "sharedfile");
lookup = input->getSharedRAbundVectors();
lastLabel = lookup[0]->getLabel();
if (lookup.size() < 2) { m->mothurOut("You have not provided enough valid groups. I cannot run the command."); m->mothurOutEndLine(); return 0; }
//used in tree constructor
m->runParse = false;
//create treemap class from groupmap for tree class to use
ct = new CountTable();
set<string> nameMap;
map<string, string> groupMap;
set<string> gps;
for (int i = 0; i < m->getAllGroups().size(); i++) {
nameMap.insert(m->getAllGroups()[i]);
gps.insert(m->getAllGroups()[i]);
groupMap[m->getAllGroups()[i]] = m->getAllGroups()[i];
}
ct->createTable(nameMap, groupMap, gps);
//clear globaldatas old tree names if any
m->Treenames.clear();
//fills globaldatas tree names
//m->Treenames = m->getGroups();
for (int k = 0; k < lookup.size(); k++) {
m->Treenames.push_back(lookup[k]->getGroup());
}
if (m->control_pressed) { return 0; }
//create tree file
makeSimsShared();
if (m->control_pressed) { for (int i = 0; i < outputNames.size(); i++) { m->mothurRemove(outputNames[i]); } return 0; }
}else{
//read in dist file
filename = inputfile;
ReadMatrix* readMatrix;
if (format == "column") { readMatrix = new ReadColumnMatrix(filename); }
else if (format == "phylip") { readMatrix = new ReadPhylipMatrix(filename); }
readMatrix->setCutoff(cutoff);
ct = NULL;
nameMap = NULL;
if(namefile != ""){
nameMap = new NameAssignment(namefile);
nameMap->readMap();
readMatrix->read(nameMap);
}else if (countfile != "") {
ct = new CountTable();
ct->readTable(countfile, true, false);
readMatrix->read(ct);
}else {
readMatrix->read(nameMap);
}
list = readMatrix->getListVector();
SparseDistanceMatrix* dMatrix = readMatrix->getDMatrix();
//clear globaldatas old tree names if any
m->Treenames.clear();
//make treemap
if (ct != NULL) { delete ct; }
ct = new CountTable();
set<string> nameMap;
map<string, string> groupMap;
set<string> gps;
for (int i = 0; i < list->getNumBins(); i++) {
string bin = list->get(i);
nameMap.insert(bin);
gps.insert(bin);
groupMap[bin] = bin;
m->Treenames.push_back(bin);
}
ct->createTable(nameMap, groupMap, gps);
vector<string> namesGroups = ct->getNamesOfGroups();
m->setGroups(namesGroups);
//used in tree constructor
m->runParse = false;
if (m->control_pressed) { return 0; }
vector< vector<double> > matrix = makeSimsDist(dMatrix);
delete readMatrix;
delete dMatrix;
if (m->control_pressed) { return 0; }
//create a new filename
map<string, string> variables;
variables["[filename]"] = outputDir + m->getRootName(m->getSimpleName(inputfile));
string outputFile = getOutputFileName("tree",variables);
outputNames.push_back(outputFile); outputTypes["tree"].push_back(outputFile);
Tree* newTree = createTree(matrix);
if (newTree != NULL) { writeTree(outputFile, newTree); delete newTree; }
if (m->control_pressed) { return 0; }
m->mothurOut("Tree complete. "); m->mothurOutEndLine();
}
//reset groups parameter
m->clearGroups();
//set tree file as new current treefile
string current = "";
itTypes = outputTypes.find("tree");
if (itTypes != outputTypes.end()) {
if ((itTypes->second).size() != 0) { current = (itTypes->second)[0]; m->setTreeFile(current); }
}
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();
return 0;
}
catch(exception& e) {
m->errorOut(e, "TreeGroupCommand", "execute");
exit(1);
}
}
int TreeGroupCommand::execute(){
try {
if (abort) { if (calledHelp) { return 0; } return 2; }
if (format == "sharedfile") {
InputData input(sharedfile, "sharedfile", Groups);
SharedRAbundVectors* lookup = input.getSharedRAbundVectors();
lastLabel = lookup->getLabel();
Groups = lookup->getNamesGroups();
if (lookup->size() < 2) { m->mothurOut("You have not provided enough valid groups. I cannot run the command.\n"); return 0; }
//create treemap class from groupmap for tree class to use
CountTable ct;
set<string> nameMap;
map<string, string> groupMap;
set<string> gps;
for (int i = 0; i < Groups.size(); i++) {
nameMap.insert(Groups[i]);
gps.insert(Groups[i]);
groupMap[Groups[i]] = Groups[i];
}
ct.createTable(nameMap, groupMap, gps);
//fills tree names with shared files groups
Treenames = lookup->getNamesGroups();
if (m->getControl_pressed()) { return 0; }
//create tree file
makeSimsShared(input, lookup, ct);
if (m->getControl_pressed()) { for (int i = 0; i < outputNames.size(); i++) { util.mothurRemove(outputNames[i]); } return 0; }
}else{
//read in dist file
filename = inputfile;
ReadMatrix* readMatrix;
if (format == "column") { readMatrix = new ReadColumnMatrix(filename); }
else if (format == "phylip") { readMatrix = new ReadPhylipMatrix(filename); }
readMatrix->setCutoff(cutoff);
ListVector* list;
if(namefile != ""){
NameAssignment* nameMap = new NameAssignment(namefile);
nameMap->readMap();
readMatrix->read(nameMap);
list = readMatrix->getListVector();
delete nameMap;
}else if (countfile != "") {
CountTable* ct = new CountTable();
ct->readTable(countfile, true, false);
readMatrix->read(ct);
list = readMatrix->getListVector();
delete ct;
}else { NameAssignment* nameMap = NULL; readMatrix->read(nameMap); list = readMatrix->getListVector(); }
SparseDistanceMatrix* dMatrix = readMatrix->getDMatrix();
Treenames.clear();
//make treemap
CountTable ct;
set<string> nameMap;
map<string, string> groupMap;
set<string> gps;
for (int i = 0; i < list->getNumBins(); i++) {
string bin = list->get(i);
nameMap.insert(bin);
gps.insert(bin);
groupMap[bin] = bin;
Treenames.push_back(bin);
}
ct.createTable(nameMap, groupMap, gps);
vector<string> namesGroups = ct.getNamesOfGroups();
if (m->getControl_pressed()) { return 0; }
vector< vector<double> > matrix = makeSimsDist(dMatrix, list->getNumBins());
delete readMatrix;
delete dMatrix;
if (m->getControl_pressed()) { return 0; }
//create a new filename
map<string, string> variables;
variables["[filename]"] = outputDir + util.getRootName(util.getSimpleName(inputfile));
string outputFile = getOutputFileName("tree",variables);
outputNames.push_back(outputFile); outputTypes["tree"].push_back(outputFile);
Tree* newTree = new Tree(&ct, matrix, Treenames);
if (m->getControl_pressed()) { delete newTree; newTree = NULL; }
else { newTree->assembleTree(); }
if (newTree != NULL) { newTree->createNewickFile(outputFile); delete newTree; }
if (m->getControl_pressed()) { return 0; } m->mothurOut("Tree complete.\n");
}
//set tree file as new current treefile
string currentName = "";
itTypes = outputTypes.find("tree");
if (itTypes != outputTypes.end()) {
if ((itTypes->second).size() != 0) { currentName = (itTypes->second)[0]; current->setTreeFile(currentName); }
}
m->mothurOut("\nOutput File Names: \n");
for (int i = 0; i < outputNames.size(); i++) { m->mothurOut(outputNames[i] +"\n"); } m->mothurOutEndLine();
return 0;
}
catch(exception& e) {
m->errorOut(e, "TreeGroupCommand", "execute");
exit(1);
}
}
