-
/*
* pcacommand.cpp
- * Mothur
+ * mothur
*
- * Created by westcott on 1/4/10.
- * Copyright 2010 Schloss Lab. All rights reserved.
+ * Created by westcott on 1/7/11.
+ * Copyright 2011 Schloss Lab. All rights reserved.
*
*/
#include "pcacommand.h"
+#include "inputdata.h"
+
+//**********************************************************************************************************************
+vector<string> PCACommand::setParameters(){
+ try {
+ CommandParameter pshared("shared", "InputTypes", "", "", "LRSS", "LRSS", "none","pca-loadings",false,false,true); parameters.push_back(pshared);
+ CommandParameter prelabund("relabund", "InputTypes", "", "", "LRSS", "LRSS", "none","pca-loadings",false,false,true); parameters.push_back(prelabund);
+ CommandParameter pgroups("groups", "String", "", "", "", "", "","",false,false); parameters.push_back(pgroups);
+ CommandParameter pmetric("metric", "Boolean", "", "T", "", "", "","",false,false); parameters.push_back(pmetric);
+ CommandParameter plabel("label", "String", "", "", "", "", "","",false,false); parameters.push_back(plabel);
+ CommandParameter pinputdir("inputdir", "String", "", "", "", "", "","",false,false); parameters.push_back(pinputdir);
+ CommandParameter poutputdir("outputdir", "String", "", "", "", "", "","",false,false); parameters.push_back(poutputdir);
+
+ vector<string> myArray;
+ for (int i = 0; i < parameters.size(); i++) { myArray.push_back(parameters[i].name); }
+ return myArray;
+ }
+ catch(exception& e) {
+ m->errorOut(e, "PCACommand", "setParameters");
+ exit(1);
+ }
+}
+//**********************************************************************************************************************
+string PCACommand::getHelpString(){
+ try {
+ string helpString = "";
+ helpString += "The pca command parameters are shared, relabund, label, groups and metric. shared or relabund is required unless you have a valid current file.";
+ helpString += "The label parameter is used to analyze specific labels in your input. Default is the first label in your shared or relabund file. Multiple labels may be separated by dashes.\n";
+ helpString += "The groups parameter allows you to specify which groups you would like analyzed. Groupnames are separated by dashes.\n";
+ helpString += "The metric parameter allows you to indicate if would like the pearson correlation coefficient calculated. Default=True";
+ helpString += "Example pca(groups=yourGroups).\n";
+ helpString += "Example pca(groups=A-B-C).\n";
+ helpString += "Note: No spaces between parameter labels (i.e. groups), '=' and parameters (i.e.yourGroups).\n";
+ return helpString;
+ }
+ catch(exception& e) {
+ m->errorOut(e, "PCACommand", "getHelpString");
+ exit(1);
+ }
+}
+//**********************************************************************************************************************
+string PCACommand::getOutputPattern(string type) {
+ try {
+ string pattern = "";
+
+ if (type == "pca") { pattern = "[filename],[distance],pca.axes"; }
+ else if (type == "loadings") { pattern = "[filename],[distance],pca.loadings"; }
+ else { m->mothurOut("[ERROR]: No definition for type " + type + " output pattern.\n"); m->control_pressed = true; }
+
+ return pattern;
+ }
+ catch(exception& e) {
+ m->errorOut(e, "PCACommand", "getOutputPattern");
+ exit(1);
+ }
+}
+//**********************************************************************************************************************
+PCACommand::PCACommand(){
+ try {
+ abort = true; calledHelp = true;
+ setParameters();
+ vector<string> tempOutNames;
+ outputTypes["pca"] = tempOutNames;
+ outputTypes["loadings"] = tempOutNames;
+ }
+ catch(exception& e) {
+ m->errorOut(e, "PCACommand", "PCACommand");
+ exit(1);
+ }
+}
//**********************************************************************************************************************
-PCACommand::PCACommand(string option){
+PCACommand::PCACommand(string option) {
try {
- abort = false;
+ abort = false; calledHelp = false;
//allow user to run help
- if(option == "help") { help(); abort = true; }
+ if(option == "help") { help(); abort = true; calledHelp = true; }
+ else if(option == "citation") { citation(); abort = true; calledHelp = true;}
else {
- //valid paramters for this command
- string Array[] = {"phylip"};
- vector<string> myArray (Array, Array+(sizeof(Array)/sizeof(string)));
+ vector<string> myArray = setParameters();
OptionParser parser(option);
map<string, string> parameters = parser. getParameters();
ValidParameters validParameter;
-
+ map<string, string>::iterator it;
+
//check to make sure all parameters are valid for command
- for (map<string, string>::iterator it = parameters.begin(); it != parameters.end(); it++) {
+ for (it = parameters.begin(); it != parameters.end(); it++) {
if (validParameter.isValidParameter(it->first, myArray, it->second) != true) { abort = true; }
}
+
+ //initialize outputTypes
+ vector<string> tempOutNames;
+ outputTypes["pca"] = tempOutNames;
+ outputTypes["loadings"] = tempOutNames;
- //required parameters
- phylipfile = validParameter.validFile(parameters, "phylip", true);
- if (phylipfile == "not open") { abort = true; }
- else if (phylipfile == "not found") { phylipfile = ""; abort = true; }
- else { filename = phylipfile; }
+ //if the user changes the input directory command factory will send this info to us in the output parameter
+ string inputDir = validParameter.validFile(parameters, "inputdir", false);
+ if (inputDir == "not found"){ inputDir = ""; }
+ else {
+ string path;
+ it = parameters.find("shared");
+ //user has given a template file
+ if(it != parameters.end()){
+ path = m->hasPath(it->second);
+ //if the user has not given a path then, add inputdir. else leave path alone.
+ if (path == "") { parameters["shared"] = inputDir + it->second; }
+ }
+
+ it = parameters.find("relabund");
+ //user has given a template file
+ if(it != parameters.end()){
+ path = m->hasPath(it->second);
+ //if the user has not given a path then, add inputdir. else leave path alone.
+ if (path == "") { parameters["relabund"] = inputDir + it->second; }
+ }
+ }
- //columnfile = validParameter.validFile(parameters, "column", true);
- //if (columnfile == "not open") { abort = true; }
- //else if (columnfile == "not found") { columnfile = ""; }
- //else { format = "column"; }
+ //check for required parameters
+ sharedfile = validParameter.validFile(parameters, "shared", true);
+ if (sharedfile == "not open") { sharedfile = ""; abort = true; }
+ else if (sharedfile == "not found") { sharedfile = ""; }
+ else { mode = "sharedfile"; inputFile = sharedfile; m->setSharedFile(sharedfile); }
- //namefile = validParameter.validFile(parameters, "name", true);
- //if (namefile == "not open") { abort = true; }
- //else if (namefile == "not found") { namefile = ""; }
+ relabundfile = validParameter.validFile(parameters, "relabund", true);
+ if (relabundfile == "not open") { relabundfile = ""; abort = true; }
+ else if (relabundfile == "not found") { relabundfile = ""; }
+ else { mode = "relabund"; inputFile = relabundfile; m->setRelAbundFile(relabundfile); }
- //error checking on files
- if (phylipfile == "") { mothurOut("You must provide a distance file before running the pca command."); mothurOutEndLine(); abort = true; }
- //if ((phylipfile == "") && (columnfile == "")) { mothurOut("You must provide a distance file before running the pca command."); mothurOutEndLine(); abort = true; }
- //else if ((phylipfile != "") && (columnfile != "")) { mothurOut("You may not use both the column and the phylip parameters."); mothurOutEndLine(); abort = true; }
+ if ((sharedfile == "") && (relabundfile == "")) {
+ //is there are current file available for any of these?
+ //give priority to shared, then list, then rabund, then sabund
+ //if there is a current shared file, use it
+ sharedfile = m->getSharedFile();
+ if (sharedfile != "") { inputFile = sharedfile; mode = "sharedfile"; m->mothurOut("Using " + sharedfile + " as input file for the shared parameter."); m->mothurOutEndLine(); }
+ else {
+ relabundfile = m->getRelAbundFile();
+ if (relabundfile != "") { inputFile = relabundfile; mode = "relabund"; m->mothurOut("Using " + relabundfile + " as input file for the relabund parameter."); m->mothurOutEndLine(); }
+ else {
+ m->mothurOut("No valid current files. You must provide a relabund or shared file."); m->mothurOutEndLine();
+ abort = true;
+ }
+ }
+ }
+
+ //if the user changes the output directory command factory will send this info to us in the output parameter
+ outputDir = validParameter.validFile(parameters, "outputdir", false); if (outputDir == "not found"){
+ outputDir = "";
+ outputDir += m->hasPath(inputFile); //if user entered a file with a path then preserve it
+ }
+
+ string temp = validParameter.validFile(parameters, "metric", false); if (temp == "not found"){ temp = "T"; }
+ metric = m->isTrue(temp);
- //if (columnfile != "") {
- // if (namefile == "") { mothurOut("You need to provide a namefile if you are going to use the column format."); mothurOutEndLine(); abort = true; }
- //}
+ label = validParameter.validFile(parameters, "label", false);
+ if (label == "not found") { label = ""; if(labels.size() == 0) { m->mothurOut("You did not provide a label, I will use the first label in your inputfile."); m->mothurOutEndLine(); } }
+ else { m->splitAtDash(label, labels); }
- //string temp = validParameter.validFile(parameters, "lt", false); if (temp == "not found") { temp = "false"; }
- //bool lt = isTrue(temp);
+ groups = validParameter.validFile(parameters, "groups", false);
+ if (groups == "not found") { groups = ""; }
+ else { m->splitAtDash(groups, Groups); }
+ m->setGroups(Groups);
- //if (lt) { matrix = 2; }
- //else { matrix = 1; }
-
-
}
-
- }
- catch(exception& e) {
- errorOut(e, "PCACommand", "PCACommand");
- exit(1);
- }
-}
-//**********************************************************************************************************************
-void PCACommand::help(){
- try {
-
- mothurOut("The pca command..."); mothurOutEndLine();
+
}
catch(exception& e) {
- errorOut(e, "PCACommand", "help");
+ m->errorOut(e, "PCACommand", "PCACommand");
exit(1);
}
}
//**********************************************************************************************************************
-PCACommand::~PCACommand(){}
-//**********************************************************************************************************************
int PCACommand::execute(){
try {
-
- if (abort == true) { return 0; }
+
+ if (abort == true) { if (calledHelp) { return 0; } return 2; }
cout.setf(ios::fixed, ios::floatfield);
cout.setf(ios::showpoint);
cerr.setf(ios::fixed, ios::floatfield);
cerr.setf(ios::showpoint);
- vector<string> names;
- vector<vector<double> > D;
-
- fbase = filename;
- if(fbase.find_last_of(".")!=string::npos){
- fbase.erase(fbase.find_last_of(".")+1);
+ //get first line of shared file
+ vector< vector<double> > matrix;
+ InputData* input;
+ if (mode == "sharedfile") {
+ input = new InputData(inputFile, "sharedfile");
+ }else if (mode == "relabund") {
+ input = new InputData(inputFile, "relabund");
+ }else { m->mothurOut("[ERROR]: filetype not recognized."); m->mothurOutEndLine(); return 0; }
+
+ vector<SharedRAbundFloatVector*> lookupFloat = input->getSharedRAbundFloatVectors();
+ string lastLabel = lookupFloat[0]->getLabel();
+
+ set<string> processedLabels;
+ set<string> userLabels = labels;
+
+ //if the user gave no labels, then use the first one read
+ if (labels.size() == 0) {
+ label = lastLabel;
+
+ process(lookupFloat);
}
- else{
- fbase += ".";
+
+ //as long as you are not at the end of the file or done wih the lines you want
+ while((lookupFloat[0] != NULL) && (userLabels.size() != 0)) {
+
+ if (m->control_pressed) { for (int i = 0; i < outputNames.size(); i++) { m->mothurRemove(outputNames[i]); } delete input; for (int i = 0; i < lookupFloat.size(); i++) { delete lookupFloat[i]; } lookupFloat.clear(); return 0; }
+
+ if(labels.count(lookupFloat[0]->getLabel()) == 1){
+ processedLabels.insert(lookupFloat[0]->getLabel());
+ userLabels.erase(lookupFloat[0]->getLabel());
+
+ process(lookupFloat);
+ }
+
+ if ((m->anyLabelsToProcess(lookupFloat[0]->getLabel(), userLabels, "") == true) && (processedLabels.count(lastLabel) != 1)) {
+ string saveLabel = lookupFloat[0]->getLabel();
+
+ for (int i = 0; i < lookupFloat.size(); i++) { delete lookupFloat[i]; } lookupFloat.clear();
+ lookupFloat = input->getSharedRAbundFloatVectors(lastLabel);
+
+ process(lookupFloat);
+
+ processedLabels.insert(lookupFloat[0]->getLabel());
+ userLabels.erase(lookupFloat[0]->getLabel());
+
+ //restore real lastlabel to save below
+ lookupFloat[0]->setLabel(saveLabel);
+ }
+
+ lastLabel = lookupFloat[0]->getLabel();
+
+ //get next line to process
+ //prevent memory leak
+ for (int i = 0; i < lookupFloat.size(); i++) { delete lookupFloat[i]; } lookupFloat.clear();
+ lookupFloat = input->getSharedRAbundFloatVectors();
}
- read(filename, names, D);
-
- double offset = 0.0000;
- vector<double> d;
- vector<double> e;
- vector<vector<double> > G = D;
- vector<vector<double> > copy_G;
- //int rank = D.size();
- cout << "\nProcessing...\n";
- for(int count=0;count<2;count++){
- recenter(offset, D, G);
- tred2(G, d, e);
- qtli(d, e, G);
- offset = d[d.size()-1];
- if(offset > 0.0) break;
- }
+ if (m->control_pressed) { for (int i = 0; i < outputNames.size(); i++) { m->mothurRemove(outputNames[i]); } delete input; for (int i = 0; i < lookupFloat.size(); i++) { delete lookupFloat[i]; } lookupFloat.clear(); return 0; }
+ //output error messages about any remaining user labels
+ set<string>::iterator it;
+ bool needToRun = false;
+ for (it = userLabels.begin(); it != userLabels.end(); it++) {
+ m->mothurOut("Your file does not include the label " + *it);
+ if (processedLabels.count(lastLabel) != 1) {
+ m->mothurOut(". I will use " + lastLabel + "."); m->mothurOutEndLine();
+ needToRun = true;
+ }else {
+ m->mothurOut(". Please refer to " + lastLabel + "."); m->mothurOutEndLine();
+ }
+ }
- output(fbase, names, G, d);
+ //run last label if you need to
+ if (needToRun == true) {
+ for (int i = 0; i < lookupFloat.size(); i++) { if (lookupFloat[i] != NULL) { delete lookupFloat[i]; } } lookupFloat.clear();
+ lookupFloat = input->getSharedRAbundFloatVectors(lastLabel);
+
+ process(lookupFloat);
+
+ for (int i = 0; i < lookupFloat.size(); i++) { if (lookupFloat[i] != NULL) { delete lookupFloat[i]; } } lookupFloat.clear();
+ }
- return 0;
- }
- catch(exception& e) {
- errorOut(e, "PCACommand", "execute");
- exit(1);
- }
-}
-/*********************************************************************************************************************************/
-
-inline double SIGN(const double a, const double b)
-{
- return b>=0 ? (a>=0 ? a:-a) : (a>=0 ? -a:a);
-}
-
-/*********************************************************************************************************************************/
-
-void PCACommand::get_comment(istream& f, char begin, char end){
- try {
- char d=f.get();
- while(d != end){ d = f.get(); }
- d = f.peek();
- }
- catch(exception& e) {
- errorOut(e, "PCACommand", "get_comment");
- exit(1);
- }
-}
-
-/*********************************************************************************************************************************/
-
-void PCACommand::read_mega(istream& f, vector<string>& name_list, vector<vector<double> >& d){
- try {
- get_comment(f, '#', '\n');
+ for (int i = 0; i < lookupFloat.size(); i++) { if (lookupFloat[i] != NULL) { delete lookupFloat[i]; } } lookupFloat.clear();
+ delete input;
- char test = f.peek();
+ if (m->control_pressed) { for (int i = 0; i < outputNames.size(); i++) { m->mothurRemove(outputNames[i]); } return 0; }
- while(test == '!'){ //get header comments
- get_comment(f, '!', ';');
- while(isspace(test=f.get())) {;}
- f.putback(test);
- test = f.peek();
- }
- while(test != '\n'){ //get sequence names
- get_comment(f, '[', ']');
- char d = f.get();
- d = f.get();
- if(d == '#'){
- string name;
- f >> name;
- name_list.push_back(name);
- while(isspace(test=f.get())) {;}
- f.putback(test);
- }
- else{
- break;
- }
- }
- int rank = name_list.size();
- d.resize(rank);
- for(int i=0;i<rank;i++){ d[i].resize(rank); }
-
- d[0][0] = 0.0000;
- get_comment(f, '[', ']');
- for(int i=1;i<rank;i++){
- get_comment(f, '[', ']');
- d[i][i]=0.0000;
- for(int j=0;j<i;j++){
- f >> d[i][j];
- if (d[i][j] == -0.0000)
- d[i][j] = 0.0000;
- d[j][i]=d[i][j];
- }
- }
+ 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) {
- errorOut(e, "PCACommand", "read_mega");
+ m->errorOut(e, "PCACommand", "execute");
exit(1);
}
}
-/*********************************************************************************************************************************/
-
-void PCACommand::read_phylip(istream& f, int square_m, vector<string>& name_list, vector<vector<double> >& d){
+/**********************************************************************************************************************
+vector< vector<double> > PCACommand::createMatrix(vector<SharedRAbundFloatVector*> lookupFloat){
try {
- // int count1=0;
- // int count2=0;
-
- int rank;
- f >> rank;
-
- name_list.resize(rank);
- d.resize(rank);
- if(square_m == 1){
- for(int i=0;i<rank;i++)
- d[i].resize(rank);
- for(int i=0;i<rank;i++) {
- f >> name_list[i];
- // cout << i << "\t" << name_list[i] << endl;
- for(int j=0;j<rank;j++) {
- f >> d[i][j];
- if (d[i][j] == -0.0000)
- d[i][j] = 0.0000;
- }
+ vector< vector<double> > matrix; matrix.resize(lookupFloat.size());
+
+ //fill matrix with shared files relative abundances
+ for (int i = 0; i < lookupFloat.size(); i++) {
+ for (int j = 0; j < lookupFloat[i]->getNumBins(); j++) {
+ matrix[i].push_back(lookupFloat[i]->getAbundance(j));
}
}
- else if(square_m == 2){
- for(int i=0;i<rank;i++){
- d[i].resize(rank);
- }
- d[0][0] = 0.0000;
- f >> name_list[0];
- for(int i=1;i<rank;i++){
- f >> name_list[i];
- d[i][i]=0.0000;
- for(int j=0;j<i;j++){
- f >> d[i][j];
- if (d[i][j] == -0.0000)
- d[i][j] = 0.0000;
- d[j][i]=d[i][j];
- }
+
+ vector< vector<double> > transposeMatrix; transposeMatrix.resize(matrix[0].size());
+ for (int i = 0; i < transposeMatrix.size(); i++) {
+ for (int j = 0; j < matrix.size(); j++) {
+ transposeMatrix[i].push_back(matrix[j][i]);
}
}
+
+ matrix = linearCalc.matrix_mult(matrix, transposeMatrix);
+
+ return matrix;
}
catch(exception& e) {
- errorOut(e, "PCACommand", "read_phylip");
+ m->errorOut(e, "PCACommand", "createMatrix");
exit(1);
}
+}*/
+//**********************************************************************************************************************
-}
-
-/*********************************************************************************************************************************/
-
-void PCACommand::read(string fname, vector<string>& names, vector<vector<double> >& D){
+int PCACommand::process(vector<SharedRAbundFloatVector*>& lookupFloat){
try {
- ifstream f;
- openInputFile(fname, f);
+ m->mothurOut("\nProcessing " + lookupFloat[0]->getLabel()); m->mothurOutEndLine();
+
+ int numOTUs = lookupFloat[0]->getNumBins();
+ int numSamples = lookupFloat.size();
- char test = f.peek();
+ vector< vector<double> > matrix(numSamples);
+ vector<double> colMeans(numOTUs);
- if(test == '#'){
- read_mega(f, names, D);
- }
- else{
- //check whether matrix is square
- char d;
- int m = 1;
- int numSeqs;
- string name;
+ //fill matrix with shared relative abundances, re-center
+ for (int i = 0; i < lookupFloat.size(); i++) {
+ matrix[i].resize(numOTUs, 0);
- f >> numSeqs >> name;
-
- while((d=f.get()) != EOF){
-
- //is d a number meaning its square
- if(isalnum(d)){
- m = 1;
- break;
- }
-
- //is d a line return meaning its lower triangle
- if(d == '\n'){
- m = 2;
- break;
- }
+ for (int j = 0; j < numOTUs; j++) {
+ matrix[i][j] = lookupFloat[i]->getAbundance(j);
+ colMeans[j] += matrix[i][j];
}
- f.close();
-
- //reopen to get back to beginning
- openInputFile(fname, f);
- read_phylip(f, m, names, D);
}
- //int rank = D.size();
- }
- catch(exception& e) {
- errorOut(e, "PCACommand", "read");
- exit(1);
- }
-}
-
-/*********************************************************************************************************************************/
-double PCACommand::pythag(double a, double b){
- return(pow(a*a+b*b,0.5));
-}
-/*********************************************************************************************************************************/
-void PCACommand::matrix_mult(vector<vector<double> > first, vector<vector<double> > second, vector<vector<double> >& product){
- try {
- int first_rows = first.size();
- int first_cols = first[0].size();
- int second_cols = second[0].size();
-
- product.resize(first_rows);
- for(int i=0;i<first_rows;i++){
- product[i].resize(second_cols);
+ for(int j=0;j<numOTUs;j++){
+ colMeans[j] = colMeans[j] / (double)numSamples;
}
- for(int i=0;i<first_rows;i++){
- for(int j=0;j<second_cols;j++){
- product[i][j] = 0.0;
- for(int k=0;k<first_cols;k++){
- product[i][j] += first[i][k] * second[k][j];
- }
+ vector<vector<double> > centered = matrix;
+ for(int i=0;i<numSamples;i++){
+ for(int j=0;j<numOTUs;j++){
+ centered[i][j] = centered[i][j] - colMeans[j];
}
}
- }
- catch(exception& e) {
- errorOut(e, "PCACommand", "matrix_mult");
- exit(1);
- }
-
-}
-
-/*********************************************************************************************************************************/
-void PCACommand::recenter(double offset, vector<vector<double> > D, vector<vector<double> >& G){
- try {
- int rank = D.size();
-
- vector<vector<double> > A(rank);
- vector<vector<double> > C(rank);
- for(int i=0;i<rank;i++){
- A[i].resize(rank);
- C[i].resize(rank);
- }
-
- double scale = -1.0000 / (double) rank;
- for(int i=0;i<rank;i++){
- A[i][i] = 0.0000;
- C[i][i] = 1.0000 + scale;
- for(int j=i+1;j<rank;j++){
- A[i][j] = A[j][i] = -0.5 * D[i][j] * D[i][j] + offset;
- C[i][j] = C[j][i] = scale;
+ vector< vector<double> > transpose(numOTUs);
+ for (int i = 0; i < numOTUs; i++) {
+ transpose[i].resize(numSamples, 0);
+
+ for (int j = 0; j < numSamples; j++) {
+ transpose[i][j] = centered[j][i];
}
}
-
- matrix_mult(C,A,A);
- matrix_mult(A,C,G);
- }
- catch(exception& e) {
- errorOut(e, "PCACommand", "recenter");
- exit(1);
- }
-
-}
-
-/*********************************************************************************************************************************/
-// This function is taken from Numerical Recipes in C++ by Press et al., 2nd edition, pg. 479
-
-void PCACommand::tred2(vector<vector<double> >& a, vector<double>& d, vector<double>& e){
- try {
- double scale, hh, h, g, f;
+ vector<vector<double> > crossProduct = linearCalc.matrix_mult(transpose, centered);
- int n = a.size();
+ vector<double> d;
+ vector<double> e;
+
+ linearCalc.tred2(crossProduct, d, e); if (m->control_pressed) { return 0; }
+ linearCalc.qtli(d, e, crossProduct); if (m->control_pressed) { return 0; }
- d.resize(n);
- e.resize(n);
+ vector<vector<double> > X = linearCalc.matrix_mult(centered, crossProduct);
- for(int i=n-1;i>0;i--){
- int l=i-1;
- h = scale = 0.0000;
- if(l>0){
- for(int k=0;k<l+1;k++){
- scale += fabs(a[i][k]);
- }
- if(scale == 0.0){
- e[i] = a[i][l];
- }
- else{
- for(int k=0;k<l+1;k++){
- a[i][k] /= scale;
- h += a[i][k] * a[i][k];
- }
- f = a[i][l];
- g = (f >= 0.0 ? -sqrt(h) : sqrt(h));
- e[i] = scale * g;
- h -= f * g;
- a[i][l] = f - g;
- f = 0.0;
- for(int j=0;j<l+1;j++){
- a[j][i] = a[i][j] / h;
- g = 0.0;
- for(int k=0;k<j+1;k++){
- g += a[j][k] * a[i][k];
- }
- for(int k=j+1;k<l+1;k++){
- g += a[k][j] * a[i][k];
- }
- e[j] = g / h;
- f += e[j] * a[i][j];
- }
- hh = f / (h + h);
- for(int j=0;j<l+1;j++){
- f = a[i][j];
- e[j] = g = e[j] - hh * f;
- for(int k=0;k<j+1;k++){
- a[j][k] -= (f * e[k] + g * a[i][k]);
- }
- }
- }
- }
- else{
- e[i] = a[i][l];
- }
-
- d[i] = h;
- }
+ if (m->control_pressed) { return 0; }
- d[0] = 0.0000;
- e[0] = 0.0000;
+ string fbase = outputDir + m->getRootName(m->getSimpleName(inputFile));
+ //string outputFileName = fbase + lookupFloat[0]->getLabel();
+ output(fbase, lookupFloat[0]->getLabel(), m->getGroups(), X, d);
- for(int i=0;i<n;i++){
- int l = i;
- if(d[i] != 0.0){
- for(int j=0;j<l;j++){
- g = 0.0000;
- for(int k=0;k<l;k++){
- g += a[i][k] * a[k][j];
- }
- for(int k=0;k<l;k++){
- a[k][j] -= g * a[k][i];
- }
- }
- }
- d[i] = a[i][i];
- a[i][i] = 1.0000;
- for(int j=0;j<l;j++){
- a[j][i] = a[i][j] = 0.0;
- }
- }
- }
- catch(exception& e) {
- errorOut(e, "PCACommand", "tred2");
- exit(1);
- }
-
-}
-
-/*********************************************************************************************************************************/
-
-// This function is taken from Numerical Recipes in C++ by Press et al., 2nd edition, pg. 479
+ if (metric) {
+
+ vector<vector<double> > observedEuclideanDistance = linearCalc.getObservedEuclideanDistance(centered);
+
+ for (int i = 1; i < 4; i++) {
+
+ vector< vector<double> > PCAEuclidDists = linearCalc.calculateEuclidianDistance(X, i); //G is the pca file
+
+ if (m->control_pressed) { for (int i = 0; i < outputNames.size(); i++) { m->mothurRemove(outputNames[i]); } return 0; }
-void PCACommand::qtli(vector<double>& d, vector<double>& e, vector<vector<double> >& z) {
- try {
- int m, i, iter;
- double s, r, p, g, f, dd, c, b;
-
- int n = d.size();
- for(int i=1;i<=n;i++){
- e[i-1] = e[i];
- }
- e[n-1] = 0.0000;
-
- for(int l=0;l<n;l++){
- iter = 0;
- do {
- for(m=l;m<n-1;m++){
- dd = fabs(d[m]) + fabs(d[m+1]);
- if(fabs(e[m])+dd == dd) break;
- }
- if(m != l){
- if(iter++ == 30) cerr << "Too many iterations in tqli\n";
- g = (d[l+1]-d[l]) / (2.0 * e[l]);
- r = pythag(g, 1.0);
- g = d[m] - d[l] + e[l] / (g + SIGN(r,g));
- s = c = 1.0;
- p = 0.0000;
- for(i=m-1;i>=l;i--){
- f = s * e[i];
- b = c * e[i];
- e[i+1] = (r=pythag(f,g));
- if(r==0.0){
- d[i+1] -= p;
- e[m] = 0.0000;
- break;
- }
- s = f / r;
- c = g / r;
- g = d[i+1] - p;
- r = (d[i] - g) * s + 2.0 * c * b;
- d[i+1] = g + ( p = s * r);
- g = c * r - b;
- for(int k=0;k<n;k++){
- f = z[k][i+1];
- z[k][i+1] = s * z[k][i] + c * f;
- z[k][i] = c * z[k][i] - s * f;
- }
- }
- if(r == 0.00 && i >= l) continue;
- d[l] -= p;
- e[l] = g;
- e[m] = 0.0;
- }
- } while (m != l);
- }
-
- int k;
- for(int i=0;i<n;i++){
- p=d[k=i];
- for(int j=i;j<n;j++){
- if(d[j] >= p){
- p=d[k=j];
- }
- }
- if(k!=i){
- d[k]=d[i];
- d[i]=p;
- for(int j=0;j<n;j++){
- p=z[j][i];
- z[j][i] = z[j][k];
- z[j][k] = p;
- }
+ double corr = linearCalc.calcPearson(PCAEuclidDists, observedEuclideanDistance);
+
+ m->mothurOut("Rsq " + toString(i) + " axis: " + toString(corr * corr)); m->mothurOutEndLine();
+
+ if (m->control_pressed) { for (int i = 0; i < outputNames.size(); i++) { m->mothurRemove(outputNames[i]); } return 0; }
}
}
+
+ return 0;
}
catch(exception& e) {
- errorOut(e, "PCACommand", "qtli");
+ m->errorOut(e, "PCACommand", "process");
exit(1);
}
}
-
/*********************************************************************************************************************************/
-void PCACommand::output(string fnameRoot, vector<string> name_list, vector<vector<double> > G, vector<double> d) {
+void PCACommand::output(string fbase, string label, vector<string> name_list, vector<vector<double> >& G, vector<double> d) {
try {
- int rank = name_list.size();
+
+ int numEigenValues = d.size();
double dsum = 0.0000;
- for(int i=0;i<rank;i++){
+ for(int i=0;i<numEigenValues;i++){
dsum += d[i];
- for(int j=0;j<rank;j++){
- if(d[j] >= 0) { G[i][j] *= pow(d[j],0.5); }
- else { G[i][j] = 0.00000; }
- }
}
- ofstream pcaData((fnameRoot+"pca").c_str(), ios::trunc);
+ ofstream pcaData;
+ map<string, string> variables;
+ variables["[filename]"] = fbase;
+ variables["[distance]"] = label;
+ string pcaFileName = getOutputFileName("pca",variables);
+ m->openOutputFile(pcaFileName, pcaData);
pcaData.setf(ios::fixed, ios::floatfield);
pcaData.setf(ios::showpoint);
+ outputNames.push_back(pcaFileName);
+ outputTypes["pca"].push_back(pcaFileName);
- ofstream pcaLoadings((fnameRoot+"pca.loadings").c_str(), ios::trunc);
+ ofstream pcaLoadings;
+ string loadingsFilename = getOutputFileName("loadings",variables);
+ m->openOutputFile(loadingsFilename, pcaLoadings);
pcaLoadings.setf(ios::fixed, ios::floatfield);
- pcaLoadings.setf(ios::showpoint);
+ pcaLoadings.setf(ios::showpoint);
+ outputNames.push_back(loadingsFilename);
+ outputTypes["loadings"].push_back(loadingsFilename);
pcaLoadings << "axis\tloading\n";
- for(int i=0;i<rank;i++){
+ for(int i=0;i<numEigenValues;i++){
pcaLoadings << i+1 << '\t' << d[i] * 100.0 / dsum << endl;
}
- pcaData << "SeqName";
- for(int i=0;i<rank;i++){
+ pcaData << "group";
+ for(int i=0;i<numEigenValues;i++){
pcaData << '\t' << "axis" << i+1;
}
pcaData << endl;
- for(int i=0;i<rank;i++){
+ for(int i=0;i<name_list.size();i++){
pcaData << name_list[i] << '\t';
- for(int j=0;j<rank;j++){
+ for(int j=0;j<numEigenValues;j++){
pcaData << G[i][j] << '\t';
}
pcaData << endl;
}
}
catch(exception& e) {
- errorOut(e, "PCACommand", "output");
+ m->errorOut(e, "PCACommand", "output");
exit(1);
}
}
-
/*********************************************************************************************************************************/
-void PCACommand::print_matrix(vector<vector<double> > A) {
- try {
- int rank = A.size();
- for(int i=0;i<rank;i++){
- for(int j=0;j<rank;j++){
- cout << A[i][j] << " ";
- }
- cout << endl;
- }
- }
- catch(exception& e) {
- errorOut(e, "PCACommand", "print_matrix");
- exit(1);
- }
-}
-/*********************************************************************************************************************************/
projects / mothur.git / blobdiff