X-Git-Url: https://git.donarmstrong.com/?a=blobdiff_plain;f=pcacommand.cpp;h=23c066b7beff9dd0d6c1da089221a07e510411ae;hb=943d3038553f82378567d3b651969146716e5862;hp=ee0a5462de87049f6b026546492203be630ea8d7;hpb=191ae1be0679d5cf4eda950b3b1bf26fb7dd503d;p=mothur.git diff --git a/pcacommand.cpp b/pcacommand.cpp index ee0a546..23c066b 100644 --- a/pcacommand.cpp +++ b/pcacommand.cpp @@ -1,19 +1,19 @@ - /* * 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 PCACommand::getValidParameters(){ try { - string Array[] = {"phylip", "outputdir","inputdir"}; + string Array[] = {"label", "groups","metric","outputdir","inputdir"}; vector myArray (Array, Array+(sizeof(Array)/sizeof(string))); return myArray; } @@ -25,10 +25,9 @@ vector PCACommand::getValidParameters(){ //********************************************************************************************************************** PCACommand::PCACommand(){ try { - abort = true; - //initialize outputTypes + abort = true; calledHelp = true; vector tempOutNames; - outputTypes["pcoa"] = tempOutNames; + outputTypes["pca"] = tempOutNames; outputTypes["loadings"] = tempOutNames; } catch(exception& e) { @@ -39,8 +38,7 @@ PCACommand::PCACommand(){ //********************************************************************************************************************** vector PCACommand::getRequiredParameters(){ try { - string Array[] = {"phylip"}; - vector myArray (Array, Array+(sizeof(Array)/sizeof(string))); + vector myArray; return myArray; } catch(exception& e) { @@ -51,7 +49,8 @@ vector PCACommand::getRequiredParameters(){ //********************************************************************************************************************** vector PCACommand::getRequiredFiles(){ try { - vector myArray; + string Array[] = {"shared","relabund","or"}; + vector myArray (Array, Array+(sizeof(Array)/sizeof(string))); return myArray; } catch(exception& e) { @@ -63,14 +62,16 @@ vector PCACommand::getRequiredFiles(){ PCACommand::PCACommand(string option) { try { - abort = false; + abort = false; calledHelp = false; + + globaldata = GlobalData::getInstance(); //allow user to run help - if(option == "help") { help(); abort = true; } + if(option == "help") { help(); abort = true; calledHelp = true; } else { //valid paramters for this command - string Array[] = {"phylip","outputdir", "inputdir"}; + string Array[] = {"label","groups","metric","outputdir", "inputdir"}; vector myArray (Array, Array+(sizeof(Array)/sizeof(string))); OptionParser parser(option); @@ -78,46 +79,47 @@ PCACommand::PCACommand(string option) { ValidParameters validParameter; map::iterator it; - + //check to make sure all parameters are valid for command for (it = parameters.begin(); it != parameters.end(); it++) { if (validParameter.isValidParameter(it->first, myArray, it->second) != true) { abort = true; } } //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("phylip"); - //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["phylip"] = inputDir + it->second; } - } - } + string inputDir = validParameter.validFile(parameters, "inputdir", false); if (inputDir == "not found"){ inputDir = ""; } //initialize outputTypes vector tempOutNames; - outputTypes["pcoa"] = 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; } + //make sure the user has already run the read.otu command + if ((globaldata->getSharedFile() == "") && (globaldata->getRelAbundFile() == "")) { + m->mothurOut("You must read a list and a group, shared or relabund file before you can use the pca command."); m->mothurOutEndLine(); abort = true; + } + + if (globaldata->getSharedFile() != "") { mode = "shared"; inputFile = globaldata->getSharedFile(); } + if (globaldata->getRelAbundFile() != "") { mode = "relabund"; inputFile = globaldata->getRelAbundFile(); } //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(phylipfile); //if user entered a file with a path then preserve it + 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); + + label = validParameter.validFile(parameters, "label", false); + if (label == "not found") { label = ""; labels = globaldata->labels; 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); } + + groups = validParameter.validFile(parameters, "groups", false); + if (groups == "not found") { groups = ""; } + else { m->splitAtDash(groups, Groups); } + globaldata->Groups = Groups; - //error checking on files - if (phylipfile == "") { m->mothurOut("You must provide a distance file before running the pca command."); m->mothurOutEndLine(); abort = true; } } - + } catch(exception& e) { m->errorOut(e, "PCACommand", "PCACommand"); @@ -127,8 +129,14 @@ PCACommand::PCACommand(string option) { //********************************************************************************************************************** void PCACommand::help(){ try { - - m->mothurOut("The pca command..."); m->mothurOutEndLine(); + m->mothurOut("The pca command can only be run after a successful read.otu command of a shared or relabund file."); m->mothurOutEndLine(); + m->mothurOut("The pca command parameters are label, groups and metric. No parameters are required."); m->mothurOutEndLine(); + m->mothurOut("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"); + m->mothurOut("The groups parameter allows you to specify which groups you would like analyzed. Groupnames are separated by dashes.\n"); + m->mothurOut("The metric parameter allows indicate you if would like the pearson correlation coefficient calculated. Default=True"); m->mothurOutEndLine(); + m->mothurOut("Example pca(groups=yourGroups).\n"); + m->mothurOut("Example pca(groups=A-B-C).\n"); + m->mothurOut("Note: No spaces between parameter labels (i.e. groups), '=' and parameters (i.e.yourGroups).\n\n"); } catch(exception& e) { m->errorOut(e, "PCACommand", "help"); @@ -140,43 +148,99 @@ 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 names; - vector > D; - - fbase = outputDir + m->getRootName(m->getSimpleName(filename)); + //get first line of shared file + vector< vector > matrix; + InputData* input; + if (mode == "shared") { + 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 lookupFloat = input->getSharedRAbundFloatVectors(); + string lastLabel = lookupFloat[0]->getLabel(); + + set processedLabels; + set userLabels = labels; - read(filename, names, D); + //if the user gave no labels, then use the first one read + if (labels.size() == 0) { + label = lastLabel; + + process(lookupFloat); + } - if (m->control_pressed) { return 0; } - - double offset = 0.0000; - vector d; - vector e; - vector > G = D; - vector > copy_G; - //int rank = D.size(); + //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++) { remove(outputNames[i].c_str()); } 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(); + } - m->mothurOut("\nProcessing...\n"); - for(int count=0;count<2;count++){ - recenter(offset, D, G); if (m->control_pressed) { return 0; } - tred2(G, d, e); if (m->control_pressed) { return 0; } - qtli(d, e, G); if (m->control_pressed) { return 0; } - offset = d[d.size()-1]; - if(offset > 0.0) break; - } + if (m->control_pressed) { for (int i = 0; i < outputNames.size(); i++) { remove(outputNames[i].c_str()); } delete input; for (int i = 0; i < lookupFloat.size(); i++) { delete lookupFloat[i]; } lookupFloat.clear(); return 0; } - if (m->control_pressed) { return 0; } + //output error messages about any remaining user labels + set::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(); + } + + for (int i = 0; i < lookupFloat.size(); i++) { if (lookupFloat[i] != NULL) { delete lookupFloat[i]; } } lookupFloat.clear(); + delete input; if (m->control_pressed) { for (int i = 0; i < outputNames.size(); i++) { remove(outputNames[i].c_str()); } return 0; } @@ -192,369 +256,124 @@ int 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) { - m->errorOut(e, "PCACommand", "get_comment"); - exit(1); - } -} - -/*********************************************************************************************************************************/ - -int PCACommand::read_phylip(istream& f, int square_m, vector& name_list, vector >& d){ +/********************************************************************************************************************** +vector< vector > PCACommand::createMatrix(vector 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> name_list[i]; - // cout << i << "\t" << name_list[i] << endl; - for(int j=0;jcontrol_pressed) { return 0; } - - f >> d[i][j]; - if (d[i][j] == -0.0000) - d[i][j] = 0.0000; - } + vector< vector > 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> name_list[0]; - for(int i=1;i> name_list[i]; - d[i][i]=0.0000; - for(int j=0;jcontrol_pressed) { return 0; } - f >> d[i][j]; - if (d[i][j] == -0.0000) - d[i][j] = 0.0000; - d[j][i]=d[i][j]; - } + + vector< vector > 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]); } } - return 0; + matrix = linearCalc.matrix_mult(matrix, transposeMatrix); + + return matrix; } catch(exception& e) { - m->errorOut(e, "PCACommand", "read_phylip"); + m->errorOut(e, "PCACommand", "createMatrix"); exit(1); } - -} - -/*********************************************************************************************************************************/ - -void PCACommand::read(string fname, vector& names, vector >& D){ +}*/ +//********************************************************************************************************************** +int PCACommand::process(vector& lookupFloat){ try { - ifstream f; - m->openInputFile(fname, f); - - //check whether matrix is square - char d; - int q = 1; - int numSeqs; - string name; - - f >> numSeqs >> name; - - while((d=f.get()) != EOF){ - - //is d a number meaning its square - if(isalnum(d)){ - q = 1; - break; - } - - //is d a line return meaning its lower triangle - if(d == '\n'){ - q = 2; - break; - } - } - f.close(); + m->mothurOut("\nProcessing " + lookupFloat[0]->getLabel()); m->mothurOutEndLine(); - //reopen to get back to beginning - m->openInputFile(fname, f); - read_phylip(f, q, names, D); - } - catch(exception& e) { - m->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 > first, vector > second, vector >& product){ - try { - int first_rows = first.size(); - int first_cols = first[0].size(); - int second_cols = second[0].size(); + vector< vector > matrix; matrix.resize(lookupFloat.size()); - product.resize(first_rows); - for(int i=0;iopenOutputFile(temp, out); + out << "matrix" << endl; - for(int i=0;igetNumBins(); j++) { + matrix[i].push_back(lookupFloat[i]->getAbundance(j)); + out << lookupFloat[i]->getAbundance(j) << '\t'; } + out << endl; } - } - catch(exception& e) { - m->errorOut(e, "PCACommand", "matrix_mult"); - exit(1); - } - -} - -/*********************************************************************************************************************************/ - -void PCACommand::recenter(double offset, vector > D, vector >& G){ - try { - int rank = D.size(); - - vector > A(rank); - vector > C(rank); - for(int i=0;i > 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]); + out << matrix[j][i] << '\t'; + } + out << endl; } - double scale = -1.0000 / (double) rank; + matrix = linearCalc.matrix_mult(matrix, transposeMatrix); - for(int i=0;ierrorOut(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 >& a, vector& d, vector& e){ - try { - double scale, hh, h, g, f; - - int n = a.size(); - - d.resize(n); - e.resize(n); - - for(int i=n-1;i>0;i--){ - int l=i-1; - h = scale = 0.0000; - if(l>0){ - for(int k=0;k= 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 d; + vector e; + vector > G = matrix; + //vector > copy_G; - d[i] = h; - } + for(int count=0;count<2;count++){ + linearCalc.recenter(offset, matrix, G); if (m->control_pressed) { return 0; } + linearCalc.tred2(G, d, e); if (m->control_pressed) { return 0; } + linearCalc.qtli(d, e, G); if (m->control_pressed) { return 0; } + offset = d[d.size()-1]; + if(offset > 0.0) break; + } - d[0] = 0.0000; - e[0] = 0.0000; - - for(int i=0;ierrorOut(e, "PCACommand", "tred2"); - exit(1); - } - -} - -/*********************************************************************************************************************************/ - -// This function is taken from Numerical Recipes in C++ by Press et al., 2nd edition, pg. 479 - -void PCACommand::qtli(vector& d, vector& e, vector >& z) { - try { - int m, i, iter; - double s, r, p, g, f, dd, c, b; + if (m->control_pressed) { return 0; } - 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=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= l) continue; - d[l] -= p; - e[l] = g; - e[m] = 0.0; - } - } while (m != l); - } + string fbase = outputDir + m->getRootName(m->getSimpleName(inputFile)); + string outputFileName = fbase + lookupFloat[0]->getLabel(); + output(outputFileName, globaldata->Groups, G, d); - int k; - for(int i=0;i= p){ - p=d[k=j]; - } - } - if(k!=i){ - d[k]=d[i]; - d[i]=p; - for(int j=0;j > EuclidDists = linearCalc.calculateEuclidianDistance(G, i); //G is the pca file + + if (m->control_pressed) { for (int i = 0; i < outputNames.size(); i++) { remove(outputNames[i].c_str()); } return 0; } + + double corr = linearCalc.calcPearson(EuclidDists, matrix); //G is the pca file, D is the users distance matrix + + m->mothurOut("Pearson's coefficient using " + toString(i) + " axis: " + toString(corr)); m->mothurOutEndLine(); + + m->mothurOut("Rsq " + toString(i) + " axis: " + toString(corr * corr)); m->mothurOutEndLine(); + + if (m->control_pressed) { for (int i = 0; i < outputNames.size(); i++) { remove(outputNames[i].c_str()); } return 0; } } } + + return 0; } catch(exception& e) { - m->errorOut(e, "PCACommand", "qtli"); + m->errorOut(e, "PCACommand", "process"); exit(1); } } - /*********************************************************************************************************************************/ -void PCACommand::output(string fnameRoot, vector name_list, vector > G, vector d) { +void PCACommand::output(string fnameRoot, vector name_list, vector >& G, vector d) { try { int rank = name_list.size(); double dsum = 0.0000; @@ -566,17 +385,17 @@ void PCACommand::output(string fnameRoot, vector name_list, vector name_list, vector