X-Git-Url: https://git.donarmstrong.com/?p=mothur.git;a=blobdiff_plain;f=matrixoutputcommand.cpp;h=310e32bf23e048c73ae37579ac2e8e2d1cb8b0ff;hp=4b92c577fa3f5ef319fb8805eaef432e793c8d2e;hb=1a20e24ee786195ab0e1cccd4f5aede7a88f3f4e;hpb=49d2b7459c5027557564b21e9487dadafbbbdc96 diff --git a/matrixoutputcommand.cpp b/matrixoutputcommand.cpp index 4b92c57..310e32b 100644 --- a/matrixoutputcommand.cpp +++ b/matrixoutputcommand.cpp @@ -13,17 +13,17 @@ //********************************************************************************************************************** vector MatrixOutputCommand::setParameters(){ try { - CommandParameter pshared("shared", "InputTypes", "", "", "none", "none", "none",false,true); parameters.push_back(pshared); - CommandParameter plabel("label", "String", "", "", "", "", "",false,false); parameters.push_back(plabel); - CommandParameter psubsample("subsample", "String", "", "", "", "", "",false,false); parameters.push_back(psubsample); - CommandParameter pgroups("groups", "String", "", "", "", "", "",false,false); parameters.push_back(pgroups); - CommandParameter pcalc("calc", "Multiple", "sharedsobs-sharedchao-sharedace-jabund-sorabund-jclass-sorclass-jest-sorest-thetayc-thetan-kstest-sharednseqs-ochiai-anderberg-kulczynski-kulczynskicody-lennon-morisitahorn-braycurtis-whittaker-odum-canberra-structeuclidean-structchord-hellinger-manhattan-structpearson-soergel-spearman-structkulczynski-speciesprofile-hamming-structchi2-gower-memchi2-memchord-memeuclidean-mempearson", "jclass-thetayc", "", "", "",true,false); parameters.push_back(pcalc); - CommandParameter poutput("output", "Multiple", "lt-square", "lt", "", "", "",false,false); parameters.push_back(poutput); - CommandParameter pmode("mode", "Multiple", "average-median", "average", "", "", "",false,false); parameters.push_back(pmode); - CommandParameter pprocessors("processors", "Number", "", "1", "", "", "",false,false); parameters.push_back(pprocessors); - CommandParameter piters("iters", "Number", "", "1000", "", "", "",false,false); parameters.push_back(piters); - CommandParameter pinputdir("inputdir", "String", "", "", "", "", "",false,false); parameters.push_back(pinputdir); - CommandParameter poutputdir("outputdir", "String", "", "", "", "", "",false,false); parameters.push_back(poutputdir); + CommandParameter pshared("shared", "InputTypes", "", "", "none", "none", "none","phylip",false,true,true); parameters.push_back(pshared); + CommandParameter plabel("label", "String", "", "", "", "", "","",false,false); parameters.push_back(plabel); + CommandParameter psubsample("subsample", "String", "", "", "", "", "","",false,false); parameters.push_back(psubsample); + CommandParameter pgroups("groups", "String", "", "", "", "", "","",false,false); parameters.push_back(pgroups); + CommandParameter pcalc("calc", "Multiple", "sharedsobs-sharedchao-sharedace-jabund-sorabund-jclass-sorclass-jest-sorest-thetayc-thetan-kstest-sharednseqs-ochiai-anderberg-kulczynski-kulczynskicody-lennon-morisitahorn-braycurtis-whittaker-odum-canberra-structeuclidean-structchord-hellinger-manhattan-structpearson-soergel-spearman-structkulczynski-speciesprofile-hamming-structchi2-gower-memchi2-memchord-memeuclidean-mempearson", "jclass-thetayc", "", "", "","",true,false,true); parameters.push_back(pcalc); + CommandParameter poutput("output", "Multiple", "lt-square-column", "lt", "", "", "","",false,false); parameters.push_back(poutput); + CommandParameter pmode("mode", "Multiple", "average-median", "average", "", "", "","",false,false); parameters.push_back(pmode); + CommandParameter pprocessors("processors", "Number", "", "1", "", "", "","",false,false,true); parameters.push_back(pprocessors); + CommandParameter piters("iters", "Number", "", "1000", "", "", "","",false,false); parameters.push_back(piters); + CommandParameter pinputdir("inputdir", "String", "", "", "", "", "","",false,false); parameters.push_back(pinputdir); + CommandParameter poutputdir("outputdir", "String", "", "", "", "", "","",false,false); parameters.push_back(poutputdir); vector myArray; for (int i = 0; i < parameters.size(); i++) { myArray.push_back(parameters[i].name); } @@ -45,7 +45,7 @@ string MatrixOutputCommand::getHelpString(){ helpString += "The iters parameter allows you to choose the number of times you would like to run the subsample.\n"; helpString += "The subsample parameter allows you to enter the size pergroup of the sample or you can set subsample=T and mothur will use the size of your smallest group.\n"; helpString += "The dist.shared command should be in the following format: dist.shared(groups=yourGroups, calc=yourCalcs, label=yourLabels).\n"; - helpString += "The output parameter allows you to specify format of your distance matrix. Options are lt, and square. The default is lt.\n"; + helpString += "The output parameter allows you to specify format of your distance matrix. Options are lt, column and square. The default is lt.\n"; helpString += "The mode parameter allows you to specify if you want the average or the median values reported when subsampling. Options are average, and median. The default is average.\n"; helpString += "Example dist.shared(groups=A-B-C, calc=jabund-sorabund).\n"; helpString += "The default value for groups is all the groups in your groupfile.\n"; @@ -61,24 +61,19 @@ string MatrixOutputCommand::getHelpString(){ } } //********************************************************************************************************************** -string MatrixOutputCommand::getOutputFileNameTag(string type, string inputName=""){ - try { - string outputFileName = ""; - map >::iterator it; +string MatrixOutputCommand::getOutputPattern(string type) { + try { + string pattern = ""; - //is this a type this command creates - it = outputTypes.find(type); - if (it == outputTypes.end()) { m->mothurOut("[ERROR]: this command doesn't create a " + type + " output file.\n"); } - else { - if (type == "phylip") { outputFileName = "dist"; } - else { m->mothurOut("[ERROR]: No definition for type " + type + " output file tag.\n"); m->control_pressed = true; } - } - return outputFileName; - } - catch(exception& e) { - m->errorOut(e, "MatrixOutputCommand", "getOutputFileNameTag"); - exit(1); - } + if (type == "phylip") { pattern = "[filename],[calc],[distance],[outputtag],dist-[filename],[calc],[distance],[outputtag],[tag2],dist"; } + else { m->mothurOut("[ERROR]: No definition for type " + type + " output pattern.\n"); m->control_pressed = true; } + + return pattern; + } + catch(exception& e) { + m->errorOut(e, "MatrixOutputCommand", "getOutputPattern"); + exit(1); + } } //********************************************************************************************************************** MatrixOutputCommand::MatrixOutputCommand(){ @@ -161,7 +156,7 @@ MatrixOutputCommand::MatrixOutputCommand(string option) { } output = validParameter.validFile(parameters, "output", false); if(output == "not found"){ output = "lt"; } - if ((output != "lt") && (output != "square")) { m->mothurOut(output + " is not a valid output form. Options are lt and square. I will use lt."); m->mothurOutEndLine(); output = "lt"; } + if ((output != "lt") && (output != "square") && (output != "column")) { m->mothurOut(output + " is not a valid output form. Options are lt, column and square. I will use lt."); m->mothurOutEndLine(); output = "lt"; } mode = validParameter.validFile(parameters, "mode", false); if(mode == "not found"){ mode = "average"; } if ((mode != "average") && (mode != "median")) { m->mothurOut(mode + " is not a valid mode. Options are average and medina. I will use average."); m->mothurOutEndLine(); output = "average"; } @@ -199,7 +194,7 @@ MatrixOutputCommand::MatrixOutputCommand(string option) { else { subsample = false; } } - if (subsample == false) { iters = 1; } + if (subsample == false) { iters = 0; } if (abort == false) { @@ -454,11 +449,9 @@ void MatrixOutputCommand::printSims(ostream& out, vector< vector >& simM try { out.setf(ios::fixed, ios::floatfield); out.setf(ios::showpoint); - - //output num seqs - out << simMatrix.size() << endl; - + if (output == "lt") { + out << simMatrix.size() << endl; for (int b = 0; b < simMatrix.size(); b++) { out << lookup[b]->getGroup() << '\t'; for (int n = 0; n < b; n++) { @@ -466,8 +459,15 @@ void MatrixOutputCommand::printSims(ostream& out, vector< vector >& simM } out << endl; } + }else if (output == "column") { + for (int b = 0; b < simMatrix.size(); b++) { + for (int n = 0; n < b; n++) { + out << lookup[b]->getGroup() << '\t' << lookup[n]->getGroup() << '\t' << simMatrix[b][n] << endl; + } + } }else{ - for (int b = 0; b < simMatrix.size(); m++) { + out << simMatrix.size() << endl; + for (int b = 0; b < simMatrix.size(); b++) { out << lookup[b]->getGroup() << '\t'; for (int n = 0; n < simMatrix[b].size(); n++) { out << simMatrix[b][n] << '\t'; @@ -486,8 +486,12 @@ int MatrixOutputCommand::process(vector thisLookup){ try { vector< vector< vector > > calcDistsTotals; //each iter, one for each calc, then each groupCombos dists. this will be used to make .dist files vector< vector > calcDists; calcDists.resize(matrixCalculators.size()); - + for (int thisIter = 0; thisIter < iters+1; thisIter++) { + map variables; + variables["[filename]"] = outputDir + m->getRootName(m->getSimpleName(sharedfile)); + variables["[distance]"] = thisLookup[0]->getLabel(); + variables["[tag2]"] = ""; vector thisItersLookup = thisLookup; @@ -630,6 +634,9 @@ int MatrixOutputCommand::process(vector thisLookup){ //Close all thread handles and free memory allocations. for(int i=0; i < pDataArray.size(); i++){ + if (pDataArray[i]->count != (pDataArray[i]->end-pDataArray[i]->start)) { + m->mothurOut("[ERROR]: process " + toString(i) + " only processed " + toString(pDataArray[i]->count) + " of " + toString(pDataArray[i]->end-pDataArray[i]->start) + " groups assigned to it, quitting. \n"); m->control_pressed = true; + } for (int j = 0; j < pDataArray[i]->thisLookup.size(); j++) { delete pDataArray[i]->thisLookup[j]; } for (int k = 0; k < calcDists.size(); k++) { @@ -673,8 +680,11 @@ int MatrixOutputCommand::process(vector thisLookup){ matrix[column][row] = dist; } - string distFileName = outputDir + m->getRootName(m->getSimpleName(sharedfile)) + matrixCalculators[i]->getName() + "." + thisLookup[0]->getLabel() + "." + output + "." + getOutputFileNameTag("phylip"); + variables["[outputtag]"] = output; + variables["[calc]"] = matrixCalculators[i]->getName(); + string distFileName = getOutputFileName("phylip",variables); outputNames.push_back(distFileName); outputTypes["phylip"].push_back(distFileName); + ofstream outDist; m->openOutputFile(distFileName, outDist); outDist.setf(ios::fixed, ios::floatfield); outDist.setf(ios::showpoint); @@ -687,72 +697,12 @@ int MatrixOutputCommand::process(vector thisLookup){ for (int i = 0; i < calcDists.size(); i++) { calcDists[i].clear(); } } - if (iters != 1) { + if (iters != 0) { //we need to find the average distance and standard deviation for each groups distance + vector< vector > calcAverages = m->getAverages(calcDistsTotals, mode); - vector< vector > calcAverages; calcAverages.resize(matrixCalculators.size()); - for (int i = 0; i < calcAverages.size(); i++) { //initialize sums to zero. - calcAverages[i].resize(calcDistsTotals[0][i].size()); - - for (int j = 0; j < calcAverages[i].size(); j++) { - calcAverages[i][j].seq1 = calcDistsTotals[0][i][j].seq1; - calcAverages[i][j].seq2 = calcDistsTotals[0][i][j].seq2; - calcAverages[i][j].dist = 0.0; - } - } - if (mode == "average") { - for (int thisIter = 0; thisIter < iters; thisIter++) { //sum all groups dists for each calculator - for (int i = 0; i < calcAverages.size(); i++) { //initialize sums to zero. - for (int j = 0; j < calcAverages[i].size(); j++) { - calcAverages[i][j].dist += calcDistsTotals[thisIter][i][j].dist; - if (m->debug) { m->mothurOut("[DEBUG]: Totaling for average calc: iter = " + toString(thisIter) + ", " + thisLookup[calcDistsTotals[thisIter][i][j].seq1]->getGroup() + " - " + thisLookup[calcDistsTotals[thisIter][i][j].seq2]->getGroup() + " distance = " + toString(calcDistsTotals[thisIter][i][j].dist) + ". New total = " + toString(calcAverages[i][j].dist) + ".\n"); } - } - } - } - - for (int i = 0; i < calcAverages.size(); i++) { //finds average. - for (int j = 0; j < calcAverages[i].size(); j++) { - calcAverages[i][j].dist /= (float) iters; - } - } - }else { //find median - for (int i = 0; i < calcAverages.size(); i++) { //for each calc - for (int j = 0; j < calcAverages[i].size(); j++) { //for each comparison - vector dists; - for (int thisIter = 0; thisIter < iters; thisIter++) { //for each subsample - dists.push_back(calcDistsTotals[thisIter][i][j].dist); - } - sort(dists.begin(), dists.end()); - calcAverages[i][j].dist = dists[(iters/2)]; - } - } - } //find standard deviation - vector< vector > stdDev; stdDev.resize(matrixCalculators.size()); - for (int i = 0; i < stdDev.size(); i++) { //initialize sums to zero. - stdDev[i].resize(calcDistsTotals[0][i].size()); - - for (int j = 0; j < stdDev[i].size(); j++) { - stdDev[i][j].seq1 = calcDistsTotals[0][i][j].seq1; - stdDev[i][j].seq2 = calcDistsTotals[0][i][j].seq2; - stdDev[i][j].dist = 0.0; - } - } - - for (int thisIter = 0; thisIter < iters; thisIter++) { //compute the difference of each dist from the mean, and square the result of each - for (int i = 0; i < stdDev.size(); i++) { - for (int j = 0; j < stdDev[i].size(); j++) { - stdDev[i][j].dist += ((calcDistsTotals[thisIter][i][j].dist - calcAverages[i][j].dist) * (calcDistsTotals[thisIter][i][j].dist - calcAverages[i][j].dist)); - } - } - } - - for (int i = 0; i < stdDev.size(); i++) { //finds average. - for (int j = 0; j < stdDev[i].size(); j++) { - stdDev[i][j].dist /= (float) iters; - stdDev[i][j].dist = sqrt(stdDev[i][j].dist); - } - } + vector< vector > stdDev = m->getStandardDeviation(calcDistsTotals, calcAverages); //print results for (int i = 0; i < calcDists.size(); i++) { @@ -776,8 +726,14 @@ int MatrixOutputCommand::process(vector thisLookup){ stdmatrix[row][column] = stdDist; stdmatrix[column][row] = stdDist; } - - string distFileName = outputDir + m->getRootName(m->getSimpleName(sharedfile)) + matrixCalculators[i]->getName() + "." + thisLookup[0]->getLabel() + "." + output + ".ave." + getOutputFileNameTag("phylip"); + + map variables; + variables["[filename]"] = outputDir + m->getRootName(m->getSimpleName(sharedfile)); + variables["[distance]"] = thisLookup[0]->getLabel(); + variables["[outputtag]"] = output; + variables["[tag2]"] = "ave"; + variables["[calc]"] = matrixCalculators[i]->getName(); + string distFileName = getOutputFileName("phylip",variables); outputNames.push_back(distFileName); outputTypes["phylip"].push_back(distFileName); ofstream outAve; m->openOutputFile(distFileName, outAve); @@ -787,7 +743,8 @@ int MatrixOutputCommand::process(vector thisLookup){ outAve.close(); - distFileName = outputDir + m->getRootName(m->getSimpleName(sharedfile)) + matrixCalculators[i]->getName() + "." + thisLookup[0]->getLabel() + "." + output + ".std." + getOutputFileNameTag("phylip"); + variables["[tag2]"] = "std"; + distFileName = getOutputFileName("phylip",variables); outputNames.push_back(distFileName); outputTypes["phylip"].push_back(distFileName); ofstream outSTD; m->openOutputFile(distFileName, outSTD);