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", "lt", "", "", "","",false,false); parameters.push_back(poutput);
+ 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);
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";
}
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"; }
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++) {
}
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{
+ 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++) {
//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++) {
if (iters != 0) {
//we need to find the average distance and standard deviation for each groups distance
+ vector< vector<seqDist> > calcAverages = m->getAverages(calcDistsTotals, mode);
- vector< vector<seqDist> > 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<double> 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<seqDist> > 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<seqDist> > stdDev = m->getStandardDeviation(calcDistsTotals, calcAverages);
//print results
for (int i = 0; i < calcDists.size(); i++) {
outAve.close();
variables["[tag2]"] = "std";
- distFileName = outputDir + getOutputFileName("phylip",variables);
+ distFileName = getOutputFileName("phylip",variables);
outputNames.push_back(distFileName); outputTypes["phylip"].push_back(distFileName);
ofstream outSTD;
m->openOutputFile(distFileName, outSTD);
projects / mothur.git / blobdiff