#include "sharedbraycurtis.h"
#include "sharedjackknife.h"
#include "whittaker.h"
+#include "odum.h"
+#include "canberra.h"
+#include "structeuclidean.h"
+#include "structchord.h"
+#include "hellinger.h"
+#include "manhattan.h"
+#include "structpearson.h"
+#include "soergel.h"
+#include "spearman.h"
+#include "structkulczynski.h"
+#include "structchi2.h"
+#include "speciesprofile.h"
+#include "hamming.h"
+#include "gower.h"
+#include "memchi2.h"
+#include "memchord.h"
+#include "memeuclidean.h"
+#include "mempearson.h"
-
+//**********************************************************************************************************************
+vector<string> SummarySharedCommand::getValidParameters(){
+ try {
+ string Array[] = {"label","calc","groups","all","outputdir","distance","inputdir", "processors"};
+ vector<string> myArray (Array, Array+(sizeof(Array)/sizeof(string)));
+ return myArray;
+ }
+ catch(exception& e) {
+ m->errorOut(e, "SummarySharedCommand", "getValidParameters");
+ exit(1);
+ }
+}
+//**********************************************************************************************************************
+SummarySharedCommand::SummarySharedCommand(){
+ try {
+ abort = true;
+ //initialize outputTypes
+ vector<string> tempOutNames;
+ outputTypes["summary"] = tempOutNames;
+ }
+ catch(exception& e) {
+ m->errorOut(e, "SummarySharedCommand", "SummarySharedCommand");
+ exit(1);
+ }
+}
+//**********************************************************************************************************************
+vector<string> SummarySharedCommand::getRequiredParameters(){
+ try {
+ vector<string> myArray;
+ return myArray;
+ }
+ catch(exception& e) {
+ m->errorOut(e, "SummarySharedCommand", "getRequiredParameters");
+ exit(1);
+ }
+}
+//**********************************************************************************************************************
+vector<string> SummarySharedCommand::getRequiredFiles(){
+ try {
+ string Array[] = {"shared"};
+ vector<string> myArray (Array, Array+(sizeof(Array)/sizeof(string)));
+ return myArray;
+ }
+ catch(exception& e) {
+ m->errorOut(e, "SummarySharedCommand", "getRequiredFiles");
+ exit(1);
+ }
+}
//**********************************************************************************************************************
SummarySharedCommand::SummarySharedCommand(string option) {
else {
//valid paramters for this command
- string Array[] = {"label","calc","groups","all","outputdir","inputdir", "processors"};
+ string Array[] = {"label","calc","groups","all","outputdir","distance","inputdir", "processors"};
vector<string> myArray (Array, Array+(sizeof(Array)/sizeof(string)));
OptionParser parser(option);
m->mothurOut("You must read a list and a group, or a shared before you can use the summary.shared command."); m->mothurOutEndLine(); abort = true;
}
+ //initialize outputTypes
+ vector<string> tempOutNames;
+ outputTypes["summary"] = tempOutNames;
+
//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 = "";
string temp = validParameter.validFile(parameters, "all", false); if (temp == "not found") { temp = "false"; }
all = m->isTrue(temp);
+ temp = validParameter.validFile(parameters, "distance", false); if (temp == "not found") { temp = "false"; }
+ createPhylip = m->isTrue(temp);
+
temp = validParameter.validFile(parameters, "processors", false); if(temp == "not found"){ temp = "1"; }
convert(temp, processors);
sumCalculators.push_back(new BrayCurtis());
}else if (Estimators[i] == "whittaker") {
sumCalculators.push_back(new Whittaker());
+ }else if (Estimators[i] == "odum") {
+ sumCalculators.push_back(new Odum());
+ }else if (Estimators[i] == "canberra") {
+ sumCalculators.push_back(new Canberra());
+ }else if (Estimators[i] == "structeuclidean") {
+ sumCalculators.push_back(new StructEuclidean());
+ }else if (Estimators[i] == "structchord") {
+ sumCalculators.push_back(new StructChord());
+ }else if (Estimators[i] == "hellinger") {
+ sumCalculators.push_back(new Hellinger());
+ }else if (Estimators[i] == "manhattan") {
+ sumCalculators.push_back(new Manhattan());
+ }else if (Estimators[i] == "structpearson") {
+ sumCalculators.push_back(new StructPearson());
+ }else if (Estimators[i] == "soergel") {
+ sumCalculators.push_back(new Soergel());
+ }else if (Estimators[i] == "spearman") {
+ sumCalculators.push_back(new Spearman());
+ }else if (Estimators[i] == "structkulczynski") {
+ sumCalculators.push_back(new StructKulczynski());
+ }else if (Estimators[i] == "speciesprofile") {
+ sumCalculators.push_back(new SpeciesProfile());
+ }else if (Estimators[i] == "hamming") {
+ sumCalculators.push_back(new Hamming());
+ }else if (Estimators[i] == "structchi2") {
+ sumCalculators.push_back(new StructChi2());
+ }else if (Estimators[i] == "gower") {
+ sumCalculators.push_back(new Gower());
+ }else if (Estimators[i] == "memchi2") {
+ sumCalculators.push_back(new MemChi2());
+ }else if (Estimators[i] == "memchord") {
+ sumCalculators.push_back(new MemChord());
+ }else if (Estimators[i] == "memeuclidean") {
+ sumCalculators.push_back(new MemEuclidean());
+ }else if (Estimators[i] == "mempearson") {
+ sumCalculators.push_back(new MemPearson());
}
}
}
void SummarySharedCommand::help(){
try {
m->mothurOut("The summary.shared command can only be executed after a successful read.otu command.\n");
- m->mothurOut("The summary.shared command parameters are label, calc and all. No parameters are required.\n");
+ m->mothurOut("The summary.shared command parameters are label, calc, distance and all. No parameters are required.\n");
m->mothurOut("The summary.shared command should be in the following format: \n");
m->mothurOut("summary.shared(label=yourLabel, calc=yourEstimators, groups=yourGroups).\n");
m->mothurOut("Example summary.shared(label=unique-.01-.03, groups=B-C, calc=sharedchao-sharedace-jabund-sorensonabund-jclass-sorclass-jest-sorest-thetayc-thetan).\n");
validCalculator->printCalc("sharedsummary", cout);
m->mothurOut("The default value for calc is sharedsobs-sharedchao-sharedace-jabund-sorensonabund-jclass-sorclass-jest-sorest-thetayc-thetan\n");
m->mothurOut("The default value for groups is all the groups in your groupfile.\n");
+ m->mothurOut("The distance parameter allows you to indicate you would like a distance file created for each calculator for each label, default=f.\n");
m->mothurOut("The label parameter is used to analyze specific labels in your input.\n");
m->mothurOut("The all parameter is used to specify if you want the estimate of all your groups together. This estimate can only be made for sharedsobs and sharedchao calculators. The default is false.\n");
m->mothurOut("If you use sharedchao and run into memory issues, set all to false. \n");
input = globaldata->ginput;
lookup = input->getSharedRAbundVectors();
string lastLabel = lookup[0]->getLabel();
-
+
/******************************************************/
//output headings for files
/******************************************************/
m->mothurOutEndLine();
m->mothurOut("Output File Names: "); m->mothurOutEndLine();
m->mothurOut(outputFileName); m->mothurOutEndLine();
- if (mult) { m->mothurOut(outAllFileName); m->mothurOutEndLine(); }
+ if (mult) { m->mothurOut(outAllFileName); m->mothurOutEndLine(); outputTypes["summary"].push_back(outAllFileName); }
+ for (int i = 0; i < outputNames.size(); i++) { m->mothurOut(outputNames[i]); m->mothurOutEndLine(); } outputTypes["summary"].push_back(outputFileName);
m->mothurOutEndLine();
return 0;
/***********************************************************/
int SummarySharedCommand::process(vector<SharedRAbundVector*> thisLookup, string sumFileName, string sumAllFileName) {
try {
+ vector< vector<seqDist> > calcDists; //vector containing vectors that contains the summary results for each group compare
+ calcDists.resize(sumCalculators.size()); //one for each calc, this will be used to make .dist files
#if defined (__APPLE__) || (__MACH__) || (linux) || (__linux)
if(processors == 1){
- driver(thisLookup, 0, numGroups, sumFileName, sumAllFileName);
+ driver(thisLookup, 0, numGroups, sumFileName+".temp", sumAllFileName+".temp", calcDists);
m->appendFiles((sumFileName + ".temp"), sumFileName);
remove((sumFileName + ".temp").c_str());
if (mult) {
remove((sumAllFileName + ".temp").c_str());
}
}else{
- int process = 0;
+ int process = 1;
vector<int> processIDS;
//loop through and create all the processes you want
processIDS.push_back(pid);
process++;
}else if (pid == 0){
- driver(thisLookup, lines[process].start, lines[process].end, sumFileName + toString(getpid()) + ".temp", sumAllFileName + toString(getpid()) + ".temp");
+ driver(thisLookup, lines[process].start, lines[process].end, sumFileName + toString(getpid()) + ".temp", sumAllFileName + toString(getpid()) + ".temp", calcDists);
+
+ //only do this if you want a distance file
+ if (createPhylip) {
+ string tempdistFileName = m->getRootName(m->getSimpleName(sumFileName)) + toString(getpid()) + ".dist";
+ ofstream outtemp;
+ m->openOutputFile(tempdistFileName, outtemp);
+
+ for (int i = 0; i < calcDists.size(); i++) {
+ outtemp << calcDists[i].size() << endl;
+
+ for (int j = 0; j < calcDists[i].size(); j++) {
+ outtemp << calcDists[i][j].seq1 << '\t' << calcDists[i][j].seq2 << '\t' << calcDists[i][j].dist << endl;
+ }
+ }
+ outtemp.close();
+ }
+
+ exit(0);
+ }else {
+ m->mothurOut("[ERROR]: unable to spawn the necessary processes."); m->mothurOutEndLine();
+ for (int i = 0; i < processIDS.size(); i++) { kill (processIDS[i], SIGINT); }
exit(0);
- }else { m->mothurOut("unable to spawn the necessary processes."); m->mothurOutEndLine(); exit(0); }
+ }
}
-
+
+ //parent do your part
+ driver(thisLookup, lines[0].start, lines[0].end, sumFileName + toString(getpid()) + ".temp", sumAllFileName + toString(getpid()) + ".temp", calcDists);
+ m->appendFiles((sumFileName + toString(getpid()) + ".temp"), sumFileName);
+ remove((sumFileName + toString(getpid()) + ".temp").c_str());
+ if (mult) { m->appendFiles((sumAllFileName + toString(getpid()) + ".temp"), sumAllFileName); }
+
//force parent to wait until all the processes are done
for (int i = 0; i < processIDS.size(); i++) {
int temp = processIDS[i];
for (int i = 0; i < processIDS.size(); i++) {
m->appendFiles((sumFileName + toString(processIDS[i]) + ".temp"), sumFileName);
remove((sumFileName + toString(processIDS[i]) + ".temp").c_str());
- if (mult) {
- if (i == 0) { m->appendFiles((sumAllFileName + toString(processIDS[i]) + ".temp"), sumAllFileName); }
- remove((sumAllFileName + toString(processIDS[i]) + ".temp").c_str());
+ if (mult) { remove((sumAllFileName + toString(processIDS[i]) + ".temp").c_str()); }
+
+ if (createPhylip) {
+ string tempdistFileName = m->getRootName(m->getSimpleName(sumFileName)) + toString(processIDS[i]) + ".dist";
+ ifstream intemp;
+ m->openInputFile(tempdistFileName, intemp);
+
+ for (int i = 0; i < calcDists.size(); i++) {
+ int size = 0;
+ intemp >> size; m->gobble(intemp);
+
+ for (int j = 0; j < size; j++) {
+ int seq1 = 0;
+ int seq2 = 0;
+ float dist = 1.0;
+
+ intemp >> seq1 >> seq2 >> dist; m->gobble(intemp);
+
+ seqDist tempDist(seq1, seq2, dist);
+ calcDists[i].push_back(tempDist);
+ }
+ }
+ intemp.close();
+ remove(tempdistFileName.c_str());
}
}
}
#else
- driver(thisLookup, 0, numGroups, (sumFileName + ".temp"), (sumAllFileName + ".temp"));
+ driver(thisLookup, 0, numGroups, (sumFileName + ".temp"), (sumAllFileName + ".temp"), calcDists);
m->appendFiles((sumFileName + ".temp"), sumFileName);
remove((sumFileName + ".temp").c_str());
if (mult) {
remove((sumAllFileName + ".temp").c_str());
}
#endif
+
+ if (createPhylip) {
+ for (int i = 0; i < calcDists.size(); i++) {
+ if (m->control_pressed) { break; }
+
+ string distFileName = outputDir + m->getRootName(m->getSimpleName(sumFileName)) + sumCalculators[i]->getName() + "." + thisLookup[0]->getLabel() + ".dist";
+ outputNames.push_back(distFileName);
+ ofstream outDist;
+ m->openOutputFile(distFileName, outDist);
+ outDist.setf(ios::fixed, ios::floatfield); outDist.setf(ios::showpoint);
+
+ //initialize matrix
+ vector< vector<float> > matrix; //square matrix to represent the distance
+ matrix.resize(thisLookup.size());
+ for (int k = 0; k < thisLookup.size(); k++) { matrix[k].resize(thisLookup.size(), 0.0); }
+
+
+ for (int j = 0; j < calcDists[i].size(); j++) {
+ int row = calcDists[i][j].seq1;
+ int column = calcDists[i][j].seq2;
+ float dist = calcDists[i][j].dist;
+
+ matrix[row][column] = dist;
+ matrix[column][row] = dist;
+ }
+
+ //output to file
+ outDist << thisLookup.size() << endl;
+ for (int r=0; r<thisLookup.size(); r++) {
+ //output name
+ string name = thisLookup[r]->getGroup();
+ if (name.length() < 10) { //pad with spaces to make compatible
+ while (name.length() < 10) { name += " "; }
+ }
+ outDist << name << '\t';
+
+ //output distances
+ for (int l = 0; l < r; l++) { outDist << matrix[r][l] << '\t'; }
+ outDist << endl;
+ }
+
+ outDist.close();
+ }
+ }
+ return 0;
}
catch(exception& e) {
m->errorOut(e, "SummarySharedCommand", "process");
}
}
/**************************************************************************************************/
-int SummarySharedCommand::driver(vector<SharedRAbundVector*> thisLookup, int start, int end, string sumFile, string sumAllFile) {
+int SummarySharedCommand::driver(vector<SharedRAbundVector*> thisLookup, int start, int end, string sumFile, string sumAllFile, vector< vector<seqDist> >& calcDists) {
try {
//loop through calculators and add to file all for all calcs that can do mutiple groups
}
for(int i=0;i<sumCalculators.size();i++) {
-
- sumCalculators[i]->getValues(subset); //saves the calculator outputs
+
+ //if this calc needs all groups to calculate the pair load all groups
+ if (sumCalculators[i]->getNeedsAll()) {
+ //load subset with rest of lookup for those calcs that need everyone to calc for a pair
+ for (int w = 0; w < thisLookup.size(); w++) {
+ if ((w != k) && (w != l)) { subset.push_back(thisLookup[w]); }
+ }
+ }
+
+ vector<double> tempdata = sumCalculators[i]->getValues(subset); //saves the calculator outputs
if (m->control_pressed) { outputFileHandle.close(); return 1; }
outputFileHandle << '\t';
sumCalculators[i]->print(outputFileHandle);
+
+ seqDist temp(l, k, tempdata[0]);
+ calcDists[i].push_back(temp);
}
outputFileHandle << endl;
}