#include "subsamplecommand.h"
#include "sharedutilities.h"
+#include "deconvolutecommand.h"
+#include "subsample.h"
//**********************************************************************************************************************
-vector<string> SubSampleCommand::getValidParameters(){
- try {
- string Array[] = {"fasta", "group", "list","shared","rabund","persample", "name","sabund","size","groups","label","outputdir","inputdir"};
- vector<string> myArray (Array, Array+(sizeof(Array)/sizeof(string)));
+vector<string> SubSampleCommand::setParameters(){
+ try {
+ CommandParameter pfasta("fasta", "InputTypes", "", "", "none", "FLSSR", "none",false,false); parameters.push_back(pfasta);
+ CommandParameter pname("name", "InputTypes", "", "", "none", "none", "none",false,false); parameters.push_back(pname);
+ CommandParameter pgroup("group", "InputTypes", "", "", "none", "none", "none",false,false); parameters.push_back(pgroup);
+ CommandParameter plist("list", "InputTypes", "", "", "none", "FLSSR", "none",false,false); parameters.push_back(plist);
+ CommandParameter pshared("shared", "InputTypes", "", "", "none", "FLSSR", "none",false,false); parameters.push_back(pshared);
+ CommandParameter prabund("rabund", "InputTypes", "", "", "none", "FLSSR", "none",false,false); parameters.push_back(prabund);
+ CommandParameter psabund("sabund", "InputTypes", "", "", "none", "FLSSR", "none",false,false); parameters.push_back(psabund);
+ CommandParameter plabel("label", "String", "", "", "", "", "",false,false); parameters.push_back(plabel);
+ CommandParameter pgroups("groups", "String", "", "", "", "", "",false,false); parameters.push_back(pgroups);
+ CommandParameter psize("size", "Number", "", "0", "", "", "",false,false); parameters.push_back(psize);
+ CommandParameter ppersample("persample", "Boolean", "", "F", "", "", "",false,false); parameters.push_back(ppersample);
+ 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, "SubSampleCommand", "getValidParameters");
+ m->errorOut(e, "SubSampleCommand", "setParameters");
exit(1);
}
}
//**********************************************************************************************************************
-SubSampleCommand::SubSampleCommand(){
+string SubSampleCommand::getHelpString(){
try {
- abort = true; calledHelp = true;
- vector<string> tempOutNames;
- outputTypes["shared"] = tempOutNames;
- outputTypes["list"] = tempOutNames;
- outputTypes["rabund"] = tempOutNames;
- outputTypes["sabund"] = tempOutNames;
- outputTypes["fasta"] = tempOutNames;
- outputTypes["name"] = tempOutNames;
- outputTypes["group"] = tempOutNames;
+ string helpString = "";
+ helpString += "The sub.sample command is designed to be used as a way to normalize your data, or create a smaller set from your original set.\n";
+ helpString += "The sub.sample command parameters are fasta, name, list, group, rabund, sabund, shared, groups, size, persample and label. You must provide a fasta, list, sabund, rabund or shared file as an input file.\n";
+ helpString += "The namefile is only used with the fasta file, not with the listfile, because the list file should contain all sequences.\n";
+ helpString += "The groups parameter allows you to specify which of the groups in your groupfile you would like included. The group names are separated by dashes.\n";
+ helpString += "The label parameter allows you to select what distance levels you would like, and are also separated by dashes.\n";
+ helpString += "The size parameter allows you indicate the size of your subsample.\n";
+ helpString += "The persample parameter allows you indicate you want to select subsample of the same size from each of your groups, default=false. It is only used with the list and fasta files if a groupfile is given.\n";
+ helpString += "persample=false will select a random set of sequences of the size you select, but the number of seqs from each group may differ.\n";
+ helpString += "The size parameter is not set: with shared file size=number of seqs in smallest sample, with all other files if a groupfile is given and persample=true, then size=number of seqs in smallest sample, otherwise size=10% of number of seqs.\n";
+ helpString += "The sub.sample command should be in the following format: sub.sample(list=yourListFile, group=yourGroupFile, groups=yourGroups, label=yourLabels).\n";
+ helpString += "Example sub.sample(list=abrecovery.fn.list, group=abrecovery.groups, groups=B-C, size=20).\n";
+ helpString += "The default value for groups is all the groups in your groupfile, and all labels in your inputfile will be used.\n";
+ helpString += "The sub.sample command outputs a .subsample file.\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, "SubSampleCommand", "GetRelAbundCommand");
+ m->errorOut(e, "SubSampleCommand", "getHelpString");
exit(1);
}
}
//**********************************************************************************************************************
-vector<string> SubSampleCommand::getRequiredParameters(){
+string SubSampleCommand::getOutputFileNameTag(string type, string inputName=""){
try {
- string Array[] = {"fasta","list","shared","rabund", "sabund","or"};
- vector<string> myArray (Array, Array+(sizeof(Array)/sizeof(string)));
- return myArray;
+ string outputFileName = "";
+ map<string, vector<string> >::iterator it;
+
+ //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 == "fasta") { outputFileName = "subsample" + m->getExtension(inputName); }
+ else if (type == "sabund") { outputFileName = "subsample" + m->getExtension(inputName); }
+ else if (type == "name") { outputFileName = "subsample" + m->getExtension(inputName); }
+ else if (type == "group") { outputFileName = "subsample" + m->getExtension(inputName); }
+ else if (type == "list") { outputFileName = "subsample" + m->getExtension(inputName); }
+ else if (type == "rabund") { outputFileName = "subsample" + m->getExtension(inputName); }
+ else if (type == "shared") { outputFileName = "subsample" + m->getExtension(inputName); }
+ 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, "SubSampleCommand", "getRequiredParameters");
+ m->errorOut(e, "SubSampleCommand", "getOutputFileNameTag");
exit(1);
}
}
+
//**********************************************************************************************************************
-vector<string> SubSampleCommand::getRequiredFiles(){
+SubSampleCommand::SubSampleCommand(){
try {
- vector<string> myArray;
- return myArray;
+ abort = true; calledHelp = true;
+ setParameters();
+ vector<string> tempOutNames;
+ outputTypes["shared"] = tempOutNames;
+ outputTypes["list"] = tempOutNames;
+ outputTypes["rabund"] = tempOutNames;
+ outputTypes["sabund"] = tempOutNames;
+ outputTypes["fasta"] = tempOutNames;
+ outputTypes["name"] = tempOutNames;
+ outputTypes["group"] = tempOutNames;
}
catch(exception& e) {
- m->errorOut(e, "SubSampleCommand", "getRequiredFiles");
+ m->errorOut(e, "SubSampleCommand", "GetRelAbundCommand");
exit(1);
}
}
//**********************************************************************************************************************
SubSampleCommand::SubSampleCommand(string option) {
try {
- globaldata = GlobalData::getInstance();
abort = false; calledHelp = false;
allLines = 1;
- labels.clear();
//allow user to run help
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[] = {"fasta", "group", "list","shared","rabund","persample", "sabund","name","size","groups","label","outputdir","inputdir"};
- vector<string> myArray (Array, Array+(sizeof(Array)/sizeof(string)));
+ vector<string> myArray = setParameters();
OptionParser parser(option);
map<string,string> parameters = parser.getParameters();
//check for required parameters
listfile = validParameter.validFile(parameters, "list", true);
if (listfile == "not open") { listfile = ""; abort = true; }
- else if (listfile == "not found") { listfile = ""; }
+ else if (listfile == "not found") { listfile = ""; }
+ else { m->setListFile(listfile); }
sabundfile = validParameter.validFile(parameters, "sabund", true);
if (sabundfile == "not open") { sabundfile = ""; abort = true; }
else if (sabundfile == "not found") { sabundfile = ""; }
+ else { m->setSabundFile(sabundfile); }
rabundfile = validParameter.validFile(parameters, "rabund", true);
if (rabundfile == "not open") { rabundfile = ""; abort = true; }
else if (rabundfile == "not found") { rabundfile = ""; }
+ else { m->setRabundFile(rabundfile); }
fastafile = validParameter.validFile(parameters, "fasta", true);
if (fastafile == "not open") { fastafile = ""; abort = true; }
else if (fastafile == "not found") { fastafile = ""; }
+ else { m->setFastaFile(fastafile); }
sharedfile = validParameter.validFile(parameters, "shared", true);
if (sharedfile == "not open") { sharedfile = ""; abort = true; }
else if (sharedfile == "not found") { sharedfile = ""; }
+ else { m->setSharedFile(sharedfile); }
namefile = validParameter.validFile(parameters, "name", true);
if (namefile == "not open") { namefile = ""; abort = true; }
else if (namefile == "not found") { namefile = ""; }
+ else { m->setNameFile(namefile); }
groupfile = validParameter.validFile(parameters, "group", true);
if (groupfile == "not open") { groupfile = ""; abort = true; }
else if (groupfile == "not found") { groupfile = ""; }
-
+ else { m->setGroupFile(groupfile); }
//check for optional parameter and set defaults
// ...at some point should added some additional type checking...
else {
pickedGroups = true;
m->splitAtDash(groups, Groups);
- globaldata->Groups = Groups;
+ m->setGroups(Groups);
}
string temp = validParameter.validFile(parameters, "size", false); if (temp == "not found"){ temp = "0"; }
- convert(temp, size);
+ m->mothurConvert(temp, size);
temp = validParameter.validFile(parameters, "persample", false); if (temp == "not found"){ temp = "f"; }
persample = m->isTrue(temp);
if ((groupfile != "") && ((fastafile != "") && (listfile != ""))) {
m->mothurOut("A new group file can only be made from the subsample of a listfile or fastafile, not both. Please correct."); m->mothurOutEndLine(); abort = true; }
+ if ((fastafile != "") && (namefile == "")) {
+ vector<string> files; files.push_back(fastafile);
+ parser.getNameFile(files);
+ }
}
}
exit(1);
}
}
-
-//**********************************************************************************************************************
-
-void SubSampleCommand::help(){
- try {
- m->mothurOut("The sub.sample command is designed to be used as a way to normalize your data, or create a smaller set from your original set.\n");
- m->mothurOut("The sub.sample command parameters are fasta, name, list, group, rabund, sabund, shared, groups, size, persample and label. You must provide a fasta, list, sabund, rabund or shared file as an input file.\n");
- m->mothurOut("The namefile is only used with the fasta file, not with the listfile, because the list file should contain all sequences.\n");
- m->mothurOut("The groups parameter allows you to specify which of the groups in your groupfile you would like included. The group names are separated by dashes.\n");
- m->mothurOut("The label parameter allows you to select what distance levels you would like, and are also separated by dashes.\n");
- m->mothurOut("The size parameter allows you indicate the size of your subsample.\n");
- m->mothurOut("The persample parameter allows you indicate you want to select subsample of the same size from each of your groups, default=false. It is only used with the list and fasta files if a groupfile is given.\n");
- m->mothurOut("persample=false will select a random set of sequences of the size you select, but the number of seqs from each group may differ.\n");
- m->mothurOut("The size parameter is not set: with shared file size=number of seqs in smallest sample, with all other files if a groupfile is given and persample=true, then size=number of seqs in smallest sample, otherwise size=10% of number of seqs.\n");
- m->mothurOut("The sub.sample command should be in the following format: sub.sample(list=yourListFile, group=yourGroupFile, groups=yourGroups, label=yourLabels).\n");
- m->mothurOut("Example sub.sample(list=abrecovery.fn.list, group=abrecovery.groups, groups=B-C, size=20).\n");
- m->mothurOut("The default value for groups is all the groups in your groupfile, and all labels in your inputfile will be used.\n");
- m->mothurOut("The sub.sample command outputs a .subsample file.\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, "SubSampleCommand", "help");
- exit(1);
- }
-}
-
-//**********************************************************************************************************************
-
-SubSampleCommand::~SubSampleCommand(){}
-
//**********************************************************************************************************************
int SubSampleCommand::execute(){
if (abort == true) { if (calledHelp) { return 0; } return 2; }
if (sharedfile != "") { getSubSampleShared(); }
- if (m->control_pressed) { for (int i = 0; i < outputNames.size(); i++) { remove(outputNames[i].c_str()); return 0; } }
+ if (m->control_pressed) { for (int i = 0; i < outputNames.size(); i++) { m->mothurRemove(outputNames[i]); return 0; } }
if (listfile != "") { getSubSampleList(); }
- if (m->control_pressed) { for (int i = 0; i < outputNames.size(); i++) { remove(outputNames[i].c_str()); return 0; } }
+ if (m->control_pressed) { for (int i = 0; i < outputNames.size(); i++) { m->mothurRemove(outputNames[i]); return 0; } }
if (rabundfile != "") { getSubSampleRabund(); }
- if (m->control_pressed) { for (int i = 0; i < outputNames.size(); i++) { remove(outputNames[i].c_str()); return 0; } }
+ if (m->control_pressed) { for (int i = 0; i < outputNames.size(); i++) { m->mothurRemove(outputNames[i]); return 0; } }
if (sabundfile != "") { getSubSampleSabund(); }
- if (m->control_pressed) { for (int i = 0; i < outputNames.size(); i++) { remove(outputNames[i].c_str()); return 0; } }
+ if (m->control_pressed) { for (int i = 0; i < outputNames.size(); i++) { m->mothurRemove(outputNames[i]); return 0; } }
if (fastafile != "") { getSubSampleFasta(); }
- if (m->control_pressed) { for (int i = 0; i < outputNames.size(); i++) { remove(outputNames[i].c_str()); return 0; } }
+ if (m->control_pressed) { for (int i = 0; i < outputNames.size(); i++) { m->mothurRemove(outputNames[i]); return 0; } }
-
+ //set fasta file as new current fastafile
+ string current = "";
+ itTypes = outputTypes.find("fasta");
+ if (itTypes != outputTypes.end()) {
+ if ((itTypes->second).size() != 0) { current = (itTypes->second)[0]; m->setFastaFile(current); }
+ }
+
+ itTypes = outputTypes.find("name");
+ if (itTypes != outputTypes.end()) {
+ if ((itTypes->second).size() != 0) { current = (itTypes->second)[0]; m->setNameFile(current); }
+ }
+
+ itTypes = outputTypes.find("group");
+ if (itTypes != outputTypes.end()) {
+ if ((itTypes->second).size() != 0) { current = (itTypes->second)[0]; m->setGroupFile(current); }
+ }
+
+ itTypes = outputTypes.find("list");
+ if (itTypes != outputTypes.end()) {
+ if ((itTypes->second).size() != 0) { current = (itTypes->second)[0]; m->setListFile(current); }
+ }
+
+ itTypes = outputTypes.find("shared");
+ if (itTypes != outputTypes.end()) {
+ if ((itTypes->second).size() != 0) { current = (itTypes->second)[0]; m->setSharedFile(current); }
+ }
+
+ itTypes = outputTypes.find("rabund");
+ if (itTypes != outputTypes.end()) {
+ if ((itTypes->second).size() != 0) { current = (itTypes->second)[0]; m->setRabundFile(current); }
+ }
+
+ itTypes = outputTypes.find("sabund");
+ if (itTypes != outputTypes.end()) {
+ if ((itTypes->second).size() != 0) { current = (itTypes->second)[0]; m->setSabundFile(current); }
+ }
+
+
m->mothurOutEndLine();
m->mothurOut("Output File Names: "); m->mothurOutEndLine();
for (int i = 0; i < outputNames.size(); i++) { m->mothurOut(outputNames[i]); m->mothurOutEndLine(); }
//takes care of user setting groupNames that are invalid or setting groups=all
SharedUtil* util = new SharedUtil();
- util->setGroups(Groups, groupMap->namesOfGroups);
+ vector<string> namesGroups = groupMap->getNamesOfGroups();
+ util->setGroups(Groups, namesGroups);
delete util;
//file mismatch quit
if (m->control_pressed) { return 0; }
- string thisOutputDir = outputDir;
- if (outputDir == "") { thisOutputDir += m->hasPath(fastafile); }
- string outputFileName = thisOutputDir + m->getRootName(m->getSimpleName(fastafile)) + "subsample" + m->getExtension(fastafile);
-
- ofstream out;
- m->openOutputFile(outputFileName, out);
- outputTypes["fasta"].push_back(outputFileName); outputNames.push_back(outputFileName);
//make sure that if your picked groups size is not too big
int thisSize = names.size();
if (thisSize < size) { size = thisSize; }
}
}else { //make sure size is not too large
- int smallestSize = groupMap->getNumSeqs(Groups[0]);
- for (int i = 1; i < Groups.size(); i++) {
+ vector<string> newGroups;
+ for (int i = 0; i < Groups.size(); i++) {
int thisSize = groupMap->getNumSeqs(Groups[i]);
- if (thisSize < smallestSize) { smallestSize = thisSize; }
+ if (thisSize >= size) { newGroups.push_back(Groups[i]); }
+ else { m->mothurOut("You have selected a size that is larger than " + Groups[i] + " number of sequences, removing " + Groups[i] + "."); m->mothurOutEndLine(); }
}
- if (smallestSize < size) { size = smallestSize; m->mothurOut("You have selected a size that is larger than your smallest sample, using your samllest sample size, " + toString(smallestSize) + "."); m->mothurOutEndLine(); }
+ Groups = newGroups;
}
m->mothurOut("Sampling " + toString(size) + " from each group."); m->mothurOutEndLine();
set<string> subset; //dont want repeat sequence names added
if (persample) {
- for (int i = 0; i < Groups.size(); i++) {
-
- //randomly select a subset of those names from this group to include in the subsample
- for (int j = 0; j < size; j++) {
-
- if (m->control_pressed) { return 0; }
+ //initialize counts
+ map<string, int> groupCounts;
+ map<string, int>::iterator itGroupCounts;
+ for (int i = 0; i < Groups.size(); i++) { groupCounts[Groups[i]] = 0; }
+
+ for (int j = 0; j < names.size(); j++) {
- //get random sequence to add, making sure we have not already added it
- bool done = false;
- int myrand;
- while (!done) {
- myrand = (int)((float)(rand()) / (RAND_MAX / (thisSize-1) + 1));
-
- if (subset.count(names[myrand]) == 0) {
-
- string group = groupMap->getGroup(names[myrand]);
- if (group == "not found") { m->mothurOut("[ERROR]: " + names[myrand] + " is not in your groupfile. please correct."); m->mothurOutEndLine(); group = "NOTFOUND"; }
-
- if (group == Groups[i]) { subset.insert(names[myrand]); break; }
- }
+ if (m->control_pressed) { return 0; }
+
+ string group = groupMap->getGroup(names[j]);
+ if (group == "not found") { m->mothurOut("[ERROR]: " + names[j] + " is not in your groupfile. please correct."); m->mothurOutEndLine(); group = "NOTFOUND"; }
+ else{
+ itGroupCounts = groupCounts.find(group);
+ if (itGroupCounts != groupCounts.end()) {
+ if (groupCounts[group] < size) { subset.insert(names[j]); groupCounts[group]++; }
}
- }
+ }
}
}else {
+
//randomly select a subset of those names to include in the subsample
- for (int j = 0; j < size; j++) {
+ //since names was randomly shuffled just grab the next one
+ for (int j = 0; j < names.size(); j++) {
if (m->control_pressed) { return 0; }
- //get random sequence to add, making sure we have not already added it
- bool done = false;
- int myrand;
- while (!done) {
- myrand = (int)((float)(rand()) / (RAND_MAX / (thisSize-1) + 1));
+ if (groupfile != "") { //if there is a groupfile given fill in group info
+ string group = groupMap->getGroup(names[j]);
+ if (group == "not found") { m->mothurOut("[ERROR]: " + names[j] + " is not in your groupfile. please correct."); m->mothurOutEndLine(); group = "NOTFOUND"; }
- if (subset.count(names[myrand]) == 0) {
-
- if (groupfile != "") { //if there is a groupfile given fill in group info
- string group = groupMap->getGroup(names[myrand]);
- if (group == "not found") { m->mothurOut("[ERROR]: " + names[myrand] + " is not in your groupfile. please correct."); m->mothurOutEndLine(); group = "NOTFOUND"; }
-
- if (pickedGroups) { //if hte user picked groups, we only want to keep the names of sequences from those groups
- if (m->inUsersGroups(group, Groups)) {
- subset.insert(names[myrand]); break;
- }
- }else{
- subset.insert(names[myrand]); break;
- }
- }else{ //save everyone, group
- subset.insert(names[myrand]); break;
- }
+ if (pickedGroups) { //if hte user picked groups, we only want to keep the names of sequences from those groups
+ if (m->inUsersGroups(group, Groups)) {
+ subset.insert(names[j]);
+ }
+ }else{
+ subset.insert(names[j]);
}
- }
+ }else{ //save everyone, group
+ subset.insert(names[j]);
+ }
+
+ //do we have enough??
+ if (subset.size() == size) { break; }
}
}
+
+ if (subset.size() == 0) { m->mothurOut("The size you selected is too large, skipping fasta file."); m->mothurOutEndLine(); return 0; }
+
+ string thisOutputDir = outputDir;
+ if (outputDir == "") { thisOutputDir += m->hasPath(fastafile); }
+ string outputFileName = thisOutputDir + m->getRootName(m->getSimpleName(fastafile)) + getOutputFileNameTag("fasta", fastafile);
+ ofstream out;
+ m->openOutputFile(outputFileName, out);
+
//read through fasta file outputting only the names on the subsample list
ifstream in;
m->openInputFile(fastafile, in);
m->mothurOut("[ERROR]: The subset selected contained " + toString(subset.size()) + " sequences, but I only found " + toString(count) + " of those in the fastafile."); m->mothurOutEndLine();
}
+ if (namefile != "") {
+ m->mothurOut("Deconvoluting subsampled fasta file... "); m->mothurOutEndLine();
+
+ string outputNameFileName = thisOutputDir + m->getRootName(m->getSimpleName(namefile)) + getOutputFileNameTag("name", namefile);
+ //use unique.seqs to create new name and fastafile
+ string inputString = "fasta=" + outputFileName;
+ m->mothurOut("/******************************************/"); m->mothurOutEndLine();
+ m->mothurOut("Running command: unique.seqs(" + inputString + ")"); m->mothurOutEndLine();
+ m->mothurCalling = true;
+
+ Command* uniqueCommand = new DeconvoluteCommand(inputString);
+ uniqueCommand->execute();
+
+ map<string, vector<string> > filenames = uniqueCommand->getOutputFiles();
+
+ delete uniqueCommand;
+ m->mothurCalling = false;
+
+ m->renameFile(filenames["name"][0], outputNameFileName);
+ m->renameFile(filenames["fasta"][0], outputFileName);
+
+ outputTypes["name"].push_back(outputNameFileName); outputNames.push_back(outputNameFileName);
+
+ m->mothurOut("/******************************************/"); m->mothurOutEndLine();
+
+ m->mothurOut("Done."); m->mothurOutEndLine();
+ }
+
+ outputTypes["fasta"].push_back(outputFileName); outputNames.push_back(outputFileName);
+
//if a groupfile is provided read through the group file only outputting the names on the subsample list
if (groupfile != "") {
string groupOutputDir = outputDir;
if (outputDir == "") { groupOutputDir += m->hasPath(groupfile); }
- string groupOutputFileName = groupOutputDir + m->getRootName(m->getSimpleName(groupfile)) + "subsample" + m->getExtension(groupfile);
+ string groupOutputFileName = groupOutputDir + m->getRootName(m->getSimpleName(groupfile)) + getOutputFileNameTag("group", groupfile);
ofstream outGroup;
m->openOutputFile(groupOutputFileName, outGroup);
int SubSampleCommand::readNames() {
try {
- ifstream in;
- m->openInputFile(namefile, in);
-
- string thisname, repnames;
- map<string, vector<string> >::iterator it;
-
- while(!in.eof()){
-
- if (m->control_pressed) { in.close(); return 0; }
-
- in >> thisname; m->gobble(in); //read from first column
- in >> repnames; //read from second column
-
- it = nameMap.find(thisname);
- if (it == nameMap.end()) {
-
- vector<string> splitRepNames;
- m->splitAtComma(repnames, splitRepNames);
-
- nameMap[thisname] = splitRepNames;
- for (int i = 0; i < splitRepNames.size(); i++) { names.push_back(splitRepNames[i]); }
-
- }else{ m->mothurOut(thisname + " is already in namesfile. I will use first definition."); m->mothurOutEndLine(); }
-
- m->gobble(in);
- }
- in.close();
-
+ nameMap.clear();
+ m->readNames(namefile, nameMap);
+
+ //save names of all sequences
+ map<string, vector<string> >::iterator it;
+ for (it = nameMap.begin(); it != nameMap.end(); it++) { for (int i = 0; i < (it->second).size(); i++) { names.push_back((it->second)[i]); } }
+
return 0;
}
int SubSampleCommand::getSubSampleShared() {
try {
- string thisOutputDir = outputDir;
- if (outputDir == "") { thisOutputDir += m->hasPath(sharedfile); }
- string outputFileName = thisOutputDir + m->getRootName(m->getSimpleName(sharedfile)) + "subsample" + m->getExtension(sharedfile);
-
- ofstream out;
- m->openOutputFile(outputFileName, out);
- outputTypes["shared"].push_back(outputFileName); outputNames.push_back(outputFileName);
-
InputData* input = new InputData(sharedfile, "sharedfile");
vector<SharedRAbundVector*> lookup = input->getSharedRAbundVectors();
string lastLabel = lookup[0]->getLabel();
if (thisSize < size) { size = thisSize; }
}
+ }else {
+ m->clearGroups();
+ Groups.clear();
+ vector<SharedRAbundVector*> temp;
+ for (int i = 0; i < lookup.size(); i++) {
+ if (lookup[i]->getNumSeqs() < size) {
+ m->mothurOut(lookup[i]->getGroup() + " contains " + toString(lookup[i]->getNumSeqs()) + ". Eliminating."); m->mothurOutEndLine();
+ delete lookup[i];
+ }else {
+ Groups.push_back(lookup[i]->getGroup());
+ temp.push_back(lookup[i]);
+ }
+ }
+ lookup = temp;
+ m->setGroups(Groups);
}
- m->mothurOut("Sampling " + toString(size) + " from " + toString(lookup[0]->getNumSeqs()) + "."); m->mothurOutEndLine();
+ if (lookup.size() == 0) { m->mothurOut("The size you selected is too large, skipping shared file."); m->mothurOutEndLine(); delete input; return 0; }
+
+ m->mothurOut("Sampling " + toString(size) + " from each group."); m->mothurOutEndLine();
//as long as you are not at the end of the file or done wih the lines you want
while((lookup[0] != NULL) && ((allLines == 1) || (userLabels.size() != 0))) {
- if (m->control_pressed) { delete input; for (int i = 0; i < lookup.size(); i++) { delete lookup[i]; lookup[i] = NULL; } out.close(); return 0; }
+ if (m->control_pressed) { delete input; for (int i = 0; i < lookup.size(); i++) { delete lookup[i]; lookup[i] = NULL; } return 0; }
if(allLines == 1 || labels.count(lookup[0]->getLabel()) == 1){
m->mothurOut(lookup[0]->getLabel()); m->mothurOutEndLine();
- processShared(lookup, out);
+ processShared(lookup);
processedLabels.insert(lookup[0]->getLabel());
userLabels.erase(lookup[0]->getLabel());
lookup = input->getSharedRAbundVectors(lastLabel);
m->mothurOut(lookup[0]->getLabel()); m->mothurOutEndLine();
- processShared(lookup, out);
+ processShared(lookup);
processedLabels.insert(lookup[0]->getLabel());
userLabels.erase(lookup[0]->getLabel());
}
- if (m->control_pressed) { out.close(); return 0; }
+ if (m->control_pressed) { return 0; }
//output error messages about any remaining user labels
set<string>::iterator it;
m->mothurOut(lookup[0]->getLabel()); m->mothurOutEndLine();
- processShared(lookup, out);
+ processShared(lookup);
for (int i = 0; i < lookup.size(); i++) { delete lookup[i]; }
}
delete input;
- out.close();
return 0;
}
}
//**********************************************************************************************************************
-int SubSampleCommand::processShared(vector<SharedRAbundVector*>& thislookup, ofstream& out) {
+int SubSampleCommand::processShared(vector<SharedRAbundVector*>& thislookup) {
try {
- int numBins = thislookup[0]->getNumBins();
- for (int i = 0; i < thislookup.size(); i++) {
- int thisSize = thislookup[i]->getNumSeqs();
-
- if (thisSize != size) {
-
- string thisgroup = thislookup[i]->getGroup();
-
- OrderVector* order = new OrderVector();
- for(int p=0;p<numBins;p++){
- for(int j=0;j<thislookup[i]->getAbundance(p);j++){
- order->push_back(p);
- }
- }
- random_shuffle(order->begin(), order->end());
-
- SharedRAbundVector* temp = new SharedRAbundVector(numBins);
- temp->setLabel(thislookup[i]->getLabel());
- temp->setGroup(thislookup[i]->getGroup());
-
- delete thislookup[i];
- thislookup[i] = temp;
-
-
- for (int j = 0; j < size; j++) {
-
- if (m->control_pressed) { delete order; return 0; }
-
- //get random number to sample from order between 0 and thisSize-1.
- int myrand = (int)((float)(rand()) / (RAND_MAX / (thisSize-1) + 1));
-
- int bin = order->get(myrand);
-
- int abund = thislookup[i]->getAbundance(bin);
- thislookup[i]->set(bin, (abund+1), thisgroup);
- }
- delete order;
- }
- }
+ //save mothurOut's binLabels to restore for next label
+ vector<string> saveBinLabels = m->currentBinLabels;
- //subsampling may have created some otus with no sequences in them
- eliminateZeroOTUS(thislookup);
+ string thisOutputDir = outputDir;
+ if (outputDir == "") { thisOutputDir += m->hasPath(sharedfile); }
+ string outputFileName = thisOutputDir + m->getRootName(m->getSimpleName(sharedfile)) + thislookup[0]->getLabel() + getOutputFileNameTag("shared", sharedfile);
+ SubSample sample;
+ vector<string> subsampledLabels = sample.getSample(thislookup, size);
+
+ if (m->control_pressed) { return 0; }
+
+ ofstream out;
+ m->openOutputFile(outputFileName, out);
+ outputTypes["shared"].push_back(outputFileName); outputNames.push_back(outputFileName);
- if (m->control_pressed) { return 0; }
+ m->currentBinLabels = subsampledLabels;
+
+ thislookup[0]->printHeaders(out);
for (int i = 0; i < thislookup.size(); i++) {
out << thislookup[i]->getLabel() << '\t' << thislookup[i]->getGroup() << '\t';
thislookup[i]->print(out);
}
+ out.close();
+
+
+ //save mothurOut's binLabels to restore for next label
+ m->currentBinLabels = saveBinLabels;
return 0;
string thisOutputDir = outputDir;
if (outputDir == "") { thisOutputDir += m->hasPath(listfile); }
- string outputFileName = thisOutputDir + m->getRootName(m->getSimpleName(listfile)) + "subsample" + m->getExtension(listfile);
-
+ string outputFileName = thisOutputDir + m->getRootName(m->getSimpleName(listfile)) + getOutputFileNameTag("list", listfile);
ofstream out;
m->openOutputFile(outputFileName, out);
outputTypes["list"].push_back(outputFileName); outputNames.push_back(outputFileName);
//takes care of user setting groupNames that are invalid or setting groups=all
SharedUtil* util = new SharedUtil();
- util->setGroups(Groups, groupMap->namesOfGroups);
+ vector<string> namesGroups = groupMap->getNamesOfGroups();
+ util->setGroups(Groups, namesGroups);
delete util;
//create outputfiles
if (thisSize < size) { size = thisSize; }
}
}else { //make sure size is not too large
- int smallestSize = groupMap->getNumSeqs(Groups[0]);
- for (int i = 1; i < Groups.size(); i++) {
+ vector<string> newGroups;
+ for (int i = 0; i < Groups.size(); i++) {
int thisSize = groupMap->getNumSeqs(Groups[i]);
- if (thisSize < smallestSize) { smallestSize = thisSize; }
+ if (thisSize >= size) { newGroups.push_back(Groups[i]); }
+ else { m->mothurOut("You have selected a size that is larger than " + Groups[i] + " number of sequences, removing " + Groups[i] + "."); m->mothurOutEndLine(); }
}
- if (smallestSize < size) { size = smallestSize; m->mothurOut("You have selected a size that is larger than your smallest sample, using your samllest sample size, " + toString(smallestSize) + "."); m->mothurOutEndLine(); }
+ Groups = newGroups;
}
m->mothurOut("Sampling " + toString(size) + " from each group."); m->mothurOutEndLine();
//randomly select a subset of those names to include in the subsample
set<string> subset; //dont want repeat sequence names added
if (persample) {
- for (int i = 0; i < Groups.size(); i++) {
+ //initialize counts
+ map<string, int> groupCounts;
+ map<string, int>::iterator itGroupCounts;
+ for (int i = 0; i < Groups.size(); i++) { groupCounts[Groups[i]] = 0; }
+
+ for (int j = 0; j < names.size(); j++) {
- for (int j = 0; j < size; j++) {
-
- if (m->control_pressed) { break; }
-
- //get random sequence to add, making sure we have not already added it
- bool done = false;
- int myrand;
- while (!done) {
- myrand = (int)((float)(rand()) / (RAND_MAX / (names.size()-1) + 1));
-
- if (subset.count(names[myrand]) == 0) { //you are not already added
- if (groupMap->getGroup(names[myrand]) == Groups[i]) { subset.insert(names[myrand]); break; }
- }
+ if (m->control_pressed) { return 0; }
+
+ string group = groupMap->getGroup(names[j]);
+ if (group == "not found") { m->mothurOut("[ERROR]: " + names[j] + " is not in your groupfile. please correct."); m->mothurOutEndLine(); group = "NOTFOUND"; }
+ else{
+ itGroupCounts = groupCounts.find(group);
+ if (itGroupCounts != groupCounts.end()) {
+ if (groupCounts[group] < size) { subset.insert(names[j]); groupCounts[group]++; }
}
- }
+ }
}
}else{
for (int j = 0; j < size; j++) {
if (m->control_pressed) { break; }
- //get random sequence to add, making sure we have not already added it
- bool done = false;
- int myrand;
- while (!done) {
- myrand = (int)((float)(rand()) / (RAND_MAX / (names.size()-1) + 1));
-
- if (subset.count(names[myrand]) == 0) { subset.insert(names[myrand]); break; }
- }
+ subset.insert(names[j]);
}
}
individual += binnames[j];
}
}
- if (subset.count(individual) != 0) { newNames += individual; }
+ if (subset.count(individual) != 0) { newNames += individual + ","; }
//if there are names in this bin add to new list
//**********************************************************************************************************************
int SubSampleCommand::getSubSampleRabund() {
try {
-
- string thisOutputDir = outputDir;
- if (outputDir == "") { thisOutputDir += m->hasPath(rabundfile); }
- string outputFileName = thisOutputDir + m->getRootName(m->getSimpleName(rabundfile)) + "subsample" + m->getExtension(rabundfile);
-
- ofstream out;
- m->openOutputFile(outputFileName, out);
- outputTypes["rabund"].push_back(outputFileName); outputNames.push_back(outputFileName);
-
InputData* input = new InputData(rabundfile, "rabund");
RAbundVector* rabund = input->getRAbundVector();
string lastLabel = rabund->getLabel();
if (size == 0) { //user has not set size, set size = 10%
size = int((rabund->getNumSeqs()) * 0.10);
- }
+ }else if (size > rabund->getNumSeqs()) { m->mothurOut("The size you selected is too large, skipping rabund file."); m->mothurOutEndLine(); delete input; delete rabund; return 0; }
m->mothurOut("Sampling " + toString(size) + " from " + toString(rabund->getNumSeqs()) + "."); m->mothurOutEndLine();
+ string thisOutputDir = outputDir;
+ if (outputDir == "") { thisOutputDir += m->hasPath(rabundfile); }
+ string outputFileName = thisOutputDir + m->getRootName(m->getSimpleName(rabundfile)) + getOutputFileNameTag("rabund", rabundfile);
+ ofstream out;
+ m->openOutputFile(outputFileName, out);
+ outputTypes["rabund"].push_back(outputFileName); outputNames.push_back(outputFileName);
+
//as long as you are not at the end of the file or done wih the lines you want
while((rabund != NULL) && ((allLines == 1) || (userLabels.size() != 0))) {
if (m->control_pressed) { delete input; delete rabund; out.close(); return 0; }
if (m->control_pressed) { delete order; return 0; }
- //get random number to sample from order between 0 and thisSize-1.
- int myrand = (int)((float)(rand()) / (RAND_MAX / (thisSize-1) + 1));
-
- int bin = order->get(myrand);
+ int bin = order->get(j);
int abund = rabund->get(bin);
rabund->set(bin, (abund+1));
//**********************************************************************************************************************
int SubSampleCommand::getSubSampleSabund() {
try {
-
- string thisOutputDir = outputDir;
- if (outputDir == "") { thisOutputDir += m->hasPath(sabundfile); }
- string outputFileName = thisOutputDir + m->getRootName(m->getSimpleName(sabundfile)) + "subsample" + m->getExtension(sabundfile);
-
- ofstream out;
- m->openOutputFile(outputFileName, out);
- outputTypes["sabund"].push_back(outputFileName); outputNames.push_back(outputFileName);
-
+
InputData* input = new InputData(sabundfile, "sabund");
SAbundVector* sabund = input->getSAbundVector();
string lastLabel = sabund->getLabel();
if (size == 0) { //user has not set size, set size = 10%
size = int((sabund->getNumSeqs()) * 0.10);
- }
+ }else if (size > sabund->getNumSeqs()) { m->mothurOut("The size you selected is too large, skipping sabund file."); m->mothurOutEndLine(); delete input; delete sabund; return 0; }
+
m->mothurOut("Sampling " + toString(size) + " from " + toString(sabund->getNumSeqs()) + "."); m->mothurOutEndLine();
+ string thisOutputDir = outputDir;
+ if (outputDir == "") { thisOutputDir += m->hasPath(sabundfile); }
+ string outputFileName = thisOutputDir + m->getRootName(m->getSimpleName(sabundfile)) + getOutputFileNameTag("sabund", sabundfile);
+ ofstream out;
+ m->openOutputFile(outputFileName, out);
+ outputTypes["sabund"].push_back(outputFileName); outputNames.push_back(outputFileName);
+
+
//as long as you are not at the end of the file or done wih the lines you want
while((sabund != NULL) && ((allLines == 1) || (userLabels.size() != 0))) {
if (m->control_pressed) { delete input; delete sabund; out.close(); return 0; }
if (m->control_pressed) { delete order; return 0; }
- //get random number to sample from order between 0 and thisSize-1.
- int myrand = (int)((float)(rand()) / (RAND_MAX / (thisSize-1) + 1));
-
- int bin = order->get(myrand);
+ int bin = order->get(j);
int abund = rabund->get(bin);
rabund->set(bin, (abund+1));
}
}
//**********************************************************************************************************************
-int SubSampleCommand::eliminateZeroOTUS(vector<SharedRAbundVector*>& thislookup) {
- try {
-
- vector<SharedRAbundVector*> newLookup;
- for (int i = 0; i < thislookup.size(); i++) {
- SharedRAbundVector* temp = new SharedRAbundVector();
- temp->setLabel(thislookup[i]->getLabel());
- temp->setGroup(thislookup[i]->getGroup());
- newLookup.push_back(temp);
- }
-
- //for each bin
- for (int i = 0; i < thislookup[0]->getNumBins(); i++) {
- if (m->control_pressed) { for (int j = 0; j < newLookup.size(); j++) { delete newLookup[j]; } return 0; }
-
- //look at each sharedRabund and make sure they are not all zero
- bool allZero = true;
- for (int j = 0; j < thislookup.size(); j++) {
- if (thislookup[j]->getAbundance(i) != 0) { allZero = false; break; }
- }
-
- //if they are not all zero add this bin
- if (!allZero) {
- for (int j = 0; j < thislookup.size(); j++) {
- newLookup[j]->push_back(thislookup[j]->getAbundance(i), thislookup[j]->getGroup());
- }
- }
- }
-
- for (int j = 0; j < thislookup.size(); j++) { delete thislookup[j]; }
- thislookup.clear();
-
- thislookup = newLookup;
-
- return 0;
-
- }
- catch(exception& e) {
- m->errorOut(e, "SubSampleCommand", "eliminateZeroOTUS");
- exit(1);
- }
-}
-
-//**********************************************************************************************************************
projects / mothur.git / blobdiff