try{
globaldata = GlobalData::getInstance();
abort = false;
+ format = "";
//allow user to run help
if(option == "help") { help(); abort = true; }
else {
//valid paramters for this command
- string Array[] = {"phylip","column","name","cutoff","precision","method","splitmethod","taxonomy","taxlevel","showabund","timing","hard","processors","outputdir","inputdir"};
+ string Array[] = {"fasta","phylip","column","name","cutoff","precision","method","splitmethod","taxonomy","taxlevel","large","showabund","timing","hard","processors","outputdir","inputdir"};
vector<string> myArray (Array, Array+(sizeof(Array)/sizeof(string)));
OptionParser parser(option);
//if the user has not given a path then, add inputdir. else leave path alone.
if (path == "") { parameters["taxonomy"] = inputDir + it->second; }
}
+
+ it = parameters.find("fasta");
+ //user has given a template file
+ if(it != parameters.end()){
+ path = hasPath(it->second);
+ //if the user has not given a path then, add inputdir. else leave path alone.
+ if (path == "") { parameters["fasta"] = inputDir + it->second; }
+ }
}
//check for required parameters
if (namefile == "not open") { abort = true; }
else if (namefile == "not found") { namefile = ""; }
+ fastafile = validParameter.validFile(parameters, "fasta", true);
+ if (fastafile == "not open") { abort = true; }
+ else if (fastafile == "not found") { fastafile = ""; }
+ else { distfile = fastafile; splitmethod = "fasta"; }
+
taxFile = validParameter.validFile(parameters, "taxonomy", true);
if (taxFile == "not open") { abort = true; }
else if (taxFile == "not found") { taxFile = ""; }
- if ((phylipfile == "") && (columnfile == "")) { m->mothurOut("When executing a cluster.split command you must enter a phylip or a column."); m->mothurOutEndLine(); abort = true; }
- else if ((phylipfile != "") && (columnfile != "")) { m->mothurOut("When executing a cluster.split command you must enter ONLY ONE of the following: phylip or column."); m->mothurOutEndLine(); abort = true; }
+ if ((phylipfile == "") && (columnfile == "") && (fastafile == "")) { m->mothurOut("When executing a cluster.split command you must enter a phylip or a column or fastafile."); m->mothurOutEndLine(); abort = true; }
+ else if ((phylipfile != "") && (columnfile != "") && (fastafile != "")) { m->mothurOut("When executing a cluster.split command you must enter ONLY ONE of the following: fasta, phylip or column."); m->mothurOutEndLine(); abort = true; }
if (columnfile != "") {
if (namefile == "") { m->mothurOut("You need to provide a namefile if you are going to use the column format."); m->mothurOutEndLine(); abort = true; }
}
+
+ if (fastafile != "") {
+ if (taxFile == "") { m->mothurOut("You need to provide a taxonomy file if you are using a fasta file to generate the split."); m->mothurOutEndLine(); abort = true; }
+ if (namefile == "") { m->mothurOut("You need to provide a names file if you are using a fasta file to generate the split."); m->mothurOutEndLine(); abort = true; }
+ }
//check for optional parameter and set defaults
// ...at some point should added some additional type checking...
temp = validParameter.validFile(parameters, "hard", false); if (temp == "not found") { temp = "F"; }
hard = isTrue(temp);
+ temp = validParameter.validFile(parameters, "large", false); if (temp == "not found") { temp = "F"; }
+ large = isTrue(temp);
+
temp = validParameter.validFile(parameters, "processors", false); if (temp == "not found"){ temp = "1"; }
convert(temp, processors);
- splitmethod = validParameter.validFile(parameters, "splitmethod", false); if (splitmethod == "not found") { splitmethod = "distance"; }
+ temp = validParameter.validFile(parameters, "splitmethod", false);
+ if (splitmethod != "fasta") {
+ if (temp == "not found") { splitmethod = "distance"; }
+ else { splitmethod = temp; }
+ }
temp = validParameter.validFile(parameters, "cutoff", false); if (temp == "not found") { temp = "10"; }
convert(temp, cutoff);
if ((method == "furthest") || (method == "nearest") || (method == "average")) { }
else { m->mothurOut("Not a valid clustering method. Valid clustering algorithms are furthest, nearest or average."); m->mothurOutEndLine(); abort = true; }
- if ((splitmethod == "distance") || (splitmethod == "classify")) { }
- else { m->mothurOut("Not a valid splitting method. Valid splitting algorithms are distance or classify."); m->mothurOutEndLine(); abort = true; }
+ if ((splitmethod == "distance") || (splitmethod == "classify") || (splitmethod == "fasta")) { }
+ else { m->mothurOut(splitmethod + " is not a valid splitting method. Valid splitting algorithms are distance, classify or fasta."); m->mothurOutEndLine(); abort = true; }
if ((splitmethod == "classify") && (taxFile == "")) { m->mothurOut("You need to provide a taxonomy file if you are going to use the classify splitmethod."); m->mothurOutEndLine(); abort = true; }
void ClusterSplitCommand::help(){
try {
- m->mothurOut("The cluster.split command parameter options are phylip, column, name, cutoff, precision, method, splitmethod, taxonomy, taxlevel, showabund, timing, hard, processors. Phylip or column and name are required.\n");
+ m->mothurOut("The cluster.split command parameter options are fasta, phylip, column, name, cutoff, precision, method, splitmethod, taxonomy, taxlevel, showabund, timing, hard, large, processors. Fasta or Phylip or column and name are required.\n");
+ m->mothurOut("The cluster.split command can split your files in 3 ways. Splitting by distance file, by classification, or by classification also using a fasta file. \n");
+ m->mothurOut("For the distance file method, you need only provide your distance file and mothur will split the file into distinct groups. \n");
+ m->mothurOut("For the classification method, you need to provide your distance file and taxonomy file, and set the splitmethod to classify. \n");
+ m->mothurOut("You will also need to set the taxlevel you want to split by. mothur will split the sequence into distinct taxonomy groups, and split the distance file based on those groups. \n");
+ m->mothurOut("For the classification method using a fasta file, you need to provide your fasta file, names file and taxonomy file. \n");
+ m->mothurOut("You will also need to set the taxlevel you want to split by. mothur will split the sequence into distinct taxonomy groups, and create distance files for each grouping. \n");
m->mothurOut("The phylip and column parameter allow you to enter your distance file. \n");
+ m->mothurOut("The fasta parameter allows you to enter your aligned fasta file. \n");
m->mothurOut("The name parameter allows you to enter your name file and is required if your distance file is in column format. \n");
m->mothurOut("The cutoff parameter allow you to set the distance you want to cluster to, default is 10.0. \n");
m->mothurOut("The precision parameter allows you specify the precision of the precision of the distances outputted, default=100, meaning 2 decimal places. \n");
m->mothurOut("The method allows you to specify what clustering algorythm you want to use, default=furthest, option furthest, nearest, or average. \n");
- m->mothurOut("The splitmethod parameter allows you to specify how you want to split your distance file before you cluster, default=distance, options distance or classify. \n");
+ m->mothurOut("The splitmethod parameter allows you to specify how you want to split your distance file before you cluster, default=distance, options distance, classify or fasta. \n");
m->mothurOut("The taxonomy parameter allows you to enter the taxonomy file for your sequences, this is only valid if you are using splitmethod=classify. Be sure your taxonomy file does not include the probability scores. \n");
m->mothurOut("The taxlevel parameter allows you to specify the taxonomy level you want to use to split the distance file, default=1. \n");
+ m->mothurOut("The large parameter allows you to indicate that your distance matrix is too large to fit in RAM. The default value is false.\n");
m->mothurOut("The cluster.split command should be in the following format: \n");
m->mothurOut("cluster.split(column=youDistanceFile, name=yourNameFile, method=yourMethod, cutoff=yourCutoff, precision=yourPrecision, splitmethod=yourSplitmethod, taxonomy=yourTaxonomyfile, taxlevel=yourtaxlevel) \n");
m->mothurOut("Example: cluster.split(column=abrecovery.dist, name=abrecovery.names, method=furthest, cutoff=0.10, precision=1000, splitmethod=classify, taxonomy=abrecovery.silva.slv.taxonomy, taxlevel=5) \n");
try {
if (abort == true) { return 0; }
-
+ time_t estart;
//****************** file prep work ******************************//
//if user gave a phylip file convert to column file
if (format == "phylip") {
-
+ estart = time(NULL);
+ m->mothurOut("Converting to column format..."); m->mothurOutEndLine();
+
ReadCluster* convert = new ReadCluster(distfile, cutoff, outputDir, false);
NameAssignment* nameMap = NULL;
}
delete listToMakeNameFile;
delete convert;
+
+ m->mothurOut("It took " + toString(time(NULL) - estart) + " seconds to convert the distance file."); m->mothurOutEndLine();
}
if (m->control_pressed) { return 0; }
- time_t estart = time(NULL);
+ estart = time(NULL);
+ m->mothurOut("Splitting the file..."); m->mothurOutEndLine();
//split matrix into non-overlapping groups
SplitMatrix* split;
- if (splitmethod == "distance") { split = new SplitMatrix(distfile, namefile, taxFile, cutoff, splitmethod); }
- else { split = new SplitMatrix(distfile, namefile, taxFile, taxLevelCutoff, splitmethod); }
+ if (splitmethod == "distance") { split = new SplitMatrix(distfile, namefile, taxFile, cutoff, splitmethod, large); }
+ else if (splitmethod == "classify") { split = new SplitMatrix(distfile, namefile, taxFile, taxLevelCutoff, splitmethod, large); }
+ else if (splitmethod == "fasta") { split = new SplitMatrix(fastafile, namefile, taxFile, taxLevelCutoff, splitmethod, processors, outputDir); }
+ else { m->mothurOut("Not a valid splitting method. Valid splitting algorithms are distance, classify or fasta."); m->mothurOutEndLine(); return 0; }
split->split();
ifstream in;
openInputFile(filename, in);
+ in >> tag; gobble(in);
+
while(!in.eof()) {
string tempName;
in >> tempName; gobble(in);
while(!in2.eof()) {
string tempName;
- in2 >> tempName; gobble(in);
+ in2 >> tempName; gobble(in2);
if (labels.count(tempName) == 0) { labels.insert(tempName); }
}
in2.close();
if (m->control_pressed) { for (int i = 0; i < listFileNames.size(); i++) { remove(listFileNames[i].c_str()); } return 0; }
+ m->mothurOut("It took " + toString(time(NULL) - estart) + " seconds to cluster"); m->mothurOutEndLine();
+
//****************** merge list file and create rabund and sabund files ******************************//
-
- mergeLists(listFileNames, singletonName, labels);
+ estart = time(NULL);
+ m->mothurOut("Merging the clustered files..."); m->mothurOutEndLine();
+
+ ListVector* listSingle;
+ map<float, int> labelBins = completeListFile(listFileNames, singletonName, labels, listSingle); //returns map of label to numBins
+
+ if (m->control_pressed) { if (listSingle != NULL) { delete listSingle; } for (int i = 0; i < outputNames.size(); i++) { remove(outputNames[i].c_str()); } return 0; }
+
+ mergeLists(listFileNames, labelBins, listSingle);
if (m->control_pressed) { for (int i = 0; i < outputNames.size(); i++) { remove(outputNames[i].c_str()); } return 0; }
- m->mothurOut("It took " + toString(time(NULL) - estart) + " seconds to cluster"); m->mothurOutEndLine();
+ m->mothurOut("It took " + toString(time(NULL) - estart) + " seconds to merge."); m->mothurOutEndLine();
m->mothurOutEndLine();
m->mothurOut("Output File Names: "); m->mothurOutEndLine();
}
}
//**********************************************************************************************************************
-int ClusterSplitCommand::mergeLists(vector<string> listNames, string singleton, set<string> userLabels){
+map<float, int> ClusterSplitCommand::completeListFile(vector<string> listNames, string singleton, set<string>& userLabels, ListVector*& listSingle){
try {
- if (outputDir == "") { outputDir += hasPath(distfile); }
- fileroot = outputDir + getRootName(getSimpleName(distfile));
-
- openOutputFile(fileroot+ tag + ".sabund", outSabund);
- openOutputFile(fileroot+ tag + ".rabund", outRabund);
- openOutputFile(fileroot+ tag + ".list", outList);
- outputNames.push_back(fileroot+ tag + ".sabund");
- outputNames.push_back(fileroot+ tag + ".rabund");
- outputNames.push_back(fileroot+ tag + ".list");
+ map<float, int> labelBin;
+ vector<float> orderFloat;
+ int numSingleBins;
//read in singletons
- ListVector* listSingle = NULL;
if (singleton != "none") {
ifstream in;
openInputFile(singleton, in);
listSingle->push_back(secondCol);
}
in.close();
- }
+ remove(singleton.c_str());
+
+ numSingleBins = listSingle->getNumBins();
+ }else{ listSingle = NULL; numSingleBins = 0; }
- vector<float> orderFloat;
-
//go through users set and make them floats so we can sort them
for(set<string>::iterator it = userLabels.begin(); it != userLabels.end(); ++it) {
- float temp;
+ float temp = -10.0;
- if ((*it != "unique") && (convertTestFloat(*it, temp) == true)){
- convert(*it, temp);
- orderFloat.push_back(temp);
- }else if (*it == "unique") { orderFloat.push_back(-1.0); }
- else {
- userLabels.erase(*it);
- it--;
- }
+ if ((*it != "unique") && (convertTestFloat(*it, temp) == true)) { convert(*it, temp); }
+ else if (*it == "unique") { temp = -1.0; }
+
+ orderFloat.push_back(temp);
+ labelBin[temp] = numSingleBins; //initialize numbins
}
//sort order
sort(orderFloat.begin(), orderFloat.end());
-
- vector<InputData*> inputs;
- vector<string> lastLabels;
- for (int i = 0; i < listNames.size(); i++) {
- InputData* input = new InputData(listNames[i], "list");
- inputs.push_back(input);
+ userLabels.clear();
- ifstream in;
- openInputFile(listNames[i], in);
- ListVector tempList(in);
- lastLabels.push_back(tempList.getLabel());
- in.close();
- }
+ //get the list info from each file
+ for (int k = 0; k < listNames.size(); k++) {
- ListVector* merged = NULL;
-
- //for each label needed
- for(int l = 0; l < orderFloat.size(); l++){
+ if (m->control_pressed) {
+ if (listSingle != NULL) { delete listSingle; listSingle = NULL; remove(singleton.c_str()); }
+ for (int i = 0; i < listNames.size(); i++) { remove(listNames[i].c_str()); }
+ return labelBin;
+ }
- string thisLabel;
- if (orderFloat[l] == -1) { thisLabel = "unique"; }
- else { thisLabel = toString(orderFloat[l], length-1); }
-
- //get the list info from each file
- for (int k = 0; k < listNames.size(); k++) {
+ InputData* input = new InputData(listNames[k], "list");
+ ListVector* list = input->getListVector();
+ string lastLabel = list->getLabel();
+
+ string filledInList = listNames[k] + "filledInTemp";
+ ofstream outFilled;
+ openOutputFile(filledInList, outFilled);
- if (m->control_pressed) {
- if (listSingle != NULL) { delete listSingle; remove(singleton.c_str()); }
- for (int i = 0; i < listNames.size(); i++) { delete inputs[i]; remove(listNames[i].c_str()); }
- delete merged; merged = NULL;
- return 0;
- }
-
- ListVector* list = inputs[k]->getListVector();
-
+ //for each label needed
+ for(int l = 0; l < orderFloat.size(); l++){
+
+ string thisLabel;
+ if (orderFloat[l] == -1) { thisLabel = "unique"; }
+ else { thisLabel = toString(orderFloat[l], length-1); }
+
//this file has reached the end
- if (list == NULL) { list = inputs[k]->getListVector(lastLabels[k], true); }
+ if (list == NULL) {
+ list = input->getListVector(lastLabel, true);
+ }else{ //do you have the distance, or do you need to fill in
- float labelFloat;
- if (list->getLabel() == "unique") { labelFloat = -1.0; }
- else { convert(list->getLabel(), labelFloat); }
-
- //check for missing labels
- if (labelFloat > orderFloat[l]) { //you are missing the label, get the next smallest one
- //if its bigger get last label, otherwise keep it
- delete list;
- list = inputs[k]->getListVector(lastLabels[k], true); //get last list vector to use, you actually want to move back in the file
- }
- lastLabels[k] = list->getLabel();
+ float labelFloat;
+ if (list->getLabel() == "unique") { labelFloat = -1.0; }
+ else { convert(list->getLabel(), labelFloat); }
- //is this the first file
- if (merged == NULL) { merged = new ListVector(); merged->setLabel(thisLabel); }
-
- for (int j = 0; j < list->getNumBins(); j++) {
- merged->push_back(list->get(j));
+ //check for missing labels
+ if (labelFloat > orderFloat[l]) { //you are missing the label, get the next smallest one
+ //if its bigger get last label, otherwise keep it
+ delete list;
+ list = input->getListVector(lastLabel, true); //get last list vector to use, you actually want to move back in the file
+ }
+ lastLabel = list->getLabel();
}
+ //print to new file
+ list->setLabel(thisLabel);
+ list->print(outFilled);
+
+ //update labelBin
+ labelBin[orderFloat[l]] += list->getNumBins();
+
delete list;
+
+ list = input->getListVector();
}
+ if (list != NULL) { delete list; }
+ delete input;
+
+ outFilled.close();
+ remove(listNames[k].c_str());
+ rename(filledInList.c_str(), listNames[k].c_str());
+ }
+
+ return labelBin;
+ }
+ catch(exception& e) {
+ m->errorOut(e, "ClusterSplitCommand", "completeListFile");
+ exit(1);
+ }
+}
+//**********************************************************************************************************************
+int ClusterSplitCommand::mergeLists(vector<string> listNames, map<float, int> userLabels, ListVector* listSingle){
+ try {
+ if (outputDir == "") { outputDir += hasPath(distfile); }
+ fileroot = outputDir + getRootName(getSimpleName(distfile));
+
+ openOutputFile(fileroot+ tag + ".sabund", outSabund);
+ openOutputFile(fileroot+ tag + ".rabund", outRabund);
+ openOutputFile(fileroot+ tag + ".list", outList);
+
+ outputNames.push_back(fileroot+ tag + ".sabund");
+ outputNames.push_back(fileroot+ tag + ".rabund");
+ outputNames.push_back(fileroot+ tag + ".list");
+
+ map<float, int>::iterator itLabel;
+
+ //for each label needed
+ for(itLabel = userLabels.begin(); itLabel != userLabels.end(); itLabel++) {
+
+ string thisLabel;
+ if (itLabel->first == -1) { thisLabel = "unique"; }
+ else { thisLabel = toString(itLabel->first, length-1); }
+
+ outList << thisLabel << '\t' << itLabel->second << '\t';
+
+ RAbundVector* rabund = new RAbundVector();
+ rabund->setLabel(thisLabel);
+
//add in singletons
if (listSingle != NULL) {
for (int j = 0; j < listSingle->getNumBins(); j++) {
- merged->push_back(listSingle->get(j));
+ outList << listSingle->get(j) << '\t';
+ rabund->push_back(getNumNames(listSingle->get(j)));
}
}
- //print to files
- printData(merged);
+ //get the list info from each file
+ for (int k = 0; k < listNames.size(); k++) {
+
+ if (m->control_pressed) { if (listSingle != NULL) { delete listSingle; } for (int i = 0; i < listNames.size(); i++) { remove(listNames[i].c_str()); } delete rabund; return 0; }
+
+ InputData* input = new InputData(listNames[k], "list");
+ ListVector* list = input->getListVector(thisLabel);
+
+ //this file has reached the end
+ if (list == NULL) { m->mothurOut("Error merging listvectors in file " + listNames[k]); m->mothurOutEndLine(); }
+ else {
+ for (int j = 0; j < list->getNumBins(); j++) {
+ outList << list->get(j) << '\t';
+ rabund->push_back(getNumNames(list->get(j)));
+ }
+ delete list;
+ }
+ delete input;
+ }
+
+ SAbundVector sabund = rabund->getSAbundVector();
+
+ sabund.print(outSabund);
+ rabund->print(outRabund);
+ outList << endl;
- delete merged; merged = NULL;
+ delete rabund;
}
- if (listSingle != NULL) { delete listSingle; remove(singleton.c_str()); }
+ outList.close();
+ outRabund.close();
+ outSabund.close();
- for (int i = 0; i < listNames.size(); i++) { delete inputs[i]; remove(listNames[i].c_str()); }
+ if (listSingle != NULL) { delete listSingle; }
+
+ for (int i = 0; i < listNames.size(); i++) { remove(listNames[i].c_str()); }
return 0;
}
exit(1);
}
}
+
//**********************************************************************************************************************
void ClusterSplitCommand::printData(ListVector* oldList){
string filename = toString(getpid()) + ".temp";
ofstream out;
openOutputFile(filename, out);
+ out << tag << endl;
for (int j = 0; j < listFileNames.size(); j++) { out << listFileNames[j] << endl; }
out.close();
globaldata->setNameFile(thisNamefile);
globaldata->setColumnFile(thisDistFile); globaldata->setFormat("column");
+ m->mothurOutEndLine(); m->mothurOut("Reading " + thisDistFile); m->mothurOutEndLine();
+
ReadMatrix* read = new ReadColumnMatrix(thisDistFile);
read->setCutoff(cutoff);
projects / mothur.git / blobdiff