#include "readphylip.h"
#include "readcolumn.h"
#include "readmatrix.hpp"
+#include "clusterdoturcommand.h"
+
//**********************************************************************************************************************
vector<string> ClusterCommand::setParameters(){
try {
- CommandParameter pphylip("phylip", "InputTypes", "", "", "PhylipColumn", "PhylipColumn", "none",false,false); parameters.push_back(pphylip);
- CommandParameter pname("name", "InputTypes", "", "", "none", "none", "ColumnName",false,false); parameters.push_back(pname);
- CommandParameter pcolumn("column", "InputTypes", "", "", "PhylipColumn", "PhylipColumn", "ColumnName",false,false); parameters.push_back(pcolumn);
- CommandParameter pcutoff("cutoff", "Number", "", "10", "", "", "",false,false); parameters.push_back(pcutoff);
- CommandParameter pprecision("precision", "Number", "", "100", "", "", "",false,false); parameters.push_back(pprecision);
- CommandParameter pmethod("method", "Multiple", "furthest-nearest-average-weighted", "average", "", "", "",false,false); parameters.push_back(pmethod);
- CommandParameter pshowabund("showabund", "Boolean", "", "T", "", "", "",false,false); parameters.push_back(pshowabund);
- CommandParameter ptiming("timing", "Boolean", "", "F", "", "", "",false,false); parameters.push_back(ptiming);
- CommandParameter psim("sim", "Boolean", "", "F", "", "", "",false,false); parameters.push_back(psim);
- CommandParameter phard("hard", "Boolean", "", "T", "", "", "",false,false); parameters.push_back(phard);
- CommandParameter pinputdir("inputdir", "String", "", "", "", "", "",false,false); parameters.push_back(pinputdir);
- CommandParameter poutputdir("outputdir", "String", "", "", "", "", "",false,false); parameters.push_back(poutputdir);
+ CommandParameter pphylip("phylip", "InputTypes", "", "", "PhylipColumn", "PhylipColumn", "none","list",false,false,true); parameters.push_back(pphylip);
+ CommandParameter pname("name", "InputTypes", "", "", "NameCount", "none", "ColumnName","rabund-sabund",false,false,true); parameters.push_back(pname);
+ CommandParameter pcount("count", "InputTypes", "", "", "NameCount", "none", "none","",false,false,true); parameters.push_back(pcount);
+ CommandParameter pcolumn("column", "InputTypes", "", "", "PhylipColumn", "PhylipColumn", "ColumnName","list",false,false,true); parameters.push_back(pcolumn);
+ CommandParameter pcutoff("cutoff", "Number", "", "10", "", "", "","",false,false,true); parameters.push_back(pcutoff);
+ CommandParameter pprecision("precision", "Number", "", "100", "", "", "","",false,false); parameters.push_back(pprecision);
+ CommandParameter pmethod("method", "Multiple", "furthest-nearest-average-weighted", "average", "", "", "","",false,false,true); parameters.push_back(pmethod);
+ CommandParameter pshowabund("showabund", "Boolean", "", "T", "", "", "","",false,false); parameters.push_back(pshowabund);
+ CommandParameter ptiming("timing", "Boolean", "", "F", "", "", "","",false,false); parameters.push_back(ptiming);
+ CommandParameter psim("sim", "Boolean", "", "F", "", "", "","",false,false); parameters.push_back(psim);
+ CommandParameter phard("hard", "Boolean", "", "T", "", "", "","",false,false); parameters.push_back(phard);
+ CommandParameter pinputdir("inputdir", "String", "", "", "", "", "","",false,false); parameters.push_back(pinputdir);
+ //CommandParameter padjust("adjust", "String", "", "F", "", "", "","",false,false); parameters.push_back(padjust);
+ 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); }
string ClusterCommand::getHelpString(){
try {
string helpString = "";
- helpString += "The cluster command parameter options are phylip, column, name, method, cuttoff, hard, precision, sim, showabund and timing. Phylip or column and name are required, unless you have a valid current file.\n";
- helpString += "The cluster command should be in the following format: \n";
+ helpString += "The cluster command parameter options are phylip, column, name, count, method, cuttoff, hard, precision, sim, showabund and timing. Phylip or column and name are required, unless you have a valid current file.\n";
+ //helpString += "The adjust parameter is used to handle missing distances. If you set a cutoff, adjust=f by default. If not, adjust=t by default. Adjust=f, means ignore missing distances and adjust cutoff as needed with the average neighbor method. Adjust=t, will treat missing distances as 1.0. You can also set the value the missing distances should be set to, adjust=0.5 would give missing distances a value of 0.5.\n";
+ helpString += "The cluster command should be in the following format: \n";
helpString += "cluster(method=yourMethod, cutoff=yourCutoff, precision=yourPrecision) \n";
helpString += "The acceptable cluster methods are furthest, nearest, average and weighted. If no method is provided then average is assumed.\n";
return helpString;
}
}
//**********************************************************************************************************************
+string ClusterCommand::getOutputPattern(string type) {
+ try {
+ string pattern = "";
+
+ if (type == "list") { pattern = "[filename],[clustertag],list-[filename],[clustertag],[tag2],list"; }
+ else if (type == "rabund") { pattern = "[filename],[clustertag],rabund"; }
+ else if (type == "sabund") { pattern = "[filename],[clustertag],sabund"; }
+ else { m->mothurOut("[ERROR]: No definition for type " + type + " output pattern.\n"); m->control_pressed = true; }
+
+ return pattern;
+ }
+ catch(exception& e) {
+ m->errorOut(e, "ClusterCommand", "getOutputPattern");
+ exit(1);
+ }
+}
+//**********************************************************************************************************************
ClusterCommand::ClusterCommand(){
try {
abort = true; calledHelp = true;
//if the user has not given a path then, add inputdir. else leave path alone.
if (path == "") { parameters["name"] = inputDir + it->second; }
}
+
+ it = parameters.find("count");
+ //user has given a template file
+ if(it != parameters.end()){
+ path = m->hasPath(it->second);
+ //if the user has not given a path then, add inputdir. else leave path alone.
+ if (path == "") { parameters["count"] = inputDir + it->second; }
+ }
}
//check for required parameters
if (namefile == "not open") { abort = true; }
else if (namefile == "not found") { namefile = ""; }
else { m->setNameFile(namefile); }
+
+ countfile = validParameter.validFile(parameters, "count", true);
+ if (countfile == "not open") { abort = true; countfile = ""; }
+ else if (countfile == "not found") { countfile = ""; }
+ else { m->setCountTableFile(countfile); }
if ((phylipfile == "") && (columnfile == "")) {
//is there are current file available for either of these?
else if ((phylipfile != "") && (columnfile != "")) { m->mothurOut("When executing a cluster command you must enter ONLY ONE of the following: phylip or column."); m->mothurOutEndLine(); abort = true; }
if (columnfile != "") {
- if (namefile == "") {
+ if ((namefile == "") && (countfile == "")){
namefile = m->getNameFile();
if (namefile != "") { m->mothurOut("Using " + namefile + " as input file for the name parameter."); m->mothurOutEndLine(); }
else {
- m->mothurOut("You need to provide a namefile if you are going to use the column format."); m->mothurOutEndLine();
- abort = true;
+ countfile = m->getCountTableFile();
+ if (countfile != "") { m->mothurOut("Using " + countfile + " as input file for the count parameter."); m->mothurOutEndLine(); }
+ else {
+ m->mothurOut("You need to provide a namefile or countfile if you are going to use the column format."); m->mothurOutEndLine();
+ abort = true;
+ }
}
}
}
+ if ((countfile != "") && (namefile != "")) { m->mothurOut("When executing a cluster command you must enter ONLY ONE of the following: count or name."); m->mothurOutEndLine(); abort = true; }
+
//check for optional parameter and set defaults
// ...at some point should added some additional type checking...
//get user cutoff and precision or use defaults
if (temp == "not found") { temp = "100"; }
//saves precision legnth for formatting below
length = temp.length();
- convert(temp, precision);
+ m->mothurConvert(temp, precision);
temp = validParameter.validFile(parameters, "hard", false); if (temp == "not found") { temp = "T"; }
hard = m->isTrue(temp);
temp = validParameter.validFile(parameters, "sim", false); if (temp == "not found") { temp = "F"; }
sim = m->isTrue(temp);
+ //bool cutoffSet = false;
temp = validParameter.validFile(parameters, "cutoff", false);
if (temp == "not found") { temp = "10"; }
- convert(temp, cutoff);
- cutoff += (5 / (precision * 10.0));
+ //else { cutoffSet = true; }
+ m->mothurConvert(temp, cutoff);
+ cutoff += (5 / (precision * 10.0));
+
+ //temp = validParameter.validFile(parameters, "adjust", false); if (temp == "not found") { temp = "F"; }
+ //if (m->isNumeric1(temp)) { m->mothurConvert(temp, adjust); }
+ //else if (m->isTrue(temp)) { adjust = 1.0; }
+ //else { adjust = -1.0; }
+ adjust=-1.0;
method = validParameter.validFile(parameters, "method", false);
if (method == "not found") { method = "average"; }
timing = validParameter.validFile(parameters, "timing", false);
if (timing == "not found") { timing = "F"; }
- }
+ }
}
catch(exception& e) {
m->errorOut(e, "ClusterCommand", "ClusterCommand");
if (abort == true) { if (calledHelp) { return 0; } return 2; }
+ //phylip file given and cutoff not given - use cluster.classic because it uses less memory and is faster
+ if ((format == "phylip") && (cutoff > 10.0)) {
+ m->mothurOutEndLine(); m->mothurOut("You are using a phylip file and no cutoff. I will run cluster.classic to save memory and time."); m->mothurOutEndLine();
+
+ //run unique.seqs for deconvolute results
+ string inputString = "phylip=" + distfile;
+ if (namefile != "") { inputString += ", name=" + namefile; }
+ else if (countfile != "") { inputString += ", count=" + countfile; }
+ inputString += ", precision=" + toString(precision);
+ inputString += ", method=" + method;
+ if (hard) { inputString += ", hard=T"; }
+ else { inputString += ", hard=F"; }
+ if (sim) { inputString += ", sim=T"; }
+ else { inputString += ", sim=F"; }
+
+
+ m->mothurOutEndLine();
+ m->mothurOut("/------------------------------------------------------------/"); m->mothurOutEndLine();
+ m->mothurOut("Running command: cluster.classic(" + inputString + ")"); m->mothurOutEndLine();
+
+ Command* clusterClassicCommand = new ClusterDoturCommand(inputString);
+ clusterClassicCommand->execute();
+ delete clusterClassicCommand;
+
+ m->mothurOut("/------------------------------------------------------------/"); m->mothurOutEndLine();
+
+ return 0;
+ }
+
ReadMatrix* read;
if (format == "column") { read = new ReadColumnMatrix(columnfile, sim); } //sim indicates whether its a similarity matrix
else if (format == "phylip") { read = new ReadPhylipMatrix(phylipfile, sim); }
read->setCutoff(cutoff);
NameAssignment* nameMap = NULL;
+ CountTable* ct = NULL;
if(namefile != ""){
nameMap = new NameAssignment(namefile);
nameMap->readMap();
- }
+ read->read(nameMap);
+ }else if (countfile != "") {
+ ct = new CountTable();
+ ct->readTable(countfile, false);
+ read->read(ct);
+ }else { read->read(nameMap); }
- read->read(nameMap);
list = read->getListVector();
- matrix = read->getMatrix();
- rabund = new RAbundVector(list->getRAbundVector());
+ matrix = read->getDMatrix();
+
+ if(countfile != "") {
+ rabund = new RAbundVector();
+ createRabund(ct, list, rabund); //creates an rabund that includes the counts for the unique list
+ delete ct;
+ }else { rabund = new RAbundVector(list->getRAbundVector()); }
delete read;
if (m->control_pressed) { //clean up
- delete list; delete matrix; delete rabund;
- sabundFile.close();rabundFile.close();listFile.close();
- for (int i = 0; i < outputNames.size(); i++) { m->mothurRemove(outputNames[i]); } outputTypes.clear();
- return 0;
+ delete list; delete matrix; delete rabund; if(countfile == ""){rabundFile.close(); sabundFile.close(); m->mothurRemove((fileroot+ tag + ".rabund")); m->mothurRemove((fileroot+ tag + ".sabund")); }
+ listFile.close(); m->mothurRemove((fileroot+ tag + ".list")); outputTypes.clear(); return 0;
}
//create cluster
- if (method == "furthest") { cluster = new CompleteLinkage(rabund, list, matrix, cutoff, method); }
- else if(method == "nearest"){ cluster = new SingleLinkage(rabund, list, matrix, cutoff, method); }
- else if(method == "average"){ cluster = new AverageLinkage(rabund, list, matrix, cutoff, method); }
- else if(method == "weighted"){ cluster = new WeightedLinkage(rabund, list, matrix, cutoff, method); }
+ if (method == "furthest") { cluster = new CompleteLinkage(rabund, list, matrix, cutoff, method, adjust); }
+ else if(method == "nearest"){ cluster = new SingleLinkage(rabund, list, matrix, cutoff, method, adjust); }
+ else if(method == "average"){ cluster = new AverageLinkage(rabund, list, matrix, cutoff, method, adjust); }
+ else if(method == "weighted"){ cluster = new WeightedLinkage(rabund, list, matrix, cutoff, method, adjust); }
tag = cluster->getTag();
if (outputDir == "") { outputDir += m->hasPath(distfile); }
fileroot = outputDir + m->getRootName(m->getSimpleName(distfile));
- m->openOutputFile(fileroot+ tag + ".sabund", sabundFile);
- m->openOutputFile(fileroot+ tag + ".rabund", rabundFile);
- m->openOutputFile(fileroot+ tag + ".list", listFile);
-
- outputNames.push_back(fileroot+ tag + ".sabund"); outputTypes["sabund"].push_back(fileroot+ tag + ".sabund");
- outputNames.push_back(fileroot+ tag + ".rabund"); outputTypes["rabund"].push_back(fileroot+ tag + ".rabund");
- outputNames.push_back(fileroot+ tag + ".list"); outputTypes["list"].push_back(fileroot+ tag + ".list");
+ map<string, string> variables;
+ variables["[filename]"] = fileroot;
+ variables["[clustertag]"] = tag;
+ string sabundFileName = getOutputFileName("sabund", variables);
+ string rabundFileName = getOutputFileName("rabund", variables);
+ if (countfile != "") { variables["[tag2]"] = "unique_list"; }
+ string listFileName = getOutputFileName("list", variables);
+
+ if (countfile == "") {
+ m->openOutputFile(sabundFileName, sabundFile);
+ m->openOutputFile(rabundFileName, rabundFile);
+ outputNames.push_back(sabundFileName); outputTypes["sabund"].push_back(sabundFileName);
+ outputNames.push_back(rabundFileName); outputTypes["rabund"].push_back(rabundFileName);
+
+ }
+ m->openOutputFile(listFileName, listFile);
+ outputNames.push_back(listFileName); outputTypes["list"].push_back(listFileName);
time_t estart = time(NULL);
loops = 0;
double saveCutoff = cutoff;
- while (matrix->getSmallDist() < cutoff && matrix->getNNodes() > 0){
+ while (matrix->getSmallDist() < cutoff && matrix->getNNodes() > 0){
if (m->control_pressed) { //clean up
delete list; delete matrix; delete rabund; delete cluster;
- sabundFile.close();rabundFile.close();listFile.close();
- for (int i = 0; i < outputNames.size(); i++) { m->mothurRemove(outputNames[i]); } outputTypes.clear();
- return 0;
+ if(countfile == "") {rabundFile.close(); sabundFile.close(); m->mothurRemove((fileroot+ tag + ".rabund")); m->mothurRemove((fileroot+ tag + ".sabund")); }
+ listFile.close(); m->mothurRemove((fileroot+ tag + ".list")); outputTypes.clear(); return 0;
}
if (print_start && m->isTrue(timing)) {
loops++;
cluster->update(cutoff);
-
- float dist = matrix->getSmallDist();
+
+ float dist = matrix->getSmallDist();
float rndDist;
if (hard) {
rndDist = m->ceilDist(dist, precision);
delete list;
delete rabund;
delete cluster;
-
- sabundFile.close();
- rabundFile.close();
+ if (countfile == "") {
+ sabundFile.close();
+ rabundFile.close();
+ }
listFile.close();
if (saveCutoff != cutoff) {
print_start = true;
loops = 0;
start = time(NULL);
-
- oldRAbund.setLabel(label);
- if (m->isTrue(showabund)) {
- oldRAbund.getSAbundVector().print(cout);
- }
- oldRAbund.print(rabundFile);
- oldRAbund.getSAbundVector().print(sabundFile);
-
+
+ oldRAbund.setLabel(label);
+ if (countfile == "") {
+ oldRAbund.print(rabundFile);
+ oldRAbund.getSAbundVector().print(sabundFile);
+ }
+
+ if (m->isTrue(showabund)) {
+ oldRAbund.getSAbundVector().print(cout);
+ }
+
oldList.setLabel(label);
oldList.print(listFile);
}
}
//**********************************************************************************************************************
+
+int ClusterCommand::createRabund(CountTable*& ct, ListVector*& list, RAbundVector*& rabund){
+ try {
+ rabund->setLabel(list->getLabel());
+ for(int i = 0; i < list->getNumBins(); i++) {
+ if (m->control_pressed) { break; }
+ vector<string> binNames;
+ string bin = list->get(i);
+ m->splitAtComma(bin, binNames);
+ int total = 0;
+ for (int j = 0; j < binNames.size(); j++) { total += ct->getNumSeqs(binNames[j]); }
+ rabund->push_back(total);
+ }
+ return 0;
+ }
+ catch(exception& e) {
+ m->errorOut(e, "ClusterCommand", "createRabund");
+ exit(1);
+ }
+
+}
+//**********************************************************************************************************************
projects / mothur.git / blobdiff