5 * Created by westcott on 10/13/09.
6 * Copyright 2009 Schloss Lab. All rights reserved.
10 #include "hclustercommand.h"
12 //**********************************************************************************************************************
13 vector<string> HClusterCommand::getValidParameters(){
15 string Array[] = {"cutoff","hard","precision","method","phylip","column","name","sorted","showabund","timing","outputdir","inputdir"};
16 vector<string> myArray (Array, Array+(sizeof(Array)/sizeof(string)));
20 m->errorOut(e, "HClusterCommand", "getValidParameters");
24 //**********************************************************************************************************************
25 HClusterCommand::HClusterCommand(){
27 //initialize outputTypes
28 vector<string> tempOutNames;
29 outputTypes["list"] = tempOutNames;
30 outputTypes["rabund"] = tempOutNames;
31 outputTypes["sabund"] = tempOutNames;
34 m->errorOut(e, "HClusterCommand", "HClusterCommand");
38 //**********************************************************************************************************************
39 vector<string> HClusterCommand::getRequiredParameters(){
41 string Array[] = {"phylip","column","or"};
42 vector<string> myArray (Array, Array+(sizeof(Array)/sizeof(string)));
46 m->errorOut(e, "HClusterCommand", "getRequiredParameters");
50 //**********************************************************************************************************************
51 vector<string> HClusterCommand::getRequiredFiles(){
53 vector<string> myArray;
57 m->errorOut(e, "HClusterCommand", "getRequiredFiles");
61 //**********************************************************************************************************************
62 //This function checks to make sure the cluster command has no errors and then clusters based on the method chosen.
63 HClusterCommand::HClusterCommand(string option) {
65 globaldata = GlobalData::getInstance();
68 //allow user to run help
69 if(option == "help") { help(); abort = true; }
72 //valid paramters for this command
73 string Array[] = {"cutoff","hard","precision","method","phylip","column","name","sorted","showabund","timing","outputdir","inputdir"};
74 vector<string> myArray (Array, Array+(sizeof(Array)/sizeof(string)));
76 OptionParser parser(option);
77 map<string,string> parameters = parser.getParameters();
79 ValidParameters validParameter;
80 map<string,string>::iterator it;
82 //check to make sure all parameters are valid for command
83 for (it = parameters.begin(); it != parameters.end(); it++) {
84 if (validParameter.isValidParameter(it->first, myArray, it->second) != true) {
89 //initialize outputTypes
90 vector<string> tempOutNames;
91 outputTypes["list"] = tempOutNames;
92 outputTypes["rabund"] = tempOutNames;
93 outputTypes["sabund"] = tempOutNames;
95 globaldata->newRead();
97 //if the user changes the input directory command factory will send this info to us in the output parameter
98 string inputDir = validParameter.validFile(parameters, "inputdir", false);
99 if (inputDir == "not found"){ inputDir = ""; }
102 it = parameters.find("phylip");
103 //user has given a template file
104 if(it != parameters.end()){
105 path = m->hasPath(it->second);
106 //if the user has not given a path then, add inputdir. else leave path alone.
107 if (path == "") { parameters["phylip"] = inputDir + it->second; }
110 it = parameters.find("column");
111 //user has given a template file
112 if(it != parameters.end()){
113 path = m->hasPath(it->second);
114 //if the user has not given a path then, add inputdir. else leave path alone.
115 if (path == "") { parameters["column"] = inputDir + it->second; }
118 it = parameters.find("name");
119 //user has given a template file
120 if(it != parameters.end()){
121 path = m->hasPath(it->second);
122 //if the user has not given a path then, add inputdir. else leave path alone.
123 if (path == "") { parameters["name"] = inputDir + it->second; }
127 //if the user changes the output directory command factory will send this info to us in the output parameter
128 outputDir = validParameter.validFile(parameters, "outputdir", false); if (outputDir == "not found"){ outputDir = ""; }
130 //check for required parameters
131 phylipfile = validParameter.validFile(parameters, "phylip", true);
132 if (phylipfile == "not open") { abort = true; }
133 else if (phylipfile == "not found") { phylipfile = ""; }
134 else { distfile = phylipfile; format = "phylip"; }
136 columnfile = validParameter.validFile(parameters, "column", true);
137 if (columnfile == "not open") { abort = true; }
138 else if (columnfile == "not found") { columnfile = ""; }
139 else { distfile = columnfile; format = "column"; }
141 namefile = validParameter.validFile(parameters, "name", true);
142 if (namefile == "not open") { abort = true; }
143 else if (namefile == "not found") { namefile = ""; }
145 if ((phylipfile == "") && (columnfile == "")) { m->mothurOut("When executing a hcluster command you must enter a phylip or a column."); m->mothurOutEndLine(); abort = true; }
146 else if ((phylipfile != "") && (columnfile != "")) { m->mothurOut("When executing a hcluster command you must enter ONLY ONE of the following: phylip or column."); m->mothurOutEndLine(); abort = true; }
148 if (columnfile != "") {
149 if (namefile == "") { cout << "You need to provide a namefile if you are going to use the column format." << endl; abort = true; }
152 //check for optional parameter and set defaults
153 // ...at some point should added some additional type checking...
154 //get user cutoff and precision or use defaults
156 temp = validParameter.validFile(parameters, "precision", false);
157 if (temp == "not found") { temp = "100"; }
158 //saves precision legnth for formatting below
159 length = temp.length();
160 convert(temp, precision);
162 temp = validParameter.validFile(parameters, "hard", false); if (temp == "not found") { temp = "F"; }
163 hard = m->isTrue(temp);
165 temp = validParameter.validFile(parameters, "cutoff", false);
166 if (temp == "not found") { temp = "10"; }
167 convert(temp, cutoff);
168 cutoff += (5 / (precision * 10.0));
170 method = validParameter.validFile(parameters, "method", false);
171 if (method == "not found") { method = "furthest"; }
173 if ((method == "furthest") || (method == "nearest") || (method == "average")) { }
174 else { m->mothurOut("Not a valid clustering method. Valid clustering algorithms are furthest, nearest or average."); m->mothurOutEndLine(); abort = true; }
176 showabund = validParameter.validFile(parameters, "showabund", false);
177 if (showabund == "not found") { showabund = "T"; }
179 sort = validParameter.validFile(parameters, "sorted", false);
180 if (sort == "not found") { sort = "F"; }
181 sorted = m->isTrue(sort);
183 timing = validParameter.validFile(parameters, "timing", false);
184 if (timing == "not found") { timing = "F"; }
187 if (abort == false) {
189 if (outputDir == "") { outputDir += m->hasPath(distfile); }
190 fileroot = outputDir + m->getRootName(m->getSimpleName(distfile));
192 if (method == "furthest") { tag = "fn"; }
193 else if (method == "nearest") { tag = "nn"; }
196 m->openOutputFile(fileroot+ tag + ".sabund", sabundFile);
197 m->openOutputFile(fileroot+ tag + ".rabund", rabundFile);
198 m->openOutputFile(fileroot+ tag + ".list", listFile);
200 outputNames.push_back(fileroot+ tag + ".sabund"); outputTypes["sabund"].push_back(fileroot+ tag + ".sabund");
201 outputNames.push_back(fileroot+ tag + ".rabund"); outputTypes["rabund"].push_back(fileroot+ tag + ".rabund");
202 outputNames.push_back(fileroot+ tag + ".list"); outputTypes["list"].push_back(fileroot+ tag + ".list");
206 catch(exception& e) {
207 m->errorOut(e, "HClusterCommand", "HClusterCommand");
212 //**********************************************************************************************************************
214 void HClusterCommand::help(){
216 m->mothurOut("The hcluster command parameter options are cutoff, precision, method, phylip, column, name, showabund, timing and sorted. Phylip or column and name are required.\n");
217 m->mothurOut("The phylip and column parameter allow you to enter your distance file, and sorted indicates whether your column distance file is already sorted. \n");
218 m->mothurOut("The name parameter allows you to enter your name file and is required if your distance file is in column format. \n");
219 m->mothurOut("The hcluster command should be in the following format: \n");
220 m->mothurOut("hcluster(column=youDistanceFile, name=yourNameFile, method=yourMethod, cutoff=yourCutoff, precision=yourPrecision) \n");
221 m->mothurOut("The acceptable hcluster methods are furthest, nearest and average.\n\n");
223 catch(exception& e) {
224 m->errorOut(e, "HClusterCommand", "help");
229 //**********************************************************************************************************************
231 HClusterCommand::~HClusterCommand(){}
233 //**********************************************************************************************************************
235 int HClusterCommand::execute(){
238 if (abort == true) { return 0; }
241 globaldata->nameMap = new NameAssignment(namefile);
242 globaldata->nameMap->readMap();
244 globaldata->nameMap = NULL;
247 time_t estart = time(NULL);
250 read = new ReadCluster(distfile, cutoff, outputDir, true);
251 read->setFormat(format);
252 read->read(globaldata->nameMap);
254 if (m->control_pressed) {
259 for (int i = 0; i < outputNames.size(); i++) { remove(outputNames[i].c_str()); } outputTypes.clear();
263 distfile = read->getOutputFile();
265 list = read->getListVector();
268 list = new ListVector(globaldata->nameMap->getListVector());
271 if (m->control_pressed) {
275 for (int i = 0; i < outputNames.size(); i++) { remove(outputNames[i].c_str()); } outputTypes.clear();
279 m->mothurOut("It took " + toString(time(NULL) - estart) + " seconds to sort. "); m->mothurOutEndLine();
282 //list vector made by read contains all sequence names
284 rabund = new RAbundVector(list->getRAbundVector());
286 m->mothurOut("Error: no list vector!"); m->mothurOutEndLine(); return 0;
289 float previousDist = 0.00000;
290 float rndPreviousDist = 0.00000;
297 cluster = new HCluster(rabund, list, method, distfile, globaldata->nameMap, cutoff);
298 vector<seqDist> seqs; seqs.resize(1); // to start loop
300 if (m->control_pressed) {
305 for (int i = 0; i < outputNames.size(); i++) { remove(outputNames[i].c_str()); } outputTypes.clear();
310 while (seqs.size() != 0){
312 seqs = cluster->getSeqs();
314 if (m->control_pressed) {
319 for (int i = 0; i < outputNames.size(); i++) { remove(outputNames[i].c_str()); } outputTypes.clear();
323 for (int i = 0; i < seqs.size(); i++) { //-1 means skip me
325 if (seqs[i].seq1 != seqs[i].seq2) {
326 cluster->update(seqs[i].seq1, seqs[i].seq2, seqs[i].dist);
328 if (m->control_pressed) {
333 for (int i = 0; i < outputNames.size(); i++) { remove(outputNames[i].c_str()); } outputTypes.clear();
340 rndDist = m->ceilDist(seqs[i].dist, precision);
342 rndDist = m->roundDist(seqs[i].dist, precision);
346 if((previousDist <= 0.0000) && (seqs[i].dist != previousDist)){
349 else if((rndDist != rndPreviousDist)){
350 printData(toString(rndPreviousDist, length-1));
353 previousDist = seqs[i].dist;
354 rndPreviousDist = rndDist;
361 if (m->control_pressed) {
366 for (int i = 0; i < outputNames.size(); i++) { remove(outputNames[i].c_str()); } outputTypes.clear();
370 if(previousDist <= 0.0000){
373 else if(rndPreviousDist<cutoff){
374 printData(toString(rndPreviousDist, length-1));
377 //delete globaldata's copy of the sparsematrix and listvector to free up memory
378 delete globaldata->gListVector; globaldata->gListVector = NULL;
380 //saves .list file so you can do the collect, rarefaction and summary commands without doing a read.list
381 if (globaldata->getFormat() == "phylip") { globaldata->setPhylipFile(""); }
382 else if (globaldata->getFormat() == "column") { globaldata->setColumnFile(""); }
384 globaldata->setListFile(fileroot+ tag + ".list");
385 globaldata->setNameFile("");
386 globaldata->setFormat("list");
393 if (m->control_pressed) {
394 for (int i = 0; i < outputNames.size(); i++) { remove(outputNames[i].c_str()); } outputTypes.clear();
399 m->mothurOutEndLine();
400 m->mothurOut("Output File Names: "); m->mothurOutEndLine();
401 for (int i = 0; i < outputNames.size(); i++) { m->mothurOut(outputNames[i]); m->mothurOutEndLine(); }
402 m->mothurOutEndLine();
404 m->mothurOut("It took " + toString(time(NULL) - estart) + " seconds to cluster. "); m->mothurOutEndLine();
408 catch(exception& e) {
409 m->errorOut(e, "HClusterCommand", "execute");
414 //**********************************************************************************************************************
416 void HClusterCommand::printData(string label){
418 if (m->isTrue(timing)) {
419 m->mothurOut("\tTime: " + toString(time(NULL) - start) + "\tsecs for " + toString(oldRAbund.getNumBins())
420 + "\tclusters. Updates: " + toString(loops)); m->mothurOutEndLine();
426 oldRAbund.setLabel(label);
427 if (m->isTrue(showabund)) {
428 oldRAbund.getSAbundVector().print(cout);
430 oldRAbund.print(rabundFile);
431 oldRAbund.getSAbundVector().print(sabundFile);
433 oldList.setLabel(label);
434 oldList.print(listFile);
436 catch(exception& e) {
437 m->errorOut(e, "HClusterCommand", "printData");
443 //**********************************************************************************************************************