5 * Created by Sarah Westcott on 5/7/09.
6 * Copyright 2009 Schloss Lab UMASS Amherst. All rights reserved.
10 #include "distancecommand.h"
11 #include "ignoregaps.h"
12 #include "eachgapdist.h"
13 #include "eachgapignore.h"
14 #include "onegapdist.h"
15 #include "onegapignore.h"
17 //**********************************************************************************************************************
19 DistanceCommand::DistanceCommand(string option) {
24 //allow user to run help
25 if(option == "help") { help(); abort = true; }
28 //valid paramters for this command
29 string Array[] = {"fasta", "output", "calc", "countends", "cutoff", "processors", "outputdir","inputdir"};
30 vector<string> myArray (Array, Array+(sizeof(Array)/sizeof(string)));
32 OptionParser parser(option);
33 map<string, string> parameters = parser.getParameters();
35 ValidParameters validParameter;
36 map<string, string>::iterator it2;
38 //check to make sure all parameters are valid for command
39 for (it2 = parameters.begin(); it2 != parameters.end(); it2++) {
40 if (validParameter.isValidParameter(it2->first, myArray, it2->second) != true) { abort = true; }
43 //if the user changes the input directory command factory will send this info to us in the output parameter
44 string inputDir = validParameter.validFile(parameters, "inputdir", false);
45 if (inputDir == "not found"){ inputDir = ""; }
48 it2 = parameters.find("fasta");
49 //user has given a template file
50 if(it2 != parameters.end()){
51 path = hasPath(it2->second);
52 //if the user has not given a path then, add inputdir. else leave path alone.
53 if (path == "") { parameters["fasta"] = inputDir + it2->second; }
57 //check for required parameters
58 fastafile = validParameter.validFile(parameters, "fasta", true);
59 if (fastafile == "not found") { m->mothurOut("fasta is a required parameter for the dist.seqs command."); m->mothurOutEndLine(); abort = true; }
60 else if (fastafile == "not open") { abort = true; }
63 openInputFile(fastafile, inFASTA);
64 alignDB = SequenceDB(inFASTA);
68 //if the user changes the output directory command factory will send this info to us in the output parameter
69 outputDir = validParameter.validFile(parameters, "outputdir", false); if (outputDir == "not found"){
71 outputDir += hasPath(fastafile); //if user entered a file with a path then preserve it
74 //check for optional parameter and set defaults
75 // ...at some point should added some additional type checking...
76 calc = validParameter.validFile(parameters, "calc", false);
77 if (calc == "not found") { calc = "onegap"; }
79 if (calc == "default") { calc = "onegap"; }
81 splitAtDash(calc, Estimators);
84 temp = validParameter.validFile(parameters, "countends", false); if(temp == "not found"){ temp = "T"; }
85 convert(temp, countends);
87 temp = validParameter.validFile(parameters, "cutoff", false); if(temp == "not found"){ temp = "1.0"; }
88 convert(temp, cutoff);
90 temp = validParameter.validFile(parameters, "processors", false); if(temp == "not found"){ temp = "1"; }
91 convert(temp, processors);
93 output = validParameter.validFile(parameters, "output", false); if(output == "not found"){ output = "column"; }
95 if ((output != "column") && (output != "lt") && (output != "square")) { m->mothurOut(output + " is not a valid output form. Options are column, lt and square. I will use column."); m->mothurOutEndLine(); output = "column"; }
97 ValidCalculators validCalculator;
99 if (isTrue(countends) == true) {
100 for (int i=0; i<Estimators.size(); i++) {
101 if (validCalculator.isValidCalculator("distance", Estimators[i]) == true) {
102 if (Estimators[i] == "nogaps") { distCalculator = new ignoreGaps(); }
103 else if (Estimators[i] == "eachgap") { distCalculator = new eachGapDist(); }
104 else if (Estimators[i] == "onegap") { distCalculator = new oneGapDist(); }
108 for (int i=0; i<Estimators.size(); i++) {
109 if (validCalculator.isValidCalculator("distance", Estimators[i]) == true) {
110 if (Estimators[i] == "nogaps") { distCalculator = new ignoreGaps(); }
111 else if (Estimators[i] == "eachgap"){ distCalculator = new eachGapIgnoreTermGapDist(); }
112 else if (Estimators[i] == "onegap") { distCalculator = new oneGapIgnoreTermGapDist(); }
120 catch(exception& e) {
121 m->errorOut(e, "DistanceCommand", "DistanceCommand");
126 //**********************************************************************************************************************
128 DistanceCommand::~DistanceCommand(){
130 for(int i=0;i<lines.size();i++){
136 //**********************************************************************************************************************
138 void DistanceCommand::help(){
140 m->mothurOut("The dist.seqs command reads a file containing sequences and creates a distance file.\n");
141 m->mothurOut("The dist.seqs command parameters are fasta, calc, countends, output, cutoff and processors. \n");
142 m->mothurOut("The fasta parameter is required.\n");
143 m->mothurOut("The calc parameter allows you to specify the method of calculating the distances. Your options are: nogaps, onegap or eachgap. The default is onegap.\n");
144 m->mothurOut("The countends parameter allows you to specify whether to include terminal gaps in distance. Your options are: T or F. The default is T.\n");
145 m->mothurOut("The cutoff parameter allows you to specify maximum distance to keep. The default is 1.0.\n");
146 m->mothurOut("The output parameter allows you to specify format of your distance matrix. Options are column, lt, and square. The default is column.\n");
147 m->mothurOut("The processors parameter allows you to specify number of processors to use. The default is 1.\n");
148 m->mothurOut("The dist.seqs command should be in the following format: \n");
149 m->mothurOut("dist.seqs(fasta=yourFastaFile, calc=yourCalc, countends=yourEnds, cutoff= yourCutOff, processors=yourProcessors) \n");
150 m->mothurOut("Example dist.seqs(fasta=amazon.fasta, calc=eachgap, countends=F, cutoff= 2.0, processors=3).\n");
151 m->mothurOut("Note: No spaces between parameter labels (i.e. calc), '=' and parameters (i.e.yourCalc).\n\n");
153 catch(exception& e) {
154 m->errorOut(e, "DistanceCommand", "help");
158 //**********************************************************************************************************************
160 int DistanceCommand::execute(){
163 if (abort == true) { return 0; }
165 int numSeqs = alignDB.getNumSeqs();
170 if (output == "lt") { //does the user want lower triangle phylip formatted file
171 outputFile = outputDir + getRootName(getSimpleName(fastafile)) + "phylip.dist";
172 remove(outputFile.c_str());
174 //output numSeqs to phylip formatted dist file
175 }else if (output == "column") { //user wants column format
176 outputFile = outputDir + getRootName(getSimpleName(fastafile)) + "dist";
177 remove(outputFile.c_str());
178 }else { //assume square
179 outputFile = outputDir + getRootName(getSimpleName(fastafile)) + "square.dist";
180 remove(outputFile.c_str());
183 #if defined (__APPLE__) || (__MACH__) || (linux) || (__linux)
184 //if you don't need to fork anything
186 driver(0, numSeqs, outputFile, cutoff);
187 }else{ //you have multiple processors
189 for (int i = 0; i < processors; i++) {
190 lines.push_back(new linePair());
191 lines[i]->start = int (sqrt(float(i)/float(processors)) * numSeqs);
192 lines[i]->end = int (sqrt(float(i+1)/float(processors)) * numSeqs);
195 createProcesses(outputFile);
197 map<int, int>::iterator it = processIDS.begin();
198 rename((outputFile + toString(it->second) + ".temp").c_str(), outputFile.c_str());
201 //append and remove temp files
202 for (; it != processIDS.end(); it++) {
203 appendFiles((outputFile + toString(it->second) + ".temp"), outputFile);
204 remove((outputFile + toString(it->second) + ".temp").c_str());
209 driver(0, numSeqs, outputFile, cutoff);
211 if (m->control_pressed) { delete distCalculator; remove(outputFile.c_str()); return 0; }
213 if (output == "square") { convertMatrix(outputFile); }
215 if (m->control_pressed) { delete distCalculator; remove(outputFile.c_str()); return 0; }
217 delete distCalculator;
219 m->mothurOutEndLine();
220 m->mothurOut("Output File Name: "); m->mothurOutEndLine();
221 m->mothurOut(outputFile); m->mothurOutEndLine();
222 m->mothurOutEndLine();
227 catch(exception& e) {
228 m->errorOut(e, "DistanceCommand", "execute");
232 /**************************************************************************************************/
233 void DistanceCommand::createProcesses(string filename) {
235 #if defined (__APPLE__) || (__MACH__) || (linux) || (__linux)
239 //loop through and create all the processes you want
240 while (process != processors) {
244 processIDS[lines[process]->end] = pid; //create map from line number to pid so you can append files in correct order later
247 driver(lines[process]->start, lines[process]->end, filename + toString(getpid()) + ".temp", cutoff);
249 }else { m->mothurOut("unable to spawn the necessary processes."); m->mothurOutEndLine(); exit(0); }
252 //force parent to wait until all the processes are done
253 for (map<int, int>::iterator it = processIDS.begin(); it != processIDS.end(); it++) {
254 int temp = it->second;
259 catch(exception& e) {
260 m->errorOut(e, "DistanceCommand", "createProcesses");
265 /**************************************************************************************************/
266 /////// need to fix to work with calcs and sequencedb
267 int DistanceCommand::driver(int startLine, int endLine, string dFileName, float cutoff){
270 int startTime = time(NULL);
273 ofstream outFile(dFileName.c_str(), ios::trunc);
274 outFile.setf(ios::fixed, ios::showpoint);
275 outFile << setprecision(4);
277 if((output == "lt") && startLine == 0){ outFile << alignDB.getNumSeqs() << endl; }
279 for(int i=startLine;i<endLine;i++){
281 string name = alignDB.get(i).getName();
282 if (name.length() < 10) { //pad with spaces to make compatible
283 while (name.length() < 10) { name += " "; }
285 outFile << name << '\t';
287 for(int j=0;j<i;j++){
289 if (m->control_pressed) { outFile.close(); return 0; }
291 distCalculator->calcDist(alignDB.get(i), alignDB.get(j));
292 double dist = distCalculator->getDist();
295 if (output == "column") { outFile << alignDB.get(i).getName() << ' ' << alignDB.get(j).getName() << ' ' << dist << endl; }
297 if (output == "lt") { outFile << dist << '\t'; }
299 if (output == "square") { //make a square column you can convert to square phylip
300 outFile << alignDB.get(i).getName() << '\t' << alignDB.get(j).getName() << '\t' << dist << endl;
301 outFile << alignDB.get(j).getName() << '\t' << alignDB.get(i).getName() << '\t' << dist << endl;
306 if (output == "lt") { outFile << endl; }
309 m->mothurOut(toString(i) + "\t" + toString(time(NULL) - startTime)); m->mothurOutEndLine();
313 m->mothurOut(toString(endLine-1) + "\t" + toString(time(NULL) - startTime)); m->mothurOutEndLine();
319 catch(exception& e) {
320 m->errorOut(e, "DistanceCommand", "driver");
324 /**************************************************************************************************/
325 int DistanceCommand::convertMatrix(string outputFile) {
328 //sort file by first column so the distances for each row are together
329 string outfile = getRootName(outputFile) + "sorted.dist.temp";
332 #if defined (__APPLE__) || (__MACH__) || (linux) || (__linux)
333 string command = "sort -n " + outputFile + " -o " + outfile;
334 system(command.c_str());
335 #else //sort using windows sort
336 string command = "sort " + outputFile + " /O " + outfile;
337 system(command.c_str());
341 //output to new file distance for each row and save positions in file where new row begins
343 openInputFile(outfile, in);
346 openOutputFile(outputFile, out);
348 out.setf(ios::fixed, ios::floatfield); out.setf(ios::showpoint);
350 out << alignDB.getNumSeqs() << endl;
352 //get first currentRow
353 string first, currentRow, second;
355 map<string, float> rowDists; //take advantage of the fact that maps are already sorted by key
356 map<string, float>::iterator it;
361 rowDists[first] = 0.00; //distance to yourself is 0.0
364 //openInputFile(outfile, in);
367 if (m->control_pressed) { in.close(); remove(outfile.c_str()); out.close(); return 0; }
369 in >> first >> second >> dist; gobble(in);
371 if (first != currentRow) {
373 out << currentRow << '\t'; //print name
376 for (it = rowDists.begin(); it != rowDists.end(); it++) {
377 out << it->second << '\t';
384 rowDists[first] = 0.00;
385 rowDists[second] = dist;
387 rowDists[second] = dist;
391 out << currentRow << '\t'; //print name
394 for (it = rowDists.begin(); it != rowDists.end(); it++) {
395 out << it->second << '\t';
402 remove(outfile.c_str());
407 catch(exception& e) {
408 m->errorOut(e, "DistanceCommand", "convertMatrix");
412 /**************************************************************************************************
413 void DistanceCommand::appendFiles(string temp, string filename) {
418 //open output file in append mode
419 openOutputFileAppend(filename, output);
420 openInputFile(temp, input);
422 while(char c = input.get()){
423 if(input.eof()) { break; }
424 else { output << c; }
430 catch(exception& e) {
431 m->errorOut(e, "DistanceCommand", "appendFiles");
435 /**************************************************************************************************/