else {
//valid paramters for this command
- string Array[] = {"phylip","column","name","cutoff","precision","method","showabund","timing","hard","processors","splitcutoff","outputdir","inputdir"};
+ string Array[] = {"phylip","column","name","cutoff","precision","method","splitmethod","taxonomy","taxlevel","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["name"] = inputDir + it->second; }
}
+
+ it = parameters.find("taxonomy");
+ //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["taxonomy"] = inputDir + it->second; }
+ }
}
//check for required parameters
if (namefile == "not open") { abort = true; }
else if (namefile == "not found") { namefile = ""; }
- if ((phylipfile == "") && (columnfile == "")) { m->mothurOut("When executing a hcluster command you must enter a phylip or a column."); m->mothurOutEndLine(); abort = true; }
- 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; }
+ 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 (columnfile != "") {
- if (namefile == "") { cout << "You need to provide a namefile if you are going to use the column format." << endl; abort = true; }
+ if (namefile == "") { m->mothurOut("You need to provide a namefile if you are going to use the column format."); m->mothurOutEndLine(); abort = true; }
}
//check for optional parameter and set defaults
temp = validParameter.validFile(parameters, "processors", false); if (temp == "not found"){ temp = "1"; }
convert(temp, processors);
- temp = validParameter.validFile(parameters, "cutoff", false);
- if (temp == "not found") { temp = "10"; }
+ splitmethod = validParameter.validFile(parameters, "splitmethod", false); if (splitmethod == "not found") { splitmethod = "distance"; }
+
+ temp = validParameter.validFile(parameters, "cutoff", false); if (temp == "not found") { temp = "10"; }
convert(temp, cutoff);
- if (!hard) { cutoff += (5 / (precision * 10.0)); }
+ cutoff += (5 / (precision * 10.0));
- temp = validParameter.validFile(parameters, "splitcutoff", false);
- if (temp == "not found") { temp = "0.10"; }
- convert(temp, splitcutoff);
- if (!hard) { splitcutoff += (5 / (precision * 10.0)); }
+ temp = validParameter.validFile(parameters, "taxlevel", false); if (temp == "not found") { temp = "1"; }
+ convert(temp, taxLevelCutoff);
- method = validParameter.validFile(parameters, "method", false);
- if (method == "not found") { method = "furthest"; }
+ method = validParameter.validFile(parameters, "method", false); if (method == "not found") { method = "furthest"; }
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 == "classify") && (taxFile == "")) { m->mothurOut("You need to provide a taxonomy file if you are going to use the classify splitmethod."); m->mothurOutEndLine(); abort = true; }
showabund = validParameter.validFile(parameters, "showabund", false);
if (showabund == "not found") { showabund = "T"; }
void ClusterSplitCommand::help(){
try {
- m->mothurOut("The cluster command can only be executed after a successful read.dist command.\n");
- m->mothurOut("The cluster command parameter options are method, cuttoff, hard, precision, showabund and timing. No parameters are required.\n");
- m->mothurOut("The cluster command should be in the following format: \n");
- m->mothurOut("cluster(method=yourMethod, cutoff=yourCutoff, precision=yourPrecision) \n");
- m->mothurOut("The acceptable cluster methods are furthest, nearest and average. If no method is provided then furthest is assumed.\n\n");
+ 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 phylip and column parameter allow you to enter your distance 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 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 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");
+
}
catch(exception& e) {
m->errorOut(e, "ClusterSplitCommand", "help");
//if user gave a phylip file convert to column file
if (format == "phylip") {
+
ReadCluster* convert = new ReadCluster(distfile, cutoff, outputDir, false);
NameAssignment* nameMap = NULL;
+ convert->setFormat("phylip");
convert->read(nameMap);
if (m->control_pressed) { delete convert; return 0; }
- string distfile = convert->getOutputFile();
-
+ distfile = convert->getOutputFile();
+
//if no names file given with phylip file, create it
ListVector* listToMakeNameFile = convert->getListVector();
if (namefile == "") { //you need to make a namefile for split matrix
delete listToMakeNameFile;
delete convert;
}
-
if (m->control_pressed) { return 0; }
time_t estart = time(NULL);
//split matrix into non-overlapping groups
- SplitMatrix* split = new SplitMatrix(distfile, namefile, splitcutoff);
+ SplitMatrix* split;
+ if (splitmethod == "distance") { split = new SplitMatrix(distfile, namefile, taxFile, cutoff, splitmethod); }
+ else { split = new SplitMatrix(distfile, namefile, taxFile, taxLevelCutoff, splitmethod); }
+
split->split();
if (m->control_pressed) { delete split; return 0; }
vector< map<string, string> > distName = split->getDistanceFiles(); //returns map of distance files -> namefile sorted by distance file size
delete split;
+ if (m->control_pressed) { return 0; }
+
m->mothurOut("It took " + toString(time(NULL) - estart) + " seconds to split the distance file."); m->mothurOutEndLine();
estart = time(NULL);
- if (m->control_pressed) { return 0; }
-
//****************** break up files between processes and cluster each file set ******************************//
vector<string> listFileNames;
set<string> labels;
dividedNames[(processToAssign-1)].push_back(distName[i]);
}
+ //not lets reverse the order of ever other process, so we balance big files running with little ones
+ for (int i = 0; i < processors; i++) {
+ int remainder = ((i+1) % processors);
+ if (remainder) { reverse(dividedNames[i].begin(), dividedNames[i].end()); }
+ }
+
createProcesses(dividedNames);
if (m->control_pressed) { return 0; }
outputNames.push_back(fileroot+ tag + ".sabund");
outputNames.push_back(fileroot+ tag + ".rabund");
outputNames.push_back(fileroot+ tag + ".list");
-
+
//read in singletons
ListVector* listSingle = NULL;
if (singleton != "none") {
it--;
}
}
-
+
//sort order
sort(orderFloat.begin(), orderFloat.end());
lastLabels.push_back(tempList.getLabel());
in.close();
}
-
+
ListVector* merged = NULL;
//for each label needed
//get the list info from each file
for (int k = 0; k < listNames.size(); k++) {
+ 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();
//this file has reached the end
}
//add in singletons
- for (int j = 0; j < listSingle->getNumBins(); j++) {
- merged->push_back(listSingle->get(j));
+ if (listSingle != NULL) {
+ for (int j = 0; j < listSingle->getNumBins(); j++) {
+ merged->push_back(listSingle->get(j));
+ }
}
-
+
//print to files
printData(merged);
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()); }
+
+ return 0;
}
catch(exception& e) {
m->errorOut(e, "ClusterSplitCommand", "mergeLists");
cluster->update(cutoff);
float dist = matrix->getSmallDist();
- float rndDist = roundDist(dist, precision);
+ float rndDist;
+ if (hard) {
+ rndDist = ceilDist(dist, precision);
+ }else{
+ rndDist = roundDist(dist, precision);
+ }
if(previousDist <= 0.0000 && dist != previousDist){
oldList.setLabel("unique");
oldList.print(listFile);
if (labels.count(toString(rndPreviousDist, length-1)) == 0) { labels.insert(toString(rndPreviousDist, length-1)); }
}
-
+
delete matrix; delete list; delete cluster; delete rabund;
listFile.close();
remove(thisNamefile.c_str());
}
-
+
return listFileNames;
}
projects / mothur.git / blobdiff