Java4Models/src/models/Worker.java at master · Saulus/Java4Models

721 lines (687 loc) · 29.6 KB
package models;
import java.io.File;
import java.io.FileReader;
import java.io.FileWriter;
import java.util.ArrayList;
import java.util.Arrays;
import java.util.Collections;
import java.util.HashMap;
import java.util.List;
import java.util.logging.Level;
import java.util.logging.Logger;
import configuration.Consts;
import configuration.Datafile;
import configuration.Filter;
import configuration.Configuration;
import configuration.Utils;
import ddi.DDIworker;
import au.com.bytecode.opencsv.CSVReader;
import au.com.bytecode.opencsv.CSVWriter;
 * The Class Worker.
 * This has three steps:
 * 	Create & Open Inputfile-Objects, incl. sorting if necessary
 * 	Create and Import Models from Model.config and Model.coeff
 * 	create and open output Model Score file
 * 	create and open output ProfilDense file (if configured)
 * 	create and open output ProfilSparse files (if configured)
 * 2. Process
 * 	Move through all inputfiles, patient by patient
 * 	Create patient object with variable/model information
 * 	save patient score (and profile) to disk
 * 3. Finish
 *  Close files, delete temps
public class Worker {
	private final static Logger LOGGER = Logger.getLogger(Worker.class.getName());
	/** The models. */
	private ArrayList<Model> models = new ArrayList<Model>();
	/** The inputfiles. */
	private ArrayList<InputFile> inputfiles = new ArrayList<InputFile>();
	/** The config. */
	private Configuration config;
	private InputFile ourLeaderfile = null;
	/** The outputfile. */
	private CSVWriter outputfile; 
	private boolean ddimode = false;
	private DDIworker ddiworker = null;
	/** The profildensefiles */
	private HashMap<Model,CSVWriter> profildensefile = new HashMap<Model,CSVWriter>();
	//private HashMap<Model,CSVWriter> profildensefile_targets = new HashMap<Model,CSVWriter>();
	/** Matlab format */
	private HashMap<Model,CSVWriter> profilsparsefile = new HashMap<Model,CSVWriter>();
	private HashMap<Model,CSVWriter> profilsparsefileRows = new HashMap<Model,CSVWriter>();
	//private HashMap<Model,CSVWriter> profilsparsefile_targets = new HashMap<Model,CSVWriter>();
	/** Svmlight format */
	private HashMap<Model,CSVWriter> profilsvmlightfile = new HashMap<Model,CSVWriter>();
	//private HashMap<Model,CSVWriter> profilsvmlightfile_targets = new HashMap<Model,CSVWriter>();
	/** The known variables. */
	private HashMap<Model,ArrayList<String>> knownVariables = new HashMap<Model,ArrayList<String>>();//holds known variables per Model    
	//private HashMap<Model,ArrayList<String>> knownVariables_targets = new HashMap<Model,ArrayList<String>>();//holds known variables per Model (targets only)
	 * Find next patient. Uses either the first available ID from all files (sorted by ID),
	 * or in case of leaderfile the next id from that... warps back all other files where last id fits
	 * @return the patient
	private Patient findNextPatient() {
		Patient newpatient= null;
		if (ourLeaderfile == null) { 
			// traditional work through; pid by pid
			List<String> pidliste = new ArrayList<String>();
			for (InputFile infile : inputfiles) {
				if (infile.hasRow()) 
					pidliste.add(infile.getID());
			if (!pidliste.isEmpty()) {
				Collections.sort(pidliste);
				newpatient= new Patient(pidliste.get(0));
			//leaderfile: row by row in that file
			//warp back all others
			if (ourLeaderfile.hasRow()) {
				newpatient= new Patient(ourLeaderfile.getID());
				for (InputFile infile : inputfiles) {
						if (infile != ourLeaderfile) infile.warpToCorrectID(newpatient.getPid());
					} catch (Exception e) {
						LOGGER.log(Level.WARNING,e.getMessage());
		return newpatient;
	 * Next row all files.
	 * @throws Exception 
	private void nextRowAllFiles () {
		for (InputFile infile : inputfiles) {
				infile.nextRow(true,ourLeaderfile != null);
			} catch (Exception e) {
				LOGGER.log(Level.WARNING,"",e);
	 * Instantiates a new worker.
	 * @param config the config
	public Worker (Configuration config) {
		this.config=config;
	 * Inits the.
	 * @return true, if successful
	public boolean init () {
		//1. Create & Open Inputfiles
		boolean worked = false; 
		ArrayList<Filter> filefilters; 
		for (Datafile nextfile : config.getInputfiles()) {
			filefilters = config.getFilters4File(nextfile.getDatentyp());
			//Sort if isSorted=false
			if (!nextfile.isSorted()) {
					Sorter sortfile = new Sorter(nextfile.getDatentyp(),nextfile.getPath(),nextfile.getFiletype(),nextfile.getIdfeld(),nextfile.getSeparator(),nextfile.getQuote(),config.getTmpPath(),filefilters);
					filefilters=null;
					LOGGER.log(Level.INFO,"Sortieren gestartet von " + nextfile.getPath());
					String newpath;
					//add new columns from addinfo?
					if (nextfile.hasZusatzinfo()) {
						CSVReader reader = new CSVReader(new FileReader(nextfile.getZusatzinfo()), ';', '"');
						List<String[]> myEntries = reader.readAll();
						reader.close();
						//first line = header-line
						String[] headerline = myEntries.get(0); myEntries.remove(0);
						for (int j=0; j<headerline.length; j++) { headerline[j] = headerline[j].toUpperCase(); }
						//create HashMap for easy translation
						HashMap<String,String[]> translator = new HashMap<String,String[]>(); 
						for (String[] nextline : myEntries) {
							translator.put(nextline[0], Arrays.copyOfRange(nextline, 1, nextline.length));
						newpath = sortfile.sortFileByID(true,headerline[0],Arrays.copyOfRange(headerline, 1, headerline.length),translator);
					} else newpath = sortfile.sortFileByID();
					LOGGER.log(Level.INFO,"Sortieren beendet von " + nextfile.getPath());
					nextfile.setPath(newpath);
					nextfile.setFiletype(Consts.csvFlag);
					nextfile.setIsSorted(true);
					worked = true; //if one files is ok, then start
				} catch (Exception e) {
					LOGGER.log(Level.SEVERE,"Fehler beim Sortieren von " + nextfile.getPath() , e);
					System.exit(1);
				LOGGER.log(Level.INFO,"Einlesen der Inputdatei "+ nextfile.getPath() + " gestartet.");
				InputFile newfile = new InputFile(nextfile.getDatentyp(),nextfile.getPath(),nextfile.getFiletype(),nextfile.getIdfeld(), nextfile.getSeparator(),nextfile.getQuote(),config.upcase(),filefilters);
				newfile.setLeader(nextfile.isLeadingTable());
				if (newfile.isLeader()) {
					if (this.ourLeaderfile != null) LOGGER.log(Level.WARNING,"Zwei leaderfiles sind nicht möglich! Das File " + nextfile.getPath() + " wird nicht als Leader genutzt." );
					else {
						ourLeaderfile=newfile;
						if (nextfile.hasSpecificColumns()) ourLeaderfile.setLeaderColnames(nextfile.getSpecificColumns());
						if (nextfile.hasNumfield()) ourLeaderfile.setLeaderNumfield(nextfile.getNumfield());
				inputfiles.add(newfile);
				worked = true; //if one files is ok, then start
			} catch (Exception e) {
				LOGGER.log(Level.SEVERE,"Fehler gefunden beim Einlesen von " + nextfile.getPath(), e);
				//e.printStackTrace();
		//1b. Init DDI
		if (worked && config.ddiconfiguration!=null) {
			ddiworker = new DDIworker();
			worked = ddiworker.init(config.ddiconfiguration, inputfiles,config.getOutputPath(),config.upcase());
			if (!worked) return worked;
			if (this.ourLeaderfile != null) LOGGER.log(Level.WARNING,"Zwei leaderfiles sind nicht möglich! Das File " + ourLeaderfile.getPath() + " wird nicht als Leader genutzt." );
			ourLeaderfile=ddiworker.getDDIInputFile();
			ddimode=true;
		//2. Create and Import Models
		List<String> processedModels = new ArrayList<String>();  
		File folder = new File(config.getModelpath());
		File[] listOfFiles = folder.listFiles();
		if (listOfFiles != null) {
			Model newmodel;
			for (File file : listOfFiles) {
			    if (file.isFile()) {
			    	String modelname = file.getName().replaceFirst("[.][^.]+$", "");
			    	if (!processedModels.contains(modelname)) {
			    		try {
			    			newmodel = new Model(modelname, inputfiles, config);
			    			knownVariables.put(newmodel,  new ArrayList<String>()); //init
			    			//knownVariables_targets.put(newmodel,  new ArrayList<String>()); //init
			    			processedModels.add(modelname);
			    			LOGGER.log(Level.INFO,"Modell "+ modelname + " konfiguriert.");
			    			models.add(newmodel);
			    			worked = true; //if one files is ok, then start
			    		} catch (Exception e) {
			    			String message = "Fehler gefunden bei Configuration des Modells " + modelname + ". Das Modell wird nicht verwendet.";
							message = message + " ...Ursache: "+ e.getMessage();
							if (e.getCause()!=null)
								message = message + " ...Ursache Detail: "+ e.getCause();
							LOGGER.log(Level.WARNING,message);
							//e.printStackTrace();
			//create and open output Score file
			if (worked && config.createScores())
				outputfile = new CSVWriter(new FileWriter(config.getOutputfile()), ';', CSVWriter.NO_QUOTE_CHARACTER);
				List<String> newline = new ArrayList<String>();
				//write header
				newline.add(Consts.idfieldheader);
				for (Model model : models) {
					newline.add(model.getName());
				outputfile.writeNext(newline.toArray(new String[newline.size()]));
				worked = true;
			} catch (Exception e) {
				LOGGER.log(Level.SEVERE,"Die Outputdatei " + config.getOutputfile() + " konnte nicht erstellt werden.", e);
				worked = false;
			//create and open output ProfilDense file (if needed)
			if (worked && config.createProfilDense()) {
				for (Model model : models) {
						profildensefile.put(model, new CSVWriter(new FileWriter(config.getProfilfileDenseTmp(model.getName(),false)), ';', CSVWriter.NO_QUOTE_CHARACTER));
						//if (model.hasTargets()) profildensefile_targets.put(model, new CSVWriter(new FileWriter(config.getProfilfileDenseTmp(model.getName(),true)), ';', CSVWriter.NO_QUOTE_CHARACTER));
						//no header here
					} catch (Exception e) {
						LOGGER.log(Level.SEVERE,"Die Outputdatei " + config.getProfilfileDense(model.getName(),false) + " konnte nicht erstellt werden.", e);
						worked = false;
			//create and open output ProfilSparse files (if needed)
			if (worked && config.createProfilSparse()) {
				for (Model model : models) {
						profilsparsefile.put(model, new CSVWriter(new FileWriter(config.getProfilfileSparse(model.getName(),false)), ';', CSVWriter.NO_QUOTE_CHARACTER));
						//if (model.hasTargets()) profilsparsefile_targets.put(model, new CSVWriter(new FileWriter(config.getProfilfileSparse(model.getName(),true)), ';', CSVWriter.NO_QUOTE_CHARACTER));
						String[] newline = {"ROW","COL","VAL"};
						//write header
						profilsparsefile.get(model).writeNext(newline);
						//if (model.hasTargets()) profilsparsefile_targets.get(model).writeNext(newline);
						//write header
						profilsparsefileRows.put(model, new CSVWriter(new FileWriter(config.getProfilfileSparseROWs(model.getName())), ';', CSVWriter.NO_QUOTE_CHARACTER));
						String[] rowline = {"ROW_NO",Consts.idfieldheader};
						profilsparsefileRows.get(model).writeNext(rowline);
					} catch (Exception e) {
						LOGGER.log(Level.SEVERE,"Die Outputdatei " + config.getProfilfileSparse(model.getName(),false) + " konnte nicht erstellt werden.", e);
						worked = false;
			//create and open output ProfilSvmlight files (if needed)
			if (worked && config.createProfilSvmlight()) {
				for (Model model : models) {
						profilsvmlightfile.put(model, new CSVWriter(new FileWriter(config.getProfilfileSvmlight(model.getName(),false)), ' ', CSVWriter.NO_QUOTE_CHARACTER));
						String[] newline1 = {"#Profile features in svmlight format, see http://svmlight.joachims.org (targets always 1, pids in #info part)"};
						profilsvmlightfile.get(model).writeNext(newline1);
						/*if (model.hasTargets()) {
							profilsvmlightfile_targets.put(model, new CSVWriter(new FileWriter(config.getProfilfileSvmlight(model.getName(),true)), ' ', CSVWriter.NO_QUOTE_CHARACTER));
							String[] newline3 = {"#Profile targets in svmlight format, see http://svmlight.joachims.org (targets always 1, pids in #info part)"};
							profilsvmlightfile_targets.get(model).writeNext(newline3);
					} catch (Exception e) {
						LOGGER.log(Level.SEVERE,"Die Outputdatei " + config.getProfilfileSvmlight(model.getName(),false) + " konnte nicht erstellt werden.", e);
						worked = false;
		} //Ende test listOfFiles!=null
		if (listOfFiles==null || models.size()==0 ) {
			LOGGER.log(Level.WARNING,"Keine korrekte Modellkonfiguration in " + config.getModelpath() + " gefunden");
			worked = false;
		return  worked;
	private void workThroughInfile (InputFile infile, Patient patient) throws Exception {
		//loop though models and process row for the patient
		for (Model model : models) {
			if (model.inputfileIsRelevant(infile) || infile.isLeader()) patient.processRow(model, infile);
		infile.nextRow(true,ourLeaderfile != null);
	 * Process.
	 * @return true, if successful
	public boolean process() { 
		LOGGER.log(Level.INFO,"Starte Profilbildung und Modellanwendung...");
		boolean worked = true;
		//add No. for rows and pids
		long pNo =0;
		long rowNo =0;
		//move to first real row for all inputfiles
		nextRowAllFiles();
		//loop through possible patients until no file has still rows
		for (Patient patient = findNextPatient(); patient != null; patient = findNextPatient()) {
			pNo++; //new patient, matrix starts from 1
			rowNo++;
		    //for each inputfiles file... 
			for (InputFile infile : inputfiles) {
				//if its the leaderfile: process only once
				//OBSOLETE: but for ddimode: loop traditional (just different decision when to stop, based on dates in DDIInputFile) 
				if (infile == ourLeaderfile) {
						workThroughInfile(infile,patient);
					} catch (Exception e) {
						LOGGER.log(Level.SEVERE,e.getMessage());
						worked=false;
						break;
				} else {
					//if no leaderfile: ..loop...that still has a valid Row and the correct Patient  
					while ((infile.hasRow()) && infile.isPatient(patient.getPid())) {
						try {
							workThroughInfile(infile,patient);
						} catch (Exception e) {
							LOGGER.log(Level.SEVERE,e.getMessage());
							worked=false;
							break;
			if (worked) {
				//save patient to file, if  includes and not excluded (per model)
				List<String> newline = new ArrayList<String>();
				// 1. Scores
				if (config.createScores()) {
					newline.add(patient.getPid());
					for (Model model : models) {
						if (model.hasCoeffs() && patient.areYouIncluded(model)) 
								newline.add(Double.toString(patient.getCoeffSum(model)));
						else newline.add(Consts.navalue); //write navalue if model has no coefficients or patient is not included
						outputfile.writeNext(newline.toArray(new String[newline.size()]));
					} catch (Exception e) {
						LOGGER.log(Level.SEVERE,"In die Outputdatei " + config.getOutputfile() + " konnte nicht geschrieben werden."+" ...Ursache:" + e.getMessage());
						worked = false;
				//2. Profiles
				if (config.createProfilDense() || config.createProfilSparse() || config.createProfilSvmlight()) {
					//if leaderfile: get all leaderfile_columns (OBSOLETE: not for ddimode -> their another options exists)
					String[] leaderrow = null;
					if (this.ourLeaderfile!=null && this.ourLeaderfile.hasLeaderCols()) {
						leaderrow = new String[this.ourLeaderfile.getLeaderColnames().length];
						//get Value not from current but from last cached row 
						for (int i=0;i<leaderrow.length;i++) leaderrow[i]=this.ourLeaderfile.getValueLastCached(this.ourLeaderfile.getLeaderColnames()[i]);
					boolean isincluded = false;
					for (Model model : models) {
						if (patient.areYouIncluded(model)) {
							isincluded = true;
							//update knownVariables
							ArrayList<String> knownVars = patient.addToKnownVariables(model,knownVariables.get(model));
							knownVariables.put(model,knownVars);
							//get ProfileValues
							ArrayList<String> profValues = patient.getProfvalues(model,knownVars);
							//same for targets 
							ArrayList<String> profValues_targets = null;
							if (model.hasTargets()) {
								//update knownVariables
								ArrayList<String> knownVars_targets = patient.addToKnownVariables(model,knownVariables_targets.get(model),true);
								knownVariables_targets.put(model,knownVars_targets);
								//get ProfileValues
								profValues_targets = patient.getProfvalues(model,knownVars_targets);
							if (config.createProfilDense()) {
								//output PID+Profilvalues
								worked = writeDenseProfileRow(profildensefile.get(model), config.getProfilfileDense(model.getName(),false), patient, profValues, leaderrow);
								/*//same for targets, w/o leaderrow
								if (worked && model.hasTargets()) 
									worked = writeDenseProfileRow(profildensefile_targets.get(model), config.getProfilfileDense(model.getName(),true), patient, profValues_targets, null);								
							} //end dense profile
							if (config.createProfilSparse()) {
								String myrowno;
								if (this.ourLeaderfile!=null && this.ourLeaderfile.hasLeaderNumfield())
									//get Value nor from current but from last chached row 
									myrowno=this.ourLeaderfile.getValueLastCached(this.ourLeaderfile.getLeaderNumfield());
								else 
									myrowno = Long.toString(rowNo);
								worked = writeMatlabProfileRow(profilsparsefile.get(model), config.getProfilfileSparse(model.getName(),false), patient, profValues, leaderrow, myrowno);
								/*//same for targets, w/o leaderrow
								if (worked && model.hasTargets()) 
									worked = writeMatlabProfileRow(profilsparsefile_targets.get(model), config.getProfilfileSparse(model.getName(),true), patient, profValues_targets, null,myrowno);
							} //end sparse profile
							if (config.createProfilSvmlight()) {
								//output: "1 column(=VariableNo.):value #PID"
								worked = writeSvmlightProfileRow(profilsvmlightfile.get(model), config.getProfilfileSparse(model.getName(),false), patient, profValues, leaderrow,config.addPidToSvm());
								//same for targets, w/o leaderrow
								/*if (worked && model.hasTargets()) 
									worked = writeSvmlightProfileRow(profilsvmlightfile_targets.get(model), config.getProfilfileSparse(model.getName(),true), patient, profValues_targets, null,config.addPidToSvm());
							} //end svmlight
							//ddi worker process
							if (ddimode) {
								worked = ddiworker.processModel(model,patient.getPid(),patient.getAllTargetVariables(model));
					} //end rolling through models
					//for sparse: write PID in row-translation.file
					if (config.createProfilSparse() && isincluded) {
						String[] pidrow = {Long.toString(rowNo), patient.getPid()};
						for (Model model : models) {
							try {
								profilsparsefileRows.get(model).writeNext(pidrow);
							} catch (Exception e) {
								LOGGER.log(Level.SEVERE,"In die Outputdatei " + config.getProfilfileSparseROWs(model.getName()) + " konnte nicht geschrieben werden.", e);
								worked = false;
					} else rowNo = rowNo-1; //if patient not included: reset no;
					//call ddiworker for patient
					if (ddiworker!=null) {
						worked = ddiworker.processPatient();
				} //end if profile needs creation
				if (pNo % 50000 == 0) {
					LOGGER.log(Level.INFO,Long.toString(pNo) + " Patienten verarbeitet.");
			} //end worked test
			else break;
		} // end for loop per patient
		if (worked) {
			LOGGER.log(Level.INFO,Long.toString(pNo) + " Patienten verarbeitet.");
		return worked;
	 * Finish.
	 * @return true, if successful
	public boolean finish() { 
		boolean worked = true;
		String[] leaderheader = null;
		if (this.ourLeaderfile!=null && this.ourLeaderfile.hasLeaderCols()) leaderheader = this.ourLeaderfile.getLeaderColnames();
		//close all files
			if (config.createScores()) {
				outputfile.close();
				LOGGER.log(Level.INFO,"Outputdatei " + config.getOutputfile() + " wurde erfolgreich geschrieben.");
			for (InputFile infile : inputfiles) {
				infile.close();
			for (Model model : models) {
				if (config.createProfilDense()) {
					//close file
					profildensefile.get(model).close();
					//add header by creating new file, and write all rows from tmpfile to new file
					worked = writeDenseHeader(config.getProfilfileDenseTmp(model.getName(),false),config.getProfilfileDense(model.getName(),false),knownVariables.get(model),leaderheader);
				    /*//same for targets
				    if (worked && model.hasTargets()) {
					  	//close file
				    	profildensefile_targets.get(model).close();
						//add header by creating new file, and write all rows from tmpfile to new file
				    	worked = writeDenseHeader(config.getProfilfileDenseTmp(model.getName(),true),config.getProfilfileDense(model.getName(),true),knownVariables_targets.get(model),null);
				    }*/
				if (config.createProfilSparse()) { 
					//close files
					profilsparsefile.get(model).close();
					LOGGER.log(Level.INFO,"Outputdatei " + config.getProfilfileSparse(model.getName(),false) + " wurde erfolgreich geschrieben.");
					profilsparsefileRows.get(model).close();
					/*if (model.hasTargets()) {
				    	//close file
						profilsparsefile_targets.get(model).close();
						LOGGER.log(Level.INFO,"Outputdatei " + config.getProfilfileSparse(model.getName(),true) + " wurde erfolgreich geschrieben.");
				    }*/
					//add header by creating new file, and write all rows from tmpfile to new file
					worked = writeMatlabHeader(config.getProfilfileSparseCOLs(model.getName(),false),knownVariables.get(model),leaderheader);
				    //same for targets
				    /*if (worked && model.hasTargets()) {
				    	//add header by creating new file, and write all rows from tmpfile to new file
					  	worked = writeMatlabHeader(config.getProfilfileSparseCOLs(model.getName(),true),knownVariables_targets.get(model),null);
				    }*/
				if (config.createProfilSvmlight()) { 
					//close file
					profilsvmlightfile.get(model).close();
					/*if (model.hasTargets()) {
				    	//close file
						profilsvmlightfile_targets.get(model).close();
						LOGGER.log(Level.INFO,"Outputdatei " + config.getProfilfileSvmlight(model.getName(),true) + " wurde erfolgreich geschrieben.");
				    }*/
					//write header
					worked = writeSvmlightHeader(config.getProfilfileSvmlightHeader(model.getName(),false),knownVariables.get(model),leaderheader);
				    /*//same for targets
				    if (worked && model.hasTargets()) {
						//add header by creating new file, and write all rows from tmpfile to new file
				    	worked = writeSvmlightHeader(config.getProfilfileSvmlightHeader(model.getName(),true),knownVariables_targets.get(model),null);
				    }*/
				if (ddimode) {
					worked = ddiworker.finish(model,config.getOutputPath());
					if (worked) LOGGER.log(Level.INFO,"DDI Statistiken geschrieben für Model " + model.getName());
		} catch (Exception e) {
			LOGGER.log(Level.SEVERE,"Fehler beim Schließen der Dateien.", e);
			worked = false;
		return worked;
	//helper functions for profile output
	private boolean writeDenseProfileRow(CSVWriter file, String filename, Patient patient, ArrayList<String> profValues, String[] leaderrow) {
		//output PID+Profilvalues only
		ArrayList<String> newline = new ArrayList<String>();
		newline.add(patient.getPid());
		newline.addAll(profValues);
			//first write leaderfile columns (only for profiles)
			if (leaderrow!=null) file.writeNext(Utils.concatArrays(leaderrow,newline.toArray(new String[newline.size()])));
			//else write rest
			else file.writeNext(newline.toArray(new String[newline.size()]));
		} catch (Exception e) {
			LOGGER.log(Level.SEVERE,"In die Outputdatei " + filename + " konnte nicht geschrieben werden.", e);
			return false;
		return true;
	private boolean writeDenseHeader(String filename_old, String filename_new, ArrayList<String> variables, String[] leaderheader) {
		ArrayList<String> header;
		//first:add leadercolumns
		if (leaderheader != null) header = new ArrayList<String>(Arrays.asList(leaderheader));
		else header = new ArrayList<String>();
		header.add(Consts.idfieldheader);
		header.addAll(variables);
			Utils.addHeaderToCsv(filename_old,header, filename_new);
		} catch (Exception e) {
			LOGGER.log(Level.SEVERE,"In die Outputdatei " + filename_new + " konnte nicht geschrieben werden.", e);
			return false;
		LOGGER.log(Level.INFO,"Outputdatei " + filename_new + " wurde erfolgreich geschrieben.");
	    return true;
	private boolean writeMatlabProfileRow(CSVWriter file, String filename, Patient patient, ArrayList<String> profValues, String[] leaderrow, String rowNo) {
		//output: row PID No, column (=Variable no.), value
		String[] sparseline = new String[3];
		int starterno=0;
		//first write leaderrow
		if (leaderrow!=null) {
			for (int i =0; i<leaderrow.length; i++) {
				if (!leaderrow[i].equals(Consts.navalue) && !leaderrow[i].isEmpty()) {
					sparseline[0]=rowNo; //Row
					sparseline[1]=Integer.toString(i+1); //Col; matrix starts from 1
					sparseline[2]=leaderrow[i]; //Val
						file.writeNext(sparseline);
					} catch (Exception e) {
						LOGGER.log(Level.SEVERE,"In die Outputdatei " + filename + " konnte nicht geschrieben werden.", e);
						return false;
					sparseline = new String[3];
			starterno = leaderrow.length;
		//Next: profvalues
		for (int i =0; i<profValues.size(); i++) {
			if (!profValues.get(i).equals(Consts.navalue)) {
				sparseline[0]=rowNo; //Row
				sparseline[1]=Integer.toString(i+1+starterno); //Col; matrix starts from 1
				sparseline[2]=profValues.get(i); //Val
				if (!sparseline.equals(new String[3])) {
						file.writeNext(sparseline);
					} catch (Exception e) {
						LOGGER.log(Level.SEVERE,"In die Outputdatei " + filename + " konnte nicht geschrieben werden.", e);
						return false;
					sparseline = new String[3];
		return true;
	private boolean writeMatlabHeader(String filename, ArrayList<String> variables, String[] leaderrow) {
		//Col-&Row-Translation File
		int starterno=0;
			CSVWriter file= new CSVWriter(new FileWriter(filename), ';', CSVWriter.NO_QUOTE_CHARACTER);
			//write header
			String[] colline = {"COL_NO","VARIABLE"};
			file.writeNext(colline);
			if (leaderrow!=null) {
				for (int i =0; i<leaderrow.length; i++) {
					String[] varcol = {Integer.toString(i+1),leaderrow[i]};
					file.writeNext(varcol);
				starterno=leaderrow.length;
			for (int i =0; i<variables.size(); i++) {
				String[] varcol = {Integer.toString(i+1+starterno), variables.get(i)};
				file.writeNext(varcol);
			file.close();
		} catch (Exception e) {
			LOGGER.log(Level.SEVERE,"In die Outputdatei " + filename + " konnte nicht geschrieben werden.", e);
			return false;
		LOGGER.log(Level.INFO,"Outputdatei " + filename + " wurde erfolgreich geschrieben.");
	    return true;
	private boolean writeSvmlightProfileRow(CSVWriter file, String filename, Patient patient, ArrayList<String> profValues, String[] leaderrow, boolean addPid) {
		//output: "1 column(=VariableNo.):value #PID"
		ArrayList<String> newline = new ArrayList<String>();
		newline.add("1");
		int starterno=0;
		//first write leaderrow
		if (leaderrow!=null) {
			for (int i =0; i<leaderrow.length; i++) {
				if (!leaderrow[i].equals(Consts.navalue) && !leaderrow[i].isEmpty()) 
					newline.add(Integer.toString(i+1) + ":" + leaderrow[i]);
			starterno = leaderrow.length;
		//Next: profvalues
		for (int i =0; i<profValues.size(); i++) {
			if (!profValues.get(i).equals(Consts.navalue))
				newline.add(Integer.toString(i+1+starterno) + ":" + profValues.get(i));
		if (addPid) newline.add("# "+patient.getPid());
			file.writeNext(newline.toArray(new String[newline.size()]));
		} catch (Exception e) {
			LOGGER.log(Level.SEVERE,"In die Outputdatei " + filename + " konnte nicht geschrieben werden."+"...Ursache: "+ e.getMessage());
			return false;
		return true;
	private boolean writeSvmlightHeader(String filename, ArrayList<String> variables, String[] leaderrow) {
		int starterno=0;
			CSVWriter file= new CSVWriter(new FileWriter(filename), ' ', CSVWriter.NO_QUOTE_CHARACTER);
			//write header
			String[] newline = {"#Header to features, separated by linespace"};
			file.writeNext(newline);
			if (leaderrow!=null) {
				for (int i =0; i<leaderrow.length; i++) {
					String[] varhead = {Integer.toString(i+1),leaderrow[i]};
					file.writeNext(varhead);
				starterno=leaderrow.length;
			for (int i =0; i<variables.size(); i++) {
				String[] varhead = {Integer.toString(i+1+starterno),variables.get(i)};
				file.writeNext(varhead);
			file.close();
		} catch (Exception e) {
			LOGGER.log(Level.SEVERE,"In die Outputdatei " + filename + " konnte nicht geschrieben werden."+"...Ursache: "+ e.getMessage());
			return false;
		LOGGER.log(Level.INFO,"Outputdatei " + filename + " wurde erfolgreich geschrieben.");
	    return true;
	public void setLeaderfile (InputFile inputfile) {
		this.ourLeaderfile=inputfile;
Provide feedback

Saved searches

Use saved searches to filter your results more quickly

FilesExpand file tree

Worker.java

Latest commit

History

Worker.java

File metadata and controls
