@Override
protected List<AlignmentFile> getAlignmentFiles(int currentIter) {
// sort input file alphabetically to look nicer
File inputDirectory = new File(this.inputDirectory);
File[] listOfFiles = inputDirectory.listFiles();
Arrays.sort(listOfFiles, new NaturalOrderFilenameComparator());
List<AlignmentFile> alignmentDataList = new ArrayList<AlignmentFile>();
List<Feature> allFeatures = new ArrayList<Feature>();
// load feature data
for (int i = 0; i < listOfFiles.length; i++) {
if (listOfFiles[i].isFile()) {
File myFile = listOfFiles[i];
String filename = myFile.getName();
String path = myFile.getPath();
String extension = filename.substring(filename.lastIndexOf('.') + 1);
// ignore other crap files in the directory
if (!"txt".equals(extension)) {
continue;
}
List<Feature> features = new ArrayList<Feature>();
try {
features = this.loadFeatures(path);
} catch (ValidityException e) {
e.printStackTrace();
} catch (ParsingException e) {
e.printStackTrace();
} catch (IOException e) {
e.printStackTrace();
}
AlignmentFile alignmentData = new AlignmentFile(i, myFile, features);
if (verbose) {
System.out.println(
"Load " + filename + " with " + alignmentData.getFeaturesCount() + " features");
}
alignmentDataList.add(alignmentData);
allFeatures.addAll(alignmentData.getFeatures());
}
}
this.alignmentFiles = alignmentDataList;
return alignmentDataList;
}
@Override
protected GroundTruth getGroundTruth() {
if (this.gtPath == null) {
return null;
}
// store only features present in files that we're aligning
List<FeatureGroup> groundTruthEntries = new ArrayList<FeatureGroup>();
Scanner in = null;
try {
// build a map from filename -> the actual feature data
Map<String, AlignmentFile> alignmentDataMap = new HashMap<String, AlignmentFile>();
for (AlignmentFile data : this.alignmentFiles) {
alignmentDataMap.put(data.getFilenameWithoutExtension(), data);
}
Map<Integer, String> tempMap = new HashMap<Integer, String>();
File inputFile = new File(this.gtPath);
in = new Scanner(inputFile);
int groupID = 1;
while (in.hasNextLine()) {
String line = in.nextLine();
Scanner lineSplitter = new Scanner(line);
while (lineSplitter.hasNext()) {
String firstToken = lineSplitter.next();
if (">".equals(firstToken)) {
int fileIdx = lineSplitter.nextInt();
String filename = lineSplitter.next();
tempMap.put(fileIdx, filename);
} else if ("#".equals(firstToken)) {
FeatureGroup gtFeatures = new FeatureGroup(groupID);
while (lineSplitter.hasNext()) {
int fileIdx = lineSplitter.nextInt();
int peakIdx = lineSplitter.nextInt();
String filename = tempMap.get(fileIdx);
AlignmentFile data = alignmentDataMap.get(filename);
if (data != null) {
Feature featureFromData = data.getFeatureByPeakID(peakIdx);
assert featureFromData != null : "fileIdx " + fileIdx + " peakIdx " + peakIdx;
gtFeatures.addFeature(featureFromData);
}
}
groupID++; // new groupID for every line in ground truth
// don't need this .. ?
// if (gtFeatures.getFeatureCount() > 1) {
groundTruthEntries.add(gtFeatures);
// }
}
}
lineSplitter.close();
} // close main while loop
} catch (FileNotFoundException e) {
e.printStackTrace();
} finally {
if (in != null) {
in.close();
}
}
System.out.println("Load ground truth = " + groundTruthEntries.size() + " rows");
// System.out.println("Retaining only entries size >= 2 = " + groundTruthEntries.size() + "
// rows");
Iterator<FeatureGroup> it = groundTruthEntries.iterator();
while (it.hasNext()) {
FeatureGroup gg = it.next();
if (gg.getFeatureCount() < 2) {
it.remove();
}
}
GroundTruth groundTruth = new GroundTruth(groundTruthEntries, gtCombinationSize, verbose);
return groundTruth;
}
