private static TrainingInstance parseTI(String line) throws InvalidFeatureVectorValueException { String[] fields = line.split(" "); boolean label = fields[0].trim().equals("+1"); FeatureVector fv = new SparseFeatureVector(); for (int i = 1; i < fields.length; i++) { String[] parts = fields[i].split(":"); int featureIndex = Integer.valueOf(parts[0]); double featureValue = Double.valueOf(parts[1]); fv.set(featureIndex, featureValue); } return new TrainingInstance(label, fv); }
public SparseFeatureVector(FeatureVector fv) throws InvalidFeatureVectorValueException { this(); for (FeatureVector.Entry entry : fv) { this.set(entry.index, entry.value); } }
SparseFeatureVector fv = new SparseFeatureVector(); for (NameNumber fve : fves) { String name = fve.name; if (expandIndex) { int i = stringMapper.getOrGenerateInteger(name); double v = fv.get(i) + value.doubleValue(); fv.set(i, v); } else { try { int i = stringMapper.getInteger(name); double v = fv.get(i) + value.doubleValue(); fv.set(i, v); } catch (UnknownKeyException e) {
private static TrainingInstance parseTI(String line) throws InvalidFeatureVectorValueException { String[] fields = line.split(" "); boolean label = fields[0].trim().equals("+1"); FeatureVector fv = new SparseFeatureVector(); for (int i = 1; i < fields.length; i++) { String[] parts = fields[i].split(":"); int featureIndex = Integer.valueOf(parts[0]); double featureValue = Double.valueOf(parts[1]); fv.set(featureIndex, featureValue); } return new TrainingInstance(label, fv); }
private void compress() throws InvalidFeatureVectorValueException { if (!(kernel instanceof LinearKernel)) return; FeatureVector newFV = new SparseFeatureVector(); for (SupportVector sv : supportVectors) { FeatureVector fv = new SparseFeatureVector(sv.featureVector); fv.multiply(sv.alpha_y); newFV.add(fv); } SupportVector newSVs[] = { new SupportVector(1, newFV) }; supportVectors = newSVs; }
private void compress() throws InvalidFeatureVectorValueException { if (!(kernel instanceof LinearKernel)) return; FeatureVector newFV = new SparseFeatureVector(); for (SupportVector sv : supportVectors) { FeatureVector fv = new SparseFeatureVector(sv.featureVector); fv.multiply(sv.alpha_y); newFV.add(fv); } SupportVector newSVs[] = { new SupportVector(1, newFV) }; supportVectors = newSVs; }
private static SupportVector readSV(SvmLightReader in) throws IOException { String[] fields = in.readLine().split(" "); double alpha_y = Double.valueOf(fields[0]); FeatureVector fv = new SparseFeatureVector(); for (int i = 1; i < fields.length; i++) { String[] parts = fields[i].split(":"); int featureIndex = Integer.valueOf(parts[0]); double featureValue = Double.valueOf(parts[1]); try { fv.set(featureIndex, featureValue); } catch (InvalidFeatureVectorValueException e) { throw new IOException(e); } } return new SupportVector(alpha_y, fv); }
private static SupportVector readSV(SvmLightReader in) throws IOException { String[] fields = in.readLine().split(" "); double alpha_y = Double.valueOf(fields[0]); FeatureVector fv = new SparseFeatureVector(); for (int i = 1; i < fields.length; i++) { String[] parts = fields[i].split(":"); int featureIndex = Integer.valueOf(parts[0]); double featureValue = Double.valueOf(parts[1]); try { fv.set(featureIndex, featureValue); } catch (InvalidFeatureVectorValueException e) { throw new IOException(e); } } return new SupportVector(alpha_y, fv); }
FeatureVector fv = new SparseFeatureVector();
FeatureVector vec = new SparseFeatureVector(); if (vectSect != null) { String[] features = vectSect.trim().split(" +");
FeatureVector fv = new SparseFeatureVector();