Package org.apache.lucene.search

Examples of org.apache.lucene.search.Similarity


    private String cachedFieldName;
    private Similarity cachedSimilarity;
   
    public byte[] norms(String fieldName) {
      byte[] norms = cachedNorms;
      Similarity sim = getSimilarity();
      if (fieldName != cachedFieldName || sim != cachedSimilarity) { // not cached?
        Info info = getInfo(fieldName);
        int numTokens = info != null ? info.numTokens : 0;
        float n = sim.lengthNorm(fieldName, numTokens);
        float boost = info != null ? info.getBoost() : 1.0f;
        n = n * boost; // see DocumentWriter.writeNorms(String segment)               
        byte norm = Similarity.encodeNorm(n);
        norms = new byte[] {norm};
       
View Full Code Here


  public void testAddDocument() throws Exception {
    Document testDoc = new Document();
    DocHelper.setupDoc(testDoc);
    Analyzer analyzer = new WhitespaceAnalyzer();
    Similarity similarity = Similarity.getDefault();
    DocumentWriter writer = new DocumentWriter(dir, analyzer, similarity, 50);
    String segName = "test";
    writer.addDocument(segName, testDoc);
    //After adding the document, we should be able to read it back in
    SegmentReader reader = SegmentReader.get(new SegmentInfo(segName, 1, dir));
View Full Code Here

      public int getPositionIncrementGap(String fieldName) {
        return 500;
      }
    };

    Similarity similarity = Similarity.getDefault();
    DocumentWriter writer = new DocumentWriter(dir, analyzer, similarity, 50);
    Document doc = new Document();
    doc.add(new Field("repeated", "repeated one", Field.Store.YES, Field.Index.TOKENIZED));
    doc.add(new Field("repeated", "repeated two", Field.Store.YES, Field.Index.TOKENIZED));
View Full Code Here

    assertEquals(0, termPositions.nextPosition());
    assertEquals(502, termPositions.nextPosition());
  }
 
  public void testPreAnalyzedField() throws IOException {
    Similarity similarity = Similarity.getDefault();
    DocumentWriter writer = new DocumentWriter(dir, new SimpleAnalyzer(), similarity, 50);
    Document doc = new Document();
   
    doc.add(new Field("preanalyzed", new TokenStream() {
      private String[] tokens = new String[] {"term1", "term2", "term3", "term2"};
View Full Code Here

   * @param doc
   * @throws IOException
   */
  public static void writeDoc(Directory dir, String segment, Document doc) throws IOException
  {
    Similarity similarity = Similarity.getDefault();
    writeDoc(dir, new WhitespaceAnalyzer(), similarity, segment, doc);
  }
View Full Code Here

  public void testSweetSpotLengthNorm() {
 
    SweetSpotSimilarity ss = new SweetSpotSimilarity();
    ss.setLengthNormFactors(1,1,0.5f);

    Similarity d = new DefaultSimilarity();
    Similarity s = ss;


    // base case, should degrade
 
    for (int i = 1; i < 1000; i++) {
      assertEquals("base case: i="+i,
                   d.lengthNorm("foo",i), s.lengthNorm("foo",i),
                   0.0f);
    }

    // make a sweet spot
 
    ss.setLengthNormFactors(3,10,0.5f);
 
    for (int i = 3; i <=10; i++) {
      assertEquals("3,10: spot i="+i,
                   1.0f, s.lengthNorm("foo",i),
                   0.0f);
    }
 
    for (int i = 10; i < 1000; i++) {
      assertEquals("3,10: 10<x : i="+i,
                   d.lengthNorm("foo",i-9), s.lengthNorm("foo",i),
                   0.0f);
    }


    // seperate sweet spot for certain fields

    ss.setLengthNormFactors("bar",8,13, 0.5f);
    ss.setLengthNormFactors("yak",6,9, 0.5f);

 
    for (int i = 3; i <=10; i++) {
      assertEquals("f: 3,10: spot i="+i,
                   1.0f, s.lengthNorm("foo",i),
                   0.0f);
    }
    for (int i = 10; i < 1000; i++) {
      assertEquals("f: 3,10: 10<x : i="+i,
                   d.lengthNorm("foo",i-9), s.lengthNorm("foo",i),
                   0.0f);
    }
    for (int i = 8; i <=13; i++) {
      assertEquals("f: 8,13: spot i="+i,
                   1.0f, s.lengthNorm("bar",i),
                   0.0f);
    }
    for (int i = 6; i <=9; i++) {
      assertEquals("f: 6,9: spot i="+i,
                   1.0f, s.lengthNorm("yak",i),
                   0.0f);
    }
    for (int i = 13; i < 1000; i++) {
      assertEquals("f: 8,13: 13<x : i="+i,
                   d.lengthNorm("foo",i-12), s.lengthNorm("bar",i),
                   0.0f);
    }
    for (int i = 9; i < 1000; i++) {
      assertEquals("f: 6,9: 9<x : i="+i,
                   d.lengthNorm("foo",i-8), s.lengthNorm("yak",i),
                   0.0f);
    }


    // steepness

    ss.setLengthNormFactors("a",5,8,0.5f);
    ss.setLengthNormFactors("b",5,8,0.1f);

    for (int i = 9; i < 1000; i++) {
      assertTrue("s: i="+i+" : a="+ss.lengthNorm("a",i)+
                 " < b="+ss.lengthNorm("b",i),
                 ss.lengthNorm("a",i) < s.lengthNorm("b",i));
    }

  }
View Full Code Here

  public void testSweetSpotTf() {
 
    SweetSpotSimilarity ss = new SweetSpotSimilarity();

    Similarity d = new DefaultSimilarity();
    Similarity s = ss;
   
    // tf equal

    ss.setBaselineTfFactors(0.0f, 0.0f);
 
    for (int i = 1; i < 1000; i++) {
      assertEquals("tf: i="+i,
                   d.tf(i), s.tf(i), 0.0f);
    }

    // tf higher
 
    ss.setBaselineTfFactors(1.0f, 0.0f);
 
    for (int i = 1; i < 1000; i++) {
      assertTrue("tf: i="+i+" : d="+d.tf(i)+
                 " < s="+s.tf(i),
                 d.tf(i) < s.tf(i));
    }

    // tf flat
 
    ss.setBaselineTfFactors(1.0f, 6.0f);
    for (int i = 1; i <=6; i++) {
      assertEquals("tf flat1: i="+i, 1.0f, s.tf(i), 0.0f);
    }
    ss.setBaselineTfFactors(2.0f, 6.0f);
    for (int i = 1; i <=6; i++) {
      assertEquals("tf flat2: i="+i, 2.0f, s.tf(i), 0.0f);
    }
    for (int i = 6; i <=1000; i++) {
      assertTrue("tf: i="+i+" : s="+s.tf(i)+
                 " < d="+d.tf(i),
                 s.tf(i) < d.tf(i));
    }

    // stupidity
    assertEquals("tf zero", 0.0f, s.tf(0), 0.0f);
  }
View Full Code Here

          return hyperbolicTf(freq);
        }
      };
    ss.setHyperbolicTfFactors(3.3f, 7.7f, Math.E, 5.0f);
   
    Similarity s = ss;

    for (int i = 1; i <=1000; i++) {
      assertTrue("MIN tf: i="+i+" : s="+s.tf(i),
                 3.3f <= s.tf(i));
      assertTrue("MAX tf: i="+i+" : s="+s.tf(i),
                 s.tf(i) <= 7.7f);
    }
    assertEquals("MID tf", 3.3f+(7.7f - 3.3f)/2.0f, s.tf(5), 0.00001f);
   
    // stupidity
    assertEquals("tf zero", 0.0f, s.tf(0), 0.0f);
   
  }
View Full Code Here

    private String cachedFieldName;
    private Similarity cachedSimilarity;
   
    public byte[] norms(String fieldName) {
      byte[] norms = cachedNorms;
      Similarity sim = getSimilarity();
      if (fieldName != cachedFieldName || sim != cachedSimilarity) { // not cached?
        Info info = getInfo(fieldName);
        int numTokens = info != null ? info.numTokens : 0;
        float n = sim.lengthNorm(fieldName, numTokens);
        byte norm = Similarity.encodeNorm(n);
        norms = new byte[] {norm};
       
        cachedNorms = norms;
        cachedFieldName = fieldName;
View Full Code Here

   * @param doc
   * @throws IOException
   */
  public static void writeDoc(Directory dir, String segment, Document doc) throws IOException
  {
    Similarity similarity = Similarity.getDefault();
    writeDoc(dir, new WhitespaceAnalyzer(), similarity, segment, doc);
  }
View Full Code Here

TOP

Related Classes of org.apache.lucene.search.Similarity

Copyright © 2018 www.massapicom. All rights reserved.
All source code are property of their respective owners. Java is a trademark of Sun Microsystems, Inc and owned by ORACLE Inc. Contact coftware#gmail.com.