Examples with DirectoryTaxonomyReader org.apache.lucene.facet.taxonomy.directory.DirectoryTaxonomyReader used on opensource projects

Example 31 with DirectoryTaxonomyReader

use of org.apache.lucene.facet.taxonomy.directory.DirectoryTaxonomyReader in project lucene-solr by apache.

the class TestTaxonomyFacetSumValueSource method testSparseFacets.

// LUCENE-5333
public void testSparseFacets() throws Exception {
    Directory dir = newDirectory();
    Directory taxoDir = newDirectory();
    // Writes facet ords to a separate directory from the
    // main index:
    DirectoryTaxonomyWriter taxoWriter = new DirectoryTaxonomyWriter(taxoDir, IndexWriterConfig.OpenMode.CREATE);
    RandomIndexWriter writer = new RandomIndexWriter(random(), dir);
    FacetsConfig config = new FacetsConfig();
    Document doc = new Document();
    doc.add(new NumericDocValuesField("num", 10));
    doc.add(new FacetField("a", "foo1"));
    writer.addDocument(config.build(taxoWriter, doc));
    if (random().nextBoolean()) {
        writer.commit();
    }
    doc = new Document();
    doc.add(new NumericDocValuesField("num", 20));
    doc.add(new FacetField("a", "foo2"));
    doc.add(new FacetField("b", "bar1"));
    writer.addDocument(config.build(taxoWriter, doc));
    if (random().nextBoolean()) {
        writer.commit();
    }
    doc = new Document();
    doc.add(new NumericDocValuesField("num", 30));
    doc.add(new FacetField("a", "foo3"));
    doc.add(new FacetField("b", "bar2"));
    doc.add(new FacetField("c", "baz1"));
    writer.addDocument(config.build(taxoWriter, doc));
    // NRT open
    IndexSearcher searcher = newSearcher(writer.getReader());
    writer.close();
    // NRT open
    TaxonomyReader taxoReader = new DirectoryTaxonomyReader(taxoWriter);
    taxoWriter.close();
    FacetsCollector c = new FacetsCollector();
    searcher.search(new MatchAllDocsQuery(), c);
    TaxonomyFacetSumValueSource facets = new TaxonomyFacetSumValueSource(taxoReader, new FacetsConfig(), c, DoubleValuesSource.fromIntField("num"));
    // Ask for top 10 labels for any dims that have counts:
    List<FacetResult> results = facets.getAllDims(10);
    assertEquals(3, results.size());
    assertEquals("dim=a path=[] value=60.0 childCount=3\n  foo3 (30.0)\n  foo2 (20.0)\n  foo1 (10.0)\n", results.get(0).toString());
    assertEquals("dim=b path=[] value=50.0 childCount=2\n  bar2 (30.0)\n  bar1 (20.0)\n", results.get(1).toString());
    assertEquals("dim=c path=[] value=30.0 childCount=1\n  baz1 (30.0)\n", results.get(2).toString());
    IOUtils.close(searcher.getIndexReader(), taxoReader, dir, taxoDir);
}

Example 32 with DirectoryTaxonomyReader

use of org.apache.lucene.facet.taxonomy.directory.DirectoryTaxonomyReader in project lucene-solr by apache.

the class TestTaxonomyFacetSumValueSource method testBasic.

public void testBasic() throws Exception {
    Directory dir = newDirectory();
    Directory taxoDir = newDirectory();
    // Writes facet ords to a separate directory from the
    // main index:
    DirectoryTaxonomyWriter taxoWriter = new DirectoryTaxonomyWriter(taxoDir, IndexWriterConfig.OpenMode.CREATE);
    RandomIndexWriter writer = new RandomIndexWriter(random(), dir);
    FacetsConfig config = new FacetsConfig();
    // Reused across documents, to add the necessary facet
    // fields:
    Document doc = new Document();
    doc.add(new NumericDocValuesField("num", 10));
    doc.add(new FacetField("Author", "Bob"));
    writer.addDocument(config.build(taxoWriter, doc));
    doc = new Document();
    doc.add(new NumericDocValuesField("num", 20));
    doc.add(new FacetField("Author", "Lisa"));
    writer.addDocument(config.build(taxoWriter, doc));
    doc = new Document();
    doc.add(new NumericDocValuesField("num", 30));
    doc.add(new FacetField("Author", "Lisa"));
    writer.addDocument(config.build(taxoWriter, doc));
    doc = new Document();
    doc.add(new NumericDocValuesField("num", 40));
    doc.add(new FacetField("Author", "Susan"));
    writer.addDocument(config.build(taxoWriter, doc));
    doc = new Document();
    doc.add(new NumericDocValuesField("num", 45));
    doc.add(new FacetField("Author", "Frank"));
    writer.addDocument(config.build(taxoWriter, doc));
    // NRT open
    IndexSearcher searcher = newSearcher(writer.getReader());
    writer.close();
    // NRT open
    TaxonomyReader taxoReader = new DirectoryTaxonomyReader(taxoWriter);
    taxoWriter.close();
    // Aggregate the facet counts:
    FacetsCollector c = new FacetsCollector();
    // MatchAllDocsQuery is for "browsing" (counts facets
    // for all non-deleted docs in the index); normally
    // you'd use a "normal" query and one of the
    // Facets.search utility methods:
    searcher.search(new MatchAllDocsQuery(), c);
    TaxonomyFacetSumValueSource facets = new TaxonomyFacetSumValueSource(taxoReader, new FacetsConfig(), c, DoubleValuesSource.fromIntField("num"));
    // Retrieve & verify results:
    assertEquals("dim=Author path=[] value=145.0 childCount=4\n  Lisa (50.0)\n  Frank (45.0)\n  Susan (40.0)\n  Bob (10.0)\n", facets.getTopChildren(10, "Author").toString());
    taxoReader.close();
    searcher.getIndexReader().close();
    dir.close();
    taxoDir.close();
}

Example 33 with DirectoryTaxonomyReader

use of org.apache.lucene.facet.taxonomy.directory.DirectoryTaxonomyReader in project lucene-solr by apache.

the class TestTaxonomyFacetCounts2 method testAllCounts.

@Test
public void testAllCounts() throws Exception {
    DirectoryReader indexReader = DirectoryReader.open(indexDir);
    TaxonomyReader taxoReader = new DirectoryTaxonomyReader(taxoDir);
    IndexSearcher searcher = newSearcher(indexReader);
    FacetsCollector sfc = new FacetsCollector();
    searcher.search(new MatchAllDocsQuery(), sfc);
    Facets facets = getTaxonomyFacetCounts(taxoReader, getConfig(), sfc);
    FacetResult result = facets.getTopChildren(NUM_CHILDREN_CP_A, CP_A);
    assertEquals(-1, result.value.intValue());
    int prevValue = Integer.MAX_VALUE;
    for (LabelAndValue labelValue : result.labelValues) {
        assertEquals(allExpectedCounts.get(CP_A + "/" + labelValue.label), labelValue.value);
        assertTrue("wrong sort order of sub results: labelValue.value=" + labelValue.value + " prevValue=" + prevValue, labelValue.value.intValue() <= prevValue);
        prevValue = labelValue.value.intValue();
    }
    result = facets.getTopChildren(NUM_CHILDREN_CP_B, CP_B);
    assertEquals(allExpectedCounts.get(CP_B), result.value);
    prevValue = Integer.MAX_VALUE;
    for (LabelAndValue labelValue : result.labelValues) {
        assertEquals(allExpectedCounts.get(CP_B + "/" + labelValue.label), labelValue.value);
        assertTrue("wrong sort order of sub results: labelValue.value=" + labelValue.value + " prevValue=" + prevValue, labelValue.value.intValue() <= prevValue);
        prevValue = labelValue.value.intValue();
    }
    IOUtils.close(indexReader, taxoReader);
}

Example 34 with DirectoryTaxonomyReader

use of org.apache.lucene.facet.taxonomy.directory.DirectoryTaxonomyReader in project lucene-solr by apache.

the class TestTaxonomyFacetSumValueSource method testNoScore.

public void testNoScore() throws Exception {
    Directory indexDir = newDirectory();
    Directory taxoDir = newDirectory();
    DirectoryTaxonomyWriter taxoWriter = new DirectoryTaxonomyWriter(taxoDir);
    IndexWriter iw = new IndexWriter(indexDir, newIndexWriterConfig(new MockAnalyzer(random())));
    FacetsConfig config = new FacetsConfig();
    for (int i = 0; i < 4; i++) {
        Document doc = new Document();
        doc.add(new NumericDocValuesField("price", (i + 1)));
        doc.add(new FacetField("a", Integer.toString(i % 2)));
        iw.addDocument(config.build(taxoWriter, doc));
    }
    DirectoryReader r = DirectoryReader.open(iw);
    DirectoryTaxonomyReader taxoReader = new DirectoryTaxonomyReader(taxoWriter);
    FacetsCollector sfc = new FacetsCollector();
    newSearcher(r).search(new MatchAllDocsQuery(), sfc);
    Facets facets = new TaxonomyFacetSumValueSource(taxoReader, config, sfc, DoubleValuesSource.fromLongField("price"));
    assertEquals("dim=a path=[] value=10.0 childCount=2\n  1 (6.0)\n  0 (4.0)\n", facets.getTopChildren(10, "a").toString());
    iw.close();
    IOUtils.close(taxoWriter, taxoReader, taxoDir, r, indexDir);
}

Example 35 with DirectoryTaxonomyReader

use of org.apache.lucene.facet.taxonomy.directory.DirectoryTaxonomyReader in project lucene-solr by apache.

the class TestTaxonomyFacetCounts2 method testBigNumResults.

@Test
public void testBigNumResults() throws Exception {
    DirectoryReader indexReader = DirectoryReader.open(indexDir);
    TaxonomyReader taxoReader = new DirectoryTaxonomyReader(taxoDir);
    IndexSearcher searcher = newSearcher(indexReader);
    FacetsCollector sfc = new FacetsCollector();
    searcher.search(new MatchAllDocsQuery(), sfc);
    Facets facets = getTaxonomyFacetCounts(taxoReader, getConfig(), sfc);
    FacetResult result = facets.getTopChildren(Integer.MAX_VALUE, CP_A);
    assertEquals(-1, result.value.intValue());
    for (LabelAndValue labelValue : result.labelValues) {
        assertEquals(allExpectedCounts.get(CP_A + "/" + labelValue.label), labelValue.value);
    }
    result = facets.getTopChildren(Integer.MAX_VALUE, CP_B);
    assertEquals(allExpectedCounts.get(CP_B), result.value);
    for (LabelAndValue labelValue : result.labelValues) {
        assertEquals(allExpectedCounts.get(CP_B + "/" + labelValue.label), labelValue.value);
    }
    IOUtils.close(indexReader, taxoReader);
}