Package com.pearson.entech.elasticsearch.search.facet.approx.date.internal

Examples of com.pearson.entech.elasticsearch.search.facet.approx.date.internal.InternalDistinctFacet$Serializer

97
98
99
100
101
102
103
104
105
106
107
108
        testSerializingDistinctFacet(counts, expectedCounts, expectedCardinalities);
    }

    private void testSerializingDistinctFacet(final ExtTLongObjectHashMap<DistinctCountPayload> counts,
            final Map<Long, Integer> expectedCounts, final Map<Long, Integer> expectedCardinalities) throws Exception {
        final InternalDistinctFacet toSend = new InternalDistinctFacet("bar", counts);
        final InternalDistinctFacet toReceive = new InternalDistinctFacet();
        serializeAndDeserialize(toSend, toReceive);
        final ExtTLongObjectHashMap<DistinctCountPayload> receivedCounts =
                new ExtTLongObjectHashMap<DistinctCountPayload>(toReceive.peekCounts());
        compareDistinctCounts(expectedCounts, expectedCardinalities, receivedCounts);
    }
View Full Code Here
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
        if(searchResponse.getFailedShards() > 0) {
            System.out.println(searchResponse);
            fail(Joiner.on(", ").join(searchResponse.getShardFailures()));
        }

        InternalDistinctFacet facet = searchResponse.getFacets().facet("stats1");
        assertThat(facet.getName(), equalTo("stats1"));
        assertThat(facet.getEntries().size(), equalTo(2));
        assertThat(facet.getEntries().get(0).getTime(), equalTo(utcTimeInMillis("2009-03-05")));
        assertThat(facet.getEntries().get(0).getTotalCount(), equalTo(2l));
        assertThat(facet.getEntries().get(0).getDistinctCount(), equalTo(2l));
        assertThat(facet.getEntries().get(1).getTime(), equalTo(utcTimeInMillis("2009-03-06")));
        assertThat(facet.getEntries().get(1).getTotalCount(), equalTo(1l));
        assertThat(facet.getEntries().get(1).getDistinctCount(), equalTo(1l));
        assertThat(facet.getTotalCount(), equalTo(3l));
        assertThat(facet.getDistinctCount(), equalTo(3l));

        // time zone causes the dates to shift by 2
        facet = searchResponse.getFacets().facet("stats2");
        assertThat(facet.getName(), equalTo("stats2"));
        assertThat(facet.getEntries().size(), equalTo(2));
        assertThat(facet.getEntries().get(0).getTime(), equalTo(utcTimeInMillis("2009-03-04")));
        assertThat(facet.getEntries().get(0).getTotalCount(), equalTo(1l));
        assertThat(facet.getEntries().get(0).getDistinctCount(), equalTo(1l));
        assertThat(facet.getEntries().get(1).getTime(), equalTo(utcTimeInMillis("2009-03-05")));
        assertThat(facet.getEntries().get(1).getTotalCount(), equalTo(2l));
        assertThat(facet.getEntries().get(1).getDistinctCount(), equalTo(2l));
        assertThat(facet.getTotalCount(), equalTo(3l));
        assertThat(facet.getDistinctCount(), equalTo(3l));

        // time zone causes the dates to shift by 2
        facet = searchResponse.getFacets().facet("stats3");
        assertThat(facet.getName(), equalTo("stats3"));
        assertThat(facet.getEntries().size(), equalTo(2));
        assertThat(facet.getEntries().get(0).getTime(), equalTo(utcTimeInMillis("2009-03-04")));
        assertThat(facet.getEntries().get(0).getTotalCount(), equalTo(1l));
        assertThat(facet.getEntries().get(0).getDistinctCount(), equalTo(1l));
        assertThat(facet.getEntries().get(1).getTime(), equalTo(utcTimeInMillis("2009-03-05")));
        assertThat(facet.getEntries().get(1).getTotalCount(), equalTo(2l));
        assertThat(facet.getEntries().get(1).getDistinctCount(), equalTo(2l));
        assertThat(facet.getTotalCount(), equalTo(3l));
        assertThat(facet.getDistinctCount(), equalTo(3l));

        // time zone causes the dates to shift by 2
        //        facet = searchResponse.getFacets().facet("stats4");
        //        assertThat(facet.getName(), equalTo("stats4"));
        //        assertThat(facet.getEntries().size(), equalTo(2));
        //        assertThat(facet.getEntries().get(0).getTime(), equalTo(utcTimeInMillis("2009-03-04")));
        //        assertThat(facet.getEntries().get(0).getTotalCount(), equalTo(1l));
        //        assertThat(facet.getEntries().get(0).getDistinctCount(), equalTo(1l));
        //        assertThat(facet.getEntries().get(1).getTime(), equalTo(utcTimeInMillis("2009-03-05")));
        //        assertThat(facet.getEntries().get(1).getTotalCount(), equalTo(2l));
        //        assertThat(facet.getEntries().get(0).getDistinctCount(), equalTo(2l));

        facet = searchResponse.getFacets().facet("stats5");
        assertThat(facet.getName(), equalTo("stats5"));
        assertThat(facet.getEntries().size(), equalTo(2));
        assertThat(facet.getEntries().get(0).getTime(), equalTo(utcTimeInMillis("2009-03-05")));
        assertThat(facet.getEntries().get(0).getTotalCount(), equalTo(2l));
        assertThat(facet.getEntries().get(0).getDistinctCount(), equalTo(2l));
        assertThat(facet.getEntries().get(1).getTime(), equalTo(utcTimeInMillis("2009-03-06")));
        assertThat(facet.getEntries().get(1).getTotalCount(), equalTo(1l));
        assertThat(facet.getEntries().get(1).getDistinctCount(), equalTo(1l));
        assertThat(facet.getTotalCount(), equalTo(3l));
        assertThat(facet.getDistinctCount(), equalTo(3l));

        facet = searchResponse.getFacets().facet("stats6");
        assertThat(facet.getName(), equalTo("stats6"));
        assertThat(facet.getEntries().size(), equalTo(2));
        assertThat(facet.getEntries().get(0).getTime(), equalTo(utcTimeInMillis("2009-03-04") - TimeValue.timeValueHours(2).millis()));
        assertThat(facet.getEntries().get(0).getTotalCount(), equalTo(1l));
        assertThat(facet.getEntries().get(0).getDistinctCount(), equalTo(1l));
        assertThat(facet.getEntries().get(1).getTime(), equalTo(utcTimeInMillis("2009-03-05") - TimeValue.timeValueHours(2).millis()));
        assertThat(facet.getEntries().get(1).getTotalCount(), equalTo(2l));
        assertThat(facet.getEntries().get(1).getDistinctCount(), equalTo(2l));
        assertThat(facet.getTotalCount(), equalTo(3l));
        assertThat(facet.getDistinctCount(), equalTo(3l));

        facet = searchResponse.getFacets().facet("stats7");
        assertThat(facet.getName(), equalTo("stats7"));
        assertThat(facet.getEntries().size(), equalTo(1));
        assertThat(facet.getEntries().get(0).getTime(), equalTo(utcTimeInMillis("2009-01-01")));
        assertThat(facet.getTotalCount(), equalTo(3l));
        assertThat(facet.getDistinctCount(), equalTo(3l));
    }
View Full Code Here
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
            System.out.println(searchResponse);
            fail(Joiner.on(", ").join(searchResponse.getShardFailures()));
        }

        // time zone causes the dates to shift by 2:00
        InternalDistinctFacet facet = searchResponse.getFacets().facet("stats1");
        assertThat(facet.getName(), equalTo("stats1"));
        assertThat(facet.getEntries().size(), equalTo(2));
        assertThat(facet.getEntries().get(0).getTime(), equalTo(utcTimeInMillis("2009-03-05")));
        assertThat(facet.getEntries().get(0).getTotalCount(), equalTo(1l));
        assertThat(facet.getEntries().get(0).getDistinctCount(), equalTo(1l));
        assertThat(facet.getEntries().get(1).getTime(), equalTo(utcTimeInMillis("2009-03-06")));
        assertThat(facet.getEntries().get(1).getTotalCount(), equalTo(2l));
        assertThat(facet.getEntries().get(1).getDistinctCount(), equalTo(2l));
        assertThat(facet.getTotalCount(), equalTo(3l));
        assertThat(facet.getDistinctCount(), equalTo(3l));

        // time zone causes the dates to shift by 1:30
        facet = searchResponse.getFacets().facet("stats2");
        assertThat(facet.getName(), equalTo("stats2"));
        assertThat(facet.getEntries().size(), equalTo(2));
        assertThat(facet.getEntries().get(0).getTime(), equalTo(utcTimeInMillis("2009-03-05")));
        assertThat(facet.getEntries().get(0).getTotalCount(), equalTo(2l));
        assertThat(facet.getEntries().get(0).getDistinctCount(), equalTo(2l));
        assertThat(facet.getEntries().get(1).getTime(), equalTo(utcTimeInMillis("2009-03-06")));
        assertThat(facet.getEntries().get(1).getTotalCount(), equalTo(1l));
        assertThat(facet.getEntries().get(1).getDistinctCount(), equalTo(1l));
        assertThat(facet.getTotalCount(), equalTo(3l));
        assertThat(facet.getDistinctCount(), equalTo(3l));
    }
View Full Code Here
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
        putSync(newID(), 1, __days[4]);
        putSync(newID(), 1, __days[6]);
        assertEquals(4, countAll());
        final SearchResponse response = getHistogram(__days[0], __days[7], "day", __userField);
        assertEquals(4, response.getHits().getTotalHits());
        final InternalDistinctFacet facet = response.getFacets().facet(__facetName);
        final List<DistinctTimePeriod<NullEntry>> facetList = facet.entries();
        // Expecting just one hit and one distinct hit per doc, for the username.
        assertEquals(4, facetList.size());
        assertEquals(__days[0], facetList.get(0).getTime());
        assertEquals(1, facetList.get(0).getTotalCount());
        assertEquals(1, facetList.get(0).getDistinctCount());
        assertEquals(__days[2], facetList.get(1).getTime());
        assertEquals(1, facetList.get(1).getTotalCount());
        assertEquals(1, facetList.get(1).getDistinctCount());
        assertEquals(__days[4], facetList.get(2).getTime());
        assertEquals(1, facetList.get(2).getTotalCount());
        assertEquals(1, facetList.get(2).getDistinctCount());
        assertEquals(__days[6], facetList.get(3).getTime());
        assertEquals(1, facetList.get(3).getTotalCount());
        assertEquals(1, facetList.get(3).getDistinctCount());
        assertThat(facet.getTotalCount(), equalTo(4l));
        assertThat(facet.getDistinctCount(), equalTo(1l)); // same user each time
    }
View Full Code Here
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
        putSync(newID(), 1, __days[4]);
        putSync(newID(), 1, __days[6]);
        assertEquals(4, countAll());
        final SearchResponse response = getHistogram(__days[0], __days[7], "day", __txtField);
        assertEquals(4, response.getHits().getTotalHits());
        final InternalDistinctFacet facet = response.getFacets().facet(__facetName);
        final List<DistinctTimePeriod<NullEntry>> facetList = facet.entries();
        // Expecting one hit for each token in the string "Document created [at] <TIMESTAMP>"
        // for each document, in this case these are unique per bucket too. The word "at"
        // is a stopword and is removed.
        assertEquals(4, facetList.size());
        assertEquals(__days[0], facetList.get(0).getTime());
        assertEquals(3, facetList.get(0).getTotalCount());
        assertEquals(3, facetList.get(0).getDistinctCount());
        assertEquals(__days[2], facetList.get(1).getTime());
        assertEquals(3, facetList.get(1).getTotalCount());
        assertEquals(3, facetList.get(1).getDistinctCount());
        assertEquals(__days[4], facetList.get(2).getTime());
        assertEquals(3, facetList.get(2).getTotalCount());
        assertEquals(3, facetList.get(2).getDistinctCount());
        assertEquals(__days[6], facetList.get(3).getTime());
        assertEquals(3, facetList.get(3).getTotalCount());
        assertEquals(3, facetList.get(3).getDistinctCount());
        assertThat(facet.getTotalCount(), equalTo(12l));
        assertThat(facet.getDistinctCount(), equalTo(6l)); // "document", "created", 4 usernames
    }
View Full Code Here
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
        putSync(newID(), 3, __days[6] + 10);
        putSync(newID(), 4, __days[6] + 20);
        assertEquals(8, countAll());
        final SearchResponse response = getHistogram(__days[0], __days[7], "day", __userField);
        assertEquals(8, response.getHits().getTotalHits());
        final InternalDistinctFacet facet = response.getFacets().facet(__facetName);
        final List<DistinctTimePeriod<NullEntry>> facetList = facet.entries();
        // Hits and distinct hits can now vary in intervals where the same user posted more
        // than once (i.e. day 0 here).
        assertEquals(4, facetList.size());
        assertEquals(__days[0], facetList.get(0).getTime());
        assertEquals(3, facetList.get(0).getTotalCount());
        assertEquals(2, facetList.get(0).getDistinctCount());
        assertEquals(__days[2], facetList.get(1).getTime());
        assertEquals(1, facetList.get(1).getTotalCount());
        assertEquals(1, facetList.get(1).getDistinctCount());
        assertEquals(__days[4], facetList.get(2).getTime());
        assertEquals(1, facetList.get(2).getTotalCount());
        assertEquals(1, facetList.get(2).getDistinctCount());
        assertEquals(__days[6], facetList.get(3).getTime());
        assertEquals(3, facetList.get(3).getTotalCount());
        assertEquals(3, facetList.get(3).getDistinctCount());
        assertThat(facet.getTotalCount(), equalTo(8l));
        assertThat(facet.getDistinctCount(), equalTo(4l)); // 4 different users
    }
View Full Code Here
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
        putSync(newID(), 3, __days[6] + 10);
        putSync(newID(), 4, __days[6] + 20);
        assertEquals(8, countAll());
        final SearchResponse response = getHistogram(__days[0], __days[7], "day", __txtField);
        assertEquals(8, response.getHits().getTotalHits());
        final InternalDistinctFacet facet = response.getFacets().facet(__facetName);
        final List<DistinctTimePeriod<NullEntry>> facetList = facet.entries();
        // Now things get a bit more complex as all the posts are identically worded apart
        // from the timestamp at the end. 3 tokens indexed per each instance of the field.
        assertEquals(4, facetList.size());
        assertEquals(__days[0], facetList.get(0).getTime());
        assertEquals(3 * 3, facetList.get(0).getTotalCount());
        assertEquals(2 + (1 * 3), facetList.get(0).getDistinctCount());
        assertEquals(__days[2], facetList.get(1).getTime());
        assertEquals(1 * 3, facetList.get(1).getTotalCount());
        assertEquals(1 * 3, facetList.get(1).getDistinctCount());
        assertEquals(__days[4], facetList.get(2).getTime());
        assertEquals(1 * 3, facetList.get(2).getTotalCount());
        assertEquals(1 * 3, facetList.get(2).getDistinctCount());
        assertEquals(__days[6], facetList.get(3).getTime());
        assertEquals(3 * 3, facetList.get(3).getTotalCount());
        assertEquals(2 + (1 * 3), facetList.get(3).getDistinctCount());
        assertThat(facet.getTotalCount(), equalTo(24l));
        assertThat(facet.getDistinctCount(), equalTo(10l)); // "document", "created", 8 usernames
    }
View Full Code Here
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
            final int totalItems = add(itemsPerDay);
            assertEquals(totalItems, countAll());

            System.out.println("Randomized testing: running facet");
            final SearchResponse response = getHistogram(__days[0], __days[7], "day", __userField, 1000);
            final InternalDistinctFacet facet1 = response.getFacets().facet(__facetName);
            final List<DistinctTimePeriod<NullEntry>> facetList1 = facet1.entries();
            assertEquals(7, facetList1.size());
            assertEquals(totalItems, facet1.getTotalCount());
            int tolerance = totalItems / 100;
            int totalDistinct = totalItems;
            assertTrue(String.format(
                    "With %d total distinct items: Estimated overall distinct count %d is not within 1%% tolerance of %d",
                    totalDistinct, facet1.getDistinctCount(), totalDistinct),
                    abs(totalDistinct - facet1.getDistinctCount()) <= tolerance);
            for(int i = 0; i < 7; i++) {
                final int exactUsers = itemsPerDay[i];
                assertEquals(exactUsers, facetList1.get(i).getTotalCount());
                tolerance = exactUsers / 100;
                final long fuzzyUsers = facetList1.get(i).getDistinctCount();
                //System.out.println("Exact user count = " + exactUsers);
                //System.out.println("Fuzzy user count = " + fuzzyUsers);
                assertTrue(String.format(
                        "With > %d terms per day: Estimated count %d is not within 1%% tolerance of %d",
                        minPerDay, fuzzyUsers, exactUsers),
                        abs(fuzzyUsers - exactUsers) <= tolerance);
            }

            final SearchResponse response2 = getHistogram(__days[0], __days[7], "day", __txtField, 1000);
            final InternalDistinctFacet facet2 = response2.getFacets().facet(__facetName);
            final List<DistinctTimePeriod<NullEntry>> facetList2 = facet2.entries();
            assertEquals(7, facetList2.size());
            assertEquals(3 * totalItems, facet2.getTotalCount());
            tolerance = totalItems / 100;
            totalDistinct = 2 + totalItems;
            assertTrue(String.format(
                    "With %d total distinct items: Estimated overall distinct count %d is not within 1%% tolerance of %d",
                    totalDistinct, facet2.getDistinctCount(), totalDistinct),
                    abs(totalDistinct - facet2.getDistinctCount()) <= tolerance);
            for(int i = 0; i < 7; i++) {
                final int exactTokens = itemsPerDay[i] * 3; // "Document created [by] <ID>"
                final int exactDistinctTokens = itemsPerDay[i] + 2;
                assertEquals(exactTokens, facetList2.get(i).getTotalCount());
                tolerance = exactDistinctTokens / 100;
View Full Code Here
129
130
131
132
133
134
135
136
            }
            _occurrences.put(fieldVal, null); // Free this up for GC immediately
        }

        _occurrences = null;
        final InternalFacet facet = new InternalDistinctFacet(facetName, counts);
        return facet;
    }
View Full Code Here
735
736
737
738
739
740
741
742
743
744
745
746
747
    HiveOutputFormat<?, ?> hiveOutputFormat = null;
    Class<? extends Writable> outputClass = null;
    boolean isCompressed = conf.getCompressed();
    TableDesc tableInfo = conf.getTableInfo();
    try {
      Serializer serializer = (Serializer) tableInfo.getDeserializerClass().newInstance();
      serializer.initialize(null, tableInfo.getProperties());
      outputClass = serializer.getSerializedClass();
      hiveOutputFormat = conf.getTableInfo().getOutputFileFormatClass().newInstance();
    } catch (SerDeException e) {
      throw new HiveException(e);
    } catch (InstantiationException e) {
      throw new HiveException(e);
View Full Code Here
TOP

Related Classes of com.pearson.entech.elasticsearch.search.facet.approx.date.internal.InternalDistinctFacet$Serializer

  • com.facebook.giraph.hive.impl.output.OutputInfo
  • com.facebook.giraph.hive.output.HiveApiOutputFormat
  • com.facebook.giraph.hive.output.OutputInfo
  • com.facebook.hiveio.output.HiveApiOutputFormat
  • com.facebook.hiveio.output.OutputInfo
  • com.google.appengine.api.datastore.Blob
  • com.google.appengine.api.datastore.Entity
  • com.google.appengine.api.datastore.Key
  • com.netflix.ice.processor.Ec2InstanceReservationPrice
  • com.netflix.ice.tag.Region
    • 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.