Added unit tests for terms and histogram sorting on sub aggregations that return NaN.

This commit is contained in:
Adrien Grand 2014-02-25 10:57:32 +01:00
parent 0e57915536
commit 66b589d58e
4 changed files with 231 additions and 15 deletions

View File

@ -0,0 +1,43 @@
/*
* Licensed to Elasticsearch under one or more contributor
* license agreements. See the NOTICE file distributed with
* this work for additional information regarding copyright
* ownership. Elasticsearch licenses this file to you under
* the Apache License, Version 2.0 (the "License"); you may
* not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing,
* software distributed under the License is distributed on an
* "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
* KIND, either express or implied. See the License for the
* specific language governing permissions and limitations
* under the License.
*/
package org.elasticsearch.common.util;
import java.util.Comparator;
/**
* {@link Comparator}-related utility methods.
*/
public enum Comparators {
;
/**
* Compare <code>d1</code> against <code>d2</code>, pushing {@value Double#NaN} at the bottom.
*/
public static int compareDiscardNaN(double d1, double d2, boolean asc) {
if (Double.isNaN(d1)) {
return Double.isNaN(d2) ? 0 : 1;
} else if (Double.isNaN(d2)) {
return -1;
} else {
return asc ? Double.compare(d1, d2) : Double.compare(d2, d1);
}
}
}

View File

@ -21,6 +21,7 @@ package org.elasticsearch.search.aggregations.bucket;
import org.elasticsearch.ElasticsearchIllegalArgumentException; import org.elasticsearch.ElasticsearchIllegalArgumentException;
import org.elasticsearch.common.text.Text; import org.elasticsearch.common.text.Text;
import org.elasticsearch.common.util.Comparators;
import org.elasticsearch.search.aggregations.Aggregation; import org.elasticsearch.search.aggregations.Aggregation;
import org.elasticsearch.search.aggregations.Aggregations; import org.elasticsearch.search.aggregations.Aggregations;
import org.elasticsearch.search.aggregations.metrics.MetricsAggregation; import org.elasticsearch.search.aggregations.metrics.MetricsAggregation;
@ -99,7 +100,7 @@ public interface MultiBucketsAggregation extends Aggregation {
public int compare(B b1, B b2) { public int compare(B b1, B b2) {
double v1 = value(b1); double v1 = value(b1);
double v2 = value(b2); double v2 = value(b2);
return asc ? Double.compare(v1, v2) : Double.compare(v2, v1); return Comparators.compareDiscardNaN(v1, v2, asc);
} }
private double value(B bucket) { private double value(B bucket) {

View File

@ -21,6 +21,7 @@ package org.elasticsearch.search.aggregations.bucket.terms;
import com.google.common.primitives.Longs; import com.google.common.primitives.Longs;
import org.elasticsearch.common.io.stream.StreamInput; import org.elasticsearch.common.io.stream.StreamInput;
import org.elasticsearch.common.io.stream.StreamOutput; import org.elasticsearch.common.io.stream.StreamOutput;
import org.elasticsearch.common.util.Comparators;
import org.elasticsearch.common.xcontent.XContentBuilder; import org.elasticsearch.common.xcontent.XContentBuilder;
import org.elasticsearch.search.aggregations.AggregationExecutionException; import org.elasticsearch.search.aggregations.AggregationExecutionException;
import org.elasticsearch.search.aggregations.Aggregator; import org.elasticsearch.search.aggregations.Aggregator;
@ -212,13 +213,7 @@ class InternalOrder extends Terms.Order {
double v2 = ((MetricsAggregator.MultiValue) aggregator).metric(valueName, ((InternalTerms.Bucket) o2).bucketOrd); double v2 = ((MetricsAggregator.MultiValue) aggregator).metric(valueName, ((InternalTerms.Bucket) o2).bucketOrd);
// some metrics may return NaN (eg. avg, variance, etc...) in which case we'd like to push all of those to // some metrics may return NaN (eg. avg, variance, etc...) in which case we'd like to push all of those to
// the bottom // the bottom
if (Double.isNaN(v1)) { return Comparators.compareDiscardNaN(v1, v2, asc);
return Double.isNaN(v2) ? 0 : 1;
}
if (Double.isNaN(v2)) {
return -1;
}
return asc ? Double.compare(v1, v2) : Double.compare(v2, v1);
} }
}; };
} }
@ -230,13 +225,7 @@ class InternalOrder extends Terms.Order {
double v2 = ((MetricsAggregator.SingleValue) aggregator).metric(((InternalTerms.Bucket) o2).bucketOrd); double v2 = ((MetricsAggregator.SingleValue) aggregator).metric(((InternalTerms.Bucket) o2).bucketOrd);
// some metrics may return NaN (eg. avg, variance, etc...) in which case we'd like to push all of those to // some metrics may return NaN (eg. avg, variance, etc...) in which case we'd like to push all of those to
// the bottom // the bottom
if (Double.isNaN(v1)) { return Comparators.compareDiscardNaN(v1, v2, asc);
return Double.isNaN(v2) ? 0 : 1;
}
if (Double.isNaN(v2)) {
return -1;
}
return asc ? Double.compare(v1, v2) : Double.compare(v2, v1);
} }
}; };
} }

View File

@ -0,0 +1,183 @@
/*
* Licensed to Elasticsearch under one or more contributor
* license agreements. See the NOTICE file distributed with
* this work for additional information regarding copyright
* ownership. Elasticsearch licenses this file to you under
* the Apache License, Version 2.0 (the "License"); you may
* not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing,
* software distributed under the License is distributed on an
* "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
* KIND, either express or implied. See the License for the
* specific language governing permissions and limitations
* under the License.
*/
package org.elasticsearch.search.aggregations.bucket;
import org.elasticsearch.action.search.SearchResponse;
import org.elasticsearch.common.settings.ImmutableSettings;
import org.elasticsearch.common.settings.Settings;
import org.elasticsearch.common.util.Comparators;
import org.elasticsearch.common.xcontent.XContentBuilder;
import org.elasticsearch.search.aggregations.Aggregation;
import org.elasticsearch.search.aggregations.bucket.histogram.Histogram;
import org.elasticsearch.search.aggregations.bucket.terms.Terms;
import org.elasticsearch.search.aggregations.metrics.MetricsAggregationBuilder;
import org.elasticsearch.search.aggregations.metrics.avg.Avg;
import org.elasticsearch.search.aggregations.metrics.stats.extended.ExtendedStats;
import org.elasticsearch.test.ElasticsearchIntegrationTest;
import org.junit.Before;
import org.junit.Test;
import static org.elasticsearch.common.xcontent.XContentFactory.jsonBuilder;
import static org.elasticsearch.search.aggregations.AggregationBuilders.*;
import static org.hamcrest.core.IsNull.notNullValue;
public class NaNSortingTests extends ElasticsearchIntegrationTest {
@Override
public Settings indexSettings() {
return ImmutableSettings.builder()
.put("index.number_of_shards", between(1, 5))
.put("index.number_of_replicas", between(0, 1))
.build();
}
private enum SubAggregation {
AVG("avg") {
@Override
public MetricsAggregationBuilder<?> builder() {
return avg(name).field("numeric_field");
}
@Override
public double getValue(Aggregation aggregation) {
return ((Avg) aggregation).getValue();
}
},
VARIANCE("variance") {
@Override
public MetricsAggregationBuilder<?> builder() {
return extendedStats(name).field("numeric_field");
}
@Override
public String sortKey() {
return name + ".variance";
}
@Override
public double getValue(Aggregation aggregation) {
return ((ExtendedStats) aggregation).getVariance();
}
},
STD_DEVIATION("std_deviation"){
@Override
public MetricsAggregationBuilder<?> builder() {
return extendedStats(name).field("numeric_field");
}
@Override
public String sortKey() {
return name + ".std_deviation";
}
@Override
public double getValue(Aggregation aggregation) {
return ((ExtendedStats) aggregation).getStdDeviation();
}
};
SubAggregation(String name) {
this.name = name;
}
public String name;
public abstract MetricsAggregationBuilder<?> builder();
public String sortKey() {
return name;
}
public abstract double getValue(Aggregation aggregation);
}
@Before
public void init() throws Exception {
createIndex("idx");
final int numDocs = randomIntBetween(2, 10);
for (int i = 0; i < numDocs; ++i) {
final long value = randomInt(5);
XContentBuilder source = jsonBuilder().startObject().field("long_value", value).field("double_value", value + 0.05).field("string_value", "str_" + value);
if (randomBoolean()) {
source.field("numeric_value", randomDouble());
}
client().prepareIndex("idx", "type").setSource(source.endObject()).execute().actionGet();
}
refresh();
ensureSearchable();
}
private void assertCorrectlySorted(Terms terms, boolean asc, SubAggregation agg) {
assertThat(terms, notNullValue());
double previousValue = asc ? Double.NEGATIVE_INFINITY : Double.POSITIVE_INFINITY;
for (Terms.Bucket bucket : terms.getBuckets()) {
Aggregation sub = bucket.getAggregations().get(agg.name);
double value = agg.getValue(sub);
assertTrue(Comparators.compareDiscardNaN(previousValue, value, asc) <= 0);
previousValue = value;
}
}
private void assertCorrectlySorted(Histogram histo, boolean asc, SubAggregation agg) {
assertThat(histo, notNullValue());
double previousValue = asc ? Double.NEGATIVE_INFINITY : Double.POSITIVE_INFINITY;
for (Histogram.Bucket bucket : histo.getBuckets()) {
Aggregation sub = bucket.getAggregations().get(agg.name);
double value = agg.getValue(sub);
assertTrue(Comparators.compareDiscardNaN(previousValue, value, asc) <= 0);
previousValue = value;
}
}
public void testTerms(String fieldName) {
final boolean asc = randomBoolean();
SubAggregation agg = randomFrom(SubAggregation.values());
SearchResponse response = client().prepareSearch("idx")
.addAggregation(terms("terms").field(fieldName).subAggregation(agg.builder()).order(Terms.Order.aggregation(agg.sortKey(), asc)))
.execute().actionGet();
final Terms terms = response.getAggregations().get("terms");
assertCorrectlySorted(terms, asc, agg);
}
@Test
public void stringTerms() {
testTerms("string_value");
}
@Test
public void longTerms() {
testTerms("long_value");
}
@Test
public void doubleTerms() {
testTerms("double_value");
}
@Test
public void longHistogram() {
final boolean asc = randomBoolean();
SubAggregation agg = randomFrom(SubAggregation.values());
SearchResponse response = client().prepareSearch("idx")
.addAggregation(histogram("histo")
.field("long_value").interval(randomIntBetween(1, 2)).subAggregation(agg.builder()).order(Histogram.Order.aggregation(agg.sortKey(), asc)))
.execute().actionGet();
final Histogram histo = response.getAggregations().get("histo");
assertCorrectlySorted(histo, asc, agg);
}
}