elasticsearch6聚合排序总结 - JavaAPI

序言:

        近来一直在javaAPI对ES数据进行统计分析, 中间也遇到不少问题, 本篇文章只针对聚合过程中的排序方法进行总结

例如:select  avg(age) as avg_age from table group by gender;

 

聚合的两种排序方式

       1. 按组内字段排序: 分组后按组内数据某个字段进行排序

       2. 按聚合后指标排序: 分组聚合后按照聚合后的某个新指标字段进行排序

 

案例:

1. 按组内字段排序

AggregationBuilder aggregation =
    AggregationBuilders
        .terms("agg").field("gender")
        .subAggregation(
            AggregationBuilders.topHits("top")
                .explain(true)
                .size(1)
                .from(10)
                .sort("sortFiled", SortOrder.ASC)
        );

输出

import org.elasticsearch.search.aggregations.bucket.terms.Terms;
import org.elasticsearch.search.aggregations.metrics.tophits.TopHits;

// sr is here your SearchResponse object
Terms agg = sr.getAggregations().get("agg");

// For each entry
for (Terms.Bucket entry : agg.getBuckets()) {
    String key = entry.getKey();                    // bucket key
    long docCount = entry.getDocCount();            // Doc count
    logger.info("key [{}], doc_count [{}]", key, docCount);

    // We ask for top_hits for each bucket
    TopHits topHits = entry.getAggregations().get("top");
    for (SearchHit hit : topHits.getHits().getHits()) {
        logger.info(" -> id [{}], _source [{}]", hit.getId(), hit.getSourceAsString());
    }
}

2. 按聚合后指标排序

 

import org.elasticsearch.search.aggregations.BucketOrder;

 doc_count以递增的方式

AggregationBuilders
    .terms("genders")
    .field("gender")
    .order(BucketOrder.count(true))

按升序方式按字母顺序按顺序排序 

AggregationBuilders
    .terms("genders")
    .field("gender")
    .order(BucketOrder.key(true))

按聚合名称标识对桶进行排序 

AggregationBuilders
    .terms("genders")
    .field("gender")
    .order(BucketOrder.aggregation("avg_height", false))
    .subAggregation(
        AggregationBuilders.avg("avg_height").field("height")
    )

 按多个聚合指标对桶进行排序

AggregationBuilders
    .terms("genders")
    .field("gender")
    .order(BucketOrder.compound( // in order of priority:
        BucketOrder.aggregation("avg_height", false), // sort by sub-aggregation first
        BucketOrder.count(true))) // then bucket count as a tie-breaker
    .subAggregation(
        AggregationBuilders.avg("avg_height").field("height")
    )

参考:https://www.elastic.co/guide/en/elasticsearch/client/java-api/current/_bucket_aggregations.html