ES的聚合主要有Bucket、Metric、Pipeline Aggregations ，本篇主要介紹Bucket、Metric這兩種。

聚合：從數(shù)據(jù)中分組和提取數(shù)據(jù)。類似于 SQL GROUP BY 和 SQL 聚合函數(shù)

聚合語法：

"aggregations" : {
    "<聚合名稱 1>" : {
        "<聚合類型>" : {
            <聚合體內(nèi)容>
        }
        [,"元數(shù)據(jù)" : {  [<meta_data_body>] }]?
        [,"aggregations" : { [<sub_aggregation>]+ }]?
    }
    [,"聚合名稱 2>" : { ... }]*
}

實(shí)例數(shù)據(jù)

Elasticsearch 示例數(shù)據(jù) accounts.json 下載

kibana中執(zhí)行

POST /bank/account/_bulk

示例 1：搜索 address 中包含 Lane的所有人的年齡分布 ( 前 10 條 ) 以及平均年齡，以及平均薪資

DSL語句

GET bank/_search
{
  "size": 0,
  "query": {
    "match": {
      "address": "Lane"
    }
  },
  "aggs": {
    "ageAggs": {
      "terms": {
        "field": "age",
        "size": 10
      }
    },
    "avgAge": {
      "avg": {
        "field": "age"
      }
    },
    "avgPrice": {
      "avg": {
        "field": "balance"
      }
    }
  }
}

java代碼

public void testSearchData() throws IOException {
        SearchRequest request = new SearchRequest();
        request.indices("bank");
        SearchSourceBuilder sourceBuilder = new SearchSourceBuilder();
        // 1.1）address 中包含 road 的所有人
        sourceBuilder.query(QueryBuilders.matchQuery("address", "road"));
        // 1.2）按照年齡分布進(jìn)行聚合
        TermsAggregationBuilder ageAgg = AggregationBuilders.terms("ageAgg").field("age").size(10);
        sourceBuilder.aggregation(ageAgg);
        // 1.3）計(jì)算平均薪資
        AvgAggregationBuilder balanceAvg = AggregationBuilders.avg("balanceAvg").field("balance");
        sourceBuilder.aggregation((balanceAvg));
        System.out.println("檢索參數(shù)：" + sourceBuilder.toString());

        request.source(sourceBuilder);

        // 2、執(zhí)行檢索
        SearchResponse response = client.search(request, RequestOptions.DEFAULT);

        // 3、分析結(jié)果
        System.out.println(response.toString());
        // 3.1）獲取查到的數(shù)據(jù)。
        SearchHits hits = response.getHits();
        // 3.2）獲取真正命中的結(jié)果
        SearchHit[] searchHits = hits.getHits();
        // 3.3）遍歷命中結(jié)果
        for (SearchHit hit : searchHits) {
            String hitStr = hit.getSourceAsString();
            BankMember bankMember = JSON.parseObject(hitStr, BankMember.class);
            System.out.println(bankMember);
        }
        // 3.4）獲取聚合信息
        // 參考文檔：https://www.elastic.co/guide/en/elasticsearch/client/java-rest/current/java-rest-high-search.html
        Aggregations aggregations = response.getAggregations();
        Terms ageAgg1 = aggregations.get("ageAgg");
        for (Terms.Bucket bucket : ageAgg1.getBuckets()) {
            String keyAsString = bucket.getKeyAsString();
            System.out.println("用戶年齡： " + keyAsString + " 人數(shù)：" + bucket.getDocCount());
        }

        Avg balanceAvg1 = aggregations.get("balanceAvg");
        System.out.println("平均薪資：" + balanceAvg1.getValue());
    }

kibana查詢結(jié)果

java執(zhí)行結(jié)果

示例 2：按照年齡分組，然后將分組后的結(jié)果按照性別分組，然后查詢出這些分組后的平均薪資

DSL語句

GET bank/_search
{
  "size": 0,
  "query": {
    "match_all": {}
  },
  "aggs": {
    "ageAggs": {
      "terms": {
        "field": "age",
        "size": 10
      },
      "aggs": {
        "genderAgg": {
          "terms": {
            "field": "gender.keyword",
            "size": 10
          },
          "aggs": {
            "avgBalance": {
              "avg": {
                "field": "balance"
              }
            }
          }
        }
      }
    }
  }
}

java代碼

 @Test
    public void testAgeAggs2() throws IOException {
        SearchRequest request = new SearchRequest("bank");

        SearchSourceBuilder searchSourceBuilder = request.source();

        TermsAggregationBuilder ageAgg = AggregationBuilders.terms("ageAgg").field("age").size(10);

        searchSourceBuilder.query(QueryBuilders.matchAllQuery());
        ageAgg.subAggregation(AggregationBuilders.terms("genderAgg")
                .field("gender.keyword")
                .subAggregation(AggregationBuilders.avg("balanceAvg")
                        .field("balance"))
        );

        searchSourceBuilder.aggregation(ageAgg);

        request.source(searchSourceBuilder);

        SearchResponse response = client.search(request, RequestOptions.DEFAULT);

        Aggregations aggregations = response.getAggregations();
        //拿到桶
        Terms ageAgg1 = aggregations.get("ageAgg");
        //最外層年齡桶
        for (Terms.Bucket bucket : ageAgg1.getBuckets()) {

            String ageString = bucket.getKeyAsString();
            long ageDocCount = bucket.getDocCount();

            Aggregations genderAggregations = bucket.getAggregations();
            //年齡桶
            Terms genderTerm = genderAggregations.get("genderAgg");
            for (Terms.Bucket genderTermbucket : genderTerm.getBuckets()) {
                String genderString = genderTermbucket.getKeyAsString();
                long genderdocCount = genderTermbucket.getDocCount();

                //年齡指標(biāo)
                Avg balanceAvg = genderTermbucket.getAggregations().get("balanceAvg");
                System.out.println("用戶年齡： " + ageString + " 人數(shù)：" + ageDocCount + "用戶性別： " + genderString + " 人數(shù)：" + genderdocCount+ "平均薪資:" + balanceAvg.getValue());

            }
        }

    }

kibana查詢視圖

java查詢結(jié)果

參考資料：
Elasticsearch原理之a(chǎn)ggregations 聚合