es 排序 聚合统计

(27条消息) es 排序 聚合统计_吐血整理：一文看懂ES的R，查询与聚合_weixin_39901439的博客-CSDN博客

模糊匹配

select * from company where name like '%康康%'
GET red/_search
{
  "query": {
    "match": {
      "companyname": "康康"
    }
  }
}
上述查询会查出姓名中带有“康康”字的文档

GET red/_search
{
  "query": {
    "match_phrase": {  #match_phrase   查询的词，顺序不能变
      "companyname": "江苏省水"
    }
  }
}

分页查询

select * from company limit 0,10

GET red/_search
{
 "from": 0,
 "size": 5
}
【注意】from+size不能大于10000，也可以进行修改，但不建议这么操作，因为es主要分片模式，其会在每个分片都会执行一样的查询，然后再进行汇总排序，如果数据太大，会撑爆内存。例如每个分片都查询出10000条，总共5个分片，最后就会进行50000条数据的排序，最后再取值。

范围查询并进行排序

select * from company where redtime>=10 and redtime<=50 order by redtime desc
GET red/_search
{
  "query": {
    "range": {
      "redtime": {
        "gte": "2012-08-23",
        "lte": "2013-01-30"
      }
    }
  },
  "sort": [
    {
      "redtime": {
        "order": "desc"
      }
    }
  ]
}

多字段匹配查询

select * from company where sex like '%男%' or name like '%男%'
GET red/_search
{
  "query": {
    "multi_match": {
      "query": "鸡",
      "fields": ["companyname","title"]
    }
  }
}

bool查询(结构化查询)

结构化查询主要有三块，分别是must，should，must_not，filter

(1)must：里面的条件都是“并”关系，都匹配

(2)should：里面的条件都是“或”关系，有一个条件匹配就行

(3)must_not：里面的条件都是“并”关系，都不能匹配

(4)filter：过滤查询，不像其它查询需要计算_score相关性，它不进行此项计算，故比query查询快



select * from company where provinceid=50
and departments like '市场部'

GET red/_search
{
  "query": {
    "bool": {
      "must": [
        {
          "term": {
              "provinceid": "330000"
          }
        }
      ],
      "filter": [
        {
          "match":{
            "companyname":"大"
          }
        }
      ]
    }
  }
}

 

cbi360.net.apiwebjzw

cbi360.net.apijzw

聚合（指标聚合）

select max(age) from company
#取最大、最小、平均、总和
GET red/_search
{
  "aggs": {
    "provinceid_Max": {
      "max": {
        "field": "provinceid"
      }
    }
  },
  "size": 1
}


#对某个字段的值进行去重之后再取总数
select count(distinct(sex)) from company
GET red/_search
{
  "aggs": {
    "provinceid_Max": {
      "cardinality": {
        "field": "provinceid"
      }
    }
  },
  "size": 1
}


#stats聚合，对某个字段一次性返回count，max，min，avg和sum五个指标
select count(distinct age),sum(age),avg(age),max(age),min(age) from company
GET red/_search
{
  "aggs": {
    "provinceid_stats": {
      "stats": {
        "field": "provinceid"
      }
    }
  },
  "size": 1
}

percentiles聚合，对某个字段的值进行百分位统计 （统计各个值出现的百分比）
GET red/_search
{
  "aggs": {
    "provinceid_stats": {
      "percentiles": {
        "field": "provinceid"
      }
    }
  },
  "size": 1
}

聚合（桶聚合）

select sex,count(1) from company group by sex     分组
GET red/_search
{
  "aggs": {
    "provinceid_stats": {
      "terms": {
        "field": "provinceid"
      }
    }
  },
  "size": 1
}

 private readonly JudgementService judgementService;
        private readonly CourtAnnouncementService courtAnnouncementService;
        private readonly CourtNoticeService courtNoticeService;
        private readonly DishonestService dishonestService;
        private readonly DishonestExecutorService dishonestExecutorService;
        private readonly BusinessRiskService businessRiskService;
        private readonly MPledgeService mPledgeService;
        private readonly PenaltyService penaltyService;
        private readonly PledgeService pledgeService;

Max:1271923

Min:502

1188

Es优化： (21条消息) ES 性能调优，这可能是全网最详细的 Elasticsearch 性能调优指南_Elastic开源社区的博客-CSDN博客_es调优