Elasticsearch系列（三十四）Elasticsearch度量聚合统计

在日常中，除了使用正常的长讯之外，我们还会涉及到在Elasticsearch中进行数据统计。在elasticsearch中支持的聚合统计种类有：度量聚合统计和多桶型聚合统计。

度量聚合统计

是指一组文档的统计分科，可以得到最大值，最小值，平均值等结果

分桶型聚合统计

是指将匹配的文档切分为一个或多个桶，然后统计每个桶里面的文档数量

本篇文章主要介绍的是度量聚合统计。下面我们来演示一下。

一、准备一份indexmapping

put /devices

{
  "mappings": {
    "properties": {
      "device_uuid": {
        "type": "keyword"
      },
      "device_name": {
        "type": "keyword"
      },
      "device_category": {
        "type": "integer"
      },
      "device_score": {
        "type": "double"
      },
      "device_cts": {
        "type": "long"
      }
    }
  },
  "settings": {
    "index": {
      "number_of_shards": "1",
      "number_of_replicas": "0"
    }
  }
}

这里我们填写了设备名称，设备分类，设备分数，设备时间等几个参数，分别作用于不同的类型，用来演示接下来的聚合计算。

二、向devices索引里面插入数据

post /devices/_doc/1

{"device_uuid":"111","device_name":"设备111","device_category":"1","device_score":"1.01","device_cts":1669433306000}

post /devices/_doc/2

{"device_uuid":"222","device_name":"设备222","device_category":"1","device_score":"1.02","device_cts":1669433306000}

post /devices/_doc/3

{"device_uuid":"333","device_name":"设备333","device_category":"1","device_score":"1.03","device_cts":1669433306000}

post /devices/_doc/4

{"device_uuid":"444","device_name":"设备444","device_category":"2","device_score":"1.04","device_cts":1669433306000}

post /devices/_doc/5

{"device_uuid":"555","device_name":"设备555","device_category":"2","device_score":"1.05","device_cts":1669433306000}

post /devices/_doc/6

{"device_uuid":"666","device_name":"设备666","device_category":"2","device_score":"1.06","device_cts":1669433306000}

post /devices/_doc/7

{"device_uuid":"777","device_name":"设备777","device_category":"3","device_score":"1.07","device_cts":1669433306000}

post /devices/_doc/8

{"device_uuid":"888","device_name":"设备888","device_category":"3","device_score":"1.08","device_cts":1669433306000}

post /devices/_doc/9

{"device_uuid":"999","device_name":"设备999","device_category":"3","device_score":"1.09","device_cts":1669433306000}

存储数据的结果是：

三、使用度量聚合计算，求整个库的device_score的最大值，最小值，平均值，总和

post /devices/_search

{
  "size": 0,
  "aggs": {
    "min_price": {
      "min": {
        "field": "device_score"
      }
    },
    "max_price": {
      "max": {
        "field": "device_score"
      }
    },
    "avg_price": {
      "avg": {
        "field": "device_score"
      }
    },
    "sum_price": {
      "sum": {
        "field": "device_score"
      }
    }
  }
}

求出结果：

可以看到数据被正确的计算出来了。

除了上面直接聚合之外，我们还可以添加查询条件，例如：

post /devices/_search

{
  "size": 0,
  "query": {
    "match_all": {}
  },
  "aggs": {
    "min_price": {
      "min": {
        "field": "device_score"
      }
    },
    "max_price": {
      "max": {
        "field": "device_score"
      }
    },
    "avg_price": {
      "avg": {
        "field": "device_score"
      }
    },
    "sum_price": {
      "sum": {
        "field": "device_score"
      }
    }
  }
}