elasticsearch-dsl聚合-1

接续上篇，本篇介绍elasticsearch聚合查询，使用python库elasticsearch-dsl进行聚合查询操作。

7.3、聚合查询

高阶概念

• Buckets(桶/集合):满足特定条件的文档的集合

• Metrics(指标):对桶内的文档进行统计计算(例如最小值,求和,最大值等)

  • 新建一张测试表

     PUT cars
     {
       "mappings": {
         "transactions":{
           "properties": {
             "price":{
               "type": "integer"
             },
             "color":{
               "type": "text",
               "fielddata": true
             },
             "make":{
               "type": "text",
               "fielddata": true
             },
             "sold":{
               "type": "date",
               "format": "yyyy-MM-dd"
             }
           }
         }
       }
     }

    插入数据

     POST /cars/transactions/_bulk
     { "index": {"_index": "cars", "_type": "transactions"}}
     { "price" : , "color" : "red", "make" : "honda", "sold" : "2014-10-28" }
     { "index": {"_index": "cars", "_type": "transactions"}}
     { "price" : , "color" : "red", "make" : "honda", "sold" : "2014-11-05" }
     { "index": {"_index": "cars", "_type": "transactions"}}
     { "price" : , "color" : "green", "make" : "ford", "sold" : "2014-05-18" }
     { "index": {"_index": "cars", "_type": "transactions"}}
     { "price" : , "color" : "blue", "make" : "toyota", "sold" : "2014-07-02" }
     { "index": {"_index": "cars", "_type": "transactions"}}
     { "price" : , "color" : "green", "make" : "toyota", "sold" : "2014-08-19" }
     { "index": {"_index": "cars", "_type": "transactions"}}
     { "price" : , "color" : "red", "make" : "honda", "sold" : "2014-11-05" }
     { "index": {"_index": "cars", "_type": "transactions"}}
     { "price" : , "color" : "red", "make" : "bmw", "sold" : "2014-01-01" }
     { "index": {"_index": "cars", "_type": "transactions"}}
     { "price" : , "color" : "blue", "make" : "ford", "sold" : "2014-02-12" }

  • 查询哪个颜色的汽车销量最好（按颜色分类）

     GET cars/transactions/_search
     {
       "size": ,
       "aggs": {
         "popular_colors": {
           "terms": {
             "field": "color"
           }
         }
       }
     }

     s = Search(index='cars')
     a = A("terms", field="color")
     s.aggs.bucket("popular_color", a)
     response = s.execute()

    或者

     s.aggs.bucket("popular_color", "terms", field="color")

  • 查询每种颜色车的平均价格

     GET cars/transactions/_search
     {
       "size": ,
       "aggs": {
         "colors": {
           "terms": {
             "field": "color"
           },
           "aggs": {
             "avg_price": {
               "avg": {
                 "field": "price"
               }
             }
           }
         }
       }
     }

     s = Search(index='cars')
     a1 = A("terms", field="color")
     a2 = A("avg", field="price")
     s.aggs.bucket("colors", a1).metric("avg_price", a2)
     response = s.execute()

    或者

     s = Search(index='cars')
     s.aggs.bucket("colors", "terms", field="color").metric("avg_price", "avg", field="price")
     response = s.execute()

  • 先按颜色分，再按品牌分，再求每种品牌的均价

     GET cars/transactions/_search
     {
       "size": ,
       "aggs": {
         "colors": {
           "terms": {
             "field": "color"
           },
           "aggs": {
             "make": {
               "terms": {
                 "field": "make"
               },
               "aggs": {
                 "avg_price": {
                   "avg": {
                     "field": "price"
                   }
                 }
               }
             }
           }
         }
       }
     }

     s = Search(index='cars')
     s.aggs.bucket("colors", "terms", field="color")
     s.aggs["colors"].bucket("make", "terms", field="make")
     s.aggs["colors"].aggs["make"].metric("avg_price", "avg", field="price")
     response = s.execute()

  • 先按颜色分，再按品牌分，再求每种品牌的最高和最低价

     GET cars/transactions/_search
     {
       "size": ,
       "aggs": {
         "colors": {
           "terms": {
             "field": "color"
           },
           "aggs": {
             "make": {
               "terms": {
                 "field": "make"
               },
               "aggs": {
                 "min_price": {
                   "min": {
                     "field": "price"
                   }
                 },
                 "max_price": {
                   "max": {
                     "field": "price"
                   }
                 }
               }
             }
           }
         }
       }
     }

     s = Search(index='cars')
     s.aggs.bucket("colors", "terms", field="color")
     s.aggs["colors"].bucket("make", "terms", field="make")
     s.aggs["colors"].aggs["make"].metric("min_price", "min", field="price")
     s.aggs["colors"].aggs["make"].metric("max_price", "max", field="price")
     response = s.execute()

  • 未完待续...