关于elasticsearch 聚合那点事1

╰半夏微凉° 2023-06-11 08:23 36阅读 0赞

# 1，根据某个字段聚合 #

如下是根据 city 字段聚合，然后按照文档的数量从大到小排序，返回5条数据。

{ 
      “query”: { 
        "bool": { 
           "filter": { 
           		.......query-dsl...... 
             }
          }
      }, 
      "aggs": { "types_count": { "term": { "field": "city", "size": 5, "ordre": { "doc_count": "desc"}}}}
    }

# 2， 分组求和： select field1, field2,sum(\*) from dd group by field1 #

{ 
      "query": { 
       ......query-dsl...... 
       },
       "aggs": { 
            "aggs-name": { 
    			"terms": { 
    				"field": "city", 
    				"size": 100,
    				"order": "totalNum"
    			},
    			"aggs": { 
    				"totalNum": { "sum": { "field": "product_num"}}
    			}
    		}
        }
    }

# 3，关于嵌套对象聚合分组求和的查询。 #

{ 
    	"query":{ 
    	......query-dsl......
    	},
    	"aggs": { 
    		"aggs-name": { 
    			"nested": { 
    				"path": "provice_nested"
    			},
    			"aggs": { 
    				"aggs-name": { 
    					"terms": { 
    						"field": "city",
    						"size": 100,
    						"order": { "total_num": "desc"}
    					},
    					"aggs": { 
    						"total_num": { "sum":{ "field": "product_num"}}
    					}
    				}
    			}
    		}
    	}
    }

4， 关于多个字段的，脚本查询，按照多个字段组合分组，并排多个聚合。

{ 
    	"query": { 
    	......query-dsl......
      	},
      	"aggs": { 
    		"aggs-name": { 
    			"terms": { 
    				"script": { 
    					"lang": "painless",
    					"source": "doc['provice'].value" + '-split-' + "doc['city'].value"
    				},
    				"size": 500
    			},
    			"aggs":{ 
    				"aggs-name1": { 
    				    "top_hits": { 
    						"_source": ["field1", "field2", ]
    						"size": 1,
    						"sort": { "timestamp": "desc"}
    					}
    				},
    				"aggs-name2": { 
    					"terms": { 
    						"script": { 
    							"lang": "painless",
    							"source": "doc['category'].value" + '-split-' + "doc['brand'].value"
    						},
    						"size": 10
    					}
    				}
    			}
    		}
    	}
    }

# 5, 关于多个字段的，脚本查询，按照多个字段组合分组，并排多个聚合，并且求和。【包含嵌套对象】 #

{ 
    	"query": { 
    	......query-dsl......
      	},
      	"aggs": { 
    		"aggs-name": { 
    			"terms": { 
    				"script": { 
    					"lang": "painless",
    					"source": "doc['provice'].value" + '-split-' + "doc['city'].value"
    				},
    				"size": 500
    			},
    			"aggs":{ 
    			    "aggs-name0-total-num": { 
    					"sum": { "field": "product-num"}
    				}
    				"aggs-name1": { 
    				    "top_hits": { 
    						"_source": ["field1", "field2", ]
    						"size": 1,
    						"sort": { "timestamp": "desc"}
    					}
    				},
    				"aggs-name2": { 
    					"terms": { 
    						"script": { 
    							"lang": "painless",
    							"source": "doc['category'].value" + '-split-' + "doc['brand'].value"
    						},
    						"size": 10
    					}
    				},
    				"aggs-name3": { 
    					"nested": { 
    						"path": "province"
    					},
    					"aggs": { 
    						"nested-agg-name": { 
    							"terms": { 
    								"script": { 
    									"lang": "painless",
    									"source": "doc['category'].value" + '-split-' + "doc['brand'].value"
    								},
    								"size": 10
    							},
    							"aggs": { "total_price": { "sum": { "field": "province.price"}}}
    						}
    					}
                    }
    			}
    		}
    	}
    }

# 6, 关于terms 的include #

{ 
    	"size": 0,
    	"query":{ ...},
    	"aggs": { 
    		"type_count": { 
                "terms": { 
                    "field": "brand",
                    "include": [
    					"nike", "qiaodan", "addas"
    				]
                 }
             }
        }
    }