MongoDB 之 aggregate $group 巧妙運用

有這樣一組數據：

{ "campaign_id": "A", "campaign_name": "A", "subscriber_id": "123" }, { "campaign_id": "A", "campaign_name": "A", "subscriber_id": "123" }, { "campaign_id": "A", "campaign_name": "A", "subscriber_id": "456" }

按照  campaign_id 與 campaign_name  分組，並查詢出每個分組下的記錄條數 及  subscriber_id  不同記錄的個數

關系型數據庫SQL示例：

select campaign_id,campaign_name,count(subscriber_id),count(distinct subscriber_id) group by campaign_id,campaign_name from campaigns;

在MongoDB下就存在兩種組合：1） campaign_id, campaign_name, subscriber_id  三個相同的分為一組，

                                                  2） campaign_id, campaign_name  兩個相同，subscriber_id 不同分為一組，

最后通過這兩種分組查詢出按照  campaign_id 與 campaign_name  分組，subscriber_id 不同記錄的個數

MongoDB示例：

db.campaigns.aggregate([ { "$match": { "subscriber_id": { "$ne": null }}}, // Count all occurrences { "$group": { "_id": { "campaign_id": "$campaign_id", "campaign_name": "$campaign_name", "subscriber_id": "$subscriber_id" }, "count": { "$sum": 1 } }}, // Sum all occurrences and count distinct { "$group": { "_id": { "campaign_id": "$_id.campaign_id", "campaign_name": "$_id.campaign_name" }, "totalCount": { "$sum": "$count" }, "distinctCount": { "$sum": 1 } }} ])

文檔結果：

第一個 group：

{ "_id" : { "campaign_id" : "A", "campaign_name" : "A", "subscriber_id" : "456" }, "count" : 1 } { "_id" : { "campaign_id" : "A", "campaign_name" : "A", "subscriber_id" : "123" }, "count" : 2 }

第二個 group：

{ "_id" : { "campaign_id" : "A", "campaign_name" : "A" }, "totalCount" : 3, "distinctCount" : 2 }

至此，我們已經查詢出一共有 3 條記錄，subscriber_id 有兩種不同的值

reference： Mongodb中Aggregation特性