MongoDB - 根據字段在小組賽中是否具有特定值來計數文檔

Question

我寫了一個 Mysql 查詢，如下所示：

SELECT account_id,date(date_posted) as date_posted,sum(if(message_format = 'e',1,0)) as email_count,sum(if(message_format = 't',1,0)) as 
sms_count,sum(if(message_format = 'p',1,0)) as push_count,sum(if(message_format = 's',1,0)) as slack_count
FROM `track_notifications` GROUP BY date(date_posted), account_id ORDER BY account_id

MySQL 結果如下圖所示：

我想得到與 MySQL 結果相同的MongoDB ，我是MongoDB的新手。 如何在 MongoDB 中編寫此查詢

我試過這樣：

db.Datasets.aggregate([{$project:{_id:0,account_id:1,date_posted:1,message_format:{$ifNull:["$message_format",0]}}},{$sort:{"account_id":1}}])

示例文檔：

_id:ObjectId(“5e8851dc53a883900d9b8e4b”)
account_id:52519
date_posted:”2018-11-19T15:38:52.000Z”
message_format:”t”

_id:ObjectId(“5e8851dc53a883900d9b8e4c”)
account_id:52519
date_posted:”2018-11-19T15:43:01.000Z”
message_format:”e”

_id:ObjectId(“5e8851dc53a883900d9b8e4d”)
account_id:52519
date_posted:”2018-11-19T15:38:51.000Z”
message_format:”e”

_id:ObjectId(“5e8851dc53a883900d9b8e4d”)
account_id:52519
date_posted:”2018-11-19T15:56:22.000Z”
message_format:”s”

_id:ObjectId(“5e8851dc53a883900d9b8e4e”)
account_id:52519
date_posted:”2018-11-19T15:56:22.000Z”
message_format:”p”

經過長時間的研究和嘗試，我被sum(if(message_format = 't',1,0)) ，請幫忙。

Answer 1

嘗試以下聚合查詢：

腳步：

1. 您將在account_id和轉換的date_posted上進行分組並對message_format字段進行條件檢查並根據字段值將1或0傳遞給$sum ，如果字段是null則將傳遞0 ，因此在對文檔進行分組時，如果 10 個文檔具有message_format == e然后email_count將是10 & rest others 將是0 。
2. 然后您可以添加從$_id中提取的account_id和date_posted字段。
3. 使用項目刪除不必要的_id字段。
4. 按account_id排序以按升序排列結果文檔。

詢問：

db.track_notifications.aggregate([
      {
        $group: {
          _id: {
            account_id: "$account_id",
            date_posted: {
              $dateToString: { format: "%Y-%m-%d", date: "$date_posted" },
            },
          },
          email_count: {
            $sum: { $cond: [{ $eq: ["$message_format", "e"] }, 1, 0] },
          },
          sms_count: { $sum: { $cond: [{ $eq: ["$message_format", "t"] }, 1, 0] } },
          push_count: {
            $sum: { $cond: [{ $eq: ["$message_format", "p"] }, 1, 0] },
          },
          slack_count: {
            $sum: { $cond: [{ $eq: ["$message_format", "s"] }, 1, 0] },
          },
        },
      },
      {
        $addFields: {
          account_id: "$_id.account_id",
          date_posted: "$_id.date_posted",
        },
      },
      { $project: { _id: 0 } },
      { $sort: { account_id: 1 } },
    ]);  

測試： MongoDB-Playground

參考：聚合管道運營商