如何在 mongoDB 中使用聚合进行多层计数细分？

Question

我有一个如下结构的文档。

{
  name: "John Doe",
  City : "OK",
  Prepaid: "Y"
},
{
  name: "Jane Doe",
  City : "CA",
  Prepaid: "N"
},
{
  name: "Jule Doe",
  City : "OK",
  Prepaid: "N"
},
{
  name: "Jake Doe",
  City : "OK",
  Prepaid: "Y"
}

我想先根据城市对它进行分组，然后再对预付费进行分组，然后分别获取每种预付费类型的计数。看起来与此类似的东西。

{
  City : OK
  Count : {
    "filter": prepaid,
    "count": {
      Y : 2
      N: 1
    }
  }
}

{
  City : CA
  Count : {
    "filter": prepaid,
    "count": {
      Y : 0
      N: 1
    }
  }
}

我尝试基于多个字段进行聚合，它给了我文档的总数，而不是细分。

这是我为聚合管道尝试的：

db.collection.aggregate([
    {$match:matchquery
    },{$group:{_id:{city: '$city', prepaid: '$prepaid' }, count:{$sum:1}
    }}
])

Answer 1

您可以先运行$group 两次以通过prepaid 计数，然后您可以应用$arrayToObject 以获取Y/N 作为对象键：

db.collection.aggregate([
    {
        $group: {
            _id: { city: "$City", prepaid: "$Prepaid" },
            total: { $sum: 1 }
        }
    },
    {
        $grop: {
            _id: "$_id.city",
            Count: {
                $push: {
                    k: "$_id.prepaid", v: "$total"
                }
            }
        }
    },
    {
        $project: {
            _id: 0,
            city: "$_id",
            Count: { $mergeObjects: [ { filter: "prepaid" }, { count: { $arrayToObject: "$Count" } } ] }
        }
    }
])

Mongo Playground