Python PyMongo按日期计算元素数

我与我的用户有数据库：

{
    "_id": {
        "$oid": "5a0decadefcb09087c08a868"
    },
    "user_id": "5b232a5a-b333-4320-ba63-722b9e167ef3",
    "email": "email@email.com",
    "password": "***",
    "registration_date": {
        "$date": "2017-11-16T19:53:17.946Z"
    },
    "type": "user"
},

{
    "_id": {
        "$oid": "5a0ded3aefcb090887d7f4fb"
    },
    "user_id": "0054bbde-3ba0-490f-8d54-ffaf72958888",
    "email": "second@gmail.com",
    "password": "***",
    "registration_date": {
        "$date": "2017-11-16T19:55:38.194Z"
    },
    "type": "user"
}

我想按每个日期（注册日期）计算用户数，并得到如下结果：

01.01.2017 – 10
01.02.2017 – 20
01.03.2017 – 15
...

我正在尝试该代码，但它不起作用：

def registrations_by_date(self):
    users = self.users_db.aggregate([
        {'$group': {
            '_id': {'registration_date':'$date'},
            'count': {'$sum':1}
        }},
    ])
    return users

---

我做错了什么？如何获取这些数据

看来你有一个额外的，

db.userReg.aggregate([
{$group: {_id: "$registration_date", count: {$sum:1}}}
])

这将给出正确的结果（基于我的McHine上的记录）：

{ “_id”：ISODate（“2017-11-15T19:55:38.194Z”）， “计数”：1.0}

{ “_id”：ISODate（“2017-11-16T19:55:38.194Z”）， “计数”：2.0}

---

似乎您有一个额外的，

db.userReg.aggregate([
{$group: {_id: "$registration_date", count: {$sum:1}}}
])

这将给出正确的结果（基于我的McHine上的记录）：

{ “_id”：ISODate（“2017-11-15T19:55:38.194Z”）， “计数”：1.0}

{ “_id”：ISODate（“2017-11-16T19:55:38.194Z”）， “计数”：2.0}

如果架构中的日期为ISODate

然后，下面的聚合查询将工作，在分组之前完成日期格式，以便在分组数据时不使用时间戳

{
        "_id" : "5a0decadefcb09087c08a868",
        "user_id" : "5b232a5a-b333-4320-ba63-722b9e167ef3",
        "email" : "email@email.com",
        "password" : "***",
        "registration_date" : ISODate("2017-11-16T19:53:17.946Z"),
        "type" : "user"
}
{
        "_id" : "5a0ded3aefcb090887d7f4fb",
        "user_id" : "0054bbde-3ba0-490f-8d54-ffaf72958888",
        "email" : "second@gmail.com",
        "password" : "***",
        "registration_date" : ISODate("2017-11-16T19:55:38.194Z"),
        "type" : "user"
}

获取结果的聚合查询是

db.userReg.aggregate([
  {$project: 
    { formattedRegDate:
      { "$dateToString": {format:"%Y-%m-%d", date:"$registration_date"}}
    }
  },
  {$group:{_id:"$formattedRegDate", count:{$sum:1}}}]);

{ "_id" : "2017-11-16", "count" : 2 }

{ "_id" : { "date" : "2017-11-16" }, "count" : 2 }

结果是

db.userReg.aggregate([
  {$project: 
    { formattedRegDate:
      { "$dateToString": {format:"%Y-%m-%d", date:"$registration_date"}}
    }
  },
  {$group:{_id:"$formattedRegDate", count:{$sum:1}}}]);

{ "_id" : "2017-11-16", "count" : 2 }

{ "_id" : { "date" : "2017-11-16" }, "count" : 2 }

如果架构中的日期为字符串

然后使用下面的方法

样本数据

{
        "_id" : "5a0decadefcb09087c08a868",
        "user_id" : "5b232a5a-b333-4320-ba63-722b9e167ef3",
        "email" : "email@email.com",
        "password" : "***",
        "registration_date" : "2017-11-16T19:53:17.946Z",
        "type" : "user"
}
{
        "_id" : "5a0ded3aefcb090887d7f4fb",
        "user_id" : "0054bbde-3ba0-490f-8d54-ffaf72958888",
        "email" : "second@gmail.com",
        "password" : "***",
        "registration_date" : "2017-11-16T19:55:38.194Z",
        "type" : "user"
}

质疑

结果是

db.userReg.aggregate([
  {$project: 
    { formattedRegDate:
      { "$dateToString": {format:"%Y-%m-%d", date:"$registration_date"}}
    }
  },
  {$group:{_id:"$formattedRegDate", count:{$sum:1}}}]);

{ "_id" : "2017-11-16", "count" : 2 }

{ "_id" : { "date" : "2017-11-16" }, "count" : 2 }

如果架构中的日期为ISODate

然后，下面的聚合查询将工作，在分组之前完成日期格式，以便在分组数据时不使用时间戳

{
        "_id" : "5a0decadefcb09087c08a868",
        "user_id" : "5b232a5a-b333-4320-ba63-722b9e167ef3",
        "email" : "email@email.com",
        "password" : "***",
        "registration_date" : ISODate("2017-11-16T19:53:17.946Z"),
        "type" : "user"
}
{
        "_id" : "5a0ded3aefcb090887d7f4fb",
        "user_id" : "0054bbde-3ba0-490f-8d54-ffaf72958888",
        "email" : "second@gmail.com",
        "password" : "***",
        "registration_date" : ISODate("2017-11-16T19:55:38.194Z"),
        "type" : "user"
}

获取结果的聚合查询是

db.userReg.aggregate([
  {$project: 
    { formattedRegDate:
      { "$dateToString": {format:"%Y-%m-%d", date:"$registration_date"}}
    }
  },
  {$group:{_id:"$formattedRegDate", count:{$sum:1}}}]);

{ "_id" : "2017-11-16", "count" : 2 }

{ "_id" : { "date" : "2017-11-16" }, "count" : 2 }

结果是

db.userReg.aggregate([
  {$project: 
    { formattedRegDate:
      { "$dateToString": {format:"%Y-%m-%d", date:"$registration_date"}}
    }
  },
  {$group:{_id:"$formattedRegDate", count:{$sum:1}}}]);

{ "_id" : "2017-11-16", "count" : 2 }

{ "_id" : { "date" : "2017-11-16" }, "count" : 2 }

如果架构中的日期为字符串

然后使用下面的方法

样本数据

{
        "_id" : "5a0decadefcb09087c08a868",
        "user_id" : "5b232a5a-b333-4320-ba63-722b9e167ef3",
        "email" : "email@email.com",
        "password" : "***",
        "registration_date" : "2017-11-16T19:53:17.946Z",
        "type" : "user"
}
{
        "_id" : "5a0ded3aefcb090887d7f4fb",
        "user_id" : "0054bbde-3ba0-490f-8d54-ffaf72958888",
        "email" : "second@gmail.com",
        "password" : "***",
        "registration_date" : "2017-11-16T19:55:38.194Z",
        "type" : "user"
}

质疑

结果是

db.userReg.aggregate([
  {$project: 
    { formattedRegDate:
      { "$dateToString": {format:"%Y-%m-%d", date:"$registration_date"}}
    }
  },
  {$group:{_id:"$formattedRegDate", count:{$sum:1}}}]);

{ "_id" : "2017-11-16", "count" : 2 }

{ "_id" : { "date" : "2017-11-16" }, "count" : 2 }

---

有一个问题，它没有按天递增进行排序，“日期”：“2017-11-27”“日期”：“2017-11-21”“日期”：“2017-11-20”“日期”：“2017-11-26”对日期字段使用$sort聚合管道，

{$sort:{date 1}}

对升序和

{$sort:{date 1}}

对降序返回相同的顺序（-1或1无差异）–在Python端修复它。再次感谢！有一个问题，它没有按天递增进行排序，“日期”：“2017-11-27”“日期”：“2017-11-21”“日期”：“2017-11-20”“日期”：“2017-11-26”对日期字段使用$sort聚合管道，

{$sort:{date 1}}

对升序和

{$sort:{date 1}}

对降序返回相同的顺序（-1或1无差异）–在Python端修复它。再次感谢！