Pymongo：具有$ lookup的子查询用于同一集合，而具有$ in的用于某些ID这是正确的方法吗？

如何解决Pymongo：具有$ lookup的子查询用于同一集合，而具有$ in的用于某些ID这是正确的方法吗？

我是MongoDB的新手，我需要在其中执行某种子查询：

• 返回每周前三名运营商（$ in中的数组）的记录计数。

我的收藏如下：

"MONTH","DAY_OF_MONTH","DAY_OF_WEEK","FL_DATE","OP_UNIQUE_CARRIER","ORIGIN_CITY_NAME","DEST_CITY_NAME","ARR_DELAY","DISTANCE",3,28,2018-03-28,"B6","Newark,NJ","Fort Myers,FL",4.00,1068.00,29,4,2018-03-29,6.00,30,5,2018-03-30,-4.00,31,6,2018-03-31,-3.00,1,2018-03-01,"New York,NY","Long Beach,CA",39.00,2465.00,2,2018-03-02,0.00,2018-03-03,2.00,7,2018-03-04,-2.00,2018-03-05,25.00,2018-03-06,-21.00,2018-03-07,20.00,8,2018-03-08,48.00,9,2018-03-09,16.00,10,2018-03-10,11,2018-03-11,-13.00,

而我当前的查询仅返回前3个中的每一个的#count，但在$ in阶段包含了变量值：

total_per_week = [
    {"$match" :{
        "OP_UNIQUE_CARRIER": { "$in": ["DL","9E","B6"]},"ORIGIN_CITY_NAME": {"$regex": "^New York"}
        }
    },{"$group" : {
        "_id" : {                   
            "week": { "$week": "$FL_DATE" },"carrier": "$OP_UNIQUE_CARRIER"
        },"total" : { "$sum": 1 } 
        }
    },#// Maybe project a prettier "flatter" output
    {"$project": {
        "_id": 0,"carrier": "$_id.carrier","week": "$_id.week","total": "$total"
    }},{"$sort": SON([("carrier",1),("week",1)])}

]

这是输出：

    carrier  week  total
0       9E     0    597
1       9E     1    964
2       9E     2    917
3       9E     3    968
4       9E     4    975
5       9E     5    927
6       9E     6    933
7       9E     7    917
8       9E     8    978
9       9E     9   1025
10      9E    10   1036
11      9E    11   1036
12      9E    12   1036
13      B6     0    797
14      B6     1    880

（...）

因此，我试图模仿这种SQL方法：

SELECT COUNT（*），星期，运营商 （）中的运营商 按运营商分组，每周

通过伪装此查询：

    lookQuery = [
    # Returns df with carrier column
    {"$match" : {"ORIGIN_CITY_NAME": {"$regex": "^New York"}} },{"$sortByCount":  "$OP_UNIQUE_CARRIER" },{"$limit": 3},#adds a new array (inner),top3,field whose elements are the matching documents from the “joined” collection
    { "$lookup": 
        {
        "from": "Flights","let": {"the_carrier": "$_id" },"pipeline": [
        {"$match" :{
            "$expr": {"$eq": ["$OP_UNIQUE_CARRIER","$$the_carrier"]},"ORIGIN_CITY_NAME": {"$regex": "^New York"}
            }
        },{"$group" : {
            "_id" : {                   
            "week": { "$week": "$FL_DATE" },"carrier": "$OP_UNIQUE_CARRIER"
            },"total_sem" : { "$sum": 1 } 
            }
        }         
      ],"as": "top3" #output of inner-array field
      }
    } #closes lookup,{"$unwind": "$top3"},{"$project": {
            "_id": 0,"Carrier": "$top3._id.carrier","Week": "$top3._id.week","Total": "$top3.total_sem"
            }
        }
        #adds a new outter array to math the carrier code in Carriers ref Table,providing the Description,{"$lookup":
          {
          "from": "Carriers","localField": "Carrier","foreignField": "Code","as": "carriers"
          }  
        },"Carrier": "$Carrier","Week": "$Week","Total": "$Total","CarrierName": "$carriers.Description"
            }
        },{"$sort": SON([("Carrier",("Week",1)])}      

]

产生所需输出：

    Carrier  Week  Total             CarrierName
0       9E     0    597     [Endeavor Air Inc.]
1       9E     1    964     [Endeavor Air Inc.]
2       9E     2    917     [Endeavor Air Inc.]
3       9E     3    968     [Endeavor Air Inc.]
4       9E     4    975     [Endeavor Air Inc.]
5       9E     5    927     [Endeavor Air Inc.]
6       9E     6    933     [Endeavor Air Inc.]
7       9E     7    917     [Endeavor Air Inc.]
8       9E     8    978     [Endeavor Air Inc.]
9       9E     9   1025     [Endeavor Air Inc.]
10      9E    10   1036     [Endeavor Air Inc.]
11      9E    11   1036     [Endeavor Air Inc.]
12      9E    12   1036     [Endeavor Air Inc.]
13      B6     0    797       [JetBlue Airways]
14      B6     1    880       [JetBlue Airways]
15      B6     2    873       [JetBlue Airways]

我的问题是：这个“查询”效率有效还是过于复杂？ 有没有更简单的方法可以达到类似或更好的性能呢？