请考虑以下学校的收藏学校:
{
"_id" : 1,
"name" : "Class 1",
"students" : [
{ "rollNo" : 10001, "name" : "Ram", "score" : 65 },
{ "rollNo" : 10002, "name" : "Shyam", "score" : 90 },
{ "rollNo" : 10003, "name" : "Mohan", "score" : 75 }
]
},
{
"_id" : 2,
"name" : "Class 2",
"students" : [
{ "rollNo" : 20001, "name" : "Krishna", "score" : 88 },
{ "rollNo" : 20002, "name" : "Sohan", "score" : 91 },
{ "rollNo" : 20003, "name" : "Radhika", "score" : 82 },
{ "rollNo" : 20004, "name" : "Komal", "score" : 55 },
{ "rollNo" : 20005, "name" : "Sonam", "score" : 91 }
]
},
{
"_id" : 3,
"name" : "Class 3",
"students" : [
{ "rollNo" : 30001, "name" : "Monika", "score" : 77 },
{ "rollNo" : 30002, "name" : "Rahul", "score" : 81 }
]
}
在这里,我的目标是从每个杂项Orderby降序得分中获取前N名学生(考虑按分数排序的前2名学生)。 我的预期结果是:
答案 0 :(得分:1)
如果您只希望学生记录而不进行分组,则可以简单地$unwind
,然后$sort
:
db.collection.aggregate([
{ $unwind: "$students" },
{ $sort: { "students.score": -1 } },
{ $limit: 2 }
])
这将留下students
对象。为了获得更清晰的输出,您可以将$replaceRoot
与$mergeObjects
结合使用:
db.collection.aggregate([
{ $unwind: "$students" },
{ $sort: { "students.score": -1 } },
{ $replaceRoot: {
newRoot: {
$mergeObjects: [ { _id: "$_id", class: "$name" }, "$students" ]
}
}
},
{ $limit: 2 }
])
这将为您提供以下输出:
[
{
"_id": 2,
"class": "Class 2",
"name": "Sonam",
"rollNo": 20005,
"score": 91
},
{
"_id": 2,
"class": "Class 2",
"name": "Sohan",
"rollNo": 20002,
"score": 91
}
]
更新:
使用它来获得每个组的前2名:
db.collection.aggregate([
{ $unwind: "$students" },
{ $sort: { "students.score": -1 }
},
{
$group: {
"_id": "$_id",
"name": { $first: "$name" },
"students": { $push: "$students" }
}
},
{
"$project": {
"top_two": { "$slice": [ "$students", 2 ] }
}
}
])