Question

我有一个包含用户名，注册日期和赚取积分的数据集，我希望看到最接近的用户在平均水平附近（分别高25％和低25％）。

我尝试在第一阶段获得平均值：

db.people.aggregate([
        {$group: {_id: null, "average_points" : { $avg : "$points" } }},
        { $project: {_id: 1,
            "name": 1,
            "points": 1,
            "avg_25_h": {$multiply: [ "$average_points", 1.25 ] },
            "avg_25_l": {$multiply: [ "$average_points", 0.75 ] } } }
         ] )

结果正确：

{
    "_id" : null,
    "avg_25_h" : 624350.9491925,
    "avg_25_l" : 374610.5695155
}

但是，当我尝试匹配时，我只能获得在迷路时达到平均水平的用户：

db.people.aggregate([
{$group: {_id: null, "average_points" : { $avg : "$points" } }},
{ $project: {_id: null,
    "name": 1,
    "points": 1,
    "avg_25_h": {$multiply: [ "$average_points", 1.25 ] },
    "avg_25_l": {$multiply: [ "$average_points", 0.75 ] } } },
    {$match: {"points": {$gte: "$avg_25_l", $lte:"avg_25_h"}}}
     ] )

我尝试使用实际的_id并淘汰了$ match阶段，但随后它计算出各个平均值并将其分别乘以0.75和1.25。

有人可以帮我吗？

谢谢！

######## UPDATE

我终于明白了。这是我所做的：

db.people.aggregate([
                {$group: {_id: null, "avg_points" : { $avg : "$points" } }},
                {$lookup:
                {
                    from: "people",
                    let: { points : "$points", avg_points: "$avg_points"},
                    pipeline: [
                    {$match: 
                        { $expr: 
                                {$and:
                                    [
                                    {$gte: ["$points", {$multiply: ["$$avg_points", 0.75]}]},
                                    {$lte: ["$points", {$multiply: ["$$avg_points", 1.25]}]}
                                    ]
                                }
                        }
                    },
                    { $project: { padding: 0, signup: 0, current_date: 0 } },
                    {$limit: 20},
                    {$sort: {"points": -1}}
                    ],
                as: "UsersAroundTheAvg"
                } },
            { $unwind: "$UsersAroundTheAvg" },
            { $project: { _id: 0, padding: 0, signup: 0, points: 0, current_date: 0, name: 0} }
            ])

MongoDB-$ project和$ group没有显示结果

0 个答案: