有没有很好的方法来相交不同的收藏

Question

我正在实施一个商店经理，我需要通过mapreduce和appgregation访问我的mongodb。我访问的这个ShopList看起来像这样，

{
    //this is one shop
    {   //key & unique
        id: '0000001',
        //device count actually
        deviceVoltage:{high:1200, mid: 2400, low: 2300},
        lastUpdatedTime: Date(2019,1,1),
        //shop name
        name:'test1'
    },
    //this is another shop
    {
        id:'0000002',
        deviceVoltage:{high:2000, mid: 100, low: 3000},
        lastUpdatedTime: Date(2019,1,2),
        name:'test2'
    },
    ...
}

由于某种原因，我必须应用过滤器，例如服务器端的高电压比过滤器。也就是说，我需要为这些属性设置上限和下限，并将其应用于每个商店。由于此操作的复杂性，我必须使用mapreduce和聚合。让我们做一个真实的案例。

task: Please filter all shops and return shops with 
      (1) device high voltage ratio>10%, 
      (2) shop name contains keyword 'pharmac' and 
      (3) last updated time is in a week.

Step1: sum the device voltage array, and calcualte the ratio of [high] 
(done by mapReduce. the collection contains an array of many maps of 
 shop_id and the calculated high-voltage device ratio)

Step2: get a collection from Step1, use mongoDb query to filter shop names and last updated time 
(done by aggregation and pipe, 
 the collection contains an array of shops objects)

Step3: intersect the collection from Step1 and the result from Step2 by the shop id 
(but, how to do this?)

有没有一种方法可以很好地完成Step3？

欢迎提出任何想法。