Question

我是Mongo的新手，并且有一个关于使用条件计算的聚合查询的问题：

我有一个评论集，每个文档都包含一个情绪评分。我想：

1）按项目分组审核

2）获取该项目的所有评论中每个项目的平均情绪分数，并按此

排序

3）获取每个项目组的评论总数

4）获取每个项目的积极情绪评论总数（例如，＃情绪评分> 75的评论）

5）获取每个项目的负面情绪评论总数（例如，情绪评分<75的＃评论）

到目前为止，我有以下查询，涵盖1-3，但不知道如何在这里获得4/5：

db.reviews.aggregate( 
    {"$group" : 
        {_id: "$item", 
        sentiment: {$avg : "$sentimentScore"}, 
        count: {$sum: 1 } 
        } 
    }, 
    {"$sort": { sentiment: -1 } } 
)

Answer 1

我假设您希望为count分别设置sentiment字段，其中包含给定阈值的负值和正值，即positive - >75和negative - <75，即总数正面情绪和负面情绪总数以及总情绪。

db.sentiments.aggregate([
    {"$group" : 
        {_id: "$item", 
        sentiment: {$avg : "$sentiment_score"}, 
        postiive_sentiments: {$sum: { $cond: { if: { $gt: [ "$sentiment_score", 75 ] }, then: 1, else: 0 } }},
        negative_sentiments: {$sum: { $cond: { if: { $lt: [ "$sentiment_score", 75 ] }, then: 1, else: 0 } }},
        count: {$sum: 1 } 
        } 
    },
     {"$sort": { sentiment: -1 } } 
])

示例数据：

{ "_id" : ObjectId("5991329ea37dbc24842a68be"), "item" : "test1", "sentiment_score" : 50 }
{ "_id" : ObjectId("599132a2a37dbc24842a68bf"), "item" : "test1", "sentiment_score" : 40 }
{ "_id" : ObjectId("599132a4a37dbc24842a68c0"), "item" : "test1", "sentiment_score" : 80 }
{ "_id" : ObjectId("599132aba37dbc24842a68c1"), "item" : "test2", "sentiment_score" : 80 }
{ "_id" : ObjectId("599132ada37dbc24842a68c2"), "item" : "test2", "sentiment_score" : 30 }
{ "_id" : ObjectId("599132b0a37dbc24842a68c3"), "item" : "test2", "sentiment_score" : 38 }
{ "_id" : ObjectId("599132b6a37dbc24842a68c4"), "item" : "test3", "sentiment_score" : 78 }
{ "_id" : ObjectId("599132b9a37dbc24842a68c5"), "item" : "test3", "sentiment_score" : 88 }
{ "_id" : ObjectId("599132bba37dbc24842a68c6"), "item" : "test3", "sentiment_score" : 58 }
{ "_id" : ObjectId("599132c4a37dbc24842a68c7"), "item" : "test3", "sentiment_score" : 98 }
{ "_id" : ObjectId("599132cba37dbc24842a68c8"), "item" : "test4", "sentiment_score" : 65 }
{ "_id" : ObjectId("599132d2a37dbc24842a68c9"), "item" : "test4", "sentiment_score" : 30 }
{ "_id" : ObjectId("599132d6a37dbc24842a68ca"), "item" : "test4", "sentiment_score" : 10 }

//结果：

{ "_id" : "test3", "sentiment" : 80.5, "negative_sentiments" : 3, "positive_sentiments" : 1, "count" : 4 }
{ "_id" : "test1", "sentiment" : 56.666666666666664, "negative_sentiments" : 1, "positive_sentiments" : 2, "count" : 3 }
{ "_id" : "test2", "sentiment" : 49.333333333333336, "negative_sentiments" : 1, "positive_sentiments" : 2, "count" : 3 }
{ "_id" : "test4", "sentiment" : 35, "negative_sentiments" : 0, "positive_sentiments" : 3, "count" : 3 }

mongodb - 如何在聚合查询中编写条件

1 个答案: