Python MongoDB聚合然后平均

我有一个使用聚合框架的MongoDB查询，如下所示：

System._get_collection().aggregate([
        { "$match": {
            "system_id": system.id,
            "utc_timestamp": {
                "$gte": datetime.datetime.now().replace(hour=0, minute=0, second=0, microsecond=0) - datetime.timedelta(days=1),
                "$lt": datetime.datetime.now().replace(hour=0, minute=0, second=0, microsecond=0)
            }
        }},
        { "$group": {
            "_id": { "$dayOfYear": "$utc_timestamp" },
            "correct": {
                "$sum": { "$cond": [
                    { "$eq": [ "$status", 455 ]}, 1, 0
                ]}
            },
            "total_count": { "$sum": 1 }
        }},
        { "$project": {
            "correctness": {
                "$cond": [
                    { "$eq": [ "$correct", 0 ] },
                    0,
                    { "$multiply" : [{"$divide": [ "$correct", "$total_count" ]}, 100] }
                ]
            }
        }}
    ])

{'result': [{'_id': 272, 'correctness': 99.89373007438896}], 'ok': 1.0}

{'result': [{'_id': 272, 'correctness': 99.89373007438896, 'delay': 5}], 'ok': 1.0}

输出如下所示：

System._get_collection().aggregate([
        { "$match": {
            "system_id": system.id,
            "utc_timestamp": {
                "$gte": datetime.datetime.now().replace(hour=0, minute=0, second=0, microsecond=0) - datetime.timedelta(days=1),
                "$lt": datetime.datetime.now().replace(hour=0, minute=0, second=0, microsecond=0)
            }
        }},
        { "$group": {
            "_id": { "$dayOfYear": "$utc_timestamp" },
            "correct": {
                "$sum": { "$cond": [
                    { "$eq": [ "$status", 455 ]}, 1, 0
                ]}
            },
            "total_count": { "$sum": 1 }
        }},
        { "$project": {
            "correctness": {
                "$cond": [
                    { "$eq": [ "$correct", 0 ] },
                    0,
                    { "$multiply" : [{"$divide": [ "$correct", "$total_count" ]}, 100] }
                ]
            }
        }}
    ])

{'result': [{'_id': 272, 'correctness': 99.89373007438896}], 'ok': 1.0}

{'result': [{'_id': 272, 'correctness': 99.89373007438896, 'delay': 5}], 'ok': 1.0}

我想在结果中添加一个字段，它是每个匹配的

System

文档中所有

delay

字段的平均值

我试图将

“delay”：{“$avg”：“$delay”}，

添加到

$group

部分，但没有改变结果

我想要这样的结果：

System._get_collection().aggregate([
        { "$match": {
            "system_id": system.id,
            "utc_timestamp": {
                "$gte": datetime.datetime.now().replace(hour=0, minute=0, second=0, microsecond=0) - datetime.timedelta(days=1),
                "$lt": datetime.datetime.now().replace(hour=0, minute=0, second=0, microsecond=0)
            }
        }},
        { "$group": {
            "_id": { "$dayOfYear": "$utc_timestamp" },
            "correct": {
                "$sum": { "$cond": [
                    { "$eq": [ "$status", 455 ]}, 1, 0
                ]}
            },
            "total_count": { "$sum": 1 }
        }},
        { "$project": {
            "correctness": {
                "$cond": [
                    { "$eq": [ "$correct", 0 ] },
                    0,
                    { "$multiply" : [{"$divide": [ "$correct", "$total_count" ]}, 100] }
                ]
            }
        }}
    ])

{'result': [{'_id': 272, 'correctness': 99.89373007438896}], 'ok': 1.0}

{'result': [{'_id': 272, 'correctness': 99.89373007438896, 'delay': 5}], 'ok': 1.0}

注意

“延迟”：5

以上应为计算的平均值

---

我是否需要以某种方式将其添加到投影中？或者，我做错了什么吗？

除了

\u id

之外，您从

$project

中省略的任何字段都不包括在输出中，因此您只需向其添加

延迟

：

{ "$project": {
    "correctness": {
        "$cond": [
            { "$eq": [ "$correct", 0 ] },
            0,
            { "$multiply" : [{"$divide": [ "$correct", "$total_count" ]}, 100] }
        ]
    },
    "delay": 1
}}