获取不同值的总数

Question

我的Couch数据库中有以下数据结构。我需要得到每天不同sessions_id的计数。记录也必须通过app_version进行筛选。

我当前的MapReduce看起来像这个

"dau": {
           "map": "function(doc) {
                  date = doc.session.timestamp.split("T")[0];
                  emit([date, doc.app_version, doc.session.session_id], 1);
                  }",
           "reduce": "_count"
       }

但这给了我以下的输出

["2014-08-20", 2, "kOnNIhCNQ31LlkpEPQ7XnN1D"] 1
["2014-08-20", 2, "TOnNIhCNQ31LlkpEPQ7XnN1D"] 2
["2014-08-21", 2, "TOnNIhCNQ31LlkpEPQ7XnN1D"] 2

我需要将其打印为-因为这是每天的唯一session_id的数量

["2014-08-20", 2] 2
["2014-08-21", 2] 1

有什么想法吗？

任何帮助都将不胜感激。

[
        {
            "app_version": 2,
            "platform": "android",
            "session": {
                "timestamp": "2014-08-20T00:00:00.000Z",
                "session_id": "TOnNIhCNQ31LlkpEPQ7XnN1D",
                "ip": "202.150.213.66",
                "location": "1.30324,103.5498"
            }
        },
    {
        "app_version": 2,
        "platform": "android",
        "session": {
            "timestamp": "2014-08-21T00:00:00.000Z",
            "session_id": "TOnNIhCNQ31LlkpEPQ7XnN1D",
            "ip": "202.150.213.66",
            "location": "1.30324,103.5498"
        }
    }
    {
        "app_version": 2,
        "platform": "ios",
        "session": {
            "timestamp": "2014-08-21T00:00:00.000Z",
            "session_id": "TOnNIhCNQ31LlkpEPQ7XnN1D",
            "ip": "202.150.213.66",
            "location": "1.30324,103.5498"
        }
    },
    {
        "app_version": 1,
        "platform": "ios",
        "session": {
            "timestamp": "2014-08-21T00:00:00.000Z",
            "session_id": "TOnNIhCNQ31LlkpEPQ7XnN1D",
            "ip": "202.150.213.66",
            "location": "1.30324,103.5498"
        }
        }
    ]

Answer 1

将app_version放在视图的第一位：

emit([doc.app_version, date, doc.session.session_id], 1);

然后使用group_level=2 进行查询