在没有 Stats Agg 或 Duplicate Aggs 错误的情况下获取范围的最小值和最大值？

Question

我觉得在实践中试图让管道聚合做我想做的事。 我会发布我所拥有的，但想法：

1. 创建一个日期范围，并为该范围内过去 10 个月的每个月创建存储桶。明白了。
2. 获取每个桶的“大小”字段的最小值和最大值。我只能弄清楚如何使用“stats”agg 来做到这一点，因为如果我尝试将两者都作为单独的 aggs 执行，则会出现重复错误。然而，我不想要其他统计数据。我可以避免这个统计数据吗？
3. 对分数求和。我到底是怎么做到的？那是在踢我的尾巴。不知道你能不能对 _score 字段求和。

所以这是我根据常见的地震概念练习的索引：

PUT _bulk
{ "index" : { "_index" : "earthquakes", "_id" : "1" } }
{ "date": "30-09-2020", "magnitude": "3.4", "lon": "74.12", "lat": "43.67" }
{ "index" : { "_index" : "earthquakes", "_id" : "2" } }
{ "date": "30-09-2020", "magnitude": "1.2", "lon": "78.02", "lat": "103.07" }
{ "index" : { "_index" : "earthquakes", "_id" : "3" } }
{ "date": "15-10-2020", "magnitude": "2.5", "lon": "178.02", "lat": "98.41" }
{ "index" : { "_index" : "earthquakes", "_id" : "4" } }
{ "date": "19-11-2020", "magnitude": "1.9", "lon": "14.67", "lat": "100.35" }
{ "index" : { "_index" : "earthquakes", "_id" : "5" } }
{ "date": "13-12-2020", "magnitude": "6.2", "lon": "123.93", "lat": "56.05" }
{ "index" : { "_index" : "earthquakes", "_id" : "6" } }
{ "date": "21-12-2020", "magnitude": "0.2", "lon": "130.31", "lat": "83.41" }
{ "index" : { "_index" : "earthquakes", "_id" : "7" } }
{ "date": "17-01-2021", "magnitude": "0.2", "lon": "10.31", "lat": "98.00" }
{ "index" : { "_index" : "earthquakes", "_id" : "8" } }
{ "date": "23-01-2021", "magnitude": "4.6", "lon": "112.31", "lat": "69.96" }
{ "index" : { "_index" : "earthquakes", "_id" : "9" } }
{ "date": "31-01-2021", "magnitude": "0.4", "lon": "79.43", "lat": "72.14" }
{ "index" : { "_index" : "earthquakes", "_id" : "10" } }
{ "date": "03-02-2021", "magnitude": "7.1", "lon": "120.80", "lat": "50.22" }

这是我汇总的内容。注意：在尝试对 _score 字段求和之前，我已将点击数设为 10...但没有发生：

GET earthquakes/_search
{
  "size": 0, 
  "aggs": {
    "range_mag": {
      "date_range": {
        "field": "date",
        "ranges": [
          {
            "from": "now-10M",
            "to": "now"
                }
              ]
            },
            "aggs": {
              "by_month_mag": {
                "date_histogram": {
                  "field": "date",
                  "calendar_interval": "month"
                },
                "aggs": {
                  "stat_mag": {
                    "stats": {
                      "field": "magnitude"
              }
            }
          }
        }
      }
    }
  }
}

^ 这可行，但要获得最小值和最大值，但会添加我不需要的数据。我没有把我的分数总和，因为它把我逼疯了。有没有更好的方法来完成我想做的事情？ 不管怎样，谢谢。在我可以轻松输入或使用文档完成的所有内容中，聚合只是我认为我会得到的一件事，但不知何故觉得卡住了。

Answer 1

对于最小值/最大值，您可以对同一字段进行单独的聚合，并将您所做的分数相加

    "score_sum":{
      "sum": {
        "script": "_score"
      }
    }

最终查询

GET earthquakes/_search
{
  "size": 0,
  "aggs": {
    "score_sum": {
      "sum": {
        "script": "_score"
      }
    },
    "range_mag": {
      "date_range": {
        "field": "date",
        "ranges": [
          {
            "from": "now-10M",
            "to": "now"
          }
        ]
      },
      "aggs": {
        "by_month_mag": {
          "date_histogram": {
            "field": "date",
            "calendar_interval": "month"
          },
          "aggs": {
            "min_mag": {
              "min": {
                "field": "magnitude"
              }
            },
            "max_mag": {
              "max": {
                "field": "magnitude"
              }
            }
          }
        }
      }
    }
  }
}