Elasticsearch 聚合类似于 group_concat

Question

我是 elasticsearch 的新手，我想创建一个 group_concat 聚合。但我不知道怎么办。有人可以帮我吗

示例数据：

POST /example_measures/_bulk
{"index":{"_id":1}}
{"id":"1","datapoint_id":"1","datetime":"1577833200000","value":"5"}
{"index":{"_id":2}}
{"id":"2","datapoint_id":"2","datetime":"1577833210000","value":"51"}
{"index":{"_id":3}}
{"id":"3","datapoint_id":"2","datetime":"1577833220000","value":"77"}

我想表达的sql：

select 
datapoint_id, 
group_concat(`datetime` order by `datetime` SEPARATOR ',' limit 5) as dt, 
group_concat(`value` order by `datetime` SEPARATOR ',' limit 5) as val 
from example_measures 
group by datapoint_id;

我希望每个数据点有 2 个数组。一个带有时间戳，一个带有值。

我使用 sql 语法没有成功，因为 sql 输入不支持 group_concat：

POST /_sql?format=txt
{
  "query":"..."
}

我使用 Kibana 和 Dev Tools 进行输入。

Answer 1

您可以通过在 datapoint_id 字段上使用 Terms Aggregation 来实现您的用例。这将创建桶 - 一个 pe 唯一值 datapoint_id。然后，您可以使用子聚合将桶进一步嵌入到这些独特的桶中。

搜索查询：

{
  "size": 0,
  "aggs": {
    "id": {
      "terms": {
        "field": "datapoint_id.keyword"
      },
      "aggs": {
        "dt": {
          "terms": {
            "field": "datetime.keyword",
            "order": { "_key" : "asc" },
            "size": 5
          }
        },
        "val": {
          "terms": {
            "field": "value.keyword",
            "size": 5
          }
        }
      }
    }
  }
}

搜索结果：

"aggregations": {
    "id": {
      "doc_count_error_upper_bound": 0,
      "sum_other_doc_count": 0,
      "buckets": [
        {
          "key": "2",
          "doc_count": 2,
          "val": {
            "doc_count_error_upper_bound": 0,
            "sum_other_doc_count": 0,
            "buckets": [
              {
                "key": "51",
                "doc_count": 1
              },
              {
                "key": "77",
                "doc_count": 1
              }
            ]
          },
          "dt": {
            "doc_count_error_upper_bound": 0,
            "sum_other_doc_count": 0,
            "buckets": [
              {
                "key": "1577833210000",
                "doc_count": 1
              },
              {
                "key": "1577833220000",
                "doc_count": 1
              }
            ]
          }
        },
        {
          "key": "1",
          "doc_count": 1,
          "val": {
            "doc_count_error_upper_bound": 0,
            "sum_other_doc_count": 0,
            "buckets": [
              {
                "key": "5",
                "doc_count": 1
              }
            ]
          },
          "dt": {
            "doc_count_error_upper_bound": 0,
            "sum_other_doc_count": 0,
            "buckets": [
              {
                "key": "1577833200000",
                "doc_count": 1
              }
            ]
          }
        }
      ]
    }
  }

Elasticsearch 聚合类似于 group_concat

Elasticsearch aggregation similar to group_concat

group-concat

aggregation

elasticsearch