Elasticsearch - Count duplicated and unique values

问题

I have the following json

[
 {"firstname": "john", "lastname": "doe"},
 {"firstname": "john", "lastname": "smith"},
 {"firstname": "jane", "lastname": "smith"},
 {"firstname": "jane", "lastname": "doe"},
 {"firstname": "joe", "lastname": "smith"},
 {"firstname": "joe", "lastname": "doe"},
 {"firstname": "steve", "lastname": "smith"},
 {"firstname": "jack", "lastname": "doe"}
]

I want to get a count of duplicate firstnames

duplicates count 3

Count of non-duplicate firstnames

non-duplicates count 2

I tried to count the number of buckets, but it seems to count all buckets whether it's duplicate or non-duplicate

GET mynames/_search
{
"aggs" : {
    "name_count" : {
        "terms" : {
            "field" : "firstname.keyword",
            "min_doc_count": 2
        }
    },
"count":{
  "cardinality": {
    "field": "firstname.keyword"
  }
}
}

回答1:

Well I've made use of several aggregations here. The below are the lists which I've used. The order of the list is the execution order of the aggregation.

For Duplicates

• Terms Aggregation
• Stats Bucket Aggregation

For Non Duplicates

• Terms Aggregation
  • Bucket Selector (As a sub aggregation)
• Sum Bucket Selector

Aggregation Query:

POST <your_index_name>/_search
{  
   "size":0,
   "aggs":{  
      "duplicate_aggs":{  
         "terms":{  
            "field":"firstname.keyword",
            "min_doc_count":2
         }
      },
      "duplicate_bucketcount":{  
         "stats_bucket":{  
            "buckets_path":"duplicate_aggs._count"
         }
      },
      "nonduplicate_aggs":{  
         "terms":{  
            "field":"firstname.keyword"
         },
         "aggs":{  
            "equal_one":{  
               "bucket_selector":{  
                  "buckets_path":{  
                     "count":"_count"
                  },
                  "script":"params.count == 1"
               }
            }
         }
      },
      "nonduplicate_bucketcount":{  
         "sum_bucket":{  
            "buckets_path":"nonduplicate_aggs._count"
         }
      }
   }
}

Response

{
  "took": 10,
  "timed_out": false,
  "_shards": {
    "total": 5,
    "successful": 5,
    "skipped": 0,
    "failed": 0
  },
  "hits": {
    "total": 8,
    "max_score": 0,
    "hits": []
  },
  "aggregations": {
    "duplicate_aggs": {
      "doc_count_error_upper_bound": 0,
      "sum_other_doc_count": 0,
      "buckets": [
        {
          "key": "jane",
          "doc_count": 2
        },
        {
          "key": "joe",
          "doc_count": 2
        },
        {
          "key": "john",
          "doc_count": 2
        }
      ]
    },
    "nonduplicate_aggs": {
      "doc_count_error_upper_bound": 0,
      "sum_other_doc_count": 0,
      "buckets": [
        {
          "key": "jack",
          "doc_count": 1
        },
        {
          "key": "steve",
          "doc_count": 1
        }
      ]
    },
    "duplicate_bucketcount": {
      "count": 3,
      "min": 2,
      "max": 2,
      "avg": 2,
      "sum": 6
    },
    "nonduplicate_bucketcount": {
      "value": 2
    }
  }
}

Notice that in the above response, we have duplicate_bucketcount.count key whose value 3 is what would display the bucket count which is the number of keys which are duplicates.

Let me know if it helps!

来源：https://stackoverflow.com/questions/53359102/elasticsearch-count-duplicated-and-unique-values