How to do multi value count aggregation in elasticsearch

Question

Is there a way to do a "multi value count" aggregation? It would work the same way as the regular value count: http://www.elasticsearch.org/guide/en/elasticsearch/reference/current/search-aggregations-metrics-valuecount-aggregation.html

But would be used to count multiple fields.

For example we have a mapping:

"properties" : {
        "lastName" : { "type" : "string" },
        "firstName" : { "type" : "string" },
        "age" : { "type" : "string" }
}

And the following 4 docs:

{"lastName":"Smith"}
{"lastName":"Jacobson", "firstName":"Scott"}
{"lastName":"Bush", "age": 68}
{"lastName":"Obama", "firstName":"Barack", "age":53}

I would like to count the number of documents with either an age OR a first name. In my case the aggregation would return 3.

Any idea ? I try a little bit using the script function:

"aggregation_name" : {
  "value_count" : {
    "script" : "doc['age'].value + doc['firstName'].value"
  }
}

but it did't seem to work.

Thanks!

Vineeth Mohan · Accepted Answer · 2015-02-05 14:34:14Z

1

You can achieve this using filter aggregation and Exists filter -

{
  "aggs": {
    "CountSpec": {
      "filter": {
        "bool": {
          "should": [
            {
              "exists": {
                "field": "age"
              }
            },
            {
              "exists": {
                "field": "firstName"
              }
            }
          ]
        }
      }
    }
  }
}

answered Feb 5, 2015 at 14:34

Vineeth Mohan

19.4k9 gold badges70 silver badges81 bronze badges

Sign up to request clarification or add additional context in comments.

2 Comments

Maxime Over a year ago

I managed to achieve what I wanted with:

"count_name" : {           "value_count" : {             "script" : "if(doc.containsKey('age') || doc.containsKey('firstName')) { 1 }"           }         }

Any idea what would be best : the filter aggregation or the value aggregation with a script ?

Vineeth Mohan Over a year ago

Certainly filter approach. Scripts are CPU intensive and filters are natively build. More importantly scripts have filter cache which means that the execution would be much faster once its cached

Collectives™ on Stack Overflow

How to do multi value count aggregation in elasticsearch

1 Answer 1

2 Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

2 Comments

Your Answer

Sign up or log in

Post as a guest

Related