elasticsearch - Average of top n terms -

in index documents have keyword, rank , timestamp. rank keyword may differ time time. means dataset may this:

{"keywords": "piano", "rank" 1, "timestamp": 1437642812} {"keywords": "piano", "rank" 2, "timestamp": 1437642813} {"keywords": "electric guitar", "rank" 5, "timestamp": 1437644326} 

i average rank of top 500 occuring keywords. cannot find out how this.

my current try-outs seem give average results individually, want average entire dataset of top results of aggregation.

post _search {     "aggs": {         "top_keywords": {             "terms": {                 "field": "keywords",                 "size": 1             }         },         "avg_rank": {             "avg": {"field": "rank"}         }     },     "size": 0 } 

attempts using top_hits haven't been successful either.

elsewhere have read splitting query separate queries, first retrieving list of top keywords , in second query filter documents keywords returned first query. feed query kibana, hope not required.

these related topics don't provide satisfying answer either.

• how run terms aggregation scoped first `n` results
• aggregation on top n results

can point me in right direction?

an elasticssearch developer told me not possible:

in current version not possible, pipeline aggregations coming in version 2.0 able use avg_bucket aggregation this: https://www.elastic.co/guide/en/elasticsearch/reference/master/search-aggregations-pipeline-avg-bucket-aggregation.html

in mean time need aggregation top 500 terms , perform average calculation on client side

and makes impossible show data in kibana:

yes work in 2.0 requests straight elasticsearch. however, take time functionality added kibana interface. kibana team thinking how add though

source: https://discuss.elastic.co/t/average-of-top-n-terms/26165