Range aggregation using elasticsearch-dsl-py

Question

POST /_search
{
"size": 0,
"aggs": {
    "by_grp" : {
        "terms": {
            "field": "grpId",
            "size": 0
        },
        "aggs": {
            "twitter_count": {
                "range": {
                    "field": "twitter.followers",
                    "ranges": [
                        { "to" : 501},
                        { "from" : 501, "to" : 1001},
                        { "from" : 1001, "to" : 5001},
                        { "from" : 5001}
                    ]
                },
                "aggs" : {
                    "email_addy": {
                        "terms" : {
                            "field": "email.value",
                            "size": 0
                        }
                    }
                }
            }
        }
    }
}
}

Using elastic-search-dsl, my python code is

from datetime import datetime
from elasticsearch_dsl import DocType, String, Date, Integer, Search, Q
from elasticsearch_dsl.connections import connections
from elasticsearch import Elasticsearch

client = connections.create_connection(hosts=['http://somehost:9200'])
s = Search(using=client, index="dexy", doc_type="grp")

s.aggs.bucket('by_grp', 'terms', field='grpId', size=0) \
.bucket('twitter_count', 'range', field='twitter.followers')

Where do I go form here to define the ranges like in the CURL code? The GIT and docs are currently sparse on this topic. No examples found.

Answer 1

This looks strange, since it seems not to full follow indenting/line continuation rules, but this works.

s.aggs.bucket('by_grp', 'terms', field='grpId', size=0) \
.bucket('twitter_count', 'range', field='twitter.followers',
    ranges=[
        {'to': 5001},
        {'from': 5001, 'to': 10001},
        {'from': 10001, 'to': 50001},
        {'from': 50001}
    ]
) \
.bucket('email_addy', 'terms', field='email.value', size=0)

As a note, the size = 0, just means the query should return all results for that item, as opposed to the default 10. Therefore, it will have a results for all grpId not just 10, and include all email that fall into the range bucket.