How do I use Elasticsearch aggregation using python elasticsearch_dsl library?

Elastic Stack Elasticsearch
1

I have a python model that creates the index as

class Telemetry(DocType):
    id = Keyword()
    imei = Keyword()
    location = field.GeoPoint()
    timestamp = Date()

I would like to query distinct count of imei. What I have done so far is, but this does not give any output.

def get_telemetry_for_license():
    docs = Telemetry.search().aggs.metric('imei_count', 'cardinality', field='imei')
    return imei_count

Any help on the above i appreciate.

I have also tried other way but it loads default 10,000 docs only and my docs size is few millions. If I increase default size my elasticsearch service does not response. Here is the code for other effort.

def get_telemetry_for_license():
    docs = Telemetry.search().source(exclude['location_shape']).extra(size=10000)
    docs = [doc.__dict__['_d_'] for doc in docs]
    imei =[]
    imei_count=0
    for doc in docs:
        if doc.get("imei") is not None:
            if doc['imei'] not in imei:
                imei_count += 1
                imei.append(doc['imei'])

    return imei_count

I would be grateful should you help me in any of my effort.

Thanks
Raf.

2

From reading the python docs it seems that you do not execute the request in the get_telemetry_for_license() method, so you are still dealing with the request object?

See Search DSL — Elasticsearch DSL 7.2.0 documentation

3

Let me check after newly Elasticsearch issue resolved. Now cant change mapping to use count, earlier i was able to.

curl -X PUT "localhost:9200/enterprise_telemetry_v1/_mapping/_doc?pretty" -H 'Content-Type: application/json' -d'
{
  "properties": {
    "imei": {
      "type": "keyword"
    }
  }
}
'
4

I have fixed mapping issue,
back to python dsl, now it raises error imei_count is not defined.

def get_telemetry_for_license():
    docs = Telemetry.search().aggs.metric('imei_count', 'cardinality', field='imei')
    docs = docs.execute()
    return imei_count
5

This topic was automatically closed 28 days after the last reply. New replies are no longer allowed.