How to add additional value in bucket aggregation?

Elastic Stack Elasticsearch
1

Hi, I have documents in the index with the field sentimental_keywords:

sentimental_keywords: [
{
    rate: 0.2857143,
    keyword: "manager"
},
...
{
    rate: -0.333333343,
    keyword: "project"
},
]

When I do terms aggregations :

"aggs": {
     sentimental_keywords": {
            "terms": {
                "field" : "sentimental_keywords.keyword.keyword",
                "size": 100
            }
}

I get the result:

'buckets': [
    {'key': 'manager', 'doc_count': 444},
    {'key': 'project', 'doc_count': 373}
]

I also want to add each term (key), the value of the rate, for example, get next result:

'buckets': [
    {'key': 'manager', 'doc_count': 444,  'rate': { 'value': 0.2857143}},
    {'key': 'project', 'doc_count': 373,  'rate': { 'value': -0.333333343}}
]

Do you have any suggestions?

2

I'm not sure if that would fit with what you're looking for as aggregations are meant to aggregate documents and not providing individual values.

So what about a avg or a max agg as a sub aggregation of your terms agg?

3

I've tried to use sub aggregation for a rate like max or avg , it returns the average or max rate for a document where term from top-level aggs meets.

4

I solved it by using sub aggregate terms for rate value

"aggs": {
    "sentimental_keywords": {
            "terms": {
                "field" : "sentimental_keywords.keyword.keyword",
                "size": 10
            },
            "aggs": {
                "rate": {
                    "terms": {
                        "field": "calculation.sentimental_keywords.rate",
                        "size": 1
                    }
                }
            }
            
        }
    }

In the result I get what I wanted:

{
     'key': 'project',
     'doc_count': 371,
     'rate': {
           'doc_count_error_upper_bound': 66,
           'sum_other_doc_count': 1789,
           'buckets': [{'key': -0.333333343, 'doc_count': 371}]
     }
}
1 Like
5

I'm surprised that this will work with more data.

6

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