How to calculate based on the result of an aggregation? - node.js

Question

For example, I can get the result of the average counts of request to a specific url base on parent id with the code below:
client.search({
index: 'console-*',
body: {
query: {
bool: {
query_string: {
query: 'meta.http.url:"https://www.google.com"'
}
}
},
aggs: {
parent_id: {
terms: {
field: 'parent_id'
}
}
}
},
size: 0
}).then(res => {
console.log(res.hits.total/res.aggregations.total.value)
})
Now let's say I have a number of urls like:
https://www.google.com
https://www.bing.com
https://www.apple.com
And I can use:
client.search({
index: 'console-*',
body: {
sort: {
'#timestamp': {
order: 'asc'
}
},
query: {
bool: {
must: [{
range: {
'#timestamp': {
gte: 'now-5m',
lte: 'now'
}
}
}, {
query_string: { query: '_exists_:meta.http.url' }
}]
}
},
aggs: {
urls: {
terms: {
field: 'meta.http.url'
},
aggs: {
total: {
cardinality: {
field: 'parent_id'
}
}
}
}
}
},
size: 0
}).then(res => {
console.log(JSON.stringify(res))
console.log(res.aggregations.urls.buckets.map(o => {
const res = {};
res[o.key] = o.doc_count / o.total.value;
return res;
}))
})
Is it possible to get the result without doing any additional calculation in Node.js?

Answer 1

Yes, you can leverage pipeline aggregations, and more specifically, the bucket_script one.
The aggs section would look like this instead and in each bucket you'll get the result of the document count divided by the total value stored in the compute section:
aggs: {
urls: {
terms: {
field: 'meta.http.url'
},
aggs: {
total: {
cardinality: {
field: 'parent_id'
}
},
compute: {
bucket_script: {
buckets_path: {
count: "_count",
total: "total"
},
script: "params.count / params.total"
}
}
}
}
}