I use Elasticsearch for my Search Service and I use this JSON request for sum_bucket aggregation:
"from": 0,
    "size": 0,
    "aggregations": {
        "aggr": {
            "type": "term",
            "field": "type.keyword",
            "size": 10,
            "order": {
                "myAggr2.value": "desc"
            },
            "aggregations": {
                "myAggr2": {
                    "type": "sum",
                    "field": "length"
                }
            }
        },
        "total_myAggr2": {
            "type": "sum_bucket",
            "buckets_path": "aggr>myAggr2"
        }
    },
    "types": [
        "cable"
    ]
}
In the response I get this:
{
    "totalHits": 4,
    "tookInMillis": 1,
    "hits": [],
    "aggregations": {
        "total_myAggr2": {
            "value": "3.7600000000000002"
        },
        "aggr": {
            "sumOtherDocCount": 0,
            "buckets": [
                {
                    "key": "cable",
                    "count": 4,
                    "aggregations": {
                        "myAggr2": {
                            "value": "3.7600000000000002"
                        }
                    }
                }
            ]
        }
    }
}
Is any possibility to format the output from "3.7600000000000002" to "3.76" ? I found attribute format in ES docu but not for number formatting...
 Why is the result both aggregations in float format? Because I have created the index for field length with scaled_float=100 so I would expected the result as 