How to take more fields when grouping - elasticsearch

Question

Trying to group data and take all of its fields by the way.
GET /testnews/default/_search
{
"size": 10,
"from":50,
"query":{
"multi_match": {
"query": "serenay",
"fields": ["Data.Title", "Data.Description", "Data.Tags.Title", "Data.MentionTitle", "Data.Program.title", "Data.Program.description", "Data.Program.original_title"]
}
},
"sort":[{
"Data.CreatedAt": {
"order": "desc"
},
"Data.ViewCount": {
"order": "desc"
}
}],
"aggs": {
"group_by_state": {
"terms": {
"field": "Data.Program.title.keyword"
}
}
}
}
But when I did it, it returns only "Program Title" in the grouped result.
Just like:
{
"key": "Kocamın Ailesi",
"doc_count": 3
}
But I just want it like:
{
"key": "Kocamın Ailesi",
"description": "blabla",
"image": "blabla.jpg",
"date": "YYYY-mm-dd",
"doc_count": 3
}
just like sql
select * from x group by field

Answer 1

Regarding the SQL example, to get the behaviour of
select a, b, count(*) from x group by a, b
you can aggregate on a, then b like this:
"aggs": {
"group_by_a": {
"terms": {
"field": "a"
},
"aggs": {
"group_by_b": {
"terms": {
"field":"b"
}
}
}
}
}
But I don't think that is what you're looking for?
If you want the full documents in aggregations you can use the "top_hits" aggregation to select the top n hits within each aggregation:
{
"aggs": {
"group_by_state": {
"terms": {
"field": "Data.Program.title.keyword"
},
"aggs": {
"state_top_hits": {
"top_hits": {
"sort": [
{ "Data.CreatedAt": { "order": "desc" } },
{ "Data.ViewCount": { "order": "desc" } }
],
"_source": {
"includes": [ "key", "description", "image", "date" ]
},
"size": 10 //Will show top 10 hits within keyword agg ordered according to the sort
}
}
}
}
}
}