Elasticsearch : How to group documents by field when field is an array - arrays

Question

My Elasticsearch index looks like that:
{
"team": ["Jane","Jason"],
"date": "2020/07/23 12:00:56",
"is_work_done": true
},
{
"team": ["Jane","Jason"],
"date": "2020/07/22 14:23:56",
"is_work_done": false
},
{
"team": ["Jane","Jason","Anna"],
"date": "2020/07/17 09:22:10",
"is_work_done": false
},
{
"team": ["Alex","George","Anna"],
"date": "2020/07/13 03:24:19",
"is_work_done": true
}
My mapping is:
{
"mappings": {
"type_name": {
"properties": {
"team": { "type": "keyword" },
"date": {"type": "date", "format": "yyyy/MM/dd HH:mm:ss"},
"is_work_done": { "type": "boolean" }
}
}
}
}
I would like to collect information on each team. How can I group my documents by team?
I created this index to solve this issue because in reality, I don't know how many members each team contain.
I tried to aggregate the documents but I don't find which type of aggregation would fit.
For example, with this query:
GET /testbench-test/_search
{
"aggs": {
"mybucket": {
"terms": { "field": "team" }
}
}
}
I get this result:
"aggregations" : {
"mybucket" : {
"doc_count_error_upper_bound" : 0,
"sum_other_doc_count" : 0,
"buckets" : [
{
"key" : "Jane",
"doc_count" : 3
},
{
"key" : "Jason",
"doc_count" : 3
},
{
"key" : "Anna",
"doc_count" : 2
},
{
"key" : "Alex",
"doc_count" : 1
},
{
"key" : "George",
"doc_count" : 1
}
]
}
}
Thanks for helping!
Edit:
With the query on my real index which contains 64,030 indexations:
POST _search
{
"aggs": {
"teams": {
"terms": {
"script": "doc['team'].join(' & ')",
"size": 10
}
}
}
}
I get this result:
{
"took" : 52,
"timed_out" : false,
"_shards" : {
"total" : 5,
"successful" : 5,
"skipped" : 0,
"failed" : 0
},
"hits" : {
"total" : 64031,
"max_score" : 1.0,
"hits" : [
{
"_index" : "test-logs",
"_type" : "log",
"_id" : "98",
"_score" : 1.0,
"_source" : {
"uuid" : "9827af",
"benchId" : "m",
"benchGroup" : "e",
"machine" : "CH",
"sha1" : "ddf380fd6a2f930813",
"date" : "2019/04/25 11:40:19",
"duration" : 0.9953742847,
"isCss" : false,
"isPassed" : true,
"finalStatus" : "OK_VNA_TEST",
"team" : [
"MCANT00013A9D00002"
]
}
},
{
"_index" : "test-logs",
"_type" : "log",
"_id" : "b0c9a8aa4",
"_score" : 1.0,
"_source" : {
"uuid" : "b0c9a1be0a8aa4",
"benchId" : "m",
"benchGroup" : "e",
"machine" : "CH",
"sha1" : "ddf385a1562d78813",
"date" : "2019/04/29 08:08:37",
"duration" : 0.4976871423,
"isCss" : false,
"isPassed" : true,
"finalStatus" : "OK_VNA_TEST",
"team" : [
"MCANT00009A9D00149"
]
}
},
{
"_index" : "test-logs",
"_type" : "log",
"_id" : "bb1d525e2a368f6d4",
"_score" : 1.0,
"_source" : {
"uuid" : "bb1da368f6d4",
"benchId" : "m",
"benchGroup" : "e",
"machine" : "CH",
"sha1" : "ddf380fd6a2f85a1562d78813",
"date" : "2019/04/29 08:09:51",
"duration" : 0.5208305083,
"isCss" : false,
"isPassed" : true,
"finalStatus" : "OK_VNA_TEST",
"team" : [
"MCANT00009A9D00179"
]
}
},
{
"_index" : "test-logs",
"_type" : "log",
"_id" : "87bec43ce2553c590b",
"_score" : 1.0,
"_source" : {
"uuid" : "87bec43c-e2553c590b",
"benchId" : "m",
"benchGroup" : "e",
"machine" : "CH",
"sha1" : "ddf380fd6a2f9302bcdf8c26e1f85a1562d78813",
"date" : "2019/04/29 08:10:10",
"duration" : 0.4629604518,
"isCss" : false,
"isPassed" : true,
"finalStatus" : "OK_VNA_TEST",
"team" : [
"MCANT00009A9D00181"
]
}
},
{
"_index" : "test-logs",
"_type" : "log",
"_id" : "3499224bdac4fa39",
"_score" : 1.0,
"_source" : {
"uuid" : "349922444bdac4fa39",
"benchId" : "m",
"benchGroup" : "e",
"machine" : "CH",
"sha1" : "ddf380fd6e1f85a1562d78813",
"date" : "2019/04/29 08:10:49",
"duration" : 0.5092588253,
"isCss" : false,
"isPassed" : true,
"finalStatus" : "OK_VNA_TEST",
"uut" : [
"MCANT00009A9D00171"
]
}
},
{
"_index" : "test-logs",
"_type" : "log",
"_id" : "f8236de0-dd6a97b7a81",
"_score" : 1.0,
"_source" : {
"uuid" : "f8236de0-add6a97b7a81",
"benchId" : "m",
"benchGroup" : "e",
"machine" : "CH",
"sha1" : "ddf380fd6a2f93085a1562d78813",
"date" : "2019/04/29 09:51:47",
"duration" : 0.6134272553,
"isCss" : false,
"isPassed" : true,
"finalStatus" : "OK_VNA_TEST",
"team" : [
"MCANT00009A9D00221"
]
}
},
{
"_index" : "test-logs",
"_type" : "log",
"_id" : "339fa5a2e-cf9f1f4738bf",
"_score" : 1.0,
"_source" : {
"uuid" : "339fa5a9f1f4738bf",
"benchId" : "m",
"benchGroup" : "e",
"machine" : "CH",
"sha1" : "ddf380fd6a2f932d78813",
"date" : "2019/04/29 09:51:57",
"duration" : 0.6249989383,
"isCss" : false,
"isPassed" : true,
"finalStatus" : "OK_VNA_TEST",
"team" : [
"MCANT00009A9D00185"
]
}
},
{
"_index" : "test-logs",
"_type" : "log",
"_id" : "dbd45ec4-f53b-4bda-9eeb-dadf2e3ab366",
"_score" : 1.0,
"_source" : {
"uuid" : "dbd45ec4-f53b-4bda-9eeb-dadf2e3ab366",
"benchId" : "m",
"benchGroup" : "e",
"machine" : "CH",
"sha1" : "ddf380fd6a2f9302bcdfd78813",
"date" : "2019/04/29 09:52:19",
"duration" : 0.5787005648,
"isCss" : false,
"isPassed" : true,
"finalStatus" : "OK_VNA_TEST",
"team" : [
"MCANT00009A9D00184"
]
}
},
{
"_index" : "test-logs",
"_type" : "log",
"_id" : "bc0d3548bed68",
"_score" : 1.0,
"_source" : {
"uuid" : "bc0d354348bed68",
"benchId" : "m",
"benchGroup" : "e",
"machine" : "CH",
"sha1" : "ddf3801562d78813",
"date" : "2019/04/29 08:08:12",
"duration" : 0.5208305083,
"isCss" : false,
"isPassed" : true,
"finalStatus" : "OK_VNA_TEST",
"team" : [
"MCANT00009A9D00160"
]
}
},
{
"_index" : "test-logs",
"_type" : "log",
"_id" : "d591b9e7-683a-4d78-be31-1b137b8a3b2b",
"_score" : 1.0,
"_source" : {
"uuid" : "d591b9e7-683a-4d78-be31-1b137b8a3b2b",
"benchId" : "m",
"benchGroup" : "e",
"machine" : "CH",
"sha1" : "ddf380fd6a2f93013",
"date" : "2019/04/29 08:08:05",
"duration" : 0.6828689948,
"isCss" : false,
"isPassed" : true,
"finalStatus" : "OK_VNA_TEST",
"team" : [
"MCANT00009A9D00146"
]
}
}
]
},
"aggregations" : {
"teams" : {
"doc_count_error_upper_bound" : 25,
"sum_other_doc_count" : 63658,
"buckets" : [
{
"key" : "SI050010AA8C00001",
"doc_count" : 86
},
{
"key" : "00845D9E0137 & 00851F9E0095 & CPPCB00608C9F00060 & MCSAS00676A9F00141_-27.8_B & MCSAS00677A9F00146_23.0_B & SI050012AG9F00060",
"doc_count" : 56
},
{
"key" : "00845D9E0275 & 00851F9G0056 & CPPCB00608C9F00127 & MCSAS00676B9G00012 & MCSAS00676B9G00012_-21.2_C & MCSAS00677B9G00005 & MCSAS00677B9G00005_32.5_C & SI050012AG9F00127",
"doc_count" : 43
},
{
"key" : "00844G9D0041 & 00847D9G0020 & MCANT00009A9G00048 & MCSAS00652F9E00091 & S20-17272 & SI050001AG9G00055 & SI050004AA9F00059",
"doc_count" : 39
},
{
"key" : "00845D9E0035 & 00851F9E0062 & CPPCB00608C9E00034 & MCSAS00676A9E00090_-35.0_B & MCSAS00677A9E00089_31.0_B & SI050012AG9E00034",
"doc_count" : 34
},
{
"key" : "IX & IX-c2-67063 & IX-x2-00511 & SI050010AA9A00002 & droneProduction",
"doc_count" : 27
},
{
"key" : "IX & IX-12-10251 & IX-x2-00484 & SI050001AF9A00020 & SI050010AA8J00154 & SI050012AG9D00082 & droneClient",
"doc_count" : 25
},
{
"key" : "MCANT00009A9G00048 & MCSAS00652F9E00091 & S20-17272 & SI050001AG9G00055",
"doc_count" : 24
},
{
"key" : "00883C0F0000",
"doc_count" : 20
},
{
"key" : "00844C8B0029 & 00847C8E0018 & 00849A8B0015 & MCANT00009A8E00017 & N/A & S20-00533 & SI050002AA8E000001 & SI050004AA8E000514",
"doc_count" : 19
}
]
}
}
}
Why not all the different 'teams' have their own bucket?

Answer 1

What you can do is to use a script in your terms aggregation like this:
POST teams/_search
{
"size": 0,
"aggs": {
"teams": {
"terms": {
"script": "doc['team'].join('-')",
"size": 10
}
}
}
}
And the result you're going to get is something like this:
"buckets" : [
{
"key" : "Jane-Jason",
"doc_count" : 2
},
{
"key" : "Alex-Anna-George",
"doc_count" : 1
},
{
"key" : "Anna-Jane-Jason",
"doc_count" : 1
}
]