Finding intersection of two buckets using Elastic - elasticsearch

Question

I have data structured as the following in an elastic index:
[ { customer_id: 1, date_of_purchase: 01-01-2022 },
{ customer_id: 2, date_of_purchase: 01-02-2022 },
{ customer_id: 1, date_of_purchase: 01-02-2022 },
....
]
I want to find the numbers of users who have bought something in both September and October, but having issues figuring out how to make a query for this. Any suggestions would rock, thanks!

Answer 1

I have used following aggregations
1. Terms aggregation
2. Bucket selector
3. Date Range
In query I have filtered all documents which either have purchase date in Jan or in Feb. This reduces number of documents for aggregation to work on. In aggregation I have done a group by(terms aggregation) on customer_id and then further grouped documents based on date ranges(1 bucket for each month). Then I have eliminated months(using bucket selector) which have zero documents i.e. with no purchase date in that month and further eliminated customers which have 1 or zero buckets
Query
{
"query": {
"bool": {
"should": [
{
"range": {
"date_of_purchase": {
"gte": "2022-01-01",
"lte": "2022-01-31"
}
}
},
{
"range": {
"date_of_purchase": {
"gte": "2022-02-01",
"lte": "2022-02-28"
}
}
}
]
}
},
"aggs": {
"cutomers": {
"terms": {
"field": "customer_id",
"size": 10
},
"aggs": {
"range": {
"date_range": {
"field": "date_of_purchase",
"ranges": [
{
"to": "2022-01-31",
"from": "2022-01-01"
},
{
"to": "2022-02-28",
"from": "2022-02-01"
}
]
},
"aggs": {
"filter_months": {
"bucket_selector": {
"buckets_path": {
"doc_count":"_count"
},
"script": "params.doc_count>=1"
}
}
}
},
"bucket_count":{
"bucket_selector": {
"buckets_path": {
"bucket_count":"range._bucket_count"
},
"script": "params.bucket_count>1"
}
}
}
}
}
}
Results
"aggregations" : {
"cutomers" : {
"doc_count_error_upper_bound" : 0,
"sum_other_doc_count" : 0,
"buckets" : [
{
"key" : 1,
"doc_count" : 2,
"range" : {
"buckets" : [
{
"key" : "2022-01-01T00:00:00.000Z-2022-01-31T00:00:00.000Z",
"from" : 1.6409952E12,
"from_as_string" : "2022-01-01T00:00:00.000Z",
"to" : 1.6435872E12,
"to_as_string" : "2022-01-31T00:00:00.000Z",
"doc_count" : 1
},
{
"key" : "2022-02-01T00:00:00.000Z-2022-02-28T00:00:00.000Z",
"from" : 1.6436736E12,
"from_as_string" : "2022-02-01T00:00:00.000Z",
"to" : 1.6460064E12,
"to_as_string" : "2022-02-28T00:00:00.000Z",
"doc_count" : 1
}
]
}
}
]
}
}