Searchkick Aggregations behavior - ruby-on-rails

Question

I am working with the Searchkick Gem and Elastic search and am trying to understand the aggregations behavior.
I have three facets (Aggregations): City, State and Company.
If I filter by any one of them, the counts of other two are reduced to reflect the total in the result set. But the selected facet comes back with all values. So say I had 100 items in the index, and I filtered by a Company that had 2 total items in the index, the City and State counts are updated to reflect no more than 2. But the Company count remains at 100.
Example (filtered to City=Atlanta)
{
"query": {
"function_score": {
"functions": [
{
"filter": {
"and": [
{
"term": {
"featured": true
}
}
]
},
"boost_factor": 1000
}
],
"query": {
"match_all": {}
},
"score_mode": "sum"
}
},
"size": 20,
"from": 0,
"post_filter": {
"bool": {
"filter": [
{
"range": {
"expiration_date": {
"from": "2016-08-18T23:07:15.670-04:00",
"include_lower": true
}
}
},
{
"range": {
"created_at": {
"to": "2016-08-18T23:07:15.670-04:00",
"include_upper": true
}
}
},
{
"term": {
"published": true
}
},
{
"term": {
"tenant_id": 4
}
},
{
"term": {
"city": "Atlanta"
}
}
]
}
},
"aggs": {
"company": {
"filter": {
"bool": {
"must": [
{
"range": {
"expiration_date": {
"from": "2016-08-18T23:07:15.670-04:00",
"include_lower": true
}
}
},
{
"range": {
"created_at": {
"to": "2016-08-18T23:07:15.670-04:00",
"include_upper": true
}
}
},
{
"term": {
"published": true
}
},
{
"term": {
"tenant_id": 4
}
},
{
"term": {
"city": "Atlanta"
}
}
]
}
},
"aggs": {
"company": {
"terms": {
"field": "company",
"size": 10
}
}
}
},
"city": {
"filter": {
"bool": {
"must": [
{
"range": {
"expiration_date": {
"from": "2016-08-18T23:07:15.670-04:00",
"include_lower": true
}
}
},
{
"range": {
"created_at": {
"to": "2016-08-18T23:07:15.670-04:00",
"include_upper": true
}
}
},
{
"term": {
"published": true
}
},
{
"term": {
"tenant_id": 4
}
}
]
}
},
"aggs": {
"city": {
"terms": {
"field": "city",
"size": 10
}
}
}
},
"state": {
"filter": {
"bool": {
"must": [
{
"range": {
"expiration_date": {
"from": "2016-08-18T23:07:15.670-04:00",
"include_lower": true
}
}
},
{
"range": {
"created_at": {
"to": "2016-08-18T23:07:15.670-04:00",
"include_upper": true
}
}
},
{
"term": {
"published": true
}
},
{
"term": {
"tenant_id": 4
}
},
{
"term": {
"city": "Atlanta"
}
}
]
}
},
"aggs": {
"state": {
"terms": {
"field": "state",
"size": 10
}
}
}
}
},
"fields": []
}
Result (2 result returned, but 58 City Aggregations come back). Note Company and City return correct # of Aggregations:
{
"took": 114,
"timed_out": false,
"_shards": {
"total": 5,
"successful": 5,
"failed": 0
},
"hits": {
"total": 2,
"max_score": 1,
"hits": [
{
"_index": "jobs_development_20160818140128648",
"_type": "job",
"_id": "457134",
"_score": 1
},
{
"_index": "jobs_development_20160818140128648",
"_type": "job",
"_id": "457137",
"_score": 1
}
]
},
"aggregations": {
"city": {
"doc_count": 58,
"city": {
"doc_count_error_upper_bound": 0,
"sum_other_doc_count": 19,
"buckets": [
{
"key": "Los Angeles",
"doc_count": 8
},
{
"key": "London",
"doc_count": 7
},
{
"key": "New York",
"doc_count": 7
},
{
"key": "Burbank",
"doc_count": 5
},
{
"key": "Pasig",
"doc_count": 3
},
{
"key": "Atlanta",
"doc_count": 2
},
{
"key": "Chicago",
"doc_count": 2
},
{
"key": "Culver City",
"doc_count": 2
},
{
"key": "London Borough of Hackney",
"doc_count": 2
},
{
"key": "Birmingham",
"doc_count": 1
}
]
}
},
"company": {
"doc_count": 2,
"company": {
"doc_count_error_upper_bound": 0,
"sum_other_doc_count": 0,
"buckets": [
{
"key": "Second Story",
"doc_count": 2
}
]
}
},
"state": {
"doc_count": 2,
"state": {
"doc_count_error_upper_bound": 0,
"sum_other_doc_count": 0,
"buckets": [
{
"key": "Georgia",
"doc_count": 2
}
]
}
}
}
}
What am I missing? Is this correct behavior?