Our elastic search document currently has following fields: Id, cityId.
It's very easy to aggregate if each document's city field has only 1 cityid i.e. cityid = 1.
But how can we aggregate on cityid if some document have multiple cityid values?
For example, if some document has cityid as (1,2), this shouldn't be treated as a new cityid i.e. (1,2) but this document should consider cityid 1 and cityid 2 as separate keys while aggregation on cityid for these documents.