With Elasticsearch, you can use the numeric properties of your data natively. You’ve seen an example of this above, where we searched for a range of dates and ratings. Elasticsearch also handles location in several formats, including geo points, geo polygons, and geo hashes. The sample movie data includes locations, drawn from 100 American cities, and assigned at random. You can run a bounding-box query by executing the following:
GET movies/_search
{
"query": {
"bool": {
"filter": {
"geo_bounding_box": {
"location": {
"top_left": {
"lat": 50.96,
"lon": -124.6
},
"bottom_right": {
"lat": 46.96,
"lon": -120.0
}
}
}
}
}
},
"sort": [
{
"_geo_distance": {
"location": {
"lat": 47.59,
"lon": -122.43
},
"order": "asc",
"unit": "km"
}
}
]
}This query searches a bounding box around Seattle and sorts the results by their distance from the center of Seattle. Even though the results themselves are not particularly meaningful (the locations are randomly assigned), the same method will let you search and sort your own data to display results that are geographically local for your customers.