发布于 2016-02-29 14:37:41 | 348 次阅读 | 评论: 0 | 来源: 网络整理
按距离聚合对于类似“找出距我1公里内的所有pizza店”这样的检索场景很适合。 检索结果需要确实地只返回距离用户1km内的文档，不过我们可以再加上一个“1-2km内的结果集”：
GET /attractions/restaurant/_search
{
"query": {
"filtered": {
"query": {
"match": { <1>
"name": "pizza"
}
},
"filter": {
"geo_bounding_box": {
"location": { <2>
"top_left": {
"lat": 40,8,
"lon": -74.1
},
"bottom_right": {
"lat": 40.4,
"lon": -73.7
}
}
}
}
}
},
"aggs": {
"per_ring": {
"geo_distance": { <3>
"field": "location",
"unit": "km",
"origin": {
"lat": 40.712,
"lon": -73.988
},
"ranges": [
{ "from": 0, "to": 1 },
{ "from": 1, "to": 2 }
]
}
}
},
"post_filter": { <4>
"geo_distance": {
"distance": "1km",
"location": {
"lat": 40.712,
"lon": -73.988
}
}
}
}上例请求的返回结果如下：
"hits": {
"total": 1,
"max_score": 0.15342641,
"hits": [ <1>
{
"_index": "attractions",
"_type": "restaurant",
"_id": "3",
"_score": 0.15342641,
"_source": {
"name": "Mini Munchies Pizza",
"location": [
-73.983,
40.719
]
}
}
]
},
"aggregations": {
"per_ring": { <2>
"buckets": [
{
"key": "*-1.0",
"from": 0,
"to": 1,
"doc_count": 1
},
{
"key": "1.0-2.0",
"from": 1,
"to": 2,
"doc_count": 1
}
]
}
}这个例子中，我们统计了落到各个环形区域中的饭店数。 当然，我们也可以使用子聚合器再在每个环形区域中进一步计算它们的平均价格，最流行，等等。