弹性搜索唯一ID中的唯一计数

Question

假设我的数据如下：

项目6000:的文件

{"项目ID"："6000"，"价格"：50万，"卧室"：2，"标题"："负担得起的奢侈品"，"面积"："A"}，

{"项目ID"："6000"，"价格"：70万，"卧室"：4，"标题"："宽敞"，"面积"："A"}

项目6001:文件

{"项目ID"："6001"，"价格"：550000，"卧室"：2，"标题"："令人惊叹"，"面积"："A"}

{"项目ID"："6001"，"价格"：650000，"卧室"：3，"标题"："美景"。"面积"："B"}

项目6002:文件

{"项目ID"："6002"，"价格"：550000，"卧室"：2，"标题"："令人惊叹"，"面积"："B"}

{"项目ID"："6002"，"价格"：650000，"卧室"：3，"标题"："美景"。"面积"："C"}

等等

现在，我想检索区域的计数，以便获得项目存在但只考虑唯一项目ID的最受欢迎的区域。我有这个问题：

query = {
"aggs": {
"most_popular": {
"terms": {
"field": "area",
"size": 5,
"order": {
"_count": "desc"
},
}
}
}

结果是：

bucket:[｛'key'："A"，doc_count:3｝，｛'key'："B"，doc-count:2｝…依此类推]但我想要的是每个项目的唯一计数，所以A的doc_count应该是2而不是3，即避免在同一项目中重复。

这在弹性搜索中是怎么可能的？

Answer 1

您需要使用作为父管道的bucket排序聚合对其父多存储桶的存储桶进行排序的聚合聚合

添加具有索引数据、搜索查询和搜索结果的工作示例

指数数据：

{"projectId": "6000", "price": 500000, "bedrooms": 2, "title": "Affordable luxury", "area": "A"},
{"projectId": "6000", "price": 700000, "bedrooms": 4, "title": "Spacious", "area": "A"}
{"projectId": "6001", "price": 550000, "bedrooms": 2, "title": "Stunning", "area": "A"}
{"projectId": "6001", "price": 650000, "bedrooms": 3, "title": "Excellent views". "area": "B"}
{"projectId": "6002", "price": 550000, "bedrooms": 2, "title": "Stunning", "area": "B"}
{"projectId": "6002", "price": 650000, "bedrooms": 3, "title": "Excellent views". "area": "C"}
{"projectId": "6002", "price": 500000, "bedrooms": 2, "title": "Affordable luxury", "area": "C"}
{"projectId": "6002", "price": 500000, "bedrooms": 2, "title": "Affordable luxury", "area": "C"}

搜索查询：

{
"size": 0,
"aggs": {
"most_popular": {
"terms": {
"field": "area",
"size": 5
},
"aggs": {
"unique": {
"cardinality": {
"field": "projectId"
}
},
"uniqueCount_bucket_sort": {
"bucket_sort": {
"sort": [
{
"unique": {
"order": "asc"
}
}
]
}
}
}
}
}
}

搜索结果：

"buckets": [
{
"key": "C",
"doc_count": 3,
"unique": {
"value": 1
}
},
{
"key": "B",
"doc_count": 2,
"unique": {
"value": 2
}
},
{
"key": "A",
"doc_count": 3,
"unique": {
"value": 2
}
}
]

Answer 2

您可以向查询中添加另一个aggregation，并获得projectId的基数，如下所示：

{
"aggs": {
"most_popular": {
"terms": {
"field": "area",
"size": 5,
"order": {
"_count": "desc"
}
},"aggs": {
"unique": {
"cardinality": {
"field": "projectId"
}
}
}
}
}
}