_doc_count 字段
编辑_doc_count 字段
编辑桶聚合总是返回一个名为 doc_count 的字段，显示每个桶中聚合和分区的文档数量。doc_count 值的计算非常简单。对于每个桶中收集的每个文档，doc_count 都会增加 1。
虽然这种简单的方法在计算单个文档的聚合时很有效，但在准确表示存储预聚合数据（例如 histogram 或 aggregate_metric_double 字段）的文档时会失败，因为一个摘要字段可能代表多个文档。
为了在处理预聚合数据时能够正确计算文档数量，我们引入了一种名为 _doc_count 的元数据字段类型。_doc_count 必须始终是一个正整数，表示在单个摘要字段中聚合的文档数量。
当字段 _doc_count 被添加到文档时，所有桶聚合都将遵守其值，并将桶的 doc_count 增加该字段的值。如果一个文档不包含任何 _doc_count 字段，则默认情况下会隐式使用 _doc_count = 1。
_doc_count字段每个文档只能存储一个正整数。不允许使用嵌套数组。- 如果文档不包含任何
_doc_count字段，聚合器将增加 1，这是默认行为。
示例
编辑以下 创建索引 API 请求创建一个新索引，其中包含以下字段映射
-
my_histogram，一个用于存储百分比数据的histogram字段 -
my_text，一个用于存储直方图标题的keyword字段
resp = client.indices.create(
index="my_index",
mappings={
"properties": {
"my_histogram": {
"type": "histogram"
},
"my_text": {
"type": "keyword"
}
}
},
)
print(resp)
response = client.indices.create(
index: 'my_index',
body: {
mappings: {
properties: {
my_histogram: {
type: 'histogram'
},
my_text: {
type: 'keyword'
}
}
}
}
)
puts response
const response = await client.indices.create({
index: "my_index",
mappings: {
properties: {
my_histogram: {
type: "histogram",
},
my_text: {
type: "keyword",
},
},
},
});
console.log(response);
PUT my_index
{
"mappings" : {
"properties" : {
"my_histogram" : {
"type" : "histogram"
},
"my_text" : {
"type" : "keyword"
}
}
}
}
以下 索引 API 请求存储两个直方图的预聚合数据：histogram_1 和 histogram_2。
resp = client.index(
index="my_index",
id="1",
document={
"my_text": "histogram_1",
"my_histogram": {
"values": [
0.1,
0.2,
0.3,
0.4,
0.5
],
"counts": [
3,
7,
23,
12,
6
]
},
"_doc_count": 45
},
)
print(resp)
resp1 = client.index(
index="my_index",
id="2",
document={
"my_text": "histogram_2",
"my_histogram": {
"values": [
0.1,
0.25,
0.35,
0.4,
0.45,
0.5
],
"counts": [
8,
17,
8,
7,
6,
2
]
},
"_doc_count": 62
},
)
print(resp1)
response = client.index(
index: 'my_index',
id: 1,
body: {
my_text: 'histogram_1',
my_histogram: {
values: [
0.1,
0.2,
0.3,
0.4,
0.5
],
counts: [
3,
7,
23,
12,
6
]
},
_doc_count: 45
}
)
puts response
response = client.index(
index: 'my_index',
id: 2,
body: {
my_text: 'histogram_2',
my_histogram: {
values: [
0.1,
0.25,
0.35,
0.4,
0.45,
0.5
],
counts: [
8,
17,
8,
7,
6,
2
]
},
_doc_count: 62
}
)
puts response
const response = await client.index({
index: "my_index",
id: 1,
document: {
my_text: "histogram_1",
my_histogram: {
values: [0.1, 0.2, 0.3, 0.4, 0.5],
counts: [3, 7, 23, 12, 6],
},
_doc_count: 45,
},
});
console.log(response);
const response1 = await client.index({
index: "my_index",
id: 2,
document: {
my_text: "histogram_2",
my_histogram: {
values: [0.1, 0.25, 0.35, 0.4, 0.45, 0.5],
counts: [8, 17, 8, 7, 6, 2],
},
_doc_count: 62,
},
});
console.log(response1);
PUT my_index/_doc/1
{
"my_text" : "histogram_1",
"my_histogram" : {
"values" : [0.1, 0.2, 0.3, 0.4, 0.5],
"counts" : [3, 7, 23, 12, 6]
},
"_doc_count": 45
}
PUT my_index/_doc/2
{
"my_text" : "histogram_2",
"my_histogram" : {
"values" : [0.1, 0.25, 0.35, 0.4, 0.45, 0.5],
"counts" : [8, 17, 8, 7, 6, 2]
},
"_doc_count": 62
}
如果我们在 my_index 上运行以下 词项聚合
resp = client.search(
aggs={
"histogram_titles": {
"terms": {
"field": "my_text"
}
}
},
)
print(resp)
response = client.search(
body: {
aggregations: {
histogram_titles: {
terms: {
field: 'my_text'
}
}
}
}
)
puts response
const response = await client.search({
aggs: {
histogram_titles: {
terms: {
field: "my_text",
},
},
},
});
console.log(response);
GET /_search
{
"aggs" : {
"histogram_titles" : {
"terms" : { "field" : "my_text" }
}
}
}
我们将得到以下响应
{
...
"aggregations" : {
"histogram_titles" : {
"doc_count_error_upper_bound": 0,
"sum_other_doc_count": 0,
"buckets" : [
{
"key" : "histogram_2",
"doc_count" : 62
},
{
"key" : "histogram_1",
"doc_count" : 45
}
]
}
}
}