# 导数聚合(Derivative Aggregation)

> 警告:此功能是实验性的，可能会在将来的版本中完全更改或删除。Elastic将采取最大的努力来解决此问题，但实验功能不受SLA官方功能的支持。

导数管道聚合，其计算父直方图（或日期 - 图形）聚合中指定度量的导数。指定的度量必须是数字，并且必须设置直方图min\_doc\_count为0（默认为直方图聚合）。

## 语法 <a href="#derivativeaggregation-dao-shu-ju-he-yu-fa" id="derivativeaggregation-dao-shu-ju-he-yu-fa"></a>

`derivative（导数）`聚合结构如下：

```
"derivative": {
  "buckets_path": "the_sum"
}
```

警告

此功能是实验性的，可能会在将来的版本中完全更改或删除。Elastic将采取最大的努力来解决此问题，但实验功能不受SLA官方功能的支持。

导数管道聚合，其计算父直方图（或日期 - 图形）聚合中指定度量的导数。指定的度量必须是数字，并且必须设置直方图min\_doc\_count为0（默认为直方图聚合）。

## 语法 <a href="#derivativeaggregation-dao-shu-ju-he-yu-fa" id="derivativeaggregation-dao-shu-ju-he-yu-fa"></a>

`derivative（导数）`聚合结构如下：

| `"derivative": {"buckets_path":"the_sum"}` |
| ------------------------------------------ |

`derivative（导数）`参数如下：

| 参数名称          | 描述                                                                                                                                                                                                                                                                                          | 是否必填 | 默认值  |
| ------------- | ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | ---- | ---- |
| buckets\_path | 想要计算导数值的桶路径，点击[the section called “`buckets_path` Syntax](https://www.elastic.co/guide/en/elasticsearch/reference/5.4/search-aggregations-pipeline.html#buckets-path-syntax)[edit](https://github.com/elastic/elasticsearch/edit/5.4/docs/reference/aggregations/pipeline.asciidoc)”查看更多细节  | 必填   |      |
| gap\_policy   | 当数据缺口出现时采用的策略，点击[the section called “Dealing with gaps in the data](https://www.elastic.co/guide/en/elasticsearch/reference/5.4/search-aggregations-pipeline.html#gap-policy)[edit](https://github.com/elastic/elasticsearch/edit/5.4/docs/reference/aggregations/pipeline.asciidoc)”查看更多细节 | 可选   | skip |
| format        | 用于规范聚合输出值的格式                                                                                                                                                                                                                                                                                | 可选   | null |

## 一级导数 <a href="#derivativeaggregation-dao-shu-ju-he-yi-ji-dao-shu" id="derivativeaggregation-dao-shu-ju-he-yi-ji-dao-shu"></a>

以下代码段计算每月总销售额的导数：

```
POST /sales/_search
{
    "size": 0,
    "aggs" : {
        "sales_per_month" : {
            "date_histogram" : {
                "field" : "date",
                "interval" : "month"
            },
            "aggs": {
                "sales": {
                    "sum": {
                        "field": "price"
                    }
                },
                "sales_deriv": {
                    "derivative": {
                        "buckets_path": "sales" #1
                    }
                }
            }
        }
    }
}
```

| 1 | buckets\_path指示这个derivative聚合是想要得到sales\_per\_month日期直方图聚合中sales聚合值的导数。 |
| - | ----------------------------------------------------------------------- |

响应可能如下所示：

```
{
   "took": 11,
   "timed_out": false,
   "_shards": ...,
   "hits": ...,
   "aggregations": {
      "sales_per_month": {
         "buckets": [
            {
               "key_as_string": "2015/01/01 00:00:00",
               "key": 1420070400000,
               "doc_count": 3,
               "sales": {
                  "value": 550.0
               } #1
            },
            {
               "key_as_string": "2015/02/01 00:00:00",
               "key": 1422748800000,
               "doc_count": 2,
               "sales": {
                  "value": 60.0
               },
               "sales_deriv": {
                  "value": -490.0 #2
               }
            },
            {
               "key_as_string": "2015/03/01 00:00:00",
               "key": 1425168000000,
               "doc_count": 2, #3
               "sales": {
                  "value": 375.0
               },
               "sales_deriv": {
                  "value": 315.0
               }
            }
         ]
      }
   }
}
```

| 1 | 由于我们至少需要2个数据点来计算导数，因此第一个桶没有值                              |
| - | --------------------------------------------------------- |
| 2 | 导数的单位默认和sales聚合以及父直方图相同。所以在这种情况下，如果价格字段的单位是美元，导数的单位就是美元/月 |
| 3 | doc\_count表示桶中的文档数                                        |

## 二级导数

可以把导数管道聚合链接到另一个管道聚合的结果，计算二级导数。如以下示例所示，它将计算总月销售额的第一和第二阶导数：

```
POST /sales/_search
{
    "size": 0,
    "aggs" : {
        "sales_per_month" : {
            "date_histogram" : {
                "field" : "date",
                "interval" : "month"
            },
            "aggs": {
                "sales": {
                    "sum": {
                        "field": "price"
                    }
                },
                "sales_deriv": {
                    "derivative": {
                        "buckets_path": "sales"
                    }
                },
                "sales_2nd_deriv": {
                    "derivative": {
                        "buckets_path": "sales_deriv" #1
                    }
                }
            }
        }
    }
}
```

| 1 | 二阶导数的buckets\_path指向一阶导数的名称 |
| - | --------------------------- |

响应可能如下所示：

```
{
   "took": 50,
   "timed_out": false,
   "_shards": ...,
   "hits": ...,
   "aggregations": {
      "sales_per_month": {
         "buckets": [
            {
               "key_as_string": "2015/01/01 00:00:00",
               "key": 1420070400000,
               "doc_count": 3,
               "sales": {
                  "value": 550.0
               } #1
            },
            {
               "key_as_string": "2015/02/01 00:00:00",
               "key": 1422748800000,
               "doc_count": 2,
               "sales": {
                  "value": 60.0
               },
               "sales_deriv": {
                  "value": -490.0
               } #2
            },
            {
               "key_as_string": "2015/03/01 00:00:00",
               "key": 1425168000000,
               "doc_count": 2,
               "sales": {
                  "value": 375.0
               },
               "sales_deriv": {
                  "value": 315.0
               },
               "sales_2nd_deriv": {
                  "value": 805.0
               }
            }
         ]
      }
   }
}
```

| 1 | 由于我们至少需要2个数据点，所以前两个桶没有二级导数 |
| - | -------------------------- |
| 2 | 一级导数计算二级导数                 |

## Units（单位） <a href="#derivativeaggregation-dao-shu-ju-he-units-dan-wei" id="derivativeaggregation-dao-shu-ju-he-units-dan-wei"></a>

导数聚合允许指定导数值的单位。这将在响应normalized\_value中返回一个额外的字段，汇报在X轴单位下的导数。在下面的例子中，我们计算出每月销售额的导数，但要求销售的导数按天计算：

```
POST /sales/_search
{
    "size": 0,
    "aggs" : {
        "sales_per_month" : {
            "date_histogram" : {
                "field" : "date",
                "interval" : "month"
            },
            "aggs": {
                "sales": {
                    "sum": {
                        "field": "price"
                    }
                },
                "sales_deriv": {
                    "derivative": {
                        "buckets_path": "sales",
                        "unit": "day" #1
                    }
                }
            }
        }
    }
}
```

| 1 | unit指定用于导数计算的X轴的单位 |
| - | ------------------ |

响应可能如下所示：

```
{
   "took": 50,
   "timed_out": false,
   "_shards": ...,
   "hits": ...,
   "aggregations": {
      "sales_per_month": {
         "buckets": [
            {
               "key_as_string": "2015/01/01 00:00:00",
               "key": 1420070400000,
               "doc_count": 3,
               "sales": {
                  "value": 550.0
               } #1
            },
            {
               "key_as_string": "2015/02/01 00:00:00",
               "key": 1422748800000,
               "doc_count": 2,
               "sales": {
                  "value": 60.0
               },
               "sales_deriv": {
                  "value": -490.0, #2
                  "normalized_value": -15.806451612903226 #3
               }
            },
            {
               "key_as_string": "2015/03/01 00:00:00",
               "key": 1425168000000,
               "doc_count": 2,
               "sales": {
                  "value": 375.0
               },
               "sales_deriv": {
                  "value": 315.0,
                  "normalized_value": 11.25
               }
            }
         ]
      }
   }
}
```

| 1，2 | value值是原单位：按月                |
| --- | ---------------------------- |
| 3   | `normalized_value值是请求的单位：按天` |
