Pipeline aggregations

edit

Pipeline aggregations work on the outputs produced from other aggregations rather than from document sets, adding information to the output tree. There are many different types of pipeline aggregation, each computing different information from other aggregations, but these types can be broken down into two families:

Parent
A family of pipeline aggregations that is provided with the output of its parent aggregation and is able to compute new buckets or new aggregations to add to existing buckets.
Sibling
Pipeline aggregations that are provided with the output of a sibling aggregation and are able to compute a new aggregation which will be at the same level as the sibling aggregation.

Pipeline aggregations can reference the aggregations they need to perform their computation by using the buckets_path parameter to indicate the paths to the required metrics. The syntax for defining these paths can be found in the buckets_path Syntax section below.

Pipeline aggregations cannot have sub-aggregations but depending on the type it can reference another pipeline in the buckets_path allowing pipeline aggregations to be chained. For example, you can chain together two derivatives to calculate the second derivative (i.e. a derivative of a derivative).

Because pipeline aggregations only add to the output, when chaining pipeline aggregations the output of each pipeline aggregation will be included in the final output.

buckets_path Syntax

edit

Most pipeline aggregations require another aggregation as their input. The input aggregation is defined via the buckets_path parameter, which follows a specific format:

AGG_SEPARATOR       =  `>` ;
METRIC_SEPARATOR    =  `.` ;
AGG_NAME            =  <the name of the aggregation> ;
METRIC              =  <the name of the metric (in case of multi-value metrics aggregation)> ;
MULTIBUCKET_KEY     =  `[<KEY_NAME>]`
PATH                =  <AGG_NAME><MULTIBUCKET_KEY>? (<AGG_SEPARATOR>, <AGG_NAME> )* ( <METRIC_SEPARATOR>, <METRIC> ) ;

For example, the path "my_bucket>my_stats.avg" will path to the avg value in the "my_stats" metric, which is contained in the "my_bucket" bucket aggregation.

Here are some more examples:

+

  • multi_bucket["foo"]>single_bucket>multi_metric.avg will go to the avg metric in the "multi_metric" agg under the single bucket "single_bucket" within the "foo" bucket of the "multi_bucket" multi-bucket aggregation.
  • agg1["foo"]._count will get the _count metric for the "foo" bucket in the multi-bucket aggregation "multi_bucket"

Paths are relative from the position of the pipeline aggregation; they are not absolute paths, and the path cannot go back "up" the aggregation tree. For example, this derivative is embedded inside a date_histogram and refers to a "sibling" metric "the_sum":

response = client.search(
  body: {
    aggregations: {
      my_date_histo: {
        date_histogram: {
          field: 'timestamp',
          calendar_interval: 'day'
        },
        aggregations: {
          the_sum: {
            sum: {
              field: 'lemmings'
            }
          },
          the_deriv: {
            derivative: {
              buckets_path: 'the_sum'
            }
          }
        }
      }
    }
  }
)
puts response
POST /_search
{
  "aggs": {
    "my_date_histo": {
      "date_histogram": {
        "field": "timestamp",
        "calendar_interval": "day"
      },
      "aggs": {
        "the_sum": {
          "sum": { "field": "lemmings" }              
        },
        "the_deriv": {
          "derivative": { "buckets_path": "the_sum" } 
        }
      }
    }
  }
}

The metric is called "the_sum"

The buckets_path refers to the metric via a relative path "the_sum"

buckets_path is also used for Sibling pipeline aggregations, where the aggregation is "next" to a series of buckets instead of embedded "inside" them. For example, the max_bucket aggregation uses the buckets_path to specify a metric embedded inside a sibling aggregation:

response = client.search(
  body: {
    aggregations: {
      sales_per_month: {
        date_histogram: {
          field: 'date',
          calendar_interval: 'month'
        },
        aggregations: {
          sales: {
            sum: {
              field: 'price'
            }
          }
        }
      },
      max_monthly_sales: {
        max_bucket: {
          buckets_path: 'sales_per_month>sales'
        }
      }
    }
  }
)
puts response
POST /_search
{
  "aggs": {
    "sales_per_month": {
      "date_histogram": {
        "field": "date",
        "calendar_interval": "month"
      },
      "aggs": {
        "sales": {
          "sum": {
            "field": "price"
          }
        }
      }
    },
    "max_monthly_sales": {
      "max_bucket": {
        "buckets_path": "sales_per_month>sales" 
      }
    }
  }
}

buckets_path instructs this max_bucket aggregation that we want the maximum value of the sales aggregation in the sales_per_month date histogram.

If a Sibling pipeline agg references a multi-bucket aggregation, such as a terms agg, it also has the option to select specific keys from the multi-bucket. For example, a bucket_script could select two specific buckets (via their bucket keys) to perform the calculation:

response = client.search(
  body: {
    aggregations: {
      sales_per_month: {
        date_histogram: {
          field: 'date',
          calendar_interval: 'month'
        },
        aggregations: {
          sale_type: {
            terms: {
              field: 'type'
            },
            aggregations: {
              sales: {
                sum: {
                  field: 'price'
                }
              }
            }
          },
          hat_vs_bag_ratio: {
            bucket_script: {
              buckets_path: {
                hats: "sale_type['hat']>sales",
                bags: "sale_type['bag']>sales"
              },
              script: 'params.hats / params.bags'
            }
          }
        }
      }
    }
  }
)
puts response
POST /_search
{
  "aggs": {
    "sales_per_month": {
      "date_histogram": {
        "field": "date",
        "calendar_interval": "month"
      },
      "aggs": {
        "sale_type": {
          "terms": {
            "field": "type"
          },
          "aggs": {
            "sales": {
              "sum": {
                "field": "price"
              }
            }
          }
        },
        "hat_vs_bag_ratio": {
          "bucket_script": {
            "buckets_path": {
              "hats": "sale_type['hat']>sales",   
              "bags": "sale_type['bag']>sales"    
            },
            "script": "params.hats / params.bags"
          }
        }
      }
    }
  }
}

buckets_path selects the hats and bags buckets (via ['hat']/['bag']`) to use in the script specifically, instead of fetching all the buckets from sale_type aggregation

Special Paths

edit

Instead of pathing to a metric, buckets_path can use a special "_count" path. This instructs the pipeline aggregation to use the document count as its input. For example, a derivative can be calculated on the document count of each bucket, instead of a specific metric:

response = client.search(
  body: {
    aggregations: {
      my_date_histo: {
        date_histogram: {
          field: 'timestamp',
          calendar_interval: 'day'
        },
        aggregations: {
          the_deriv: {
            derivative: {
              buckets_path: '_count'
            }
          }
        }
      }
    }
  }
)
puts response
POST /_search
{
  "aggs": {
    "my_date_histo": {
      "date_histogram": {
        "field": "timestamp",
        "calendar_interval": "day"
      },
      "aggs": {
        "the_deriv": {
          "derivative": { "buckets_path": "_count" } 
        }
      }
    }
  }
}

By using _count instead of a metric name, we can calculate the derivative of document counts in the histogram

The buckets_path can also use "_bucket_count" and path to a multi-bucket aggregation to use the number of buckets returned by that aggregation in the pipeline aggregation instead of a metric. For example, a bucket_selector can be used here to filter out buckets which contain no buckets for an inner terms aggregation:

response = client.search(
  index: 'sales',
  body: {
    size: 0,
    aggregations: {
      histo: {
        date_histogram: {
          field: 'date',
          calendar_interval: 'day'
        },
        aggregations: {
          categories: {
            terms: {
              field: 'category'
            }
          },
          min_bucket_selector: {
            bucket_selector: {
              buckets_path: {
                count: 'categories._bucket_count'
              },
              script: {
                source: 'params.count != 0'
              }
            }
          }
        }
      }
    }
  }
)
puts response
POST /sales/_search
{
  "size": 0,
  "aggs": {
    "histo": {
      "date_histogram": {
        "field": "date",
        "calendar_interval": "day"
      },
      "aggs": {
        "categories": {
          "terms": {
            "field": "category"
          }
        },
        "min_bucket_selector": {
          "bucket_selector": {
            "buckets_path": {
              "count": "categories._bucket_count" 
            },
            "script": {
              "source": "params.count != 0"
            }
          }
        }
      }
    }
  }
}

By using _bucket_count instead of a metric name, we can filter out histo buckets where they contain no buckets for the categories aggregation

Dealing with dots in agg names

edit

An alternate syntax is supported to cope with aggregations or metrics which have dots in the name, such as the 99.9th percentile. This metric may be referred to as:

"buckets_path": "my_percentile[99.9]"

Dealing with gaps in the data

edit

Data in the real world is often noisy and sometimes contains gaps — places where data simply doesn’t exist. This can occur for a variety of reasons, the most common being:

  • Documents falling into a bucket do not contain a required field
  • There are no documents matching the query for one or more buckets
  • The metric being calculated is unable to generate a value, likely because another dependent bucket is missing a value. Some pipeline aggregations have specific requirements that must be met (e.g. a derivative cannot calculate a metric for the first value because there is no previous value, HoltWinters moving average need "warmup" data to begin calculating, etc)

Gap policies are a mechanism to inform the pipeline aggregation about the desired behavior when "gappy" or missing data is encountered. All pipeline aggregations accept the gap_policy parameter. There are currently two gap policies to choose from:

skip
This option treats missing data as if the bucket does not exist. It will skip the bucket and continue calculating using the next available value.
insert_zeros
This option will replace missing values with a zero (0) and pipeline aggregation computation will proceed as normal.
keep_values
This option is similar to skip, except if the metric provides a non-null, non-NaN value this value is used, otherwise the empty bucket is skipped.