approximate

Usage

view: view_name {
  measure: field_name {
    approximate: yes 
  }
}

Definition

See the Dialect support for approximate section on this page for the list of dialects that support indexes.

The approximate parameter lets you use approximate counting with measures of type: count and type: count_distinct. With large datasets, approximate counts can be much faster than exact counts and are typically within a few percent of the actual value. Please check your SQL dialect's documentation to understand the speed and accuracy tradeoffs of this method.

measure: apx_unique_count {
  type: count_distinct
  approximate: yes   # default value is no
  sql: ${id} ;;
}

-

Turning on approximate with a measure of type: count might seem unnecessary, because the approximate counting feature applies only to distinct counts. However, there are some situations when Looker automatically turns measures of type: count into a distinct count of a primary key to provide accurate results for joined views. In those situations, approximate counting may be useful.

Dialect support for approximate

The ability to use approximate depends on the database dialect your Looker connection is using. In the latest version of Looker, the following dialects support approximate: