APPROX_COUNT_DISTINCT

Counts the approximate number of unique or not NULL values. APPROX_COUNT_DISTINCT uses the HLL algorithm with the default parameter to control the sketch size set to 17. This is subject to change.

Syntax

APPROX_COUNT_DISTINCT ( <expr> )

Parameter	Description
`<expr>`	Valid values for the expression include column names or functions that return a column name.

By default, APPROX_COUNT_DISTINCT and COUNT(DISTINCT) return the same, approximate results. If you require a precise result for COUNT(DISTINCT) (with a performance penalty), please contact Firebolt Support through the Help menu support form.

Example

To understand the difference between COUNT(DISTINCT pk) with exact precision enabled, APPROX_COUNT_DISTNCT(pk), and HLL_COUNT_DISTINCT(pk, <precision>), consider a table, count_test with 8,388,608 unique pk values.

SELECT
	COUNT(DISTINCT pk) as count_distinct,
	APPROX_COUNT_DISTINCT(pk) as approx_count
	HLL_COUNT_DISTINCT(pk, 12) as hll12_count,
	HLL_COUNT_DISTINCT(pk, 20) as hll20_count
FROM
	count_test;

Returns:

Assuming 8,388,608 unique pk values, we will see results like:

' +----------------+--------------+-------------+-------------+
' | count_distinct | approx_count | hll12_count | hll20_count |
' +----------------+--------------+-------------+-------------+
' |      8,388,608 |    8,427,387 |   8,667,274 |   8,377,014 |
' +----------------+--------------+-------------+-------------+

where approx_count is using precision 17, hll12_count is using precision 12, and hll20_count is using precision 20, the most precise.