approx_percentile_rank()

Since 1.0.0

Estimate the percentile at which a given value would be located.

Samples

Estimate the percentile rank of the value 99, given a sample containing the numbers from 0 to 100.

SELECT
  approx_percentile_rank(99, tdigest(data))
FROM generate_series(0, 100) data;

approx_percentile_rank
----------------------------
        0.9851485148514851

Arguments

The syntax is:

approx_percentile_rank(
  value DOUBLE PRECISION,
  digest TDigest
) RETURNS DOUBLE PRECISION

Name	Type	Default	Required	Description
value	DOUBLE PRECISION	-	✔	The value to estimate the percentile of
digest	TDigest	-	✔	The tdigest aggregate

Returns

Column	Type	Description
approx_percentile_rank	DOUBLE PRECISION	The estimated percentile associated with the provided value.