Skip to content

SQL - PostgreSQL - Creating Equal-Sized Buckets using ntile

Source: sql-snippets/ntile.md at main · count/sql-snippets (github.com)

View an interactive version of this snippet here.

Description

When you want to create equal sized groups (i.e. buckets) of rows, there is a SQL function for that too. The function NTILE lets you do exactly that:

WITH users AS (
    SELECT md5(RANDOM()::TEXT) AS name_hash,
           CASE WHEN RANDOM() <= 0.5 THEN 'male' ELSE 'female' END AS gender,
           CASE WHEN RANDOM() <= 0.33 THEN 'student' WHEN RANDOM() <= 0.66 THEN 'professor' ELSE 'alumni' END AS status
    FROM generate_series(1, 10)
)

SELECT 
       name_hash,
       gender,
       status,
       NTILE(3) OVER(PARTITION BY status ORDER BY status) AS ntile_bucket
FROM
       users;

Output:

name_hash gender status ntile_bucket
8c9882e8cc6f748c06b3a20ee7af9787 female alumni 1
032dffdb7cb67a5f1a6b77b713043508 female alumni 2
72559a7e25935004d172329a47cb3db5 male alumni 3
284d1337a3d30433fe17ad2110d27655 male professor 1
3797bb418424d7ced117e8817f398fea male professor 2
e56d1e6e53f12c6a133c4f47a962754c female professor 3
8713e6b5a1c359f42e960c05460c8e14 male student 1
af8ea7ea643fbbfefe7c845e77c9e4d3 male student 1
033161ae441c347c4aa43acd46347836 female student 2
a476b8c2cfa8df1c9494c9851cbd4e66 female student 3

Backlinks:

list from [[SQL - PostgreSQL - Creating Equal-Sized Buckets using `ntile`]] AND -"Changelog"