Window

CUME_DIST

PostgreSQL

Menghitung cumulative distribution (0 < x <= 1). Formula: jumlah baris dengan nilai <= baris ini / total baris. Berbeda dengan PERCENT_RANK, CUME_DIST tidak pernah 0.

Tipe hasil: double precision

Syntax

SQL

CUME_DIST() OVER (
  [PARTITION BY partition_expression]
  ORDER BY sort_expression
)

Parameter

ORDER BYexpressionwajib

Menentukan urutan distribusi

PARTITION BYexpressionopsional

Membagi data menjadi grup terpisah

Contoh Penggunaan

Cumulative Distribution

SQL

1 SELECT 
2   name,
3   score,
4   CUME_DIST() OVER (ORDER BY score) AS cume_dist
5 FROM students;

Persentase siswa dengan skor <= skor ini.

Hasil

name	score	cume_dist
Andi	60	0.2
Budi	70	0.4
Citra	80	0.6
Dewi	80	0.6
... 1 baris lainnya

Analisis Distribusi Harga

SQL

1 SELECT 
2   price_range,
3   COUNT(*) as products,
4   MAX(cume_dist) as cumulative_pct
5 FROM (
6   SELECT 
7     product_name,
8     price,
9     CASE 
10       WHEN price < 100000 THEN < 100K
11       WHEN price < 500000 THEN 100K-500K
12       ELSE > 500K
13     END AS price_range,
14     CUME_DIST() OVER (ORDER BY price) AS cume_dist
15   FROM products
16 ) t
17 GROUP BY price_range;

Analisis distribusi produk per range harga.

Filter Bottom 20%

SQL

1 SELECT *
2 FROM (
3   SELECT 
4     employee_id,
5     performance_score,
6     CUME_DIST() OVER (ORDER BY performance_score) AS cume_dist
7   FROM employees
8 ) ranked
9 WHERE cume_dist <= 0.2;

Mencari karyawan di bottom 20% performance.

Fungsi Equivalen di Database Lain

CUME_DIST

MySQL

CUME_DIST

Snowflake

CUME_DIST

BigQuery

CUME_DIST

DuckDB

Fungsi Terkait

ROW_NUMBERRANKDENSE_RANKPERCENT_RANKNTILE

Kembali ke PostgreSQL Functions

1	SELECT
2	name,
3	score,
4	CUME_DIST() OVER (ORDER BY score) AS cume_dist
5	FROM students;

1	SELECT
2	price_range,
3	COUNT(*) as products,
4	MAX(cume_dist) as cumulative_pct
5	FROM (
6	SELECT
7	product_name,
8	price,
9	CASE
10	WHEN price < 100000 THEN < 100K
11	WHEN price < 500000 THEN 100K-500K
12	ELSE > 500K
13	END AS price_range,
14	CUME_DIST() OVER (ORDER BY price) AS cume_dist
15	FROM products
16	) t
17	GROUP BY price_range;

1	SELECT *
2	FROM (
3	SELECT
4	employee_id,
5	performance_score,
6	CUME_DIST() OVER (ORDER BY performance_score) AS cume_dist
7	FROM employees
8	) ranked
9	WHERE cume_dist <= 0.2;