# name: test/sql/window/test_window_distinct.test
# description: Windowed distinct aggregates functionality
# group: [window]

statement ok
PRAGMA enable_verification

query I
SELECT COUNT(DISTINCT 42) OVER ()
----
1

query IIII
WITH t AS (
	SELECT col0 AS a, col1 AS b 
	FROM (VALUES
		(1,2),
		(1,1),
		(1,2),
		(2,1),
		(2,1),
		(2,2),
		(2,3),
		(2,4)
	) v) 
SELECT *, COUNT(b) OVER(PARTITION BY a), COUNT(DISTINCT b) OVER(PARTITION BY a) 
FROM t
ORDER BY 1, 2
----
1	1	3	2
1	2	3	2
1	2	3	2
2	1	5	4
2	1	5	4
2	2	5	4
2	3	5	4
2	4	5	4

statement ok
CREATE TABLE figure1 AS 
	SELECT * 
	FROM VALUES 
			(1, 'a'),
			(2, 'b'),
			(3, 'b'),
			(4, 'c'),
			(5, 'c'),
			(6, 'b'),
			(7, 'c'),
			(8, 'a')
		v(i, s);

query III
SELECT i
	, s
	, COUNT(DISTINCT s) OVER( ORDER BY i ROWS BETWEEN 2 PRECEDING AND 2 FOLLOWING) AS c
FROM figure1
ORDER BY i
----
1	a	2
2	b	3
3	b	3
4	c	2
5	c	2
6	b	3
7	c	3
8	a	3

query III
WITH uncascaded AS (
	SELECT i, i % 29 AS v 
	FROM range(1000) tbl(i)
)
SELECT i
	, v
	, COUNT(DISTINCT v) OVER (ORDER BY i ROWS BETWEEN 25 PRECEDING AND 25 FOLLOWING) AS w
FROM uncascaded
ORDER BY i
----
3000 values hashing to cb9c296986f7b9eaeee380bbc049ab39

query III
WITH cascaded AS (
	SELECT i, i % 29 AS v 
	FROM range(10000) tbl(i)
)
SELECT i
	, v
	, COUNT(DISTINCT v) OVER (ORDER BY i ROWS BETWEEN 25 PRECEDING AND 25 FOLLOWING) AS w
FROM cascaded
ORDER BY i
----
30000 values hashing to 673869e81fecab82f0bcec032236115a

# Exclude falls back to naïve
query IIII
SELECT i
	, s
	, i // 2 AS o
	, COUNT(DISTINCT s) OVER(
		ORDER BY i // 2
		ROWS BETWEEN 2 PRECEDING AND 2 FOLLOWING 
		EXCLUDE TIES
	) AS c
FROM figure1
ORDER BY i
----
1	a	0	2
2	b	1	3
3	b	1	3
4	c	2	2
5	c	2	2
6	b	3	3
7	c	3	2
8	a	4	3

# DISTINCT aggregate with NULL values in the dataset
statement ok
INSERT INTO figure1 VALUES 
	(9, NULL),
	(NULL, 'b'),
	(NULL, NULL),
;

query III
SELECT i
	, s
	, COUNT(DISTINCT s) OVER( ORDER BY i, s NULLS LAST ROWS BETWEEN 2 PRECEDING AND 2 FOLLOWING) AS c
FROM figure1
ORDER BY i, s NULLS LAST
----
1	a	2
2	b	3
3	b	3
4	c	2
5	c	2
6	b	3
7	c	3
8	a	3
9	NULL	3
NULL	b	2
NULL	NULL	1

# DISTINCT over nested types, e.g. LIST/STRUCT
statement ok
CREATE TABLE nested AS
	SELECT 
		i, 
		s, 
		{"m": i % 2, "s": s} AS n,
		[(i % 2)::VARCHAR, s] AS l,
		i * i AS r
	FROM figure1

query III
SELECT i
	, n
	, COUNT(DISTINCT n) OVER( ORDER BY i, s NULLS LAST ROWS BETWEEN 2 PRECEDING AND 2 FOLLOWING) AS c
FROM nested
ORDER BY i, s NULLS LAST
----
1	{'m': 1, 's': a}	3
2	{'m': 0, 's': b}	4
3	{'m': 1, 's': b}	5
4	{'m': 0, 's': c}	4
5	{'m': 1, 's': c}	4
6	{'m': 0, 's': b}	4
7	{'m': 1, 's': c}	4
8	{'m': 0, 's': a}	5
9	{'m': 1, 's': NULL}	5
NULL	{'m': NULL, 's': b}	4
NULL	{'m': NULL, 's': NULL}	3

query III
SELECT i
	, l
	, COUNT(DISTINCT l) OVER( ORDER BY i, s NULLS LAST ROWS BETWEEN 2 PRECEDING AND 2 FOLLOWING) AS c
FROM nested
ORDER BY i, s NULLS LAST
----
1	[1, a]	3
2	[0, b]	4
3	[1, b]	5
4	[0, c]	4
5	[1, c]	4
6	[0, b]	4
7	[1, c]	4
8	[0, a]	5
9	[1, NULL]	5
NULL	[NULL, b]	4
NULL	[NULL, NULL]	3

# DISTINCT with RANGE instead of ROWS
query III
SELECT r
	, s
	, COUNT(DISTINCT s) OVER( ORDER BY r RANGE BETWEEN 10 PRECEDING AND 10 FOLLOWING) AS c
FROM nested
ORDER BY i, s NULLS LAST
----
1	a	2
4	b	2
9	b	3
16	c	2
25	c	1
36	b	1
49	c	1
64	a	1
81	NULL	0
NULL	b	1
NULL	NULL	1

# DISTINCT with an aggregate with a destructor (e.g. LIST or STRING_AGG)
query III
SELECT i
	, s
	, STRING_AGG(DISTINCT s, ', ') OVER( ORDER BY i, s NULLS LAST ROWS BETWEEN 2 PRECEDING AND 2 FOLLOWING) AS c
FROM nested
ORDER BY i, s NULLS LAST
----
1	a	a, b
2	b	a, b, c
3	b	a, b, c
4	c	b, c
5	c	b, c
6	b	c, b, a
7	c	c, b, a
8	a	b, c, a
9	NULL	c, a, b
NULL	b	a, b
NULL	NULL	b

# DISTINCT MEDIAN, or distinct for aggregates that have a special window function?
query III
SELECT i
	, s
	, MEDIAN(DISTINCT s) OVER( ORDER BY i, s NULLS LAST ROWS BETWEEN 2 PRECEDING AND 2 FOLLOWING) AS c
FROM nested
ORDER BY i, s NULLS LAST
----
1	a	a
2	b	b
3	b	b
4	c	b
5	c	b
6	b	b
7	c	b
8	a	b
9	NULL	b
NULL	b	a
NULL	NULL	b

# DISTINCT FILTER
query III
SELECT i
	, s
	, COUNT(DISTINCT s) 
		FILTER (WHERE i % 3 = 0) 
		OVER( ORDER BY i, s NULLS LAST ROWS BETWEEN 2 PRECEDING AND 2 FOLLOWING) 
		AS c
FROM nested
ORDER BY i, s NULLS LAST
----
1	a	1
2	b	1
3	b	1
4	c	1
5	c	1
6	b	1
7	c	1
8	a	1
9	NULL	0
NULL	b	0
NULL	NULL	0

query III
SELECT i
	, s
	, COUNT(DISTINCT s) FILTER (WHERE i % 3 = 1) OVER( ORDER BY i ROWS BETWEEN 2 PRECEDING AND 2 FOLLOWING) AS c
FROM nested
ORDER BY i
----
1	a	1
2	b	2
3	b	2
4	c	1
5	c	1
6	b	1
7	c	1
8	a	1
9	NULL	1
NULL	b	0
NULL	NULL	0

query III
SELECT i
	, s
	, COUNT(DISTINCT s) FILTER (WHERE i % 3 = 2) OVER( ORDER BY i ROWS BETWEEN 2 PRECEDING AND 2 FOLLOWING) AS c
FROM nested
ORDER BY i
----
1	a	1
2	b	1
3	b	2
4	c	2
5	c	1
6	b	2
7	c	2
8	a	1
9	NULL	1
NULL	b	1
NULL	NULL	0