In a table, there are ids with multiple values
id|value
--|------
1 |a
1 |b
2 |a
2 |b
3 |a
3 |c
4 |a
5 |a
5 |b
...
so how can i select all id's that have the same values as 1? What I want should look like this
id
--
1
2
5
Add a comment
|
2 Answers
If you want exactly the same values, you can use a set-based approach:
select t.id
from t join
t t1
on t.value = t1.value and t1.id = 1
group by t.id
having count(*) = (select count(*) from t where t.id = 1);
Assuming no duplicates, this counts the number of values that matches each id and then checks that there are the same number.
I should admit that the the string_agg() approach is also elegant:
select id
from (select id, string_agg(value, ',' order by value) as values,
max(case when id = 1 then string_agg(value, ',' order by value)) over () as values_1
from t
group by id
) t
where values = values_1;
Comments
Also using arrays:
SELECT id
FROM (SELECT DISTINCT id FROM t ORDER BY 1) ids
WHERE ARRAY(SELECT value FROM t WHERE id = ids.id ORDER BY 1) =
ARRAY(SELECT value FROM t WHERE id = 1 ORDER BY 1);
This should work even for duplicated and/or NULL values.
