如何 select 只有一个名称与 Redshift SQL 中的 ID 配对

Question

I have a dataset in which each id is paired with more than 1 name and in some cases the name is null. I need to decouple it to select just once the ID and the most recent updated name.

id	name	last_updated
001	name1	04/01/2022
001	name2	05/01/2022
001	NULL	06/01/2022
002	name3	04/01/2022
002	name4	05/01/2022
002	NULL	06/01/2022

In this case, I need to select the distinct IDs (001, 002) and the most recent name which is not null

id	name	last_updated
001	name2	05/01/2022
002	name4	05/01/2022

Answer 1

It is a typical task using SQL window functions.

We rank every row based on id and order them by date column. After it, we choose only rows with the first rank

with temp as (select *,
                     rank() over (partition by id order by last_updated desc) as rank
              from test.temp
              where name is not null)
select id, name, last_updated
from temp
where rank = 1