Postgresql如何将select列匹配条件？

Question

我有一个 table 这样的：

inventory_id | customer_id |         max         
--------------+-------------+---------------------
         4497 |           1 | 2005-07-28 00:00:00
         1449 |           1 | 2005-08-22 00:00:00
         1440 |           1 | 2005-08-02 00:00:00
         3232 |           1 | 2005-08-02 00:00:00
         3418 |           2 | 2005-08-02 00:00:00
          654 |           2 | 2005-08-02 00:00:00
         3164 |           2 | 2005-08-21 00:00:00
         2053 |           2 | 2005-07-27 00:00:00

我想要 select 行，其中最近的日期和相应的列，这就是我想要实现的：

inventory_id | customer_id |         max         
--------------+-------------+---------------------
         1449 |           1 | 2005-08-22 00:00:00
         3164 |           2 | 2005-08-21 00:00:00

我尝试使用聚合，但我需要 inventory_id 和 customer_id 同时出现。有没有什么方法可以做到这一点？

Answer 1

使用distinct on:

select distinct on (customer_id) t.*
from t
order by customer_id, max desc;

distinct on 是一个 Postgres 扩展，根据括号中的任何内容，行上 returns。此行基于 order by -- 排序集中出现的第一个。

Answer 2

    SELECT inventory_id, customer_id, max FROM 
    (SELECT inventory_id, customer_id, max, 
    ROW_NUMBER() OVER(PARTITION BY customer_id ORDER BY max DESC) AS ROWNO 
    FROM inventory_table) AS A
    WHERE ROWNO=1

Postgresql如何将select列匹配条件？

Postgresql how to select columns where it matches conditions?

sql

postgresql

greatest-n-per-group