PostgreSQL：在GROUPBY之后使用LEAST（）命令实现第一个事务

我正在使用magento表，如下所示：

+-----------+--------------+------------+--------------+----------+-------------+
| date      | email        | product_id | product_type | order_id | qty_ordered |
+-----------+--------------+------------+--------------+----------+-------------+
| 2017/2/15 | x@y.com      | 18W1       | custom       | 12       | 1           |
+-----------+--------------+------------+--------------+----------+-------------+
| 2017/2/15 | x@y.com      | 18W2       | simple       | 17       | 3           |
+-----------+--------------+------------+--------------+----------+-------------+
| 2017/2/20 | z@abc.com    | 22Y34      | simple       | 119      | 1           |
+-----------+--------------+------------+--------------+----------+-------------+
| 2017/2/20 | z@abc.com    | 22Y35      | custom       | 31       | 2           |
+-----------+--------------+------------+--------------+----------+-------------+

+-----------+--------------+------------+--------------+----------+-------------+
| date      | email        | product_id | product_type | order_id | qty_ordered |  
+-----------+--------------+------------+--------------+----------+-------------+
| 2017/2/15 | x@y.com      | 18W1       | custom       | 17       | 1           |
+-----------+--------------+------------+--------------+----------+-------------+
| 2017/2/15 | z@abc.com    | 18W2       | simple       |   31     | 3           |
+-----------+--------------+------------+--------------+----------+-------------+

我想通过电子邮件分组来创建一个新视图，然后只获取order_id最少的行

因此，从上面执行此操作后，我的最终表格应如下所示：

+-----------+--------------+------------+--------------+----------+-------------+
| date      | email        | product_id | product_type | order_id | qty_ordered |
+-----------+--------------+------------+--------------+----------+-------------+
| 2017/2/15 | x@y.com      | 18W1       | custom       | 12       | 1           |
+-----------+--------------+------------+--------------+----------+-------------+
| 2017/2/15 | x@y.com      | 18W2       | simple       | 17       | 3           |
+-----------+--------------+------------+--------------+----------+-------------+
| 2017/2/20 | z@abc.com    | 22Y34      | simple       | 119      | 1           |
+-----------+--------------+------------+--------------+----------+-------------+
| 2017/2/20 | z@abc.com    | 22Y35      | custom       | 31       | 2           |
+-----------+--------------+------------+--------------+----------+-------------+

+-----------+--------------+------------+--------------+----------+-------------+
| date      | email        | product_id | product_type | order_id | qty_ordered |  
+-----------+--------------+------------+--------------+----------+-------------+
| 2017/2/15 | x@y.com      | 18W1       | custom       | 17       | 1           |
+-----------+--------------+------------+--------------+----------+-------------+
| 2017/2/15 | z@abc.com    | 18W2       | simple       |   31     | 3           |
+-----------+--------------+------------+--------------+----------+-------------+

我正在尝试使用以下查询（但不起作用）：

---

我真的很想得到你的帮助，谢谢

我想你需要在上有明显的

：

select distinct on (email) t.*
from t
order by email, order_id;

distinct on
是Postgres的扩展。根据
order by
子句，它为括号中的所有键组合获取一条记录。在这种情况下，它是每
电子邮件一行，第一行是最小的
订单id
（因为
订单人
）。
select
中的键也必须是
order by
中的第一个键。最后一个表示例没有意义。这里面有错误吗？