計算以下項的不同值：A列和一對列（A，B）

Question

假設我們有以下數據：

    +-----------+----------+---------+
    | user_code | order_id | line_id |
    +-----------+----------+---------+
    | ezz       | 1        | 1       |
    +-----------+----------+---------+
    | ezz       | 1        | 2       |
    +-----------+----------+---------+
    | ezz       | 1        | 2       |
    +-----------+----------+---------+
    | ezz       | 2        | 1       |
    +-----------+----------+---------+
    | ezz       | 2        | 2       |
    +-----------+----------+---------+
    | ezz       | 2        | 2       |
    +-----------+----------+---------+
    | ezz       | 2        | 2       |
    +-----------+----------+---------+
    | ezz       | 2        | 3       |
    +-----------+----------+---------+
    | ezz       | 3        | 1       |
    +-----------+----------+---------+

對於給定的user_code ，我們如何計算它具有多少個唯一的order_id以及它具有多少個唯一對（order_id，line_id） ？

所需的結果應類似於：

+-----------+-------------+------------------+
| user_code | order_count | order_line_count |
+-----------+-------------+------------------+
| ezz       | 3           | 6                |
+-----------+-------------+------------------+

Answer 1

您可以使用：

SELECT user_code,
     COUNT(DISTINCT order_id) AS order_count,
     COUNT(DISTINCT order_id || '^' || line_id) AS order_line_count
                    -- building single value from 2 columns
FROM tab
GROUP BY user_code

Answer 2

不幸的是，Oracle不支持帶有多個參數的count(distinct) 。 您可以使用字符串串聯：

select user_code, count(distinct order_id) as num_orders,
       count(distinct order_id || ':' || line_id) as num_order_lines
from t
group by user_code;

Answer 3

一種方法是先對數據進行重復數據刪除，然后使用count（distinct）和標准count。

select   user_code, count(distinct order_id) as ct_ord_id, 
                    count(line_id)           as ct_line_id
from     (select distinct user_code, order_id, line_id from TABLE_NAME)
group by user_code
;