如果行中的初始值在 SQL 中为 < 或 > 0，如何 Return 分隔列的总和

Question

我正在尝试编写一个 SQL 查询，该查询将 return 根据值是负数还是正数，为每个客户提供独特的装运零件号总和行。例如：

如果我们两次将 20 个零件 Z 运送给客户 A，他们 return 将 6 个零件退回给我们，它会显示：

+-----------------+-------------+---------+
| Customer_Number | Part_Number | Shipped |
+-----------------+-------------+---------+
| A               | Z           |      20 |
| A               | Z           |      20 |
| A               | Z           |      -6 |
+-----------------+-------------+---------+

经过运行查询，预期结果：

+-----------------+-------------+---------+
| Customer_Number | Part_Number | Shipped |
+-----------------+-------------+---------+
| A               | Z           |      40 |
| A               | Z           |      -6 |
+-----------------+-------------+---------+

我提出的查询是：

SELECT 
    Customer_Number,
    Part_Number,
    CASE
        WHEN Ship_Quantity > 0 THEN SUM(Ship_Quantity)
        WHEN Ship_Quantity < 0 THEN SUM(Ship_Quantity)
    END
FROM Sales_Line
GROUP BY
    Customer_Number,
    Part_Number

但是我得到错误：

Column invalid. Must be a group by column: Ship_Quantity in SELECT LIST.

当我将“Ship_Quantity”添加到我的 GROUP BY 时，它没有给出准确的结果：

原始输入：

+-----------------+---------------+---------+
| Customer_Number |  Part_Number  | Shipped |
+-----------------+---------------+---------+
| A080            | C76           | -11.0   |
| A080            | C76           | -1.0    |
| A080            | C76           | -2.0    |
| A080            | C76           | -1.0    |
| A080            | C76           | -1.0    |
| A080            | C76           | 21.0    |
| A080            | C76           | 79.0    |
| A080            | C76           | 1.0     |
| A080            | C76           | 11.0    |
| A080            | C76           | 99.0    |
| A045            | X150          | -6.0    |
| A045            | X150          | -1.0    |
| A045            | X150          | -11.0   |
| A045            | X150          | -2.0    |
| A045            | X150          | -1.0    |
| A045            | X150          | -1.0    |
| A045            | X150          | -1.0    |
| A045            | X150          | 373.0   |
| A045            | X150          | 12.0    |
| A045            | X150          | 1.0     |
| A045            | X150          | 300.0   |
| A045            | X150          | 146.0   |
| A045            | X150          | 150.0   |
| A045            | X150          | 150.0   |
| A045            | X150          | 200.0   |
| A045            | X150          | 150.0   |
| A045            | X150          | 150.0   |
+-----------------+---------------+---------+

查询后：

+-----------------+---------------+---------+
| Customer_Number |  Part_Number  | Shipped |
+-----------------+---------------+---------+
| A045            | X150          | 300.0   |
| A045            | X150          | 373.0   |
| A080            | C76           | -11.0   |
| A080            | C76           | -2.0    |
| A080            | C76           | -3.0    |
+-----------------+---------------+---------+

我该怎么做？

Answer 1

在GROUP BY子句中使用函数SIGN()：

SELECT Customer_Number, Part_Number,
       SUM(Shipped) Shipped
FROM Sales_Line
GROUP BY Customer_Number, Part_Number, SIGN(Shipped)

如果您的数据库不支持 SIGN() 函数，请使用 CASE 表达式：

SELECT Customer_Number, Part_Number,
       SUM(Shipped) Shipped
FROM Sales_Line
GROUP BY Customer_Number, Part_Number, 
         CASE WHEN Shipped < 0 THEN -1 ELSE 1 END

参见demo。

如果行中的初始值在 SQL 中为 < 或 > 0，如何 Return 分隔列的总和

How to Return Separate Sums of Columns if Initial Value in row is < or > 0 in SQL

sql

group-by

row

sum

case