sum 函数的用法

SUM()函数用于计算一组值或表达式的总和,SUM()函数的语法如下:

SUM(DISTINCT expression)

SUM()函数是如何工作的?

  • 如果在没有返回匹配行 SELECT 语句中使用SUM函数,则SUM函数返回NULL,而不是0
  • DISTINCT运算符允许计算集合中的不同值。
  • SUM函数忽略计算中的NULL值。

SUM() 函数示例

MySQL 示例数据库模式由以下表组成:

  • customers: 存储客户的数据。
  • products: 存储汽车的数据。
  • productLines: 存储产品类别数据。
  • orders: 存储客户订购的销售订单。
  • orderDetails: 存储每个销售订单的订单产品数据项。
  • payments: 存储客户订单的付款数据信息。
  • employees: 存储所有员工信息以及组织结构,例如,直接上级 (谁向谁报告工作)。
  • offices: 存储销售处数据,类似于各个分公司。
    526150748_17922.png

可以使用SUM()函数来计算订单编号10100的总金额,如下查询所示:

SELECT FORMAT(SUM(quantityOrdered * priceEach),2) total FROM orderdetails WHERE orderNumber = 10100;

执行上面查询语句,得到以下结果 -

mysql> SELECT FORMAT(SUM(quantityOrdered * priceEach),2) total FROM orderdetails WHERE orderNumber = 10100;  
+-----------+  
| total     |  
+-----------+  
| 10,223.83 |  
+-----------+  
1 row in set

请注意,FORMAT()函数用于格式化SUM()函数的返回值。

SUM 与 GROUP BY 子句

当与GROUP BY子句组合时,SUM()函数计算 GROUP BY 子句中指定的每个分组的总和。

例如,可以使用具有GROUP BY子句的SUM函数计算每个订单的总金额,如下所示:

SELECT orderNumber,  
 FORMAT(SUM(quantityOrdered * priceEach),2) total  
FROM orderdetails  
GROUP BY orderNumber  
ORDER BY SUM(quantityOrdered * priceEach) DESC;

执行上面查询语句,得到以下结果 -

+-------------+-----------+  
| orderNumber | total     |  
+-------------+-----------+  
|       10165 | 67,392.85 |  
|       10287 | 61,402.00 |  
|       10310 | 61,234.67 |  
|       10212 | 59,830.55 |  
*** 此处省略了一大波数据 *****  
|       10116 | 1,627.56  |  
|       10158 | 1,491.38  |  
|       10144 | 1,128.20  |  
|       10408 | 615.45    |  
+-------------+-----------+  
327 rows in set

SUM 与 HAVING

您可以使用 HAVING 子句中在SUM函数中来根据特定条件过滤结果。例如,您可以计算总订单量,只能选择总金额大于60000的订单。如下查询语句 -

SELECT orderNumber,  
 FORMAT(SUM(quantityOrdered * priceEach),2)  
FROM orderdetails  
GROUP BY orderNumber  
HAVING SUM(quantityOrdered * priceEach) > 60000  
ORDER BY SUM(quantityOrdered * priceEach);

执行上面查询语句,得到以下结果 -

mysql> SELECT orderNumber,  
 FORMAT(SUM(quantityOrdered * priceEach),2)  
FROM orderdetails  
GROUP BY orderNumber  
HAVING SUM(quantityOrdered * priceEach) > 60000  
ORDER BY SUM(quantityOrdered * priceEach);  
+-------------+--------------------------------------------+  
| orderNumber | FORMAT(SUM(quantityOrdered * priceEach),2) |  
+-------------+--------------------------------------------+  
|       10310 | 61,234.67                                  |  
|       10287 | 61,402.00                                  |  
|       10165 | 67,392.85                                  |  
+-------------+--------------------------------------------+  
3 rows in set

SUM 与 LIMIT

假设您想要计算products表中前十名最昂贵的产品的总和,可以提出以下查询:

SELECT SUM(buyprice)  
FROM products  
ORDER BY buyprice DESC  
LIMIT 10;

执行上面查询语句,得到以下结果 -

mysql> SELECT SUM(buyprice)  
FROM products  
ORDER BY buyprice DESC  
LIMIT 10;  
+---------------+  
| SUM(buyprice) |  
+---------------+  
| 5983.47       |  
+---------------+  
1 row in set

它不起作用,因为具有SUM函数的SELECT语句只返回一行,LIMIT子句约束要返回的行数无效。

要解决此问题,请使用以下子查询:

SELECT FORMAT(SUM(buyprice),2) FROM  
(SELECT buyprice  
FROM products  
ORDER BY buyprice DESC  
LIMIT 10) price;

执行上面查询语句,得到以下结果 -

+-------------------------+  
| FORMAT(SUM(buyprice),2) |  
+-------------------------+  
| 958.71                  |  
+-------------------------+  
1 row in set

上面语句是怎么运行的?

  • 子查询选择十大价格最高的产品。
  • 外部查询计算从子查询返回的前10个价格最高的产品的总和。

SUM 与 NULL

如果没有匹配的行,则SUM函数返回NULL值。 有时,您希望SUM函数返回0而不是NULL。 在这种情况下,可以使用COALESCE函数。COALESCE函数接受两个参数,如果第一个参数为NULL,则返回第二个参数,否则返回第一个参数; 参考以下查询语句:

SELECT COALESCE(SUM(quantityOrdered * priceEach),0)  
FROM orderdetails  
WHERE productCode = 'S1_212121';

执行上面查询语句,得到以下结果 -

mysql> SELECT COALESCE(SUM(quantityOrdered * priceEach),0)  
FROM orderdetails  
WHERE productCode = 'S1_212121';  
+----------------------------------------------+  
| COALESCE(SUM(quantityOrdered * priceEach),0) |  
+----------------------------------------------+  
| 0.00                                         |  
+----------------------------------------------+  
1 row in set

SUM 与连接语句

您可以使用 SELECT JOIN 语句中的SUM函数根据另一个表中的值指定的条件来计算表中的值的总和。

例如,要计算取消订单金额的总和,请使用以下语句:

SELECT FORMAT(SUM(quantityOrdered * priceEach),2) loss  
FROM orderdetails  
INNER JOIN orders USING(orderNumber)  
WHERE status = 'Cancelled'

SUM(布尔表达式)

创建表

DROP TABLE IF EXISTS `a`;
CREATE TABLE `a`  (
  `日期` date NULL DEFAULT NULL,
  `结果` varchar(255) CHARACTER SET utf8 COLLATE utf8_general_ci NULL DEFAULT NULL
) ENGINE = InnoDB CHARACTER SET = utf8 COLLATE = utf8_general_ci ROW_FORMAT = Compact;

插入数据

INSERT INTO `a` VALUES ('2015-05-09', '胜');
INSERT INTO `a` VALUES ('2015-05-09', '胜');
INSERT INTO `a` VALUES ('2015-05-09', '负');
INSERT INTO `a` VALUES ('2015-05-09', '负');
INSERT INTO `a` VALUES ('2015-05-10', '胜');
INSERT INTO `a` VALUES ('2015-05-10', '负');
INSERT INTO `a` VALUES ('2015-05-10', '负');

查询

select 日期,sum(结果='胜') as 胜,sum(结果='负') as 负 from a group by 日期;

F7771C6BDE814290A0260E0558F8D8FD.jpg

SUM(if 条件)

select  source.to_els_account as elsAccount ,sum(if(source.performance_deduce=0 and source.grade='III',source.fbk2,0))as significantQualityLevelThreeScore,
    sum(if(source.performance_deduce=0 and source.grade='II',source.fbk2,0))as significantQualityLevelTwoScore,sum(if(source.performance_deduce=0 and source.grade='I',source.fbk2,0))as significantQualityLevelOneScore,
    sum(source.performance_deduce=0 and source.grade='III')as significantQualityLevelThreeNum,
    sum(source.performance_deduce=0 and source.grade='II')as significantQualityLevelTwoNum,sum(source.performance_deduce=0 and source.grade='I')as significantQualityLevelOneNum,
    DATE_FORMAT(source.oa_syn_date,'%Y-%m') as supplierPerformanceSourceDate,
    info.org_name as companyName,info.org_code as company
    FROM purchase_material_major_quality_fault_source source,purchase_organization_info info
    where 1=1
    and info.org_code=source.company and info.org_category_code='companyCode'
    and source.oa_syn_date between #{startTime} and #{endTime}
group by source.to_els_account,company,supplierPerformanceSourceDate