运行总 MySQL 显示行的总和答案

【问题标题】：Running total MySQL showing the sum of row运行总 MySQL 显示行的总和
【发布时间】：2018-04-26 14:46:03
【问题描述】：

我有下表：

+-------------+-------+-------+
|    date     | agent |  hold |
+-------------+-------+-------+
| 2017-01-01  | A     |   100 |
| 2017-01-01  | B     |   200 |
| 2017-01-02  | C     |   400 |
+-------------+-------+-------+

我的查询要添加运行总计：

SET @runtot:=0;
SELECT
 date,
  SUM(CASE WHEN agent = 'A' THEN hold ELSE 0 END) AS A,
  SUM(CASE WHEN agent = 'B' THEN hold ELSE 0 END) AS B,
 SUM(CASE WHEN agent = 'C' THEN hold ELSE 0 END) AS C,
SUM(CASE WHEN agent = 'D' THEN hold ELSE 0 END) AS D,
sum(hold) as Total,
(@runtot := @runtot + sum(hold)) AS rt
FROM daily_results
group by date

但我在 rt 列中得到每行的总和（保持）。我做错了什么？

【问题讨论】：

在 sqlfiddle.com 上提供表结构和示例数据。同时根据该示例数据添加预期结果。
你在做什么？好像你需要WITH ROLLUP
我已经添加了表结构和结果
您有 3 行。这里肯定没有什么可以总结的。

标签： mysql sql cumulative-sum

【解决方案1】：

按照建议使用汇总或将聚合移动到子查询并在主查询中计算运行总计。例如给定

+------------+------------+---------+
| tanggal    | iddowntime | idshift |
+------------+------------+---------+
| 2017-10-03 | MOR01      |       1 |
| 2017-10-03 | NF001      |       2 |
| 2017-10-03 | MOR01      |       2 |
| 2017-10-02 | MOR01      |       2 |
| 2017-10-03 | NF001      |       2 |
| 2017-10-02 | MOR01      |       1 |
| 2017-10-02 | NF001      |       1 |
| 2017-10-02 | NF001      |       1 |
+------------+------------+---------+
8 rows in set (0.00 sec)

select s.*,
        @rt:=@rt + total RunningTotal
from
(
SELECT
  tanggal,
  SUM(CASE WHEN iddowntime = 'mor01' THEN idshift ELSE 0 END) AS A,
  SUM(CASE WHEN iddowntime = 'nf001' THEN idshift ELSE 0 END) AS B,
  #SUM(CASE WHEN agent = 'C' THEN hold ELSE 0 END) AS C,
  #SUM(CASE WHEN agent = 'D' THEN hold ELSE 0 END) AS D,
  sum(idshift) as Total
FROM trans_lhpdtdw
where iddowntime in('mor01','nf001')
group by tanggal 
) s
cross join
(select @rt:=0) rt

+------------+------+------+-------+--------------+
| tanggal    | A    | B    | Total | RunningTotal |
+------------+------+------+-------+--------------+
| 2017-10-02 |    3 |    2 |     5 |            5 |
| 2017-10-03 |    3 |    4 |     7 |           12 |
+------------+------+------+-------+--------------+
2 rows in set (0.01 sec)

【讨论】：