Question

我有两个柜台，一个是单位，一个是废物。我需要提取计时器的总和并按年份分开 Left Join有点起作用，但是只要我把一个where条件分开定时器，那么思想就会崩溃。
有人可以查看一下，并指出我正确的方向吗？
谢谢。杰米。

 SELECT year(counts.ts) as SalesYear, Q1.Res1 AS units, Q2.Res2 as rate
 FROM 
 (counts
 LEFT JOIN (SELECT (counts.ts), SUM(counts.count) AS Res1 
 FROM counts where counts.`name` = 104 and counts.count > 0 ) AS Q1 ON (counts.ts) = Q1.ts) 

 LEFT JOIN (SELECT (counts.ts), Sum(counts.count) AS Res2
 FROM counts where counts.`name` = 107 and counts.count > 0) AS Q2 ON (counts.ts) = Q2.ts

 GROUP BY year(counts.ts)
 order by year(counts.ts)

Answer 1

我相信你可以像这样使用“条件聚合”：

SELECT
      YEAR(counts.ts) as SalesYear
    , SUM(case when counts.count = 104 then counts.count else 0 end) AS Res1 
    , SUM(case when counts.count = 107 then counts.count else 0 end) AS Res2
FROM counts 
WHERE counts.name IN (104,107)
      AND counts.count > 0 
GROUP BY YEAR(counts.ts) 
ORDER BY YEAR(counts.ts)
;

由于样本数据现已通过sqlfiddle实例提供，因此以下内容现已相关。请注意，上面的查询与问题一样，问题是条件错误地基于列counts.count而不是counts.name。一旦纠正，完全相同的查询结构就会提供结果。（请注意，如果没有看到数据，我就无法知道这一点。）

SQL Fiddle

MySQL 5.6架构设置：

DROP TABLE IF EXISTS `counts`;
CREATE TABLE `counts` (
  `id` int(15) NOT NULL AUTO_INCREMENT,
  `name` varchar(16) DEFAULT NULL,
  `ts` datetime DEFAULT NULL,
  `count` int(11) DEFAULT NULL,
  PRIMARY KEY (`id`),
  UNIQUE KEY `UniqueDataIndex` (`name`,`ts`,`count`) USING BTREE
) ENGINE=InnoDB AUTO_INCREMENT=475152 DEFAULT CHARSET=utf8;

INSERT INTO `counts` VALUES ('475026', '107             ', '2015-01-27 10:18:00', '40');
INSERT INTO `counts` VALUES ('475032', '104             ', '2015-01-27 10:20:00', '11');
INSERT INTO `counts` VALUES ('475039', '107             ', '2015-01-27 10:22:00', '109');
INSERT INTO `counts` VALUES ('475046', '104             ', '2015-01-27 10:24:00', '92');
INSERT INTO `counts` VALUES ('475052', '107             ', '2015-01-27 10:26:00', '124');
INSERT INTO `counts` VALUES ('475058', '104             ', '2016-01-27 10:28:00', '98');
INSERT INTO `counts` VALUES ('475069', '107             ', '2016-01-27 10:32:00', '77');
INSERT INTO `counts` VALUES ('475075', '104             ', '2016-01-27 10:34:00', '44');
INSERT INTO `counts` VALUES ('475081', '107             ', '2016-01-27 10:36:00', '43');
INSERT INTO `counts` VALUES ('475087', '104             ', '2016-01-27 10:38:00', '89');
INSERT INTO `counts` VALUES ('475093', '107             ', '2016-01-27 10:40:00', '89');

查询1 ：

SELECT
      YEAR(counts.ts) as SalesYear
    , SUM(case when counts.name = '104' then counts.count else 0 end) AS Res1 
    , SUM(case when counts.name = '107' then counts.count else 0 end) AS Res2
FROM counts 
WHERE counts.name IN ('104','107')
      AND counts.count > 0 
GROUP BY YEAR(counts.ts) 
ORDER BY YEAR(counts.ts)

<强> Results ：

| SalesYear | Res1 | Res2 |
|-----------|------|------|
|      2015 |  103 |  273 |
|      2016 |  231 |  209 |

左加入不符合条件

1 个答案: