SQL联表查询以及在查询的结果最后添加合计

最新推荐文章于 2023-06-07 17:13:20 发布

ITKouSyou

最新推荐文章于 2023-06-07 17:13:20 发布

阅读量3.8k

点赞数

CC 4.0 BY-SA版权

分类专栏： SQL 文章标签：联表查询合计 SQL

本文链接：https://blog.youkuaiyun.com/hx765287443/article/details/80744631

SQL 专栏收录该内容

2 篇文章

订阅专栏

本文介绍了一种使用SQL查询特定时间段内不同渠道的访问数和注册数的方法，并提供了如何针对特定渠道进行查询及添加合计数的示例。

摘要生成于 C知道，由 DeepSeek-R1 满血版支持，前往体验 >

先创建两个示例表，并插入测试数据。

source表：

插入数据

channel表：

插入数据

查询每个channel对应的访问数和注册数：

SELECT DISTINCT(channel_name) as cname, identification as iden,
(select count(*) from source 
WHERE create_time >= '2018-06-12' and create_time < '2018-06-17' and type="visit" and a.id=channel_id)as uv ,
(select count(*) from source 
WHERE create_time >= '2018-06-12' and create_time < '2018-06-17' and type="register" and a.id=channel_id) as reg
from source
RIGHT JOIN channel a
on a.id=channel_id;

结果为：

如果想要只查询测试1则在最后加个查询条件：

SELECT DISTINCT(channel_name) as cname, identification as iden,
(select count(*) from source 
WHERE create_time >= '2018-06-12' and create_time < '2018-06-17' and type="visit" and a.id=channel_id)as uv ,
(select count(*) from source 
WHERE create_time >= '2018-06-12' and create_time < '2018-06-17' and type="register" and a.id=channel_id) as reg
from source
RIGHT JOIN channel a
on a.id=channel_id
where channel_id=66;

结果：

想要在此基础上再加个合计，方法有两种，

第一种

在之前基础上再加上个合计的查询，拼起来

SELECT '所有' as cname, '-' as iden,
(select count(*) from source 
where create_time >= '2018-06-12' and create_time < '2018-06-17' and type="visit" )as uv ,
(select count(*) from source 
WHERE create_time >= '2018-06-12' and create_time < '2018-06-17' and type="register" ) as reg
from source  
GROUP BY cname;

加条件的话是这样：

SELECT '所有' as cname, '-' as iden,
(select count(*) from source 
where create_time >= '2018-06-12' and create_time < '2018-06-17' and channel_id=66   and type="visit" )as uv ,
(select count(*) from source 
WHERE create_time >= '2018-06-12' and create_time < '2018-06-17' and channel_id=66  and type="register" ) as reg
from source  
GROUP BY cname;

第二种：

select   
     case when c_name is not null then c_name else '所有渠道' end c_name,  
     sum(uv) as uvs,  
     sum(reg) as regs  
from (SELECT DISTINCT(channel_name) as c_name, 
(select count(*) from source 
WHERE create_time >= '2018-06-12' and create_time < '2018-06-17' and type="visit" and a.id=channel_id)as uv ,
(select count(*) from source 
WHERE create_time >= '2018-06-12' and create_time < '2018-06-17' and type="register" and a.id=channel_id) as reg
from source b
RIGHT JOIN channel a
on a.id=channel_id ) as c
group by c_name
with rollup;