Question

我有一个客户端表，我需要按以下顺序选择数据：

Gender      Count
-----------------
Female      10
Male        50
Unknown     2

（上表数据仅作为示例）

为实现这一目标，我使用了以下SQL：

SELECT decode(gender, 'F', 'Female', 'M', 'Male', 'Unknown') "Gender"
, count(ID) "Count"
FROM client c
WHERE c.created <= '01-JAN-2000'
GROUP BY decode(gender, 'F', 'Female', 'M', 'Male', 'Unknown')
ORDER BY decode(gender, 'F', 'Female', 'M', 'Male', 'Unknown')
;

但是，问题是在1-JAN-2000之前没有创建客户端！因此，该表显示为空白：

Gender     Count
----------------

但是，我想在这种情况下表明：

Gender     Count
----------------
Female     0
Male       0
Unknown    0

请告诉我该怎么做？

Answer 1

您可以从性别上选择（使用内联视图）开始，然后对结果执行left outer join。这将保证在第一列中始终显示可用的性别列表：

SELECT g.gender
,      count(ID) "Count"
FROM   ( select 'Female' gender from dual
         union all
         select 'Male' gender from dual
         union all
         select 'Unknown' gender from dual
       ) g
left
outer
join   client c
on     decode(c.gender, 'F', 'Female', 'M', 'Male', 'Unknown') = g.gender
and    c.created <= '01-JAN-2000'
GROUP
BY     g.gender
ORDER
BY     g.gender
;

Answer 2

您应该规范化数据以保存性别表。或至少在您的查询中使用一个（视图），如下所示，并进行OUTER连接。

并使用标准日期格式..

编辑：为了更好的可读性和性能，您必须仔细分类性别。这里假设一个额外的类别'U'用于未知案例。

SELECT  g.GENDER "Gender"
, count(ID) "Count"
FROM client c, 
     ( SELECT 'Female' As Gender,'F' as GenderId FROM DUAL
       UNION ALL
       SELECT 'Male','M' FROM DUAL
       UNION ALL
       SELECT 'Unknown','U' FROM DUAL
     ) g
WHERE c.created(+) <= DATE '2000-01-01'
    AND c.Gender(+) = g.genderId
GROUP BY g.gender
ORDER BY g.gender
;

Answer 3

这是怎么做的

SELECT 
   'Female' Gender
   , Count(c.ID) Count
FROM 
   Client c
WHERE 
   c.Gender = 'F'

UNION ALL

SELECT 
   'Male' Gender
   , Count(c.ID) Count
FROM 
   Client c
WHERE 
   c.Gender = 'M'

UNION ALL

SELECT 
   'Unknown' Gender
   , Count(c.ID) Count
FROM 
   Client c
WHERE 
   c.Gender is NULL

如何在组中找不到记录时编写返回count = 0的SQL查询

3 个答案: