如何在 1 列中计算多个事物答案

【问题标题】：How to count multiple things in 1 column如何在 1 列中计算多个事物
【发布时间】：2014-08-05 17:48:21
【问题描述】：

我有一张桌子empname、age 和gender

     empname   age    gender
     Nana      24       F
     Jillian   28       M
     Sally     29       F
     David     30       M
     Mike      35       M
     Daisy     37       F

问题是我如何从名为employee 的表中计算女性30、男性30 的性别

     female<30  female>30  male<30  male>30
         2          1         1        2

【问题讨论】：

标签： sql

【解决方案1】：

最简单的方法是使用多个子查询：

SELECT
(SELECT COUNT(*) FROM employee WHERE gender = F AND age<30) AS 'Female Below 30',
(SELECT COUNT(*) FROM employee WHERE gender = F AND age>=30) AS 'Female Above 30',
(SELECT COUNT(*) FROM employee WHERE gender = M AND age<30) AS 'Male Below 30',
(SELECT COUNT(*) FROM employee WHERE gender = M AND age>=30) AS 'Male Above 30'

我认为在这种情况下使用 PIVOT 更好：

SELECT  [FB30],
        [FA30],
        [MB30],
        [MA30]
FROM    (
            SELECT  empname, 
                    gender + CASE WHEN age<30 THEN 'B30' ELSE 'A30' END AS SexAge 
            FROM    Employee
        ) AS SourceTable
        PIVOT
        (
            COUNT(empname)
            FOR SexAge IN ([FB30],[FA30],[MB30],[MA30])
        ) AS PivotTable;

看看这个SQL SERVER Fiddle Demo

经过一番搜索（我没有进入 Oracle ......）我最终得到了 PIVOT：

WITH
   T
AS
(
    SELECT "empname" ,
           "gender" || CASE WHEN "age"<30 THEN 'B30' ELSE 'A30' END AS "SexAge"
    FROM  Employee
)
SELECT
   SUM(FB30) AS FB30,SUM(FA30) AS FA30,SUM(MB30) AS MB30,SUM(MA30) AS MA30
FROM
   T
PIVOT
(
   COUNT(*)
   FOR
      ("SexAge")
   IN
      ('FB30' AS FB30,'FA30' AS FA30,'MB30' AS MB30,'MA30' AS MA30)
);

Oracle PIVOT Demo

既然您要求提供DECODE 示例：

WITH
   T
AS
(
    SELECT "empname" ,
           "gender" || DECODE(TRUNC("age"/30),0,'B30','A30') AS "SexAge"
    FROM  Employee
)
SELECT
   SUM(FB30) AS FB30,SUM(FA30) AS FA30,SUM(MB30) AS MB30,SUM(MA30) AS MA30
FROM
   T
PIVOT
(
   COUNT(*)
   FOR
      ("SexAge")
   IN
      ('FB30' AS FB30,'FA30' AS FA30,'MB30' AS MB30,'MA30' AS MA30)
);

Oracle DECODE PIVOT Demo

【讨论】：

这个工作正常，但我需要把不同的，所以数据不会重复
你想要Distinct在哪里？
没关系。我们可以使用 decode 作为解决方案吗？
你在Oracle吗？您要使用DECODE 代替大小写吗？
是的，我在Oracle，我想找到其他方法来获得结果

【解决方案2】：

你可以这样查询：

select
  count(case when gender='F' and age >= 30 then 1 end) "Female>=30",
  count(case when gender='F' and age < 30 then 1 end) "Female<30",
  count(case when gender='M' and age >= 30 then 1 end) "Male>=30",
  count(case when gender='M' and age < 30 then 1 end) "Male>30"
from ...

您可以只阅读一次表格以获得您想要的结果。当表很大时，这可能会有很大的不同。

【讨论】：

为什么要放1？如果我把 0 它也工作了
您可以输入除null 之外的任何值。 count 函数计数不是空值。

【解决方案3】：

请尝试：

select
    sum(case when gender='F' and age<30 then 1 end) [female<30],
    sum(case when gender='F' and age>=30 then 1 end) [female>30],
    sum(case when gender='M' and age<30 then 1 end) [male<30],
    sum(case when gender='M' and age>=30 then 1 end) [male>30]
from
    tbl

【讨论】：

【解决方案4】：

你可以这样做：

SELECT  SUM(CASE
               WHEN gender = "F" and age < 30
               THEN 1
               ELSE 0
        END ) AS female_lt_30
,
        SUM(CASE
               WHEN gender = "F" and age >= 30
               THEN 1
               ELSE 0
        END ) AS female_ge_30
,
        SUM(CASE
               WHEN gender = "M" and age < 30
               THEN 1
               ELSE 0
        END ) AS male_lt_30
,
        SUM(CASE
               WHEN gender = "M" and age >= 30
               THEN 1
               ELSE 0
        END ) AS female_ge_30

FROM the_table;

【讨论】：