Question

当我执行此查询时，depart_decode列中显示的所有离开的数字都显示为= Nil

select D.depart_decode 
(select count (Staff_NO) from AA.table where Depart_code=D.Depart_code and Depart_code=151 and Staff_Sub_NO=5) Manager,
(select count (Staff_NO) from AA.table where Depart_code=D.Depart_code and Depart_code=151 and Staff_Sub_NO=4) HOD,
(select count (Staff_NO) from AA.table where Depart_code=D.Depart_code and Depart_code=151 and Staff_Sub_NO=3) Head,
(select count (Staff_NO) from AA.table where Depart_code=D.Depart_code and Depart_code=151 and Staff_Sub_NO in(1,2)) staff
 from AA.Departments D

Answer 1

您的查询在每个子选择中包含此where子句：

where Depart_code=D.Depart_code and Depart_code=151

如果表Departments中存在Depart_code 151并且仅针对该行，则此计算结果为true。

删除and Depart_code=151，您应该得到结果。

Answer 2

在correlated subqueries中使用select clause通常是性能问题的根源。我建议使用子查询来准备您想要的计数，然后将其加入到部门表中。

SELECT
      D.depart_decode, c.Manager, c.Hod, c.Head, c.staff
FROM AA.Departments D
INNER JOIN (
            SELECT
                  Depart_code
                , COUNT(CASE WHEN Staff_Sub_NO = 5 THEN Staff_NO END) Manager
                , COUNT(CASE WHEN Staff_Sub_NO = 4 THEN Staff_NO END) HOD
                , COUNT(CASE WHEN Staff_Sub_NO = 3 THEN Staff_NO END) Head
                , COUNT(CASE WHEN Staff_Sub_NO < 3 THEN Staff_NO END) staff
            FROM AA.table
            GROUP BY
                  Depart_code
            ) C ON D.Depart_code = C.Depart_code
;

来自查询的子选择的意外答案

2 个答案: