SQL分組和總數超過科目的計數

Question

我還沒有設法找到一個與我正在嘗試做的完全匹配的答案。

我不完全確定如何去做到這一點，相當誠實！

我有一個學生在各學科取得的成績表：

我想是一個表，其中顯示每個科目的成績分佈，每個年級的總獎勵：

到目前爲止，我的查詢是：

SELECT subjectcode, 
(SELECT COUNT(grade) FROM results WHERE grade = "A") as A, 
(SELECT COUNT(grade) FROM results WHERE grade = "B") as B, 
(SELECT COUNT(grade) FROM results WHERE grade = "C") as C, 
(SELECT COUNT(grade) FROM results WHERE grade = "D") as D, 
(SELECT COUNT(grade) FROM results WHERE grade = "E") as E 
FROM results 
GROUP BY grade 
ORDER BY subjectcode ASC; 

任何意見或建議，將不勝感激。

謝謝。

Answer 1

使用條件聚集，而不是多個子查詢：

SELECT subjectcode, 
     SUM(CASE WHEN grade = 'A' THEN 1 END) as A, 
     SUM(CASE WHEN grade = 'B' THEN 1 END) as B, 
     SUM(CASE WHEN grade = 'C' THEN 1 END) as C, 
     SUM(CASE WHEN grade = 'D' THEN 1 END) as D, 
     SUM(CASE WHEN grade = 'E' THEN 1 END) as E 
FROM results 
GROUP BY subjectcode 
ORDER BY subjectcode ASC; 

注：

• 的group by子句是固定的。您在group by中有grade，但仍然按subjectcode排序，這沒有意義。這在大多數數據庫中都是不允許的。
• case上沒有else條款，所以不符合收益率NULL。
• 字符串常量由單引號而不是雙引號分隔。單引號是ANSI標準，應該用於字符串和日期常量。
• 這被稱爲數據透視查詢，並且一些數據庫具有直接支持它們的語法。

Answer 2

declare @grades table(result_id int identity, upn int, grade varchar(2), subjectcode varchar(2)) 

insert into @grades (upn,grade,subjectcode) values 
(1,'D','Ar'), 
(1,'A','Bi'), 
(1,'C','Ch'), 
(2,'A*','Ar'), 
(2,'B','Ch'), 
(2,'A*','PE'), 
(3,'C','Ar'), 
(3,'C','Ph'), 
(3,'A','Bi') 


SELECT subjectcode, 
    ISNULL(SUM(CASE WHEN grade = 'A*' THEN 1 END),'') 'A*', 
    ISNULL(SUM(CASE WHEN grade = 'A' THEN 1 END),'') A, 
    ISNULL(SUM(CASE WHEN grade = 'B' THEN 1 END),'') B, 
    ISNULL(SUM(CASE WHEN grade = 'C' THEN 1 END),'') C, 
    ISNULL(SUM(CASE WHEN grade = 'D' THEN 1 END),'') D, 
    ISNULL(SUM(CASE WHEN grade = 'E' THEN 1 END),'') E, 
    ISNULL(SUM(CASE WHEN grade = 'F' THEN 1 END),'') F, 
    ISNULL(SUM(CASE WHEN grade = 'G' THEN 1 END),'') G, 
    ISNULL(SUM(CASE WHEN grade = 'U' THEN 1 END),'') U, 
    ISNULL(SUM(CASE WHEN grade = 'X' THEN 1 END),'') X 

FROM @grades 
GROUP BY subjectcode 
ORDER BY subjectcode ASC;