根据其他两个表更新列
Updating column based on other two tables
我有如下三个table,
select * from student_grade;
+--------------+----------+----------+-------+--------------------------+
| enrollmentid | courseid | personid | grade | credits_earned |
+==============+==========+==========+=======+==========================+
| 1 | 1001 | 1 | A | null |
| 2 | 1002 | 1 | B | null |
| 3 | 1003 | 1 | A | null |
| 4 | 3001 | 3 | A | null |
| 5 | 3001 | 2 | B | null |
| 6 | 4001 | 4 | A | null |
| 7 | 4002 | 4 | A | null |
+--------------+----------+----------+-------+--------------------------+
7 tuples
sql>select * from course;
+----------+-------------------+---------+--------------+
| courseid | title | credits | departmentid |
+==========+===================+=========+==============+
| 1001 | Data structures | 12 | 101 |
| 1002 | Algorithms | 12 | 101 |
| 1003 | Graphics | 20 | 101 |
| 2001 | DSP | 20 | 102 |
| 2002 | Matlab | 20 | 102 |
| 2003 | Maths | 10 | 102 |
| 3001 | CAD | 10 | 104 |
| 4001 | Power electronics | 10 | 103 |
| 4002 | Semi conductors | 20 | 103 |
+----------+-------------------+---------+--------------+
9 tuples
sql>select * from grade_to_credits;
+-----------+--------------------------+
| gradechar | credits |
+===========+==========================+
| A | 1 |
| B | 0.9 |
+-----------+--------------------------+
我想做的是:
我正在使用 course
table 的 credits
列和 credits
的 credits
列更新 student_grade
table 的 credits_earned
列18=] table.
就这样,
select c.credits * gc.credits
from course c, grade_to_credits gc, student_grade sg
where sg.courseid = c.courseid and sg.grade = gc.gradechar;
+--------------------------+
| L2 |
+==========================+
| 12 |
| 10.8 |
| 20 |
| 10 |
| 9 |
| 10 |
| 20 |
+--------------------------+
我在单独执行时得到了上述值,但现在我想使用 update 查询在 student_grade
table 中更新这些值。
现在我将查询用作:
update student_grade
set credits_earned = (select c.credits * gc.credits
from course c, grade_to_credits gc, student_grade sg
where sg.courseid = c.courseid
and sg.grade = gc.gradechar);
但是上面的查询不起作用,我得到的错误是:
Cardinality violation, scalar value expected
我知道,我可以单独设置值,但我想通过更新命令来设置。
不对的地方请指正
注意: 下面的解决方案适用于 MySQL。 (OP 稍后 retagged/changed 从 MySQL 到 MonetDB 的问题)
- 您尝试的问题是内部子查询返回多行。
- 即使它 returns 只有一行,它的另一个问题是用相同的值更新 table
student_grade
中的所有 credits_earned
字段(输出内部子查询)。
- 您应该避免使用隐式联接。他们太老了,不是一个好习惯。
你应该 INNER JOIN 这些 table,然后使用
加入 table 的列以更新按行数据。尝试以下操作:
update student_grade sg
inner join course c on c.courseid = sg.courseid
inner join grade_to_credits gc on gc.gradechar = sg.grade
set sg.credits_earned = c.credits * gc.credits;
如果我没理解错的话,你只需要从子查询中删除student_grade
:
update student_grade st
set credits_earned = (select sc.credits * gc.credits
from course c cross join
grade_to_credits gc
where sg.courseid = c.courseid and
sg.grade = gc.gradechar
);
无需使事情过于复杂,使用 LEFT JOIN
即可实现结果。集合运算通常比逐行运算更有效。
update student_grade
SET credits_earned = a.credits_earned
from (
select s.enrollmentid,
s.courseid,
s.personid,
s.grade,
(g.credits * c.credits) as credits_earned
from student_grade s
LEFT OUTER JOIN @grade_to_credits g ON g.gradechar = s.grade
LEFT OUTER JOIN @course c ON c.courseid = s.courseid
) as a
where student_grade.enrollmentid = a.enrollmentid
结果:
1 1001 1 A 12
2 1002 1 B 10,8
3 1003 1 A 20
4 3001 3 A 10
5 3001 2 B 9
6 4001 4 A 10
7 4002 4 A 20
我有如下三个table,
select * from student_grade;
+--------------+----------+----------+-------+--------------------------+
| enrollmentid | courseid | personid | grade | credits_earned |
+==============+==========+==========+=======+==========================+
| 1 | 1001 | 1 | A | null |
| 2 | 1002 | 1 | B | null |
| 3 | 1003 | 1 | A | null |
| 4 | 3001 | 3 | A | null |
| 5 | 3001 | 2 | B | null |
| 6 | 4001 | 4 | A | null |
| 7 | 4002 | 4 | A | null |
+--------------+----------+----------+-------+--------------------------+
7 tuples
sql>select * from course;
+----------+-------------------+---------+--------------+
| courseid | title | credits | departmentid |
+==========+===================+=========+==============+
| 1001 | Data structures | 12 | 101 |
| 1002 | Algorithms | 12 | 101 |
| 1003 | Graphics | 20 | 101 |
| 2001 | DSP | 20 | 102 |
| 2002 | Matlab | 20 | 102 |
| 2003 | Maths | 10 | 102 |
| 3001 | CAD | 10 | 104 |
| 4001 | Power electronics | 10 | 103 |
| 4002 | Semi conductors | 20 | 103 |
+----------+-------------------+---------+--------------+
9 tuples
sql>select * from grade_to_credits;
+-----------+--------------------------+
| gradechar | credits |
+===========+==========================+
| A | 1 |
| B | 0.9 |
+-----------+--------------------------+
我想做的是:
我正在使用 course
table 的 credits
列和 credits
的 credits
列更新 student_grade
table 的 credits_earned
列18=] table.
就这样,
select c.credits * gc.credits
from course c, grade_to_credits gc, student_grade sg
where sg.courseid = c.courseid and sg.grade = gc.gradechar;
+--------------------------+
| L2 |
+==========================+
| 12 |
| 10.8 |
| 20 |
| 10 |
| 9 |
| 10 |
| 20 |
+--------------------------+
我在单独执行时得到了上述值,但现在我想使用 update 查询在 student_grade
table 中更新这些值。
现在我将查询用作:
update student_grade
set credits_earned = (select c.credits * gc.credits
from course c, grade_to_credits gc, student_grade sg
where sg.courseid = c.courseid
and sg.grade = gc.gradechar);
但是上面的查询不起作用,我得到的错误是:
Cardinality violation, scalar value expected
我知道,我可以单独设置值,但我想通过更新命令来设置。 不对的地方请指正
注意: 下面的解决方案适用于 MySQL。 (OP 稍后 retagged/changed 从 MySQL 到 MonetDB 的问题)
- 您尝试的问题是内部子查询返回多行。
- 即使它 returns 只有一行,它的另一个问题是用相同的值更新 table
student_grade
中的所有credits_earned
字段(输出内部子查询)。 - 您应该避免使用隐式联接。他们太老了,不是一个好习惯。
你应该 INNER JOIN 这些 table,然后使用 加入 table 的列以更新按行数据。尝试以下操作:
update student_grade sg
inner join course c on c.courseid = sg.courseid
inner join grade_to_credits gc on gc.gradechar = sg.grade
set sg.credits_earned = c.credits * gc.credits;
如果我没理解错的话,你只需要从子查询中删除student_grade
:
update student_grade st
set credits_earned = (select sc.credits * gc.credits
from course c cross join
grade_to_credits gc
where sg.courseid = c.courseid and
sg.grade = gc.gradechar
);
无需使事情过于复杂,使用 LEFT JOIN
即可实现结果。集合运算通常比逐行运算更有效。
update student_grade
SET credits_earned = a.credits_earned
from (
select s.enrollmentid,
s.courseid,
s.personid,
s.grade,
(g.credits * c.credits) as credits_earned
from student_grade s
LEFT OUTER JOIN @grade_to_credits g ON g.gradechar = s.grade
LEFT OUTER JOIN @course c ON c.courseid = s.courseid
) as a
where student_grade.enrollmentid = a.enrollmentid
结果:
1 1001 1 A 12
2 1002 1 B 10,8
3 1003 1 A 20
4 3001 3 A 10
5 3001 2 B 9
6 4001 4 A 10
7 4002 4 A 20