我正在尝试创建一个查询,该查询会将一个现有表中的数据输入到另一个现有表中。两者之间没有共同的ID字段。
我有以下现有表t1
----------+------+-------+
|user | criteria | record|
----------+------+-------+
| 1 | 11 | K |
----------+------+-------+
| 1 | 12 | L |
----------+------+-------+
| 1 | 13 | M |
----------+------+-------+
| 1 | 16 | P |
----------+------+-------+
| 1 | 18 | R |
----------+------+-------+
| 1 | 20 | T |
----------+------+-------+
| 2 | 11 | K |
----------+------+-------+
| 2 | 12 | L |
----------+------+-------+
| 2 | 13 | M |
----------+------+-------+
| 2 | 16 | P |
----------+------+-------+
| 2 | 18 | R |
----------+------+-------+
| 2 | 20 | T |
----------+------+-------+
用户数量众多,(约有100个标准)有6个标准
我有兴趣插入以下现有表格中
table t2
+----------+----------+----------+----------+----------+----------+
| Label u | Label v | Label w | Label x | Label y | Label z |
+----------+----------+----------+----------+----------+----------+
| record K | record L | record M | record P | record R | record T|
| record K | record L | record M | record P | record R | record T|
+----------+----------+----------+----------+----------+----------+
where Criteria number 11 = Label u
Criteria number 12 = Label v
Criteria number 13 = Label w
Criteria number 16 = Label x
Criteria number 18 = Label y
Criteria number 20 = Label z
Note* Line 1 in t2 corresponds to user 1
Line 2 in t2 corresponds to user 2
There is no "user" column in t2 for the fields "user 1, user 2"
t2 already contains data in other columns
t1用户与user中t2中的列相同。从t1插入的t1用户的数据必须与t2已经存在的同一用户的数据匹配。
我的最终查询是这样的(不起作用)
INSERT INTO t2 a
SELECT
GROUP_CONCAT(IF (criteria = 11,record,NULL)) AS Label_u,
GROUP_CONCAT(IF (criteria = 12,record,NULL)) AS Label_v,
GROUP_CONCAT(IF (criteria = 13,record,NULL)) AS Label_w,
GROUP_CONCAT(IF (criteria = 14,record,NULL)) AS Label_x,
GROUP_CONCAT(IF (criteria = 15,record,NULL)) AS Label_y,
GROUP_CONCAT(IF (criteria = 16,record,NULL)) AS Label_z
FROM mytable b
WHERE a .user field = b.user
GROUP BY USER
ORDER BY USER;;
I have tried many - Can someone help me find a functional query to do this?
最佳答案
这是您的数据透视表。如果您不想拥有用户ID,则可以删除2.行
SELECT
user,
GROUP_CONCAT(IF (criteria = 11,record,NULL)) AS Label_u,
GROUP_CONCAT(IF (criteria = 12,record,NULL)) AS Label_v,
GROUP_CONCAT(IF (criteria = 13,record,NULL)) AS Label_w,
GROUP_CONCAT(IF (criteria = 14,record,NULL)) AS Label_x,
GROUP_CONCAT(IF (criteria = 15,record,NULL)) AS Label_y,
GROUP_CONCAT(IF (criteria = 16,record,NULL)) AS Label_z
FROM mytable
WHERE user IN (1,2)
GROUP BY USER
ORDER BY USER;
结果
+------+---------+---------+---------+---------+---------+---------+
| user | Label_u | Label_v | Label_w | Label_x | Label_y | Label_z |
+------+---------+---------+---------+---------+---------+---------+
| 1 | P1 | P1 | P1 | P1 | R1 | T1 |
| 2 | P2 | P2 | P2 | P2 | R2 | T2 |
+------+---------+---------+---------+---------+---------+---------+
2 rows in set (0.00 sec)
MariaDB >
关于mysql - MySQL Pivot查询到现有表,我们在Stack Overflow上找到一个类似的问题:https://stackoverflow.com/questions/33266805/