我有一张表格,其中存储了测量温度和湿度的数据。带温度的记录将字段meas_kind
设置为" T",对于湿度,将其设置为" H"。
mysql> describe meteo;
+-----------+---------------------+------+-----+---------+----------------+
| Field | Type | Null | Key | Default | Extra |
+-----------+---------------------+------+-----+---------+----------------+
| id | bigint(20) unsigned | NO | PRI | NULL | auto_increment |
| meas_time | timestamp | YES | | NULL | |
| room | varchar(10) | NO | | NULL | |
| meas_kind | char(1) | NO | | NULL | |
| value | double | NO | | NULL | |
+-----------+---------------------+------+-----+---------+----------------+
mysql> select * from meteo;
+----+---------------------+------+-----------+-------+
| id | meas_time | room | meas_kind | value |
+----+---------------------+------+-----------+-------+
| 35 | 2017-05-24 16:51:47 | 123 | T | 22.61 |
| 36 | 2017-05-24 16:51:47 | 123 | H | 36.93 |
| 37 | 2017-05-24 16:51:51 | 123 | T | 22.61 |
| 38 | 2017-05-24 16:51:51 | 123 | H | 36.94 |
| 39 | 2017-05-24 16:51:56 | 123 | T | 22.61 |
| 40 | 2017-05-24 16:51:56 | 123 | H | 36.94 |
+----+---------------------+------+-----------+-------+
同时测量温度和湿度,所以我希望这个表格如下:
+---------------------+------+-------+-------+
| meas_time | room | Temp | Humid |
+---------------------+------+-------+-------+
| 2017-05-24 16:51:47 | 123 | 22.61 | 36.93 |
| 2017-05-24 16:51:51 | 123 | 22.61 | 36.94 |
| 2017-05-24 16:51:56 | 123 | 22.61 | 36.94 |
+---------------------+------+-------+-------+
我已尝试进行此查询,但它给了我一个错误:
错误1242(21000):子查询返回超过1行
请帮助我获得正确的结果。
select
m.meas_time,
m.room,
(select value from meteo m1 where m1.meas_kind='T' and m.meas_time=m1.meas_time) as Temp,
(select value from meteo m2 where meas_kind='H' and m.meas_time=m2.meas_time) as Humid
from meteo m
join meteo m1
on m.meas_time=m1.meas_time
and m.room=m1.room
join meteo m2
on m.meas_time=m2.meas_time
and m.room=m2.room
group by m.room, m.meas_time;
答案 0 :(得分:1)
你太复杂了。简单的数据透视查询可以为您提供所需的结果。
SELECT
meas_time,
room,
MAX(CASE WHEN meas_kind = 'T' THEN value END) AS Temp,
MAX(CASE WHEN meas_kind = 'H' THEN value END) AS Humid
FROM meteo
GROUP BY meas_time, room