我有简单的postgresql查询,在最近的日期之前进行查询事件,但是我想从结果中删除这个不同的列(时间)
SELECT DISTINCT time, sensor_id, event_type, value from events ORDER BY sensor_id
+---------------------+---+---+-----+
| 2014-02-13 12:42:00 | 2 | 2 | 5 |
| 2014-02-13 13:19:57 | 2 | 4 | -42 |
| 2014-02-13 13:32:36 | 2 | 3 | 54 |
| 2014-02-13 14:48:30 | 2 | 2 | 2 |
| 2014-02-13 12:54:39 | 3 | 2 | 7 |
+---------------------+---+---+-----+
需要这样的结果
+---+---+-----+
| 2 | 2 | 5 |
| 2 | 4 | -42 |
| 2 | 2 | 2 |
| 2 | 3 | 54 |
| 3 | 2 | 7 |
+---+---+-----+
您可以使用PostgreSQL DISTINCT ON (...)
功能:
SELECT DISTINCT ON (time, sensor_id, event_type, value)
sensor_id, event_type, value from events
ORDER BY sensor_id;
我认为您的意思是:
SELECT sensor_id, event_type, value
from (
SELECT DISTINCT time, sensor_id, event_type, value from events
) A
ORDER BY sensor_id
您可以使用汇总函数按时间的最大值订购。
SELECT sensor_id, event_type, value
FROM events
GROUP BY sensor_id, event_type, value
ORDER BY MAX(time) DESC
使用CTE和等级窗口函数
with temp as
( select time,sensor_id, event_type, value , rank() OVER (PARTITION BY ensor_id, event_type, value order by 'time') as rnk
)
select time,sensor_id, event_type, value from temp where rnk =1