我有一个以下SQL语句，它正确生成相关输出（我想每3分钟对值进行分组）：
SELECT date_time date, UNIX_TIMESTAMP(date_time) AS time_value,
FLOOR((MINUTE(date_time) + (HOUR(date_time)*60))/3) AS minute_value, ph1_active_power AS p1
FROM powerpro1
GROUP BY date_time
生成输出：
+-----------+---------------------+------------+--------------+---------+
| record_no | date | time_value | minute_value | p1 |
+-----------+---------------------+------------+--------------+---------+
| 1 | 2014-12-01 00:00:00 | 1417372200 | 0 | 73.0767 |
| 2 | 2014-12-01 00:01:00 | 1417372260 | 0 | 73.0293 |
| 3 | 2014-12-01 00:02:00 | 1417372320 | 0 | 72.9818 |
| 4 | 2014-12-01 00:03:00 | 1417372380 | 1 | 72.9343 |
| 5 | 2014-12-01 00:04:00 | 1417372440 | 1 | 72.8868 |
| 6 | 2014-12-01 00:05:00 | 1417372500 | 1 | 72.8392 |
| 7 | 2014-12-01 00:06:00 | 1417372560 | 2 | 72.7916 |
| 8 | 2014-12-01 00:07:00 | 1417372620 | 2 | 72.744 |
| 9 | 2014-12-01 00:08:00 | 1417372680 | 2 | 72.6963 |
| 10 | 2014-12-01 00:09:00 | 1417372740 | 3 | 72.6486 |
| 11 | 2014-12-01 00:10:00 | 1417372800 | 3 | 72.6009 |
| 12 | 2014-12-01 00:11:00 | 1417372860 | 3 | 72.5531 |
| 13 | 2014-12-01 00:12:00 | 1417372920 | 4 | 72.5053 |
| 14 | 2014-12-01 00:13:00 | 1417372980 | 4 | 72.4575 |
| 15 | 2014-12-01 00:14:00 | 1417373040 | 4 | 72.4096 |
| 16 | 2014-12-01 00:15:00 | 1417373100 | 5 | 72.3617 |
| 17 | 2014-12-01 00:16:00 | 1417373160 | 5 | 72.3137 |
| 18 | 2014-12-01 00:17:00 | 1417373220 | 5 | 72.2657 |
| 19 | 2014-12-01 00:18:00 | 1417373280 | 6 | 72.2177 |
| 20 | 2014-12-01 00:19:00 | 1417373340 | 6 | 72.1697 |
| 21 | 2014-12-01 00:20:00 | 1417373400 | 6 | 72.1216 |
| 22 | 2014-12-01 00:21:00 | 1417373460 | 7 | 72.0734 |
| 23 | 2014-12-01 00:22:00 | 1417373520 | 7 | 72.0253 |
| 24 | 2014-12-01 00:23:00 | 1417373580 | 7 | 71.9771 |
+-----------+---------------------+------------+--------------+---------+
但是，我想获得time_value的平均值和p1的平均值，然后需要GROUP by minute_值。如果我使用上面的查询进行相关更改，如下所示，
SELECT date_time date, AVG(UNIX_TIMESTAMP(date_time)) AS time_value, FLOOR((MINUTE(date_time) + (HOUR(date_time)*60))/3) AS minute_value, ROUND(AVG(ph1_active_power),4) AS p1
FROM powerpro1
GROUP BY minute_value
我得到了如下所述的错误输出。
+-----------+---------------------+-----------------+--------------+--------+
| record_no | date | time_value | minute_value | p1 |
+-----------+---------------------+-----------------+--------------+--------+
| 1 | 2014-12-01 00:00:00 | 1418754688.6364 | 0 | 2.2622 |
| 4 | 2014-12-01 00:03:00 | 1418754868.6364 | 1 | 2.2541 |
| 7 | 2014-12-01 00:06:00 | 1418755048.6364 | 2 | 2.246 |
| 10 | 2014-12-01 00:09:00 | 1418755228.6364 | 3 | 2.2378 |
| 13 | 2014-12-01 00:12:00 | 1418755408.6364 | 4 | 2.2297 |
| 16 | 2014-12-01 00:15:00 | 1418755588.6364 | 5 | 2.2216 |
| 19 | 2014-12-01 00:18:00 | 1418755768.6364 | 6 | 2.2134 |
| 22 | 2014-12-01 00:21:00 | 1418755948.6364 | 7 | 2.2052 |
+-----------+---------------------+-----------------+--------------+--------+
必需输出：
+-----------+---------------------+--------------+------------+---------+
| record_no | time_value | minute_value | time_value | p1 |
+-----------+---------------------+--------------+------------+---------+
| 2 | 2014-12-01 00:01:00 | 0 | 1417372260 | 73.0293 |
| 5 | 2014-12-01 00:04:00 | 1 | 1417372440 | 72.8868 |
| 8 | 2014-12-01 00:07:00 | 2 | 1417372620 | 72.744 |
| 11 | 2014-12-01 00:10:00 | 3 | 1417372800 | 72.6009 |
| 14 | 2014-12-01 00:13:00 | 4 | 1417372980 | 72.4575 |
+-----------+---------------------+--------------+------------+---------+
可能是错的。
任何人都可以帮助我利用宝贵的时间和知识。
答案 0 :(得分：1)
SELECT date_time date, SUM(UNIX_TIMESTAMP(date_time))/COUNT(record_no) AS time_value, FLOOR((MINUTE(date_time) + (HOUR(date_time)*60))/3)*3 AS minute_value, ROUND((SUM(ph1_active_power)/COUNT(record_no)),4) AS p1
FROM powerpro1
GROUP BY minute_value
我通过以下查询完成了它：
SELECT record_no, date_time,
ROUND(AVG(UNIX_TIMESTAMP(date_time))) AS time_value,
ROUND(AVG(ph1_active_power),4) AS p1
FROM powerpro1
WHERE date_time <= '2014-12-20 00:00:00'
GROUP BY date_time DIV 300