所以我选择每15分钟将数据添加到数据库中的最新x小时范围。我想通过平均过去那天当时添加的所有数据，在这个数据预测中添加下一个x小时的计数。
以下是我正在使用的两个表的示例。我无法在计数表中提供足够的数据来表示我想要做的事情。
NAME Table
+---------+-------+
| NAME_ID | NAME |
+---------+-------+
| 1 | name1 |
+---------+-------+
COUNT Table
+----+---------+-------+---------------------+
| ID | NAME_ID | COUNT | DTM |
+----+---------+-------+---------------------+
| 1 | 1 | 4 | 2015-03-08 23:45:00 |
+----+---------+-------+---------------------+
| 2 | 1 | 5 | 2015-03-09 00:00:00 |
+----+---------+-------+---------------------+
| 3 | 1 | 6 | 2015-03-09 00:15:00 |
+----+---------+-------+---------------------+
| 4 | 1 | 2 | 2015-03-09 00:30:00 |
+----+---------+-------+---------------------+
| 5 | 1 | 5 | 2015-03-09 00:45:00 |
+----+---------+-------+---------------------+
| 6 | 1 | 6 | 2015-03-09 01:00:00 |
+----+---------+-------+---------------------+
| 7 | 1 | 6 | 2015-03-09 01:15:00 |
+----+---------+-------+---------------------+
| 8 | 1 | 3 | 2015-03-09 01:30:00 |
+----+---------+-------+---------------------+
以下是我想要的结果示例。
+----+-------+-------+---------------------------------------------------------------------------+
| ID | NAME | COUNT | DTM |
+----+-------+-------+---------------------------------------------------------------------------+
| 1 | name1 | 5 | 2015-03-09 00:45:00 <-----Past Data |
+----+-------+-------+---------------------------------------------------------------------------+
| 2 | name1 | 6 | 2015-03-09 01:00:00 <-----Past Data |
+----+-------+-------+---------------------------------------------------------------------------+
| 3 | name1 | 6 | 2015-03-09 01:15:00 <-----Past Data |
+----+-------+-------+---------------------------------------------------------------------------+
| 4 | name1 | 3 | 2015-03-09 01:30:00 <-----CURRENT TIME |
+----+-------+-------+---------------------------------------------------------------------------+
| 5 | name1 | 4 | 2015-03-09 01:45:00 <-----Average count for all counts with matching day of the week and time |
+----+-------+-------+---------------------------------------------------------------------------+
| 6 | name1 | 5 | 2015-03-09 02:00:00 <-----Average count for all counts with matching day of the week and time|
+----+-------+-------+---------------------------------------------------------------------------+
| 7 | name1 | 6 | 2015-03-09 02:15:00 <-----Average count for all counts with matching day of the week and time|
+----+-------+-------+---------------------------------------------------------------------------+
| 8 | name1 | 2 | 2015-03-09 02:30:00 <-----Average count for all counts with matching day of the week and time|
+----+-------+-------+---------------------------------------------------------------------------+
这是我目前的尝试，但没有得到我需要的结果。
SELECT * FROM (
select c.count, c.dtm
from count c
join name n on c.name_id = n.name_id
where NAME Like @name
union
SELECT AVG(c.count), c.dtm
from count c
join name n on c.name_id = n.name_id
where (n.NAME Like @name)
AND (c.DTM < NOW())
LIMIT 20
) a
ORDER BY DTM DESC