18

我有一张表,其中包含有关参加某些活动的数据。我在表中有每次用户发送新考勤的考勤数据,信息是这样的:

mysql> SELECT id_branch_channel, id_member, attendance, timestamp, id_member FROM view_event_attendance WHERE id_event = 782;
+-------------------+-----------+------------+------------+-----------+
| id_branch_channel | id_member | attendance | timestamp  | id_member |
+-------------------+-----------+------------+------------+-----------+
|              1326 |    131327 |        459 | 1363208604 |    131327 |
|              1326 |    131327 |        123 | 1363208504 |    131327 |
|              1326 |    131327 |          1 | 1363208459 |    131327 |
|              1326 |     93086 |          0 |       NULL |     93086 |
|              1326 |     93087 |          0 |       NULL |     93087 |
|              1326 |     93088 |          0 |       NULL |     93088 |
|              1326 |     93093 |          0 |       NULL |     93093 |
|              1326 |     99113 |          0 |       NULL |     99113 |
|              1326 |     99135 |          0 |       NULL |     99135 |
|              1326 |     99199 |          0 |       NULL |     99199 |
|              1326 |     99200 |          0 |       NULL |     99200 |
|              1326 |    131324 |          0 |       NULL |    131324 |
|              1326 |     85850 |          0 |       NULL |     85850 |
|              1326 |     93085 |          0 |       NULL |     93085 |
+-------------------+-----------+------------+------------+-----------+
14 rows in set (0.00 sec)

(这实际上是一个视图,因此某些字段为空)。

我可以按 id_member 分组,因此每个成员只能获得一行(即,只有用户发送的最后一次出席)。但是,当我这样做时,我收到了用户发送的第一个出席信息,而不是最后一个。

mysql> SELECT id_branch_channel, id_member, attendance, timestamp, id_member FROM view_event_attendance WHERE id_event = 782 GROUP BY id_event,id_member;
+-------------------+-----------+------------+------------+-----------+
| id_branch_channel | id_member | attendance | timestamp  | id_member |
+-------------------+-----------+------------+------------+-----------+
|              1326 |    131327 |          1 | 1363208459 |    131327 |
|              1326 |     93086 |          0 |       NULL |     93086 |
|              1326 |    131324 |          0 |       NULL |    131324 |
|              1326 |     93087 |          0 |       NULL |     93087 |
|              1326 |     93088 |          0 |       NULL |     93088 |
|              1326 |     93093 |          0 |       NULL |     93093 |
|              1326 |     99113 |          0 |       NULL |     99113 |
|              1326 |     99135 |          0 |       NULL |     99135 |
|              1326 |     85850 |          0 |       NULL |     85850 |
|              1326 |     99199 |          0 |       NULL |     99199 |
|              1326 |     93085 |          0 |       NULL |     93085 |
|              1326 |     99200 |          0 |       NULL |     99200 |
+-------------------+-----------+------------+------------+-----------+
12 rows in set (0.00 sec)

我已经尝试添加 ORDER BY 子句,但它们根本不起作用......有什么想法吗?

提前致谢!

编辑:这是创建表的脚本

CREATE OR REPLACE VIEW view_event_attendance 
    AS
        SELECT 
            tbl_event.id_event,
            tbl_member_event.id_member,
            tbl_event.id_branch_channel,
            tbl_member_event_attendance.id_member_event_attendance,
            IF(ISNULL(tbl_member_event_attendance.attendance), 0, tbl_member_event_attendance.attendance) AS attendance,
            tbl_member_event_attendance.timestamp
        FROM 
            tbl_event
            INNER JOIN 
                tbl_member_event ON tbl_member_event.id_event = tbl_event.id_event
                LEFT OUTER JOIN
                    tbl_member_event_attendance ON tbl_member_event_attendance.id_member_event = tbl_member_event.id_member_event
        ORDER BY 
            tbl_member_event_attendance.timestamp DESC;

编辑2:

非常感谢 MichaelBenjamin,但是使用子查询时的问题是视图的大小:

mysql> DESCRIBE SELECT id_branch_channel, id_member, attendance, timestamp, id_member 
    -> FROM (select * from view_event_attendance order by timestamp desc) as whatever
    -> WHERE id_event = 782 
    -> GROUP BY id_event,id_member;
+----+-------------+-----------------------------+--------+-----------------+-----------------+---------+------------------------------------------------+-------+----------------------------------------------+
| id | select_type | table                       | type   | possible_keys   | key             | key_len | ref                                            | rows  | Extra                                        |
+----+-------------+-----------------------------+--------+-----------------+-----------------+---------+------------------------------------------------+-------+----------------------------------------------+
|  1 | PRIMARY     | <derived2>                  | ALL    | NULL            | NULL            | NULL    | NULL                                           | 16755 | Using where; Using temporary; Using filesort |
|  2 | DERIVED     | tbl_member_event            | index  | id_event        | id_event        | 8       | NULL                                           | 16346 | Using index; Using temporary; Using filesort |
|  2 | DERIVED     | tbl_event                   | eq_ref | PRIMARY         | PRIMARY         | 4       | video_staging.tbl_member_event.id_event        |     1 |                                              |
|  2 | DERIVED     | tbl_member_event_attendance | ref    | id_event_member | id_event_member | 4       | video_staging.tbl_member_event.id_member_event |     1 | Using index                                  |
+----+-------------+-----------------------------+--------+-----------------+-----------------+---------+------------------------------------------------+-------+----------------------------------------------+
4 rows in set (0.08 sec)

如您所见,我的表中有很多行,因此我不想使用子查询...

编辑 3:

但是将 WHERE 添加到子查询中看起来更好......

mysql> DESCRIBE SELECT id_branch_channel, id_member, attendance, timestamp, id_member 
    -> FROM (select * from view_event_attendance where id_event = 782 order by timestamp desc) as whatever
    -> WHERE id_event = 782 
    -> GROUP BY id_event,id_member;
+----+-------------+-----------------------------+-------+-----------------+-----------------+---------+------------------------------------------------+------+----------------------------------------------+
| id | select_type | table                       | type  | possible_keys   | key             | key_len | ref                                            | rows | Extra                                        |
+----+-------------+-----------------------------+-------+-----------------+-----------------+---------+------------------------------------------------+------+----------------------------------------------+
|  1 | PRIMARY     | <derived2>                  | ALL   | NULL            | NULL            | NULL    | NULL                                           |   14 | Using where; Using temporary; Using filesort |
|  2 | DERIVED     | tbl_event                   | const | PRIMARY         | PRIMARY         | 4       |                                                |    1 | Using temporary; Using filesort              |
|  2 | DERIVED     | tbl_member_event            | ref   | id_event        | id_event        | 4       |                                                |   12 | Using index                                  |
|  2 | DERIVED     | tbl_member_event_attendance | ref   | id_event_member | id_event_member | 4       | video_staging.tbl_member_event.id_member_event |    1 | Using index                                  |
+----+-------------+-----------------------------+-------+-----------------+-----------------+---------+------------------------------------------------+------+----------------------------------------------+
4 rows in set (0.01 sec)

如果我找不到其他不使用子查询的东西,我想我会选择这个作为答案......

编辑 4

在看到答案中的评论后,我决定选择另一个作为答案。这是两个查询的 DESCRIBE,我认为最好的解决方案是显而易见的:

mysql> DESCRIBE SELECT 
    ->   id_branch_channel,
    ->   id_member, 
    ->   attendance, 
    ->   timestamp,
    ->   id_member
    -> FROM view_event_attendance AS t1
    -> WHERE id_event = 782
    -> AND timestamp = (SELECT MAX(timestamp)
    ->                  FROM view_event_attendance AS t2 
    ->                  WHERE t1.id_member = t2.id_member 
    ->                    AND t1.id_event = t2.id_event 
    ->                  GROUP BY id_event, id_member)
    -> OR timestamp IS NULL
    -> GROUP BY id_event, id_member;
+----+--------------------+-----------------------------+--------+--------------------+--------------------------+---------+------------------------------------------------+------+-----------------------------------------------------------+
| id | select_type        | table                       | type   | possible_keys      | key                      | key_len | ref                                            | rows | Extra                                                     |
+----+--------------------+-----------------------------+--------+--------------------+--------------------------+---------+------------------------------------------------+------+-----------------------------------------------------------+
|  1 | PRIMARY            | tbl_event                   | index  | PRIMARY            | id_member_branch_channel | 4       | NULL                                           |  208 | Using index; Using temporary; Using filesort              |
|  1 | PRIMARY            | tbl_member_event            | ref    | id_event           | id_event                 | 4       | video_staging.tbl_event.id_event               |   64 | Using index                                               |
|  1 | PRIMARY            | tbl_member_event_attendance | ref    | id_event_member    | id_event_member          | 4       | video_staging.tbl_member_event.id_member_event |    1 | Using where; Using index                                  |
|  2 | DEPENDENT SUBQUERY | tbl_event                   | eq_ref | PRIMARY            | PRIMARY                  | 4       | func                                           |    1 | Using where; Using index; Using temporary; Using filesort |
|  2 | DEPENDENT SUBQUERY | tbl_member_event            | eq_ref | id_event,id_member | id_event                 | 8       | video_staging.tbl_event.id_event,func          |    1 | Using where; Using index                                  |
|  2 | DEPENDENT SUBQUERY | tbl_member_event_attendance | ref    | id_event_member    | id_event_member          | 4       | video_staging.tbl_member_event.id_member_event |    1 | Using where; Using index                                  |
+----+--------------------+-----------------------------+--------+--------------------+--------------------------+---------+------------------------------------------------+------+-----------------------------------------------------------+
6 rows in set (0.00 sec)


mysql> DESCRIBE SELECT *
    -> FROM (SELECT id_branch_channel, id_member, attendance, timestamp, id_event 
    ->       FROM view_event_attendance 
    ->       WHERE id_event = 782 
    ->       ORDER BY timestamp desc
    ->      ) as whatever
    -> GROUP BY id_event,id_member;
+----+-------------+-----------------------------+-------+-----------------+-----------------+---------+------------------------------------------------+------+---------------------------------+
| id | select_type | table                       | type  | possible_keys   | key             | key_len | ref                                            | rows | Extra                           |
+----+-------------+-----------------------------+-------+-----------------+-----------------+---------+------------------------------------------------+------+---------------------------------+
|  1 | PRIMARY     | <derived2>                  | ALL   | NULL            | NULL            | NULL    | NULL                                           |   14 | Using temporary; Using filesort |
|  2 | DERIVED     | tbl_event                   | const | PRIMARY         | PRIMARY         | 4       |                                                |    1 | Using temporary; Using filesort |
|  2 | DERIVED     | tbl_member_event            | ref   | id_event        | id_event        | 4       |                                                |   12 | Using index                     |
|  2 | DERIVED     | tbl_member_event_attendance | ref   | id_event_member | id_event_member | 4       | video_staging.tbl_member_event.id_member_event |    1 | Using index                     |
+----+-------------+-----------------------------+-------+-----------------+-----------------+---------+------------------------------------------------+------+---------------------------------+
4 rows in set (0.00 sec)
4

6 回答 6

10

使用 id_member 的简单组,但选择:

substring(max(concat(from_unixtime(timestamp),attendance)) from 20) as attendance

这会将出勤附加到组中每一行的时间戳,以便能够使用 max() 选择所需的时间戳/出勤,然后仅提取出勤。

返回的是concat()19 个字符的格式化时间戳 (YYYY-mm-dd HH:MM:SS),从字符 20 开始附加出勤;substring(... from 20)仅从该组的(字符串方式)最大出席人数中获得出席人数。您可以删除组

select concat(from_unixtime(timestamp),attendance), timestamp, attendance

更好地了解它如何使用 max 来获得正确的出勤率。

于 2013-03-14T23:29:21.370 回答
3
SELECT id_branch_channel, id_member, attendance, timestamp, id_member 
FROM (select * from view_event_attendance order by timestamp desc) as whatever
WHERE id_event = 782 
GROUP BY id_event,id_member;

编辑:这可能会产生更好的性能:

SELECT *
FROM (SELECT id_branch_channel, id_member, attendance, timestamp, id_member 
      FROM view_event_attendance 
      WHERE id_event = 782 
      ORDER BY timestamp desc
     ) as whatever
GROUP BY id_event,id_member;

只要结果集可以放入 Innodb_buffer_pool,您就不会看到显着的性能下降。

于 2013-03-14T22:56:52.810 回答
3

我看到了JOINSand的答案Subquerys,但我相信一个简单的HAVING子句应该可以解决问题:

SELECT 
  id_branch_channel,
  id_member, 
  attendance, 
  timestamp,
  id_member
FROM view_event_attendance
WHERE id_event = 782 
GROUP BY id_event, id_member
HAVING MAX(timestamp) OR timestamp IS NULL;

编辑:如果您还想包含这些行,请添加对 IS NULL 的检查。

编辑 2:当您已经将其过滤为 1 个事件时,是否甚至需要按 id_event 分组?

编辑 3:不知道为什么不赞成,这个 sql fiddle显示它有效。

编辑 4:我必须道歉,@ysth 是正确的,SQL Fiddle 无法正常工作。我应该得到-1,但是当你投反对票时,至少解释一下为什么这样我也可以自己学习一些东西。

以下工作,但不幸的是它再次有一个子查询,并且不会比此处发布的其他解决方案更好。

SELECT 
  id_branch_channel,
  id_member, 
  attendance, 
  timestamp,
  id_member
FROM view_event_attendance AS t1
WHERE id_event = 782
AND timestamp = (SELECT MAX(timestamp)
                 FROM view_event_attendance AS t2 
                 WHERE t1.id_member = t2.id_member 
                   AND t1.id_event = t2.id_event 
                 GROUP BY id_event, id_member)
OR timestamp IS NULL
GROUP BY id_event, id_member;
于 2013-03-14T23:08:45.357 回答
3
SUBSTRING_INDEX(SUBSTRING_INDEX(group_concat(%requiredfield%), ',', count(*)),',',-1)

这将从任何 group_concat 获取“必填字段”的最后一个值,如果未排序,默认情况下它将是表中的最后一个值。

可以使用 group_concat_ws 来说明可能的空字段。

于 2015-03-06T14:51:47.547 回答
1

这是一个选项(未经测试):

SELECT v.id_branch_channel, v.id_member, v.attendance, v.timestamp, v.id_member 
FROM view_event_attendance v
    JOIN (
        SELECT id_event, id_member, MAX(attendance) maxattendance
        FROM view_event_attendance 
        GROUP BY id_event, id_member ) m ON 
            v.id_event = m.id_event AND
            v.id_member = m.id_member AND
            v.attendance = m.maxattendance
WHERE v.id_event = 782 
GROUP BY v.id_member;

这个概念是获取MAX()时间戳并将该字段用于JOIN您的视图。您可能不需要所有字段——实际上取决于您的表结构。但这应该让你朝着正确的方向前进。

于 2013-03-14T22:55:34.087 回答
-1

一种方法是使用窗口函数和子查询,如果您将条目添加到选择列表,因为row_number() over (partition by id_member order by timestamp desc)这将解析为按每个 id_member 组中分组的时间戳(1 是最旧的)对行进行排序的数字(运行如果这没有意义,它会很清楚)。然后,您可以从中选择作为子查询,其中额外的列 = 1,它将仅选择每个组中具有最高时间戳的行。

于 2013-03-14T22:51:38.187 回答