有下表：event_table，其中包含有关object_name，event_number，event_supplementary_info，event_time的信息。我想有一个SQL查询，它显示每天最后七天的事件数和总数。
我需要这样的东西
select Object_name, event_number
max(decode(trim(dow),'MONDAY',totalquantity,0)) Mon,
max(decode(trim(dow),'TUESDAY',totalquantity,0)) Tue,
max(decode(trim(dow),'WEDNESDAY',totalquantity,0)) Wed,
max(decode(trim(dow),'THURSDAY',totalquantity,0)) Thu,
max(decode(trim(dow),'FRIDAY',totalquantity,0)) Fri,
max(decode(trim(dow),'SATURDAY',totalquantity,0)) Sat,
max(decode(trim(dow),'SUNDAY',totalquantity,0)) Sun,
(
max(decode(trim(dow),'MONDAY',totalquantity,0)) +
max(decode(trim(dow),'TUESDAY',totalquantity,0)) +
max(decode(trim(dow),'WEDNESDAY',totalquantity,0)) +
max(decode(trim(dow),'THURSDAY',totalquantity,0)) +
max(decode(trim(dow),'FRIDAY',totalquantity,0)) +
max(decode(trim(dow),'SATURDAY',totalquantity,0)) +
max(decode(trim(dow),'SUNDAY',totalquantity,0))
) TOTAL
from
(
select event_name,
to_char(event_time, 'DAY') as dow,
sum(event_time) as totalquantity
from event_table a
where a.event_time >= trunc(sysdate-7,'D')
and a.tradedate <= trunc(sysdate-7,'D') + 7
group by a.event_name, alarm_time
)
group by Object_name, event_number;
答案 0 :(得分：2)
尝试使用此示例数据。下次请自己提供一些样本数据。
SQL> create table event_table (event_number,object_name,event_time,event_supplementary_info)
2 as
3 select 1, 'A', sysdate - 7, 'info' from dual union all
4 select 1, 'B', sysdate - 7, 'info' from dual union all
5 select 1, 'B', sysdate - 7, 'info' from dual union all
6 select 2, 'A', sysdate - 6, 'info' from dual union all
7 select 2, 'B', sysdate - 6, 'info' from dual union all
8 select 2, 'B', sysdate - 6, 'info' from dual union all
9 select 3, 'A', sysdate - 5, 'info' from dual union all
10 select 3, 'A', sysdate - 5, 'info' from dual union all
11 select 4, 'C', sysdate - 4, 'info' from dual union all
12 select 4, 'C', sysdate - 4, 'info' from dual union all
13 select 4, 'C', sysdate - 4, 'info' from dual union all
14 select 4, 'C', sysdate - 4, 'info' from dual union all
15 select 4, 'C', sysdate - 4, 'info' from dual union all
16 select 4, 'D', sysdate - 4, 'info' from dual union all
17 select 5, 'A', sysdate - 3, 'info' from dual union all
18 select 6, 'D', sysdate - 2, 'info' from dual union all
19 select 6, 'D', sysdate - 2, 'info' from dual union all
20 select 7, 'A', sysdate - 1, 'info' from dual union all
21 select 7, 'A', sysdate - 1, 'info' from dual union all
22 select 7, 'A', sysdate - 1, 'info' from dual union all
23 select 7, 'A', sysdate - 1, 'info' from dual
24 /
Table created.
您的查询不起作用：有多余的右括号。如果我删除它们，它仍然无效，因为您要将日期与varchar2进行比较：
SQL> select object_name,
2 event_number,
3 count(*),
4 event_supplementary_info
5 from event_table
6 where event_time between to_char(sysdate -7, 'YYYY-MM-DD')
7 and to_char(sysdate , 'YYYY-MM-DD')
8 group by object_name, event_number, event_supplementary_info
9 /
no rows selected
所以我的基本查询就是这个，我将日期与日期进行比较：
SQL> select object_name
2 , event_number
3 , count(*)
4 , event_supplementary_info
5 from event_table
6 where event_time between sysdate -7 and sysdate
7 group by object_name
8 , event_number
9 , event_supplementary_info
10 order by object_name
11 , event_number
12 /
O EVENT_NUMBER COUNT(*) EVEN
- ------------ ---------- ----
A 1 1 info
A 2 1 info
A 3 2 info
A 5 1 info
A 7 4 info
B 1 2 info
B 2 2 info
C 4 5 info
D 4 1 info
D 6 2 info
10 rows selected.
我将您的问题解释为您希望记录每个object_name的事件总数。对于此示例数据，对于对象A，B，C和D，您需要4个额外记录。为此，我在object_name上添加了另一个分组集。为了清楚起见，我将trunc（event_time）包含在现有的分组集中。
SQL> select object_name
2 , event_number
3 , count(*)
4 , event_supplementary_info
5 , trunc(event_time)
6 from event_table
7 where event_time between sysdate -7 and sysdate
8 group by grouping sets
9 ( ( object_name
10 , event_number
11 , event_supplementary_info
12 , trunc(event_time)
13 )
14 , ( object_name )
15 )
16 order by object_name
17 , event_number
18 /
O EVENT_NUMBER COUNT(*) EVEN TRUNC(EVENT_TIME)
- ------------ ---------- ---- -------------------
A 1 1 info 15-08-2010 00:00:00
A 2 1 info 16-08-2010 00:00:00
A 3 2 info 17-08-2010 00:00:00
A 5 1 info 19-08-2010 00:00:00
A 7 4 info 21-08-2010 00:00:00
A 9
B 1 2 info 15-08-2010 00:00:00
B 2 2 info 16-08-2010 00:00:00
B 4
C 4 5 info 18-08-2010 00:00:00
C 5
D 4 1 info 18-08-2010 00:00:00
D 6 2 info 20-08-2010 00:00:00
D 3
14 rows selected.
编写此类查询的简短方法是将分组集转换为汇总：
SQL> select object_name
2 , event_number
3 , count(*)
4 , event_supplementary_info
5 , trunc(event_time)
6 from event_table
7 where event_time between sysdate -7 and sysdate
8 group by object_name
9 , rollup ((event_number,event_supplementary_info,trunc(event_time)))
10 order by object_name
11 , event_number
12 /
O EVENT_NUMBER COUNT(*) EVEN TRUNC(EVENT_TIME)
- ------------ ---------- ---- -------------------
A 1 1 info 15-08-2010 00:00:00
A 2 1 info 16-08-2010 00:00:00
A 3 2 info 17-08-2010 00:00:00
A 5 1 info 19-08-2010 00:00:00
A 7 4 info 21-08-2010 00:00:00
A 9
B 1 2 info 15-08-2010 00:00:00
B 2 2 info 16-08-2010 00:00:00
B 4
C 4 5 info 18-08-2010 00:00:00
C 5
D 4 1 info 18-08-2010 00:00:00
D 6 2 info 20-08-2010 00:00:00
D 3
14 rows selected.
此致 罗布。
答案 1 :(得分：0)
或许这就是你要找的东西：http://psoug.org/reference/rollup.html？