個人認為，這個在財務計算中肯定有用：
A),
select tc_orc002, tc_orc017, avg(tc_orc018)
from ds4.tc_orc_file
where tc_orc016 > sysdate - 2
and tc_orc003 is not null
group by rollup(tc_orc002, tc_orc017)；
TN70BS228B 2MN70-61B 0.01
TN70BS228B 0.01
PU90UHS030G 2MU90-04G 1
PU90UHS030G 1
VN60V-060SB 2MN60-01B 0.01
VN60V-060SB 0.01
PU90UHS022.4G 2MU90-04G 1
PU90UHS022.4G 1
RE70J-S018B0100 2ME70-01B 0.09
RE70J-S018B0100 0.09
RN70J-G210B0100 2MN70-01B 28.04
RN70J-G210B0100 28.04
分組小計，說白了，就是把平均值給出後，把平均的結果中平均一遍，最後再把每條記錄求和後除以總數；
B):
select tc_orc002, tc_orc017, SUM(tc_orc018),AVG(tc_orc018)
from ds4.tc_orc_file
where tc_orc016 > sysdate -0.5
and tc_orc003 is not null AND TC_ORC017 LIKE'2MN%'
group by cube(tc_orc002, tc_orc017)
個人感覺就是在rollup的基礎上，再把每類統計一下：
28.22 2.56545454545455
2MN50-01B 0.25 0.25
2MN70-01B 25.2 3.15
2MN90-01B 2.77 1.385
RN90J-G085B0100 1.94 1.94
RN90J-G085B0100 2MN90-01B 1.94 1.94
RN90J-S040B0100 0.83 0.83
RN90J-S040B0100 2MN90-01B 0.83 0.83
RN70AS568-129B0100 0.53 0.53
RN70AS568-129B0100 2MN70-01B 0.53 0.53
XN70AS568-312B0100 11.11 11.11
XN70AS568-312B0100 2MN70-01B 11.11 11.11
QN70M01002005B0100A 0 0
QN70M01002005B0100A 2MN70-01B 0 0
RN50M008000350B0100 0.25 0.25
RN50M008000350B0100 2MN50-01B 0.25 0.25
RN70M003200120B0100 0.03 0.03
RN70M003200120B0100 2MN70-01B 0.03 0.03
C):Grouping:
在統計的時候是否用到了特定列：
select tc_orc002, tc_orc017, SUM(tc_orc018),AVG(tc_orc018),grouping(tc_orc017)
from ds4.tc_orc_file
where tc_orc016 > sysdate -0.5
and tc_orc003 is not null AND TC_ORC017 LIKE'2MN%'
group by cube(tc_orc002, tc_orc017);
D):GROUPING SETS:
SELECT deptno,job,avg(sal), FROM emp
group by grouping sets(deptno,job);
這個是把 部門的平均工資，和崗位工資 合起來而已。
deptno job avg(sal)
-----------------------------------------------
10 2916.6
20 2175
30 1566.7
analyst 3000
cleark 1037.5
manager 2758.33
president 5000
salesman 1400