摘自 http://blog.itpub.net/26977915/viewspace-734114/
在报表语句中经常要使用各种分组汇总,rollup和cube就是常用的分组汇总方式。
第一:group by rollup
1、如果使用诸如group by rollup(A,B,C)的方式分组,那么返回的分组结果是
(A,B,C) (A,B) (A) (NULL) 一共四种结果。即从右到左递减,最后来个合计。
例如:
SQL> select * from t;
YEARS MONTHS PRODUCT_NA SALES
---------- ---------- ---------- ----------
2008 1 A 1000
2008 1 B 1500
2008 2 A 2000
2008 2 B 3000
2008 2 C 1000
2008 3 A 3000
已选择6行。
SQL> select years,months,product_name,sum(sales) sum_sales from t
2 group by rollup(years,months,product_name)
3 /
YEARS MONTHS PRODUCT_NA SUM_SALES
---------- ---------- ---------- ----------
2008 1 A 1000 ----------group by (years,months,product_name)
2008 1 B 1500
2008 1 2500 ----------group by (years,months)
2008 2 A 2000
2008 2 B 3000
2008 2 C 1000
2008 2 6000 ----------group by (years,months)
2008 3 A 3000
2008 3 3000 ----------group by (years,months)
2008 11500 ----------group by (years)
11500 ----------group by (NULL)
已选择11行。
2、如果使用诸如group by A,ROLLUP(B,C) 那么返回的分组方式是:
(A,B,C) (A,B) (A,NULL) 及在这种情况下,先计算rollup里面的分组情况,再与A组合。
例如:
SQL> select years,months,product_name,sum(sales) sum_sales from t
2 group by years,rollup(months,product_name)
3 /
YEARS MONTHS PRODUCT_NA SUM_SALES
---------- ---------- ---------- ----------
2008 1 A 1000 ----------group by (years,months,product_name)
2008 1 B 1500
2008 1 2500 ----------group by (years,months)
2008 2 A 2000
2008 2 B 3000
2008 2 C 1000
2008 2 6000
2008 3 A 3000
2008 3 3000
2008 11500 ----------group by (years)
已选择10行。
第二:group by cube
1、如果使用诸如cube(A,B,C)的方式,那么返回的分组组合是
(A) (A,B) (A,C) (A,B,C) (B) (B,C) (C) (null) 共8种组合方式
例如:
SQL> select years,months,product_name,sum(sales) sum_sales from t
2 group by cube(years,months,product_name)
3 /
YEARS MONTHS PRODUCT_NA SUM_SALES
---------- ---------- ---------- ----------
11500 ----------group by (null)
A 6000 ----------group by (product_name)
B 4500
C 1000
1 2500
1 A 1000
1 B 1500
2 6000
2 A 2000
2 B 3000
2 C 1000
3 3000 ----------group by (months)
3 A 3000 ----------group by (months,product_name)
2008 11500 ----------group by (years)
2008 A 6000
2008 B 4500
2008 C 1000 ----------group by (years,product_name)
2008 1 2500
2008 1 A 1000
2008 1 B 1500
2008 2 6000
2008 2 A 2000
2008 2 B 3000
2008 2 C 1000
2008 3 3000 ----------group by (years,months)
2008 3 A 3000 ----------group by (years,months,product_name)
已选择26行。
2、如果使用GROUP BY A,CUBE(B,C),那么返回的分组组合为:
(A,B) (A,B,C) (A,C) (A)
例如:
SQL> select years,months,product_name,sum(sales) sum_sales from t
2 group by years,cube(months,product_name)
3 /
YEARS MONTHS PRODUCT_NA SUM_SALES
---------- ---------- ---------- ----------
2008 11500 ----------group by (years)
2008 A 6000 ----------group by (years,product_name)
2008 B 4500
2008 C 1000
2008 1 2500 ----------group by (years,months)
2008 1 A 1000 ----------group by (years,months,product_name)
2008 1 B 1500
2008 2 6000
2008 2 A 2000
2008 2 B 3000
2008 2 C 1000
2008 3 3000
2008 3 A 3000
已选择13行。
3、如果使用GROUP BY A,ROLLUP(B,C),CUBE(D,E),那么返回的分组组合为:
先分解cube:
a,rollup(b,c),d,e
a,rollup(b,c),d
a,rollup(b,c),e
a,rollup(b,c)
再分解ROLLUP而得到最终所有情况为:
a,b,c,d,e
a,b,d,e
a,d,e
a,b,c,d
a,b,d
a,d
a,b,c,e
a,b,e
a,e
a,b,c
a,b
a
例如:
SQL> select years,months,product_name,sum(sales) sum_sales from t
2 group by years,rollup(months),cube(product_name)
3 /
YEARS MONTHS PRODUCT_NA SUM_SALES
---------- ---------- ---------- ----------
2008 1 A 1000 ----------group by (years,months,product_name)
2008 2 A 2000
2008 3 A 3000
2008 1 B 1500
2008 2 B 3000
2008 2 C 1000
2008 A 6000 ----------group by (years,product_name)
2008 B 4500
2008 C 1000
2008 1 2500 ----------group by (years,product_name)
2008 2 6000
2008 3 3000
2008 11500 ----------group by (years)
已选择13行。
第三:grouping sets
如果使用group by A,grouping sets(B,C) 那么相当于group by A,B UNION ALL group by A,C
例如:
SQL> select years,months,product_name,sum(sales) sum_sales from t
2 group by years,grouping sets(months,product_name)
3 /
YEARS MONTHS PRODUCT_NA SUM_SALES
---------- ---------- ---------- ----------
2008 2 6000 ----------group by (years,months)
2008 1 2500
2008 3 3000
2008 B 4500 ----------group by (years,product_name)
2008 C 1000
2008 A 6000
已选择6行。
-----------------------------------------------------------------------------------------------------------------华丽的分割线!
现实中可能希望出现小计、合计等字样的报表,那么可以使用grouping函数来达到美化的效果!
第三:grouping(exp),当没有对exp分组汇总时,便返回1;
例如:
SQL> select months,product_name,sum(sales) sum_sales,grouping(product_name) from t
2 group by rollup(months,product_name)
3 /
MONTHS PRODUCT_NA SUM_SALES GROUPING(PRODUCT_NAME)
---------- ---------- ---------- ----------------------
1 A 1000 0 ----------group by (months,product_name)
1 B 1500 0
1 2500 1 ----------group by (months)
2 A 2000 0
2 B 3000 0
2 C 1000 0
2 6000 1 ----------group by (months)
3 A 3000 0
3 3000 1 ----------group by (months)
11500 1 ----------group by (null)
已选择10行。
第四:GROUPING_ID(exp1,exp2,…,expN)={GROUPING(exp1)||GROUPING(exp2)||…||GROUPING(expN)}变成十进制数,如:
如果GROUPING(A)=1,GROUPING(B)=0,GROUPING(C)=1,那么
GROUPING_ID(A,B,C) = [101]二进制 = 5,
GROUPING_ID(B,A,C) = [011]二进制 = 3.
例如:
SQL> select years,months,product_name,sum(sales) sum_sales,grouping_id(years,months,product_name) g_id from t
2 group by rollup(years,months,product_name)
3 /
YEARS MONTHS PRODUCT_NA SUM_SALES G_ID
---------- ---------- ---------- ---------- ----------
2008 1 A 1000 0
2008 1 B 1500 0
2008 1 2500 1 ----------group by (years,months) 001=1
2008 2 A 2000 0
2008 2 B 3000 0
2008 2 C 1000 0
2008 2 6000 1
2008 3 A 3000 0
2008 3 3000 1
2008 11500 3 ----------group by (years) 011=3
11500 7 ----------group by (null) 111=7
已选择11行。
了解了grouping和grouping_id函数后,便可以结合decode函数来生成小计合计的效果了;
SQL> select decode(grouping(months)+grouping(product_name),1,'月份小计',2,'合计:',months) months,
2 product_name,sum(sales) sum_sales from t
3 group by rollup(months,product_name)
4 /
MONTHS PRODUCT_NA SUM_SALES
---------------------------------------- ---------- ----------
1 A 1000
1 B 1500
月份小计 2500
2 A 2000
2 B 3000
2 C 1000
月份小计 6000
3 A 3000
月份小计 3000
合计: 11500
已选择10行。
SQL> select decode(grouping_id(months,product_name),1,'月份小计:',2,'产品小计:',3,'合计:',months) months,
2 product_name,sum(sales) sum_sales from t
3 group by cube(months,product_name)
4 order by 2
5 /
MONTHS PRODUCT_NA SUM_SALES
---------------------------------------- ---------- ----------
1 A 1000
2 A 2000
3 A 3000
产品小计: A 6000
1 B 1500
2 B 3000
产品小计: B 4500
2 C 1000
产品小计: C 1000
月份小计: 2500
月份小计: 6000
月份小计: 3000
合计: 11500
已选择13行。
点评:group by rollup、group by cube、grouping sets、grouping函数、grouping_id函数这些属于报表常用函数,要灵活运用!