rollup 和 cube 的区别

select  列1,sum(计算列) from  table_name t group by  rollup(列1,列2..列n);
会返回 【列1】 【列1，列2】【列1，列2，列3】...【列1，列2，列3...列n】的每种排列的sum(计算列)  和 全部记录的sum(计算列)  ，一共是n+1种排列

select  列1,...列n,sum(计算列) from  table_name t group by  cube(列1,列2...列n);
会返回 【列1】，【列2】，【列3】，【列4】...【列n】，【列1，列2】，【列1，列3】，【列1，列4】...【列1，列n】，【列2,列3】，【列2，列4】...【列2，列n】...【列n-1,列n】 
每个分组的sum(计算列)  和 全部记录的sum(计算列)，一共是 C 0 n+C 1 n+C 2 n+...C n n种排列。
如下演示（因为多条数据会比较杂乱，不易说明，所以我们选择极端情况，就认为表里只有一条记录）：
--group by
select flag1, flag2, flag3, flag4, sum(flag5)

  from (select 'column1' as flag1,

               'column2' as flag2,

               'column3' as flag3,

               'column4' as flag4,

               1 flag5

          from dual)

 group by flag1, flag2, flag3, flag4；

--group by cobe
select flag1, flag2, flag3, flag4, sum(flag5)

  from (select 'column1' as flag1,

               'column2' as flag2,

               'column3' as flag3,

               'column4' as flag4,

               1 flag5

          from dual)

 group by cube(flag1, flag2, flag3, flag4)；

--group by rollup

select flag1, flag2, flag3, flag4, sum(flag5)

  from (select 'column1' as flag1,

               'column2' as flag2,

               'column3' as flag3,

               'column4' as flag4,

               1 flag5

          from dual)

 group by rollup(flag1, flag2, flag3, flag4)；

rollup的分组排列种类为 n+1 ，上述例子group by 为4列，所以 是 4+1 种情况

cube的分组排列种类是 C 0 n+C 1 n+C 2 n+...C n n 上述例子group by 为4列，所以是C 0 4+C 1 4+C 2 4+C 3 4+C 4 4=1+4+6+4+1=16种情况

上面两种计算都是计算排列种类，并非是行数，行数由具体被分组的该列的值所确定的，这个虽然可以计算，但是比较麻烦也无意义，所以我的理解cube和rollup其实相当于对group 的多列字段进行排列，然后将每种排列进行展示或者进行聚合计算。