来源:http://www.cnblogs.com/linjiqin/archive/2012/04/06/2434806.html
分析函数是什么?
分析函数是Oracle专门用于解决复杂报表统计需求的功能强大的函数,它可以在数据中进行分组然后计算基于组的某种统计值,并且每一组的每一行都可以返回一个统计值。
分析函数和聚合函数的不同之处是什么?
普通的聚合函数用group by分组,每个分组返回一个统计值,而分析函数采用partition by分组,并且每组每行都可以返回一个统计值。
分析函数的形式
分析函数带有一个开窗函数over(),包含三个分析子句:分组(partition by), 排序(order by), 窗口(rows) ,他们的使用形式如下:over(partition by xxx order by yyy rows between zzz)。
-----first_value()与last_value():求最值对应的其他属性
FIRST_VALUE(AREA_CODE) OVER(PARTITION BY BILL_MONTH ORDER BY SUM(LOCAL_FARE) DESC
ROWS BETWEEN UNBOUNDED PRECEDING AND UNBOUNDED FOLLOWING) FIRSTVAL,
LAST_VALUE(AREA_CODE) OVER(PARTITION BY BILL_MONTH ORDER BY SUM(LOCAL_FARE) DESC
ROWS BETWEEN UNBOUNDED PRECEDING AND UNBOUNDED FOLLOWING) LASTVAL
-----rank(),dense_rank()与row_number():求排序
rank()值相同时排名相同,其后排名跳跃不连续
RANK() OVER(PARTITION BY DEPTNO ORDER BY SAL DESC) RW
dense_rank()值相同时排名相同,其后排名连续不跳跃
DENSE_RANK() OVER(PARTITION BY DEPTNO ORDER BY SAL DESC) RW
row_number()值相同时排名不相等,其后排名连续不跳跃
ROW_NUMBER() OVER(PARTITION BY DEPTNO ORDER BY SAL DESC) RW
-----------lag()与lead():求之前或之后的第N行
lag和lead函数可以在一次查询中取出同一字段的前n行的数据和后n行的值。这种操作可以使用对相同表的表连接来实现,不过使用lag和lead有更高的效率。
lag(arg1,arg2,arg3)
第一个参数是列名,
第二个参数是偏移的offset,
第三个参数是超出记录窗口时的默认值。
举例如下:
SQL> select * from kkk;
ID NAME
---------- --------------------
1 1name
2 2name
3 3name
4 4name
5 5name
SQL> select id,name,lag(name,1,0) over(order by id) from kkk;
ID NAME LAG(NAME,1,0)OVER(ORDERBYID)
---------- -------------------- ----------------------------
1 1name 0
2 2name 1name
3 3name 2name
4 4name 3name
5 5name 4name
SQL> select id,name,lead(name,1,0) over(order by id) from kkk;
ID NAME LEAD(NAME,1,0)OVER(ORDERBYID)
---------- -------------------- -----------------------------
1 1name 2name
2 2name 3name
3 3name 4name
4 4name 5name
5 5name 0
SQL> select id,name,lead(name,2,0) over(order by id) from kkk;
ID NAME LEAD(NAME,2,0)OVER(ORDERBYID)
---------- -------------------- -----------------------------
1 1name 3name
2 2name 4name
3 3name 5name
4 4name 0
5 5name 0
SQL> select id,name,lead(name,1,'linjiqin') over(order by id) from kkk;
ID NAME LEAD(NAME,1,'ALSDFJLASDJFSAF')
---------- -------------------- ------------------------------
1 1name 2name
2 2name 3name
3 3name 4name
4 4name 5name
5 5name linjiqin
-------------rollup()与cube():排列组合分组
1)、group by rollup(a, b, c):
首先会对(a、b、c)进行group by,
然后再对(a、b)进行group by,
其后再对(a)进行group by,
最后对全表进行汇总操作。
2)、group by cube(a, b, c):
则首先会对(a、b、c)进行group by,
然后依次是(a、b),(a、c),(a),(b、c),(b),(c),
最后对全表进行汇总操作。
-----------max(),min(),sum()与avg():求移动的最值总和与平均值
SUM(LOCAL_FARE) OVER(PARTITION BY AREA_CODE ORDER BY TO_NUMBER(BILL_MONTH) RANGE BETWEEN 1 PRECEDING AND 1 FOLLOWING)
AVG(LOCAL_FARE) OVER(PARTITION BY AREA_CODE ORDER BY TO_NUMBER(BILL_MONTH) RANGE BETWEEN 1 PRECEDING AND 1 FOLLOWING)
MAX(LOCAL_FARE) OVER(PARTITION BY AREA_CODE ORDER BY TO_NUMBER(BILL_MONTH) RANGE BETWEEN 1 PRECEDING AND 1 FOLLOWING)
MIN(LOCAL_FARE) OVER(PARTITION BY AREA_CODE ORDER BY TO_NUMBER(BILL_MONTH) RANGE BETWEEN 1 PRECEDING AND 1 FOLLOWING)
----unbounded preceding and unbouned following针对当前所有记录的前一条、后一条记录,也就是表中的所有记录
--unbounded:不受控制的,无限的
--preceding:在...之前
--following:在...之后
----ROWS BETWEEN unbounded preceding AND current row 是指第一行至当前行的汇总
----ROWS BETWEEN current row AND unbounded following 指当前行到最后一行的汇总
----ROWS BETWEEN 1 preceding AND current row 是指当前行的上一行(rownum-1)到当前行的汇总
----ROWS BETWEEN 1 preceding AND 1 following 是指当前行的上一行(rownum-1)到当前行的下辆行(rownum+2)的汇总
----------累加
SELECT AREA_CODE, BILL_MONTH, LOCAL_FARE, SUM(LOCAL_FARE) OVER(PARTITION BY AREA_CODE ORDER BY BILL_MONTH ASC) "last_sum_value" FROM (SELECT T.AREA_CODE, T.BILL_MONTH, SUM(T.LOCAL_FARE) LOCAL_FARE FROM T GROUP BY T.AREA_CODE, T.BILL_MONTH) ORDER BY AREA_CODE, BILL_MONTH
------------排序
ROW_NUMBER() over (PARTITION byhexuncookieid order by cycookied) order
-------------无需排序给行号
select row_number() over () as rowid, cust_idfrom dms.fund_trans_log2 limit 10;