SQL:累计求和的例子

例一、譬如实现如下结果

月份(month) 数量(number)   累计数量(total)

2015-03               1                   1

2015-04               2                   3

2015-05               3                   7

2015-03               1                   1

2015-04               2                   3

2015-05               3                   7

-------------------------------------------------------

sql示例如下：

select

 b.month,

 b.number,

 sum(a.month）

from TABLE a

join TABLE b

on a.month=b.month

group by b.month,b.number

小结：其实实现累计的核心就是将相同的表在做一次关联，让里面的值作为比较参数。

 

 

 

例二、有以下表

日期 增加 减少

2015-04-24 18709 12608

2015-04-25 28508 11412

2015-04-26 39092 20858

2015-04-27 80146 57995

2015-04-28 53581 19584

2015-04-29 50609 26319

2015-04-30 52969 28884

2015-05-01 79146 70007

2015-05-02 50536 38031

2015-05-03 58161 42329

2015-05-04 27287 14137

我需要用SQL语句得到下面的东西:

日期 增加 减少 留存率

2015-04-24 18709 12608 =(18709-12608)

2015-04-25 28508 11412 =(18709-12608)+(28508-11412)

2015-04-26 39092 20858 =(18709-12608)+(28508-11412)+( 39092-20858)

2015-04-27 80146 57995 下面依次类推

2015-04-28 53581 19584 …

2015-04-29 50609 26319 …

2015-04-30 52969 28884 …

2015-05-01 79146 70007 …

2015-05-02 50536 38031 …

2015-05-03 58161 42329 …

2015-05-04 27287 14137 …

非递归的写法（应该所有的数据库都支持）

SELECT t1.日期, t1.增加, t1.减少,
           SUM(t2.增加-t2.减少) 留存率
      FROM table1 t1
      JOIN table1 t2
        ON t2.日期 <= t1.日期
  GROUP BY t1.日期, t1.增加, t1.减少

 递归的写法（SQL Server）

WITH t0 AS (
    SELECT *,
           ROW_NUMBER() OVER(ORDER BY 日期) rn
      FROM table1
)
,t AS (
    SELECT rn,
           日期, 增加, 减少,
           (增加-减少) 留存率
      FROM t0
     WHERE rn = 1

    UNION ALL

    SELECT t2.rn,
           t2.日期, t2.增加, t2.减少,
           t1.留存率 + (t2.增加-t2.减少) 留存率
      FROM t t1
      JOIN t0 t2
        ON t2.rn = t1.rn + 1
)
SELECT 日期, 增加, 减少, 留存率
  FROM t

 

 

 

例三（自己写的）：

create table #p  
(  
    id int,  
    year varchar(4),  
    month varchar(2),  
    qty int  
)  
   
insert into #p values (1,'2012','1',10);  
insert into #p values (2,'2012','2',15);  
insert into #p values (3,'2012','3',20);  
insert into #p values (4,'2013','5',30);  
insert into #p values (5,'2013','6',35);  
insert into #p values (6,'2013','7',40);  
insert into #p values (7,'2013','8',45)  
insert into #p values (8,'2013','9',50)  
insert into #p values (9,'2013','12',100)  
insert into #p values (10,'2014','1',10);  
insert into #p values (11,'2014','3',15);  
insert into #p values (12,'2014','4',20);  
insert into #p values (13,'2014','5',30);  
insert into #p values (14,'2014','7',40);  
insert into #p values (15,'2014','8',45);  
insert into #p values (16,'2014','9',50);  
insert into #p values (17,'2015','5',30);  
insert into #p values (18,'2016','7',40);  
insert into #p values (19,'2017','8',45);  
insert into #p values (120,'2017','9',50);  
  
select * from #p  

select t1.year,t1.month,SUM(t2.qty) from #p t1 join #p t2 on t1.year=t2.year and cast(t2.month as int)<=cast(t1.month as int) group by t1.year,t1.month order by cast(t1.year as int),cast(t1.month as int)