Oracle模式SQL基础知识

--创建用于测试的表及数据
create table t603
(
  code  varchar2(10),
  year  varchar2(10),
  month varchar2(10),
  p_id  varchar2(7),
  v1    number(10),
  v2    number(10)
);
comment on column t603.code is '工厂';
comment on column t603.year is '年份';
comment on column t603.month is '月份';
comment on column t603.p_id is '产品';
comment on column t603.v1 is '生产量';
comment on column t603.v2 is '销售量';
 
insert into t603 values('600001',2012,11,'30',1,1);
insert into t603 values('600001',2012,11,'31',1,1);
insert into t603 values('600001',2012,11,'10',2,2);
insert into t603 values('600002',2012,12,'10',3,2);
insert into t603 values('600002',2012,12,'31',2,1);
insert into t603 values('600002',2012,12,'30',2,1);
 
select * from t603;

 

--按自定义的规则聚合

 

select code,
       p_id,
       v1 
  from t603
 where code in ('600001', '600002') 
       model return updated rows --MODEL后面的RETURN UPDATED ROWS子句将结果限制为在该查询中创建或更新的那些行，使结果集只包含新计算的值。
       partition by(code)        --分区
       dimension by(p_id)        --维度
       measures(v1)              --度量
       rules (v1['err1'] = v1['30'] + v1['31'] - v1['10']) --计算规则
 order by code, p_id;

  

输出   CODEP_IDV1

1600001err10

2600002err11

 

 

 

--按多个规则聚合，返回包括非聚合行

 

select code,
       year,
       p_id, 
       month,
       v1
  from t603
 where year = '2012'
       model
       partition by (code,year)
       dimension by (p_id,month)
       measures (v1,v2)
       rules (
             v1['err1',11] = v1['30',11] + v1['31',11] - v1['10',11],
             v2['err1',12] = v2['30',12] + v2['31',12] - v2['10',12])
 order by code, p_id;

  

输出   CODEYEARP_IDMONTHV1

1600001201210112

2600001201230111

3600001201231111

46000012012err112

56000012012err1110

6600002201210123

7600002201230122

8600002201231122

96000022012err112

106000022012err111

 

 

 

--指定要聚合的维值

 

select code,
       p_id, 
       month,
       v1
  from t603
 where year = '2012'
       model return updated rows
       partition by (code)
       dimension by (p_id,month)
       measures (v1)
       rules (
             v1['err1',for month in(11,12)] = v1['30',cv(month)] + v1['31',cv(month)] -v1['10',cv(month)])
 order by code, p_id, month;

  

输出  CODEP_IDMONTHV1

1600001err1110

2600001err112

3600002err111

4600002err1121

 

--CV()函数获取多行统计记录的当前值，可以使用表达式进行计算

--为数值型可用语法：for month from 1 to 12 increment 1

--还可以使用子查询：for year in (select year from t603)