Oracle消除重復記錄的幾種方式

消除重復數據是數據清洗的頭等大事，下面介紹比較常用的幾種去重方式。

Distinct

刪除重復的員工信息

select distinct(empno) from emp

Rowid

利用rowid結合max和min函數快速去重

select e.* from emp e where e.rowid = (select max(e.rowid) from emp e 

利用rowid結合max和min函數快速刪除重復數據

delete e.* from emp e where e.rowid < (select max(e.rowid) from emp e

Group by

select deptno from emp group by deptno;

Row_number()

row_number是通過標記排號方式去重，如果有2條或以上的重復數據，直接篩選刪除即可。

1.查看重復數據

select d.id,d.outer_code from  dict_depts_source d order by outer_code

2.標識重復數據

select d.id,d.outer_code,row_number() over(partition by outer_code order by outer_code) row_flag from  dict_depts_source d

3.刪除重復數據

delete from dict_depts_source where id in(
select id from(select d.id,d.outer_code,row_number() over(partition by outer_code order by outer_code) row_flag from  dict_depts_source d)t
where t.row_flag > 1)

4.檢查刪除效果

select d.id,d.outer_code,row_number() over(partition by outer_code order by outer_code) row_flag from  dict_depts_source d