Sql 删除保留最新记录的重复记录

在一个表中，我有两列，IPaddress和date。存在ip地址重复值。但日期是独一无二的。我需要删除所有重复的IP，保留最新的IP

假设您没有空值并且分配了主键，请尝试此查询

  DELETE FROM MyTable
LEFT OUTER JOIN (
   SELECT MIN(RowId) as RowId, ipaddress, date
   FROM MyTable 
   GROUP BY ipaddress, date
) as KeepRows ON
   MyTable.RowId = KeepRows.RowId
WHERE
   KeepRows.RowId IS NULL

根据您的要求进行编辑

首先，需要将标识设置到表中，然后使用此用户设置标识

 ALTER TABLE [tablename] ADD Id INT IDENTITY(1,1)  

一旦设置了标识，请执行以下查询以删除重复的IP地址

DELETE FROM dbo.DataTime WHERE ColumnID NOT IN (SELECT MIN(ColumnID) _
FROM dbo.DataTime group by (ipaddress))

---

如果使用mysql，您可以尝试

DELETE JOIN

：

delete t1
from yourtable t1
join yourtable t2
on t1.ip = t1.ip
and t1.`date` < t2.`date`

删除t1
从您的表t1
加入您的表t2
在t1.ip=t1.ip上
和t1.`date`

试试这个

---

您希望删除所有存在较新条目（即具有相同IP地址和较高日期的条目）的记录：

---

此查询显示了一种为ip选择除最新行以外的所有行的方法：

WITH 
 --this bit of code just creates some data we can play with because I
--don't have your table structure in my database
 ip_data (ip, creation_date)
 AS
 (SELECT '1.2.3.4',sysdate-1 FROM dual UNION ALL
  SELECT '1.2.3.4',sysdate-2 FROM dual UNION ALL
  SELECT '1.2.3.4',sysdate-3 FROM dual UNION ALL
  SELECT '1.2.3.4',sysdate-4 FROM dual UNION ALL
  SELECT '1.2.3.4',sysdate-5 FROM dual UNION ALL
  SELECT '1.2.3.9',sysdate-1 FROM dual UNION ALL
  SELECT '1.2.3.9',sysdate-2 FROM dual UNION ALL
  SELECT '1.2.3.9',sysdate-3 FROM dual UNION ALL
  SELECT '1.2.3.9',sysdate-4 FROM dual 
  ) 
--this query takes our data with row numbers and excludes any with row number
--1 (that is, the most recent row for each ip)
SELECT
 ip
,creation_date
FROM
  --this query assigns a row number to each row.  The latest row for an ip
  --get row number 1.  Numbering restarts for each ip (PARTITION BY ip)
 (SELECT
   ip  
  ,creation_date
  ,ROW_NUMBER() OVER (PARTITION BY ip ORDER BY creation_date DESC)    rn
  FROM
   ip_data
)
WHERE rn > 1
;

您需要为您的表结构重新编写此注释，但希望注释有意义。一旦您了解了它的工作原理，您就可以将其应用到以下内容中：

DELETE FROM <your table>
WHERE (ip, creation_date) IN
 (<select statement similar to the above>)

从中删除
其中（ip、创建日期）
()

尝试使用distinct您想只获取唯一的ip地址，还是要删除它？首先确定您使用的是哪种RDBMS。然后看我在使用oracle sql devoolper不要按注释编辑答案，而是将注释添加到注释框中。他没有两个表。@user7417866这是自联接。@user7417866、t1和t2是所谓的表别名，这里引用了同一基表的两个实例。您好，谢谢您的快速响应。实际上没有行ID。这是一个带有列的表-IP地址、区号、添加的日期我想删除重复的IP地址，只保留最新的日期谢谢您的快速响应请解释结构中的aa和bb值。感谢you@Ashfen99它们只是别名

WITH 
 --this bit of code just creates some data we can play with because I
--don't have your table structure in my database
 ip_data (ip, creation_date)
 AS
 (SELECT '1.2.3.4',sysdate-1 FROM dual UNION ALL
  SELECT '1.2.3.4',sysdate-2 FROM dual UNION ALL
  SELECT '1.2.3.4',sysdate-3 FROM dual UNION ALL
  SELECT '1.2.3.4',sysdate-4 FROM dual UNION ALL
  SELECT '1.2.3.4',sysdate-5 FROM dual UNION ALL
  SELECT '1.2.3.9',sysdate-1 FROM dual UNION ALL
  SELECT '1.2.3.9',sysdate-2 FROM dual UNION ALL
  SELECT '1.2.3.9',sysdate-3 FROM dual UNION ALL
  SELECT '1.2.3.9',sysdate-4 FROM dual 
  ) 
--this query takes our data with row numbers and excludes any with row number
--1 (that is, the most recent row for each ip)
SELECT
 ip
,creation_date
FROM
  --this query assigns a row number to each row.  The latest row for an ip
  --get row number 1.  Numbering restarts for each ip (PARTITION BY ip)
 (SELECT
   ip  
  ,creation_date
  ,ROW_NUMBER() OVER (PARTITION BY ip ORDER BY creation_date DESC)    rn
  FROM
   ip_data
)
WHERE rn > 1
;

DELETE FROM <your table>
WHERE (ip, creation_date) IN
 (<select statement similar to the above>)