Question

这是我的问题我有1+ mil行和50 +列的表，其中保留所有记录更改因为需要删除列的表大小并且需要删除所有重复的记录

----------------------------------------
ID   otherID   a     b   |    c  
1     10       1     2   |    1
2     10       1     2   |    2
3     20       2     2   |    3
4     10       1     2   |    4
5     10       10    2   |    5
6     10       10    2   |    6
7     10       1     2   |    7
---------------------------------------

所以我要说删除C行，我只需要保留第一条唯一记录我需要的记录是1,3,5,7 我的问题是我如何分组1和2和4而不是第6行

Answer 1

您可以使用row_number()方法的差异：

select min(Id), otherId, a, b
from (select t.*,
             (row_number() over (order by id) - 
              row_number() over (partition by otherId, a, b order by id)
             ) as grp
     ) t
group by otherId, a, b, grp;

对于相邻行中具有相同值的行，行号的不同是常量。

Answer 2

U可以使用以下查询获得第1,3和5行

 SELECT * FROM (SELECT tbl.*,
              ROW_NUMBER() OVER (PARTITION BY otherId, a, b ORDER BY id)
             AS rn
             FROM [tablename] tbl
     ) t
WHERE rn = 1

但是获得第7行的要求并不明确，因为它不是唯一的行（与第1,2,4行相同

Answer 3

如果我做对了

declare  @T TABLE 
    (ID int, otherID int, a int, b int)
;

INSERT INTO @T(ID, otherID, a, b)
VALUES
    (1, 10, 1, 2),
    (2, 10, 1, 2),
    (3, 20, 2, 2),
    (4, 10, 1, 2),
    (5, 10, 10, 2),
    (6, 10, 10, 2),
    (7, 10, 1, 2)
;

with grp as (
select ID, otherID,  a,  b,
   grp = ROW_NUMBER()OVER(PARTITION BY otherID ORDER BY (id))
   - ROW_NUMBER()OVER(PARTITION BY otherID, a, b ORDER BY (id))
   from @T
), filter as (
select ID, otherID,  a,  b, grp,
   rn = ROW_NUMBER()OVER(PARTITION BY grp, otherID, a, b ORDER BY (id))
from grp
)
delete from filter 
where rn >1;

select * 
from @T 
order by id

Sql组只有连续的行

3 个答案: