Reputation: 2380
I have the following table:
id - home id - home status - previous home status
1 - a - occupied - null
2 - a - abandoned - null
3 - a - foreclosure - null
4 - b - occupied - null
5 - b - occupied - null
6 - c - occupied - null
I want to generate an update statement that updates the table as follows:
1 - a - occupied - abandoned
2 - a - abandoned - foreclosure
3 - a - foreclosure - null
4 - b - occupied - occupied
5 - b - occupied - null
6 - c - occupied - null
I'm using the following algorithm:
set @id = 1
set @max_id = (select MAX(id) from home)
select @homeid = homeid from home where id = @id
while @id <= @max_id
begin
select
@previous_homeid = home,
@previous_home_status = @home_status
from
home where id = (@id + 1)
if(@previous_homeid = @homeid)
begin
update
home
set
home_status = @previous_home_status
where
id = @id
end
set @homeid = @previous_homeid
set @previous_homeid = null
set @previous_home_status = null
set @id = @id + 1;
end
Is there a more efficient algorithm? I don't like the loop efficiency
Upvotes: 0
Views: 3538
Reputation: 96055
This is untested, but should get you what you're after:
WITH CTE AS(
SELECT id, [home id], [home status], [previous home status],
LEAD([previous home status]) OVER (PARTITION BY [home id] ORDER BY id ASC) AS NextStatus
FROM home)
UPDATE CTE
SET [previous home status] = NextStatus;
Upvotes: 2
Reputation: 49270
You can use lead
to get the value on the next row.
with cte as (select t.*
,lead(home_status) over(partition by home_id order by id) as next_home_status
from tbl t)
update cte set previous_home_status=next_home_status
Upvotes: 2