I have a table with locations and there are duplicates because there are locations with same name that belongs to different counties.
+------+-----------+-----------
| ID | location | parent_id |
+------+-----------+------------+
| 1 | Country | 0 |
| 2 | County 1 | 1 |
| 3 | County 2 | 1 |
| 4 | A | 2 |
| 5 | B | 2 |
| 6 | C | 2 |
| 7 | A | 3 |
| 8 | E | 3 |
| 9 | B | 3 |
+---- -+-----------+------------+
I would like to create and insert in column duplicate for each duplicate ID value . The result should look like this: ( Would be better to add ID in duplicate column to all duplicates excepting the one with smallest ID)
| ID | location | parent_id | duplicate|
+------+-----------+------------+----------+
| 1 | Country | 0 | |
| 2 | County 1 | 1 | |
| 3 | County 2 | 1 | |
| 4 | A | 2 | 4 |
| 5 | B | 2 | 5 |
| 6 | C | 2 | |
| 7 | A | 3 | 7 |
| 8 | E | 3 | |
| 9 | B | 3 | 9 |
+---- -+-----------+------------+---------+
I use this to list the rows:
SELECT a.* FROM mytable a
JOIN (SELECT location, COUNT(*)
FROM mytable
GROUP BY location
HAVING count(*) > 1 ) b
ON a.location= b.location
ORDER BY a.location
It looks like a window count can do what you want:
select t.*,
case when count(*) over(partition by location) > 1 then id end duplicate
from mytable t
This requires MySQL 8.0. In earlier versions, an alternative is a correlated subquery:
select t.*,
case when (select count(*) from mytable t1 where t1.location = t.location) > 1 then id end duplicate
from mytable t