如何在（子查询）中选择 Where column_name

Question

I'd like to select all the rows of a table where the value of a column is in the results of a query. conceptually the code below seems like it should work but I'm not sure whether it's a syntax error

SELECT * FROM generated.existing_conditions ec 
WHERE st_to IN (
    SELECT st_to FROM 
        (SELECT st_to AS st_to, COUNT(*) AS total_count 
         FROM generated.existing_conditions ec GROUP BY st_to) AS source_1
    WHERE total_count > 1
) source_2
;

Is this simply not allowed? can I rewrite it as a WITH query AS ?

My goal is to select the unique id's of any row with an attribute value that is repeated as this is likely to be an error in the data.

Answer 1

Your query is fine, although the two levels of subquery are not needed. You could rewrite it as:

SELECT *
FROM generated.existing_conditions ec 
WHERE st_to IN (SELECT ec2.st_to FROM 
                FROM generated.existing_conditions ec2
                GROUP BY ec2.st_to
                HAVING COUNT(*) > 1
               ) ;

However, I would suggest window functions instead:

select ec.*
from (select ec.*, count(*) over (partition by st_to) as cnt
      from generated.existing_conditions ec 
     ) ec
where cnt > 1;

Both the subqueries could be written using CTEs if you prefer CTEs.

Answer 2

compared to what I posted, the error is fixed below by removing the source_2 alias.

SELECT * FROM generated.existing_conditions ec 
WHERE st_to IN (
    SELECT st_to FROM 
        (SELECT st_to AS st_to, COUNT(*) AS total_count FROM generated.existing_conditions ec GROUP BY st_to) AS source_1
    WHERE total_count > 1
);