PostgreSQL: detect duplicates

Detect duplicates in at_file after a given date:

1 2 3 4 5 6 7 8 9

select * from monitor_db.at_file  where exists (  select 'x'                   from monitor_db.at_file i                  where i.file_id = at_file.file_id                    and i.process_id = at_file.process_id                    and i.ctid > at_file.ctid                    and log_date > '2017-01-31'              )