What about a simple count?

WITH dups AS (SELECT f1 [, f2, ...]
                FROM mylib/myfile AS mf1
                GROUP BY f1 [, f2, ...]
                HAVING COUNT(*) > 1
             )
SELECT RRN(mf), mf.*
  FROM       mylib/myfile AS mf
  INNER JOIN dups         AS d  ON  d.f1=mf.f1 [AND d.f1=mf.f2, ...]

You could easily rewrite this as an EXISTS instead of the CTE and the INNER JOIN, but it would come out the same.

I'll have to give that one a try.

We've actually done this in the INSERT TO table, as an up front way to avoid generating a dup key error that would otherwise have to be handled.

If you're loading from a spreadsheet, you probably don't have a huge quantity of data. But, generally speaking, if f1 [, f2, ...] are high order keys in any index, the GROUP BY shold be an index only operation, so fast. The larger the table, the more important this becomes.

To avoid duplicate keys/entries on an INSERT, I tend to use the WHERE NOT EXISTS clause.

Something like this:

 EXEC SQL INSERT INTO customer (dbField1, dbField2)           
           (select  :hostVar1, :hostVar2 from sysibm/SYSDUMMY1   
            WHERE NOT exists                                   
       (select * from customer where dbField1 = :hostVar1)); 

The problem I've found with WHERE NOT EXISTS is that applies to what is in the destination file at the time you run the insert. In many cases the data I'm given is repleat with duplicates. The result is that I run the insert and get 30 records added and a duplicate at 31, so I run the SQL !--script-- again (first 31 records are skipped and I get another 20 inserted before the next duplicate). So I run the same SQL !--script-- over and over, until I reach the end of the file - and I don't have foreknowledge of which records are duplicate and able to make a predetermination of whether one is incomplete, whether it's a true duplicate (all columns) or only the key is duplicate.

By checking for duplicates before I start, I can make an eyeball predetermination of the obvious bad rows, and can highlight duplicates back to the data owner for clarification.

Just to mention another coding option to avoid adding dupes, you can do an EXCEPTION JOIN between the source and target tables to INSERT all the rows except the ones that already match on keys.  This tends to perform well when you have many rows to process becasue it does an index scan to see the match.

To avoid adding SOME dupes from the source table, you can add the word to DISTINCT to your SELECT statement.  If everything about the columns in your select list are identical, the SQL processor will automatically coalesce multiple rows into a single row, avoiding dupes for this type of error.