SQL does not allow for the substitution of column names in an expression. Given the table with column names, a SQL can be written with string concatenation a which results in one SQL statement per column name. That result can be written to a file and then executed to provide the desired result.
Trying every column and and combination of columns is probably overkill. It is probably better to focus on the few that would also provide value as join columns or direct access columns. If there are no columns that have access value and all that is desired is even distribution, the a NOPI is a more direct solution.
ToddAWalter can you please provide some sample script to use NOPI.
I understand from you comment that NOPI will distributed data better but how will i come to know that based on which column data was distributed for NOPI?
See the documentation under create table, it will tell you all about the syntax for NO PRIMARY INDEX (NOPI).
With NOPI, column data is not used for distribution, it is a round robin/random algorithm instead.
If you are using Create Table As or Insert Select to populate this table, you may also want to read about the HASH BY RANDOM clause if your source data is skewed.