I have a very large dataframe (original_df) with columns of codes
14 15 21 22 18 16
And a second dataframe (crosswalk) which maps \'old_codes\' t