which_columns = ... # specify the columns whose unique values you want here uniques = {col: df[col].unique() for col in which_columns}