I would like to know if there is any difference if I save dataframe during tranformation to itself as first code or to new dataframe as second example.

Thanks

log_df = log_df.withColumn("process_timestamp",from_utc_timestamp(lit(current_timestamp()),"Europe/Amsterdam")) 

or 

new_df = log_df.withColumn("process_timestamp",from_utc_timestamp(lit(current_timestamp()),"Europe/Amsterdam"))