pub fn max_by(value_col: &Column, ord_col: &Column) -> Column
Value of value_col in the row where ord_col is maximum (PySpark max_by). Use in groupBy.agg().