I have a large dataframe (df) with 10 columns. I am interested in making a new dataframe, one column (col2) is sorted according to values of another column (col4). To use a
