How come in the train data we have customers buying say 0.5 , 0.2 , 0.3 , 0.4 of a particular product. These values aren't intuitive for the 'qty_this_week' column.