At a very simple level, it will run through and count how many times each different combination of items appears. In this case, I will simply remove all rows with non-positive Quantity. From experience this is consistent with the state in which an analyst is most likely to receive data from a clients they are to conduct a piece of analysis for. These are compiled into a user-item ratings matrix , with items e. It is also used in recommending books or music based on other ones people have liked.
nest...