Hi! Thanks for hosting this competition. I'm looking forward to taking part.
I noticed in the training data that some of the buses leaving Kisii going to Nairobi have shuttles with identical ride ID's. See for example trips with ride ID's 7942, 8125 and 8524. I was wondering how we should treat these cases when aggregating the training data. Does a bus + shuttle with the same ride ID count as one trip and therefore we should count their number of tickets together? Or should we treat them separately and create a new unique ID for either the bus or the shuttle?
In any case, it shouldn't make much of a difference, there are only 7 such examples. Thought I just point it out.
@janmarais, great catch! Well-spotted! You are right, those should be different ride IDs. We will correct the training set and re-post it as a different version. But for all of you out there that have already started working, just note that the rides should be unique for travel_time, travel_to, travel_from, and car_type. Keep up the good work!