Predicting Bike Rentals¶
The data we will use in this chapter is used with the permission of Capital Bikeshare. You can download the data from their website. We are using a prepared version of this data that has already been augmented with additional weather data which you can download from the UCI Machine Learning Repository.
Predicting bike rental trends is very important from both an operational and planning perspective. Bikeshare companies need to stay up to date on rental trends to know where they should add new facilities, and how to reposition bikes to get them to the locations with the highest demand. They do not want to wait until all of the bikes are rented at a particular location before moving additional bikes into position, as that is lost revenue for them.
Both hour.csv and day.csv have the following fields (with the exception
of hr which is not available in day.csv).
instant: record indexdteday: dateseason: season (1:spring, 2:summer, 3:fall, 4:winter)yr: year (0: 2011, 1:2012)mnth: month (1 to 12)hr: hour (0 to 23)holiday: whether day is holiday or notweekday: day of the weekworkingday: 0 if day is either weekend nor holiday is 1, otherwise 1weathersit:1: Clear, Few clouds, Partly cloudy, Partly cloudy
2: Mist + Cloudy, Mist + Broken clouds, Mist + Few clouds, Mist
3: Light Snow, Light Rain + Thunderstorm + Scattered clouds, Light Rain + Scattered clouds
4: Heavy Rain + Ice Pellets + Thunderstorm + Mist, Snow + Fog
temp: Normalized temperature in Celsiusatemp: Normalized feeling temperature in Celsiushum: Normalized humiditywindspeed: Normalized wind speedcasual: count of casual usersregistered: count of registered userscnt: count of total rental bikes including both casual and registered
You can read about UCI’s work with this data set here. <https://link.springer.com/article/10.1007/s13748-013-0040-3>