I chose Ford GoBike System Dataset between January 2019 and December 2019 for my Anaylsis. I was interested in trying to understand more about user preferences, help to discover usage pattern, rider characteristics.
There are two datsets in the branch:
fordgo_bike_master.csv: The original dataset from merging every month of 2019 datsets.
bikes_master_ordered.csv: The cleaned dataset from the original one.
- The most usage of bikes were in thee morning between 7 - 9 AM in the morning and 16 - 19 which is a tpyical workingday
- Thursdays and Tuesdays are the most popular days for using bikes, however Wednesday, Mondays and Fridays are very close to Tues & Weds numbers. All of that suggests that Bike usage are more usable on the workdays than weekends.
The Multivariate Exploration strengthened most of the patterns from my previous findings. The short period of bike usage are for subscribers that use it for a very short duration between Monday and Friday, which indicates to that bikes are used for work/study commute. On the other hand, plots are showing that customers have a very flexiable duration and also they use bikes more often on weekends, which suggests that they use for tours or lesisure.
Subscribers are using the bike sharing system more than casual customers. Subscribers hd the most rides during the spring season and the least during the winter months. On the other hand, customers got the most in late summer and in Autumn, also as subscribers they got the least in winter. Obviously the usage of both types are differet and they have different patterns and riding habits. Subscribers used the system heavily on work days around 7-9AM and 16-18PM for work commute or study, whereas customers ride a lot over weekends and in the afternoon for leisure/touring purposes. Subscribers tended to have much shorter/quicker trips compared to customers which makes subscriber usage more efficient.