Below are the key findings about temporal patterns in bike share ridership that were observed when exploring all available historical processed data on bike share trips between Jan 1, 2018 and March 31, 2023. Due to the seasonality in bike share ridership, trends shown below are separated by weekdays and weekends.
<aside> 💡 The two types of stations discussed earlier during EDA are considered separately when discussing insights from the data. They are top-performing stations and all other stations. Attributes were extracted separately for each type of station.
</aside>
<aside> 💡 Bike share ridership is represented by the average ridership per station. Since the 100 highest-demand stations were considered to be top-performers, there are many more stations that were not top-performers than those that were. So, the total number of bike share trips taken from the top-performers will be smaller than the total from all other stations. This would suggest weaker relative performance at the top-performing stations, which is counterintuitive. For this reason, the average ridership per station was used to compare temporal patterns in ridership at both types of stations. Average ridership per station is defined as the ratio of total number of trips to total number of stations used in trips.
</aside>