This dataset contains bikeshare system data from January 2017 until August 2020 in zipped seperate files. 2017 was in one file, every other month was in a separate file. The data before cleaning had these features:
- Duration in seconds
- Start time
- End time
- Start station id
- Start station name (dropped in cleaning)
- Start latitude
- Start longitude
- End station id
- End station name (dropped in cleaning)
- End latitude
- End longitude
- Bike id (dropped in cleaning)
- User type - subscriber or casual customer - .
- Bike share for all trip or not (dropped in cleaning)
- Rental access method (dropped in cleaning)
- Ride id (dropped in cleaning)
- Rideable type (dropped in cleaning)
- Is equity or not (dropped in cleaning)
We notice a slightly high rent count in July and September and a little higher counts in August and October
Weekends have significantly less rent counts
Weekends have more rental durtion
We notice that most of rent durations are less than 34 minutes and most of them are between 0 and 16 minutes. And the mode is about 5 to 6 minutes
We notice two areas with highest density. The area from -122.396 to -122.394 lat, 37.775 to 37.780 lng and the area from -122.94 to -122.392 lat, 37.790 to 37.795 lng
A significant number of stations have very low number of rentals
Subscribers has a lot more rentals than casual cutomers.
For customers there are more trips in January, February and August. But for subscribers there are more trips August and October then in July and September
casual customers tend to have more rent duration than subscribers
We notice that weekends have less trips than weekdays but they have more rental durtion
We surely notice that subscribers has a lot more rentals than casual cutomers. We also notice that for customers there are more trips in January, February and August. But for subscribers there are more trips August and October then in July and September