Hourly and daily bike rental counts from Washington DC's Capital Bikeshare, with weather and calendar features.
About
Two years of Capital Bikeshare usage aggregated by hour and by day, joined with weather (temperature, feels-like temperature, humidity, wind speed, weather situation) and calendar information (season, holiday, working day). The target is the rental count, split into casual and registered users.
Files in the zip
| File | Rows |
|---|---|
hour.csv |
17,379 |
day.csv |
731 |
Readme.txt |
Column definitions |
Temperatures and other weather values are normalised (divided by their maximum); the Readme gives the scale factors.
Source and licence
Licensed CC BY 4.0.
Fanaee-T, H. (2013). Bike Sharing. UCI Machine Learning Repository. https://doi.org/10.24432/C5W894