5 ms·
IPython notebook to play with Citi Bike data
- paddy_m 13y agoThis is a very simplistic notebook. I just wanted to play with the data a bit. There are about 250 data points included in the bundle. I would love to see what else people will do with the data and notebook.
- JeffJenkins 13y agoWhere does the data come from? The "System Data" on the website doesn't have anything yet
- apples2apples 13y agoIn function currently_available_bikes, you see him curling: https://citibikenyc.com/stations/json https://citibikenyc.com/stations/json
- paddy_m 13y agoimport json def bikes_available_for_set(stats): available_count = 0 for s in stats['stationBeanList']: available_count += s['availableBikes'] return available_count def currently_available_bikes(): !curl https://citibikenyc.com/stations/json https://citibikenyc.com/stations/json > latest.json stats = json.loads(open("latest.json").read()) available_count = bikes_available_for_set(stats) print "there are currently %r bikes in stations" % available_count currently_available_bikes() That's the code block that I use to figure out how many bikes are currently available. I pull from stations/json url.
- JeffJenkins 13y agoThanks! This is awesome.
- runjake 13y agoFor others as confused as I was, Citibike is apparently a (NYC-based?) bicycle sharing service. http://citibikenyc.com http://citibikenyc.com
- topherjaynes 13y agoRight, it's literally just rolling out as it's in beta at the moment and opens full time in June for non-year members. It's going to be some interesting data as it's NYC's first bike share program. Looking forward to some cool visualizations!