Jauntbox / SurgeTrends

Collect and plot Uber surge multipliers as a function of time

SurgeTrends is a Python script to collect and plot Uber surge multipliers as a function of time.

Instructions:

• Register for an API key with Uber: https://developer.uber.com/
• Create a file called config.py with a single line specifying your server token server_token="string you got when registering with Uber's API"
• Define a set of location endpoints you'd like you use for your trips in the locations dictionary in the surge.py file
• There is example code for either hard-coding trips between selected endpoints or generating all possible trips between your defined endpoints. Note that Uber limits you to 1000 API calls per hour so keep this in mind when specifying your trips and time between queries.
• Choose an output file to store your data in (look for the output_file variable)
• Run the code via python surge.py
• Look through the iPython notebook file make_plots.ipynb for examples on how to plot the results. Hopefully it's documented well enough to follow.

For running the dashboard (currently super-basic):

• You'll need to have a data file already generated from surge.py, e.g. called data.txt
• Next, convert the text file to serialized json format with python txt_to_json.py (you'll need to open the python script to edit the file it opens)
• Once you have a data.json file, edit dashboard.html so it knows to open it
• To view the webpage locally, you need to run a local webserver, e.g. with python -m SimpleHTTPServer 8001
• Open your web browser and type in http://localhost:8001/ which will show the contents of the folder than the webserver was started in
• Navigate to your SurgeTrends folder and open up dashboard.html - you should see a plot of your data