Created
October 20, 2015 20:45
-
-
Save jakevdp/ce1faeb01776f8160c32 to your computer and use it in GitHub Desktop.
Fremont Bike Counts 2015
Yep, looks like the dataset was updated, and they cleaned the results somehow. Thanks for pointing it out!
Oh, and the annual model is just fitting the phase and amplitude of a sine wave: y(x) = Asin(x) + Bcos(x). Not sure why you're getting a flat line, though...
Weird, I'll take a another look. Thanks!
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Jake, I was trying to work out how the annual model worked but if I run the notebook locally, I get a very different looking annual curve (see below) - guessing it's a minor issue! Any ideas - I'd love to get my head round how it works. Thanks.
ps, looks like they might have fixed the dodgy data :-)