04:13 <nuck> Hey, we have some math nerds here right?

04:14 <nuck> Trying to figure out the best way to calculate "compatibility" between two Users, given a list of items and ratings

04:14 <nuck> Somebody suggested using vectors but frankly I just have no idea how to do that

04:16 <nuck> It's basically just comparing two weighted lists, but the lists could be tens of thousands of weighted rows

05:23 <jfhbrook> nuck: I think you want cosine similarity

05:23 <jfhbrook> nuck: https://en.wikipedia.org/wiki/Cosine_similarity

05:24 <jfhbrook> nuck: where a dot product is something like def dot(x, y): x[0] * y[0] + x[1] * y[1] + x[2] * y[2]

05:25 <jfhbrook> nuck: and the length is ye olde pythagorean equation

05:26 <jfhbrook> nuck: you can do this with apis instead of "manually" if you bust out a linear algebra lib and vectors, which for your purposes are collections that support certain kinds of operations, like dot products

05:27 <jfhbrook> (there are of course other industry grade ways of doing recommendations if that's what you're into, but this is a good, not too confusing start