We just published this tutorial about ARMA(p,q) models for modeling time series, and how to fit them using Python. But while it’s a tutorial, it has a few twists. First, it’s interactive: you’ll learn by solving problems and making choices. Second, it’s a story: you play a character in a plot that gives you real-life problems to solve. And third, it’s illustrated: we spent many hours hacking with Stable Diffusion, GIMP, and matplotlib.
This is chapter 3 in our interactive course, Everyday Data Science. [1] The first half of the chapter is free. You can get the whole course forever for $29. These chapters are a lot of effort to produce, so please let us know what you think :-)
- Andrew Carr [2] and Jim Fisher [3]
[1]: https://news.ycombinator.com/item?id=32118530
[2]: https://twitter.com/andrew_n_carr
[3]: https://jameshfisher.com/
I'm not sure which is more annoying:
-going through years of education in pure mathematics without this kind of tutorial
-not having come up with this method myself, when it's so close to the dumbed-down edutainment software of the 90's
Every university course should start with at least a day of instruction written in this style.
[kicks dirt]