These are just some experiments which helped me to get to grips both with using vegalite in Clerk notebooks (in clojure). I was playing with the equations for calculating a binomial distribution in order to get a better grip on "bayesian updating".
Changes pushed to github will automatically be compiled by a github action and deployed to github pages you may need to enable this for your repository if you fork this one.
See built static pages here.
Notebooks include:
- A slideshow of notes I made to clarify my thinking about Statistical Rethinking Chapter 2.
- An attempt at animating Bayesian updating
- A slidehow of graphs showing a model adapting to evidence, basically the same as the above animated example but as a series of slides.
TODO:
- Others might be able to improve on my explanations in the first slideshow.
- Not happy with the symbol representing land and water sample W and L also being used for parameter name.
- The animation I tried to do in notebooks/bayes_posterior.clj doesn't work on github pages, want to experiment with cljs next.
- Structure of code could probably be improved.