Thank you, that's very helpful!

You mention that using R and Julia gives a wider array of approaches -including Bayesian models. I'm curious what you mean by this - will Julia make my Bayesian models better/easier?

I've tried for a while, but I couldn't get the REPL to work in a convenient way and there wasn't a way to build quarto with a key combo. I've gotten spoiled by R-specific editors.

And trees are made of what you breathe out!

I choose to use the terminal because I like it! For most of what I do (updates, moving files, running code) Kubuntu tries to make me use a GUI (I've disabled Discover updater so many times yet it keeps coming back). Otherwise my workflow could be gui-only if I weren't a cli masochist.

Are there any functions to "untidy" the data (i.e. pivot so that rows contain multiple observations), especially with prefixes or messy colnames only for the new columns? Although I guess this is pretty case-specific and I should just do this myself 😊

Why oh why did I only learn about this now, after teaching a whole course in data wrangling and visualization?!? This looks awesome!

Thank you!

Wait, is there a reasoning behind the colours?

Ome-ome is actually one of the keywords!

I agree with this, but also please indent your R code well!

I switched to the "aes as layer" idea a couple months ago and I love it! As a corollary, it makes a lot of code "one line = one idea". The data is declared inside ggplot, then we declare the aes in a different line. It makes it super easy to scan the code and see exactly what variables are at play!