tjmahr / tjmahr.github.io

A responsive Jekyll theme with clean typography and support for large full page images.

https://tjmahr.github.io

MIT License

9 stars 8 forks source link

random-effects-penalized-splines-same-thing/ #2

Open utterances-bot opened 3 years ago

utterances-bot commented 3 years ago

Random effects and penalized splines are the same thing - Higher Order Functions

Weighted wiggles and smoothed categories

https://tjmahr.github.io/random-effects-penalized-splines-same-thing/

tjmahr commented 3 years ago

Tristan, this post was buttery smooth. (I'm testing out comments. I just installed them.)

mbedward commented 3 years ago

Brilliant post - thanks Tristan. This is the first explanation of this topic that I've been able to get my head around in any meaningful way.

tjmahr commented 3 years ago

Thank you!

cdriveraus commented 3 years ago

Nice. Thinking along these lines also made the connect between splines and stochastic differential equations clearer for me.

mtthwastn commented 3 years ago

It's always good to see this connection explained clearly. "Smoothing and mixed models" (2003) by Matt Wand is an elegant paper on the subject that you may find interesting (link is from his personal webpage).

adbpatagonia commented 2 years ago

Thanks! This is brilliant It integrated several concepts I knew, but did not know quite well how they related to each other. I heard the “random effects are penalized smooths “ from Eric Pedersen a few years ago, and at the time I thought “I’ll never be smart enough to understand this”. Great explanation, thank you

TaymAlsalti commented 1 year ago

Hi Tristan,

It's been a while, but do you remember whether you found a way to extract details about fit splines from fit brm models the same way you did from the fit gam models?

tjmahr commented 1 year ago

Nope, I can’t find anything.