Gavin Simpson

@gsimpson.bsky.social

(Palaeo)[ecologist | limnologist] & #fakeStatistican, #rstats user, wielder of #GAMs. He/him/his. Opinions mine…

It is not possible for you to comprehend the utter loathing I have for LaTeX (specifically endfloat & longtable) + Pandoc + Quarto + knitr::kable() currently Something changed somewhere in that chain in the last 7 mo that broke my manuscript & I've spent the day figuring out why it fails #RStats

The Lego for an in class experimental design activity just arrived. So I guess I’m done for the day; time to play! The various plates are for blocking structures, and the 1x1 pink blocks are animals ( 🐷 )

Bild

Just bought 2000 Kr. (~300 USD) of #Lego for an in-class activity on the experimental design course I'm developing for our 3rd year Animal Science students Can't wait to reconcile this transaction on my university credit card with the finance folks 😬

I’m looking to recruit a postdoc to join my research group at University of Iowa to work on disturbance ecology and biodiversity! Start date flexible, lots of opportunity for skill development, collaboration, and mentorship. Will review applicants in mid-August! Link: gavinmjones.com/opportunities/

Opportunities

BIG NEWS! The Jones Lab will be relocating to the School of Earth, Environment, and Sustainability (SEES) at the University of Iowa in Fall 2026. Our work on disturbance ecology and conservation sc…

gavinmjones.com

WTH @posit.co ! I just updated my positron install and now I'm getting spammed by copilot suggested text while I'm trying to write prose. I am not subscribed to Posit AI or to Copilot though I am logged in to GitHub. Why do you turn this sh*t on by default in a new version!? Antisocial BS 🤬

🇺🇸Fun Facts: 1) The bald eagle LOVES landfills & 2) most tv/movies play a red-tailed hawk sound when showing a bald eagle, since eagles sound ridiculously silly. So, bald eagles are the perfect USA mascot: v pretty but secretly just a trash animal who steals other people’s things. Happy 250! 🦅🌎🧪

a bald eagle stands atop a pile of trash in a landfill, with other birds (likely also eagles) flying in the backdrop.
text on the image reads “©️Andrea Westmoreland”

Stan embedded Laplace was a long project and happy that it's finally released. We are not competing with INLA and TMB software, as they are orders of magnitude faster and scale better with data size for models that you can implement with them 1/

MC Stan@mc-stan.org · 3mo ago

CmdStan 2.39 has been released! - embedded Laplace approximation - new yule_simon distribution - gamma_lccdf more stable - optimization convergence status in output - length-1 tuples See more at blog.mc-stan.org/2026/05/19/r...

Thanks ❤️ Writing this paper with Eric, Noam & Dave (order determined by `sample()`) was a pleasure & has resulted in not just a thing that has proven useful to a wide range of folk, but to long lasting friendships with some genuinely nice people who also happen to be very knowledgeable about stuff

Francisco Rodriguez-Sanchez@frodsan.bsky.social · 3mo ago

#StatsPubs No matter how many times you read it, this paper on hierarchical GAMs keeps giving. Clear model explanations together with R code (mgcv). So useful! doi.org/10.7717/peer... Thanks for writing @ericjpedersen.bsky.social, D. Miller, @gsimpson.bsky.social @noamross.net

In this paper, we discuss an extension to two popular approaches to modeling complex structures in ecological data: the generalized additive model (GAM) and the hierarchical model (HGLM). The hierarchical GAM (HGAM), allows modeling of nonlinear functional relationships between covariates and outcomes where the shape of the function itself varies between different grouping levels. We describe the theoretical connection between HGAMs, HGLMs, and GAMs, explain how to model different assumptions about the degree of intergroup variability in functional response, and show how HGAMs can be readily fitted using existing GAM software, the mgcv package in R. We also discuss computational and statistical issues with fitting these models, and demonstrate how to fit HGAMs on example data. All code and data used to generate this paper are available at: github.com/eric-pedersen/mixed-effect-gams.