Profile avatar
ctrlalttim.bsky.social
Almost certainly one of the ecologists of all time. AI/ML, ecological networks, biogeography, virome prediction, open science, methodological anarchism he/they πŸ§ͺ https://poisotlab.io/ πŸ“° https://buttondown.email/ctrl-alt-tim
1,570 posts 17,115 followers 1,392 following
Regular Contributor
Active Commenter

Not true, it's 20 minutes actually

Typst is genuinely game changing. I'm using it for a 50k+ words document with hundreds of refs and dozens of figures, and it renders it almost instantly. And the math syntax is really great. This is what LaTeX was supposed to feel like.

Curves bend - or why I think that when talking about climate, biodiversity, and disease, it's important to remember how we managed to progress in the past. πŸ§ͺ🌎 buttondown.com/ctrl-alt-tim...

I love this.

New preprint: a toolkit for species distribution modelling, in #JuliaLang. The result of many years of work with fantastic people including @gabdans.bsky.social and @mdcatchen.bsky.social - a quick thread πŸ§ͺ🌏 ecoevorxiv.org/repository/v...

Curves bend - or why I think that when talking about climate, biodiversity, and disease, it's important to remember how we managed to progress in the past. πŸ§ͺ🌎 buttondown.com/ctrl-alt-tim...

I'm pretty sure that the reason haversine is so often written as Haversine is that it just reads better as a last name, rather than what it actually is: half a versed sine. And I'm even more sure that I've made the mistake myself, despite knowing what it is.

Paul's Fundamental Immunology

nevermind I forgot to rescale my features, a thing I TEACH to a whole classroom every year, and yet

guys I don't think I'm doing gradient descent well

Where do we look to monitor biodiversity? In a new preprint, @karinorman.bsky.social brings some good news: the different site selection algorithms tend to reach a similar efficiency, even though they recommend different locations! πŸ§ͺ🌎 ecoevorxiv.org/repository/v...

Global Ecology 🌐 starter pack vol 1 & 2 are full and curated !! βœ¨πŸ’š Please share so they will fly into the blue sky πŸ¦‹πŸŒˆ Vol1 πŸ‘‰ bsky.app/starter-pack... Vol2 πŸ‘‰ bsky.app/starter-pack... Will start a Vol. 3 soon, please reply if you want to be in ! 🌐πŸ§ͺπŸŒπŸ¦€πŸπŸ¦‘πŸͺ΄πŸ¦‰πŸπŸŒΎ

We finally got a ChatGPT generated review on a manuscript. So I wrote about how we handle these reviews is a sign of whether we have given up on peer review altogether.

Welcome, 11 y.o., to the open source movement

@buttondown.com is great!

When the reviewer wants you to start a fight

The Data Life Cycle is a useful narrative device to understand how data flow within a project. But understanding how it fits with complex model pipelines has been awkward. So we wrote a thing! It's called the Model Life Cycle, and it's now out in PLOS Comp Biol: πŸ§ͺ journals.plos.org/ploscompbiol...

Thank you, reviewer who made a bad faith argument about semantics. It had been too long since I had used the word β€œsynecdoche”.

πŸ“–PublishedπŸ“– Our new review highlights useful links (and differences) between Bayesian and frequentist approaches to smoothing, in an accessible way, with a focus on the mgcv implementation. Read more here πŸ‘‡

In two weeks, as part of my #SimulatingLife class, we're talking about ant colony optimization - it's a pretty fun algorithm to write, runs quickly enough, and gives nice plots.