Step 1: code takes too long to run Step 2: spend 6 hours optimizing until the memory footprint is halved, the GC calls are gone, and it runs 25% faster Step 3: run the project again, notice 0 improvement in performance for the analysis Step 4: ??? Step 5: actually merge the branch with the new code

I like to think of myself as "not overly pedantic", and yet today I entered a shop and immediately turned around because they were playing Joshua Bell. Absolutely not. Under NO circumstances.

This is a real concern. I can think of at least one paper that would make most of the key players in US biodiversity sciences ineligible for funding. Collaborative papers that set the agenda for an entire field are now preventing this research agenda from being implemented (in the US).

It's good that I get to moderate a panel on biodiversity and one health, a topic on which I have 0 inflammatory opinions at all.

Pray for my undergrads, nothing wrong with them but they're learning about pairwise invasibility plots today

For my Data Science for Biodiversity class (I get to teach this! And also Simulating Life in the winter! I have the coolest job!), students must use @gbif.org to get data. Proper use of DOI is part of the assessment. They learn (i) good data citation practices and (ii) how amazing GBIF is. 🧪

Excited to do grading because I've received my new gel pens. So if you're wondering whether I'm still cool, I'd greatly appreciate that you keep the answer to yourselves.

A student wrote to me after the midterm, and they said they should have gotten a lower grade. I think this is important, because I fundamentally disagree with this student, and it took me a while to articulate why. This is my rough draft. 🧵

More raccoons? Turns out it would be bad news. We're predicting where key reservoirs of the raccoon rabies virus (🦝 and 🦨) may be in Québec by 2100, and if we don't act on climate change, the answer may be: everywhere. 🧪

This is correct, with one important caveat - most decisions on funding, admin, publishing, career, are made (and can be derailed) by people who by definition benefited the most from these metrics and incentives, and stand to lose the most by changing them.

Join the ENDEMIC Network on 4 November for a webinar that focuses on how to make reasonable adjustments in the workplace and at university for neurodivergent ecologists at any career stage. Register now 👉 https://f.mtr.cool/rkknflximp

🚨 NEW: The majority holder of the world's genetic sequence data is a bad actor who can cut off access to critics and competing services. We've tolerated this for years, and now it threatens the pandemic treaty. Time for WHO to step in. With @ctrlalttim.com: www.thinkglobalhealth.org/article/to-f...

As one of the greatest writers of our age said, “venom with a new twang, same swagger, new thang” - the lab has changed a lot over ten years, and after thinking about who we became, we decided to re-do the website. And write about our theory of change. Follow us - @epic-biodiversity.org !

When sequence databases can unilaterally cut off access to analytic platforms, we are left unprepared for the next pandemic. In @thinkglobalhealth.org, @colincarlson.bsky.social and I highlight the severity of the situation, and what it means for @who.int negotiations on the Pandemic agreement. 🧪😷🧵

The Canadian government has seemingly never read a single one of its own commissioned reports on what Canadian science needs. (It's not more people.)

He's got a law degree and writes editorials about how the Nagoya protocol was a bad idea

"opening Word and opining" is also not how you write a literature review, and honestly a baffling (and dismissive) oversimplification. In my experience, it's more along the lines of "spending months reading hundreds of papers and synthesizing them". Reviews are not opinions.

"The pupils do not learn the scribal craft" is the new "the midterm average was 62%"

I submitted a grant, reviewed over 130 applications for a position, and returned a big deliverable for a contract over the past 48 hours. When Lou Reed said "I am tired, I am weary, I could sleep for a thousand years", honestly, same.

Rough draft, obviously, but I think I like this approach to bivariate legends.

OK, that's enough dataviz for tonight.

I was out of the country for less than 3 days. My reimbursement claim is 24 pages long, for 7 items. By the time I was done, I had spent about 2 hours on it. 🧪

I was listening to a podcast on Paulo Freire, and the part about critical consciousness was immediately followed by "anyway thanks to better help for sponsoring this episode". 90% of comedy really is timing.

I surprisingly enjoyed writing my NSERC discovery application this time. Now I need to clean up my CCV. I will enjoy this a lot less.

Typst is, as far as im concerned, at the center of latex (rendering), markdown (ease to learn), and HTML (incredibly powerful styling abilities. It's quite incredible, and it's so, so, so fast. I have a 60k words document with dozens of figures and hundreds of refs. It renders _in real time_.

Sometimes a reviewer will write something that is so wild that I will genuinely start panicking, fearing having gotten something fundamentally wrong. It turned out that yes, I did write the softmax function correctly. Thanks for the panic attack, reviewer 1.

#LivingData #DatosLivos is starting three days, and the lab (and friends) will be well represented. Here are some presentations you simply can't miss. 🧪🌎 On Oct 21, @crcruzr.bsky.social will tell you everything about socio-ecological indicators and what they don't yet do well. 🧵 (1/n)

Me, a gainfully employed computational ecologist, while teaching: "Biostatistics are pâtisserie, but data science is cuisine"

We went to see The Tallest Man on Earth tonight, and in lieu of a review: I bit my lip so hard I drew blood. What an incredible artist.

I just finished Metal From Heaven by August Clarke, which is a great read - and it also contains a plot twist that made me go "no, no way" out loud. I love it when a plot twist is very transparently foreshadowed, but only in retrospect.

Inspired by an event I attended this week: the number one issue for adoption of data sharing practices is that part of the open science community is unable to conceptualize that incentives won't ever fix the fact that some people just don't want to share.

Very sincerely, we don't do the "lay piles of paper on the ground" type of research enough anymore. This was peak research.

I spent the last 3 years or so thinking that Obsidian was meh. I was wrong. I was so deeply, fundamentally wrong. Obsidian is my life now.

"You don't have wifi because you're on Linux" No, IT guy, I don't have wifi because my desktop doesn't have a wifi card. I don't think running windows would help with that.

_scratch.whatever

I did, in fact, ask for this.

I went to sit in the forest and enjoy the autumn sun, and did one (1) math. What I suspected might be a vaguely interesting idea is now turning into an actual draft. Touching grass works.

hey i just got an alert saying your personal info was found on the dork web

In community ecology, this is called the "Appendix XVII of a Peter Abrams paper" effect

Me staring at the ceiling at night: Maybe I'm done having ideas? What if one day I have no idea what the next research question is. Does everyone know I have peaked long ago and didn't realize it? Me planning a grant: In conclusion, this is why I need 12 PhD students over the next 5 years.

Our software paper on SpeciesDistributionToolkit.jl is now published in @peercomjournal.bsky.social - peercommunityjournal.org/articles/10....

I just finished "The West Passage" by Jared Pechačekz, and this may be my favorite fantasy book of the year so far. One of the things it does really well is making a world so chaotic and impredictable feel so consistent and real. Impeccable worldbuilding.

This morning in class, I was showing a map of temperature in Mexico, and I asked students what type of problem it was. The correct answer was "clustering", but one student answered "a basic cartography one, because I don't see a legend or an orientation marker on that map". Honestly? Respect.

New preprint! 🥳🎉 We looked at viral coinfection patterns at the largest scale ever in wildlife. We found a strong association among CoVs, PMVs, and influenza A, and higher coinfection rates in wildlife trade; plus, evidence that bats accumulate persistent infections. www.biorxiv.org/content/10.1...

The @panaction.bsky.social just published a fantastic briefing paper about the Pathogen Access and Benefit Sharing system - take a look if you want the full perspective on what PABS is, and why it matters for both health equity and fair data governance. www.pandemicactionnetwork.org/news/briefin...

11y.o. showed up with his copy of Fellowship of the Ring, absolutely livid, demanding to know why they cut Tom Bombadil from the movies.

lol. lmao even

Today I learned that the rain cover for my saddle bags is great at keeping the water IN. In particular, in the bag that contained an iPad and the entirety of my notes for population dynamics, which I am teaching tomorrow at 9am.

Yo come over! We got some of the master's tools and we're gonna wreck his shitty house hahaa

Site selection algorithms for optimal ecological monitoring design, by @karinorman.bsky.social - now out 🧪 www.sciencedirect.com/science/arti... Long story short: most site selection algorithms for biodiversity monitoring have similar performances - users can pick them based on features!