Computational neuroscientist. Senior Lecturer at Ulster University. http://odonnellgroup.github.io
@adredish @cian @elduvelle Re hardware support, I think things are getting there slowly. Here are some forum discussions, including a link to a talk from JuliaCon.
https://discourse.julialang.org/t/does-julia-support-any-hardware-developement/105024/4
My first publication as a solo author is out in Physical Review Letters! 🎉🎉
Here I introduce a novel mathematical formulation to deal with systems of coupled oscillators which are subject to noise, bringing a lot of new physics to the table! I worked it out for the neuroscience applications but the theory is very general.
Link: https://journals.aps.org/prl/abstract/10.1103/PhysRevLett.134.197201
The punchline of the paper is that I am able to compute exactly the fluctuations to the Kuramoto order parameters, which encode all the information about the system's synchronization state. Those equations have very important practical consequences! Let me do a small explanatory thread about it :)
Cellular and subcellular specialization enables biology-constrained deep learning https://www.biorxiv.org/content/10.1101/2025.05.22.655599v1?med=mas
I've also heard very good things about Julia. I'm tempted to take the time to rebuild my intralab Matlab codeset (which is very large) in Julia. Or at least to start looking into it.
How stable is it?
Does it have the python ecosystem problem?
Does it have reasonable hardware support? (In the sense of being able to access arduinos with serial ports and cameras with video?)
@adredish @elduvelle and sorry I have no idea about the hardware support!
@adredish @elduvelle id say worth implementing some routines you use a lot and doing a speed comparison with MATLAB to see if it's worth the effort for your lab.
Julia was annoyingly unstable until v1.0, now it is pretty good. Of course sometimes get dependency issues though.
The Julia package manager is excellent. They keep it a lot more locked down than conda or pip seem to.
@elduvelle @adredish have you tried Julia? I wish it had more uptake in neuroscience. Seems like exactly what we want
@adredish Julia has been my go-to language for a few years now. Very easy switch from MATLAB/Python, and ime you often get big speed ups 'for free'. Miles ahead for differential equation solving especially. Only downside is the poor documentation. But LLMs largely debug well for me these days
Does anyone know what happened to the Mathworks website?
What alternatives are people using for Matlab?
(And don't say Python. In my experience, python is on the order of 1000 times slower than Matlab for the neuroscience that I do, even using NumPy, even when written in an unreadable manner (q.v. https://mastodon.social/@dynomight/114512683967447209). Moreover Python is particularly inefficient because it is always copying objects. I can run Matlab code on my desktops and the equivalent Python code requires a supercomputer. And don't get me started on the frustration and inefficiency of that stupid f***ing conda ecosystem one has to run to match all the different versions of all the different codes. <rant off>)
PS. Yes, Matlab has its limitations (objects are very much a hack), but IME nothing is as efficient for large matrix operations, which is what neural ensemble analysis is.
Announcing: https://justaqrcode.com.
Tired of "free" QR code generators that are full of ads and trackers, that share your data, and that want to sell you something? Me too. Here's my act of resistance: I made a one-page site that works entirely in your browser to generate a simple QR code. And that's all it does. You can download the HTML page and run it locally, even. Read the source; nothing up my sleeves. Just a QR code.
My offer to you -- I will continue to pay for the domain name and web hosting for it, myself. If you find it valuable, you can pay it back by creating your own useful thing for the world and releasing it for free. Let's take back the friendly web, one vexingly-monetized utility at a time!
Here it is, in a clear graphic from the NYTimes. The GOP budget destruction bill will add $3.3 TRILLION to the national debt.
And how? Entirely by tax cuts to the rich, paid for by cutting health, food for the poor, climate science, and Medicaid.
@petergleick doesn't sound very paid for?
Currently reflecting on how much change the Covid-19 pandemic drove in higher education (good and bad) and how much of that has persisted.
Just before I took over as Dean the pandemic forced my institution to finally accept (almost overnight) electronic submission of doctoral theses, which had been a red line for our academics for many years (and a very emotive one). We also immediately allowed online vivas and supervision and most community activities and training either shifted online (sometimes very successfully, sometimes not) or withered.
I effectively became Dean of an entirely online operation for a short while, until we re-opened labs and specialist facilities. And all our support for that had to be maned without a single F2F meeting for quote some time. It took me nearly two years to move into my new office… not the job I was expecting or a way of working I was used to (although I got a bit of practice at the end of my previous role).
Since then we have had to roll back on some of our online offer (especially around supervision) to my immense regret.
We still only accept electronic thesis submission and no-one ever asks to reintroduce the hard copy even as an option.
today I heard for the first time about "STEAM" education, the attempt to bring back Arts into STEM training. Which sounds great! 🚂
@jonny LA traffic nexus + Shell reminded me of my favorite portrait of Los Angeles I took a few years ago
Charting human migration over deep time
(via https://outofedenwalk.nationalgeographic.org/media/2013-01-a-walk-through-time)
I've felt for a while that a mainstream method, reverse engineering, in cognitive science & AI is incompatible w computationalism‼️ So I wrote "Modern Alchemy: Neurocognitive Reverse Engineering" w the wonderful Natalia Scharfenberg & @Iris to elaborate: https://philsci-archive.pitt.edu/25289/
1/n
The long-sought pinnacle, the very apex, of a well-rounded education, the platonic ideal of the life of the mind, is to become a full-time fact-checker and nonsense-detector for devices that extrude plausible-sounding slop for the benefit of people who are trying to get away with something.
growing up in Ireland, we learned that the Great Famine was due to the bad luck of potato blight
later we were told it was due to negligence by the British government
here Colm Tóibín argues that it was an intentional annihilation of a class of people
https://open.spotify.com/episode/5O6AjDI0sG9oVJmVwsKH4a?si=w5UOIIsPQZeaJiaNJIktuQ
highly recommend people of ireland/Britain/beyond listen to these 2 Empire pod episodes on the Famine (plus ~10 others on Irish colonialism)
Absolutely heartbreaking to listen to while a modern version plays out in Gaza.
it's marking season and every academic I know is complaining about ChatGPT use among students
