warrickball

@warrickball@mas.to

asteroseismologist turned research software engineer

boosts & toots about #Academia, #Astrophysics, #OpenSource, #OpenScience, #Python, #Fortran

This profile is from a federated server and may be incomplete. Browse more on the original instance.

astro_jcm, to Astro
@astro_jcm@mastodon.online avatar

1/ Most massive stellar in our galaxy found! With 33 times the mass of the Sun, this is the most massive black hole formed after the collapse of a star that we've found so far in the .

ESA's Gaia mission found it via the wobble it induces on a star orbiting it, and data from ground-based telescopes helped confirm its mass and elucidate how it formed.

https://www.eso.org/public/news/eso2408/

warrickball,

@astro_jcm

Though it still feels a long way from release, the epoch data in Gaia DR4 is going to be absolutely incredible.

warrickball, to random

An interesting development: UKRI has "pre-announced" a round of research grants that "will support projects using scientific methods to deepen our understanding of how different structures, incentives, and funding practices within the Research and Development (R&D) system shape scientific research outputs and career outcomes".

https://www.ukri.org/opportunity/ukri-metascience-research-grants/

warrickball, to random

If you're interested in how we model stellar evolution, take note!

In this recent article, Manchon et al. incorporate information about convection from fluid dynamical simulations by matching the entropy at the bottom of the simulations in their global models. It's a straightforward idea but tricky to implement and the article explores how much it matters. The spread of curves in this plot roughly shows the theoretical uncertainty about how hot red giants should be.

https://arxiv.org/abs/2401.15172

warrickball, to Astronomy

It's still #TESStuesday and I wanted to highlight a series of papers by John Southworth in The Observatory, titled "Rediscussion of Eclipsing Binaries". He's using TESS data to reanalyse known eclipsing binary systems, most recently the ~F1V+F2IV binary GK Dra. Here's his amplitude spectrum after removing the eclipses, clearly showing pulsations (probably δ Sct+γ Dor). The main period is attributed to the secondary but the primary could be pulsating too!

https://arxiv.org/abs/2401.13396

#astronomy

warrickball, to random

I'd like to tell you about what it's like being a research software engineer () but that might not make sense if I don't first explain what we mean by "research software".

While studying everything from ancient manuscripts and historical market sentiments to novel battery technologies and how the Universe formed, researchers use niche software and technologies. What's more, they often write their own programs & scripts to genereate or analyse data. This, in essence, is research software.

warrickball, to random

Boost good content.

Search engines serve a lot of SEO garbage and LLMs will only make it worse, so I'm finding personal recommendation increasing in value again.

So, boost good content.

warrickball, to academia

"Most scientists don’t enjoy writing grants."

💯

"Here’s how to change that."

😃

"Specialists share how to make the experience more enjoyable and foster a sense of belonging."

😑

https://www.nature.com/articles/d41586-023-03871-0

warrickball, to random

Recently discovered the --word-diff option for git diff and it's amazingly useful for various plain text/Markdown/LaTeX documents I work with in repos.

In certain situations it's very useful on code too!

franco_vazza, to Astro
@franco_vazza@mastodon.social avatar

I now supervise 3 young (<25yr) students on different numerical projects, from master to PhD, we exchange and discuss codes daily.

It is quite striking that these new "native digital" (term allowed for people born in ~2000?) actually have HUGE gaps, and it's not their fault (neither is mine).

Examples:

  • "it doesn't work" - with 0 attached error message, plot, information.

  • assuming a .exe compiled on a machine would just work elsewhere (like an app?).

warrickball,

@franco_vazza

I've also come across what seems like a decline in understanding how computers actually work. But we should be doing a much better job of filling in this gap in undergraduate programmes. What really got me was that we'd see 3rd-year students not vectorising Python code or knowing how to use anything besides a Jupyter Notebook, despite two years of "Scientific Computing with Python".

This isn't a problem with the students; it's a problem with our curriculum.

warrickball, to academia

Interesting article on in The Economist:

https://www.economist.com/science-and-technology/2023/11/15/new-ways-to-pay-for-research-could-boost-scientific-progress

Draws—and takes its opening thought experiment—from the excellent essay "A Vision of Metascience" by Nielsen & Qiu:

https://scienceplusplus.org/metascience/

It's a longish read (~a PhD thesis) but I found it wonderfully thought provoking. Research funding in is horribly monoculture and its effectiveness questionable. But there are so many other interesting ideas we could try!

j_bertolotti, to random
@j_bertolotti@mathstodon.xyz avatar

The best covers are those that take a song and change the style dramatically, while keeping them perfectly recognizable. Bonus points if they are a lot of fun.
https://www.youtube.com/watch?v=hPjJIujUSOo

warrickball,

@j_bertolotti

Reminds me of Richard Cheese, though I think that's from too long ago for official HD video:

https://youtu.be/EIVrEGj5fC0?t=59

pyOpenSci, to python
@pyOpenSci@fosstodon.org avatar

question: how do you get your team, PI, manager to understand and embrace the IMPORTANCE of documentation & software best practices for usability?

Our thoughts: If you want people to use / adopt your tools, documentation and accessibility is crucial.

What do you think?

warrickball,

@pyOpenSci :100: but I don't think software (re-)use is rewarded well enough in research software.

I think another possible hook is to have people think of themselves picking up their own software after a 3+ month break.

"How do I install this again?"
"What was that function interface?"
"Why did I write this function like this?

Some people may have already had this experience!

b0rk, to random
@b0rk@jvns.ca avatar

we initially printed 1500 copies of The Pocket Guide to Debugging and in less than a year they all sold out!!!

we've ordered new copies so they're back in stock now :)

get it here: https://wizardzines.com/zines/debugging-guide/

warrickball,

@b0rk Gosh, the table of contents page alone is one of the best (and definitely the most succinct) resources on debugging I've seen. 🤯

franco_vazza, to llm
@franco_vazza@mastodon.social avatar


there is a lot of talking about this I have just discovered:
https://arxiv.org/pdf/2310.01783.pdf

"Can large language models provide useful feedback
on research papers? A large-scale empirical analysis"

the paper seems well written (by humans?? 🤔 ) and careful, yet I do not like the implications many want to draw from this
"let's have all papers refereed by AI and be free from journals"

image/png

warrickball,

@franco_vazza Often when I hear "Look, here's something AI does well", I'm starting to take it as evidence that the human input isn't (or, perhaps, isn't consistently) adding much value.

E.g., that AI can write convincing mission statements strikes me as evidence that many mission statements have little meaning.

I'd sooner conclude an issue with current peer review than decide that we can replace human reviewers with AI.

warrickball,

@franco_vazza

Relevant opinion in Nature yesterday:

https://www.nature.com/articles/d41586-023-03238-5

warrickball, to Astronomy

Just uploaded the "asteroseismic" HR diagram I created for my review in A&G to the Wikimedia Commons and added it to the @wikipedia article on . Use it and edit it to suit your needs!

It's also taught me that SVG is a bit of a weird format. I can't paste it here and there are a few flaws in the online preview version that I can't reproduce...

https://commons.wikimedia.org/wiki/File:Asteroseismic_HR_diagram.svg

j_bertolotti, to random
@j_bertolotti@mathstodon.xyz avatar

After all these years, I am still finding new typos in my PhD thesis 🤦🏻‍♂️

warrickball,

@j_bertolotti @SvenGeier This is why I wish more PhD theses were easier to access. They're often a goldmine for explanations from people who were also, back then, first learning their fields.

tpoisot, to random

Our academic credentials have not been renewed by @github for several months now, despite many tickets and emails, for no apparent reason.

This is starting to interfere with research and teaching, and my take home messages are:

  1. It will likely have to move our stuff out of GitHub
  2. It will definitely stop teaching GitHub as they're not offering correct academic support

It sucks

warrickball,

@tpoisot

While I also was never able to renew my academic discount, I started moving many projects off GitHub after the Copilot beta.¹ I fear GitHub is poised for Cory Doctorow's "enshittification".²

To teach students without forcing them to choose an external provider, I'd use our institution's local GitLab instance.

¹
https://sfconservancy.org/GiveUpGitHub/
² https://en.wikipedia.org/wiki/Cory_Doctorow#Enshittification

warrickball,

@jaztrophysicist @tpoisot

💯 Institutional GitLab repos are definitely not the answer for everything! It's something we run into at @joss because we require that the software can take contributions in a system that doesn't control or restrict registration, and institutional repos often do.

I wish there was something like Zenodo+git (gitlab.cern.ch?) but sadly we still seem to need proprietary providers for pan-institutional hosting. There is @Codeberg but it's behind in features (e.g. CI).

astro_jcm, to coffee
@astro_jcm@mastodon.online avatar

To be honest, the main virtue of is that when I take it it increases your chances of surviving if you dare talk to me in the morning.

“The Virtues of Coffee” Explained in 1690 Advertisement: The Cure for Lethargy, Scurvy, Dropsy, Gout & More — by @openculture

https://www.openculture.com/2023/09/the-virtues-of-coffee-explained-in-1690-advertisement-the-cure-for-lethargy-scurvy-dropsy-gout-more.html

warrickball,

@astro_jcm @openculture That's some drop cap to start the text!

Is there a package for this style? Or maybe a template? 😛

warrickball,

@dockaba @astro_jcm @openculture

Oh, wow... 🤯 🤩

gvwilson, to random
@gvwilson@mastodon.social avatar

"Experts, who are deeply familiar with the product…are often defensive and resistant to the suggestion that the design has flaws. In contrast, less intensive users, who have given up on understanding the product, and rely on only a handful of memorized commands, are so frustrated by their experience that an analysis like ours seems to them belaboring the obvious." https://neverworkintheory.org/2016/09/30/rethinking-git.html 2/2

warrickball,

@gvwilson Great article. Reading section 3 ("Operational misfits") made it feel like the authors had watched me use . I run into the first four misfits routinely...

gvwilson, to random
@gvwilson@mastodon.social avatar

"Ten Simple Rules for Writing a Technical Book" https://journals.plos.org/ploscompbiol/article?id=10.1371/journal.pcbi.1011305 many thanks to my co-author Jess Haberman - hope people find it useful

warrickball,

@gvwilson

"Rule 6: Avoid common mistakes

  • Avoid banal advice (like “avoid common mistakes”)."

Nice. 👍

coreyspowell, to science
@coreyspowell@mastodon.social avatar

There's a dent in Earth's gravitational field, centered on the Indian Ocean, and nobody really knows why.

(Put another way: "There's a hole in the bottom of the sea. There's a hole. There's a hole...")
https://eos.org/science-updates/seismologists-search-for-the-indian-oceans-missing-mass

warrickball,

@coreyspowell One of my personal favourites, written by Douglas Gough for the proceedings of a conference in 1997:

https://ui.adsabs.harvard.edu/abs/1997IAUS..181..397G

warrickball, to Astronomy

Before I even applied for my current RSE role, I was asked if I'd like to briefly review some recent results in #asteroseismology for #Astronomy & Geophysics, the magazine of the Royal Astronomical Society (RAS). The article, which they dubbed "Finger on the pulse of asteroseismology", has now been published:

https://academic.oup.com/astrogeo/article/64/2/2.30/7077792

There wasn't nearly enough space (nor did I have time) to be exhaustive but it offers a personal view into the state of asteroseismology. Hope you enjoy!

warrickball,

At the suggestion of some friends and collaborators, my review of #asteroseismology for the April edition of #Astronomy & Geophysics, magazine of the RAS, is now available on arXiv:

https://arxiv.org/abs/2306.15070

It lacks the formatting and some extra images added by A&G's production team, so I recommend reading the published version if you have access (e.g. through your University).

  • All
  • Subscribed
  • Moderated
  • Favorites
  • •
  • megavids
  • khanakhh
  • mdbf
  • ethstaker
  • magazineikmin
  • cubers
  • rosin
  • thenastyranch
  • Youngstown
  • InstantRegret
  • slotface
  • osvaldo12
  • kavyap
  • DreamBathrooms
  • JUstTest
  • Durango
  • everett
  • tacticalgear
  • modclub
  • normalnudes
  • ngwrru68w68
  • cisconetworking
  • tester
  • GTA5RPClips
  • Leos
  • anitta
  • provamag3
  • lostlight
  • All magazines
    •