Lightnews — Scholar-powered news

Reposted by Thomas Lumley

Nic Rawlence @nicrawlencenz.bsky.social · 2h

Here we go again 😡. I'm really concerned about this & that the mistakes of the previous disastrous restructure seven years ago will be repeated. A museum is much more than its buildings. It's people that matter. Is central government going to properly fund Te Papa? www.rnz.co.nz/news/nationa... 1/3

Te Papa announces plans for restructure in an effort to cut rising costs

It comes after the museum started charging international visitors to the site due to rising costs of energy, insurance and staffing.

www.rnz.co.nz

1 2 4

Thomas Lumley @tslumley.bsky.social · 8h

Someone needs to explore the boundary between the names. Tryptophan Cloudberry is clearly Space Twink and Bandicoot Calabash is the other guy, but can we find a sort of Buridan colloquy that wavers between them.

2

Thomas Lumley @tslumley.bsky.social · 11h

The sedevacantists may be nuts, but at least they're nuts in an internally coherent way

1

Thomas Lumley @tslumley.bsky.social · 15h

There's a family of geneticists who make it rhyme with 'sang'

1

Reposted by Thomas Lumley

Jörg Vogel @jorg-vogel-lab.bsky.social · 1d

Two group leader positions available in the broader areas of RNA science, RNA technologies, and RNA medicine. Attractive packages and a great environment. Come and join us at Helmholtz RNA Würzburg, Bavaria.

72 61

Thomas Lumley @tslumley.bsky.social · 1d

*speed advantage

1

Thomas Lumley @tslumley.bsky.social · 1d

Even in 2001 I was saying "the advantage of the *apply functions is legendary: that is, partly history and partly fiction"

1 5

Thomas Lumley @tslumley.bsky.social · 1d

It's probably slow because of the nested loops, but leaving the braces off won't make it any slower

3

Thomas Lumley @tslumley.bsky.social · 1d

It's a cognitive dissonance thing for me

On one hand, I hate how the server knows perfectly well that I have access because who else will want a math stat paper from 1973 but I still have to do the whole login dance

On the other hand, I remember having to walk across campus to photocopy the paper

1 5

Thomas Lumley @tslumley.bsky.social · 1d

"writ in burnished rows of steel", too

Reposted by Thomas Lumley

Kieran Healy @kjhealy.co · 1d

“Habeas corpus” is what Sam Altman joyfully shouts when he thinks he has finally secured all the training-data legal rights to other people’s copyrighted work.

3 12 80

Thomas Lumley @tslumley.bsky.social · 1d

Middle Earth

Thomas Lumley @tslumley.bsky.social · 1d

Middle Earth

3

Thomas Lumley @tslumley.bsky.social · 1d

Ok, now I know who someone was subtweeting

1

Thomas Lumley @tslumley.bsky.social · 1d

Train cancelled for the second day in a row. Despite the 1-hour gap between trains, AT Mobile app still says "Good Service" on the Onehunga line

1 1

Reposted by Thomas Lumley

Robert “The Bobby Yaga” McNees @mcnees.bsky.social · 1d

Astronomer Ejnar Hertzsprung was born #OTD in 1873. 🧪 🔭

H-R diagram:

🔵 🟦 🟥
🔵 🟧
⚪️ 🟨
⚪️
🟡🟡
⚬ ⚬ 🟠🟠
⚬ ⚬ ⚬ 🔴
⚬ 🔴
⚫️

3 42 170

Reposted by Thomas Lumley

Russell Brown @publicaddress.bsky.social · 2d

Fucksake. It's quite clear now that they only looked at this benefit policy as political marketing and took no advice on who it would impact – which includes young people with serious health issues and disabilities. Just feckless and cynical. Shame on them. www.rnz.co.nz/news/politic...

Hundreds of teens with a health condition, disability may be cut from Jobseeker benefit

The government is ending Jobseeker payments to 18- and 19 year olds whose parents earn more than $65,000.

www.rnz.co.nz

37 150 460

Reposted by Thomas Lumley

Michael Friendly @datavisfriendly.bsky.social · 2d

Great post by @andrew.heiss.phd on animating #rstats {dplyr} operations

www.andrewheiss.com/blog/2024/04...

Visualizing {dplyr}’s mutate(), summarize(), group_by(), and ungroup() with animations | Andrew Heiss

Visually explore how {dplyr}’s more complex core functions work together to wrangle data

www.andrewheiss.com

1 8 53

Thomas Lumley @tslumley.bsky.social · 2d

Halloween coming up soon. Maybe you'll find out then why the garden is unexpectedly fertile.

5

Thomas Lumley @tslumley.bsky.social · 2d

I've only read them via Nils Hjort's book

1

Thomas Lumley @tslumley.bsky.social · 2d

Also too, research impact statements

From Matt Levine:
I used to be a corporate equity derivatives investment banker, and I once watched with uncomfortable awe as my boss explained that a derivatives deal we did for a biotech company lowered its cost of capital and thus allowed it to develop a promising cancer treatment, so if you really thought about it we were basically curing cancer. Did I believe that? Did he believe that? I mean. Perhaps his grandparents did.

1 4

Thomas Lumley @tslumley.bsky.social · 2d

There is a bit of a gray area in citations that are purely to give prior credit in academia. I haven't read Bonferroni on the union bound in probability or Takeuchi on model-agnostic versions of AIC (because languages). I'm not citing them as evidence for the propositions they contain, though.

1 3

Thomas Lumley @tslumley.bsky.social · 2d

The standard in law (which lawyers are actually being sanctioned for violating) is that including a citation is an explicit claim that you, *personally*, have checked the citation, it says what you are claiming, and it hasn't been overridden by later decisions.

Olivier Simard-Casanova @o.simardcasanova.net · 2d

It's not the first time that I'm seeing this

I'm afraid that hallucinated citations is an issue that scientists and experts will have to deal with from now on

Ulrike Franke @rikefranke.bsky.social · 3d

And here we go. I never wrote this article, and yet it is cited here.

www.liberalbriefs.com/geopolitics/...

And of course, it sounds so plausible, I seriously checked whether I had forgotten it, or the footnote was slightly wrong.

#AIisnotresearch

2 4 16

Reposted by Thomas Lumley

John Pfaff @johnpfaff.bsky.social · 2d

Not the main point, but when you're missing ~20 yr of data, I think you should just leave it blank.

Iit could be that it rose for the next five years, then plummeted and kept going! Maybe it's roughly linear! Maybe something else.

I mean, it's rising in the 1970s: you sure it pivoted RIGHT THERE?

Charlotte Clymer @charlotteclymer.bsky.social · 2d

Gallup has been surveying Americans' trust in legacy media since 1972. This year, a mere 28 percent of the public have "a great deal" or "a fair amount" of trust in legacy media, its lowest level in more than five decades of tracking.

news.gallup.com/poll/695762/...

3 10 48

Thomas Lumley @tslumley.bsky.social · 2d

survey::rake does iterative proportional fitting but it might not be exposed in a way you can use it

2 1