Thomas Lumley

Beautiful. I must go back to Tiritiri again sometime soon

My guess is if you write fairly boring code and don't get too clever, it's likely to keep working with feasible levels of maintenance.

... On the minus side, low-level code makes you subject to hardware (Apple ARM, RISC-V) or language (C++ standard) changes as well as R changes. For example, Apple ARM defaulted to not distinguishing NA and NaN and RISC-V completely won't.

It depends on your time horizon and how much maintenance you're willing to do. On the plus side, R Core is pretty conservative about backwards compatibility and will try not to break a package like Rcpp with lots of dependencies, especially if it's not being actively maintained...

Which education results?
(I don't have Twitter any more)

I'm following the government accounts because it's literally my job to monitor what government agencies say.

We are really excited to have one or more slots this year. I'm personally very aware of how much incredible talent is coming up on the market -- the fact that we can make one or more hires this round is great news for us. Juniors are what make a department exciting, relevant and good. Apply!

Don't we already have "truthiness" as a word for that concept?

The only one I've noticed was for the student hardship fund, but I might just have ignored other solicitations

Climate Analyst Climate Analyst

New Zealand - Aotearoa folks and otherwise interested folks: The Auckland Council has an opening for a Climate Analyst careers.aucklandcouncil.govt.nz/job/Central-...

From the applicant perspective, let me add one more: a lack of clarity in postings about exactly when and in what stage letters will be requested!

I schedule my time to put in applications trying to take into account lead-time for my letter writers but I can't do that if the postings aren't clear!

Writing letters of recommendation is a responsibility that I take very seriously.

But though I have my "system" for keeping on top of the many formats and scattered deadlines, I live in a constant, low-level panic that I'm going to miss one.

The current practice for academic LORs is a TRAINWRECK.

Rare traits like support for violence, conspiracy beliefs, and unsafe health behaviors are really hard to measure! On Thursday November 13, learn from experts about the problem and solutions.

Free for students and AAPOR members, $5 for others. Sign up here: www.eventbrite.com/e/the-rare-t...

Counterpoint: Reasonable people dismiss concerns about subway crime because *those concerns are factually baseless.*

That's a very good reason to dismiss them!

Flat, graphic illustrated poster entitled "Red-Eyed Birds of North America" featuring shadowy gray-blue silhouetted birds punctuated by bright red eyes. 23 birds are depicted in total, some highlights being the Snail Kite, Roseate Spoonbill, Common Loon, and Wood Duck.

Bro! Tell me we still know how to talk about sports! In the old days,
everyone saw that golf sucks: posh, prolonged, PGA-owned.

The place where people drop of their votes after voting at home and sealing their envelopes? Probably not an especially high risk of voter interference.

Another one for the list might be "Thorne–Żytkow object", for some sort of demi-metal band

It's a low-level offense that would typically get tried without a jury -- the penalties are lower and the judicial process faster and less expensive than for fraud per se.

I've been looking through this as an online resource, and I think it is a nice resource for those out there handling ecology type data. Well done on the authors.

They aren't claiming to do it by woo, which is what's banned (this whole thread started with witchcraft laws)

In NZ fortune-telling is only illegal for money (and only when done with intent to deceive)
www.legislation.govt.nz/act/public/1...

NHANES, too?

"Ford: who will you truckle to?"

The Role of Congeniality in Multiple Imputation for Doubly Robust Causal Estimation | Lucy D'Agostino McGowan This talk provides clear and practical guidance on the specification of imputation models when multiple imputation is used in conjunction with doubly robust estimation methods for causal inference. Th...

And for my next talk, “The Role of Congeniality in Multiple Imputation for Doubly Robust Causal Estimation” #IDWSDS2025

In which I insist imputation models MUST include:

✅outcome
✅exposure
✅all in PS model
✅all in outcome models
*in the same functional form*

🎞️: www.lucymcgowan.com/talk/idwsds_...

Opening for an epidemiologist at my organisation, the Pacific Community (SPC). Based in Suva; not in my Statistics division but in the Public Health Division next door. Applications close 26 October Fiji time. careers.spc.int/job/epidemio...

Big shakeup for NZ's research funding - Expert Reaction - Science Media Centre The Government will consolidate decision-making for New Zealand's science research funding as part of its new science system strategy. Research Funding New Zealand will be a new independent board that...

Big shakeup for NZ’s research funding – Expert Reaction

The film didn't exist when I were young

I think the only two assigned high school authors that I went out to read more of at the time were Shakespeare and Stoppard.