Emily Riederer

Scientists no Longer Find Twitter Professionally Useful, and have Switched to Bluesky Synopsis. Social media has become widely used by the scientific community for a variety of professional uses, including networking and public outreach. For

Bluesky is the new science Twitter, new study by @whysharksmatter.bsky.social and Julia Wester concludes!

"Results show that for every reported professional benefit that scientists once gained from Twitter, scientists can now gain that benefit more effectively on Bluesky than on Twitter."

Plenty that others will say about all the semi-dystopic Super Bowl ads, but Monday morning the one in my head is every employee was vibe coding their own apps

Sprawling ungoverned internal tool dev has notably worked out so well in corporate BI tools 😣

The Death of Direct File Tech for good not welcome in the Trump administration

Reminder when you watch the Intuit Super Bowl ads, they have Super Bowl ad money because they are a successful rent seekers who have worked for decades to kill free, quality, public tax preparation services donmoynihan.substack.com/p/the-death-...

Got #data ideas to share? Two awesome confs have CFPs ongoing through February!

@posit.co ::conf submissions open thru Fri 2/20 posit.co/blog/posit-c...

Applied ML Conf (feat @vickiboykis.com !) submissions open thru Su 2/22 appliedml.us/2026/

Once we start moving left I’ve got bigger problems

The weirdest thing about reading financial journalist is their love of the phrase “up and to the right”. I’m no time series expert but I guess I wasn’t worried about the “to the right” bit. Thought that was kind of non-negotiable

Congratulations on the release! I’m just hear to say the cover art choice is 💯

We've both got this! 💪

It is tragic that we are about 250mi too far apart to have coffee dates to panic…err I mean prepare

Congratulations!! Whoever they are they are so lucky!

Had a great time on @posit.co 's Test Set pod w/ @mchow.com @hadley.nz @wesmckinney.com!

We talk about moving between R, SQL, python and the strengths of different analytical tools for diff data tasks. You won't believe what proprietary language gets a shout-out (Stata!)

posit.co/thetestset/e...

Thank you so much for the kind words and for sharing!

"Tacit knowledge" is more charitable than how I describe my niche as "things too small or post-hoc obvious for anyone to care". Definitely a remnant of "growing up" during peak data twitter and being such a huge beneficiary of all the sharing

Thank you so much for taking the time to explain! That’s very interesting. My “understanding” was probably based mostly on the mere existence of “Bayesian trial design” textbooks and not the real world. Appreciate it!

For the curious but uninitiated, could you please say a few words about what part of this is new? I thought I’d seen plenty of books in the past about Bayesian methods for trials. Were they really not used / accepted for official FDA submissions at all?

Thanks for this! Missed the cast but absolutely love learning about others’ favorite extensions especially. Great Comments and Pastum look fantastic

Cannot confirm, but if it can help me get a seat I’ll read up!

Never underestimate the quest for knowledge no matter how irrelevant it may seem in the moment

If I hadn’t gone to a public university, I honestly wouldn’t have ever bothered to watch a basketball game. But now? “Boxing out” is a top ten skills for boarding public transportation

🤯 So fun! (Also “caveat” has to be my all time favorite whimsical font)

Which contains a comparison to renv a2-ai.github.io/rv-docs/refe...

Which contains a comparison to packrat rstudio.github.io/renv/article...

Its tricky stuff, folks

Weeks after writing about how the best ideas bleed btwn R and python, I find out from @stephenturner.us 's excellent blog that the #rstats world is attempting their own Rust-based analog to `uv` (for deps mgmt not packaging)

Haven't tried but neat to watch

a2-ai.github.io/rv-docs/

Thanks for reading! Coming out of that session, I totally expected it to be interesting and insightful, but it felt *important* in a way I couldn’t even articulate. Was fun to look back!

GitHub - socialfoundations/folktables: Datasets derived from US census data Datasets derived from US census data. Contribute to socialfoundations/folktables development by creating an account on GitHub.

I’ve not tried it but I quite like the idea of the folktables package which pulls micro data from the ACS for suggested regression and classification problems

However lots of Census fields more categorical which could be a trick for many intro methods

github.com/socialfounda...

In our 30s we get plausible deniability that we meant Hinge Health

Whoa new form of social anxiety just dropped

Liked this on LinkedIn and hopped over here to see if I could like it again. Thrill of the chase..

R + Python: From polyglot to crosspolination | Emily Riederer A combined reflection on 2025, posit::conf(2025), and the necessity of diversity in open source

Last 2025 post, on @python.org & @carpentries.carpentries.org values, a great @posit.co conf session feat @mchow.com @richmeister.bsky.social @davisvaughan.bsky.social, the magic of #rstats #python dev cultures sharing best practices, and random #rstats history

www.emilyriederer.com/post/py-rgo-...

It’s almost like shockingly good! Beautifully written — I tried to summarize it to someone today but it just didn’t work wit any other words

There Is No Mary Problem in ‘It’s a Wonderful Life’ George’s vision of his wife without him is essential to the film, but critics continue to miss its true—and profound—meaning.

Ha librarian part always bugged me too! Life goals

If you want to spend way more time thinking about it than needed, I just saw a “think piece” relitigating that very plot point

www.thebulwark.com/p/there-is-n...

Only Christmas movie I’ll watch every year! (And honestly realized last night it may have been the only movie I watched all year period)

Importance of community and causal inference is like the perfect rstats holiday movie (plus, bonus points for banking regulation, imho!)

I know of a 40,000 patient clinical trial where the SAS programmer determined the key secondary outcome as positive if Y=time to stroke < maximum follow-up time. Y=missing -> positive outcome because with SAS missing < anything. The error still stands.