New tutorial: Bayesian Dynamic Path Analysis in @pymc.io Why do New Year's resolutions fail so reliably? The statistical answer involves time-varying causal effects and masked mediation. Full tutorial (code + math + Odyssey metaphor): nathanielf.github.io/posts/post-w... 🧵 Thread Below

Final blogpost of the year: in which I enthuse about Alex Meehan and Snow Zhang’s excellent recent paper ‘Bayes is Back’!

Big new blogpost! My guide to data visualization, which includes a very long table of contents, tons of charts, and more. --> Why data visualization matters and how to make charts more effective, clear, transparent, and sometimes, beautiful. www.scientificdiscovery.dev/p/salonis-gu...

The next online installment of ESTIMATE: The Reduced Form is coming on Dec 11-12. I've continued to unify and expand regression-based methods to apply to exit, non-binary treatments, DDD, discrete outcomes, and more. All proceeds to the MSU economics PhD program. econ.msu.edu/academics/es...

🧵 I spoke yesterday at @python.org's PyCon Ireland 2025 at UCD on Bayesian workflows + Structural Equation Models (SEMs) — exploring how we can infer latent preferences from noisy behavioural signals. youtu.be/NS7SjHDSdmA?...

TLDR; The PSF has made the decision to put our community and our shared diversity, equity, and inclusion values ahead of seeking $1.5M in new revenue. Please read and share. pyfound.blogspot.com/2025/10/NSF-... 🧵

This may just be the best CS paper I’ve read this year. Just read the abstract and first para of the intro! The rest of the intro is really wild too, but very very good: dl.acm.org/doi/pdf/10.1...

This is a very nice demonstration of the power of GAMs.

Books Upstairs is a fine store with a particularly excellent poetry section. Why not support them by ordering your books online instead: booksupstairs.ie

It was fun (painful 😅) to implement VAR(p) models from scratch juanitorduz.github.io/var_numpyro/

The Aaron Swartz production function www.programmablemutter.com/p/the-aaron-...

Very glad to see that someone is doing the important work of aligning AI alignment. alignmentalignment.ai

"Open-mindedness is not incompatible with having opinions."

Boosting this live tweeting of the excellent Emily Adlai’s new book to #philsci #philsky #quantum – thanks @jeffreymepstein.bsky.social for this public service! :)

I’m especially proud of this article I wrote about Gaussian Processes for the Recast blog! 🥳 GPs are super interesting, but it’s not easy to wrap your head around them at first 🤔 This is a medium level (more intuition than math) introduction to GPs for time series. getrecast.com/gaussian-pro...

```On Monday, Pirro told a judge in a separate federal criminal case in D.C. that “an Indictment has not been returned” after “a __third__ grand jury returned a no true bill.” The file drawer problem but for guilty verdicts

Market share predictions fail when models ignore how markets really work. At #PyDataBerlin, @nathanielforde.bsky.social shows how PyMC Marketing’s new Consumer Choice module uses Bayesian models for realistic substitution. 🗓️ Sept 1, 4:20 PM | 📍 Room B05-B06 #BayesianStats #PyMC

#rstats #statistics I've released my new open source book, "Powered by Linear Algebra: the role of matrices and vector space in data science," at matloff.github.io/WackyLinearA.... Turns the classic LA course on its head! Still proves the theorems, but with a deep emphasis on applications.

I'd be interested in reading an essay which traces a direct line from Krug's "Don't make me think" to Trump. en.wikipedia.org/wiki/Don%27t...

New Blog Post: VAEs vs Bayesian Multivariate models – modelling the complexity of job satisfaction. nathanielf.github.io/posts/post-w... It’s not one score. It’s a bundle: how we feel ❤️, how we work 💼, how we think 🧠. Here’s what happened when we tried to compress that complexity… 🧵

Looking forward to @pyconde.bsky.social in Berlin this Year! Will be speaking about @pymc.io -marketing and consumer choice models. Hope to see you there! berlin.pydata.org/conferences/...

🚀 Just added support for multinomial logit & nested logit product choice models to pymc-marketing! Use a formula API to model product choices & customer preferences — all in a Bayesian workflow 🧠📊 Docs: 🔗 l1nq.com/KRb64 🔗 encr.pw/oa49I

 @pymc.io #discrete-choice #bayesian #stats #causal-inference

New post: David Blackwell on the Comparison of Experiments

www.wheresyoured.at/the-era-of-t... This from @edzitron.com is great!

This is an excellent illustration of interaction effects with a (far too) resonant example.

Python Ireland just released their PyCon 2024 recordings! 📽️ My talk on Causal Inference with @pymc-labs.bsky.social CausalPy package is here. Can we trust individual IV designs. What's the role of CI in industry? Recording: youtu.be/-C4p4b2cUp8?... Deck: nathanielf.github.io/talks/pycon_...

Curious about Multi-level Bayesian Regression Models? Sceptical of Priors? Dubious about of weighted outcomes? Come along to the Python Ireland Meet-up in May, we'll talk survey data, population weights and choosing your own adventure with @pymc.io. www.meetup.com/pythonirelan...

Looking forward to this! Talking about the do-operator and structural causal models in PyMC

"Devise a system that scores philosophy papers on a scale of zero to ten based on how well their conclusions would withstand this sort of back-and-forth analysis and then apply it to Nguyen's paper." What could go wrong? www.truesciphi.ai/p/value-capt... #philsky

Powerful piece. The thread about the discomfort of thinking and the search for safety blanket like-assurances in prediction and asymptotic guarantees, especially resonates. So many cases where a bias for action sees any brittle assurance as a rubber stamp for the decisions we want to make.

I have a paper forthcoming in Ergo where I lay out my game-theoretic solution to the rule-following paradox: www.asgeirberg.org/papers/privi... It's a community solution and I argue that it is nevertheless objective; that the assumption that social accounts of meaning are not objective is false.

Big Irish Head - BT Young Scientist Awards The world is not ready for gen z. They are shitposters in real life.

Lots of folks are discovering the relevance of decision theory to the practical problem of analyzing an A/B. A quick🧵

1/N Some New Years reading to share! In this post, we dive into Cronbach's alpha, Fisher info, KL divergence, and Bayes factors as measures of item informativeness. We then use these metrics to reduce a large 100 item pool down to just 15 items while maximizing information 🤖

This is an excellent piece on the efficacy of policy innovation in large organisations. Sobering in a time when so many hope to simply "pivot to AI".

The distinction @jessicahullman.bsky.social makes between 1st and 2nd order uncertainty is important to understanding prediction vs inference. Conditional predictive performance (1st order) can be checked, but whether we’ve learned the right distribution (2nd order) is what we need to generalize.

This last 12 months or so was intense and fun both professionally and personally. Thanks owed @benvincent.bsky.social , @twiecki.bsky.social, @alex-andorra.bsky.social and all the pymc devs! Looking forward to 2025 @pymc-devs.bsky.social My #pymc, #opensource year-in-review:

On February 6th we will have a workshop on Structural Equation Models and the Do-Operator in PyMC by @nathanielforde.bsky.social More info: bit.ly/3BNn3tO Please share! #EconSky #Python #AcademicSky

This is an excellent piece on Netflix's impact on content creation and the incentives at play: www.nplusonemag.com/issue-49/ess...

Merry Christmas from all of us here at the Dublin City Library and Archive. We hope you have a wonderful Christmas filled with love, presents and fun. We look forward to welcoming you all back to the Reading Room in the New Year Merry Christmas 📷 Cuala Press

Had my mind expanded by this podcast on CFA SEM and causal inference with a Bayesian twist. Added bonus examples of employee disengagement and burnout!

"Bayesian analysis adds nothing to" . . . what process? You're talking as if there's a better way than Bayesianism. But in truth there's no alternative. Science has got into terrible trouble pretending that there is. www.the-tls.co.uk/regular-feat...

I have this artwork in my house and I'm not even sorry

🎙️ Episode #121 is live! 🎧 learnbayesstats.com/episode/121-... In this episode, @alex-andorra.bsky.social & @nathanielforde.bsky.social talk about: 📊 The flexibility of Bayesian approaches in modelling, key steps in model validation. ⚖️ Balancing model complexity, factor analysis and more...

🎧 Podcast Milestone 🗣 I recently had the pleasure of joining @alex-andorra.bsky.social on the @learnbayesstats.com podcast. We discussed SEM, CFA and People Analytics Surveys. Check it out here: youtu.be/6mL1J_b7_E0?...

Hard to overstate how revelatory this paper was for me — one takeaway I had from it was "small fixed costs can really deter people from optimizing." Which made me think about a few other recent papers where fixed optimization costs are a key ingredient:

I love all of these, though I'm not sure I'd use any of them if I were just introducing someone to causal methods. Here are a few I find myself recommending a lot:

#DublinCentral - if you want to add to your stress, watch the live tallies come in on this spreadsheet (hat-tip @breakaway-ian.bsky.social ) #GE24 Hutch now ahead of Lab, GP, PBP, FF, and Daly (but will fade when Cabra/Glasnevin boxes are opened (I hope)) docs.google.com/spreadsheets...