I want to share a very raw project, “NumPyro Forecast”. It is based on Pyro’s forecasting module docs.pyro.ai/en/dev/contr... (an actual port to Jax and NumPyro). I plan to extract many of the custom models I have from my blog and other places :) github.com/juanitorduz/...

Since more than 3 months I haven’t used Bluesky much…I am not finding stats Twitter from 2021 😔. I might just close the account and safe some phone battery …

Absolutely heartbreaking. So many mathematicians I know studied there, men AND women. Maryam Mirzakhani studied there. If this isn't enough to cancel/relocate from the US the International Congress of Mathematicians #ICM2026, nothing will be.

ArviZ 1.0 brings a lot of new plotting functionality. Check our gallery for a glimpse of what's new: python.arviz.org/projects/plo...

Statistical Rethinking 2026 is done: 20 new lectures emphasizing logical and critical statistical workflow, from basics of probability theory to causal inference to reliable computation to sensitivity. It's all free, made just for you. Lecture list and links: github.com/rmcelreath/s...

@rmcelreath.bsky.social I followed all the B course of statistical rethinking 2026. It was amazing! Thank you much for sharing! (I’m said it is over 🥲)

An MMM can predict perfectly, and still get causality wrong. This chart shows it: uncalibrated ROAS posteriors miss reality. Add one lift test → estimates snap back to truth. Join our MMM Calibration deep dive Webinar (Apr 8): dub.link/Nax7e44

Nutpie: state-of-the-art mass matrix adaptation for HMC statmodeling.stat.columbia.edu/2026/03/20/n...

ArviZ 1.0 is out! We have refactored it to be more modular, flexible & lightweight. For an overview of the changes, check the migration guide. python.arviz.org/en/stable/us...

I am super excited to join @pymc-labs.bsky.social full-time! I will be working on open-source projects, helping companies leverage Bayesian methods for decision-making, developing tailored educational workshops for industry practitioners, and serving as a product manager for our AI products.

Here are two examples on causal inference and through the lens of probabilistic programming languages (PPLs): - Introduction to Causal Inference with PPLs juanitorduz.github.io/intro_causal... - Causal Inference with Multilevel Models: juanitorduz.github.io/ci_multilevel/ Implementations in PyMC.

Here is the recording of my talk PyData Berlin 2025: Introduction to Stochastic Variational Inference with NumPyro Notebook: juanitorduz.github.io/intro_svi/ youtu.be/wG0no-mUMf0?... #pydata #berlin #bayes

Open Science and Open Source only with Diversity, Equity, Inclusion and Accessibility. Inclusion is essential to science, and science is only worthwhile if it lifts everyone up together. ropensci.org/blog/2025/02... #OpenSource #OpenScience

I got mail! I can’t not wait @vincentab.bsky.social I’ll try to do many of these examples by “hand” (learning by doing).

7 reasons to use Bayesian inference! statmodeling.stat.columbia.edu/2025/10/11/7...

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

Festival der Riesendrachen #Berlin

Kudos to @sethaxen.com for implementing the Pyro wrapper that makes this possible (shipped in sbi v0.25)! And thanks to @juanitorduz.bsky.social sharing the cookie factory example—it's a great accessible example for hierarchical inference. Everything runs in Colab 📊

A nice primer on normalizing flows by PyMC/PyTensor devs Ricardo and Jesse. pytensor.readthedocs.io/en/latest/ga...

Here are the materials for the PyData Berlin 2025 talk on Stochastic Variational Inference with NumPyro: - Slides: juanitorduz.github.io/html/intro_s... - Notebook; juanitorduz.github.io/intro_svi/

The ArviZ core devs have done tremendous work on an improved API with a lot of novel improvements. They have put together a great migration guide: python.arviz.org/en/stable/us... If you are an ArviZ user please take a look at it and provide feedback. Open source is all about the community 🫶

Los Amigos Invisibles #Berlin

I’ll be giving a talk ok variational inference (VI) at PyData Berlin 2025 🙂! I’ll focus on some learnings of using VI for forecasting models at scale. If you are around come and say hi. #PyDataBerlin

I used to experience this and it is fucking horrible 🫠

A beautiful read! @markhoppus.bsky.social #blink182

Forecasting isn’t just about prediction — it’s about decision-making under uncertainty. @juanitorduz.bsky.social shows how Bayesian models help: 🔹 Sparse data? Use hierarchies 🔹 Stockouts? Use censored likelihoods 🔹 Messy demand? Use priors + state spaces Practical guide 👉 dub.sh/prob-forecas...

Freedom of science and research is one of Europe's great strengths. It’s how excellence and innovation thrive. We’ll make proposals to help scientists and researchers ‘Choose Europe’. The best and brightest from around the world. 
 To make Europe the home of innovation again. → europa.eu/!JFF7jm

Here is a first intro notebook on Bayesian Power Analysis following the method described in the paper "The Bayesian New Statistics: Hypothesis testing, estimation, meta-analysis, and power analysis from a Bayesian perspective" (link.springer.com/content/pdf/...) juanitorduz.github.io/power_sample...

Here is a new blog post with PyMC Labs: "Probabilistic Time Series Analysis: Opportunities and Applications." We provide a collection of business cases where probabilistic methods excel in real-world applications of probabilistic time series methods. www.pymc-labs.com/blog-posts/p...

I decided to wrap these cohort modeling techniques in a little pre-print 🤗 arxiv.org/abs/2504.16216

Time Series Analysis with StatsModels Video from my PyData Global tutorial is up now: www.youtube.com/watch?v=foMb...

A new Python edition of "Forecasting: Principles and Practice" is now available online at otexts.com/fpppy/. Thanks to @azulgarza.bsky.social, Cristian Challu, Max Mergenthaler, Kin Olivares & Nixtla for making this happen. #forecasting #python

📚😅🎉 Yay!! I just submitted the complete manuscript of my upcoming book to the publisher! Learn to easily and clearly interpret (almost) any stats model w/ R or Python. Simple ideas, consistent workflow, powerful tools, detailed case studies. Read it for free @ marginaleffects.com #RStats #PyData

This is my current readingI had it in my mind for many years and I finally decided to read it! So far so good!

I’ll be attending the 3rd Vienna Workshop on Economic Forecasting 2025 where I will have a poster on “Probabilistic Forecasting at Scale with NumPyro” 🙂 www.ihs.ac.at/current/even...

At PyMC Labs I've been working with a group developing synthetic consumers for marketing research. We just published this white paper with an overview of work in this space -- and we have a blog post coming next week with some experimental results. www.pymc-labs.com/blog-posts/s...

We just launched our first white paper! By 2027, AI-generated consumers could power 50%+ of market research data. 📄 What’s inside? ✅ What synthetic consumers are ✅ How businesses use them and more.... 🔗 dub.sh/61AEavU 📩 Curious how this could benefit your org? 👉 [email protected] #GenAI

We just released PyMC-Marketing with a new multidimensional MMM class to allow for custom hierarchical model across multiple dimensions (e.g. geographies). You can specify hierarchical components easily so the opportunities are huge. www.pymc-marketing.io/en/stable/no... #mmm #pymc #marketing

Open source is draining. One's todo-list is open to world. People seldom realize the cost of what they get for free. Unpleasant comments do happen. Cost of maintenance is not understood. opensource.com/article/17/2...

It was suggested by @juanitorduz.bsky.social that I add some simulation proof to some claims I make in my blog post over at geteppo.com By all means, have some simulation proof (plus a little explanation of what I said re: pulling estimates apart) dpananos.github.io/posts/2025-0...

🔥🎉New library: boosting for survival analysis, including multiclass (competing risks) Survival = missing outcomes because limited observation window (common in medicine, marketting...) soda-inria.github.io/hazardous Gives very fast boosted-trees for survival

My favorites: Challenges and opportunities... proceedings.neurips.cc/paper/2021/h... Robust, accurate stochastic optimization... papers.nips.cc/paper/2020/h... ...improving the reliability of black-box VI jmlr.org/papers/v25/2... Yes, but Did It Work? ... proceedings.mlr.press/v80/yao18a.h...

Hi! Here is a question for all the #bayesian folks! Can you recommend references on evaluation and diagnostics of (stochastic) variational inference? Anything would be much appreciated 🙏

"Have directed acyclic graphs fullfilled their promise?" - the recording of my debate with @margaritamb.bsky.social at the World Congress of Epidemiology 2024 is now available on YouTube! www.youtube.com/watch?v=FG79... #EpiSky #CausalSky #WCE2025

𝐏𝐲𝐌𝐂-𝐌𝐚𝐫𝐤𝐞𝐭𝐢𝐧𝐠 𝐢𝐬 𝐠𝐫𝐨𝐰𝐢𝐧𝐠 𝐟𝐚𝐬𝐭. We’ve crossed 175,000 𝐭𝐨𝐭𝐚𝐥 𝐝𝐨𝐰𝐧𝐥𝐨𝐚𝐝𝐬 & now hit 20,000 𝐦𝐨𝐧𝐭𝐡𝐥𝐲 𝐝𝐨𝐰𝐧𝐥𝐨𝐚𝐝𝐬. 🤔𝐓𝐡𝐢𝐧𝐤𝐢𝐧𝐠 𝐨𝐟 𝐁𝐮𝐢𝐥𝐝𝐢𝐧𝐠 𝐈𝐧-𝐇𝐨𝐮𝐬𝐞? Developing a custom solution can take 𝐦𝐨𝐧𝐭𝐡𝐬, 𝐏𝐲𝐌𝐂-𝐌𝐚𝐫𝐤𝐞𝐭𝐢𝐧𝐠 gets you there 𝐟𝐚𝐬𝐭𝐞𝐫 📅 𝐒𝐜𝐡𝐞𝐝𝐮𝐥𝐞 𝐚 𝐅𝐫𝐞𝐞 𝐒𝐭𝐫𝐚𝐭𝐞𝐠𝐲 𝐂𝐨𝐧𝐬𝐮𝐥𝐭𝐚𝐭𝐢𝐨𝐧 : calendly.com/niall-oulton

PyMC-Marketing 0.11.0 is out! It is a huge release with more than 80 PRs from the community 🚀. Checkout the release notes: github.com/pymc-labs/py...

"Masculine Energy"

Its much more pleasant to write a latex article with a linter, git, girhub actions, and LLMs thank in a simple tex editor (10 years ago 😅)