My first actual talk in front of a ton of people 🙃

TIL about lava lamp encryption www.cloudflare.com/learning/ssl...

Do you have to deal with numerical features that involve large outliers, and need to train linear models or neural networks? Then you might want to try the skrub SquashingScaler. The SquashingScaler behaves like scikit-learn RobustScaler, but smoothly clips outliers to predefined boundaries.

Working hard on the next @skrub-data.bsky.social slide deck...

Today at #EuroScipy2025, @glemaitre58.bsky.social and I presented a tutorial on pitfalls of machine learning for imbalanced classification problems. We discussed what (not) to do when fitting a classifier and obtaining degenerate precision or recall values. probabl-ai.github.io/calibration-...

📢 Talk Announcement "Skrub: machine learning for dataframes", by Guillaume Lemaitre, Jérôme Dockès and @riccardocappuzzo.com. @skrub-data.bsky.social 📜 Talk info: pretalx.com/pydata-paris-2025/talk/T9KTPU 📅 Schedule: pydata.org/paris2025/schedule 🎟 Tickets: pydata.org/paris2025/tickets

Attending the @skrub-data.bsky.social tutorial by @riccardocappuzzo.com and @glemaitre58.bsky.social at #EuroScipy2025. They introduce the new DataOps feature released in skrub 0.6. Here is the repo with the material for the tutorial: github.com/skrub-data/E...

Heads Up, #Python Developers! There is an active phishing attack targeting PyPI users. • Threat: Emails from [email protected] (with a 'j') link to a fake login page. • Action: Do not click any links. If you already did, change your PyPI password ASAP. • Note: PyPI itself has not been breached.

Huge release, and the first one where I felt like I actually contributed a lot to the final result. I really think DataOps are a game changer, and I can't wait to see what people come up with with them. I also ended up rewriting most of the user guide, hopefully improving it along on the way 😂

I got 16/26 on the fstrings.wtf quiz. Can you do better? fstrings.wtf

David R. Hagen just solved a small mystery that I mentioned 13 years ago in the mouseover text of a comic drhagen.com/blog/the-mis...

Really cool graffiti I spotted while walking around in the town where I live

classic arstechnica.com/ai/2025/06/g...

Out of all the features in the expressions, this may be my personal favorite. I always end up adding too many configurations just because the syntax is so convenient.

📝 The skrub TextEncoder brings the power of HuggingFace language models to embed text features in tabular machine learning, for all those use cases that involve text-based columns.

The Skrub Cleaner is a lightweight transformer that performs consistency checks on a dataframe: 🔍 It gives a uniform representation of null values, converting those represented as strings (such as "N/A") 🗑️ It drops columns that contain too many null values (according to a user-defined threshold)

Now that the paper is out, I can finally share the totally-not-confusing script/plot/table map I made to track which scripts prepare which figures and tables and from what data. If it wasn't clear, don't do this. If you *really* have to, I used the @obsidian.md canvas for this.

Work done with @gaelvaroquaux.bsky.social and @papotti.bsky.social

A bit of a mess up with this figure! This is what it's supposed to look like 🙈

🌟 New paper alert! 🌟 Our paper, "Retrieve, Merge, Predict: Augmenting Tables with Data Lakes", has been published in TMLR! In this work, we created YADL (a semi-synthetic data lake), and we benchmarked methods for augmenting user-provided tables given information found in data lakes. 1/

More fun digging around my @last.fm scrobbles using with @matplotlib.org I had no idea how much of a difference changing fonts and background color could make

I haven't been using a lot of Copilot until very recently, so I'm still learning what it can do. It just blew my mind by autocompleting the dictionary "release_dates" with the correct dates for Muse albums based on the fact I am looking at data about Muse in the script. wow

First experiment plotting my Last.fm scrobbles With 10 years worth of data, I'll be working on this for a while. Also first time working with @matplotlib.org stackplots, much finagling was involved

👀 This week's post is a sneak peek into the next major Skrub feature, Skrub expressions 🚀 As this is a preview of an upcoming feature, we are looking for your thoughts and feedback before release.

Biggest commitment to a 3 second joke I've ever seen

And if you're not familiar with what Skrub is all about, you might want to check out our introductory slide deck here: skrub-data.org/skrub-materi...

🚀⚡ Release: 0.5.3 Check out the release notes: skrub-data.org/stable/CHANG... Highlights below ⤵️

My first blog post for Skrub 🎉 Not pictured, a full day of time and memory profiling to spot the difference between two slightly different StringEncoder PRs

Trying something new: A 🧵 on a topic I find many students struggle with: "why do their 📊 look more professional than my 📊?" It's *lots* of tiny decisions that aren't the defaults in many libraries, so let's break down 1 simple graph by @jburnmurdoch.bsky.social 🔗 www.ft.com/content/73a1...

FBI Uncovers Al-Qaeda Plot To Just Sit Back And Enjoy Collapse Of United States

OpenAI right now

My first major contribution to Skrub has now been released 🎉

Discover the CIA’s Simple Sabotage Field Manual: A Timeless Guide to Subverting Any Organization with “Purposeful Stupidity” (1944)

Donate now, if anything for the added bonus of sticking it to the muskman

More shady practices from OpenAI, why am I not surprised www.lesswrong.com/posts/8ZgLYw...

OpenAI funded FrontierMath, forbade them from disclosing it before o3 was revealed, and had access to the problem set excluding a separate hold out blind set. They updated their Arxiv to reflect this after Dec 20th 👇 www.lesswrong.com/posts/cu2E8w...

On today's episode of "why is this a thing and why do I have to waste time to find a fix": If you open a folder through a symlink in VSCode, there is the chance that the git indicators for changes in a file may stop working, and the fix is opening the folder directly. github.com/microsoft/vs...

I was wondering, is it (still?) possible to jailbreak LLMs that let you upload PDFs by hiding prompts in the text? Or is the text parsing is sanitized enough for that to not happen?

I feel better about seeing gymnasts as "morphing alien shoggoths" when they were doing their routines and bending in shapes in ways I never thought could be achieved by a human body arstechnica.com/information-...

Slides for "Table Foundation Models" I explain why these models can strongly outperform tree-based models, what are the intuitions, hopefully pointing to ways forward for more improvement speakerdeck.com/gaelvaroquau...

One good thing about living in Paris is that, well, you're living in Paris.

The good thing about running experiments in France is that I can feel slightly less guilty about my emissions, but still yikes

Today I had to work with SQL again for the first time in some time, and I felt a tinge of nostalgia of the begone days of my bachelor's. Yep, even after all those years, my hate for SQL hasn't changed a bit.

Well this is grim

We had a great time at the @pyladiescon.bsky.social #PyLadiesCon 2024! Over the course of the mentored sprint we were able to merge five challenging PRs that will be added to the next release, and welcomed three new contributors 🎉🎉🎉 See you there next year!

Any suggestion for a light theme (I know, I know) for VSCode? I'm currently using Quiet Light, but I'm open to suggestions

Sister's Christmas cat

This is a far more nuanced and informative thread than what I posted earlier on the use of social media data for academic work (and more), definitely worth reading