iiinova — where your digital self keeps evolving

Riccardo Cappuzzo @riccardocappuzzo.com · 12/03/2026

What a world we are already living in www.adriankrebs.ch/blog/dead-in...

a screenshot of an email that reads "hey sorry - my agent got a mind of its own and started applying for jobs for me"

100

Riccardo Cappuzzo @riccardocappuzzo.com · 21/02/2026

And this is the result recorded with asciinema Selecting a file opens it in VS Code at the given line, very convenient

000

Riccardo Cappuzzo @riccardocappuzzo.com · 18/02/2026

Rabbit hole of the day: writing a command that fuzzy searches in the repository for any substring, shows me a preview of the line with context and opens the file at the given line in VS Code. Requires fzf, universal-ctags and batcat

the screenshot of a shell script that uses fzf and ripgrep to find substrings, classes, and files in the skrub repository

100

Riccardo Cappuzzo @riccardocappuzzo.com · 12/01/2026

Funny bug of the day: if you try to use pandas' "guess_datetime_format" with datetimes where the hour and minute are the same as the year (like 1959 and 19:59), the parser will fail and return None. This bug is present in pandas 2.3.3, but has been fixed in the dev version.

A short script demonstrating how the `guess_datetime_format` function of pandas does not work as intended when trying to parse the datetime "1959-01-01 19:59:16": it returns none instead of returning the correct datetime format.

100

Riccardo Cappuzzo @riccardocappuzzo.com · 09/10/2025

"ok the test run is done, let's see" ... "this will be hard to debug"

000

Riccardo Cappuzzo @riccardocappuzzo.com · 13/06/2025

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

010

Riccardo Cappuzzo @riccardocappuzzo.com · 20/05/2025

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.

050

Riccardo Cappuzzo @riccardocappuzzo.com · 19/05/2025

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

010

Riccardo Cappuzzo @riccardocappuzzo.com · 19/05/2025

⏱️ Complex aggregation methods are slower and don't significantly boost prediction performance. 6/

101

Riccardo Cappuzzo @riccardocappuzzo.com · 19/05/2025

⚖️ Beware of diminishing returns! Performance plateaus as more candidates are retrieved, while resource costs (time and RAM) keep rising. 5/

121

Riccardo Cappuzzo @riccardocappuzzo.com · 19/05/2025

🎯 Simple metric-based retrieval and candidate selection methods often outperform complex methods and are more efficient. 4/

101

Riccardo Cappuzzo @riccardocappuzzo.com · 19/05/2025

🔍 Good table retrieval is crucial helps finding candidates with useful features and fewer missing values. Jaccard containment is helpful but has its limits. 3/

101

Riccardo Cappuzzo @riccardocappuzzo.com · 19/05/2025

🌳 Tree-based models offer better prediction and computational performance than deep learning-based methods in our setting, which involves training models over features that contain a large fraction of missing values. 2/

121

Riccardo Cappuzzo @riccardocappuzzo.com · 19/05/2025

🌟 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/

263

Riccardo Cappuzzo @riccardocappuzzo.com · 02/05/2025

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

041

Riccardo Cappuzzo @riccardocappuzzo.com · 01/05/2025

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

A dictionary that contains various Muse albums and their release dates, and a suggestion for the release date of an album that hasn't been added yet

000

Riccardo Cappuzzo @riccardocappuzzo.com · 30/04/2025

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

032

Riccardo Cappuzzo @riccardocappuzzo.com · 14/12/2024

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

040

Riccardo Cappuzzo @riccardocappuzzo.com · 12/12/2024

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

110

Riccardo Cappuzzo @riccardocappuzzo.com · 27/11/2024

Some of these samples are deeply, deeply unsettling From fugatto.github.io

110

Riccardo Cappuzzo @riccardocappuzzo.com · 19/11/2024

This is a very simple example of what I am working with, only I have potentially thousands of lines like this. Looking at the documentation, it does look like I wouldn't need a lot of the features of SSSOM (and it might just add overhead in my scenario). Still, thanks for the clarification 👍

110