#python

A neat collections of tips by Tommy Odland to avoid common linear algebra mistakes when using NumPy. Should be a useful read or refresher for all but the most expert of NumPy users.

We are done with a lovely 3-day sprint in Vienna this August. Next to the Prater, with a view of people being flung via slingshot into the sky, we tinkered with a strategy to improve the API of scikit-image in difficult places.

I’m looking forward to an exciting week at the EuroSciPy in Szczecin – one of my favorite conferences! scikit-image will host a tutorial on image processing. And on the weekend afterward, we’ll have a rare in-person scikit-image sprint with multiple maintainers on site!

Last month, I got to enjoy the SciPy 2024 conference again and had a great time meeting people, discussing scikit-image and scientific Python in general there. Here’s my slightly delayed recounting of it.

During the last two weeks, I had the good fortune to attend two open source related events: a very collaborative developer summit in Seattle and a very educational meeting on open science in Boston.

Patterns and tips around Python’s dataclass that may not be obvious when first using the feature.

Python libraries usually alert users in advance, when they use arguments that will be removed in future versions. Using a simple placeholder as a temporary default value for deprecated arguments is often the simplest and most expressive way to implement this.

The stacklevel argument is important to determine which line of code is shown as the origin of a warning and how easy it is to understand its cause. If warnings are meant for your users, then they are part of your public API, and it is a good idea to check their origin in your test suite. Here’s how.

With the release of NumPy 2 getting closer, there are a few changes that maintainers of Python code might have to address. This post lists a few takeaways I collected while we prepared scikit-image for the new API.

When a GitHub label is deleted, it is removed from every issue and pull request that it was associated with. Unfortunately, it is not as easy to revert this action if the label was accidentally deleted. I recently had to figure out an automated way to restore the label for 100+ issues, as I certainly didn’t want to do so by hand. So here is a quick “how-to” in case you ever encounter a similar problem.

Four really cool interactive articles (4x ~10 min) by Alexander Putilin that explain how Python’s dictionary works under the hood. Very easy to follow as concepts are introduced successively, visualized and put into context.