#pypistats has been down for two weeks, but now PSF is working on taking stewardship over the service
github.com/crflynn/pypi...
#python #pythonnews
15.08.2025 19:08
👍 0
🔁 0
💬 0
📌 0
#pypistats has been down for two weeks, but now PSF is working on taking stewardship over the service
github.com/crflynn/pypi...
#python #pythonnews
AWS Cloud service users: Be careful with S3 bucket naming
A private empty S3 bucket just caused a 1300 USD bill overnight. Someone knowing your bucket name can generate PUT requests towards it and basically do a "Denial of Wallet" (DoW) attack against you.
youtu.be/OWggTcVgiNg?...
Heatmap showing data for years 2018-2022. One pixel per day, organized weekly.
dayplot looks like a great library for getting a yearly/monthly/weekly view on time-series data! 🚀📅
josephbarbierdarnal.github.io/dayplot/tuto...
#python #datascience #datavisualization
Faster* static type checking for Python is coming!
#python #redknot #red-knot
(*To be seen, but very, very likely)
#FreeJavaScript update: Oracle has informed us they won’t voluntarily withdraw their trademark on "JavaScript". Next: they’ll file their Answer and we’ll start discovery to show how "JavaScript" is widely recognized as a generic term and not controlled by Oracle.
Interesting. Do you know if there's some use case where this would be practical?
Just bumped up to FastHTML. It's a new python web framework for creating HTML apps with HTMX and ASGI.
www.fastht.ml
As a side note. Interesting that they chose to use package name "python-fasthtml" and not "fasthtml" (which is still available on PyPI).
#python #htmx #FastHTML
Spicy food leads to road fatalities! /s
Few other interesting ones at www.tylervigen.com/spurious-cor...
Image source: @cremieuxrecueil (X)
#datascience #datascientist #science #data #graphs #causation #correlation
Created a small python package for making monotonically increasing smoothing B-spines: github.com/fohrloop/pen...
#python #datascience #bsplines
Well said. The problem that data is not normalized ("per person" or otherwise) is actually pretty common.
Panel Graphic Walker
A new way for creating Tableau like data exploration interfaces with #python and just a few lines of code!
It's a basically a Panel (Holoviz) widget which brings the JS project graphic-walker into python.
github.com/panel-extens...
#datascience #dataanalysis #eda
Filtering data with few peaks, like a spectrum? Weighted Whittaker Smoother seems like a good alternative:
- Filters noise out effectively
- Can be adjusted on peak wavelengths to filter less
- Quite simple
eigenvector.com/wp-content/u...
#datascience #filtering #datafiltering
Just watched an astonishing video explaining splines both visually and at some level mathematically. The video covers Bezier curves, Hermite Splines, Catmull-Rom Spline, Linear Spline and B-Spline. No wonder it has 1.5M views!
www.youtube.com/watch?v=jvPP...
#math #splines #scientist #datascience
For Maths fans, 2025 is a square.
45² = 45 x 45 = 2025
Also,
9² x 5² = 2025
40² + 20² + 5² = 2025
My favourite?
1³+2³+3³+4³+5³+6³+7³+8³+9³ = 2025
#Mathematics #teaching #education
.. and I can't be *that guy* at home. Otherwise my peaceful moment at night with just me and my computer might be at danger. 😂
Awesome! I have the glove80 with Kailh Brown v1's and I'm just optimizing the layout for it. The Shhhocks are something that I definitely should buy! Did you try the keycap tilters with the glove80 thumbs?
#keyboards #ergomechkeyboard
variable names *are* code comments
What might be surprising, sometimes the unary addition does something, though (if the object implements __pos__(self))
#python
The slides for my lectures on (Bayesian) Active Learning, Information Theory, and Uncertainty are online now 🥳 They cover quite a bit from basic information theory to some recent papers:
blackhc.github.io/balitu/
and I'll try to add proper course notes over time 🤗
matplotlib tip: It's possible to move back and forward zoom levels in the GUI with left and right arrow keys (and C and V for left hand use).
That makes looking around data much faster (no need to hit the home button and start over)
#python #datascience #matplotlib #plots
Monte Carlo methods use random sampling to solve tough problems (optimization, integration, ...)
@13:55 When your simulations are expensive, you want to be clever about sampling (Quasi Monte Carlo) -> less samples, same info
@32:00 cleverness of different sampling methods compared
#datascience
You're absolutely right! The example is perhaps too simple, and many simpler cases should be handled with a @pytest.fixture instead.
Looks like a library that could be useful when writing unit tests!
It's for creating and updating inline snapshots in tests.
#python #pytest #pythontesting #unittests
Also just discovered that the download stats by python version has a bug: github.com/ClickHouse/c..., so for accurate information about it I would still use pypistats.org - But the full history is a killer feature. Compare to 6 months of pypistats or 3 months of (the free) pepy.tech.
Screenshot from the ClickPy showing two time series graphs of download statistics.
Just learned about ClickPy: an alternative to pypistats.org & pepy.tech — #python package download stats 🐍📊
Highlights:
- 📊 FULL history of the package
- 🔍 Filter by time range and see widgets updating (YES!😍)
- ⭐ GitHub star history
- 📦 Dependents
Example:
clickpy.clickhouse.com/dashboard/wa...
Today I learned about PSQL for DuckDB. It looks like SQL, but with none of the nested bits.
https://buff.ly/3ZV3p8I
Black is awesome, but nowadays I'm using ruff's formatter as it's faster and does the same thing. It can also replace isort, which is convenient.
#python #ruff #pythonruff
This would be perfect!
Nice visualization for telling the story! I love the one Asian country swimming out of the picture with more than 2.5 cell phone subscriptions per person.
Something seems to be off for sure 😄 The free-threading python is still in experimental phase though.
