Posts on Potentially Useful

I recently learned how to use some newer Git features (partial clones and sparse checkout with sparse indexes) to effectively wield a brand new “monorepo” for my team’s data science work.

I was proposing my doctoral dissertation fifteen years ago when neural network machine learning models entered a second Renaissance.

They say it takes six weeks to form a habit, but as far as I can tell, the research suggests an average closer to two months—depending on the habit being undertaken and the individual being studied.

I’m not making as many data graphics at work as I had in previous positions, and I sometimes miss it.

At work I’m helping my team transition a large code base (really, a collection of code bases) from a “legacy” closed-source data analysis tool to a popular “big data platform” which presents an interface based on computational notebooks.

Most folks who would ever read this weeknote probably already have a good understanding of how LLMs work, but this blog post from Josh Sharp is a great explainer I foresee sharing with others.

I learned a lot from this piece on “Expert Generalists”, certainly (at least in part) because I identify as such.

Microsoft will stop supporting Windows 10 after October this year.

This will be brief, but I’m proud of this one.

I have become obsessed with this puzzle website: Every 5x5 Nonogram.

Just a QR Code It does exactly what it says on the tin.

It’s all stats and data science for this post folks!

I broke my streak of “weeknotes”1 and missed last week’s update.

Making more informative maps in R I love making maps in R—I was the go-to map-maker a couple jobs ago, and I believe that making a decent-looking map in a take-home coding assignment helped me land my current gig.

Managing reproducible environments in R I saw a toot on Mastodon bemoaning the experience of using renv to manage “reproducible environments” in R.

FFmpeg’s ASM lessons Assembly language (specifically the x86 instruction set) was actually one of the first “programming languages” that I learned, having come to computers through engineering.

The whys and hows of alt text This piece nominally targets journalists posting to social media, but it does a great job of explaining why alt text is important and has guidelines for composing good descriptions.

This post is a major departure from what I usually write.

pyOpenSci statement on DEIA A couple weeks back I shared a statement from rOpenSci reaffirming their commitment to diversity, inclusivity, equity, and accessibility.

Multiprocessing in Python is pretty broken A common scenario: you have some code Python that’s taking too long.

How to move R projects to Codeberg Last week I shared a link collaborating with Git and no code forge, but now I want to share a new (to me) resource for doing R on Codeberg, “a democratic community-driven, non-profit software development platform.”

rOpenSci statement on DEIA A thing that drew me to switch to R as my primary programming language for data analysis 11 years ago was the community of R enthusiasts that I found online1; people were friendly and inclusive, and the vibes—in general—were exc…

Personal websites are in This week’s zeitgeist brought a bunch of new (to me) stuff about indie and personal websites to my attention.

Purrr goes parallel Literally the only thing I miss from my days of programming MATLAB is parfor, which made it dead simple to massively hasten slow but parallelizable algorithms.

Astroterm: a terminal-based star map I walk my dog after dark most evenings, and I enjoy the excuse to look at the sky and try to identify the planets and constellations I can.

Tinyplot 0.3.0 is on CRAN Tinyplot is “a lightweight extension of the base R graphics system, with support for automatic grouping, legends, facets, themes, and various other enhancements.”

Constructive pessimism beats false hope Do you feel like the state of the world is bad, is nearly certain to get worse, and that there’s virtually nothing you can do to change any of it?

uBlock Origin is more than an ad-blocker This post from my friend @slgr@scholar.social does a great job explaining how useful uBlock Origin is for more than just blocking ads.

Fight Flash Fraud I came into possession of a fraudulent USB “thumb” drive, fully aware of what I was getting.

1001 Album Generator In 2005, a panel of music critics published the first edition of 1001 Albums You Must Hear Before You Die.

Happy Public Domain Day!

Henry Leach — Lego Trains on Brio Track My youngest child is three and obsessed with the toys in question, so when Henry posted a picture of a train made of LEGO-style plastic bricks running on Brio-style wooden tracks, I had to know his secret.

Yesterday was dark and chilly, demanding a one-pan casserole for dinner.

When I present data to non-technical stakeholders, I sometimes express differences (typically changes over time) in terms of percentage change.

I have added a now page to my site and a link to it in the navigation footer below.

The complexity of our world is beyond the limits of human comprehension.

To view a list of vignettes associated with a package (for example, dplyr), enter the following command in any R console: vignette(package = "dplyr") You can also use: browseVignettes("dplyr") To see a specific vignette, use: vignette("dplyr", package =…

This is just a collection of interesting (and not necessarily recent) Git posts from around the web.

SourceHut, a hacker1-oriented code forge, recently suffered a couple days’ downtime due to a sustained DDoS attack, and parts of it are still coming online.

The R package ecosystem has several packages1 that were created to help with the design and layout of tables in documents. condformat (Code forge) DT (Code forge, Website) flextable (Code forge, Website) ftextra (Code forge, Website) formattable (Code for…

Today, I followed these instructions and “hibernated” my LinkedIn account.

I was wondering how to incorporate a test for equality into a pipeline using R’s new(ish) native forward pipe operator1, so I asked the Fediverse and got great advice2.

I participated in December Adventure this year and posted nearly daily updates on this page.

I needed a few pages of graph paper, and it seemed easier and more economical1 to print some at home than to buy a whole pad of it.

This is a departure from the things I normally post here, but I want to put an important PSA on the internet… and hopefully prevent some folks from getting ripped off.

One of my top pet-peeves on internet discussion and help boards (dating back to the Usenet days, covering the web forum era, and continuing in this current age of Stack Overflow hegemony) occurs when a person replies to a request for help or information w…

I have a Python script that needs to be run from within a specific Anaconda environment (which I’ll call environment_name) for reasons that aren’t relevant here.

Andy Bell from Set Studio shared the outcome of a study of browser viewport sizes; the area of the screen (measured in pixels) available for web pages.

Every nerd knows that ISO 8601 is the “correct” way to display dates, but those who dig into the standard know that it encompasses more than YYYY-MM-DD format1.

Google is getting worse1.