Grant McDermott
LightNews LightNews
gmcd.bsky.social
Grant McDermott
@gmcd.bsky.social
5.1K followers 160 following 660 posts
Economics and data science | Former academic now working in the tech sector | Views my own | 🇿🇦 in 🇺🇸 | https://grantmcdermott.com/
Posts Replies Media Videos
Pinned
gmcd.bsky.social
Grant McDermott @gmcd.bsky.social · Aug 23
A few months ago, I gave a workshop on “(Pretty) big data wrangling with DuckDB and Polars”.

Slides, notebooks etc. are all available here: grantmcdermott.com/duckdb-polars/

#EconSky
We just read a ~180 million row dataset (from disk!) and did a group-by aggregation on it. In < 1 second. On a laptop. Benchmark plot showing minimal data I/O cost of duckdb and polars relative to other options (alongside very fast compute time)
gmcd.bsky.social
#rstats In very non-Superbowl news, here's an experimental 📦 for nice graphics display and UX in VS Code (and other potential front-ends).

github.com/grantmcdermo...
GitHub - grantmcdermott/jgd: Lightweight R graphics device
Lightweight R graphics device. Contribute to grantmcdermott/jgd development by creating an account on GitHub.
github.com
February 9, 2026 at 12:47 AM
gmcd.bsky.social
#rstats I've been testing out `arf`, a new R console written in Rust. Early days, but a v. smooth experience so far. Much simpler to install and link from VS Code than radian, for example, with better features too. github.com/eitsupi/arf
GitHub - eitsupi/arf: Alternative R Frontend — a modern R console written in Rust
Alternative R Frontend — a modern R console written in Rust - eitsupi/arf
github.com
January 31, 2026 at 4:48 PM
gmcd.bsky.social
Please consider citing if you are a fixest user!
paperposterbot.bsky.social ArXiv Paperboy (Stat.ME+Econ.EM) @paperposterbot.bsky.social · 11d
arXiv📈🤖
Fast and user-friendly econometrics estimations: The R package fixest
By Berg\'e, Butts, McDermott
fixest is an R package for fast and flexible econometric estimation, providing a comprehensive toolkit for applied researchers. The package particularly excels at fixed-effects estimation, supported by a novel fixed-point acceleration algorithm implemented in C++. This algorithm achieves rapid convergence across a broad class of data contexts and further enables estimation of complex models, including those with varying slopes, in a highly efficient manner. Beyond computational speed, fixest provides a unified syntax for a wide variety of models: ordinary least squares, instrumental variables, generalized linear models, maximum likelihood, and difference-in-differences estimators. An expressive formula interface enables multiple estimations, stepwise regressions, and variable interpolation in a single call, while users can make on-the-fly inference adjustments using a variety of built-in robust standard errors. Finally, fixest provides methods for publication-ready regression tables and coefficient plots. Benchmarks against leading alternatives in R, Python, and Julia demonstrate best-in-class performance, and the paper includes many worked examples illustrating the core functionality.
January 30, 2026 at 8:09 PM
Reposted by Grant McDermott
paulgp.com
One of the most important statistical packages made in Econ in the last decade
paperposterbot.bsky.social ArXiv Paperboy (Stat.ME+Econ.EM) @paperposterbot.bsky.social · 11d
arXiv📈🤖
Fast and user-friendly econometrics estimations: The R package fixest
By Berg\'e, Butts, McDermott
fixest is an R package for fast and flexible econometric estimation, providing a comprehensive toolkit for applied researchers. The package particularly excels at fixed-effects estimation, supported by a novel fixed-point acceleration algorithm implemented in C++. This algorithm achieves rapid convergence across a broad class of data contexts and further enables estimation of complex models, including those with varying slopes, in a highly efficient manner. Beyond computational speed, fixest provides a unified syntax for a wide variety of models: ordinary least squares, instrumental variables, generalized linear models, maximum likelihood, and difference-in-differences estimators. An expressive formula interface enables multiple estimations, stepwise regressions, and variable interpolation in a single call, while users can make on-the-fly inference adjustments using a variety of built-in robust standard errors. Finally, fixest provides methods for publication-ready regression tables and coefficient plots. Benchmarks against leading alternatives in R, Python, and Julia demonstrate best-in-class performance, and the paper includes many worked examples illustrating the core functionality.
January 30, 2026 at 5:03 PM
gmcd.bsky.social
ICYMI with a quick screenshot as sales pitch. Simply feed your estimation a bunch of .parquet files on disk and run it as normal.

This example takes 180m rows of NYC taxi data and computes the model in < 1 second.

On a laptop.
dbreg example showing how to estimate a model against a large dataset, comprising parquet files on disk. The model takes < 1 to run, including the time needed to read the data from disk. dbreg( tip_amount ~ fare_amount + passenger_count | month + vendor_name, path = "read_parquet('nyc-taxi/**/*.parquet')", vcov = ~month, # clustered SEs strategy = "compress" # skip auto strategy overhead )
January 28, 2026 at 5:20 PM
gmcd.bsky.social
Things are grim. But in more frivolous news...

@jamesbrandecon.bsky.social and I have been chipping away at `dbreg`, a 📦 for running big regression models on database backends. For the right kinds of problems, the speed-ups are near magical.

Website: grantmcdermott.com/dbreg/

#rstats

[1/2]
dbreg
grantmcdermott.com
January 26, 2026 at 4:57 PM
gmcd.bsky.social
Closing the work week out with Minneapolis's finest. Solidarity and massive respect. youtu.be/pbbIAwAqBTk?...
Bastards of Young
YouTube video by The Replacements - Topic
youtu.be
January 24, 2026 at 3:33 AM
gmcd.bsky.social
With the fine weather this MLKD weekend, I took the opportunity to do something economically useful: digging holes and filling them back up again.
A post in hole Same post and hole, now filled with concrete
January 20, 2026 at 12:55 AM
Reposted by Grant McDermott
paulgp.com
Great cogent post on the issues of agentic coding (Eg our buddy Claude Code) and the spillover costs on everyone else

lucumr.pocoo.org/2026/1/18/ag...

@gmcd.bsky.social hits on the things you were complaining about
Agent Psychosis: Are We Going Insane?
What’s going on with the AI builder community right now?
lucumr.pocoo.org
January 18, 2026 at 12:51 PM
gmcd.bsky.social
Me, reading LinkedIn posts.
Brandolini's law: The amount of energy needed to refute bullshit is an order of magnitude bigger than that needed to produce it.
January 14, 2026 at 4:07 PM
gmcd.bsky.social
I see a lot of econ/socsci Bluesky asking for AI starter advice. IMO the highest impact, low cost tweak to your workflow is training an AI to follow *your* coding/writing prefs & habits.

Literally, tell Claude: "Look at <XYZ dirs>. ID common themes, styles & conventions. Write them to a .md file."
January 12, 2026 at 9:14 PM
gmcd.bsky.social
Same for many African nations (and everywhere else, which is Juan's point of course). There is no exceptionalism to be had here.
January 7, 2026 at 10:24 PM
gmcd.bsky.social
Moment 89 for me 🥹

An all-round, joyous and encompassing celebration of this game we love.
squidgerugby.bsky.social Squidge Rugby @squidgerugby.bsky.social · Dec 23
2025. An incredible year for rugby union. We count down every moment that made it special, silly, strange and so much more in our definitive celebration of the 288 Greatest Rugby Moments of 2025...
youtu.be/7hVloJg_FsE?...
The 288 Greatest Rugby Moments of 2025
YouTube video by Squidge Rugby
youtu.be
December 24, 2025 at 8:05 PM
Reposted by Grant McDermott
s3alfisc.bsky.social
`maketables` now has plug-in tooling - this means you can support maketables with your library without changing the maketables code base. This will hopefully be pretty useful for package developers, but also researchers developing their own estimators / in their own codebase.
December 20, 2025 at 2:01 PM
gmcd.bsky.social
github.com/grantmcdermo... 👀
Clustered standard errors by grantmcdermott · Pull Request #29 · grantmcdermott/dbreg
Closes #11 This PR adds support for analytic clustered standard errors. MWE NYC taxi example, taken from the (updated) README: dbreg( tip_amount ~ fare_amount + passenger_count | month + vendor_...
github.com
December 20, 2025 at 12:16 AM
Reposted by Grant McDermott
easystats.github.io
🎉 Great news for #rstats users! If you love the native R graphics feel of #tinyplot AND you're a fan of the powerful #easystats #modelbased package, this is for you!

Thanks to @gmcd.bsky.social, we significantly enhanced the tinyplot integration.

🔗 Read more: easystats.github.io/modelbased/a...
Plotting estimated marginal means with tinyplot
easystats.github.io
December 12, 2025 at 7:22 AM
gmcd.bsky.social
If you see this, repost with your model of positive masculinity.
Bandit. The dad we all want to be when we grow up.
December 6, 2025 at 10:02 PM
gmcd.bsky.social
#rstats Any sparklyr / Kubernetes experts able to help me with this issue? github.com/sparklyr/spa... 🙏
December 5, 2025 at 4:10 PM
Reposted by Grant McDermott
paulgp.com
Please enjoy this blog post based on my colleague Kyle Jensen's talk surveying different approaches to AI coding abilities

paulgp.com/ai-coding/20...
December 2, 2025 at 8:01 PM
gmcd.bsky.social
Hearts on the vine?
A houseplant casts a shadow of heart-shaped climbers in the background
November 29, 2025 at 3:48 AM
gmcd.bsky.social
Happy `tinyplot` v0.6.0 (codename "Thanksgiving") release day to all those that celebrate. Some new features, but mostly bug fixes and internal improvements.
grantmcdermott.com/tinyplot/NEW...

#rstats
Example of dodged pointrange (tiny)plot, with aligned ribbons layered on top.
November 27, 2025 at 5:26 PM
gmcd.bsky.social
Extraordinary. Doris MoM by miles for me.
November 22, 2025 at 7:47 PM
gmcd.bsky.social
What is the Irish word incredible courage combined with extreme stupidity.
November 22, 2025 at 6:40 PM
gmcd.bsky.social
First 20 min of this game almost as intense as @kjhealy.co's posit::conf keynote.
November 22, 2025 at 6:09 PM