iiinova — where your digital self keeps evolving

Art Steinmetz @apsteinmetz.bsky.social · 27/08/2026

Should we assume the remaining hubs have been emptied out by now?

Art Steinmetz @apsteinmetz.bsky.social · 28/03/2026

Off-topic-ish: when I enlisted in the USAF in 1981, the recruiter (a sergeant with quotas) ask asked me a about weed use. I said "yes" and he said, "from now on the answer is 'no'"

000

Art Steinmetz @apsteinmetz.bsky.social · 12/02/2026

Seriously, Jake, engagement is not worth your time. There is an endless supply of idiots.

000

Art Steinmetz @apsteinmetz.bsky.social · 09/02/2026

See the #rstats Shiny app many are calling the best #bobsburgers Shiny app ever. Give it a spin! H/T @sponce1.bsky.social bobsburgersR package for inspiration! viztoy.shinyapps.io/bobs_burger/

020

Art Steinmetz @apsteinmetz.bsky.social · 25/01/2026

Very excited to try this,https://www.jottr.org/2026/01/22/futurize-0.1.0/

001

Art Steinmetz @apsteinmetz.bsky.social · 16/01/2026

Both are terrific but I prefer the georeferenced one. You could toggle a satview underneath it.

010

Art Steinmetz @apsteinmetz.bsky.social · 09/01/2026

New year update of my wfmu.playlistanalyzer.net. Wherein you will find minutes of diverting amusement! Now with: Spins= 3,686,922 Songs= 1,070,901 Artists= 526,464 Shows= 108,549 DJs= 474 Earliest= 1982-04-02 @dianekamikaze.bsky.social! Latest= 2026-01-05 Days= 10,261 @wfmu.bsky.social #WFMU

020

Art Steinmetz @apsteinmetz.bsky.social · 07/01/2026

Google meet seems the smoothest. You can have a persistent link address that any group of people can use at will.

010

Art Steinmetz @apsteinmetz.bsky.social · 30/10/2025

I would like to hear about best practices also. The context window fills up pretty fast in my work.

010

Art Steinmetz @apsteinmetz.bsky.social · 14/09/2025

@amigabill.bsky.social Hi, Bill. I'm hoping your vast network might be able to help me to extract what's on this Ricoh 500mb MO drive cart. It was in my A2000 with a GVP SCSI controller. Powers up but I've been unsuccessful in recognizing the disk. I'd be happy to find a recovery service. THX!

010

Art Steinmetz @apsteinmetz.bsky.social · 12/09/2025

I'm a Windows person who also did a Mint install or R and RStudio. I didn't find *much* more work. R2U is a must. eddelbuettel.github.io/r2u/

eddelbuettel.github.io

CRAN as Ubuntu Binaries - r2u

Easy, fast, reliable -- pick all three!

000

Art Steinmetz @apsteinmetz.bsky.social · 07/08/2025

Does anyone have an opinion on whether Claude is better at vibe coding in #Python than #Rstats? My assumption would be Python because of a larger training set but I can't speak with any authority on the subject.

000

Art Steinmetz @apsteinmetz.bsky.social · 07/08/2025

Very cool, Jake. Too bad you can't climb the pyramid any more. The view from the top is something. Though that rule is in place for a good reasons - preservation and danger.

000

Art Steinmetz @apsteinmetz.bsky.social · 07/08/2025

Great stuff! Let me point out you should always use the duckplyr package on top of duckdb. With it you never have to do SQL translations yourself, just dplyr syntax and it’s even faster! outsiderdata.netlify.app/posts/2024-0...

outsiderdata.netlify.app

The Truth About Tidy Wrappers – Outsider Data Science

Are Tidyverse wrappers around powerful database engines really worth it?

010

Art Steinmetz @apsteinmetz.bsky.social · 06/08/2025

The funny thing is Trump could end this war "tommorow." The Russians don't need to lose they "just" need to be convinced they will never win. And that is in Trump's power, he's just too dumb to see it.

130

Art Steinmetz @apsteinmetz.bsky.social · 06/08/2025

In all seriousness, I have noticed that RStudio has starting hanging for me, often, in the last two weeks.2025.05.1

100

Art Steinmetz @apsteinmetz.bsky.social · 21/07/2025

Let's view the #NJ flooding on a map. Sooo many #Rstats #ggplot2 graphic elements in this plot and moving from mapping coordinate system to cartesian and back again. Crazy. Code here: gist.github.com/apsteinmetz/...

040

Art Steinmetz @apsteinmetz.bsky.social · 21/07/2025

I should also point out this is gage height CHANGE from baseline before storms.

000

Art Steinmetz @apsteinmetz.bsky.social · 21/07/2025

repost since animation of previous post was not...animated, for some reason.

000

Art Steinmetz @apsteinmetz.bsky.social · 21/07/2025

I continue to explore visualizing flooding with #rstats, #ggplot with stream gage data from the USGS. Here are the USGS stream gage data for the flooding in NJ on 7/14. You'll note that "Bound Brook in Middelsex" is the station downstream from Plainfield, where the worst flooding occured.

310

Art Steinmetz @apsteinmetz.bsky.social · 14/07/2025

The USGS monitors rivers and streams in near-real time. The gage closest to #CampMystic on the #GuadalupeRiver "pinned the meter" around 5AM, July 4 and stopped reporting. Some perspective visualizations using #Rstats. Claude helped with the USGS API. Code here: gist.github.com/apsteinmetz/...

$plots of USGS stream gage separation vs. elevation, river height vs. time and map positions of gages. Camp Mystic is upstream from all USGS sites. The first downstream site is 6 miles below at Hunt. At nightfall July 3 the river height at Hunt was 7 feet and the flow was 8 feet per second. Hunt stopped reporting around 5 AM on July 4 when the river height was 38 feet and the flow was 120,000 cubic feet per second. It took 2 hours go from 8 feet to 38 feet at Hunt.\ Judging from the other sites, the wave would have hit the camp around 4 AM. The flood wave moved quickly downstream. The dam at Sattler was not breached."$

000

Art Steinmetz @apsteinmetz.bsky.social · 21/06/2025

The numbers don't lie, but....all languages are "dead." Long live "vibe." I'm an old R-dog who can't write "Hello World" in Python, so I asked Claude to make a Shiny4Python app to display historical yield curves, just because. A few prompts and viola! Sort of takes the fun out of life.

Screen shot of a Shiny dashboard that allows the user to select a date range and displays Treasury yield curves.

121

Art Steinmetz @apsteinmetz.bsky.social · 03/06/2025

excuse the typo. I am estimating the year the book was published, not when the author reached "pubish."

010

Art Steinmetz @apsteinmetz.bsky.social · 03/06/2025

Here's the trivial #rstats code. I crudely estimate the publish date as birth year + 18 + half the lifespan after 18. Opinions may vary.

110

Art Steinmetz @apsteinmetz.bsky.social · 03/06/2025

Project Gutenberg is a fantastic resource but I've been frustrated by the absence of an original publication date. The metadata on the site shows the date it was posted to PG. #tidytuesday was a great excuse to "fix" this. A few lines with the #rstats gutenbergr package and presto!

170

Art Steinmetz @apsteinmetz.bsky.social · 11/05/2025

Seems like some mention of Positron, based on VS code should be made. Is this a better solution than "base" VS code?

000

Art Steinmetz @apsteinmetz.bsky.social · 14/04/2025

The Pogues were never my thing but "Fairytale of New York" but whenever I hear "Galway Bay" I tear up. Slowly, it's worming it's way into the Christmas canon, right up there with "Santa Baby."

010

Art Steinmetz @apsteinmetz.bsky.social · 13/04/2025

I agree and it scares the hell out of me.

010

Art Steinmetz @apsteinmetz.bsky.social · 13/04/2025

At this point Trump will hire a literal sack of bricks if it has a maga sticker on it.

020

Art Steinmetz @apsteinmetz.bsky.social · 13/04/2025

Or else, what? Nothing is what. Shameful.

0100

Art Steinmetz @apsteinmetz.bsky.social · 22/02/2025

Thanks to your distro shaming ;) I found the the minty version of "noble" and installed that. Cool runnings.

100

Art Steinmetz @apsteinmetz.bsky.social · 21/02/2025

Again, mint. Current release is a focal derivative.

100

Art Steinmetz @apsteinmetz.bsky.social · 20/02/2025

The only twist with a derivative is if you try to setup a repo with "add-apt-repository "deb cloud.r-project.org/bin/linux/ub... $(lsb_release -cs)-cran40/" is "lsb_release -cs" will return an unrecognized flavor. You have to manually insert the Ubuntu flavor. "focal", not "ulyana" in my case.

cloud.r-project.org

Ubuntu Packages For R - Brief Instructions

100

Art Steinmetz @apsteinmetz.bsky.social · 20/02/2025

@eddelbuettel.com I just did a linux (mint) install, again, so I'll thank you, again, for the fantastic r2u #rstats repo and the very clear setup instructions on github!

283

Art Steinmetz @apsteinmetz.bsky.social · 03/02/2025

...and, while climate is what you expect and weather is what you get, the years have been less icy of late.

000

Art Steinmetz @apsteinmetz.bsky.social · 03/02/2025

Lake Erie is pretty frosty now. A couple #ggplot2 extensions illustrate the context. The first plot duplicates what's at the NOAA using gghighlight. The second is fun with ggridges! #rstats

plot showing ice cover on lake erie with current year and average highlighted.

plot showing ice cover on lake Erie for the each of the last 50 years in a ridgeline format.

111

Art Steinmetz @apsteinmetz.bsky.social · 24/01/2025

How have the optimal starting words evolved as the original corpus has been "consumed?" Your guess #1 words are substantially different from what are popular with players.

000

Art Steinmetz @apsteinmetz.bsky.social · 24/01/2025

If we're talking about #rstats then Duckdb with Duckplyr is the superior #tidyverse solution. I've road tested "all the things."

191

Art Steinmetz @apsteinmetz.bsky.social · 20/12/2024

That is a dream of mine, because I still don't.

130

Art Steinmetz @apsteinmetz.bsky.social · 06/12/2024

Combining two of my favorite things. #rstats and oysters. My latest blog post is a project to predict New York Harbor water quality using data from Billionoysterproject.org and #tidymodels. outsiderdata.netlify.app/posts/2024-1...

outsiderdata.netlify.app

Predicting Water Quality in New York Harbor – Outsider Data Science

Using Tidymodels and data from the Billion Oyster Project

180

Art Steinmetz @apsteinmetz.bsky.social · 18/11/2024

Nice. It's mind blowing how few lines of code accomplish this. I also appreciate the many, many lines of code the package authors created to make it look easy.

000

Art Steinmetz @apsteinmetz.bsky.social · 17/11/2024

As a side note, '.by=' is the way to group when using duckDB's tidyverse wrapper, `duckplyr`, otherwise duckDB will drop back to using `dplyr` and lose the speed advantage.

150

Art Steinmetz @apsteinmetz.bsky.social · 17/11/2024

Some good ones in there! By "good," I mean useful beyond novelty value. Here's how you can use frame grabs to extract palettes. with R.

outsiderdata.netlify.app

Outsider Data Science - Rick and Morty Palettes

000

Art Steinmetz @apsteinmetz.bsky.social · 17/11/2024

Loops vs. base R (apply,sapply,lapply,mapply) argument predates tidyverse.

110