Tessera
Folgen

Allison Horst

@allisonhorst.bsky.social

Artist. Previously DevRel & teaching faculty (Environmental Data Science) at UC Santa Barbara. PhD, Env Science. Data science | R | data viz | education | art | allisonhorststudio.com (landscapes) | www.allisonhorst.com (data science)

70 Posts5728 Follower506 Folgt
Allison Horst@allisonhorst.bsky.social · 4mo ago

Day 7 #30DayChartChallenge. Multi...scales? Questionable whether this fits the theme, but anyways here are 3 cool perceptual scales. Learned about the Bortle Dark Sky scale last month. Spent 10 years in EnvSci & found the Forel-Ule scale yesterday. Imagine what color scale I'll learn tomorrow.

A watercolor and ink sketch showing three color scales: the urine color scale (to measure dehydration), the Bortle Dark Sky scale (or sky darkness / pollution), and the Forel Ule Scale for water color.
Allison Horst@allisonhorst.bsky.social · 4mo ago

Day 3 #30DayChartChallenge: Mosaic. This is a fish tale, based on how Greg and I remember time spent fly fishing on 4 rivers we love. In the making of it we looked through old photos, revisited great memories, and more than once said “We’ve gotta get back there.” Watercolor on cold press paper.

A hand-painted watercolor mosaic chart representing personal experiences fly fishing on four rivers. Each segment is filled with a different fish pattern (brown trout, rainbow rout, mountain whitefish).
Allison Horst@allisonhorst.bsky.social · 2y ago

I recently had the opportunity to interview Ben Welsh (Reuters news applications editor), during a very interesting time for data journalists. Check out the story below! 👇

Observable@observablehq.com · 2y ago

We chatted with award-winning journalist and Reuters applications editor Ben Welsh about building back-end data pipelines for data journalism. Learn how he's cutting down the time data journalists spend writing code, to make time for more important work: https://buff.ly/4fJFIF8

Katharine Hayhoe@katharinehayhoe.com · 2y ago

I'm a climate scientist living in Texas who served as a lead author on the National Climate Assessment under the previous Trump administration. I witnessed first-hand its chilling impact on access to data, funding & publication of research, implementation of solutions, and scientists themselves. 🧵

Trump victory has sweeping climate change consequences

From the endorsement of climate denial to a brain drain of climate scientists, Trump's victory will bring major changes.

axios.com

Allison Horst@allisonhorst.bsky.social · 2y ago

👻 Here's the ✨ new 2024 art, feat. community responses to the question "What's something in R that feels tempting, but is actually dangerous?" I couldn't fit them all, so I've added a blank version 👇🧵 so you can make your own (could be a fun class activity 😉) Happy Halloween, #rstats world! 🤡🎈

An illustration recreating the It movie poster, where a monster version of Pennywise peers out of a storm drain holding red balloons and hex stickers, enticing a little monster in a yellow raincoat who looks on. On the balloons are written different things to be cautious about in R / stats. On the raincoat are different hex stickers (for janitor, naniar, usethis, dplyr, renv...).
Allison Horst@allisonhorst.bsky.social · 2y ago

The responses are so good, & I learned quite a few new things! Thanks everybody, keep 'em coming -- the thread feels like the start of a fun blog post. 👻🎀👾🎀👾👻 Here's one that wasn't specifically made for Halloween but feels like the right vibes, from early 2021.

Two smiling green monsters in matching blue dresses (like the scary twins in The Shining), standing in a hallway with hex wallpaper. Behind them is spraypaint-style text reading "come play with us" with "forevR" in a hex logo behind them.
Allison Horst@allisonhorst.bsky.social · 2y ago

I'm working on my 2024 Halloween R monsters illustration and #rstats folks, I need your help 👻🎃 What's something in data science, R, statistics, etc. that feels tempting but is very, very dangerous? In the meantime I'll be sharing some Halloween R art from previous years 🕷️🧟☠️ The first (2019?):

Two fuzzy monsters, one dressed as David Bowie and one as a witch, run screaming out of a haunted forest with spooky trees and red eyes watching them. On the tree branches are things like "setwd", fatal error, merge conflict, aborted R session, no metadata, etc.
Jenny Bryan@jennybryan.bsky.social · 2y ago

If you're an #rstats user, one way to protect yourself from yourself is to do: usethis::git_vaccinate() which will setup or modify your user-level (aka global) gitignore file to handle the most obvious R-associated files that probably don't belong in git. usethis.r-lib.org/reference/gi...

Vaccinate your global gitignore file — git_vaccinate

Adds .Rproj.user, .Rhistory, .Rdata, .httr-oauth, .DS_Store, and .quarto to your global (a.k.a. user-level) .gitignore. This is good practice as it decreases the chance that you will accidentally leak...

usethis.r-lib.org

Jenny Bryan@jennybryan.bsky.social · 2y ago

git commit -am "This is how you push your tokens and R history to GitHub for the whole world to enjoy!"

Allison Horst@allisonhorst.bsky.social · 2y ago

I'm working on my 2024 Halloween R monsters illustration and #rstats folks, I need your help 👻🎃 What's something in data science, R, statistics, etc. that feels tempting but is very, very dangerous? In the meantime I'll be sharing some Halloween R art from previous years 🕷️🧟☠️ The first (2019?):

Two fuzzy monsters, one dressed as David Bowie and one as a witch, run screaming out of a haunted forest with spooky trees and red eyes watching them. On the tree branches are things like "setwd", fatal error, merge conflict, aborted R session, no metadata, etc.