This is how you can visualize gene expression data in #tidyheatmaps 🤩 https://jbengler.github.io/tidyheatmaps/ #rstats #dataviz #phd
L. Eleanor "Moose" Nguyen
@moosewin.bsky.social
DOGE-displaced former civil servant technologist. Now does blood draws and physicals for a living. they/she. 🚄🏳️🌈 https://lucenguyen.com
The "scientific establishment" is always a funny thing to say because there's not one singular place where science happens. A tenure-hungry microbiologist at a state university or an early career fellow in Infectious Diseases would love to find the origin of a global pandemic, actually!
Ernie Piper@whoisernie.bsky.social · 7d agoZeynep Tufekci is still completely insane over the lab leak idea. "There is a conspiracy to cover this up" is not evidence for "covid was a bioweapon that escaped from a lab". It could also be evidence for my theory, "please stop talking about this stupid idea"
This is how you can add the mean and standard deviation to summarize the raw data points in your #tidyplots 🤩 Alternatively, you can also display the median, standard error of the mean, 95% confidence interval, and more. #rstats #dataviz #phd
the always-insightful Christian Hellwig has a (recently posted) Chicago Fed working paper that looks provocative and interesting: www.chicagofed.org/publications...
OH MY GOD (again) I GET TO TALK ABOUT MORE OF MY DISSERTATION! This is the second time this has happened in a week and has never happened before. So…verification at odds with a decentralized ethos, you say? Step into my office. 1/
pork, cheese, broccoli rabe@garlicbuffalo.gobirds.biz · 2mo agowe should judge anybody who has been verified by bluesky extremely harshly. besides being nerds, they are diametrically opposed to the decentralized promises of this platform.
Utilities are apparently not sharing smart meter data across PJM’s grid territory. That’s holding back grid-relieving demand-response programs.
Utilities aren’t sharing smart meter data in PJM. That’s a problem.
Demand response companies say utilities in the biggest U.S. energy market are blocking gigawatts of grid relief by withholding data meant to help…
canarymedia.com
Amongst all the rubbish that I have seen in peer reviewed papers, this data is the worst. There are sketches, images of children (with no mention of ethics) and multiple celebrities, whilst the paper states: "A smart camera located in patient rooms has been used to collect images".
Retraction Watch@retractionwatch.com · 3mo agoScrolling through an online image dataset, Adrian Barnett pointed out a few familiar faces: Sylvester Stallone, George Clooney, Angelina Jolie and Daniel Craig. “You can see,” Barnett said, “this is just a comically bad dataset.” retractionwatch.com/2026/05/18/k...
Craziest thing about lean.py? You can roll your own (almost) feature-parity successor to Dafny. Python programs. Lean obligations. Grind as a solver. Proof complete. Program verified. Python code. Real.
Kiran@kirancodes.me · 3mo agoPython doesn't have a refinement type system---except now it does~ Combining algebraic effects with Lean, we can create a lightweight verification framework embedded into python~ We send the VCs off to lean, the function is verified if we can close the proof! Liquid Haskellers on suicide watch
![def positive_increment(x: Annotated[int, Gt(0)]):
y = x + 3
assert_refined(y, Gt(3)) # x > 0 → x + 3 > 3 ✓
z = x + 10
assert_refined(z, Gt(10)) # x > 0 → x + 10 > 10 ✓
assert verify(positive_increment)
def bounded_sum(x: Annotated[int, Gt(0)], y: Annotated[int, Gt(0)]):
s = x + y
assert_refined(s, Gt(1)) # x > 0 ∧ y > 0 → x + y > 1 ✓
assert_refined(s, Ge(2)) # x > 0 ∧ y > 0 → x + y >= 2 ✓
assert verify(bounded_sum)
def failing(x: Annotated[int, Gt(0)]):
y = x + 1
assert_refined(y, Gt(10)) # x > 0 → x + 1 > 10 ✗
assert not verify(failing)](https://cdn.bsky.app/img/feed_thumbnail/plain/did:plc:i32jjsch6xqcguzsf2lgbfyu/bafkreih65hwqerzcwynj4j5me5efv6vyf52dsecn3y27ojqg3elqsow6nu)
Got a notification from a medieval copyright lawyer
Linear Regression: A Visual Introduction to (Almost) Everything You Should Know 👉 mlu-explain.github.io/linear-regre... By @jdwlbrv — a must-read for anyone into #rstats, #dataviz, #datascience, #machinelearning, #LinearRegression, and #stats.
Linear Regression
A visual, interactive explanation of linear regression for machine learning.
mlu-explain.github.io
The iron-ish law of YIMBYism: Markets that spend decades below 6% rental vacancy are almost always expensive. Markets that maintain healthy vacancy rates are usually cheap
arXiv📈🤖 int3ract: Johnson-Neyman Technique and its Three-Way Extension for Frequentist and Bayesian Models in R By Krause
I've seen enough, US Postal Inspection Service needs to take over presidential protection and leave the Secret Service to go back to their niche of chasing down counterfeiting.
Mike Black@mikeblack114.bsky.social · 3mo agoanyway we won't know what actually happened until the books come out next year, but regardless the next dem admin has to just completely destroy the USSS and task someone else with protection duties
> 50 years of stability 50 years ago we were 3 years away from the Iranian Revolution and 4 years away from the Iraq-Iran War. Anyone who thinks that 50 years of stability in geopolitics is even possible in this domain — well, I have a bridge in Brooklyn to sell them.
Luke Kawa@ljkawa.bsky.social · 4mo agoThe view from Apollo chief economist Torsten Slok: “Markets are overreacting to what will likely be a 4- to 6-week period of volatility, which will ultimately result in 50 years of stability in oil markets, supply chains and geopolitics.”
You cannot be saying this kind of thing when your primary job is podcaster.
Every day is a good day for sharing one of the most useful papers about research data ever written. PLEASE get your people to understand and follow this advice. www.tandfonline.com/doi/full/10....
More details and context about what being a woman math major in the early-2000s was like and why it was so damaging here: cooperativeoverlapping.substack.com/p/a-fuller-s...
A Fuller Statement About My Bluesky Posts
Let me paint a picture of what it was like to be a woman math major at the University of Chicago circa 2004.
cooperativeoverlapping.substack.com
As a 17-year-old commege sophomore, I was sexually harassed by my math professor. It changed, markedly, the trajectory of my life. I'd earned a full scholarship to study science & engineering at 15. That became impossible when I couldn't even sit in a room alone with a professor without panicking.
my bloomberg terminal is just showing the scene where wile e. coyote runs off past the edge of a cliff, is this good?
Name of the hospital that saved my life (Lucile Packard Children's Hospital) and famous historical figure (Eleanor Rykener) got me Lucille Eleanor.
Samberry/サムベリー🍓@killingangel.bsky.social · 5mo agohi lil trans ppl
Run from it, hide from it, the Haber-Bosch Process comes for you all the same.
Maia@maiamindel.bsky.social · 5mo agobro i did not know the gulf states had an even bigger share of the global fertilizer market than of the oil market this is going to be a disaster
Throughout industrialized nations, there's been a long-run decline in hours worked annually. This time is, in fact, used on leisure and other non-work activities. This has been concurrent with mechanization, automated manufacturing, and the decline in agricultural share of total employment.
Dr. Hilary Agro, PhD 🍄@hilaryagro.com · 5mo agoIt's bizarre to me how many people in the comments here seem to think that automation is going to translate into more freedom for the working class. When has that ever panned out? How much leisure time do you have, after years of these "innovations"? We are Charlie Brown trying to kick the football
Is the only way we can create algorithms that people understand to make them trivially simple? We argue, no. People can predict the behavior of algorithms that are arbitrarily complex, if and only if they are available, compact and aligned. arxiv.org/abs/2601.18966
I just created a series of seven deep-dive videos about AI, which I've posted to youtube and now here. 😊 Targeted to laypeople, they explore how LLMs work, what they can do, and what impacts they have on learning, well-being, disinformation, the workplace, the economy, and the environment.
Part 1: How do LLMs work?
YouTube video by Andrew Perfors
youtube.com
Writing a GeoPackage(GPKG) library from scratch taught me a lot about some friction between the spec and the modern geospatial ecosystem🦀🌍 GPKG is SQLite. We take that for granted, but have you ever stopped to think about the advantages and disadvantages? Here’s my take. dev.to/yutannihilat...
How it feels to write a GPKG library in 2026 (in Rust!)
There are a lot of existing libraries to deal with GeoPackage (GPKG). GDAL is the undisputed champion...
dev.to
The Spring 2026 course website for my grad causal inference class is up and running. Was inspired by @andrew.heiss.phd and @mattblackwell.bsky.social to move my materials over to a standalone site and use as little of Canvas as possible. www.antonstrezhnev.com/ps813/
PoliSci 813
antonstrezhnev.com
🔥I am super excited for the official release of an open-source library we've been working on for about a year! 🪄interpreto is an interpretability toolbox for HF language models🤗. In both generation and classification! Why do you need it, and for what? 1/8 (links at the end)
1/ We found that deep sequence models memorize atomic facts "geometrically" -- not as an associative lookup table as often imagined. This opens up practical questions on reasoning/memory/discovery, and also poses a theoretical "memorization puzzle."