Randy Boyes
@randy.pub
Analytics at @presage.group | PhD Epidemiology | Wolfe Island Volunteer Firefighter | #julialang #python #rstats | publications at www.randy.pub | GitHub @rdboyes
This joke goes nuts if you, like me, are the son of a evangelical pastor who became an epidemiologist
Julia M. Rohrer@dingdingpeng.the100.ci · 6d agoNew blog post! Odds ratios are...a bit weird. Here I explain two problems (interpretation, non-collapsibility) and provide some simple recommendations for improved reporting. www.the100.ci/2026/07/30/r...
I wish claude would just admit it doesn't know instead of attempting to gaslight me
Bruce only likes to chew toys while holding them over the edge of the couch, and gets very upset when he drops them
“Hey codex what are your ‘likes’”
a man with a beard is sitting on a couch and talking about ghouls .
ALT: a man with a beard is sitting on a couch and talking about ghouls .
media.tenor.com
WIRED@wired.com · 3mo ago“Never talk about goblins, gremlins, raccoons, trolls, ogres, pigeons, or other animals or creatures unless it is absolutely and unambiguously relevant,” reads OpenAI's coding agent instructions. www.wired.com/story/openai...
The flexibility of `geom_polygon` lets you do basically whatever you want with your plots, e.g. (code: gist.github.com/rdboyes/7a7f...)
Information is Beautiful @infobeautiful.bsky.social · 4mo agoBeautiful vintage #infographic on the ceiling & speed of different aircraft (via Reddit bit.ly/4c2bMCs)
Cool study, and I’m convinced they’re generally correct. Very high lizardman quotient in this figure though - if you’re not getting 100% strong agreement that “betting (real money)” is a form of gambling, something is off, and the SVK sample “strong agree” for that item is <50%
Veli-Matti Karhulahti@mkarhulahti.bsky.social · 6mo agobehold, we found great variation in how people think! Many activities that we thought would be “gaming” weren't & vice versa, eg half of the participants interpreted ‘gambling’ to be ‘gaming’. Ergo: surveying ‘gaming’ without defining it creates data mess
Average programming language enjoyer: there are two hard things, cache invalidation and naming things #julialang programmers:
Nathan Goldwag @goldwagnathan.bsky.social · 6mo agomods are asleep post forbidden greek letters
Only had time to finish part 1 of #adventofcode today - hoping to get back to p2 on the weekend though!
Two #adventofcode solutions for the price of one today, since I was too busy working yesterday. TypeScript feeling ok, but how is array.sum() not a thing!?
![function solve(line: string, bat_length: number) {
const arr = line.split("").map(function (char: string) {
return Number(char);
});
let end: number = arr.length - bat_length + 1;
let start: number = 0;
let digits: number[] = [];
while (digits.length < bat_length) {
const selected = Math.max(...arr.slice(start, end));
digits.push(selected);
end += 1;
start += 1 + arr.slice(start, end).indexOf(selected);
}
return digits
.reverse()
.map(function (n, i) {
return n * 10 ** i;
})
.reduce((a, b) => {
return a + b;
}, 0);
}
console.log(
require("fs")
.readFileSync("data/2025/3.txt", "utf-8")
.split(/\r?\n/)
.filter((line) => line.trim().length > 0)
.map(function (line: string) {
return [solve(line, 2), solve(line, 12)];
})
.reduce(function (a, b) {
return [a[0] + b[0], a[1] + b[1]];
}),
);](https://cdn.bsky.app/img/feed_thumbnail/plain/did:plc:2h5e6whhbk5vnnerqqoi256k/bafkreiguiw3f25b5gl5t2usqsebbfgkmulqaz5ywg3sufsz35q7hyp3rnq)
![var input_matrix = require("fs")
.readFileSync("data/2025/4.txt", "utf-8")
.split(/\r?\n/)
.filter((line) => line.trim().length > 0)
.map(function (line: string) {
return line.split("").map(function (char) {
if (char == "@") {
return 1;
} else {
return 0;
}
});
});
var count = 0;
var remove = [];
var exit = false;
while (exit == false) {
for (const target of remove) {
input_matrix[target[0]][target[1]] = 0;
}
var remove = [];
for (var i = 1; i < input_matrix.length - 1; i++) {
for (var j = 1; j < input_matrix.length - 1; j++) {
if (
input_matrix[i - 1]
.slice(j - 1, j + 2)
.concat(
input_matrix[i].slice(j - 1, j + 2),
input_matrix[i + 1].slice(j - 1, j + 2),
)
.reduce((a, b) => {
return a + b;
}, 0) <= 4
) {
if (input_matrix[i][j] == 1) {
count++;
remove.push([i, j]);
}
}
}
}
if (remove.length == 0) {
exit = true;
}
}
console.log(count);](https://cdn.bsky.app/img/feed_thumbnail/plain/did:plc:2h5e6whhbk5vnnerqqoi256k/bafkreicvj3id64a43rroadtecprreznncjbvcgnvdsc25ekqmvyiqzzf4m)
TypeScript #adventofcode Day 2 complete! Rocky one, really missing some of the builtin functions from my "home languages", but we got there
![const range = (start: number, stop: number): number[] =>
Array.from({ length: 1 + stop - start }, (_, i) => start + i);
function chunkString(str: string, length: number) {
return str.match(new RegExp(".{1," + length + "}", "g"));
}
const divisors = (num: number) =>
[...Array(num)].map((x, i) => i).filter((x) => num % x === 0);
function solve(id_range: string): number {
const arr_range = id_range.split("-");
const possibles = range(Number(arr_range[0]), Number(arr_range[1]));
const ans = possibles
.map(function (element: number): number {
let id_string = `${element}`;
const checks: number[] = divisors(id_string.length).map((num) => {
return new Set(chunkString(id_string, num)).size;
});
if (Math.min(...checks) == 1) {
return element;
} else {
return 0;
}
})
.reduce((a, b) => {
return a + b;
}, 0);
return ans;
}
console.log(
require("fs")
.readFileSync("data/2025/2.txt", "utf-8")
.split(",")
.map(function (id_range: string) {
return solve(id_range);
})
.reduce((a, b) => {
return a + b;
}, 0),
);](https://cdn.bsky.app/img/feed_thumbnail/plain/did:plc:2h5e6whhbk5vnnerqqoi256k/bafkreidp2p5ferqbjmivb7rn2i6cwlrvnyzvi3ikab36earezakwqkdspm)
Doing #adventofcode in TypeScript this year since I need to learn it for work... missing #julialang but it's not too bad. Day 1:
![Typescript solution to advent 2025 day 1. Code: class Dial {
current: number = 50;
zeros: number = 0;
pass_zero: number = 0;
turn(s: string): void {
if (s[0] == "R") {
this.current += Number(s.substring(1));
if (this.current >= 100) {
this.pass_zero += Math.trunc(this.current / 100);
this.current = this.current % 100;
}
} else {
if (this.current == 0) {
this.pass_zero -= 1;
}
this.current -= Number(s.substring(1));
if (this.current <= 0) {
this.pass_zero += -Math.trunc(this.current / 100) + 1;
this.current = 100 + (this.current % 100);
}
}
if (this.current == 0 || this.current == 100) {
this.zeros += 1;
this.current = 0;
}
}
}
var p1 = new Dial();
require("fs")
.readFileSync("data/2025/1.txt", "utf-8")
.split(/\r?\n/)
.forEach(function (line: string) {
p1.turn(line);
});
console.log(p1.zeros);
console.log(p1.pass_zero);](https://cdn.bsky.app/img/feed_thumbnail/plain/did:plc:2h5e6whhbk5vnnerqqoi256k/bafkreienupynbsg3lshxcxg4u4haqgzpvdyn2stgg7netyvpb5uvevi5me)
I assume the reviewers didn’t like the original title, “oh no: we have to throw out all survey data collected in the last 1,084 days”
Josh McCrain@joshmccrain.bsky.social · 9mo agonew paper by Sean Westwood: With current technology, it is impossible to tell whether survey respondents are real or bots. Among other things, makes it easy for bad actors to manipulate outcomes. No good news here for the future of online-based survey research
I am giving 2 talks abt making your data analysis outputs into polished products with the power of #RStats and #Python pkgs + my pkg {brandthis}. Talk 1: DevFest 2025. Register here: wedevfest.com Talk 2: R+AI 2025 organized by R Consortium. Register here: rconsortium.github.io/RplusAI_webs...
This may just be the best CS paper I’ve read this year. Just read the abstract and first para of the intro! The rest of the intro is really wild too, but very very good: dl.acm.org/doi/pdf/10.1...
Exploring some ideas for how to create a macro-free (or macro-light) implementation of the tidyverse in #julialang (get out the jolly roger 🏴☠️): www.randy.pub/posts/piracy/
You wouldn't pirate a type
randy.pub
I’m doing my honest best to listen to this with an open mind but hooiiioill There are so many words said that have absolutely no bearing on the facts of the case podcasts.apple.com/ca/podcast/t...
The Law, such as it is
Government Podcast · Monthly Series · An exploration of a broken, if beautiful, institution, the law, and tradition, America. Season one told the story of a case, LESSIG v. NYT. Season two begins the ...
podcasts.apple.com
I miss being able to just buy a thing and have it, I hate paying for subscriptions
erika@yeeeerika.bsky.social · 2y agoi don't want to hear your most boomer complaint. what's your most millennial complaint?
I really enjoyed giving a brief introduction to Julia for statisticians yesterday - thanks to Jake Grainger and Charles Dufour for organising such a great session! Slides: nrennie.rbind.io/talks/rss-in... #RStats #JuliaLang #ggplot2 #RSS2025Conf @tidierjl.bsky.social
No wonder they were having a blast
Jason Schreier@jasonschreier.bsky.social · 12mo agoBREAKING: Silksong will be out on September 4. Two weeks from today. Really. Often, games that take 7+ years to make are plagued by mismanagement and painful burnout. But for Silksong? Team Cherry was having a blast. They still are. This is their story: www.bloomberg.com/news/newslet...
Nutrition research continues to fill the library’s hexagonal galleries
Health Nerd@gidmk.bsky.social · 12mo agoSomeone just sent me this study, and all I can say is that I don't understand why anyone would bother investigating this question. 1/n
Another #juliacon in the books! Check out my talk on TidierPlots.jl here: www.youtube.com/live/HMdBi9L... and the State of Tidier.jl (by Daniel Rizk) here: www.youtube.com/live/_Uf2KQX... to see what's been going on with @tidierjl.bsky.social in #julialang
JuliaCon Global 2025 | Else Room | Day 2
YouTube video by The Julia Programming Language
youtube.com
These values are required to do science - acting like this is a problem is like saying that CEOs valuing “professionalism” or “punctuality” more than the average person means they aren’t value-aligned with the average person
Matthew Yglesias @mattyglesias.bsky.social · last yr.There’s a big gap on *values* (not facts) between scientific experts and the mass public, which is driving a lot of the crisis of confidence in institutions and expertise. www.slowboring.com/p/the-crisis...
Easily GOTY so far, and it’s hard to imagine anything knocking it out of that spot
Skill Up@skillup.bsky.social · last yr.After 2 months of work, the Clair Obscur Expedition 33 video is here. I interviewed most of the game's senior leadership team asking one simple question: how did you do this? Their answers were inspiring and moving, and it's no surprise that COE33 is as remarkable as it is: youtu.be/mXLOLgC2V2Q?...
![[grossed out] --dangerously-skip-permissions
[interested] auto mode](https://cdn.bsky.app/img/feed_thumbnail/plain/did:plc:2h5e6whhbk5vnnerqqoi256k/bafkreic55bdcbgjbn2ost4vqgnelavotkabhf4djy5pjuik3n7salenlra)
