Stephen Olivier 🇿🇦
LightNews LightNews
banner
stephenolivier.bsky.social
Stephen Olivier 🇿🇦
@stephenolivier.bsky.social
180 followers 290 following 150 posts
Statistician, Africa Health Research Institute. Views are my own.
Posts Replies Media Videos
stephenolivier.bsky.social
Ha, everything just keeps getting more and more stupid www.theguardian.com/business/202...
Disney to invest $1bn in OpenAI, allowing characters in Sora video tool
Agreement comes amid anxiety in Hollywood over impact of AI on the industry, expression and rights of creators
www.theguardian.com
December 11, 2025 at 5:10 PM
Reposted by Stephen Olivier 🇿🇦
jessicahullman.bsky.social
I would like to be able to take papers on LLM deception seriously, but the tip-toeing around problems caused by sycophancy makes me cringe.

statmodeling.stat.columbia.edu/2025/12/09/i...
“it has been argued that current chatbots may pose a risk of amplifying delusional thinking in vulnerable users, due to their tendency to sycophantic and overly validating behaviour” | Statistical Mo...
statmodeling.stat.columbia.edu
December 9, 2025 at 5:37 PM
Reposted by Stephen Olivier 🇿🇦
madhupai.bsky.social
“tuberculosis is not a relic of medical history. It remains the leading infectious cause of death worldwide”

www.nybooks.com/articles/202...
The Plague That Won’t Die | Pria Anand
As my recent diagnosis shows, tuberculosis is not a relic of medical history. It remains the leading infectious cause of death worldwide—and America is hardly immune.
www.nybooks.com
December 5, 2025 at 11:22 PM
Reposted by Stephen Olivier 🇿🇦
harriet-mason.bsky.social
HEY! YOU! ggdibbler 0.6.1 is finally up on CRAN!

harriet-mason.github.io/ggdibbler/in...

You can now pass random variables to any aesthetic in any ggplot geom/stat. If you can express it as a distribution, ggdibbler can plot it. It also works with ggplot extensions, like gganimate. #rstats
December 6, 2025 at 3:16 PM
Reposted by Stephen Olivier 🇿🇦
libbyheeren.bsky.social
We're getting a native R %notin%!! "after many years of private definitions mostly hidden in packages" 😂😂

#rstats #databs
December 6, 2025 at 7:49 PM
Reposted by Stephen Olivier 🇿🇦
jenvictor.bsky.social
I cannot evangelize enough for @zotero.org for citation management. It's a game changer. If you're a student, and you're writing a paper right now, and you're NOT using Zotero, I ask you...why? Why are you making things harder than they need to be?
a woman is sitting at a desk with her arms outstretched in front of a map on the wall .
ALT: a woman is sitting at a desk with her arms outstretched in front of a map on the wall .
media.tenor.com
December 6, 2025 at 9:41 PM
stephenolivier.bsky.social
I wish that some mathematician could prove that playing 3D chess is actually much easier than normal chess. Just because I want that phrase to die
December 7, 2025 at 12:54 PM
Reposted by Stephen Olivier 🇿🇦
andrew.heiss.phd
I recently discovered Conventional Comments (conventionalcomments.org) for providing a pseudo-standard set of labels for feedback and just tried it for an article review and it was really helpful to specify issues vs. thoughts vs. suggestions, etc. Hopefully it's helpful for the authors too!
We strongly suggest using the following labels: praise: Praises highlight something positive. Try to leave at least one of these comments per review. Do not leave false praise (which can actually be damaging). Do look for something to sincerely praise. nitpick: Nitpicks are trivial preference-based requests. These should be non-blocking by nature. suggestion: Suggestions propose improvements to the current subject. It’s important to be explicit and clear on what is being suggested and why it is an improvement. Consider using patches and the blocking or non-blocking decorations to further communicate your intent. issue: Issues highlight specific problems with the subject under review. These problems can be user-facing or behind the scenes. It is strongly recommended to pair this comment with a suggestion. If you are not sure if a problem exists or not, consider leaving a question. todo: TODO’s are small, trivial, but necessary changes. Distinguishing todo comments from issues: or suggestions: helps direct the reader’s attention to comments requiring more involvement. question: Questions are appropriate if you have a potential concern but are not quite sure if it’s relevant or not. Asking the author for clarification or investigation can lead to a quick resolution. thought: Thoughts represent an idea that popped up from reviewing. These comments are non-blocking by nature, but they are extremely valuable and can lead to more focused initiatives and mentoring opportunities. chore: Chores are simple tasks that must be done before the subject can be “officially” accepted. Usually, these comments reference some common process. Try to leave a link to the process description so that the reader knows how to resolve the chore. note: Notes are always non-blocking and simply highlight something the reader should take note of.
November 17, 2025 at 3:52 PM
Reposted by Stephen Olivier 🇿🇦
wiringthebrain.bsky.social
New paper with @statsepi.bsky.social and @deevybee.bsky.social in which we show there's really no evidence for a link between the gut microbiome and autism www.cell.com/neuron/fullt...
Conceptual and methodological flaws undermine claims of a link between the gut microbiome and autism
Claims that the gut microbiome causally contributes to autism regularly appear in the scientific literature and popular press. Mitchell et al. critically examine influential studies underpinning these...
www.cell.com
November 13, 2025 at 4:10 PM
Reposted by Stephen Olivier 🇿🇦
caulfieldtim.bsky.social
Sum: Nah.

Gut #microbiome tests are everywhere. Should you get one? medicalxpress.com/news/2025-11... #GutHealth

- Little evidence to support use
- Often associated with sale of unproven supplements
- And even if worked, probably wouldn't change approach (healthy diet, exercise, sleep)!
Gut microbiome tests are everywhere. Should you get one?
At-home gut microbiome tests are trending, selling insights to the curious, empowerment to the chronically ill and a claimed path to longevity to the wellness-conscious.
medicalxpress.com
November 11, 2025 at 2:15 PM
Reposted by Stephen Olivier 🇿🇦
weare.rladies.org
A little bit of self-promotion, I @ivelasq3.bsky.social will be giving a talk on Positron for RStudio Users!

Hoping to share some perspectives, extensions, and other tips in case you're interested in exploring Positron.

Sign up here: www.meetup.com/rladies-gabo...
September 25, 2025 at 5:57 PM
stephenolivier.bsky.social
It's almost as if maths gives the illusion that statistics is simple
September 25, 2025 at 10:29 AM
stephenolivier.bsky.social
Always read Danielle (ARD)
djnavarro.net Danielle Navarro @djnavarro.net · Sep 21
Against my better instincts, I have written some notes on how human probability judgements work and what you should expect from surveys that ask people to guess what proportion of the population is transgender. I hope never to speak of this matter again
Some notes on probability judgement – Notes from a data witch
For the love of fuck, literally nobody thinks that 20% of the population is transgender. Please stop sharing that ridiculous YouGov statistic
blog.djnavarro.net
September 24, 2025 at 9:03 AM
Reposted by Stephen Olivier 🇿🇦
rbly.bsky.social
Prediction modelling hat on: this great post extends to how humans interact with predictions. Generally we appear to be good at discrimination, ie ranking things by risk. Calibration OTOH: cooked. Maybe we should focus decision support on realistic risk preds, w less focus on AUROC?
#statsky #rstats
djnavarro.net Danielle Navarro @djnavarro.net · Sep 21
Against my better instincts, I have written some notes on how human probability judgements work and what you should expect from surveys that ask people to guess what proportion of the population is transgender. I hope never to speak of this matter again
Some notes on probability judgement – Notes from a data witch
For the love of fuck, literally nobody thinks that 20% of the population is transgender. Please stop sharing that ridiculous YouGov statistic
blog.djnavarro.net
September 24, 2025 at 4:45 AM
Reposted by Stephen Olivier 🇿🇦
boschbot.bsky.social
Temptation of St Anthony
Temptation of St Anthony
September 24, 2025 at 7:17 AM
Reposted by Stephen Olivier 🇿🇦
datavisfriendly.bsky.social
#rstats An interesting post if you're interested in learning how things work in {ggplot2}
-> layer_data() give you a view of what goes on inside the box
mattansb.msbstats.info Mattan S. Ben-Shachar @mattansb.msbstats.info · Sep 15
New blog post!

Ever wonder what geom_histogram is actually doing? How about geom_boxplot?

In celebration of the release of #ggplot2 4.0.0 (ggplot8?), I explore the relationships between the “geoms” and “stats” offered by the core {ggplot2} functions.

#rstats
Exploring {ggplot2}’s Geoms and Stats – Stat’s What It’s All About
blog.msbstats.info
September 16, 2025 at 12:12 AM
Reposted by Stephen Olivier 🇿🇦
triadsou.bsky.social
GLMcat: An R Package for Generalized Linear Models for Categorical Responses. Lorena León, Jean Peyhardi, Catherine Trottier. Journal of Statistical Software. www.jstatsoft.org/article/view...
GLMcat: An R Package for Generalized Linear Models for Categorical Responses by Lorena León, Jean Peyhardi, Catherine Trottier
<p>In statistical modeling, there is a wide variety of generalized linear models for categorical response variables (nominal or ordinal responses); yet, there is no software embracing all these models...
www.jstatsoft.org
September 12, 2025 at 4:28 AM
stephenolivier.bsky.social
A poor worm once had the life sucked out of it by mistakenly seeking refuge in the brain of a parasitic RFK Jr
September 5, 2025 at 9:16 PM
stephenolivier.bsky.social
I'm not shocked at all
September 3, 2025 at 8:11 PM
Reposted by Stephen Olivier 🇿🇦
steamtraen.eu
Separated at birth
Prince Andrew and the "guilty monkey"
September 3, 2025 at 6:00 PM
Reposted by Stephen Olivier 🇿🇦
grimalkina.bsky.social
CHANGE TECHNICALLY IS ON THE CASE

www.changetechnically.fyi/2396236/epis...
September 3, 2025 at 3:43 PM
stephenolivier.bsky.social
Apparently Hari's books are edible
Facebook post by disgraced journalist Johann Hari ignorantly holding up a Asian cuisine book, "Cooking with Poo"
September 3, 2025 at 9:20 AM
stephenolivier.bsky.social
Remember the first person you must fool is yourself. Then you can seemingly get away with anything
August 28, 2025 at 2:15 PM
Reposted by Stephen Olivier 🇿🇦
emilhvitfeldt.bsky.social
Happy to announce ✨quarto-revealjs-editable✨

This fully supersedes the imagemover extension, as I back then didn't realize the potential. You can now also move, resize, change font size and alignment for text in your slides

github.com/EmilHvitfeld...
#quarto #slidecrafting
August 20, 2025 at 5:39 PM
Reposted by Stephen Olivier 🇿🇦
cameronpat.bsky.social
these were new to me, look very useful though!
cghlewis.bsky.social Crystal Lewis @cghlewis.bsky.social · Aug 16
I often need to filter data in #rstats based on many variables. The dplyr::if_any() and dplyr::if_all() functions are really handy for this.

More examples here: cghlewis.github.io/data-wrangli...
library(tidyverse) # Review the data (df). Missing data is denoted by “MR”. df # A tibble: 6 x 6 id form var1 var2 var3 var4 <dbl> <chr> <chr> <chr> <chr> <chr> 1 10 "a" 3 4 3 3 2 11 "b" 3 MR 2 4 3 12 "b" 1 3 1 MR 4 13 "" MR MR MR MR 5 14 "" MR MR MR MR 6 15 "" 1 2 3 MR # (01) Only keep rows where respondent has missing responses for AT LEAST ONE variable (var1:var4). df %>% filter(if_any(var1:var4, ~ . == "MR")) # A tibble: 5 x 6 id form var1 var2 var3 var4 <dbl> <chr> <chr> <chr> <chr> <chr> 1 11 "b" 3 MR 2 4 2 12 "b" 1 3 1 MR 3 13 "" MR MR MR MR 4 14 "" MR MR MR MR 5 15 "" 1 2 3 MR # (02) Only keep rows where respondent has missing responses for ALL variables (var1:var4). df %>% filter(if_all(var1:var4, ~ . == "MR")) # A tibble: 2 x 6 id form var1 var2 var3 var4 <dbl> <chr> <chr> <chr> <chr> <chr> 1 13 "" MR MR MR MR 2 14 "" MR MR MR MR # (03) Remove rows where respondent has missing responses for ALL variables (var1:var4). df %>% filter(!if_all(var1:var4, ~ . == "MR")) # A tibble: 4 x 6 id form var1 var2 var3 var4 <dbl> <chr> <chr> <chr> <chr> <chr> 1 10 "a" 3 4 3 3 2 11 "b" 3 MR 2 4 3 12 "b" 1 3 1 MR 4 15 "" 1 2 3 MR
August 17, 2025 at 8:15 AM