Eduardo Leoni
@domleoni.bsky.social
When I am feeling down, I turn to two songs help me. They have basically the same ideas, from two separate timepoints, one in language, one in Portuguese. First, this banger by @themountaingoats.bsky.social www.youtube.com/watch?v=H7fB... "I am going to make it through this year if it kills me ..."
the Mountain Goats - This Year (Official Video)
YouTube video by the Mountain Goats
www.youtube.com
November 14, 2025 at 1:19 PM
When I am feeling down, I turn to two songs help me. They have basically the same ideas, from two separate timepoints, one in language, one in Portuguese. First, this banger by @themountaingoats.bsky.social www.youtube.com/watch?v=H7fB... "I am going to make it through this year if it kills me ..."
Reposted by Eduardo Leoni
In an Australia first, I used FOI to get the AI chat logs of Australia's top national security bureaucrat.
They show how he used Microsoft Copilot to ghostwrite speeches and messages to colleagues.
This comes as the government looks to push AI use in the public service.
www.crikey.com.au/20...
They show how he used Microsoft Copilot to ghostwrite speeches and messages to colleagues.
This comes as the government looks to push AI use in the public service.
www.crikey.com.au/20...
November 12, 2025 at 2:01 AM
In an Australia first, I used FOI to get the AI chat logs of Australia's top national security bureaucrat.
They show how he used Microsoft Copilot to ghostwrite speeches and messages to colleagues.
This comes as the government looks to push AI use in the public service.
www.crikey.com.au/20...
They show how he used Microsoft Copilot to ghostwrite speeches and messages to colleagues.
This comes as the government looks to push AI use in the public service.
www.crikey.com.au/20...
Reposted by Eduardo Leoni
I've found similar numerical errors/inconsistencies in 3 other peer-reviewed papers by subsets of these authors. (So far!)
So this is clearly not a one-off error caused by janky formatting or something. It's a systematic workflow problem.
So this is clearly not a one-off error caused by janky formatting or something. It's a systematic workflow problem.
November 2, 2025 at 4:06 PM
I've found similar numerical errors/inconsistencies in 3 other peer-reviewed papers by subsets of these authors. (So far!)
So this is clearly not a one-off error caused by janky formatting or something. It's a systematic workflow problem.
So this is clearly not a one-off error caused by janky formatting or something. It's a systematic workflow problem.
Reposted by Eduardo Leoni
Fun fact: it has been 56 days since I notified the editors of Neurology about glaring statistical errors in this peer-reviewed study on sweeteners and cognitive health.
No expression of concern, no correction, no retraction.
No expression of concern, no correction, no retraction.
"Sweeteners can harm cognitive health equivalent to 1.6 years of ageing, study finds"
Or does it? Let's take a look at this "study"...
www.theguardian.com/food/2025/se...
Or does it? Let's take a look at this "study"...
www.theguardian.com/food/2025/se...
Sweeteners can harm cognitive health equivalent to 1.6 years of ageing, study finds
Researchers say low- and no-calorie sweeteners appear to affect thinking and memory in middle age
www.theguardian.com
November 2, 2025 at 4:06 PM
Fun fact: it has been 56 days since I notified the editors of Neurology about glaring statistical errors in this peer-reviewed study on sweeteners and cognitive health.
No expression of concern, no correction, no retraction.
No expression of concern, no correction, no retraction.
Reposted by Eduardo Leoni
I was loving Claude Code... until I tried it with #rstats. Constant errors, wouldn't use the tidyverse even when asked, "optimized" functions were slower.
Frustrated, I started a session just to teach R to Claude and summarize what it learned into a CLAUDE.md file gist.github.com/sj-io/3828d6...
Frustrated, I started a session just to teach R to Claude and summarize what it learned into a CLAUDE.md file gist.github.com/sj-io/3828d6...
Claude R Tidyverse Expert
Claude R Tidyverse Expert. GitHub Gist: instantly share code, notes, and snippets.
gist.github.com
August 21, 2025 at 10:18 AM
I was loving Claude Code... until I tried it with #rstats. Constant errors, wouldn't use the tidyverse even when asked, "optimized" functions were slower.
Frustrated, I started a session just to teach R to Claude and summarize what it learned into a CLAUDE.md file gist.github.com/sj-io/3828d6...
Frustrated, I started a session just to teach R to Claude and summarize what it learned into a CLAUDE.md file gist.github.com/sj-io/3828d6...
Reposted by Eduardo Leoni
Hard to compete with this.
October 7, 2025 at 3:25 PM
Hard to compete with this.
Reposted by Eduardo Leoni
If memory serves I added this at some point on user request, so not sure how many years we had this but `anytime` does this by default (and `anytime()` does the same returning `POSIXct` instead of `Date`).
![R code showing month and day imputation for incomplete dates:
> anytime::anydate(c("2015", "2015-09", "2015-09-19"))
[1] "2015-01-01" "2015-09-01" "2015-09-19"
>](https://cdn.bsky.app/img/feed_thumbnail/plain/did:plc:hdbguwort62mhodqy7vjspud/bafkreidbxrcl3tazbdfh3jcwb4iha2bovy2a3bjx3cop3mocckubj3icmi@jpeg)
October 2, 2025 at 6:55 PM
If memory serves I added this at some point on user request, so not sure how many years we had this but `anytime` does this by default (and `anytime()` does the same returning `POSIXct` instead of `Date`).
Reposted by Eduardo Leoni
No longer on a weekly basis, but I could not resist ....
open.substack.com/pub/adamprze...
open.substack.com/pub/adamprze...
Musings 1
To readers of my weekly comments
open.substack.com
September 30, 2025 at 1:44 PM
No longer on a weekly basis, but I could not resist ....
open.substack.com/pub/adamprze...
open.substack.com/pub/adamprze...
Reposted by Eduardo Leoni
An abbreviation (ABB) in a journal article (JA) or Grant Application (GA) is rarely worth the words it saves. Every ABB requires cognitive resources (CR) and at my age by the time I'm halfway through a JA or GA I no longer have the CR to remember what your ABB stood for.
August 15, 2025 at 9:39 AM
An abbreviation (ABB) in a journal article (JA) or Grant Application (GA) is rarely worth the words it saves. Every ABB requires cognitive resources (CR) and at my age by the time I'm halfway through a JA or GA I no longer have the CR to remember what your ABB stood for.
Reposted by Eduardo Leoni
This is so very not good.
I have said before that even the President can't pervert top-tier official statistics secretly, but doing it publicly is certainly within his capabilities
I have said before that even the President can't pervert top-tier official statistics secretly, but doing it publicly is certainly within his capabilities
Some prescient person on here joked about this just an hour ago.
Trump demands firing of BLS chief after soft jobs report
The monthly jobs report showed just 73,000 jobs in July, with big reductions to May's and June's numbers
www.politico.com
August 1, 2025 at 7:15 PM
This is so very not good.
I have said before that even the President can't pervert top-tier official statistics secretly, but doing it publicly is certainly within his capabilities
I have said before that even the President can't pervert top-tier official statistics secretly, but doing it publicly is certainly within his capabilities
Reposted by Eduardo Leoni
Good stuff: Paul Krugman's general theory of enshittification even includes a little dynamic programming model.
The General Theory of Enshittification
It isn’t a new phenomenon, but it seems to matter more
paulkrugman.substack.com
July 26, 2025 at 6:12 AM
Good stuff: Paul Krugman's general theory of enshittification even includes a little dynamic programming model.
Reposted by Eduardo Leoni
1. Here's a fish swimming upstream. Nothing unusual about that.
What's unusual is that this particular fish is *dead*. Vortices in the water as it flows past the fish cause the fish's body to flex, maintaining orientation and actually propelling it forward.
(D. N. Beal et al 2006 J. Fluid. Mech.)
What's unusual is that this particular fish is *dead*. Vortices in the water as it flows past the fish cause the fish's body to flex, maintaining orientation and actually propelling it forward.
(D. N. Beal et al 2006 J. Fluid. Mech.)
July 26, 2025 at 7:34 AM
1. Here's a fish swimming upstream. Nothing unusual about that.
What's unusual is that this particular fish is *dead*. Vortices in the water as it flows past the fish cause the fish's body to flex, maintaining orientation and actually propelling it forward.
(D. N. Beal et al 2006 J. Fluid. Mech.)
What's unusual is that this particular fish is *dead*. Vortices in the water as it flows past the fish cause the fish's body to flex, maintaining orientation and actually propelling it forward.
(D. N. Beal et al 2006 J. Fluid. Mech.)
Reposted by Eduardo Leoni
The call for papers for LatinR 2025 (online) is now open! You can present in English, Spanish, or Portuguese 🗣️ #RStats latinr.org/en/blog/en/2...
Call for papers – LatinR 2024
latinr.org
July 25, 2025 at 10:15 AM
The call for papers for LatinR 2025 (online) is now open! You can present in English, Spanish, or Portuguese 🗣️ #RStats latinr.org/en/blog/en/2...
Reposted by Eduardo Leoni
This is actually a `data.table` idiom I use frequently: 't(ransposed)strsplit'.
Makes it a one-line / single statement. (I know `data.table` is not for everyone, but it concise, efficient, fast, ... and liked by some.) #rstats
Makes it a one-line / single statement. (I know `data.table` is not for everyone, but it concise, efficient, fast, ... and liked by some.) #rstats
![> df <- data.frame(Place = c("New York, NY", "Boston, MA", "Chicago, IL"))
> setDT(df) # data.table already loaded in this session
> df[, c("City", "State") := tstrsplit(Place, ",")]
> df
Place City State
<char> <char> <char>
1: New York, NY New York NY
2: Boston, MA Boston MA
3: Chicago, IL Chicago IL
>](https://cdn.bsky.app/img/feed_thumbnail/plain/did:plc:hdbguwort62mhodqy7vjspud/bafkreighech6cpu2uuoz24qwexsnhmoixjjj4go72c3uukzr72fio3dzbi@jpeg)
June 26, 2025 at 12:56 AM
This is actually a `data.table` idiom I use frequently: 't(ransposed)strsplit'.
Makes it a one-line / single statement. (I know `data.table` is not for everyone, but it concise, efficient, fast, ... and liked by some.) #rstats
Makes it a one-line / single statement. (I know `data.table` is not for everyone, but it concise, efficient, fast, ... and liked by some.) #rstats
Reposted by Eduardo Leoni
Better world maps for mobile!
Despite years of mobile-first design, there’s been little experimentation in this area. So we designed our own at NZZ – and validated it in a large user study.
Summary of the paper: jonasoesch.ch/articles/the... 1/3
Despite years of mobile-first design, there’s been little experimentation in this area. So we designed our own at NZZ – and validated it in a large user study.
Summary of the paper: jonasoesch.ch/articles/the... 1/3
May 14, 2025 at 9:38 AM
Better world maps for mobile!
Despite years of mobile-first design, there’s been little experimentation in this area. So we designed our own at NZZ – and validated it in a large user study.
Summary of the paper: jonasoesch.ch/articles/the... 1/3
Despite years of mobile-first design, there’s been little experimentation in this area. So we designed our own at NZZ – and validated it in a large user study.
Summary of the paper: jonasoesch.ch/articles/the... 1/3
Re: Implement a 100% tariff and ruin a movie
I am the only one confused by all the responses? With 100% tariffs you would get *less* of something, not *more*!
I am the only one confused by all the responses? With 100% tariffs you would get *less* of something, not *more*!
May 5, 2025 at 11:12 AM
Re: Implement a 100% tariff and ruin a movie
I am the only one confused by all the responses? With 100% tariffs you would get *less* of something, not *more*!
I am the only one confused by all the responses? With 100% tariffs you would get *less* of something, not *more*!
Reposted by Eduardo Leoni
Congratulations to Stefanie Stantcheva (@s-stantcheva.bsky.social) of @harvard.edu, winner of the 2025 John Bates Clark Medal! #econsky www.aeaweb.org/about-aea/ho...
Stefanie Stantcheva, Clark Medalist 2025
www.aeaweb.org
April 22, 2025 at 12:05 PM
Congratulations to Stefanie Stantcheva (@s-stantcheva.bsky.social) of @harvard.edu, winner of the 2025 John Bates Clark Medal! #econsky www.aeaweb.org/about-aea/ho...
Reposted by Eduardo Leoni
I was given a cracking fact today- the single largest reason for the declining birth rate is the successful prevention of teenage pregnancies, compounded with the fact that those who have their first child as a teenager are more likely to have bigger families overall.
April 12, 2025 at 4:30 PM
I was given a cracking fact today- the single largest reason for the declining birth rate is the successful prevention of teenage pregnancies, compounded with the fact that those who have their first child as a teenager are more likely to have bigger families overall.
Reposted by Eduardo Leoni
This is fantastic! This website allows you to watch TV stations from around the world. I just watched 2 minutes of Afghan volleyball. All the small regional TV stations from Germany are here too. Spread the word! Source: tv.garden
March 29, 2025 at 6:26 AM
This is fantastic! This website allows you to watch TV stations from around the world. I just watched 2 minutes of Afghan volleyball. All the small regional TV stations from Germany are here too. Spread the word! Source: tv.garden
Reposted by Eduardo Leoni
The key point is that with tariffs you may be primarily hurting yourself. And if indeed you think trump won’t feel the pain or will deny it, or will just use it then better to focus on growth strategies for Europe than follow down his path
March 12, 2025 at 2:34 PM
The key point is that with tariffs you may be primarily hurting yourself. And if indeed you think trump won’t feel the pain or will deny it, or will just use it then better to focus on growth strategies for Europe than follow down his path
Reposted by Eduardo Leoni
JD Vance talking about housing as a supply issue while part of the Tariffs and Construction Worker Deportations admin is basically the whole NatCon shtick writ large: you accurately diagnose the problem, talk eloquently about it’s impacts, then enthusiastically commit yourself to being part of it.
March 13, 2025 at 12:34 AM
JD Vance talking about housing as a supply issue while part of the Tariffs and Construction Worker Deportations admin is basically the whole NatCon shtick writ large: you accurately diagnose the problem, talk eloquently about it’s impacts, then enthusiastically commit yourself to being part of it.
Reposted by Eduardo Leoni
Now compare that to this cleaned version from the Global Tariff Database:
March 11, 2025 at 5:34 PM
Now compare that to this cleaned version from the Global Tariff Database:
Reposted by Eduardo Leoni
🚨 New paper (with Kasey Rhee & Nico Studen). We use a new within-precinct design to isolate how ideology affects vote choice holding turnout fixed, analyzing 3.4M precinct observations across state & fed elections (2016-2022).
tldr: Ideological moderation affects vote shares, but not by much. 🧵⬇️
tldr: Ideological moderation affects vote shares, but not by much. 🧵⬇️
The Electoral Consequences of Ideological Persuasion: Evidence from a Within-Precinct Analysis of U.S. Elections
Most research on the electoral penalty of candidate ideology relies on betweendistrict or longitudinal comparisons, which are confounded by turnout and ballot c
papers.ssrn.com
March 11, 2025 at 8:58 PM
🚨 New paper (with Kasey Rhee & Nico Studen). We use a new within-precinct design to isolate how ideology affects vote choice holding turnout fixed, analyzing 3.4M precinct observations across state & fed elections (2016-2022).
tldr: Ideological moderation affects vote shares, but not by much. 🧵⬇️
tldr: Ideological moderation affects vote shares, but not by much. 🧵⬇️
Reposted by Eduardo Leoni
Come work with us! In my research group, we have two four-year PhD positions available in comparative politics. Happy to answer questions per email. Please share the ad.
You can find the ad here: stellen.uni-konstanz.de/jobposting/d...
You can find the ad here: stellen.uni-konstanz.de/jobposting/d...
February 25, 2025 at 2:54 PM
Come work with us! In my research group, we have two four-year PhD positions available in comparative politics. Happy to answer questions per email. Please share the ad.
You can find the ad here: stellen.uni-konstanz.de/jobposting/d...
You can find the ad here: stellen.uni-konstanz.de/jobposting/d...