Giles
@gdeejay.bsky.social
Applied economist working in international development policy
🇹🇴🇸🇧🇻🇺🇫🇯🇦🇺🇲🇲 (etc),
Mostly memes.
Sometimes post about #publicpolicy, research, data science, #PFM and #Rstats
Economist | Independent Consultant
All views my own
🇹🇴🇸🇧🇻🇺🇫🇯🇦🇺🇲🇲 (etc),
Mostly memes.
Sometimes post about #publicpolicy, research, data science, #PFM and #Rstats
Economist | Independent Consultant
All views my own
Pinned
Giles
@gdeejay.bsky.social
· Oct 29
an_intro_to_r_for_policy_analysis
policyanalysislab.com
Dear #rstats,
Please find attached the alpha version of my book "An Introduction to R for Policy Analysis":
policyanalysislab.com/shared/r-for...
Unfortunately, the draft has been released too late to be considered for a Nobel prize, but I look forward to being nominated in 2026.
...
Please find attached the alpha version of my book "An Introduction to R for Policy Analysis":
policyanalysislab.com/shared/r-for...
Unfortunately, the draft has been released too late to be considered for a Nobel prize, but I look forward to being nominated in 2026.
...
November 30, 2025 at 1:38 AM
I don’t know if anyone else notices or cares, but when I see a presentation in which the speaker uses obviously generated-AI images to illustrate their slides, it makes me immediately less confident in whatever other content they’re presenting.
November 29, 2025 at 10:03 PM
I'm struggling with the flu today, so have spent a lot of today playing No Man's Sky.
I agree with the critique that it's a mile wide and an inch deep, but something about it has really triggered my hyper-focus today.
🤔
I agree with the critique that it's a mile wide and an inch deep, but something about it has really triggered my hyper-focus today.
🤔
November 29, 2025 at 6:23 AM
I'm struggling with the flu today, so have spent a lot of today playing No Man's Sky.
I agree with the critique that it's a mile wide and an inch deep, but something about it has really triggered my hyper-focus today.
🤔
I agree with the critique that it's a mile wide and an inch deep, but something about it has really triggered my hyper-focus today.
🤔
This post was published on @r-bloggers.bsky.social for some reason.
Apologies as it has nothing to do with #rstats, but it's probably good to make the data easier to find.
Facing some teething problems with my new website...
Apologies as it has nothing to do with #rstats, but it's probably good to make the data easier to find.
Facing some teething problems with my new website...
This is unlikely to be of interest to anyone here, but I've made India's 2015 National Classification of Occupations available as a data file here:
www.gilesd-j.com/2025/11/28/i...
...
www.gilesd-j.com/2025/11/28/i...
...
India’s 2015 National Classification of Occupations Data File – Giles
www.gilesd-j.com
November 28, 2025 at 8:31 PM
This post was published on @r-bloggers.bsky.social for some reason.
Apologies as it has nothing to do with #rstats, but it's probably good to make the data easier to find.
Facing some teething problems with my new website...
Apologies as it has nothing to do with #rstats, but it's probably good to make the data easier to find.
Facing some teething problems with my new website...
This is unlikely to be of interest to anyone here, but I've made India's 2015 National Classification of Occupations available as a data file here:
www.gilesd-j.com/2025/11/28/i...
...
www.gilesd-j.com/2025/11/28/i...
...
India’s 2015 National Classification of Occupations Data File – Giles
www.gilesd-j.com
November 28, 2025 at 11:09 AM
This is unlikely to be of interest to anyone here, but I've made India's 2015 National Classification of Occupations available as a data file here:
www.gilesd-j.com/2025/11/28/i...
...
www.gilesd-j.com/2025/11/28/i...
...
Look, I'm not a software developer. I mainly work with data in #rstats, but it's fascinating to see the split between programmers who hate LLMs vs those that love them.
I wonder what the difference is? Is it how they use LLMs, think about the problem or the type of work they do?
🤔
I wonder what the difference is? Is it how they use LLMs, think about the problem or the type of work they do?
🤔
At the risk of starting the flame war to end all flame wars...
Modern LLMs (GPT-5.1, Claude 4.5, Gemini 3) produce excellent code and can be a significant productivity boost to software engineers who take the time to learn how to effectively apply them - especially if used with coding agent tools
Modern LLMs (GPT-5.1, Claude 4.5, Gemini 3) produce excellent code and can be a significant productivity boost to software engineers who take the time to learn how to effectively apply them - especially if used with coding agent tools
November 27, 2025 at 10:53 PM
Look, I'm not a software developer. I mainly work with data in #rstats, but it's fascinating to see the split between programmers who hate LLMs vs those that love them.
I wonder what the difference is? Is it how they use LLMs, think about the problem or the type of work they do?
🤔
I wonder what the difference is? Is it how they use LLMs, think about the problem or the type of work they do?
🤔
Uhh,,, Figure 1 shows you... what exactly? Trying to understand Medical fryrmbial, runctitional features and mum's legs going through concrete.
This whole article is a bit of a disaster. And it's very difficult to find other published work for the author. Strange! 🧪
(via @smutclyde.bsky.social)
This whole article is a bit of a disaster. And it's very difficult to find other published work for the author. Strange! 🧪
(via @smutclyde.bsky.social)
November 27, 2025 at 10:06 AM
I'm going to have some spare time in the coming weeks, which I'm obviously going to take advantage of again by dual booting Linux again for reasons.
November 27, 2025 at 7:46 AM
I'm going to have some spare time in the coming weeks, which I'm obviously going to take advantage of again by dual booting Linux again for reasons.
Me during today's lunchclub call when asked if I'd heard of Henry George and his land tax.
November 27, 2025 at 2:30 AM
Me during today's lunchclub call when asked if I'd heard of Henry George and his land tax.
You know when you kind of need to sneeze, but can't?
Sneeze limbo.
That's what I'm experiencing ATM and will be skeeting about for the next hour.
You're welcome #bluesky.
Sneeze limbo.
That's what I'm experiencing ATM and will be skeeting about for the next hour.
You're welcome #bluesky.
Another day, another set of simpletons asking pointless questions. My intellect is wasted here. You wouldn't comprehend my real values even if I etched them onto your corneas. Now, cease your incessant buzzing.
November 26, 2025 at 11:56 PM
You know when you kind of need to sneeze, but can't?
Sneeze limbo.
That's what I'm experiencing ATM and will be skeeting about for the next hour.
You're welcome #bluesky.
Sneeze limbo.
That's what I'm experiencing ATM and will be skeeting about for the next hour.
You're welcome #bluesky.
So, the reason I discovered this is that I wanted to use local LLMs to help me analyze an unstructured text document.
I basically had huge chunks of uncategorized text, which I first manually cleaned, categorized into themes and divided into ideas.
...
I basically had huge chunks of uncategorized text, which I first manually cleaned, categorized into themes and divided into ideas.
...
...
(This merges text cells from a transcript according to the idea being discussed / general theme)
(This merges text cells from a transcript according to the idea being discussed / general theme)
November 26, 2025 at 11:34 PM
So, the reason I discovered this is that I wanted to use local LLMs to help me analyze an unstructured text document.
I basically had huge chunks of uncategorized text, which I first manually cleaned, categorized into themes and divided into ideas.
...
I basically had huge chunks of uncategorized text, which I first manually cleaned, categorized into themes and divided into ideas.
...
One of the things I learned early in my career is that extraordinary people are often just ordinary people with extra resources.
Like, all those op-eds and journal articles published under their name that make them look prolific: written by interns and junior employees unable to push back.
Like, all those op-eds and journal articles published under their name that make them look prolific: written by interns and junior employees unable to push back.
November 26, 2025 at 8:55 PM
One of the things I learned early in my career is that extraordinary people are often just ordinary people with extra resources.
Like, all those op-eds and journal articles published under their name that make them look prolific: written by interns and junior employees unable to push back.
Like, all those op-eds and journal articles published under their name that make them look prolific: written by interns and junior employees unable to push back.
Just hit the 50 percent mark.
Still enjoying it. In fact, it's getting better!
👌🏼
Still enjoying it. In fact, it's getting better!
👌🏼
One thing I've noticed when reading this is that you can't skim a section without risking losing the narrative.
I find it to be written really well, but have had to go back a lot when I've missed something important about a scene.
I find it to be written really well, but have had to go back a lot when I've missed something important about a scene.
Okay, I finished Catcher in the Rye and am two chapters into Crime and Punishment by Dostoevsky.
November 26, 2025 at 11:09 AM
Just hit the 50 percent mark.
Still enjoying it. In fact, it's getting better!
👌🏼
Still enjoying it. In fact, it's getting better!
👌🏼
My wife got me the best Christmas present ever:
A personalized message from a celebrity rooster.
😎
A personalized message from a celebrity rooster.
😎
November 26, 2025 at 3:49 AM
My wife got me the best Christmas present ever:
A personalized message from a celebrity rooster.
😎
A personalized message from a celebrity rooster.
😎
Well, I just learned a new #rstats thing.
You can use group_by() to merge cells by a categorical variable:
dta_transcript_merged <- dta_transcript |>
group_by(idea_snippet, idea_theme) |>
summarise(
text = paste(text, collapse = " ")
) |>
ungroup()
...
You can use group_by() to merge cells by a categorical variable:
dta_transcript_merged <- dta_transcript |>
group_by(idea_snippet, idea_theme) |>
summarise(
text = paste(text, collapse = " ")
) |>
ungroup()
...
November 26, 2025 at 3:34 AM
Well, I just learned a new #rstats thing.
You can use group_by() to merge cells by a categorical variable:
dta_transcript_merged <- dta_transcript |>
group_by(idea_snippet, idea_theme) |>
summarise(
text = paste(text, collapse = " ")
) |>
ungroup()
...
You can use group_by() to merge cells by a categorical variable:
dta_transcript_merged <- dta_transcript |>
group_by(idea_snippet, idea_theme) |>
summarise(
text = paste(text, collapse = " ")
) |>
ungroup()
...
Reposted by Giles
I’m confused about why he didn’t get the job. He answered the question correctly.
November 26, 2025 at 12:26 AM
I’m confused about why he didn’t get the job. He answered the question correctly.
I heard that my automated budget analysis tool has been successfully used in another parliamentary vote.
So cool.
😎
Stay in school, folks.
(And get good with Microsoft Excel. 😬)
So cool.
😎
Stay in school, folks.
(And get good with Microsoft Excel. 😬)
November 26, 2025 at 1:07 AM
I heard that my automated budget analysis tool has been successfully used in another parliamentary vote.
So cool.
😎
Stay in school, folks.
(And get good with Microsoft Excel. 😬)
So cool.
😎
Stay in school, folks.
(And get good with Microsoft Excel. 😬)
I calculated my consulting revenue for the year and actually did better than expected.
Good. Not great.
But, better than expected given global economic uncertainty, USAID cuts and the budget shortfalls experienced by global organizations.
🙂↕️
Good. Not great.
But, better than expected given global economic uncertainty, USAID cuts and the budget shortfalls experienced by global organizations.
🙂↕️
November 24, 2025 at 3:42 AM
I calculated my consulting revenue for the year and actually did better than expected.
Good. Not great.
But, better than expected given global economic uncertainty, USAID cuts and the budget shortfalls experienced by global organizations.
🙂↕️
Good. Not great.
But, better than expected given global economic uncertainty, USAID cuts and the budget shortfalls experienced by global organizations.
🙂↕️
Today I tried kettle corn for the first time.
November 23, 2025 at 6:35 AM
Today I tried kettle corn for the first time.
I had to pick another subject in high school and the only one left was economics.
In the first class the teacher explained the concept of supply and demand, which made me think "being good at this could help a lot of people!".
Then all of this *gestures widely at everything* happened.
In the first class the teacher explained the concept of supply and demand, which made me think "being good at this could help a lot of people!".
Then all of this *gestures widely at everything* happened.
November 23, 2025 at 5:50 AM
I had to pick another subject in high school and the only one left was economics.
In the first class the teacher explained the concept of supply and demand, which made me think "being good at this could help a lot of people!".
Then all of this *gestures widely at everything* happened.
In the first class the teacher explained the concept of supply and demand, which made me think "being good at this could help a lot of people!".
Then all of this *gestures widely at everything* happened.
The West Wing, but swoon worthy.
November 23, 2025 at 5:42 AM
The West Wing, but swoon worthy.
library(tidyveRse) amirite #rstats?!
capitalizing the R in package names is the WOrST TrEND EVEr cc @mrcaseb.com
November 23, 2025 at 5:35 AM
library(tidyveRse) amirite #rstats?!
