Randy Boyes
@randy.pub
Analytics at presagegroup.com | PhD Epidemiology | #julialang #rstats | publications at www.randy.pub | GitHub @rdboyes, currently working on https://github.com/TidierOrg/TidierPlots.jl | 🐦@randyboyes
I typically just use `factor(x)` but if you want names other than TRUE and FALSE I don't think you're getting a cleaner solution than what you have, unless you prefer something like `factor(c(TRUE, FALSE)) |> fct_recode(Yes = "TRUE", No = "FALSE")`
November 17, 2025 at 11:18 AM
I typically just use `factor(x)` but if you want names other than TRUE and FALSE I don't think you're getting a cleaner solution than what you have, unless you prefer something like `factor(c(TRUE, FALSE)) |> fct_recode(Yes = "TRUE", No = "FALSE")`
As someone who occasionally administers tests like this - it’s a negative signal, but not a dealbreaker necessarily. It suggests bad practices though
November 17, 2025 at 12:10 AM
As someone who occasionally administers tests like this - it’s a negative signal, but not a dealbreaker necessarily. It suggests bad practices though
would have been extremely relevant. Nothing beats R for quick visualizations, and doing ML in anything other than Python is painful. Right tool for the right task, basically, and the syntax for whatever language you use most is going to look best
(Unless that is python, ew)
(Unless that is python, ew)
November 14, 2025 at 11:15 AM
would have been extremely relevant. Nothing beats R for quick visualizations, and doing ML in anything other than Python is painful. Right tool for the right task, basically, and the syntax for whatever language you use most is going to look best
(Unless that is python, ew)
(Unless that is python, ew)
I think “data science” is too broad for there to be such a thing as “a good language for data science” in general. He says performance considerations are secondary, but when I was doing my PhD, I had R code that would take multiple days to run, and the 100x speedup you can get in Julia …
November 14, 2025 at 11:12 AM
I think “data science” is too broad for there to be such a thing as “a good language for data science” in general. He says performance considerations are secondary, but when I was doing my PhD, I had R code that would take multiple days to run, and the 100x speedup you can get in Julia …
Isn’t this just regular recursion? Either way, I don’t think you’re missing anything from a practical point of view - writing a loop is almost always going to be clearer
November 6, 2025 at 5:24 PM
Isn’t this just regular recursion? Either way, I don’t think you’re missing anything from a practical point of view - writing a loop is almost always going to be clearer
You've already extracted a lot of the actual "learning" benefit - I guess the question is how much you need the piece of paper. Do you want to be an academic (more to the point - do you *only* want to be an academic)? I finished mine years late and running on spite; I don't think it was all upside
November 3, 2025 at 1:50 PM
You've already extracted a lot of the actual "learning" benefit - I guess the question is how much you need the piece of paper. Do you want to be an academic (more to the point - do you *only* want to be an academic)? I finished mine years late and running on spite; I don't think it was all upside
Ah - I misunderstood what you meant by "cars" and thought "why would anyone buy those"
November 3, 2025 at 1:38 PM
Ah - I misunderstood what you meant by "cars" and thought "why would anyone buy those"
The only reasonable rail route in ON is between GTA-Montreal, and even that is just “as fast as driving”
November 3, 2025 at 1:26 AM
The only reasonable rail route in ON is between GTA-Montreal, and even that is just “as fast as driving”
Project big life (www.projectbiglife.ca) has some nice examples of public-facing prediction model interfaces
Project Big Life
Health Calculators for a Big Life Powered by Big Data
www.projectbiglife.ca
October 31, 2025 at 12:51 PM
Project big life (www.projectbiglife.ca) has some nice examples of public-facing prediction model interfaces
Technically the question was “can you import pickles” not “can you do anything with the objects once they’re in R” ;)
October 31, 2025 at 12:42 PM
Technically the question was “can you import pickles” not “can you do anything with the objects once they’re in R” ;)
It’s so good - I’ll be shocked if it doesn’t win. The only thing that brought it down a little for me was act 3 felt a little weak compared to the first 2
October 31, 2025 at 9:33 AM
It’s so good - I’ll be shocked if it doesn’t win. The only thing that brought it down a little for me was act 3 felt a little weak compared to the first 2
Left is a Table, right could be either
October 31, 2025 at 9:28 AM
Left is a Table, right could be either
Should be able to do that with reticulate rstudio.github.io/reticulate/r...
Save and Load Python Objects — py_save_object
Save and load Python objects.
rstudio.github.io
October 31, 2025 at 12:55 AM
Should be able to do that with reticulate rstudio.github.io/reticulate/r...
Wait what was the product they had that was supposed to compete with steam? I thought they just made games (poorly)
October 29, 2025 at 2:05 PM
Wait what was the product they had that was supposed to compete with steam? I thought they just made games (poorly)
There’s an @aes user that I’ve tagged a bunch by accident while making tidierplots - sorry Anders
October 28, 2025 at 8:44 PM
There’s an @aes user that I’ve tagged a bunch by accident while making tidierplots - sorry Anders
̶c̶o̶n̶f̶o̶u̶n̶d̶e̶r̶ covariate
- we thank the reviewer for their insightful suggestion
- we thank the reviewer for their insightful suggestion
October 28, 2025 at 5:02 PM
̶c̶o̶n̶f̶o̶u̶n̶d̶e̶r̶ covariate
- we thank the reviewer for their insightful suggestion
- we thank the reviewer for their insightful suggestion
Getting a “side quest” in R is the best! The hard part is always convincing people that no actually this is fun for me
October 27, 2025 at 11:39 AM
Getting a “side quest” in R is the best! The hard part is always convincing people that no actually this is fun for me
Imagine this has all been a long con and ends with him dumping satoshis ~million bitcoin on the market
Would not make it worth it but that’s a pretty good silver lining
Would not make it worth it but that’s a pretty good silver lining
October 25, 2025 at 11:04 PM
Imagine this has all been a long con and ends with him dumping satoshis ~million bitcoin on the market
Would not make it worth it but that’s a pretty good silver lining
Would not make it worth it but that’s a pretty good silver lining
The only thing that forces me to keep a windows 11 drive is office (for work, it’s not really optional) - ironically if the web versions of office weren’t such garbage I would be able to cut ties completely
October 24, 2025 at 5:50 PM
The only thing that forces me to keep a windows 11 drive is office (for work, it’s not really optional) - ironically if the web versions of office weren’t such garbage I would be able to cut ties completely
I look forward to this every year, but I usually get distracted from it around day 12 anyway with holiday commitments. Good change, and thanks for the puzzles!
October 23, 2025 at 10:28 PM
I look forward to this every year, but I usually get distracted from it around day 12 anyway with holiday commitments. Good change, and thanks for the puzzles!
Should be a join that only grabs a single column from the second dataframe
October 22, 2025 at 5:35 PM
Should be a join that only grabs a single column from the second dataframe