Daniel Anderson

Feel like tonight was a bit of momentous occasion. My daughter Malia wrote her first ever function(s)! Written in 🐍 (I showed her #rstats first but she liked python more 🤷‍♂️). She already has the programmer look down too!

@dgkeyes My preference is to prefer the former, and use the latter when passing to an argument other than the first. For example, splt_mtcars <- split(mtcars, mtcars$cyl) map(splt_mtcars, ~lm(mpg ~ hp, data = .x)) You could still do `map(splt_mtcars, lm, formula = mpg ~ hp)`, but 😵‍💫

@dgkeyes I think I would describe it by saying that in the first you're directly passing the function to be applied to each thing. In the second, the open parens is making it expect something to be passed to it. So you'd have to do, e.g., `pwalk(~rmarkdown::render(.x))`.

Why do functions used within walk/map functions from purrr not use open and closed parentheses? #rstats Why this? reports %>% pwalk(rmarkdown::render) And not this? reports %>% pwalk(rmarkdown::render())

Today is the day I’m finally going to delete the bird app from my phone. I’m not leaving here entirely yet but I plan to soon-ish. I’m mostly on post now. Find me there. post.news/datalorax

@healthandstats @tjmahr @rbjanis5 Post is the way!

It's early but I am really liking the feel of post. Feels much more intuitive/natural than mastodon (to me anyway). Find me here: post.news/1456e2a3-0640-…

@healthandstats Yeah minus the underscore (which I didn’t really want here either but 🤷‍♂️)

@healthandstats I have high hopes for it. I haven’t fully setup my profile but it “feels” better to me than mastodon.

@JosiahParry @andrewheiss @tjmahr *built this into a package...

@JosiahParry When I need to do this I use a function to preview it (over and over and over again). @andrewheiss has built this into a function, and I believe I first saw it from @tjmahr github.com/andrewheiss/ra…

I just signed up for the waitlist for Post as another possible alternative to Twitter. I think it has potential. Join me? post.news/?r=rRZeZ

Hi all! I’m supporting an amazing doc student who wants to apply for this @NIDAnews @NIHFunding R36. Does anyone have experience with this mechanism? Any tips are appreciated; it seems similar to an R21 grants.nih.gov/grants/guide/p…

@andrewheiss Yup, I think that's right.

@andrewheiss I think you can only do that after making the prediction for the individual unit. So add in the deviation for the specific year or state or whatever and then plogis.

@andrewheiss I believe it's the variance in the log-odds. I have some slides on this that used an example from Gelman and Hill, if they are helpful. #165" target="_blank" rel="nofollow noopener">mlm2-2021.netlify.app/slides/w8p1.ht…

Well I'm officially on mastodon @datalorax@mastodon.cloud. I have no idea what I'm doing and it feels very strange, but if you're there too I hope we can connect! I'm hoping to leave twitter because I despise musk. Hoping the #rstats community there can grow.

@healthandstats @VincentAB @tjmahr @CMastication When I left UO, we were finally talking about abandoning the two-course sequence in ANOVA (!) and teaching everything through GLM, so there is hope. It’s just amazing how slow we move sometimes.

@healthandstats @VincentAB @tjmahr @CMastication I always tried to teach everything through the GLM, but in my field it’s almost required that you know what an ANOVA is because it’s still very common practice. And the equivalence to regression is still not as widely known as it should be, which is depressing.

@hardsci Congratulations! My career stage was obviously much different from yours but, FWIW, making the switch for me has been really reinvigorating. I hope you have a similar experience!

@dpshow Ducks - Medium!

We wanna give out some Merch... If you're a fan of any of the following College Football teams and plan on attending a game this season, respond to this tweet with your team and shirt size. We'll pick a couple people at random #DanPatrickSaturdays (Teams in Next Tweet)

Associated Press is seeking candidates to help scrape election results on Nov. 8. Experience with web scraping required; Python skills preferred. Work to be done remotely on Nov 8 from 5pm ET to 3 am. Candidates will attend 2-hr training. Pay starts at $20/hr, students welcome