Educational attainment in #TidyTuesday UK towns | Julia Silge

[utterances-bot]

Educational attainment in #TidyTuesday UK towns | Julia Silge

A data science blog

https://juliasilge.com/blog/educational-attainment/

[gunnergalactico]

Hello Julia, not sure if this question has been asked before, I couldn't find anything about it online. Is it possible to use Apple Silicon's GPU to train models in RStudio?

Thanks

[juliasilge]

@gunnergalactico I know you can when you are using Tensorflow (for example, via the keras R package):
https://developer.apple.com/metal/tensorflow-plugin/

I don't think GPUs help much with other kinds of models, like linear or tree-based models, so folks don't really bother to set up explicit support for using GPUs.

[olusolaxx]

Hi Julia,

Am reading your book: Tidy modeling with R. In Chapter 6, you ran the code as

tidy(lm_form_fit)

but I dd as shown below
lm_form_fit |>
extract_fit_engine() |>
summary() |>
coef() |>
tidy()
I got the warning (Warning: 'tidy.numeric' is deprecated. See help("Deprecated"))

Another issue in Chapter 8, when I ran the code shown below.

lm_wflow <-
lm_wflow |>
remove_variables() |>
add_recipe(simple_ames)

I saw warning below. What could be responsible?

Warning: The workflow has no variables preprocessor to remove.
Error in add_recipe():
! A recipe cannot be added when a formula already exists.
Backtrace:

1. workflows::add_recipe(remove_variables(lm_wflow), simple_ames)

[juliasilge]

@olusolaxx Yep, using tidy() on numeric data is deprecated so I would recommend doing something like tidy(lm_fit).

I am able to run the code that does remove_variables() without any problems. I recommend that you create a reprex (a minimal reproducible example) showing the problem you are experiencing, then posting on Posit Community. It's a great forum for getting help with these kinds of modeling questions. Good luck! 🙌