Closed deblnia closed 4 years ago
Definitely in my thoughts (see #230). I'm trying to re-envision the whole stat learning module. It covers two days and I presume no prior math background, so I have to start from square one essentially. I got rid of cross-validation and tree-based models because most students didn't seem to really understand how they worked (so much copy and pasting from the lecture notes). But I really like the tidymodels
interface and I think it will be very useful for modelers moving forward.
That said, I'm not entirely comfortable shifting to it entirely. I still think students need to understand how to call some of the core model functions like lm()
and interact with those objects.
Closing to avoid redundancy!
Not sure how you feel about this -- definitely a difficult change to make, presumes some stat background, and probably goes along the lines of updating to encompass more of dplyr 1.0 -- but I've found the pre-processing framework fairly lovely, the documentation is good, and there are a few nice tutorials out there.
Talk soon 😄