step_discretize in recipes: can I supply my own cuts?

Hi all,

I am learning how to integrate recipes/parsnip into my ML workflow, and I have a recipes related question:

If I want to discretize the target variables into 4 equal buckets, I see that I can use "cuts = 4". Is there a way for me to manually select the cuts (say, >0 vs <= 0 , or cut at -10, -1, 1, 10) -- or to get 4 equally sized cuts (e.g. the length of x in each should be equal)?


Right now the cuts generate equal frequencies in the training set.

But there is a GH issue for what you are interested in (so "not yet")

1 Like