Closed vincentgaul closed 4 years ago
If you are using a parsnip model (i.e. not tune) you might look into data descriptors such as .cols()
. You can specify a model such as:
library(parsnip)
rand_forest(mode = "classification", mtry = .cols() - 2)
There is not support for this in tune and dials currently.
This issue has been automatically locked. If you believe you have found a related problem, please file a new issue (with a reprex: https://reprex.tidyverse.org) and link to this issue.
mtry() for example is dependant on data size
I don't know if this is currently possible, but would it be useful to reference a recipe so grid can be made of the number of predictors? For example my df may have 50 columns, but 5 might have an "ID" role