spatialsample::spatial_clustering_cv() for Tidymodels -- coords removed for prediction or no?

[PathosEthosLogos]

I was just wondering if setting spatial_clustering_cv(v, coords = c(x, y) would keep x and y as predictors or would be removed (like id or strata would be).

How can I check this?

[hfrick]

I'm not entirely sure what you mean here. rsample and spatialsample do not have predictors, they only resample the data. No columns get removed. Does this answer help you?

[PathosEthosLogos]

I'm not entirely sure what you mean here. rsample and spatialsample do not have predictors, they only resample the data. No columns get removed. Does this answer help you?

Thanks for the reply @hfrick. Sorry it was unclear.
The scenario is:

• set up a spatial cross validation with the two columns for the argument coords
• set up recipe (y ~ .)
• tune hyperparameters for elastic net (or any other algorithm)

One example I would like to cite is update_role(..., new_role = 'ID'). Once role is applied to a particular column, they are removed as predictors. I was wondering if coords argument does something similar.

[hfrick]

Ah, I see, thanks. Resampling only splits the data. The columns you declare as coords will still be part of the data and their role will be what you declare in the recipe.

[mikemahoney218]

Let us know if you have further clarifications or questions 😄

[github-actions]

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.