Roadmap: parsnip support for sparse tibbles

[EmilHvitfeldt]

What we need:

• fit() to take sparse tibbles as data Allow sparse tibbles in fit() and fit_xy() #1165
• fit() to take {Matrix} sparse matrix as data let fit_xy() take dgCMatrix input #1121
  • turn them into sparse tibbles early
• predict() to take sparse tibbles as data Make sure sparse matrices can be used with predict() #1167
• predict() to take {Matrix} sparse matrix as data Make sure sparse matrices can be used with predict() #1167
  • turn them into sparse tibbles early
• look into if we document which engines are sparse friendly document sparse data usage in parsnip #1171
• special cases for some model types make sure xgboost works with sparse data #1173
  • {xgboost} with xgboost::xgb.DMatrix() and xgb_train() (only special case engine that allows sparse data)
• Better error if sparse matrix is used with fit() Make sure all sparse data errors look nice #1174
• Make sure all error messages are piped through correctly Make sure all sparse data errors look nice #1174

I think we could use a option() of some kind to unit test that the data passed is passed around in a way that keeps the sparsity.

Adding all of this will give us

• standalone usage of sparse matrices in {parsnip}
• everything it needs to be able to work with the rest of {tidymodels} in regards to sparse tibbles

[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.