LoginSign Up
← Back to all sparks
R

recipes

AI-ASSISTANTS
Velocity0.0

Preprocessing and feature engineering for R

github.com/tidymodels/recipes ↗

tidymodels' preprocessing engine learned sparsity, then settled into deprecations.

tidymodelspreprocessingsparse-datadeprecationsr-language
◆Current state
recipes is at 1.3.3, whose entire changelog is one suggested-package declaration. The substantive release in the window is 1.2.0, which taught recipe, prep and bake to work with sparse tibbles and sparse matrices, added a sparse argument to eight dummy and indicator steps, and made seventeen more steps preserve sparsity they receive. Since then the work has been deprecations and bug fixes.
◆Where it's heading
The direction is consolidation of a large step catalogue rather than growth. step_select and step_nnmf have entered deprecation, arguments across nine steps moved from strings and vars() calls to bare names, and all steps now require the same four arguments. The sparse work stands as the last structural change; what follows tidies the surface around it.
◆Prediction
With step_select mid-deprecation and step_nnmf newly deprecated in favour of step_nnmf_sparse, the next release most likely advances those deprecations rather than adding steps.

◆Recent moves

  1. 2mo ago

    Declares mixOmics as a suggested package

    A one-line release moving mixOmics to suggested packages. Dependency metadata only.

    View source ↗
  2. 4mo ago

    step_nnmf() deprecated in favour of step_nnmf_sparse()

    step_nnmf() is deprecated in favour of step_nnmf_sparse(), folding the older step into the sparse line 1.2.0 established. A step_pls() bake failure with scale = FALSE is fixed.

    View source ↗
  3. 1y ago

    Fixes tune_args() with tuned parsnip arguments

    A single fix for tune_args() erroring when a step argument held a parsnip object with tuned arguments. Narrow and internal.

    View source ↗
  4. 1y ago

    Bare-name arguments across nine steps; step_select() deprecated

    A broad tidy-up: nine steps now accept bare column names instead of strings or vars() calls, step_select() begins deprecation, step_impute_bag() prepares with a much smaller memory footprint, and several steps stop erroring on empty selections and NA values.

    View source ↗
  5. 1y ago

    Fixes sparsity steps applied to derived variables

    One bug fix, for sparsity-creation steps erroring on variables produced by earlier steps. A follow-up to the 1.2.0 sparse work.

    View source ↗
  6. 1y ago

    Sparse tibbles and sparse matrices supported end to end

    ⚡ SPARK

    The structural release of the window: sparse data becomes a first-class input and output across the preprocessing pipeline, not a conversion users perform around it. Everything after this consolidates rather than extends.

    View source ↗