fix: gg_variable.randomForest classification — yhat.* columns + smooth bugs (#87)

[ehrlinger]

Summary

• `gg_variable.randomForest` for classification forests now stores per-class OOB vote fractions as `yhat.` columns (from `object$votes`), matching the rfsrc path. Previously a single `yhat` factor (class labels from `object$predicted`) was stored, preventing the multi-class pivot in `plot.gg_variable` from firing. Vote fractions are row-normalised to [0, 1] even when fit with `norm.votes = FALSE`.
• Binary classification `smooth = TRUE`: added explicit `aes(x, y)` to `geom_smooth` so `layer_data()` no longer errors.
• Multi-class numeric path: added missing `smooth = TRUE` block.
• 6 new unit tests + 2 vdiffr snapshots.

Housekeeping

Closes #81, Closes #82 (both were fully fixed by PR #83 but auto-close never fired).

Test plan

• `devtools::test()` — all tests pass, 0 failures
• `devtools::check(args="--as-cran")` — 0 errors, 0 warnings, 0 notes

🤖 Generated with Claude Code

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 87.11%. Comparing base (cf6229a) to head (99be00e).

Additional details and impacted files

@@            Coverage Diff             @@
##             main      #88      +/-   ##
==========================================
+ Coverage   86.39%   87.11%   +0.71%     
==========================================
  Files          38       38              
  Lines        3022     3043      +21     
==========================================
+ Hits         2611     2651      +40     
+ Misses        411      392      -19     

Files with missing lines	Coverage Δ
R/gg_variable.R	85.41% <100.00%> (+2.85%)	⬆️
R/plot.gg_variable.R	82.13% <100.00%> (+5.94%)	⬆️

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[Copilot]

Pull request overview

Fixes the randomForest classification path for variable-dependence extraction/plotting by switching to per-class vote fractions (yhat.<class>) and addressing smooth=TRUE issues in plot.gg_variable, with accompanying regression tests and snapshot coverage.

Changes:

• gg_variable.randomForest() (classification) now emits per-class yhat.<class> columns derived from object$votes, normalized to probabilities.
• plot.gg_variable() multi-class reshape now uses class names for the outcome facet column; classification smooth=TRUE now supplies explicit aes(x, y) (binary) and adds the missing smooth layer (multi-class numeric).
• Adds new unit tests + vdiffr snapshots; bumps package version/date and documents the fixes in NEWS.

Reviewed changes

Copilot reviewed 6 out of 6 changed files in this pull request and generated 2 comments.

Show a summary per file

File	Description
R/gg_variable.R	Adjusts randomForest classification extraction to store per-class vote fractions as yhat.<class> columns.
R/plot.gg_variable.R	Fixes multi-class pivot outcome labels and ensures smooth=TRUE works in affected classification paths.
tests/testthat/test_gg_variable.R	Adds tests validating yhat.<class> columns, normalization, patchwork return, and smooth/layer_data() behavior.
tests/testthat/test_snapshots.R	Adds vdiffr snapshots for randomForest classification gg_variable plots.
NEWS.md	Documents the gg_variable.randomForest + plot.gg_variable fixes and increments dev version.
DESCRIPTION	Bumps package version and date.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.