Consistent variable and partial dependence yhat values.

[ehrlinger]

This is an rfsrc issue. For survival and classification forests, variable dependence is returned on [0,1], partial dependence for survival is returned as [0,100]. Should be able to detect the difference and convert in gg_variable.

[ehrlinger]

Triaged 2026-05-29 during a docs/cleanup pass. Empirical check on a current rfsrc survival forest (pbc):

• gg_variable() survival yhat comes back on [0,1] (survival-probability scale) — confirmed.
• gg_partial() on a survival plot.variable(partial=TRUE) result returns a structured gg_partial object (a {continuous, categorical} grouping), not a flat yhat column, so its effective scale isn't directly comparable without unpacking the sub-frames. The original "[0,100]" report predates ~11 years of randomForestSRC changes and may reflect an older partial-output scale.

Not closing yet. Remaining work to resolve: unpack gg_partial's survival sub-frame, confirm whether its yhat scale still differs from gg_variable's [0,1], and if so add detect-and-convert in the extractor plus a regression test. Scoped as a candidate fix for the v3.1.0 documentation/cleanup line (where doc reconciliation surfaced it) or a later patch. Keeping open with this head-start.