Tidy hazard and cumulative-hazard curves from a Random Hazard Forest

Extracts case-specific ensemble hazard and cumulative-hazard estimates from a fitted randomForestRHF::rhf() object into a tidy long data frame, one row per (case, time) pair on the forest's time.interest grid.

Usage

gg_rhf(object, ...)

# S3 method for class 'rhf'
gg_rhf(object, source = c("oob", "inbag"), ...)

Arguments

object: A fitted rhf object from randomForestRHF.
...: Not currently used.
source: Which ensemble estimate to extract: "oob" (default, out-of-bag) or "inbag". Falls back to the other when the requested one is absent (e.g. bootstrap = "none" has no OOB estimate).

Value

A data.frame of class c("gg_rhf", "data.frame") with columns id, time, hazard, chf, source, an integer ntime attribute (number of grid points), and a provenance attribute.

The frame always has ntime rows per case. Both hazard and chf may be NA, and they are masked on different rules. From randomForestRHF 2.0.0 the pointwise hazard is defined only where a grid point falls inside one of the case's supplied (start, stop] intervals, and is NA in gaps and after the final stop time. From 2.0.3 the cumulative hazard is NA after each case's final stop, but stays flat through an internal gap rather than going missing there. So on a fit whose cases each have a single interval the two masks look the same, and with time-dependent covariates they do not. Earlier versions carried chf across the whole grid. The values are passed through unchanged; use na.rm = TRUE when summarising either column, or see plot.gg_rhf(), which drops the masked cells before drawing.

References

Ishwaran H, Hsich EM, Kogalur UB, Lee DKK (2026). Random Hazard Forests. arXiv:2608.21597. doi:10.48550/arXiv.2608.21597 .

Ishwaran H, Kogalur UB (2026). randomForestRHF: Random Hazard Forests. R package version 2.0.3. https://CRAN.R-project.org/package=randomForestRHF.

Examples

# \donttest{
if (requireNamespace("randomForestRHF", quietly = TRUE)) {
  data(pbc, package = "randomForestSRC")
  d <- randomForestRHF::convert.counting(
    survival::Surv(days, status) ~ ., na.omit(pbc))
  o <- randomForestRHF::rhf("Surv(id, start, stop, event) ~ .", d, ntree = 30)
  gg <- gg_rhf(o)
  plot(gg, idx = c(1, 5, 10))
}

# }