ROC plot generic function for a gg_roc object.

ROC plot generic function for a gg_roc object.

Usage

# S3 method for class 'gg_roc'
plot(x, which_outcome = NULL, ..., panel = c("overlay", "facet"))

Arguments

x

A gg_roc object, or a raw rfsrc or randomForest classification forest.

A raw forest is accepted, but plain plot(forest) does not arrive here. Both randomForestSRC and randomForest register their own plot methods, so S3 dispatch sends a raw forest to plot.rfsrc or plot.randomForest instead. This branch is reached only by naming the method outright, as plot.gg_roc(forest).

That branch also does not use gg_roc's own default for which_outcome: given a multi-class forest and which_outcome = NULL it calls gg_roc once per class and overlays the one-vs-rest curves, where gg_roc(x) alone returns a single curve (a macro-average for randomForest, or class 1 with a warning for rfsrc). Prefer plot(gg_roc(x, which_outcome)), which is explicit about both the class and the engine. Issue #72 tracks reconciling the entry points.

which_outcome

Integer; for multi-class problems, the index of the class to plot. When NULL (default) and the forest has more than two classes, the curves for all classes are overlaid in one plot. For binary forests, NULL defaults to class index 2.

...

Additional arguments passed to gg_roc when x is a raw forest (e.g. oob = FALSE).

panel

Character; layout for per-class ROC objects, the ones from gg_roc(..., per_class = TRUE). "overlay" (default) draws every class curve in one panel, colored by class; "facet" gives each class its own panel. Ignored for single-class gg_roc objects.

Value

A ggplot object. The x-axis is 1 - Specificity (FPR), the y-axis is Sensitivity (TPR), and a dashed red diagonal marks the random-classifier baseline. Single-class curves carry the AUC as an annotation; multi-class plots color and style each class curve distinctly.

References

Breiman L. (2001). Random forests, Machine Learning, 45:5-32.

Ishwaran H. and Kogalur U.B. (2007). Random survival forests for R, Rnews, 7(2):25-31.

Ishwaran H, Kogalur U (2026). Fast Unified Random Forests for Survival, Regression, and Classification (RF-SRC). R package version 3.6.2. https://cran.r-project.org/package=randomForestSRC

See also

gg_roc calc_roc calc_auc rfsrc randomForest

Examples

## ------------------------------------------------------------
## classification example
## ------------------------------------------------------------
## -------- iris data
# Build a small classification forest (ntree=50 keeps example fast)
set.seed(42)
rfsrc_iris <- randomForestSRC::rfsrc(Species ~ ., data = iris, ntree = 50)

# ROC for setosa (outcome index 1)
gg_dta <- gg_roc(rfsrc_iris, which_outcome = 1)
plot(gg_dta)


# ROC for versicolor (outcome index 2)
gg_dta <- gg_roc(rfsrc_iris, which_outcome = 2)
plot(gg_dta)


# ROC for virginica (outcome index 3)
gg_dta <- gg_roc(rfsrc_iris, which_outcome = 3)
plot(gg_dta)


# Plot all three ROC curves in one call by iterating over outcome indices
n_cls <- ncol(rfsrc_iris$predicted)
for (i in seq_len(n_cls)) print(plot(gg_roc(rfsrc_iris, which_outcome = i)))