plot.gg_survival: Plot a 'gg_survival' object.

In ggRandomForests: Visually Exploring Random Forests

Plot a gg_survival object.

Description

Draws a Kaplan-Meier (or Nelson-Aalen) survival curve from a gg_survival object. You can overlay a confidence envelope around the curve using error: "shade" fills the area between the pointwise confidence limits, "lines" draws them as dashed step functions, and "bars" shows them as error bars. When gg_survival was called with a by argument, each group gets its own step function and the label argument renames the legend.

Usage

## S3 method for class 'gg_survival'
plot(
  x,
  type = c("surv", "cum_haz", "hazard", "density", "mid_int", "life", "proplife"),
  error = c("shade", "bars", "lines", "none"),
  label = NULL,
  ...
)

Arguments

x	gg_survival or a survival gg_rfsrc object created from a rfsrc object
type	"surv", "cum_haz", "hazard", "density", "mid_int", "life", "proplife"
error	"shade", "bars", "lines" or "none"
label	Modify the legend label when gg_survival has stratified samples
...	Additional arguments forwarded to geom_step(). When alpha is supplied it is passed to the step geom and a halved value is used for the ribbon overlay.

Details

The type argument selects which quantity to plot on the y-axis – survival probability ("surv") is the default, but cumulative hazard, density, and several transformed scales are available for the cases where a linear scale reveals more about the tails.

Value

A ggplot object. The y-axis shows the chosen type (e.g. survival probability for "surv") and the x-axis shows time. Confidence shading, bars, or lines are added when the input object carries confidence-interval columns.

See Also

gg_survival, kaplan, nelson, gg_rfsrc

Examples

## -------- pbc data
data(pbc, package = "randomForestSRC")
pbc$time <- pbc$days / 364.25

# This is the same as kaplan
gg_dta <- gg_survival(
  interval = "time", censor = "status",
  data = pbc
)

plot(gg_dta, error = "none")
plot(gg_dta)

# Stratified on treatment variable.
gg_dta <- gg_survival(
  interval = "time", censor = "status",
  data = pbc, by = "treatment"
)

plot(gg_dta, error = "none")
plot(gg_dta)
plot(gg_dta, label = "treatment")

# ...with smaller confidence limits.
gg_dta <- gg_survival(
  interval = "time", censor = "status",
  data = pbc, by = "treatment", conf.int = .68
)

plot(gg_dta, error = "lines")
plot(gg_dta, label = "treatment", error = "lines")

# ...with smaller confidence limits.
gg_dta <- gg_survival(
  interval = "time", censor = "status",
  data = pbc, by = "sex", conf.int = .68
)

plot(gg_dta, error = "lines")
plot(gg_dta, label = "sex", error = "lines")

ggRandomForests documentation built on June 13, 2026, 5:07 p.m.