Skip to contents

Product-limit geoms for survival and competing risks curves

Source: R/geom_prodlim.R
geom_prodlim.Rd

Draw Kaplan-Meier or Aalen-Johansen curves with optional confidence bands. For competing risks with multiple causes, the default display shows one curve per cause. When `cause = "stacked"`, causes are drawn as stacked filled rectangles instead of step curves with confidence shadows.

Usage

geom_prodlim(
  mapping = NULL,
  data = NULL,
  position = "identity",
  na.rm = FALSE,
  show.legend = NA,
  inherit.aes = TRUE,
  type = "risk",
  cause = NULL,
  conf_int = TRUE,
  conf_int_alpha = 0.2,
  percent = TRUE,
  timeconverter = NULL,
  times = NULL,
  cens.code = "0",
  ...
)

Arguments

mapping

Set of aesthetic mappings created by [ggplot2::aes()].

data

Data frame.

position

Position adjustment.

na.rm

If `FALSE`, missing values are removed with a warning.

show.legend

Logical. Should this layer be included in the legends?

inherit.aes

If `FALSE`, overrides the default aesthetics.

type

Passed to [prodlim::prodlim()], usually `"risk"` or `"surv"`.

cause

Cause(s) for competing risks. Can be a vector of causes or the string `"stacked"`.

conf_int

Logical. Draw confidence intervals.

conf_int_alpha

Alpha level for confidence shadows.

percent

Logical. Passed to [prodlim::summary.prodlim()].

timeconverter

Optional time conversion string.

times

Optional evaluation times passed to [prodlim::summary.prodlim()].

cens.code

Censoring code passed to [prodlim::Hist()].

...

Further arguments passed to [ggplot2::layer()].

Details

When multiple causes are specified, `fill`/`colour` aesthetics are ignored and replaced by a cause-based mapping so that all causes are shown in a single legend. When `cause = "stacked"`, causes are stacked on top of each other and confidence shadows are not drawn.

Examples

library(riskRegression)
#> riskRegression version 2026.02.13
library(data.table)
library(ggplot2)
#> 
#> Attaching package: ‘ggplot2’
#> The following object is masked from ‘package:lava’:
#> 
#>     vars
data(Melanoma)
# Kaplan-Meier
ggplot(data = Melanoma,aes(x = time, event = 1*(status != 0)))+
       geom_prodlim(type = "surv")

# stratified Kaplan-Meier inherited aes
ggplot(data = Melanoma,aes(x = time, event = 1*(status != 0),fill = sex,color = sex))+
       geom_prodlim(type = "surv")

# stratified Kaplan-Meier geom aes
ggplot(data = Melanoma,aes(x = time, event = 1*(status != 0)))+
       geom_prodlim(aes(fill = sex,color = sex),type = "surv")

# facet
ggplot(data = Melanoma,aes(x = time, event = 1*(status != 0)))+
       geom_prodlim(type = "surv")+facet_grid(~sex)

# stratified and facet
ggplot(data = Melanoma,aes(x = time, event = 1*(status != 0),fill = epicel,color = epicel))+
       geom_prodlim(type = "surv")+facet_grid(~sex)


# Aalen-Johansen
ggplot(data = Melanoma,aes(x = time, event = status))+
       geom_prodlim(type = "surv",cause = 1:2)

# stratified Aalen-Johansen inherited aes
ggplot(data = Melanoma,aes(x = time, event = status,fill = sex,color = sex))+
       geom_prodlim(type = "surv")

# stratified Aalen-Johansen geom aes
ggplot(data = Melanoma,aes(x = time, event = status))+
       geom_prodlim(aes(fill = sex,color = sex),type = "surv")

# facet
ggplot(data = Melanoma,aes(x = time, event = status))+
       geom_prodlim(type = "surv")+facet_grid(~sex)

# stratified and facet
ggplot(data = Melanoma,aes(x = time, event = status,fill = epicel,color = epicel))+
       geom_prodlim(type = "surv")+facet_grid(~sex)

# stacked
ggplot(data = Melanoma,aes(x = time, event = status))+
       geom_prodlim(cause = "stacked")+facet_grid(~sex)

# stacked with cens.code option
ggplot(data = Melanoma,aes(x = time, event = event))+
       geom_prodlim(cens.code="censored",cause = "stacked")+facet_grid(~sex)