kmplot: Plot survival functions by discrete categories
In tobyjohnson/gtx: Genetics ToolboX

Description Usage Arguments Details Value Author(s) Examples

Plots survival functions by discrete categories (such as genotype).

## S3 method for class 'character'
kmplot(object, x, data, ylab, xlab,
                           ylim, xlim, col, lty,
                           legend.location = "topright", legend.cex = 1,
                           legend.median = FALSE, digits = 1,
                           atrisk = FALSE, ...)
## S3 method for class 'Surv'
kmplot(object, x, data, ylab, xlab,
                      ylim, xlim, col, lty,
                      legend.location = "topright", legend.cex = 1,
                      legend.median = FALSE, digits = 1,
                      atrisk = FALSE, ...)

`object`	The name or values of the survival variable
`x`	The name or values of the discrete category variable
`data`	A data frame containing values
`ylab`	A label for the y axis
`xlab`	A label for the x axis
`ylim`	Range for the y-axis
`xlim`	Range for the x-axis
`col`	A vector of colours
`lty`	A vector of line types
`legend.location`	Location for legend
`legend.cex`	Character expansion (size) for legend
`legend.median`	Whether to calculate and include median survival times in legend
`digits`	Number of digits for legend data
`atrisk`	Logical, whether to display at risk numbers
`...`	Other arguments

This function calculates Kaplan–Meier estimates of survival functions and draws an annotated plot. The arguments object and x specify the survival data and the levels of a discrete factor, either directly (as variables of class Surv and factor respectively), or as names of variables in the data frame data.

Several options for annotating the plot are supported, which provide improvements over the default print.survfit method. A legend with the number of subjects in each category is automatically added, this may be disabled by setting legend.location="none". The legend may be automatically annotated with the median survival times and 95% confidence limits. Numbers of subjects at risk, at each tick on the x-axis, may be added.

Returns an invisible null. The plot is generated as a side effect.

Toby Johnson Toby.x.Johnson@gsk.com

library(survival)
data(aoex1)
aoex1 <- within(aoex1, srvDays <- Surv(SRVDY, SRVCFLCD))
## show two different ways of calling, either:
with(aoex1, kmplot(srvDays, rs123456, xlab = "Days"))
## or:
kmplot("srvDays", "rs123456", data = aoex1, xlab = "Days")
## show additional annotation
kmplot("srvDays", "rs123456", data = aoex1, xlab = "Days",
       legend.median = TRUE, atrisk = TRUE)
## show correct behaviour when one level has entirely missing survival data
with(aoex1, kmplot(srvDays, STUDYID, xlab = "Days"))