Zelig (version 3.5.1)

plot.surv: Plotting Confidence Intervals for Survival Curves

Description

The plot.surv command generates confidence intervals for Kaplan-Meier survival curves

Usage

plot.surv(x, duration, censor, type = "line", plotcensor=TRUE,
               plottimes = FALSE, int = c(0.025,0.975), ...)

Arguments

x

output from sim stored as a list. Each element of the list is the sim output for a particular survival curve.

duration

the duration variable (e.g. lifetime, survival, etc.).

censor

the censored data

type

the type of confidence interval. Defaults to "line", which draws vertical confidence intervals at observed event times. "poly" draws confidence regions using polygons.

plotcensor

default is TRUE. Plots censoring times as a rug object.

plottimes

default is FALSE. Plots step function with indicators at observed event times.

int

vector of quantile limits for the confidence interval. Default is 95% interval.

...

Additional parameters passed to plot.

Value

• For survival models, plot.surv() returns vertical confidence intervals or polygon survival regions for Kaplan-Meier survival curves. You may save this plot using the commands described in the Zelig manual (http://gking.harvard.edu/zelig).

See Also

The full Zelig manual is available at http://gking.harvard.edu/zelig, and users may also wish to see plot, lines.

Examples

data(coalition)
z.out1 <- zelig(Surv(duration,ciep12)~invest+numst2+crisis,
robust=TRUE,cluster="polar",model="coxph",data=coalition)
low <- setx(z.out1,numst2=0)
high <- setx(z.out1,numst2=1
# Simulate Survival Curves for Each Group
s.out1 <- sim(z.out1,x=low) 
s.out2 <- sim(z.out1,x=high)

# Organize simulated output as a list
out <- list(s.out1,s.out2)

plot.surv(x = out, duration = coalition$duration, censor=coalition$ciep12,
          type="line", plottimes=FALSE, plotcensor=FALSE,
          main="Survival", xlab="Time", ylab="Survival")