Differentiating each Line with different type in `ggsurv` plots (or in `plot`)

Question

I am using Rstudio. I am using ggsurv function from GGally package for drawing Kaplan-Meier curves for my data (for survival analysis), from tutorial here. I am using it instead of plot because ggsurv takes care of legends by itself.

As shown on the link, multiple curves are differentiated by color. I want to differentiate based on linetype. The tutorial does not seem to have any option for that. Following is my command:

surv1 <- survfit(Surv(DaysOfTreatment,Survived)~AgeOnFirstContactGroup)
print(ggsurv(surv1, lty.est = 3)+ ylim(0, 1))

lty.est=3(or 2) gives same dashed lines for all the lines. I want differently dashed line for each line. Using lty=type gives error:object 'type' not found. And lty=type would work in ggplot but ggplot does not directly deal with survfit plots.

Please show me how to differentiate curves by linetype in either ggsurv or simple plot (although I would prefer ggsurv because it takes care of legends)

Answer 1

From the documentation for ggsurv

lty.est: linetype of the survival curve(s). Vector length should be either 1 or equal to the number of strata.

So, to get a different line type for each stratum, set lty.est equal to a vector of the same length as the number of lines you are plotting, with each value corresponding to a different line type.

For example, using the lung data from the survival package

library(GGally)
library(survival)
data(lung)
surv1 <- survfit(Surv(time,status) ~ sex, data = lung)
ggsurv(surv1, lty.est=c(1,2), surv.col = 1)

Gives the following plot

You can add ggplot themes or other ggplot elements to the plot too. For example, we can improve the appearance using the cowplot theme as follows

library(ggplot2)
library(cowplot)
ggsurv(surv1, lty.est=c(1,2), surv.col = 1) + theme_cowplot()

If you need to change the legend labels after differentiating by linetype, then you can do it this way

ggsurv(surv1, lty.est=c(1,2), surv.col = 1) +
  guides(colour = FALSE) +
  scale_linetype_discrete(name   = 'Sex', breaks = c(1,2), labels = c('Male', 'Female'))