Kerri Colman Kerri Colman - 3 months ago 19
R Question

kolmogorov-smirnov plot in R ggplot

I am trying to do a KS plot in r and all seems to be going well - except for the fact that i can only use colour to visualise the two different samples and not line type.

I have tried the following:

sample1<-SD13009
sample2<-SD13009PB

group <- c(rep("sample1", length(sample1)), rep("sample2", length(sample2)))
dat <- data.frame(KSD = c(sample1,sample2), group = group)
cdf1 <- ecdf(sample1)
cdf2 <- ecdf(sample2)

minMax <- seq(min(sample1, sample2), max(sample1, sample2), length.out=length(sample1))
x0 <- minMax[which( abs(cdf1(minMax) - cdf2(minMax)) == max(abs(cdf1(minMax) - cdf2(minMax))) )]
y0 <- cdf1(x0)
y1 <- cdf2(x0)


#attempt 1

plot<-ggplot(dat, aes(x = KSD, group = group, colour = group, linetype=group))+
stat_ecdf(size=1) +
mytheme + xlab("mm") +scale_x_continuous(limits=c(0,1))+
ylab("Cumulitive Distibution") +
#geom_line(aes(group=group,size=1)) +
geom_segment(aes(x = x0[1], y = y0[1], xend = x0[1], yend = y1[1]),
linetype = "dashed", color = "red") +
geom_point(aes(x = x0[1] , y= y0[1]), color="red", size=1) +
geom_point(aes(x = x0[1] , y= y1[1]), color="red", size=1) +
ggtitle("K-S Test: Sample 1 / Sample 2")


#attempt 2

cdf <- ggplot(dat, aes(x=KSD, group=group,linetype=group)) + stat_ecdf(aes(linetype=group)) + coord_cartesian(xlim = c(0, 0.8)) + geom_segment(aes(x = x0[1], y = y0[1], xend = x0[1], yend = y1[1]),
linetype = "dashed", color = "red") +
geom_point(aes(x = x0[1] , y= y0[1]), color="red", size=1) +
geom_point(aes(x = x0[1] , y= y1[1]), color="red", size=1) +
ggtitle("K-S Test: Sample 1 / Sample 2")


This is what i get:

enter image description here

Answer

I cannot reproduce this, with the following code:

# Make two random samples
sample1<-rnorm(1000)
sample2<-rnorm(1000, 2, 2)

group <- c(rep("sample1", length(sample1)), rep("sample2", length(sample2)))
dat <- data.frame(KSD = c(sample1,sample2), group = group)
cdf1 <- ecdf(sample1) 
cdf2 <- ecdf(sample2) 

minMax <- seq(min(sample1, sample2), max(sample1, sample2), length.out=length(sample1)) 
x0 <- minMax[which( abs(cdf1(minMax) - cdf2(minMax)) == max(abs(cdf1(minMax) - cdf2(minMax))) )] 
y0 <- cdf1(x0) 
y1 <- cdf2(x0) 


ggplot(dat, aes(x = KSD, group = group, colour = group, linetype=group))+
  stat_ecdf(size=1) +
  xlab("mm") +
  ylab("Cumulitive Distibution") +
  geom_segment(aes(x = x0[1], y = y0[1], xend = x0[1], yend = y1[1]),
               linetype = "dashed", color = "red") +
  geom_point(aes(x = x0[1] , y= y0[1]), color="red", size=1) +
  geom_point(aes(x = x0[1] , y= y1[1]), color="red", size=1) +
  ggtitle("K-S Test: Sample 1 / Sample 2")

enter image description here

It seems that in your plot the lines are so close together that you can't see that they are different linetypes, but they are.