[R] P-values Kolmogorov–Smirnov test

[Boo G.]

Hello,

I am trying to perform a Kolmogorov–Smirnov test to assess the difference between a distribution and samples drawn proportionally to size of different sizes. I managed to compute the Kolmogorov–Smirnov distance but I am lost with the p-value. I have looked into the ks.test function unsuccessfully. Can anyone help me with computing p-values for a two-tailed test?

Below a simplified version of my code.

Thanks in advance.
Gianluca


library(spatstat)

#reference distribution
d_1 <- sort(rpois(1000, 500))
p_1 <- d_1/sum(d_1)
m_1 <- data.frame(d_1, p_1)

#data frame to store the values of the siumation
d_stat <- data.frame(1:1000, NA, NA)
names(d_stat) <- c("sample_size", "ks_distance", "p_value")

#simulation
for (i in 1:1000) {
  #sample from the reference distribution
  m_2 <-m_1[(sample(nrow(m_1), size=i, prob=p_1, replace=F)),]
  m_2 <-m_2[order(m_2$d_1),]
  d_2 <- m_2$d_1
  p_2 <- m_2$p_1

  #weighted ecdf for the reference distribution and the sample
  f_d_1 <- ewcdf(d_1, normalise=F)
  f_d_2 <- ewcdf(d_2, 1/p_2, normalise=F, adjust=1/length(d_2))

  #kolmogorov-smirnov distance
  d_stat[i,2] <- max(abs(f_d_1(d_2) - f_d_2(d_2)))
}


	[[alternative HTML version deleted]]