tranSurv (version 1.1-4)

uncondKendall: Computes the unconditional Kendall's tau

Description

This function computes the unconditional Kendall's tau, or the Kendall rank correlation coefficient for two variables. The implementation is done in C and performs a lot faster than cor with method = "kendall".

Usage

uncondKendall(x, y)

Arguments

x

a numeric vector

y

a numeric vector with compatible length to 'y'

Value

The output is a numeric value for the unconditional Kendall's tau.

References

Kendall, M. G. (1938), A new measure of rank correlation, Biometrika, 81-93.

Examples

## Prepare to generate from normal copula
pho <- .25
A <- matrix(c(1, pho, pho, sqrt(1 - pho^2)), 2)

## Generating truncated survival data from normal copula
set.seed(123)
n <- 10000
dat <- matrix(NA, n, 2)
for (i in 1:n) dat[i,] <- pnorm(A %*% rnorm(2))
dat <- data.frame(dat)
colnames(dat) <- c("x", "y")
system.time(print(uncondKendall(dat$x, dat$y)))
system.time(print(cor(dat$x, dat$y, method = "kendall")))