XploRe Help : rankcorr

Keywords - Function groups - @ A B C D E F G H I J K L M N O P Q R S T U V W X Y Z

Library:	stats
See also:	corr relation

Quantlet:		rankcorr
Description:		computes rank correlation coefficients according to Spearman and Kendall. In the case of ties, corrected versions are computed.

Reference(s):

"Verteilungsfreie Methoden in der Biostatistik" Bortz/Lienert/Boehnke (1990) Springer-Verlag

Usage:	`rc = rankcorr (x {,opt})`
Input:
	`x`	n x 2 matrix, the data.
	`opt`	string, either "spearman" or "kendall".
Output:
	`rc`	rank correlation between the columns of x.

Note:

It is not necessary that the input x are ranks. The macro computes the ranks internally.

Example:

library("stats") randomize(0) x = normal(200, 2) rankcorr(x, "spearman") rankcorr(x, "kendall") x =(1:7)~(7:1) rankcorr(x, "spearman") rankcorr(x, "kendall")

Result:

Contents of rc
[1,] -0.00040501
Contents of rc
[1,] -0.0001005
Contents of rc
[1,]       -1
Contents of rc
[1,]       -1

Author: Marlene Mueller, 990218