XploRe Help : relationcorrcont

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:	relation relationchi2 relationcont relationrank relationcorr

Quantlet:		relationcorrcont
Description:		Computes the corrected contingency coefficient for discrete data.

Usage:	`rel = relationcorrcont (x {, colname {, opt}})`
Input:
	`x`	n x p variables
	`colname`	n x p variable names
	`opt`	q x 1 text vector of optional parameters
Output:
	`rel.r`	p x p matrix of corrected contingency coefficients
	`rel.pval`	p x p significance level of chi^2 statistics

Note:

To compute the corrected contingency coefficient non-interactively you need to set opt = "automatic". In the interactive mode you will get a menu sorted after the largest coefficients. If you click on the coefficient you will get the crosstable of the corresponding variables.

Example:

; loads the library stats library("stats") ; read swiss banknote data ; actually these data are continuous, but the first ; three variables have approx. 20 realizations x = read("bank2") x = x[,1:3] ; compute the corrected contingency coefficients and the ; pvalues of the chi^2 statistic automatically colname = string("X%0.f", 1:cols(x)) relationcorrcont(x, colname, "automatic")

Result:

Contents of rel.r
[1,]        1  0.81833  0.85195
[2,]  0.81833        1  0.86405
[3,]  0.85195  0.86405        1
Contents of rel.pval
[1,]        0  0.20152  0.0024184
[2,]  0.20152        0  3.8681e-11
[3,]  0.0024184  3.8681e-11        0

Author: S. Klinke, 19981112