Minimum and Maximum of Possible Correlations

Description

This program minmax.ordcorr checks the first condition of the feasibility of a correlation matrix of ordinal random numbers. A mean vector (as list) needs to be specified. It returns yes/no if also a correlation matrix was given and in either case the Min-Max Correlation Matrix, which has the minimum correlation in the lower triangular matrix and the maximum correlation in the upper triangular matrix.

Usage

1
minmax.ordcorr(probs, Cor = 0, n = 1e+06, showX = FALSE)

Arguments

probs

List of probabilities for the variables, length of probability equals number of items, length of list equals number of variables

Cor

Correlation matrix

n

Number of Observations

showX

If TRUE resulting correlation matrix is shown

Value

It returns yes/no if also a correlation matrix was given and in either case the Min-Max Correlation Matrix, which has the minimum correlation in the lower triangular matrix and the maximum correlation in the upper triangular matrix.

Author(s)

Dominik Traeger and Sebastian Kaiser

See Also

check.ordcorr

Examples

1
minmax.ordcorr(list(c(1,1,1,1)/4,c(1,1,1,1)/4), cbind(c(0.5, 0.4), c(0.4, 0.8)), n = 1000, showX = TRUE)