Transformation of a Distance Matrix for becoming Euclidean

Share:

Description

transforms a distance matrix in a Euclidean one.

Usage

1
lingoes(distmat, print = FALSE, tol = 1e-07, cor.zero = TRUE)

Arguments

distmat

an object of class dist

print

if TRUE, prints the eigenvalues of the matrix

tol

a tolerance threshold for zero

cor.zero

if TRUE, zero distances are not modified

Details

The function uses the smaller positive constant k which transforms the matrix of sqrt(dij² + 2*k) in an Euclidean one

Value

returns an object of class dist with a Euclidean distance

Author(s)

Daniel Chessel
Stéphane Dray stephane.dray@univ-lyon1.fr

References

Lingoes, J.C. (1971) Some boundary conditions for a monotone analysis of symmetric matrices. Psychometrika, 36, 195–203.

Examples

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
data(capitales)
d0 <- capitales$dist
is.euclid(d0) # FALSE
d1 <- lingoes(d0, TRUE)
# Lingoes constant = 2120982
is.euclid(d1) # TRUE
plot(d0, d1)
x0 <- sort(unclass(d0))
lines(x0, sqrt(x0^2 + 2 * 2120982), lwd = 3)
 
is.euclid(sqrt(d0^2 + 2 * 2120981), tol = 1e-10) # FALSE
is.euclid(sqrt(d0^2 + 2 * 2120982), tol = 1e-10) # FALSE
is.euclid(sqrt(d0^2 + 2 * 2120983), tol = 1e-10) 
    # TRUE the smaller constant

Want to suggest features or report bugs for rdrr.io? Use the GitHub issue tracker.