Eigendecomposition of \(B = -J D^2 J / 2\); coordinates are the leading
eigenvectors scaled by the square roots of their (non-negative)
eigenvalues, as stats::cmdscale (Torgerson 1952); stress-1 of the
fitted distances as the Python arm morie.fn.mds.mds.
Usage
ClassicalMds(D, n_dims = 2)
Arguments
- D
Distance matrix.
- n_dims
Dimensions.
Value
List with coordinates, eigenvalues, stress.
References
Torgerson, W. S. (1952). Multidimensional scaling: I. Theory
and method. Psychometrika 17, 401-419.
Examples
ClassicalMds(rbind(c(0, 3, 4, 5), c(3, 0, 5, 4), c(4, 5, 0, 3), c(5, 4, 3, 0)))$eigenvalues
#> [1] 1.600000e+01 9.000000e+00 1.265235e-15 -2.571184e-15
