
XGvis uses Multidimensional Scaling (MDS) 
  to map a set of N objects 
  to N points in k-space 
  such that a given set of target distances or dissimilarities are 
  approximated as well as possible:
   
    DISSIMILARITY( object i, object j ) ~ DISTANCE( point i, point j )

  The pairwise dissimilarities are the input data;
  the positions of the points are the parameters to be optimized.

MDS optimizes the point positions by minimizing a 
STRESS FUNCTION or a STRAIN FUNCTION.  
See the help windows for the formulae.

The matrix of target distances or dissimilarities should be symmetric 
of size NxN.

It is either 

    - read from a file containing such a matrix 
        ("file.dist"), or

    - calculated as the shortest-path distances of a discrete graph  
        ("file.edges"), or

    - calculated as the Euclidean distances of a multivariate dataset
        ("file.pos" or "file.dat")

See the help window for input file formats.

