Home > brede > brede_mat_dist.m

brede_mat_dist

PURPOSE ^

brede_mat_dist - Distance between objects

SYNOPSIS ^

function D = brede_mat_dist(M, varargin)

DESCRIPTION ^

 brede_mat_dist       - Distance between objects
 
       function D = brede_mat_dist(M, varargin)

       Input:    M    'Mat' structure or matrix

       Property: Distance [ {InnerProduct} | Covariance | CorrCoef |
                          RawCorrCoef ] Distance or similarity
                          measure 

       Output:   D    'Mat' structure with a symmetric distance
                      matrix. 
    
       Calculate the distance between the objects, ie, between one
       row and an other. If the input matrix is of size NxP the the
       distance matrix will be NxN. The input might also be an
       ordinary numeric matrix. If the input is a structure and the
       'rows' field exists the these are copied to the 'rows' and
       'columns' field of the output structure.

       'Distance'='innerproduct' the distance matrix is X * X'. 

       'Distance'='covariance' is the same a the inner product but
       with center data.

       'Distance'='RawCorrCoef' is the correlation coefficient
       without centering.

       'Distance'='CorrCoef' is the correlation coefficient with
       centering. 

       NaN elements are ignored. If there are missing values and the
       'CorrCoef' distance is used then the distance matrix might
       have elements larger than one.
 
       'Covariance' are normalized with the number of elements (that
       are not NaN) in each row of M.matrix.

       See also BREDE, BREDE_MAT, BREDE_MAT_PLOT_MDS,
                BREDE_MAT_PLOT_COUNT, BREDE_MAT_TRANSPOSE,
                BREDE_BIB_BIB2MAT, BREDE_EXP_EXP2MAT.

 $Id: brede_mat_dist.m,v 1.10 2005/05/09 19:16:06 fnielsen Exp $

CROSS-REFERENCE INFORMATION ^

This function calls: This function is called by:
Generated on Fri 27-Nov-2009 18:11:22 by m2html © 2005