SOM Toolbox Online documentation



 Calculate the response of the given data on the map.



 Returns a vector indicating the response of the map to the data.
 The response of the data on the map can be calculated e.g. in
 three ways, selected with the mode argument: 
  'crisp'    traditional hit histogram: how many times each map unit 
             was the BMU for the data set
  'kernel'   a sum of neighborhood kernels, where a kernel
             is positioned on the BMU of each data sample. The 
             neighborhood function is sMap.neigh and the
             neighborhood width is sMap.trainhist(end).radius_fin
             or 1 if this is not available 
  'fuzzy'    fuzzy response calculated by summing 

                       1 +  (q/a)^2

             for each data sample, where q is a vector containing
             distance from the data sample to each map unit and 
             a is average quantization error

Required input arguments

   sMap              The vectors from among which the BMUs are searched
                     for. These must not have any unknown components (NaNs).
            (struct) map struct
            (matrix) codebook matrix, size munits x dim

   sData             The data vector(s) for which the BMUs are searched.
            (struct) data struct
            (matrix) data matrix, size dlen x dim

Optional input arguments

   mode     (string) The respond mode: 'crisp' (default), 'kernel'
                     or 'fuzzy'. 'kernel' can only be used if 
                     the first argument (sMap) is a map struct.                     

Output arguments

   hits     (vector) The number of hits in each map unit.


  hits = som_hits(sM,D);
  hits = som_hits(sM,D,'kernel');
  hits = som_hits(sM,D,'fuzzy');

See also

som_bmus Find BMUs and quantization errors for a given data set.