Christoph Budziszewski
starting som prediction fine-tuned class-performance visualisation
Christoph Budziszewski commited 4dbef18 at 2009-01-21 16:34:25
som_dendrogram.m
function [h,Coord,Color,height] = som_dendrogram(Z,varargin)
%SOM_DENDROGRAM Visualize a dendrogram.
%
% [h,Coord,Color,height] = som_dendrogram(Z, [[argID,] value, ...])
%
% Z = som_linkage(sM);
% som_dendrogram(Z);
% som_dendrogram(Z,sM);
% som_dendrogram(Z,'coord',co);
%
% Input and output arguments ([]'s are optional):
% h (vector) handle to the arc lines
% Z (matrix) size n-1 x 1, the hierarchical cluster matrix
% returned by functions like LINKAGE and SOM_LINKAGE
% n is the number of original data samples.
% [argID, (string) See below. The values which are unambiguous can
% value] (varies) be given without the preceeding argID.
% Coord (matrix) size 2*n-1 x {1,2}, the coordinates of the
% original data samples and cluster nodes used
% in the visualization
% Color (matrix) size 2*n-1 x 3, the colors of ...
% height (vector) size 2*n-1 x 1, the heights of ...
%
% Here are the valid argument IDs and corresponding values. The values
% which are unambiguous (marked with '*') can be given without the
% preceeding argID.
% 'data' *(struct) map or data struct: many other optional
% arguments require this
% (matrix) data matrix
% 'coord' (matrix) size n x 1 or n x 2, the coordinates of
% the original data samples either in 1D or 2D
% (matrix) size 2*n-1 x {1,2}, the coordinates of both
% original data samples and each cluster
% *(string) 'SOM', 'pca#', 'sammon#', or 'cca#': the coordinates
% are calculated using the given data and the
% required projection algorithm. The '#' at the
% end of projection algorithms refers to the
% desired output dimension and can be either 1 or 2
% (2 by default). In case of 'SOM', the unit
% coordinates (given by SOM_VIS_COORDS) are used.
% 'color' (matrix) size n x 3, the color of the original data samples
% (matrix) size 2*n-1 x 3, the colors of both original
% data samples and each cluster
% (string) color specification, e.g. 'r.', used for each node
% 'height' (vector) size n-1 x 1, the heights used for each cluster
% (vector) size 2*n-1 x 1, the heights used for both original
% data samples and each cluster
% *(string) 'order', the order of combination determines height
% 'depth', the depth at which the combination
% happens determines height
 
הההההההההההההההההההההההההההההההההההההההההההההההההההההההההההההההההההההההההההההההההההההההההההההההההההההההההההההההההההההההההההההההההההההההההההההההההההההההההההההההההההההההההההההההההההההההההההההההההההההההההההההההההההההההההההההההההההההההההההההההההההההההההההההההה
XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX