Project high dim. data unto principal components (PCA) for visualization. Optionally IDX can be specified to indicate different classes for the points; in this case points in different classes are displayed using different colors. Up to 12 types are handled (for technical reasons involving plot), any cluster with a label>12 is assigned the label 12. USAGE visualizeData( X, k, [IDX], [types], [C] ) INPUTS X - column vector of data - N vectors of dimension p (X is Nxp) k - dimension to which to reduce data (2 or 3) IDX - [] cluster membership [see kmeans2.m] types - [] cell array of length ntypes of text labels for each type C - [] cluster centers (Kxp) OUTPUTS EXAMPLE X = [randn(100,5); randn(100,5)+4]; C = [mean(X(1:100,:)); mean(X(101:200,:))]; IDX = [ones(100,1); 2*ones(100,1)]; visualizeData( X, 2, IDX, {'type1','type2' }, C); See also KMEANS2, DEMOCLUSTER Piotr's Computer Vision Matlab Toolbox Version 2.0 Copyright 2014 Piotr Dollar. [pdollar-at-gmail.com] Licensed under the Simplified BSD License [see external/bsd.txt]

Generated by