%PREX-MCPLOT PRTools example on a multi-class classifier plot
help prex_mcplot
echo on
gridsize(100)
% Generate twice normally distributed 2-class data in 2D
a = +gendath([20,20]); % data only
b = +gendath([20,20]); % data only
% Shift the second data by a vector [5,5]
% and combine it with the first dataset in to A
A = [a; b+5];
% Generate 4-class labels
lab = genlab([20 20 20 20],[1 2 3 4]');
% Construct a 4-class dataset A
A = prdataset(A,lab);
A = setname(A,'4-class dataset')
% Plot this 4-class dataset
figure
% Make a scatter-plot of the right size
scatterd(A,'.'); drawnow;
% Compute normal densities based quadratic classifier
w = qdc(A);
% Plot filled classification regions
plotc(w,'col'); drawnow;
hold on;
% Redraw the scatter-plot
scatterd(A);
hold off
echo off