hello everyone. How to divide a group of points named data into two parts, dataA and dataB, as shown in the figure. Thank you very much.

[Wesley]

Answer 1

[Stephan]

% divide data
idx = clusterdata(data,2);
dataA = data(idx==2,:);
dataB = data(idx==1,:);

% plot results
scatter(dataA(:,1),dataA(:,2))
hold on
scatter(dataB(:,1),dataB(:,2))
hold off

result is:

Note that this needs Statisitcs and Machine Learning Toolbox.

[Rik]

The question and plot screamed kmeans to me, but that also requires this toolbox. And this is such a clear division that every clustering function should get this result.

[Stephan]

result of kmeans:

[Wesley]

It is very helpful to me. Thank you for your warm-hearted reply.

Answer 2

[KSSV]

x = data(:,1) ;
y = data(:,2) ;

mx = mean(x) ; my = mean(y) ; 
idx = y < my ; 

x1 = x(idx) ; y1 = y(idx) ; 
x2 = x(~idx) ; y2 = y(~idx) ; 
plot(x,y,'.r')
hold on
plot(mx,my,'*b')
plot(x1,y1,'.b')
plot(x2,y2,'.g')

[Wesley]

Thank you for your warm-hearted reply.